'^%^- ^. ;a6«-*-»«3Jw*-^^'^i Ho riot drciiT-»fc HARVARD UNIVERSITY. LIBRARY MUSEUM or OOMPAEATIVE ZOOLOGY 3_m fyjyO(MJJ/YHJ ^JnA^OML ^0, lf5-~ UJaaI h l^l(o. Mrrt 1 lyio BULLETIN MUSEUM OF COMPAEATTVE ZOfiLOGY HARVARD COLLEGE, LN CAMBRIDGE. VOL. LIX. CAMBRIDGE, MASS., U. S. A 1915. The Cosmos Press: E. W. Wheeler, Cambridge, U. S. A. CONTENTS Page No. 1. — Mammals obtained by the Phillips Palestine expedition. By Glover M. Allen. February, 1915 1 No. 2. — The cranial nerves of AnoHs carolinensis. By William A. WiLLARD. (7 plates). July, 1915 15 No. 3. — Relics of Peale's Museum. By Walter Faxon. July, 1915 . 117 No. 4. — Exploration of the coast water between Nova Scotia and Chesapeake Bay, July and August, 1913, by the U. S. Fisheries Schooner Grampus. Oceanography and plankton. By Henry B. BiGELOW. (2 plates). September, 1915 149 ft- No. 5. — Notes on birds from East Siberia and Arctic Alaska. By W. Sprague Brooks. September, 1915 361 No. 6. — A revision of the hzards of the genus Ameiva. By Thomas Barbour and G. Kingsley Noble. October, 1915. .... 415 No. 7. — Two new genera of myrmicine ants from Brazil. By William Morton Wheeler. November, 1915 481 No. 8. — New chilopods from Mexico and the West Indies. By Ralph V. Chamberlin. (5 plates). November, 1915 493 3lV^ Bulletin of the Museum of Compai-ative Zoology AT HARVARD CO I LEGE. Vol. LIX. No. 1. MAMMALS OBTAINED BY THE PHILLIPS PALESTINE EXPEDITION. By Glover M. Allen. CAMBRIDGE, MASS., U. S. A.: PRINTED FOR THE MUSEUM. February, 1915. Reports on the Scientific Results of the Expedition to the East- ern Tropical Pacific, , in charge of Alexander Agassiz, by the U. S. Fish CoMMisaibx Steamer "Albatross," from October, 1904, to IMarch, 1905, L^i!;mt:^'K^'r Comiviander L. M. Garrett, U. S. N., CoMMA>a)i:NG, PUBLISHED OP. IN prkparation: — A. AGASSIZ. V.6 General Report on the Expedition. A. AGASSIZ. I.> Three Letters to Geo. I M. Bowers, U. S. Fish Com. A. AGASSIZ and H. L. CLARK. The Echini. H. B. BIGELOW. XVI." The Medusae. H. B. BIGELOW. XXIIL^' The Sipho- nophores. H. B. BIGELOW. XXVI.26 The Cteno- R. P. BIGELOW. The Stomatopods. O. CARLGREN. The Actinaria. R. V. CHAMBERLIN. The Annelids. H. L. CLARK. The Holothurians. H. L. CLARK. The Starfishes. H. L. CLARK. The Opliiurans. S. F. CLARKE. VIII.s The Hydroids. W. R. COE. The Nemerteans. L. J. COLE. XIX.i' The Pycnogonida. W. H. BALL. XIV." The Molluslis. C. R. EASTMAN. VII.' The Sharks- Teeth. S. GARMAN. XII." The Reptiles. H. J. HANSEN. The Cirripeds. H. J. HANSEN. XXVII." The Schi- zopods. S. HENSHAW. The Insects. W. E. HOYLE. The Cephalopods. W. C. KENDALL and L! RADCLIFFE. XXV." The Fishes. C.A. KOFOID. III.> IX.« XX.^o The Protozoa. C. A. KOFOID and J. R. MICHENER. XXII." The Protozoa. C. A. KOFOID and E. J. RIGDEN. XXIV.24 The Protozoa. P. KRUMBACH. The Sagittae. R. VON LENDENFELD. XXI." The Siliceous Sponges. G. W. MiJLLER. The Ostracods. JOHN MURRAY and G. V. LEE. XVII." The Bottom Specimens. MARY J. RATHBUN. X.i" The Crus- tacea Decapoda. HARRIET RICHARDSON. 11.' The Isopods. W. E. RITTER. IV.« The Tunicates. B. L. ROBINSON. The Plants. G. O. SARS. The Copepods. F. E. SCHULZE. XL" The Xenophyo- phoras. HARRIET R. SEARLE. XXVIII.m Isopods. H. R. SIMROTH. Pteropods, Hetero- pods. E. C. STARKS. XIII. 13 Atelaxla, TH. STUDER. The Alcyonaria. JH. THIELE. XV.15 Bathysciadium. T. W. VAUGHAN. VI.« The Corals. R. WOLTERECK. XVIII. >8 The Am- phipods. 1 Bull. M. C. z. Vol. » Bull. M. C. z. Vol. > BiUl. M. C. z. Vol. * Bull. M. C. z. Vol. 6 Mem M C. z. Vol. 6 Bull. M. C. z. Vol. ' Bull.' M. C. z. Vol. 8 Mem M. C. z. Vol. 9 Bull. M. c. z. Vol. 10 Mem M. c. z. Vol. " Bull. M. c. z. Vol. ■» Bull. M. c. z. Vol. " Bull. M. c. z. Vol. » Bull. M. c. z. Vol. " Bull. M. c. z. Vol. i« Mem M. c. z. Vol. I'Mem M. c. z. Vol. 18 Bull. M. c. z. Vol. » Bull. M. c. z. Vol. »" Bull. M. c. z. Vol. '1 Mem M. c. z. Vol. " Bull. M. c. z. Vol. "Mem M c z. Vol. " Bull. M c. z. VoL MMem M c. z. Vol. M Bull. M c. z. Vol. "Mem M c. z. Vol. w Bull. M. c. z. Vol. XLVL, No. 4. April. 1905, 22 pp. XLVL, No. 6. July. 1905, 4 pp., 1 pi. XLVI., No. 9, September, 1905, 5 pp., 1 pi. XLVL, No. 13, January, 1906, 22 pp., 3 pis. XXXIII., January, 1906, 90 pp., 96 pis. L.. No. 3, August, 1906, 14 pp., 10 pis. L., No. 4, November, 1906, 26 pp., 4 pis. XXXV., No. 1, February, 1907, 20 pp., 15 pis. L., No. 6, February, 1907, 48 pp., 18 pis. XXXV. No. 2, August, 1907. 56 pp., 9 pis. LI., No. 6,. November, 1907, 22 pp.. 1 pi. LIL, No. 1, June, 1908, 14 pp., 1 pi. LIL, No. 2, July, 1908, 8 pp.. ^ pis. XLIIL. No. 6, October, 1908, 285 pp., 22 pis. LIL, No. 5, October, 1908, 11 pp., 2 pis. XXXVII. , February, 1909, 243 pp., 48 pis. XXXVIII., No. 1, June, 1909, 172 pp., 5 pis.. 3 LIL, No. 9, June. 1909, 26 pp., 8 pis. LIL, No. 11, August, 1909, 10 pp., 3 pis. LIL, No. 13, September, 1909, 48 pp.. 4 pis. XLL, August, September, 1910, 323 pp., 56 pis. LIV.. No. 7, August, 1911, 38 pp. XXXVIII., No. 2, December, 1911, 232 pp., 32 LIV., No. 10, February, 1912. 16 pp., 2 pis. XXXV., No. 3, April. 1912, 98 pp., S pis. LIV, No. 12. April. 1912, 38 pp.. 2 pis. XXXV., No. 4, July. 1912, 124 pp., 12 pis. LVIIL, No. 8. August, 1914, 14 pp. Bulletin of the Museum of Comparative ZoSlogy AT HARVARD COLLEGE. Vol. LIX. No. 1. MAMMALS OBTAINED BY THE PHILLIPS PALESTINE EXPEDITION. By Glover M. Allen. CAMBRIDGE, MASS., U. S. A.: PRINTED FOR THE MUSEUM. February, 1915. No. 1. — Mammals obtained by the Phillips Palestine Expedition. By Glover M. Allen. The collection here reported on consists of about one hundred and fifty small mammals, mostly skins with skulls, presented to the Mu- seum of Comparative Zoology by Dr. John C. Phillips, who obtained them during his expedition (March- June, 1914) to the Sinai Peninsula and Palestine. His route was from Cairo eastward along the west coast of the Gulf of Suez, to Mt. Sinai, thence northward to Akaba, at the head of the Gulf of Akaba, and on to the region of the Dead Sea. Mr. William M. Mann, who accompanied him, made further collections about the base of Mt. Hermon. Twenty-four species were obtained, some of which are yet very imperfectly known. The ranges of others are slightly extended by Dr. Phillips's collections. The country to the south of Syria is inhabited by a typical desert fauna of genera which for the most part are not found north of the Dead Sea region, where they give place to more strictly Palaearctic types, as Apodemus, Microtus, Eliomys. No new species were discovered, but the fine series of Apodemus mystacinus from the vicinity of Mt. Hermon makes it possible to determine the status of the form found in the Black Sea forest of Asia Minor, and it is here described as new. Crocidura russula (Hermann). White-toothed Shrew. Sorex russulus Hermann, Zimmermann's Geogr. geschichte, 1780, 2, p. 382. No specimens of this genus were met with except in the country about the western base of Mt. Hermon, at Rasheya, Baniyas, Ammik, and Aithenit. The series of seven skins is of very uniform appearance and seems to be indistinguishable from typical russula. PiPISTRELLUS KUHLII (Kuhl). Kuhl's Bat. Vespertilio kuhlii Kuhl, Ann. Wetterau. ges. naturk., 1819, 4, p. 199. A single specimen of this small species was taken at Shtora, Syria. 4 bulletin: museum of comparative zoology. Eptesicus serotinus (Schreber). Serotine Bat. Vespertilio serotinus Schreber, Saugethiere, 1774, 1, pi. 53; 1775, 1, p. 167 (description). This common species of southern and central Europe was taken once at Shtora, Syria. Taphozous nudiventris Cretzschmar. Tomb-haunting Bat. Taphozous nudiventris Cretzschmar, Riippell's Atlas reise nordl. Afrika. Siiugeth., 1826, p. 70, fig. 27b. A small series was shot at evening from the walls of Jericho, which must be near its northw^ard limit in Palestine. Canis aureus Linne. Jackal. Canis aureus Linne, Syst. nat., ed. 10, 1758, 1, p. 40. Throughout much of the country traversed, jackals were common and frequently proved a great nuisance by following the lines of traps and carrying off both trap and mouse. One specimen brought back has a deformed lower jaw, which is so much shorter than the upper, that the lower canines close behind the upper ones. All the teeth seem normally formed, however, but on account of the shortness of the ramus the premolar series is greatly crowded. Eliomys melanurus Wagner. Black-tailed Dormouse. Eliomys melanurus Wagner, Abh. K. Bayer, akad. Miinchen, Math.-phys. cl., 1843, 3, p. 176, pi. 3, fig. 1. This beautiful dormouse was met with near the west base of Mt. Hermon, where at Ain.Hersha and Rasheya, three specimens were taken. One from the former locality. May 31, is a young individual. It is known also from the Sinai region. ALLEN: MAMMALS OF THE PHILLIPS PALESTINE EXPEDITION. 5 PsAMMOMYS OBESUS Cretzschmar. Sand Mouse. Psammomys obesus Cretzschmar, Riippell's Atlas reise nordl. Afrika. Saugeth., 1826, p. 58, pi. 22, 23. At Ain xA.bu Heran, to the north of Akaba, Dr. PhilHps shot a sub- adult male which seems referable to this species. The type locality is Alexandria, so that this record extends its range well to the eastward. The Museum has also a skin from Palestine, without definite locality, but probably from near the southwest coast. The long-tailed P. terraesandae of the Dead Sea region seems to be a species distinct from the shorter-tailed obesus and algiricus. Meriones tristrami Thomas. Tristram's Gerbil. Meriones tristrami Thomas, Ann. mag. nat. hist., 1892, ser. 6, 9, p. 148. This species was first described on the basis of specimens from the Dead Sea region and Mt. Carmel collected by Canon Tristram, who referred it to M. tamaricinus. Nehring (Sitzb. Ges. naturf. freunde Berlin, 1901, p. 171) records a specimen from the south of Jaffa on the coast, west of the Dead Sea. On his way north from Akaba, Dr. Phillips first met with this gerbil at Shobek, and on successive days, trapped it at Ain Gleidat and Tafileh. The locality first-named, probably represents nearly its southern limit, as it was not found on the high plateau country to the south. Three individuals, not quite fully grown, lack the bright sandy color of the adults, and are decidedly grayer. A young one from Beir el Doleh, Syria, is more fulvous. Meriones crassus Sundevall. Silky Gerbil. Meriones crassus Sundevall, K. Vet. akad. Handl. for 1842, 1843, p. 233, pi. 2, fig. 4, a-d. A single specimen sent by the Swedish traveller Hedenborg, served as Sundevall's type of this remarkable species. Hedenborg's note bulletin: museum of comparative zoology. accompanying it gives its habitat as the Sinai desert, " Ad vias circa fontes Mosis." The Wells of Moses (Ain Musa) near the west shore of the Gulf of Suez, may therefore be considered the type locality. Dr. Phillips obtained two adults near Mt. Sinai, at Wady Feiran and Um Shomer respectively, and a third on the eastern side of the Sinai peninsula, at Suweira, slightly to the north of Akaba. It therefore probably ranges over the greater part of the Sinai desert. Bonhote (Proc. Zool. soc. London, 1912, p. 226) has recorded a specimen from Tor in Sinai, collected by Capt. S. S. Flower. The peculiar inflation of the auditory meatus causing it to touch the angle of the squamosal process, and the posterior enlargement of the bullae, so that they extend behind the supraoccipital and notably constrict the exoccipitals, may prove to be characters of generic value, when the time comes for a revision of the group. The pelage is extremely soft and silky ; the pale, sand-colored hairs of the upper surface of the body are minutely black-tipped. Gerbillus calurus Thomas. Bushy-tailed Gerbil. Gerbillus calurus Thomas, Ann. mag. nat. hist., 1892, ser. 6, 9, p. 76. One of the most interesting of Dr. Phillips's captures is a fine adult male of this rare gerbil. Hitherto but three specimens seem to have been recorded, all of which are in the British Museum. The original specimen is an alcoholic without locality ; the second, also an alcoholic, is from Sinai, and unfortunately in poor condition ; the third is a skin with imperfect skull, from Wady Sikait, south of Gebel Sebara, eastern Egypt. Dr. Phillips's specimen (the fourth to be recorded) is from the Sinai region at Wady Sa'al. The type was for many years in the British Museum before it was made known by Thomas, and it was not till the publication of the two other records by x\nderson in his Zoology of Egypt, 2, Mammals, in 1902, that its probable range was indicated. The squirrel-like tail is a remarkable feature in the genus, but the skull seems sufficiently typical. The measurements of the present specimen are: — head and body 118 mm., tail 145, hind foot 33, ear 22; skull, greatest length 36.5, basal length 30, palatal length 19.3, diastema 8.5, zygomatic width 18.7, mastoid width 18.9, inter- orbital constriction 5.5, bullae 14 X 7.5, upper molar row (alveoli) ALLEN: MAMMALS OF THE PHILLIPS PALESTINE EXPEDITION. 7 5.6. Posteriorly the inflated mastoids project slightly beyond the occipital region. Gerbillus gerbillus (Olivier). Tawny Gerbil. Dipus gerbillus Olivier, Voy. Egypt., 1801, 3, p. 157, pi. 28; Bull. Soc. philom. Paris, 1801, 2, p. 121. This brightly colored gerbil was first trapped at Wady Shurandel in the Sinai region. Other specimens were taken at the head of the Gulf of Akaba to the northeast, namely at Akaba and Suweira, but none has been recorded to the northward of these places. DiPODiLLUS QUADRiMACULATUS Lataste. Four-spotted Gerbil. Dipodillus quadrimaculatus Lataste, Le naturaliste, 1882, 4, p. 27. A series of six specimens from Akaba, at the head of the Gulf of the same name, appears to represent this species, and extends its known range somewhat to the eastward. Its apparent absence from the high rough country of the interior of the Sinai peninsula may indicate that it is confined to the low sandy areas along the coast. Dipodillus dasyuroides Nehring. Nehring's Smooth-footed Gerbil. Dipodillus dasyuroides Nehring, Sitzb. Ges. naturf. freunde Berlin, 1901, p. 173. A series of thirteen skins, old and young, seems referable to Neh- ring's species, the type of which is from Mount Moab, east of the southern end of the Dead Sea. The chief color character distinguish- ing it from Wagner's dasyurus of western Arabia is said to be the yellowish instead of pure white area above the eyes. In the series before me there is some variation in tint, chiefly due to the greater or less suffusion of the upper parts with buffy. This seems partly a matter of age, since the young and subadults are less buffy, the pale area above the eyes is dirty white, and the ventral side of the tail is 5 bulletin: museum of comparative zoology. white. In the adults with worn teeth the entire pelage above is huffier, as well as the eye spots and ventral side of the tail. The adult males are brighter buff or fulvous as compared with the females, which even in the adult, seem grayer, like the young. One specimen has the extreme tip of the tail white. The type locality, Moab, must be near the northern limit of its range. Dr. Phillips obtained it at Suweira, Nuheibeh, and Um Shomer in the Sinai region, then farther north at Petra, and in Syria at Wady Hesa, Wady Ain Musa, and Beir el Doleh. Several young specimens from one third to one half the adult size were collected in late April and early May at Petra and Beir el Doleh. DiPODiLLUS mariae Bonhote. Mrs. Bonhote's Pigmy Gerbil. Dipodillus mariae Bonhote, Proc. Zool. soc. London, 1909, p. 792. This minute grayish species was but recently described on the basis of two specimens from the Mokattam Hills, near Cairo, Egypt. A single male collected by Dr. Phillips at Wady Feiran, Sinai, corre- sponds completely with the published description, and seems thus to represent the third recorded specimen. The known range of the species is extended considerably to the eastward by this capture. MiCROTUS GUENTHERi (Danford and Alston). Guenther's Vole. Arvicola guentheri Danford and Alston, Proc. Zool. soc. London, 1880, p. 62, pi. 5. Eight specimens of a short-tailed yellowish vole I have provision- ally referred to guentheri, with the description of which they seem to agree. All are from localities in the valley west of Mt. Hermon. In the original diagnosis, the presence of five plantar tubercles is given as a chief distinguishing character, but in some specimens there seems to be a minute sixth one indicated. The ears project distinctly from the fur of the head, and instead of being well haired near their margins as stated by the describers of guentheri, they are clothed with very minute hairs and appear nearly naked unless narrowly examined. The relationship of this species to M. socialis is apparently close. ALLEN: MAMMALS OF THE PHILLIPS PALESTINE EXPEDITION. U Apodemus sylvaticus tauricus (Barrett-Hamilton). Taurus Wood Mouse. Mus sylvaticus tauricus Barrett-Hamilton, Proc. Zool. soc. London, 1900, p. 412. Three specimens from Shiba, Rasheya, and Ain Hersha respectively, localities near the southwest base of Mt. Hermon, belong to the syl- vaticus group, and here represent nearly the southern limit of its range in Palestine. Two of these, though nearly growTi, are in the immature slaty gray pelage, and the third is an adult male. Barrett-Hamilton's name tauricus probably applies to these specimens, though he gives no description beyond the length and breadth of the type skull (23 X 12 mm.). His type is an alcoholic in the British Museum from Zebil in the Bulgar Dagh, southern Asia Minor, and thus not very far from Palestine. The adult skin brought back by Dr. Phillips's expedition is a very pale buffy animal uniformly 'lined' above with black hairs and quite without the russet tints of true sylvaticus. Judging from descriptions alone it is nearly indistinguishable from A. s. dichrurus of the European Mediterranean region and in its measurements it shows no appreciable differences. The skull of the adult male is 24.6 mm. in greatest length as against 23 mm. given for the type of tauricus, but the latter measurement may well be within the limits of varia- bility for an immature individual. - Apodemus flavicollis (Melchior). Yellow-collared Mouse. Mus flavicollis Melchior, Den Danske Staats og Norges pattedyr, 1834, p. 99. Two specimens, one adult, the other immature, from Ain Hersha near the base of Mt. Hermon extend the recorded range of this species well into Palestine and probably indicate nearly the southeastern limit of its distribution. Through the kindness of Mr. Gerrit S. Miller, Jr., I have been able to compare these with a series of European flavicollis, including topotypes from Denmark, in the U. S. N. M. The adult, in russet pelage, is a mere shade paler than any of the 10 bulletin: museum of comparative zoology. European skins yet probably falls Avithin the limits of individual variation. Skins froni the Harz Moinitains of Germany and others from Switzerland match it very closely. The feet are a little small and the skull, compared with those from Europe having equally worn teeth, is a trifle smaller, yet in both these respects it can be duplicated in the European series. The braincase seems smaller, howe\er, and the angle formed by the sides of the frontoparietal suture is more acute Additional specimens from Palestine may show that the local representatives of the species are entitled to rank as a separate race. The immature specimen is in the slaty gray pelage, and though taken June 1st, is fairly well grown (total length 190 mm.), indicating as Barrett-Hamilton has suggested, that it breeds early in the year. Apodemus myst acinus (Danford and Alston). Gray ^Yood IMouse. Mus mystacinns Daiiford and Alston, Proc. Zool. soc. London, 1877, p. 279. A series of fourteen skins, young and adult, represents this species, which seems to be rare in collections. All are from the region about the base of Mt. Hermon, and correspond in all details with the original description. The young, unlike those of the sylvaticus group, are col- ored practically like the adults, though the fawn tints on the sides of the face and body brighten slightly with age. The original series in the British jSIuseum comprised three specimens, two at least in alcohol, collected in the Bulgar Dagh region of southern Asia Minor. The pale coloration is typical of the dry country in which this mouse lives, and Mr. Oldfieid Thomas (Ann.' mag. nat. hist., 1903, ser. 7, 12, p. 18S) has lately described an even paler race, A. m. smyniouns, from ex- treme western Asia ISIinor at Smyrna. In this race the hairs of the lower surfaces are pure white to the roots instead of having slaty bases. Through Mr. Thomas's kirrdness the M. C. Z. has received in ex- change a specimen referred to mystacinus taken in the forest belt bordering the Black Sea, an area \ery different faimally from the arid country to the south. Dr. Phillips's fine series representing typical mystacinus shows that the Black Sea animal, as might be expected, is very different in color. It is much darker, and almost without the buffv tints of the former. It mav bo known as ALLEN: MAMMALS OF THE PHILLIPS PALESTINE EXPEDITION. 11 Apodemus mystacinus euxinus, suhsp. nov. Black Sea Wood Mouse. T\jpc.— Skin and skull 14,887 M. C. Z., from Scalita (near Trebi- zond), Asia Minor; male, collected November 25, 1905, by A. Robert, altitude 1,000 meters. General Characters. — Similar to typical mystacinua but much darker, the back blacker, the buffy tints of face and sides replaced by grayish. Description. — The type specimen is subadult, and in coniparison with specimens of similar age from Palestine, is much darker through- out. The entire dorsal surfaces are grayish, heavily washed with black which predominates in the middle of the back. The sides of the head and body are paler gray very faintly washed with " pinkish- buff" but in much less degree than in the typical race so that the general appearance is dark gray. Along the sides of the body a faint band of 'pinkish buff' delimits the color of the dorsal surface from the white of the belly. The slaty bases of the hairs of the ventral surface show through sufficiently to darken the entire imderparts except on the forearms, which are pure white below. Feet and hands white, ankles slaty, with a dusky prolongation reaching the calcaneum behind, though the tarsal joint is white on its upper surface. Tail sharply bicolor, blackish above, white below. Measurements. — Head and body 94 mm., tail 109, hind foot 24, ear 18. The skull shows no appreciable differences from that of typical mystacinus; condylobasal length 26, palatal length 14, zygo- matic width 14, upper cheek teeth (alveoli) 4.5. Remarks. — Mr. Thomas has already described several new forms of mammals from the forest belt along the northern coast of Asia Minor on the Black Sea. Here, he sa^s, " there is a strip of forest, some 50 miles wide, sloping northwards to the Black Sea from an altitude of 1500 to 2000 metres at its southern edge. The forest then abruptly disappears and an open steppe country commences, inhabited by Hamsters and Spermophiles, and continuous with the more desert countries further south. Compared with this more open and desert country the coast-forest has a very different fauna, of a distinctly northern character. * * * Mr. Robert's work was done at two localities in the heart of the forest-strip — Sumela * * * and Scalita * * * a village in the same valley as Sumela but about 3000 m. [ = 300 m.?] lower," and some 30 or more miles south of Trebizond (Ann. mag. nat. hist., 1906, ser. 7, 17, p. 415). 12 bulletin: museum of comparative zoology. The dark color and lack of buffy patches on the sides of the head and behind the ears in the Black Sea form are no doubt correlated with life in this coastal forest. Apparently A. mystacinus is not closely related to A. epimelas of Europe which is sharply distinguished by the presence of a fourth minute tubercle at the posteroexternal margin of the first and second upper molars. Mus MuscuLUS Linne. House Mouse. Mus musculus Linn6, Syst. nat., ed. 10, 1758, 1, p. 62. Three skins from Akaba do not seem different from the form of House Mouse introduced into the eastern United States. Probably at Akaba the typical variety has been introduced by the shipping. Mus MUSCULUS ORiExNTALis Cretzschmar. Mus orientalis Cretzschmar, Riippell's Atlas reise nordl. Afrika. Saugeth., 1826, p. 76, pi. 30, fig. a. Four skins are pale-bellied, yet with conspicuous d'usky bases to the white-tipped hairs, and with a buffy line along the sides of the body. They are to be considered as representing orientalis though it seems questionable if they are not better referred to gentilis, of which they would be reckoned a dark extreme. The four specimens are from Akaba, Arabia, and from Rasheya, Hasbeiya, and Shiba, Syria (near Mt. Hermon). Mus MUSCULUS GENTILIS Brants. WTiite-bellied House Mouse. Mus gentilis Brants, Muizen, 1827, p. 126. This pale, white-bellied form was taken at Shobek in Arabia and at Wady Kerak and El Kerak in Syria. The hairs of the belly are clear white to the base, or with the very base only light plumbeous. Proba- bly these are the native form of House Mouse. ALLEN: MAMMALS OF THE PHILLIPS PALESTINE EXPEDITION. 13 AcoMYS RUSSATUS Wagner. Short-tailed Spiny Mouse. Acomys russatus Wagner, Abh. K. Bayer, akad. Miinchen, Math.-phys. cl., 1843, 3, p. 195, pi. 3, fig. 2. Of this rare species, two specimens were procured at Wady Feiran, in the dry rocky country of Sinai, and so are practically topotypes. Nehring (Sitzb. Ges. naturf. freunde Berlin, 1901), records one each from Moab and Engeddi, Palestine, and Tristram had previously found it at Massada at the south end of the Dead Sea. In describing as a distinct race the specimens he found in the Mokattam Hills, near Cairo, Bonhote (Proc. Zool. soc. London, 1912, p. 229) also mentions a pair from Sinai that he kept alive. The known range of the typical form is thus from the region of the Dead Sea through the Sinai peninsula. Acomys dimidiatus (Cretzschmar). Desert Spiny Mouse. Mus dimidiatus Cretzschmar, Riippell's Atlas reise nordl. Afrika. Saugeth., 1826, p. 37, pi. 13, fig. a. This is the commonest small rodent in the collection. ISIany specimens were taken in the Sinai region, at Akaba (head of the Gulf of Akaba) and northward at Petra and Tafileh. The most northerly specimen is from Wady Kerak at the southern end of the Dead Sea. Jaculus macrotarsus (Wagner). Long-footed Jerboa. Dipus macrotarsus Wagner, Abh. K. Bayer, akad. Miinchen, Math.-phys. cl., 1843, 3, p. 214, pi. 4, fig. 2. A single specimen from Wady Feiran, Mt. Sinai, is practically a topotype of this species, which was originally described from speci- mens sent from Mt. Sinai. Nehring (Sitzb. Ges. naturf. freunde Berlin, 1901, p. 163), in naming schliiferi from southwestern Palestine, compared it with examples from western Arabia, which he took to represent macrotarsus. It seems likely that in this he was correct. 14 bulletin: museum of comparative zoology. At all events the ventral hook-like process of the jugal is lacking in the Sinai specimen as in these, and they have two perforations of the angle of the jaw instead of one as in the other species. In color J . macrotarsus seems to be very much darker than J. jaculus by reason of the many dark-tipped hairs among the pale buffy fur of the back. These dark tips also extend to the sides of the belly, and give a soiled appearance to the white of this area. Among the vibrissae is a single one of great length on each side (some 104 mm.). The collector's measurements are: — total length 300 mm., tail 180, hind foot 55, ear 22.5. The skull measures: greatest median length 31.7 mm., basal length 28, palatal length 18.5, diastema 8.8, zygomatic breadth 21.5, width across malars 20.5, mastoid width 23, upper tooth row (alveoli) 5.1. Capra nubiana sinaitica Hemprich and Ehrenberg. Sinai Ibex. Capra sinaitica Hemprich and Ehrenberg, Symb. phys. zool., 1828, 1, pi. 18. Dr. Phillips supplies the following interesting note as to the present status of this animal. "The Sinai Ibex still persists over all the rugged parts of the Sinai peninsula, near Akaba and up at least as far as the northeast end of the Dead Sea. Although undoubtedly greatly reduced in numbers since Tristram's time (1884), it manages to persist in spite of the fact that every hand is against it during the entire year, and its freshly dropped kids are eagerly hunted by the natives with dogs. I hunted three days and saw only four smallish animals, but signs were fairly numer- ous. The Ibex appears to be independent of water, at least during winter and spring. The leopard hunts these Ibexes and presumably kills a good many, as various sportsmen have testified. We obtained a new born kid at Feiran, March 30th, and another at Akaba, April 16th." Dr. Phillips found evidence that they frequent caves among the rocks as hiding places. The following Publications of the Museum in preparation; - of Comparative Zoology are LOUIS CABOT. Immature State of the Odonata, Part IV. E. L. MARK. Studies on Lepidosteus, continued. E. L. MARK. On Arachnactis. A. AGASSIZ and C. O. WHITMAN. Pelagic Fislies. Part II., with 14 Plates. H. L. CLARK. The "Albatross" Hawaiian Echini. Reports on the Results of Dredging Operations in 1877. 1878. 1879, and 1880, in charge of Alexander Agassiz, by the U. S. Coast Survey Steamer "Blake," as follows: — A. MILNE EDWARDS and E. L. BOUVIER. The Crustacea of the "Blake." A. E. VERRILL. The Alcyonaria of the "Blake." Reports on the Results of the Expedition of 1891 of the U S. Fish Commission Steamer "Albatross," Lieutenant Commander Z. L. Tanner, U. S. N., Commanding, in charge of Alexander Agassiz, as follows:— K. BRANDT. The Sagittae. K. BRANDT. The Thalassicolae. O. CARLGREN. The Actinarians. R. V. CHAMBERLIN. The Annelids. W. R. COE. The Nemerteans. REINHARD DOHRN. The Eyes of Deep-Sea Crustacea. H. J. HANSEN. The Cirripeds. H. J. HANSEN. The Schizopods. HAROLD HEATH. Solenogaster. Reports on the Scientific Results of the Expedition to the Tropical Pacific, in charge of Alexander Agassiz, on the U. S. Fish Commission Steamer "Albatross," from August, 1899, to March, 1900, Commander Jefferson F. Moser, U. S. N., Com- manding, as follows: — W. A. HERDMAN. The Ascidians. S. J. HICKSON. The Antipathids. E. L. MARK. Branchiocerianthus. JOHN MURRAY. The Bottom Speci- mens. P. SCHIBMENZ. The Pteropods and Heteropods. THEO. STUDER. The Alcyonarians. The Salpidae and Doliolidae. H. B. WARD. The Sipunculids. R. V. CHAMBERLIN. The Annelids. H. L. CLARK. The Holothurians. H. L. CLARK. The Ophiurans. The Volcanic Rocks. The Coralliferous Limestones. S. HENSHAW. The Insects. R. VON LENDENFELD. The Siliceous Sponges. G. W. MtJLLER. The Ostracods. MARY J. RATHBUN. The Crustacea Decapoda. G. O. SARS. The Copepods. L. STEJNEGER. The Reptiles. C. H. TOWNSEND. The Mammals, Birds, and Fishes. T. W. VAUGHAN. Thei Corals, Recent and Fossil. PUBLICATIONS OF THE MUSEUM OF COMPARATIVE ZOOLOGY AT HARVARD COLLEGE. There have been published of the Bulletin Vols. I. to LIV., and Vol. LVIIL; of the Memoirs, Vols. I. to XXIV., and also Vols. XXVI. to XXIX., XXXI. to XXXIV., XXXVI. to XXXVIIL, XLI., and XLIV. Vols. LV. to LVII. and LIX. of the Bulletin, and Vols. XXV., XXX., XXXV., XXXIX., XL., XLIL, XLIIL, XLV. to XLVIII. of the Memoirs, are now in course of publication. The Bulletin and Memoirs are devoted to the publication of original work by the Officers of the Museum, of investigations carried on by students and others in the different Laboratories of Natural History, and of work by specialists based upon the Museum Collections and Explorations. The following publications are in preparation : — Reports on the Results of Dredging Operations from 1877 to 1880, in charge of Alexander Agassiz, by the U. S. Coast Survey Steamer "Blake," Lieut. Commander C. D. Sigsbee, U. S. N., and Commander J. R. Bartlett, U. S. N., Commanding. Reports on the Results of the Expedition of 1891 of the U. S. Fish Commission Steamer "Albatross," Lieut. Commander Z. L. Tanner, U. S. N., Com- manding, in charge of Alexander Agassiz. Reports on the Scientific Results of the Expedition to the Tropical Pacific, in charge of Alexander Agassiz, on the U. S. Fish Commission Steamer "Albatross," from August, 1899, to March, 1900, Commander Jefferson F. Moser, U. S. N., Commanding. Reports on the Scientific Results of the Expedition to the Eastern Tropical Pacific, in charge of Alexander Agassiz, on the U. S. Fish Commission Steamer "Albatross," from October, 1904, to April, 1905, Lieut. Com- mander L. M. Garrett, U. S. N., Commanding. Contributions from the Zoological Laboratory, Professor E. L. Mark, Director Contributions from the Geological Laboratory, Professor R. A. Daly, in charge. These publications are issued in numbers at irregular intervals. Each number of the Bulletin and of the Memoirs is sold separately. A price list of the publications of the Museum will be sent on appli- cation to the Director of the Museum of Comparative Zoology, Cambridge, Mass. iV^^ Bulletin of the Museum of Comparative Zoology AT HARVARD COLLEGE. Vol. LIX. No. 2. 9 THE CRANIAL NERVES OF ANOLIS CAROLINENSTS. By William A. Willard. With Seven Plates. CAMBRIDGE, MASS., U. S. A.: PRINTED FOR THE MUSEUM. July, 1915. Reports on the Scientific Results of the Expedition to the East- ern Tropical Pacific, in charge of Alexander Agassiz, by the U. S. Fish Commission Steamer "Albatross," from October, 1904, TO March, 1905, Lieutenant Commander L. M. Garrett, U. S. N., Commanding, published or in preparation: — A. AGASSIZ. V.6 General Report on the Expedition. A. AGASSIZ. I.i Tliree Letters to Geo. M. Bowers. U. S. Fisli Com. A. AGASSIZ and H. L. CLARK. The Echini. H. B. BIGELOW. XVI. »» The Medusae. H. B. BIGELOW. XXIII.23 The Sipho- nophores. H. B. BIGELOW. XXVI.28 The Cteno- phores. R. P. BIGELOW. The Stomatopods. O. CARLGREN. . The Actinaria. R. V. CHAMBERLIN. The Annelids. H. L. CLARK. The Hoiothurians. H. L. CLARK. The Starfishes. H. L. CLARK. The Ophiurans. S. F. CLARKE. VIII.' The Hydroids. W. R. COE. The Nemerteans. L. J. COLE. XIX." The Pycnogonida. W. H. DALL. XIV.H The Mollusks. C. R. EASTMAN. VII.' The Sharks' Teeth. S. GARMAN. XII." The Reptiles. H. J. HANSEN. The Cirripeds. H. J. HANSEN. XXVII." The Schi- zopods. S. HENSHAW. The Insects. W. E. HOYLE. The Cephalopods. W. C. KENDALL and L. RADCLIPFE. XXV.JB The Fishes. C. A. KOFOID. III.3 IX.« XX. 2« The Protozoa. C. A. KOFOID and J. R. MICHENER. XXII.22 The Protozoa. C. A. KOFOID ana E. J. RIGDEN. XXIV.24 The Protozoa. P. KRUMBACH. The Sagittae. R. VON LENDENFELD. XXI." The Siliceous Sponges. R. VON LENDENFELD. XXIX.^' HexactineUida. G. W. MULLER. The Ostracods. JOHN MURRAY and G. V. LEE. XVII." the Bottom Specimens. MARY J. RATHBUN. X.«> The Crus- tacea Decapoda. HARRIET RICHARDSON. II.2 The Isopods. W. E. RITTER. IV.< The Tunicates. B. L. ROBINSON. The Plants. G. O. SARS. Th e Copepods. F. E. SCHULZE. XI." The Xenophyo- phoras. HARRIET R. SEARLE. XXVIII. =8 Isopods. H. R. SIMROTH. Pteropods, Hetero- pods. E. C. STARKS. XIII.13 Atelaxia TH. STUDER. The Alcyonaria. JH. THIELE. XV.16 Bathysciadium. T. W. VAUGHAN. VI.« The Corals. R. WOLTERECK. • XVIII. "s The Am- phipods. > Bull. M. C. Z.. « Bull. M. C. Z., 'Bull. M. C. Z., 4 Bull. M. C. Z.. 'Mem. M. C. Z., 6 Bull. M. C. Z., ' Bull. M. C. Z. 8 Mem M. C. Z. »Bull. M. C. Z., "Mem M. C. Z. n Bull. M. C. Z. 12 Bull. M. C. Z. 13 Bull. M. C. Z. H Bull. M. C. Z. 's Bull. M. C. Z. "Mem M. C. Z. I'Mem M. C. Z. '8 Bull. M. C. Z. 19 Bull. M. C. Z. »» Bull. M. C. Z. 21 Mem M. C. Z. " Bull. M. C. Z. "Mem M. C. Z. " Bull. M. C. Z. ssMem M. C. Z. 29 Bull. M. C. Z. 2' Mem M. C. Z. 28 Bull. M. C. Z. 2»Mcm . M. C. Z. Vol. XLVI., No. 4, April, 1905, 22 pp. Vol. XLVI., No. 6, July. 1905, 4 pp., 1 pi. Vol. XLVI., No. 9, September, 1905, 5 pp., 1 pi. Vol. XLVI.. No. 13, January, 1906, 22 pp., 3 pis. Vol. XXXIII.. January, 1906, 90 pp., 96 pis. Vol. L.. No. 3, August, 1906, 14 pp., 10 pis. Vol. L., No. 4, November, 1906, 26 pp., 4 pis. Vol. XXXV., No. 1, February, 1907. 20 pp., 15 pis. Vol. L., No. 6, February, 1907, 48 pp., 18 pis. Vol. XXXV. No. 2, August, 1907. 56 pp., 9 pis. Vol. LI., No. 6, November, 1907, 22 pp., 1 pi. Vol. LIL, No. 1, June, 1908, 14 pp., 1 pi. Vol. LIL, No. 2, July, 1908, 8 pp., 5 pis. Vol. XLIIL. No. 6, October, 1908, 285 pp., 22 pis. Vol. LII., No. 5, October, 1908, 11 pp., 2 pis. Vol. XXXVII., February, 1909, 243 pp., 48 pis. Vol. XXXVIII. , No. 1, June, 1909. 172 pp., 5 pis., 3 maps. . Vol. LIL, No. 9, June, 1909, 26 pp., 8 pis. Vol. LIL, No. 11, August, 1909, 10 pp., 3 pis. Vol. LIL, No. 13, September, 1909. 48 pp.. 4 pis. Vol. XLI., August, September, 1910, 323 pp.. 56 pis. , Vol. LIV.. No. 7, August, 1911, 38 pp. , Vol. XXXVIIL, No. 2, December, 1911, 232 pp., 32 pl.s. , Vol. LIV., No. 10. February. 1912, 16 pp.. 2 pis. , Vol. XXXV., No. 3. April, 1912, 98 pp., 8 pis. , Vol. LIV, No. 12, April, 1912, 38 pp., 2 pis. , Vol. XXXV.. No. 4, July. 1912. 124 pp.. 12 pis. Vol. LVIII , No. 8. August. 1914. 14 pp. , Vol. XLII.. June. 1915. 397 pp., 109 pis. JUl 2o ;,;:; Bulletin of the Museum of Comparative Zoology AT HARVARD COLLEGE. Vol. LIX. No. 2. THE CRANIAL NERVES OF ANOLIS CAROLINENSIS. By William A. Willard. With Seven Plates. CAMBRIDGE, MASS., U. S. A.: PRINTED FOR THE MUSEUM. July, 1915. No. -2. — The Cranial Nerves of Anolis carolinensis. CONTRIBUTIONS FROM THE ZOOLOGICAL LABORATORY OF THE MUSEUM OF COMPARATIVE ZOOLOGY AT HARVARD COLLEGE. NO. 261. By William A. Willard. TABLE OF CONTENTS. PAGE A. Introduction .......... 18 B. Scope of the paper 19 C. Material and methods 20 D. Systematic position of Anolis ....... 22 E. Historical statement 23 F. Skeletal structures 24 1. Skull 24 2. Hyoid apparatus 25 G. Cutaneous sense organs 25 H. Visceral end organs 30 I. Muscles 31 1. Muscles of the orbit 31 2. Muscles of the tongue 34 3. Muscles of the hyoid apparatus 35 4. Muscles of the jaw and neck 38 J. Oculomotor nerve 44 K. Trochlear nerve .......... 45 L. Abducent nerve .......... 45 Ciliary ganglion and nerves ....... 45 M. Trigeminal nerve 48 N. Facial nerve 62 O. Glossopharyngeal nerve 71 P. Vagus nerve .......... 73 1. Relation between nerves IX and X .... 76 2. A vestigial dorsal ganglion on the roots of the vagus . 77 3. Terminal distribution of nerves IX and X ... 80 Q. Spinal accessory nerve 85 R. Hypoglossal nerve 88 S. Summary on the distribution of nerves IX, X and XII . . 92 '1\ Additional innervation to the ventral head region ... 92 18 bulletin: museum of comparative zoology. U. Spinal nerves 93 1. Third spinal nerve 93 2. Second spinal nerve ....... 95 3. First spinal nerve ........ 95 4. Connections of first three spinal nerves with sympathetic. 95 V. General considerations on nerve XII and the spinal nerves . . 96 W. General summary 98 Bibliography 101 Explanation of plates 109 A. INTRODUCTION. It is now generally recognized that the comparative anatomy of the nervous system can not be profitably studied without recoui'se to microscopic analysis in the peripheral region as well as in the central organs. It is further recognized that, for their final explanation, the highly complex nervous organs of mammals, including man, must await the revisions and additions which such a method of study, applied to the lower vertebrates, will suggest for their interpretation. Inasmuch as the nervous system is essentially a correlating structure, it follows naturally that changes in its mechanism should be studied in connection with a study of the correlated organs and functions. This gives to the study of the nervous system a broader biological significance than is possessed by other organs. But this relation adds much to the technical difficulties and makes it almost impossible to bring within the scope of a single research the complete analysis of even a single form. The shortcomings of the present study from that point of view are recognized by the author. However, in the methods pursued, he has had constantly in mind the larger problem, which does not recognize the arbitrary distinction between central and pe- ripheral nervous systems. In fact, he has already collected much material and information on the central nervous system of Anolis; but it seems advisable to reserve that material for a more complete account of the Anolis brain. The present investigation is carried far enough toward the nerve centers to articulate easily with such a study. A large part of this work was done a number of years ago in the Zoological Laboratory of Harvard University. Since then it has been continued in the Laboratory of Histology and Embryology of the University of Nebraska, College of Medicine. In this connection I wish to acknowledge my gratitude to Dr. E. L. Mark for placing every will.'S.rd: cranial nerves of anolis carolinensis. 19 ■facility at my disposal and for his continued interest and encourage- ment. Also the work has been aided not a little through the courtesy of Professor H. W. Norris of Grinnell College and Dr. C. J. Herrick of Chicago University, who on different occasions have given me work- ing facilities in their laboratories. B. SCOPE OF THE PAPER. This paper includes, to some extent, the general anatomy of the head of Anolis; much of this, however, is of quite secondary impor- tance and there has been no attempt to treat critically anything but the nervous organs; even within this field certain features have been omitted, since they can better be considered in connection with the central nervous system. This is true of the larger sense organs and their nerves, i. e., eye, ear, and olfactory organs. Also the distribu- tion of the vagus nerve is not carried into the trunk region beyond the limits of the series of sections figured. Of the non-nervous structures, the skeleton is included for the purpose of more exact topography. The muscles and glands, and the integument and mucous membrane with their sense organs are included on account of their relation to the peripheral distribution of neurones. The aim of this work is to give as complete an account as possible of a single reptilian form, which may serve as a basis for further com- parative study. The entire absence of such an account within the whole sauropsidan group is believed to be sufficient justification for the publication of the present paper. But this plan is not compatible with the intensive treatment of many of the problems that arise in connection with various details of the work. Much of the literature that has been consulted in the course of the study has not been specifi- cally cited as it would have been, had the field been more limited. This is particularly true of most of the reptilian studies whose results are based on dissections alone and are therefore open to more than one interpretation, because an attempt at detailed comparisons in such cases would serve only to impair the usefulness of the present study. In regard to the Icthyopsida, where more exact work has been done, it seems premature to go far with comparisons until the study of the Sauropsida has covered several forms. Such comparisons as are made should, therefore, be considered tentative. 20 bulletin: museum of comparative zoology. C. MATERIAL AND METHODS. The adults of Anolis carolinensis were obtained from Jacksonville, Florida, and later from Colmesneil, Texas. They were sent alive by express, being received in good condition. After arriving they could be kept alive for any length of time by feeding with living flies. With such material constantly on hand, the use of the various special nerve methods was practicable. The methods of staining found best adapted to the purpose of the present study were Weigert's and Vom Rath's. The projection drawings of the peripheral distribution of the cranial nerves were made from a head prepared by the Vom Rath method, which was found the best for this material. The animals were killed with chloroform; the head, with the anterior part of the body as far back as the region of the fore legs, was removed and, after some of the tissue on the right side had been cut away so as to expose the brain and insure more rapid penetration, was put into a Vom Rath's solution (formula for vertebrates). The length of time necessary in this fluid depends upon the requirements of decalcification, for the fluid serves the double purpose of decalcify- ing the bone and impregnating the medullary sheaths of the nerve 'fibers. I found that a week to ten days served the purpose well.. After the required length of time the specimen was brought through the grades of alcohol and left in 80% alcohol long enough to remove as much of the picric acid as would readily come out. The after treatment with pjToligneous acid was not used. From 100% alcohol it was cleared in cedar oil, then passed into xylol and, finally, embedded in paraffin. For better embedding it was found ad\-antageous to use the air pump, either while the object was in the melted paraffin or before, while in the oil. It was alwa^'s found necessary to re-embed several times during the cutting of a complete series, as there were cavities not filled at first. This resulted at several places in the series in 'partial sections which, however, were numbered in sequence with the others. The series from which the large plots were made was cut transversely 10 micra thick. In place of artificial orientation lines the median plane was used for projection on the frontal plane, but for projection on the sagittal plane there was no natural line a^■ailable for the whole distance. The border of the upper lip served for this purpose as far back as the corner of the mouth. From this point caudad the orientation was determined from a comparison with a drawing of the lateral external view of the same head, and also with sagittal sections of heads of the same size. willard: cranial nerves of anolis carolinensis. 21 The projection plots were made from camera drawings magnified 3T diameters and, except in regions of greatest complication, each nerve was projected accurately throughout its entire course. In a few cases some of the nerves are displaced a little from the position they really occupy in the middle of their course in order to avoid too great confusion of lines. Of course the essential points such as central and peripheral endings and relationship of branches given off in the intermediate course are strictly adhered to. Even after the most careful study of serial sections and reconstruc- tions, one is likely to fall into error unless the work is supplemented by free-hand dissections. This is particularly true in case of very small branching rami or fine plexuses. Animals that are small enough to be practicable for sectioning are generally too small for entire dis- sections, but by the following method it was possible to make accurate and fairly complete dissections of the head : — the integument of the part wanted for dissection was carefully removed to avoid cutting, any of the underlying muscles. Then the animal was put into the Vom Rath fluid, as for sectioning, but treated a shorter time (24 to 36 hours). This decalcifies sufficiently well, hardens the muscles and leaves them well defined because of a slight shrinkage; it also hardens and blackens the nerves. Instead of further hardening in alcohol, the specimen was washed out in water and put into a mixture of alcohol and glycerine. It was afterwards dissected in water under a lens. For best results, however, it should be dissected soon. Although the whole muscle is much darkened, there is contrast enough between it and the nerves to allow the identification of the finest branches of the blackened nerves. The muscles are also more or less brittle, whereas the nerves retain their characteristic toughness. As the dissection progressed portions were removed and cleared for permanent mounts in balsam. I consider the making of balsam preparations a very valuable part of the technique of this work, for in no other way can the fiber course in the finer anastomoses be made out. For ascertain- ing the courses of nerves and for topographical relations this short Vom Rath method is a valuable adjunct and the two methods — sections and dissections — were constantly used together. For the central relations I found that properly prepared Weigert preparations gave more satisfactory results than the Vom Rath, but the latter method allowed the roots to be followed well into the brain and in some instances served even there better than the Weigert. Portions of the courses brought out by one method overlap those by the other, so there could be no chance of error in combining these two methods of study. 22L bulletin: museum of comparative zoology. For the best Weigert preparations, Herriek's method of fixation for fishes with Flemming's fluid was found the best for the lizard also, although I was unsuccessful in my attempts to get the best results in l)Oth the central and peripheral fibers by a single treatment. That is to say, when the sections were decolorized properly for the central nervous fibers, the peripheral nerves were decolorized too far. On the whole I found it much more satisfactory to base the study of the peripheral nerves on the Vom Rath series, and the present paper chiefly rests on the findings in such material. D. SYSTEMATIC POSITION OF AXOLIS. Following Cope (:00, p. 158, ff.), we have the following division of the living Reptilia: — Class Monocondylia. Subclass Reptilia Orders Testudinata Loricata Rh}Tichocephalia Squamata Suborders Ophidia Sauria The group characters of the Sauria are as follows: "Quadrate bone articulating with the exoccipital ; parietal bones not closing the brain case in front; generally an epipterygoid and sternum; teeth with dentinal roots; phalanges with condyles" (p. 178). The family Iguanidae is represented in North America by twelve genera; it is subdivided as follows : — Subfamilies Anolinae Basiliscinae Iguaninae The genus Anolis falls under the first of these three subfamilies, which includes six know^n genera, Anolis being the only one foimd in the United States. Cope (p. 233) describes Atiolis carolinensis and states that it " is distributed from the Rio Grande to Florida, inclusive, and as far north as Kinston, North Carolina. It is, moreover, com- mon in the Bahama Islands and Cuba, where it reaches a size rather superior to what is usual in the United States." Among other Iguan- idae found in the United States are the well-known forms Sceloporus willard: cranial nerves of anolis carolinensis. 23 and Phrynosoma. As to general external features and habits of life, the members of this family are most diverse, as will be seen by com- paring two such forms as Phrynosoma and Anolis. E. HISTORICAL STATEMENT. The Sauropsida have been quite generally neglected as regards the anatomy of the nervous system in the adult, most of the descriptive accounts having been written more than fifty years ago. This applies equally well to other reptiles and to birds, notwithstanding the fact that the latter have long since become of classic forms for embryological study. Aside from the embryological studies on birds and reptiles, such work as has been done has used exclusively the methods of gross anat- omy and must necessarily be incomplete. It might be added that those who ha^•e more recently contributed to our knowledge of the saurop- sidan nervous system and who use only the methods of the older anatomists increase our knowledge comparatively little. Much of the older work cannot be excelled within its limitations. The modern investigator should profit by modern methods and remove some of these limitations. The first important account of reptilian anatomy is by Bojanus ('19) in his monograph on the anatomy of the turtle. This is a classic, and is still the authority for much of the subject it covers. No other investigator has gone over the same field in so thorough a manner. Vogt ('39) gives us, in his dissection of PytJion tigris, our first in- formation on the cranial nerves of the Ophidia. This, however, is incomplete, nerves IV, VI, and IX not being found at all, and III only partially dissected out. Miiller ('40) about the same time discussed the nervous system of reptiles in his Neurologic der Myxinoiden. Bendz ('43) made a comparative study of the last four cranial nerves (or the vagus group) in reptiles, including among those studied two saurians, the alligator and the chamaeleon. Fischer's ('52) paper on the saurians is still the most important descriptive work on the cranial nerves of reptiles. He studied eleven species of lizards, two of crocodiles and the alli- gator. He treated the subject quite exhaustively and made use of comparative methods to establish certain homologies. Rabl- Riickhard ('78) gave in his description of the alligator's central nervous system, the first account of a reptilian brain. He inciden- tallv mentioned the roots of the cranial nerves. Hoffmann ('79-90), 24 bulletin: museum of comparative zoology. in the part of Bronn's Thier-Reich devoted to the Reptilia, based his description of the cranial nerves of lizards largely upon the work Fischer published forty years previously. More recently Osawa ('98) has given a quite full account of the anatomy of Hatteria, in which the cranial nerves are described. This work fails of much of its use- fulness because of lack of plates ; the text figures are too obscure to be of much use. Watkinson ( :06) describes J^aranus bii'ittatus, includ- ing the skull and musculature among her drawings. In matters of close comparison, however, her work has proved to be inadequately illustrated properly to supplement her description. In the same manner Cords ( :04) attempts to meet a long felt need in the anat- omy of the nervous system of birds; but here, again, the lack of a diagram of the complete system, for the purpose of comparison, is felt. In addition to the works mentioned, the cranial nervous system of reptiles and birds has been touched upon by various authors in the comparati\'e treatment of certain nerves or groups of nerves. In the following work on Anolis there is presented for the first time a description of the cranial nerves of an adult amniote form based upon a complete series of sections. F. SKELETAL STRUCTURES. 1. Skull. The skeletal parts involved in this account are those which are related topographically to the cranial nerve roots or serve for the attachment of muscles whose description follows, or have some other important, though indirect, relation to the main subject. INIany points of secondary importance in regard to length of rami, points of branching, and course of peripheral nerves are explained when refer- ence is made to the skull and its foramina. The skeletal parts appear- ing in the transverse sections are not readily understood without reference to the entire structure, therefore the three drawings of the skull are made with accuracy with a view to their permanent value in any problem involving the comparative anatomy of the reptilian head. Cope (:00) is followed in naming the parts of the skull and hyoid apparatus. As he separates the Reptilia altogether on osteo- logical characters, the basis for a large part of his definition of the Sauria may be recognized in this skull, c. g., "Quadrate bone articu- lating with exoccipital; parietal bones not closing the brain case in willard: cranial nerves of anolis carolinensis. 25 front, generally an epipterygoid present". . . . (p. 178). The labelled drawings of the skull (Plate 1, figs. 1-3), and the cross-section drawings (Plates 4-7) furnish all the description called for in this connection. Certain cartilaginous elements appearing in the sections were not preserved in the preparation of the skull. 2. Hyoid Apparatus. The hyoid apparatus is entirely free from any cranial attachments. "The hyoid system in lizards consists of a glossohyal, which is con- tinuous with a basihyal tract, a hj^pohyal, often continuous with the basihyal tract, a ceratohyal, a iirst ceratobranchial, and a second ceratobranchial, which is always continuous with the basihyal tract. There may be in addition an epibranchial, which belongs to the first ceratobranchial" (Cope, p. 189). Taking Anolis carolinensis as typical of the genus in this respect. Cope (p. 232) says, "the hyoid apparatus has the extreme development seen in all the lizards with a gular compressed pouch or fan. That is, the ceratobranchials of the second pair are closely appressed and produced to a great length. First pair of ceratobranchials and ceratohyals simple, the latter at- tached to the extremities of the moderately developed hypohyals." G. CUTANEOUS SENSE ORGANS. The cutaneous innervation and the epidermal sense organs deserve exhaustive study in the reptiles in view of the fact that these sense organs are apparently absent as such in mammals, although abundant in the Amphibia and fishes. On the other hand, mammals possess dermal tactile organs of problematic origin. Inasmuch as some new facts are presented in this general account of Anolis, a Jjrief statement of what has already been described within the group of reptiles is necessary. Maurer ('95, p. 228) refers to Reinhardt's ('61) article in which 191 species of snakes were examined for the epidermal sense organs. In 85 of these nothing of this nature was found; in 44 species there was one organ, and in 62 species two organs, to each scale. Maurer's ('95, p. 17, 196-239) own work covers the field sufficiently for our purpose. This can best be understood by reference to his text figures (9-14), which include most of his illustrations relating to the distribution of these sense organs. Besides Hatteria, only one 26 bulletin: museum of comparative zoology. lizard (Anguis fragilis) is shown. Maurer studied Lacerta, but does not mention the occurrence of the " Tastflecken." He found them, however, in Chamaeleo, but does not figure them. He calls attention to their large number, as many as six on one scale, in the primitive reptile Hatteria, and to their reduction to one to a scale in some of the snakes and in the crocodiles. His study of the minute structure of these organs results in his putting forth a theory that the epi- dermal organs are evolving into dermal organs, the cells of which come in all cases from the epidermis. The condition found in Coro- nella (Maurer, '95, Taf. viii, fig. 2, /.) shows one stage in the process; the crocodile (Taf. vii, fig. 12, t) a more advanced stage. Moreover the crocodile's single organ in the middle of the scale is really a multi- ple organ, as is indicated by the number of these tactile bodies, as many as six being found under one of the tactile spots. In Hatteria, which he regards as most primitive, the subepidermal tissue is involved in the organ (Taf. vii, fig 11, s.). In all cases he represents nerves going to these organs. The tactile organs of AnoUs. It is assumed that the organs in question are tactile in this animal for the reason that structurally there is more evidence for this view than has been put forth for any other related form. We find projecting from the center of each tactile spot a slender "tactile bristle" of considerable length. The distribution of these sense organs, provided with tactile bristles or " hairs," is very readily and accurately made out through the study of the moulted horny layer, of the skin. From a lot of individuals pro- cured in early spring a number showed a tendency to " shed" the skin. From an animal showing the beginning of this process it is possible to strip off artificially the whole corneous layer, and before it dries and curls it may be spread out on a slide and covered so as to remain in a perfectly flat condition. By mounting the whole cast, or at least half of it, piecemeal, on different slides, any part of the body surface can be readily referred to. While there is a variation in the number and size of these sensory organs there is no part of the body lacking them. Attention was first directed to a closer examination of the cast through the discovery of the central bristle in the sections. Under a low power (X 37), which was used for general drawings, these fine structures escaped notice, but a higher power never failed to bring them out. In the several camera drawings (Figures A-D) are shown some of the differences in arrangement and distribution that occur in several regions of the body. The sense organs are most abundant, although of inferior size, in the large pavement scales covering the willard: cranial nerves of anolis carolinensis. 27 Fig. a. — Pavement scales covering the right half of the anterior end of lower jaw; the position of sense buds is shown by the circles: each bud !_ possesses a tactile "hair," which is visible under higher magnification. Fig. B. — Pavement scales from frontal region of the head, showing that the sense organs are arranged near the margin of the scales. 28 bulletin: museum of comparative zoology. upper and lower jaws (Figure A). Where the scales show a tendency to be imbricated, as they are on most of the body region, the organs are situated on the thicker caudal Iwrder of each scale, the bristles projecting backward. From one to three were found on each scale Fig. C. — Imbricated scales from the ventral tail region, showing sense organs at the posterior margin of the scales. Fig. D. — Ventral surface of a portion of a fore-foot digit. The terminal part shows the "hairs "projecting from the distal edges of the scales; the subterminal part, where the tactile "hairs" are absent, shows the position of the adhesive organs. in the dorsal cervical region, while they are more abundant in the dorsal tail region (Figure C). The scales from the frontal region of the head (Figure B) show a distribution of sense organs intermediated willard: cranial nerves of anolis carolinensis. 2& between Figures A and C. An exception should possibly be made to the statement of their universal distribution, for they are apparently absent from that part of the digit which is provided wuth adhesive organs, "suckers," but they appear again on the small terminal part of the digit, where the bristle projects well beyond the imbricated margin of the scales (Figure D). The structure of this tactile organ is described here only in so far as it was shown in the material prepared for general study. The fixation in Vom Rath's fluid is of course excellent, but it precludes varied stain- ing methods, and certain differences that might thus have been brought out have not been observed. In section these organs are very incon- spicuous, and if they did not interrupt slightly the free and basal margins of the epidermal layer, they would escape notice in the un- stained osmic-fixed sections. In nearly all regions the corneous layer of the skin is slightly parted from the underlying epidermis and the abrupt thinning of this layer where it covers the tactile spot is more conspicuous than the organ itself. The modification in the horny layer, aside from the bristle, is effected more by the projection of the sense bud into this layer than by a depression of the circular area covering the bud. The cuticular bristle projects from the center of this circular area. It is slightly tapering, ending in a fine point. No structure could be recognized within this, nor any indication of how it is produced. The full length of these structures is not often pre- served in sections. Those found in sections measure up to 20 micra. in length. The surface view shows them somewhat longer; however, the text figures do not represent them with accuracy on this point. The structure of the organ itself agrees closely with that of Coronella (Maurer, '95), with this difference, however, that the column of cells forming the tactile end organ (/) could not be identified as a definite and separate structure. The continuity of the basal layer of the epidermal cells is interrupted by a ball of very small cells with rounded nuclei forming a bud which projects beyond the limits of the other cells and this approaches nearer to the free surface of the horny layer. Should the corneum be stripped off in Coronella (see Maurer), the tac- tile spots would appear in it as in Anolis. Despite the fact that all the other published figures show it, the innervation was not determined in Anolis beyond the fact that medullated fibers are often found directed through the corium toward the organ. Pieces of integument w^ere stripped off in the fresh condition, stretched out, and stained with osmic acid, but with no more definite results. The skin is seen to be richly supplied with medullated nerve fibers, but their distribution 30 bulletin: museum of comparative zoology. is not of such pattern as to point out these organs. It seems probable that the nerves innervating these organs lose their sheaths before reaching the epidermis. The exact innervation of these organs should not be difficult to determine by one of the finer histological methods, and oflPers an interesting problem. H. VISCERAL END ORGANS. The organs in which visceral nerves terminate, both efferent and afferent, include glands and smooth muscle fibers for the former and specific visceral sense organs and mucous membrane epithelium for the latter. As no methods were employed to determine the actual nerve terminations, a detailed description of these structures at this time would have no significance for the general descriptive anatomy of the cranial nerves which follows. In Anolis the mucous membrane of the mouth cavity, the pharynx, and the nasopharynx presents a variety of conditions in different regions which are demonstrable without the use of special staining methods. It is throughout richly glandular, the glands having the simple vesicular or tubular type along the gums and the tongue, while in the postlingual region the epithelium is ciliated and has a rich supply of unicellular glands of the goblet-cell type. The taste buds are confined almost entirely to the mouth cavity proper, although an occasional bud was found in the region of the larynx (Plate 4, Fig. 11, gm. gus.). They have the structure which is typical for these organs elsewhere, possessing a well-defined gustatory pit, in the base of which the sensory cells terminate. These taste buds are distributed along areas which stand out as sensory -glandular patches along the roof of the mouth and inner gums of both upper and lower jaws (Plates 4, 5, Figs. 9-12, gm. gus. m., gm. gus. I.). Their position is shown by the course of the sensory rami of nerve VII (palatine and chorda tympani). No taste buds were found on the tongue itself. Whether there are other sensory buds besides taste buds, was a question that suggested itself through the presence of clumps of cells which lacked the gustatory pit and were less sharply defined but were not like glands in their structure. The fact of importance in connection with visceral end organs is that typical gustatory buds are readily demonstrated, and their distribution fully determined in the series of sections from which the study was made. Upon this are based certain conclusions as to the nature of some of the branches of nerve VII. willard: cranial nerves of anolis carolinensis. 31 I. MUSCLES. 1. Muscles of the Orbit. This group includes the muscles of the eyeball proper and those of the eyelid, or their derivatives. The former are the dorsal, ventral, anterior, and posterior recti, and the dorsal and ventral oblique, to- gether with the m. retractor oculi and m. bursalis. Those of the eyeHd consist of the m. depressor palpebrae inferioris and its specialized part, m. protrusor oculi (Bruner). Mm. rectus posterior {externus), bursalis, and retractor oculi. Not only in their innervation but also in their skeletal connections, these, muscles form one group, having a more posterior origin than any of the other muscles of the eyeball. The bursalis (brs.) and the retractor oculi {rtr. oc.) arise close together within the basisphenoid bone (Plate 6, fig. 16), that of the bursalis being on the innfr lateral surface of the bony cap containing the distal end of the pituitary body. The area of origin of the fibers of the retractor forms a median forward continuation of that of the bursalis ; the two muscles then run rostrad together into the orbit, where they separate (Plate 5, figs. 14, 1.5). The bursalis fibers bend rather sharply dorsolaterad within the orbit to be inserted on the posterior median side of the eyeball, thus offering a sort of loop through which passes the tendon from the nic'-itating membrane (Plate 5, fig. 14, tnd. mb. nic). The retractor oculi passes directly forward across the floor of the orbit as a relatively slender muscle to be inserted on the median side of the eyeball anterior and ventral to the optic nerve (Plate 5, figs. 13, 14).^ Somewhat anterior to the origin of these the posterior rectus arises from the presphenoid bone (prcsph.) along the median line and passes directly out around the posterior side of the orbit to its insertion on the eyeball (Plate 3, fig. 7 and Plate 5, fig. 15, rt. p.). Mm. recti dorsalis and ventralis. These muscles arise at the same cross-section level as the rectus posterior, but dorsal to it, from a median fascia between the presphenoid and a cartilaginous rod which continues forward from the inner lamellae of the basisphenoid bone (Plate 5, fig. 15, rt. d., rt. v.). The one passes dorsal to the optic nerve, the other \entral, and each spreads out into a broad fan-like insertion on the dorsal and ventral sides of eyeball respectively (Plate 5, figs. 12, 13). 32 bulletin: museum of comparative zoology, .1/. rectus anterior {internus) has its origin around the cartilaginous rod mentioned in the preceding paragraph, a Httle anterior to the others and dorsal to the optic chiasma. This muscle is flattened out against the median connective-tissue septum which forms the common floor of the orbits, so that the muscles of the two sides are close together and parallel, inclosing between them the cartilaginous bar. It has a wide insertion on the anterior side of the eyeball (Plate 5, figs. 12, 13; Plate 4, fig. 11, rt. a.). *Mm. ohliquus dorsalis and obUquus voitralis. Both oblique muscles (ob. d., oh. V.) arise at about the same level along the cartilaginous bar which is a continuation of the interorbital septum, their fibers di^•erg- ing to their respective insertions on the eye ball (Plate 4, figs. 10, 11; Plate 5, figs. 12-14). The dorsal oblique takes origin from two dis- tinct bundles (Plate 4, fig. 11), the ventral thickened edge of the muscle beginning as a separate bundle, which has a tendinous origin an- terior to that of the inferior oblique (Fig. 10 and Fig. 7). As nerve IV reaches the dorsd oblique, crossing the orbit mesad to the dorsal rectus, it divides to innervate by one part this ventral portion, while the rest of the nerve passes across the muscle a little distance to innervate the more dorsal and anterior part. There is in the muscle, however, no apparent separation into two bundles at the level where this distinc- tion in innervation occurs (Plate 5, fig. 12, IV). Somewhat in contrast with the uniformity of the six principal e\-e muscles of vertebrates, is the variability in occurrence and structure of the accessory muscles which arise from them. Osawa ('98, p. 536) describes for Hatteria a retractor oculi muscle only, but from his description it is evident that he has found practically the same con- ditions as exist in Anolis, for he states that it is inserted "mit zwei Portionen an der Umgebung der Eintrittsstelle des N. opticus und an der laterale Flache des Bulbus etwa in der Gegend des Aequators." This would indicate the presence of a bursalis muscle included in his retractor oculi. He ('98, p. 537) describes the innervation of m. re- tractor oculi in Hatteria as follows: "Der M. retractor oculi weist in der Innervation eine Eigenthiimlichkeit auf insofern, als er am Ursprung einen Zweig aus dem N. abducens und in seinem weiteren Verlauf mehrere kleine Zweige vom Ganglion ciliare bekommt." The probability of such innervation in Anolis is discussed in connec- tion with the ciliary nerves (p. 46). M. depressor palpebrae inferioris. This is the muscle first described by Bojanus ('19) for the turtle as the m. palpebralis, and later by Fischer ('52) as the m. adductor maxillae, and finally by Weber ('77) willard: cranial nerves of anolis carolinensis. 33 as the m. depressor palpebrae inferioris. Bradley (:03), in analyzing the musculature with a ^•iew to explaining its relation to the mastica- tory movements, recognizes in several of the lizards a double function for this muscle, as suggested by the fact that some of its fibers are inserted on the lower eyelid, while others appear to have their inser- tion in the " fibrous membrane circumscribed by the pterygoid, pala- tine and the transverse bones " (p. 481) . This is also brought out more fully by Bruner ( :07),^ who, in looking in the head musculature for an apparatus to control the flooding and distension of the orbital blood sinuses, discovered that a part of this muscle was completely differ- entiated into a m. protrusor oculi, and records its occurrence in eleven lizards, including iVnolis. It is apparent, however, from a study of several series of sections that the two are not completely distinct morphologically in Anolis, as the following description will indicate. It seems probable, however, that the two functions as described by Bruner are here subserved. The following is the condition in Anolis : — the thin layer of striated muscle (Plate 3, fig. 7; Plates 4, 5, figs. 11-15, protru. oc, dep. palh. if.) which lines the floor of the orbit ventromesad to all other orbital structures has two distinct origins; one (protru. oc.) is from a slender tendon attached to the membranous wall of the cranium just anterior to the bony part on a level with the Gasserian ganglion (Plate 3, fig. 7; Plate 6, fig. 16). This tendon passes down mesad to the columella (epipterygoid) and is continuous with a muscle band which lies closely applied to the membranous region (Plate 5, figs. 14-15) referred to by Bradley. \Mule some of its fibers may be inserted here, most of them continue and spread out to form the anterior part of the broad palpebral muscle with insertion on the lower lid. This is best shown in the dissection (Plate 3, fig. 7). The second origin, ventral to that of the first, is from the fascia on the ventral face of the bursalis muscle; this band crosses ventral to the posterior rectus muscle (Plate 3, fig. 7; Plate 5, fig. 15, dcp. palp, if.) and turns sharply to form the posterior half, or more, of the palpebral muscle. The two bundles of origin include between them the jugular vein and suborbital sinus of the same side. The innervation of these muscles favors the view of a double function, and suggests for a part of it at least a more visceral function, such as that assigned to it by Bruner. This will be described in detail in connection with the ac- count of the nerves (p. 50). ■ Bruner does not mention the earlier recognition of this muscle by Bradley. 34 bulletin: museum of comparative zoology. 2. Muscles of the Tongue. The tongue musculature consists of a set of longitudinal muscles and an interlacing system of intrinsic fibers inserted on its mucous membrane perpendicularly to the surface. The longitudinal muscles consist of m. genioglossus {gen-gls.) and some others, to which specific names will not be given. The genioglossus muscle is a paired structure having its origin along a considerable extent of the cerato-branchial I. (Plates 5-7, figs. 13-22; Fig. E). These bands passing forward form rounded bundles extending the full length of the tongue and are more or less surrounded by the other tongue muscles. Ventrally, however, sphlcoll. , omo-hy.prf. .Tirnd. ker-md i 'dep.md. myl-hy. Fig. E. — Ventral aspect of head and neck after complete removal of the mm. mylo-hyoideus, sphincter colli, cerato-mandibularis 1 and 2, and omo-hyoideus (superflcialis). The m. omo-hyoideus profundus and the m. cerato-mandibularis 1 of the left side remain to show more clearly their position and e.xtent. The origin of m. cerato-mandibularis 2. indicated by dotted lines, covers that of a part of the longitudinal tongue muscle. they are bounded by large lymph sinuses, which leave the tongue attached by only its lateral muscular walls and thus permit its free movement and change of form (Plate 4, figs. 9-11). Two other longitudinal muscles may be recognized. These are not separable in their origin, but differ in their anterior attachment. They arise on the ventral face of the genioglossus (Fig. E) at about the level of the lar\Tix, and the bundles are increased by fibers springing apparently from the connective tissue underlying the mucous mem- brane lateral to this muscle (Plate 5, fig. 12, Ing. Ig.) Passing forward this divides into two liundles with different relations to the tongue. willard: cranial nerves of anolis carolinensis. 35 One (Plate 4, fig. 9, lug. Ig. 1) forms the lateral wall of the tongue at its base and continues forward in the same relative position finally to dwindle out in the mucous membrane anterior to the sublingual gland (Plate 4, fig. 8). For practically its whole distance it is closely ad- herent to the mucous membrane lateral to the tongue proper. The second part of this muscle {Ing. Ig. 2) separates from the ventro-median side of the common bundle and has nothing to do directly with the tongue or mucous membrane. It passes forward as a free bundle to be attached to the anterior end of the mandibular ramus (Fig. E). Its action is to pull the Ijase of the tongue forward. In addition to these longitudinal muscles there is an intrinsic musculature. A detailed analysis of this is not attempted, but it is somewhat as fol- lows : — At about the level of the larynx there appears in cross sections a mass of transverse fibers {Ing. t.) applied to the dorsal face of the m. genioglossus and a mass of vertical fibers (Ing. vrt.) on the median side of the same muscle (Plate 4, figs. 10, 11). Farther cephalad this intrinsic tongue musculature becomes what might be called a vertical decussating, system; these fibres, originating along the median side of m. genioglossus and around the glossohyal, cross the median plane just dorsal to the latter and radiate up into the long lingual papillae, which are well developed in the subterminal region of the tongue (Plate 4, fig. 9). The innervation of the tongue musculature is from the main bundle of nerve XII, although a part of this innervation occurs distal to the anastomosis of XII with lingual V, making an analysis by means of sections impossible. However, other facts, to be referred to later, support the assumption that lingual V provides a general cutaneous sensory component to the papillae and mucous surface in the anterior part of the mouth cavity, that the chorda tympani comes in with it to innervate the taste buds, and that XII innervates the tongue musculature. 3. Muscles of the Hyoid Apparatus. M. ccrato-mandibularis 1 (kcr-md.-^ = mylohyoides, Sanders). This is a broad, rather thin muscle (Figs. E, F, G) having its origin along the inner margin of the dentary bone and inserted along the whole extent of the cerato-branchials I. M. cerato-mandibularis 2 (kcr-md.^). This is a compact bundle (Fig. F) having its origin on the ventral face of the anterior end of the 36 bulletin: museum of comparative zoology mandibular ramus near the median line and its insertion on the basi- hyal.^f In section it appears as a median part of m. cerato-mandi- bularisj,!, except anteriorly, where it extends forward beyond the origin'of the latter. ■ kpr-md. i sin ky gen gh ker-brn. z omo-hy. prf. Fig. F. — Ventral aspect of the head and neck with integument removed to show the superficial muscles. The right half of the mylo-hyoideus muscle is also removed exposing the underlying muscles. sln/iy omo-hy. omo-hy prf. kiT-md . Fig. G. — Right lateral aspect of the liead with the mm. mylo-hyoideus and .sphincter colli removed; digastric muscle (rfep. md.) is cut at its insertion and tiu-ned up. M. cerato-mandibularis 3 (ker-md.^). This is a very small muscle band, to which the above name is applied because, like the other two, it extends from the mandible to the hyoid apparatus. This, however. willard: cranial nerves of anolis carolinensis. 37 is a perfectly distinct muscle, having its origin under cover of m. cerato-mandibularis 1 about half way along the jaw, and being in- serted near the distal end of the cerato-hyal instead of the cerato- branchial (Plates 5, 6, figs. 13-19). No corresponding muscle has been found in any of the descriptions of saurians. The innervation of the above three muscles is from nerve XII, which passes across the deeper surface of the first described portion (ker-mdJ); two branches are given oft' here, the anterior of which runs cephalad to supply also the anterior muscle (kcr-md.^). The small bimdle (kcr-md/) is inner\ated by an independent ramus, directly from XII. M. ceratohyoideus = hyoglossus (hy-gls.). This is a muscle (Plates 6, 6, figs. 13-19) extending between the first and second horns of the hyoid apparatus (i. e., between cerato-hyal and cerato-branchial I). The attachment on the latter is more extended than that on the cerato- hyal, so that the muscle sheet is much thinned posteriorly, consisting of fine bands somewhat separated from one another. The pharyngo- laryngeal branch of IX + X (phx-Iar.^) passes along the superficial face of this muscle between it and m. genioglossus. It is innervated by motor fibers carried in this branch (Plate 2, fig. 5), which might indicate either a visceral origin for the muscle or the presence of fibers from XII in the nerve. M. omohyoideus (omo-hy.). This (Figs. F-G; Plates 6, 7, figs. 16-24) has its origin on the shoulder girdle along the clavicle, ex- tending to near the median line, and passes forward as a broad sheet slightly converging with its mate to be inserted on the cerato-branchial I along almost its entire length and also on the basihyal. The fibers with the latter insertion form a bundle which, through a difference in direction, differentiates the superficial part (omo-hy.) from the deeper part {omo-hy. prf.), a distinction of slight importance in this form. M. sterno-hyoideus {sin-hy.). The removal of m. omohyoideus uncovers a muscle (Plate 5, figs. 13-15) which takes its origin from the sternum, the muscles of the two sides having a common median origin. The insertion is along the cerato-branchial I, co-extensive with that for the omohyoideus, but at a deeper level. At the angle of cerato- branchials I and II (Plate 5, fig. 13) the fibers of the median edge of sterno-hyoideus turn mesad to be inserted on the base of cerato- branchial II, on both the ventral and dorsal sides. The median edges of these muscles are the thickest and they spread out fan-shaped from median origin to insertion on the cerato-branchial I. A little of the trachea {tr.) is exposed just caudad to the l^asihyal, as is shown in Figure E. 38 bulletin: museum of comparative zoology. The innervation of the two preceding muscles is from the ventral ramus of the first spinal nerve, which passes between m. sterno- hyoideus and m. omohyoideus to supply both (Plate 7, fig. 22; Plates 2, 3). 4. Muscles of the Jaw and Neck. M. pferygo-mandibidaris (Bradley). This is the muscle (Plates 5, 6, figs. 14-19, pt-md.) generally described as the external ptery- goid. It is best seen from the ventral side with the roof of the pharynx removed (Fig. 7). Its origin is along the entire posterior edge of the main portion of the pterygoid bone, and chiefly by means of a strong tendinous fascia that is continuous with the muscle sheath on the pl-md. pt-5ph.]i.- plpar. prcplqd ' jitjphp ba'occ. Fig. H. Fig. I. Fig. H. — Right lateral aspect of a portion of the skull to show attachments of the deeper muscles, m. pterygo-parietalis and m. pterygo-sphenoidalis posterior. Fig. I. — Ventral aspect of a portion of the right half of the skull with man- dible removed to show m. pterygo-mandibularis. This really covers the m. -J pterygo-sphenoidalis posterior, but is slightly displaced to show the position of the latter. ventral side. The most posterior[fibers have a fleshy origin along the posterior wing of the pterygoid. The insertion of this muscle is on the angulare and articulare over the entire ventral plate as well as on the dorsal face of the median extension of these bones (Plate 6, fig. 19). Those muscle fibers that pass around the ventral side of the jaw to an extreme lateral insertion have their origin along the surface of the tendinous sheath instead of extending to the pterygoid bone. The innervation (Plate 3, fig. 6, pt-md.) is from a ventral twig of the same ramus of mandibular V which innervates the deeper portion of the m. pterygoideus. M. pterygo-sphenoidalis posterior {pt-sph. p., Figs. H, I, PlateT6, figs. 16, 17) of Bradley. The following is Bradley's (:03, p. 478) ■will.\rd: cranial xerves of anolis carolinensis. 39 description for this muscle in Varanus: — "This is a triangular muscle arising by its apex from that part of the basisphenoid which forms the lower boundary of the notch whose upper limit is established by the prootic bone and into which the Gasserian ganglion projects, .... Inser- tion to the full extent of the upper and lower surfaces of the pterygoid bone from its posterior extremity to as far foi-ward as a level, with the articulation between the basipterygoid process of the sphenoid and the pterygoid." With this Anolis agrees quite closely; in the latter the origin is also from the basipterygoid process of the sphenoid, and its insertion is along the whole length of the pterygoquadrate. The innervation is from a motor ramus separating itself from the other motor components at the Gasserian ganglion (Plate 3, fig. 6; Plate 6, figs. 16, 17). M. ptcrygo-paricfaJis of Bradley. (Fig. //; Plate 6, fig. 16, pt-par.) This muscle lies just posterior to the epipterygoid bone and its tendi- nous origin passes mesad of the latter to attach to the edge of the parie- tal, while its fibers take a direction that diverges somewhat from the epipterygoid and find insertion on the pterygoid just mesad to that of the deeper part of the pterygoideus, that is, on the upper surface of the pterygoid bone immediately caudad to the articulation of the epipterygoitl. Bradley homologized these two muscles, pterygo-parietalis and pterygo-sphenoidalis posterior, with muscles described by Katheriner ( :00) in the snakes under the same name, and calls attention to the fact that there are but two references to them found by him in the literature of the Saurians, viz., Stanius ('56) and Sanders ('70). Hoffmann ('79-90) does not refer to. them in his description of Reptilia given in Bronn's Thier-reich. Bradley, therefore, concludes that these muscles are peculiar to snakes and to those lizards (Kiokrania) which have a columella (epipterygoid). In Anolis m. pterygo-parietalis also has a special motor-nerve ramus (Plate 3, fig. 6, pt-par.) leaving the main motor bundle through the ganglion. M. capiti mandibularis (temporalis). (Plates 5, 6, figs. 14-19, cap. md.) The origin of this large muscle is from the postfrontal, jugal, postorbital, supratemporal, parietal, prootic, and quadrate bones. The superficial part of the muscle shows a parallel sheet of fibers running diagonally down to the lower jaw. The deeper por- tions, however, show toward their insertions a tendency to differen- tiate into several bundles. When the quadrato-jugal arch is removed, it is shown that the fibers having origin on the median face of the jugal 40 bulletin: museum of comparative zoology. and postfrontal converge to a tendinous insertion on the coronoid bone of the lower jaw. There is also a tendinous fascia at the dorsal end of the quadrate, which extends fonvard on to the parietal. In large part the fibers have a fleshy insertion along the complementare. The innervation (Plate 3, fig. 6, cap. md} and cap. md.^) is from several motor rami which leave ramus mandibularis V just distal to the Gasserian ganglion. M. pterygoideus (internal pterygoid). It is difficult clearly to delimit this muscle from the foregoing, as that part of its origin on the parietal is simply a continuation of the area of origin of m. capiti mandibularis; the part of it originating along the epipterygoid (Plate 6, fig. 16, pt.^), known as the "deeper portion," forms a mass of fibers more vertical in direction and innervated by a branch of the ramus which also innervates the m. pterygo-mandibularis. The portion described by Bradley as the "superficial" part (Plates 5, 6, figs'. 14-16, pt.) originates from the parietal and upper end of the columella, its fibers converging to a tendinous insertion on the coronoideum mesad to that of m. capiti mandibularis. With the pterygo-man- dil^ularis removed, the deeper part of the pterygoideus is well demonstrated from the ventral aspect, where it appears as a flat band extending from nearly the whole length of the epipterygoid to be inserted mesad to the superficial part. M. sphincter colli (spht. coll., Fig. F). This is not a well de\'eloped muscle in Anolis and is very easily torn off with the skin. Its fibers are so little massed that the innervation is very difficult to determine, except by means of the dissection of preparations blackened in osmic acid. The fibers have their origin superficial to the digastric, they cover this muscle to some extent and, forming an extremely thin band, pass as a continuous bundle to the opposite side. The distinctness of this muscle is emphasized in the figure. In the median ventral region its fibers pass dorsal to cerato-branchials II, where they are interrupted by fascia, and some of them dip under the fibers of the ventral longitudinal muscles so that it becomes impossible to dissect the sphincter colli free along the ventral median line. Its innervation (Fig. J) is from a very fine bundle of the ventral division of motor Yll. In Anolis this muscle does not take any part in bounding the external auditory meatus, nor does it form a "Schliessmuskel," as described by Versluys ('98) for some of the Geckoniden. M. mylo-hyoideus (intermandilmlaris). This muscle (Fig. F, myl-hy.) is also very thin and similar to the sphincter colli, although not so attenuated as the latter. It forms a sheet of muscle extending from willard: cranial nerves of anolis carolinensis. 41 one ramus of the mandible to the other, and reaching from the articu- Uire to the anterior fourth of the mandible, thus covering all of the intermandibular space except a small anterior area. From the figure it is seen that the posterior half of the sheet has its origin on the ventral side of the mandible, while the anterior half is inserted on the median face of that bone. ]Many of the fibers from the latter pass underneath those of the cerato-mandibularis, so that there is an interlacing of these two muscles along the inner border of the mandible. The innervation is from the anterior fibers of ramus hyoideus VII and from two or three rami of the mandibular V (myl-hy., Fig. ./.), which leave the*main trunk as mixed nerves (Plate 2, fig. 5; Plate 3, fig. (i, myl-hy. ).%^^ j^Versluys ('98, fig. 60-62) figures the m. mylo-hyoideus of Mahuia inulfifasciatajis dovetailing with the cerato-mandibularis, as in Anolis. i.md"— sphLcoU. myl-hy. Fig. J. — Camera drawing of the distribution of the motor rami of the facial nerve and the motor rami of the trigeminal nerve that are carried in the ramtis manditaularis. From a preparation made by removing sphincter, digastric, and mylo-hyoideiis muscles from a specimen fixed in vom Rath's soltition; the muscles are retained in their relative positions to one another and mounted for microscopic examination. The nerve fibers are all well blackened as far as the myelin extends. The preparation is viewed from its deeper surface. Three rami from the fifth nerve are shown in this figure, whereas only two appear in the plotting (figures 5 and 6 of Plates 2 and 3). Although not demonstrated, it is probable that the two more anterior rami arise from a common branch of the mandibular ramus, which corresponds to the second ramus of the mylo-hyoideus slwjwn in figs. 5 and 6, Plates 2 and 3. The muscle described as the posterior part of the mylo-hyoideus is designated by Versluys as the m. intermandibularis. This is the part innervated by a branch of VH. M. depressor mandibulae (digastric). This (dep. rnd.) is a well- developed, but not a powerful muscle in Anolis. It is broad and fan 42 bulletin: museum of comparative zoology. shaped, having its origin from the parietal ridge and neck fascia along an irregular line which anteriorly lies near the median plane, but posteriorly assumes a more lateral position (Fig. G). All the fibers converge to the articulare of the lower jaw, where most of them are inserted at the extreme posterior point by means of a ligamentous attachment. A thin band {dep. vidS), however, passes around to the ventral side and a little farther cephalad finds attachment in the fascia which also serves for the posterior fibers of m. mylohyoides (Figs. F and J). The anterior border of this muscle (Plate 3, fig. 7, dep. md.) forms the posterior border of the external auditory meatus and is the thickest and strongest part of the muscle. This portion is roughly separable from the sheet-like part behind it and has a firmer origin in the skull, the thinner portions behind arising from the neck. The tympanic chamber extends backward underneath this muscle (Plates 6, 7, figs. 19-24). The innervation of this muscle is wholly from motor VII, which supplies it by two main branches (Fig. J). Versluys ('98, p. 285) refers to the thinner posterior portion as occurring in many Lacertilia. It corresponds to the "superficial" portion and Ca md of Ruge ('97, p. 326-331) for Varanus. The thicker part is the " Hauptportion " of Versluys. M. episterno-cleido-mastoideus (Versluys) capiti-cleido-episternalis, (Watkinson). This muscle (c'sfn-clei-mast.), which extends from the skull, having its origin underneath the main part of the m. depressor mandibulae, passes caudo-ventrad to be inserted on the sternum (Fig. G). It forms a thickened ventral border to the cucullaris muscle, next to be described, and has the same innervation (Plate 7, fig. 24). M. capiti-dorso-clavicidaris (trapezius, or cucullaris). The fibers of this muscle (Fig. G, cap. d'clav.), have the same direction as the preceding, but both origin and insertion are different. It is an extremely thin sheet of fibers having origin underneath the thin posterior part of the m. depressor mandibulae and being inserted on the pectoral girdle. The line of origin extends well back past the level of insertion, so that the most posterior fibers are directed forward. It does not form a continuous sheet, but between the successive bundles are spaces which leave exposed the underlying lymph sac (Plate 7, fig. 24, sac. cn'lym.). These spaces are not indicated in figure G. The innervation of the two foregoing muscles is from two or more spinal nerves, (Plates 2 and 3, figs. 4, 6, spi. v. 3) the first muscle being supplied wholly from the motor part of the third spinal nerve, which comes out to the muscle as a mixed lateral ramus. The sensory part willard: cranial nerves of anolis carolinensis. 43 passes through this muscle and is distributed to the ventral skin region. Three branches of spinal nerves are shown to innervate the cucullaris muscle; their precise relation to particular spinal nerves was not determined owing to apparent anastomoses of rami. No relation that would suggest an innervation from a spinal accessory nerve was established with any nerves anterior to spinal II. M. laxator tyvipani of Versluys {lax. tym.). A very small muscle, less than one half mm. in length, extends caudad from the insertion of the tympanic ligament on the extra-columella (Plate 7, figs. 21, 22). Its fibers end on the connective-tissue covering of the parotic process. The motor component of nerve VII passes ventral to the posterior end of this small bundle and in contact with it. The series of sections did not show with certainty the innervation of this muscle from motor VII, but a few fibers are given off from the main nerve bundle which in all probability accomplish such innervation. This is the more probable because there is no other nerve in the vicinity. Owang to its minute size this muscle was not isolated in dissection, but is a constant feature of the sections. This muscle was first described by Versluys ('98) and given the name descriptive of its function. He found it in a number of lizards, all members of the family Gecko- nidae, while he failed to discover ^pJ' .m^phxhr. it in as many others, which be- longed to diiferent families. No more positi\-e statement of its in- nervation than is here given for Anolis is contained in Versluys's account. M. constrictor venae jugularis internac of Bruner (co'st. vn. j. i.). In Anolis this muscle is found in relations similar to those described by Bruner ( : 07, p. 42) for Phr\Tio- soma. These striate muscle fibers surround the internal jugular vein for a distance of .84 mm. The most anterior fibers, for a distance of .14 mm., have their origin from the most posterior portion of the parotic process (Plate 7, fig. 23, jjrc. pa'ot.) and from the ligamentous extension of it. These anterior r/.M 3, Z.I. Fig. K. — Frontal projection, dorsal as- pect, of nerves IX and X of the left side, showing the region of the roots and ganglia 'only, together with their connections. Plotted from Anolis, transverse series 48a. Two tumefac- tor nerves (rn{. ini. j.) are given oflf from the ramus pharyngo-larjngeus and one from the ramus superior laryngeus. 44 bulletin: museum of comparative zoology. fibers surround the vessel in a circular manner. For the remainder of the distance the fibers are diagonal or longitudinal to the wall of the vessel. Here the latter lies in immediate contact with the thymus gland. The muscle is innervated by some very fine rami (not included in plotting) from X and IX + X (Fig. K, rm. vn. j.), as was determined by study of the sections. Some undetermined rami appearing in the dissection (Plate 3, fig. 7) are probably of like function. J. OCULOMOTOR NERVE. The oculomotor nerve (///) arises from its nucleus of origin in the floor of the aqueduct of Sylvius and emerges as one large root from the ventral side of the mesencephalon (Plate 2, fig. 4). It swings laterad and cephalad to pass out of the cranium through the membranous wall, being covered laterally at this point by the bursalis muscle. (Plate 6, fig. 16). Posterior to the origin of the recti muscles the nerve divides into three large rami; the dorsal one goes to the dorsal rectus; a large ventral one, which later divides (Plate 5, fig. 14, ///), supplies the ventral rectus, the anterior rectus and the ventral oblique; be- tween these two large rami is the short root of the ciliary nerve. The latter is composed chiefly of fine neuraxons, which may be recognized as a distinct bundle in the center of the main trunk from the point where it emerges from the brain to its separation from the somatic motor components as the ciliary root. Within the brain itself it could not be independently traced in this series. The branch of nerve III to the dorsal rectus muscle (Plate 2, fig. 4, rt. d.) is a large one. It follows the lateral face of this muscle near its anterior ventral edge (Plate 5, fig. 14), several successive bundles of fibers being given ofi^ to accomplish the innervation of the muscle. The large ventral ramus (Plate 5, fig. 15, ///) passes mesad of the retractor oculi to attain a position on the ventral face of the ventral rectus muscle. It is in this part of its course that a group of long rami separate from it, and penetrate the ventral rectus from both the dorsal and ventral surfaces (Plate 2, fig. 4). The remainder of the nerve (Plate 5, fig. 14, ///) passes cephalad close to the median line, dividing to send more than half of its fibers dorsally into the anterior rectus (Plate 5, fig. 13, rt. a.), the remainder passing on to the ventral oblique (Fig. 13, oh. v.). willard: cranial nerves of anolis carolinensis. 45 K. TROCHLEAR NERVE. Xerve IV (Plate 2, fig. 4, /I'), from its nucleus of origin on the side of the aqueduct floor opposite to that of its emergence, passes dorsad of the mesocoele directly to the surface, the whole central course showing in a single transverse section. Peripherally it turns cephalad for a short intracranial course, lying between the brain and the Gas- serian ganglion. It then passes through the membranous cranivim, but keeps a median position throughout its course, to the dorsal oblique muscle, which receives all of its fibers (Plates 5, 6, figs. 12-17). L. ABDUCENT NERVE. The central origin of nerve VI (Plate 2, fig. 4, F/) in the meten- cephalon was readily found. The neuraxons leave this part of the brain in small groups as large heavily medullated fibers. These emerge from the ventral side of the brain in small rootlets, which immediately combine into one main trunk. After a short intra- cranial course nerve VI enters the sphenoid bone and takes a course cephalad through a special foramen between the outer and inner lamellae of this bone, and dorsal 'to the beginnings of the basi- pterygoid process of the sphenoid (Plate 6, fig. 17). This foramen opens mesally into the bony pocket protecting the ventral end of the pituitary body and at the point of origin of the bursalis and retractor oculi muscles. Nerve VI passes through these muscles to reach the posterior rectus, which it innervates (Plate 5, 6, figs. 15, 16). The bursalis and retractor oculi muscles are innervated by a small ramus of not more than 25 fibers which is given off from VI as it passes between them. This is shown in Plate 6, fig. 16 (brs.), but in the plotting (Plate 2, fig. 4) it is covered by the main trunk. It is difficult to explain the very pronoimced disproportion between this small ramus and the remainder of the abducent nerve. No connection was found which would relate the abducent nerve to any structures other than the three muscles named. Ciliary Ganglion and Nerves. In their macroscopic features the ciliary nerves and ganglion and their relation to the trigeminal and oculo motor nerves are compara- 46 bulletin: museum of comparative zoology. lively simple. The ganglion (Plate 5, fig. 15, gn. cil.) lies between the membranous cranium on one side and the bursalis and retractor oculi muscles on the other, and is separated from both III and V by roots of considerable size. It is connected with nerve III, as is generally the case in other forms, by a shorter, thicker root, radix brevis (Plate 2, fig. 4, rx. cil. Ill), while the root from V {rx. cil. V) is longer and more slender. Only the proximal end of this root is shown in the figure. Both roots communicate with the ganglion directly and in like manner the two ciliary ner^'es arise directly from its distal end. In the plotting (Plate 2, fig. 4) this point is hidden by nerve III. These ciliary nerves (Plate 5, fig^. 13_, 14, cil.) cross dorsad of the optic nerve to enter the sclerotic coat of the eyeball. They keep close together in their course as they pass laterad and cephalad around the eyeball to their terminations in the striated muscles of the ciliary body and epithelial surfaces. From the smaller of these ciliary nerves is given off a very fine branch, which follows the others for a distance but is lost before its entrance into the eyeball. In sections individual medullated fillers are seen to be given off from this small ramus. These are lost along the walls of the small blood vessels. The pres- ence of these fine rami in immediate contact with the retractor oculi muscle at its end of insertion could readily give rise to a misinter- pretation in regard to their distribution (see p. 32). It is possible that Osawa ('98, p. 537), basing his conclusions on dissections alone, made such an error in his account of Hatteria. In Anolis every recognized fiber leaving the ciliary nerves could be traced cephalad beyond the most anterior extent of this muscle {rtr. oc), and in no case w^ere these fibers distributed to the muscle. In many points the microscopic evidence was far from conclusive for determining the relations of the ciliary roots and nerves to the ganglion, but they offer certain facts worth recording. The short root is a large one and is principally composed of fibers of very light medullation, but not of the smallest caliber, i. c, they are larger than those of the visceral sensory system as shown in palatine VII. Among these are a very few coarser fibers of a medullation sufficiently heavier to make them conspicuous. These are as large as many that remain in the somatic motor rami, but do not equal in size th6^e which pre- dominate in these motor nerves, nor are they segregated into a single group to be readily followed through the ganglion. Howe^'er, the fact that every section through the ganglion shows about the same number of these coarser fibers points to the possibility of their unin- terrupted passage. The ganglion is a uniformly oval structure_, the cells of which entirelv surround the short root. willard: cranial nerves of anolis carolinensis. 47 The long root {rx. cil. V.), from the nasalis division of the opthahnic, is composed of both medullated and non-medullated fibers and joins the ganghon on its dorsal side (Plate 5, fig. 15) about midway of its length, many of its fibers immediately mingling with the ganglion cells. The independence of the bundle is, however, preserved through- out by the continuity of the non-medullated components of this root, which passes across the dorsal side of the anterior half of the ganglion and on into the large ciliary nerve, in which it appears for a con- siderable distance as a lighter area in the cross section. These facts do not preclude the possibility that many of the non-medullated fibers end in the ganglion, nor that a part of the postganglionic bundle of fine fibers takes origin in the ganglion. Further analysis of the ciliary nerves discloses some of the coarse fibers of the short root in each. The rest are of the smaller medullated variety and appear to take their rise as peripheral neuraxons of the cells of the ciliary ganglion, since they do not occur in the ciliary roots. In comparing Anolis with other Sauropsida we find that the arrange- ment of roots, ganglion and ciliary nerves is that which Fischer ('52, p. 117) describes as typical for reptiles. As an exception he mentions Salvator merianae, where the trigeminal and oculomotor roots join proximal to the ganglion. Other accounts agree with Anolis. Osawa ('98, p. 602), in describing Hatteria, establishes another exception, wherein the ganglion itself is not connected with the trigeminal nerve by an independent root, but, if at all, through recurrent fibers, as in the fowl, the ciliary rami from V joining the ciliary nerve distal to the ganglion. Carpenter's (:06, p. 158) careful analysis of the ciliary ganglion and its connections in the adult fowl is the only basis we have for a comparison of histological features with conditions in birds. In the fowl, the short root from the oculomotor being much the same as for Anolis, one main ciliary nerve leaves this ganglion. This contains all the well medullated fibers. Another small bundle, of feebly medul- lated fibers, leaves the ganglion dorsal to the large one. A third (small) ramus accompanies the other two; microscopic study, how- ever, showed Carpenter that it contains no fibers from the ganglion, but is merely a communicating ramus from the trigeminal, which meets the ciliary nerve distal to the ganglion. The same ramus sends some recurrent fibers back to the ganglion. Other fine rami may be given off from the communicating ramus. All the trigeminal elements are non-medullated. The ciliary ganglion itself is divisible into a sympathetic and a cerebro-spinal part. 48 bulletin: museum of comparative zoology. As compared with this, we find in Anolis the following differences : — (a) nerve V connects directly with the ciliary ganglion by a branch which contains a large proportion of medullated fibers; (b) the non- medullated components from the trigeminal are carried in the large ciliary nerve and form no small non-medullated rami ; (c) the ganglion is not divisible into two parts and has more the appearance of a cerebro- spinal ganglion. It should be noted, however, in regard to this last point (c), that the complete Vom Rath method was not used in the preparation of the Anolis material, the treatment with pyroligneous acid being omitted. This omission may have resulted in less differen- tiation within the ganglion and a consequent failure to discover the differences recorded by Carpenter. M. TRIGEMINAL NERVE. The afferent and eft'erent neurons that go to make up the mixed root of nerve V form one large bundle, in the cross section of which there is shown no segregation of components at the point of emergence from the brain. More centrally, however, the main sources of these components may readily be found. More complete study of the brain will no doubt result in some addition to this account of peri- pheral structure and possibly to some revision. The central relations of the neurons are here referred to briefly and only in so far as they were brought out in the series of sections from which the plottings were made. The efferent neurons. The motor components were seen, in sec- tions posterior to the superficial origin of the root, to arise from two sources : (a) a group of cells lying a short distance mesad and slightly dorsad to the connection of the root with the brain, and (b) from a region just laterad to the median longitudinal fasciculus. From both these sources the bundles of heavily medullated fibers pass directly out to the ventro-median part of the root. Here they form with the. afferent components a common root-bundle, in which it is difficult to distinguish the two. At the point of superficial origin the mixed root-bundle is covered dorsally by the ganglion of nerve VIII (Plate 6, fig. 18, rx. V; gn. VIII). The root passes cephalad between the prootic bone and the skull to the Gasserian ganglion (Plate 6, fig. 17, gn. V), which lies in the foramen prooticum (Gaupp). The foramen is represented in the skull of Anolis by a notch in the anterior })order willard: cranial nerves of anolis carolinensis. 49 of the prootic bone, being bounded anteriorly by the membranous part of the cranial wall (Plate 1, fig. 2; for. V). Before the ganglion is reached, the root shows a distinct separation into three parts, which in cross section appear as three unequal seg- ments of a circle. The largest is the dorsal segment, which passes into the main part {gn. V) of the ganglion (semilunar of Fischer); the ventral comprises the motor neurons, while the median division from this point on is free from all other connections and passes into the ophthalmic ganglion (Plate 2, fig. 4; Plate 3, figs. 6, 7, gn. V; gn. opth.). The motor components, thus segregated just proximal to the gan- glion, are almost exclusively distributed in small pure rami directly from the ganglion (Plate 3, fig. 6). A few motor fibers are, however, included in mandibular V and reappear in several small mixed rami innervating a part of m. mylo-hyoideus (Plates 2 and 3, figs. 5 and 6). Those rami which supply the dorsal and lateral jaw musculature arise from the dorsal division of the main motor bundle. This passes across the ventral side of the ganglion and the base of mandibular V to be split into two rami, one of which (cap. md. 1 and 2) supplies chiefly the m. capiti mandibularis, and the other (pt.) the greater part of the m. pter,ygoideus. Of the former a part passes directly dorsad, while the rest follows the course of maxillaris V as several slender rami to supply the anterior part of these muscles (Plates 2 and 3, figs. 4 and 6). The motor components which do not enter into this dorsal and lateral distribution pass directly cephalad on the ventral face of the ganglion. A part follows for a short distance mandibular V on its ventral side and then divides, one branch (pt-m.d.) going to innervate the m. pterygo-mandibularis, the other (pt.) to innervate the deeper part of the m. pterygoideus. The rest separates into three rami, two of which are very small. Of these one (pt-par.) innervates m. pterygo- parietalis, the other (pt-sph. p.) m. pterygo-sphenoidalis posterior. The third ramus (protru. oc.) is larger than either of these two. It passes cephalad to the orbit to innervate m. protrusor oculi and m. depressor palpebrae inferioris. Because of its special functional and structural relations, this nerve demands a more detailed account. Ramus ad m. depressor palpebrae inferioris (dep. palh. if.). This ramus is recognizable on the ventral side of the main motor bundle opposite the proximal end of the ganglion (Plate 6, fig. 17). The bundle as a whole has a characteristic appearance, the fibers being well medullated but somewhat finer than those of the other motor 50 bulletin: museum of comparative zoology. nerves. Upon closer examination it is found to contain fibers of still finer caliber and of lighter meduUation and also possibly some non- meduUated fibers. The latter point is difficult to establish where such fibers are relatively few and not grouped into a bundle. The course of this nerve is directly cephalad to the orbit keeping to the ventro-lateral side of all the eye muscles (Plates 3, 5, 6). Soon after this nerve has begun its peripheral course its lightly medullated fibers become grouped into a bundle on its ventral side. These leave the main ramus about midway between the Gasserian ganglion and the orbit, and form a communicating ramus (comn.), which joins palatine VII a short distance cephalad of this point (Plate 3, figs. 6, 7; Plate 5, figs. 14, 15). In the palatine nerve its identity is wholly lost, although the nature of its union strongly indicates that it does not form a recurrent bundle, but continues its course cephalad. Two very small but noteworthy twigs (Plate 3, figs. 6, 7, protru. oc.) are given off from this communicating ramus to innervate the muscle which has been described as the protrusor oculi. In the dissection (Plate 3, fig. 7), where this was clearly worked out, it will be noticed that one twig is given oft' from the rm. palpebralis inferior itself and only one from the communicating ramus. The fibers innervating this muscle are of the same character as those of the communicating ramus and in distinct contrast to those remaining in the main motor ramus, which, farther cephalad, innervate the depressor muscle of the lower eyelid. A comparison with the opposite side of the head and with other series of sections shows practically the same relation, al- though on the opposite side in the same series a twig is given off proximal to the communicating rami, as in Plate 3, fig. 7. It comes, however, from the ventral lightly medullated bundle, which is as clearly marked off as though it were a separate ramus. Although Fischer ('52) describes the innervation of the m. depressor palpebrae inferioris by a ramus coming directly from the motor root, As in Anolis, he makes no mention of a communicating ramus between this nerve and palatine VII. The one mention of it met with is by Watkinson (:06, p. 457, 463) in Varanus, where it is described as a communicating ramus between palatine VII and the Gasserian ganglion by way of this motor nerve. From dissections alone it would appear to be mere assumption that it takes this course. From the sections of Anolis it seems quite clear that the components of this connecting ramus have a distal existence in the palatine. Such an anastomosis between a pure motor ramus and the viscero-sensory is not met with in Anolis in any other connection. No reference is made willard: cranial nerves of anolis carolinensis. 51 by any writer to a muscle innervation similar to the one above de- scribed. This peculiarity, coupled with the described function (p. 33) of this muscle, offers a problem for comparative study. It is to be noted that no ganglion cells were found at the point of union either on V or VII. Afferent neurons. The central course of these fibers is so involved in other brain tracts that the present description will begin at the superficial origin of the root. The descending root of the trigeminal nerve traced cephalad approaches the surface of the medulla and produces a swelling of the surface beneath the roots of nerves VII and VIII (Plate 6, fig. 18, rx. V). The motor neurons, as before described, join this bundle rather abruptly from a more median position, and the two together leave the brain as a single root. The fibers making up the sensory components of the trigeminal lack uniformity of size and medullation. They are, however, of a size approximately that of the motor components of the trigeminal and facial, but also have among them both larger and smaller fibers. A characteristic feature of the cross sections is the presence of a few very large, strongly medullated fibers. These are recognizable in the brain, and a central connection may possibly be established for them. Peripherally they seem to offer no special relation to end organs. They occur in limited num- bers and are always scattered as isolated fibers. It is possible they are motor in their character and are transferred to the motor rami. The ophthalmic, maxillary, and mandibular divisions of the trigemi- nal are recognizable even proximal to the ganglion {gn. V; gn. opth.), the two ganglia being entirely distinct and scarcely in contact in series 30. The relative sizes and positions of the ganglia are best seen from the figures on Plates 2, 3. The cross section (Plate 6, fig. 17) shows the segregation of the two cell groups. In addition to these two ganglia, one of the figures (Fig. 6) shows a third ; a very small ven- tral group of cells entirely similar in the sections to those of the larger ganglia. Careful study was made of this small ganglion to discover any structural features that would indicate a sympathetic character, but comparison showed no such group on the opposite side of the same individual ; in this case therefore it is probably a separation (in- constant) of a few cells of the main ganglion. From what we know of other forms, it seems probable that the sympathetic rami of the lachrymal plexus send fibers to the Gasserian ganglion, but that could not be demonstrated in Anolis, and there is no deep sympathetic trunk connecting the Gasserian with more posterior ganglia. A detailed account of the peripheral distribution of the sensory 52 bulletin: museum of comparative zoology. components of the trigeminal nerve will be preceded by a more general statement. The main branches of this nerve correspond quite closely with the typical condition found in lower vertebrates. The ophthalmi- cus profundus, proceeding from its owti independent ganglion, is distributed to the skin of the dorsal surface of the head over an area extending from a post orbital region forward to the tip of the snout, the frontal region receiving its fibers over a special ramus, which leaves the ophthalmic at the ganglion; the rest of the nerve, as the nasalis branch, takes the course through the orbit typical of the profundus, receiving the long root of the ciliary nerve on its way, and supplies the skin of the dorsal surface of the snout and the epithelium of ante- rior nasal chamber. The second and third (maxillary and mandibular) branches of the trigeminus join the main trigeminal ganglion as large branches from the upper and lower jaw regions. The maxillary supplies the skin around the eye and the upper and lower lids ; it crosses the floor of the orbit to the upper jaw, in which it runs forward to the tip of the snout, supplying the skin lateral to the jaw and the epithelium of the gums and the teeth. The mandibular branch passes down to the lower jaw, which it enters, supplying the skin, gums and teeth in a manner similar to that 'described for the upper jaw. It also supplies the anterior regions of mouth and tongue. In this latter distribution it is closely associated with the somatic motor components of XII and the viscero-sensory components of the chorda tympani. Certain plexuses and anasto- moses omitted from this general account will be included in the de- tailed description which follows. I. Nervus ophthalmicus yrofundus. This term, as including all the fibers entering the ophthalmic ganglion, has no application in Anolis, as the two main branches (r. nasalis and r. frontalis) do not unite distal to the ganglion. These two branches will be described as the frontal and nasal rami. la. Ramus frontalis (/.). — The point where in different reptiles the frontal ramus joins the nasalis is determined somewhat by the relation of the ganglion to the orbit. In Anolis (Plates 2, 3) the orbit is large and the frontal ramus takes rather an abrupt dorsal course to reach the skin posterior to it; the ramus therefore joins the ganglion directly. The frontalis is about half the size of the nasalis. Both are composed of the same kind of fibers, except that the nasalis carries the non-meduUated fibers of the radix longa of the ciliary nerve, which collectively can be recognized from the beginning of the willard: cranial nerves of anolis carolinexsis. 53 ramus. The frontal ramus lies dorsal to the other and both run cephalad for some distance, passing between the membranous crani- um and the ligamentous origin of the protrusor oculi muscle (Plate 6, fig. 16, /. and na.) . A little farther forward the frontalis turns abruptly dorsad to the posterior angle of the orbit (Plate 5, fig. 15). In this course it keeps next the membranous cranium and attains a position just caudad to the lachr^^nal gland. As it turns dorsally it gives off a branch that becomes involved in the sympathetic plexus (Plate 3, figs. 6, 7), but, like the other branches, this contains cutaneous sen- sory elements that can be traced to the skin. The main branch con- tinues between the brain and the large blood sinus which lies just posterior to the lachrymal gland (Plate 5, fig. 15). A little farther forward, and at a level where the gland is beginning to be cut, there are given off a number of branches, which anastomose with the sympathetic, but for the most part supply the upper lid. These lie just mesad of the lachr;yaual gland (Plate 5, fig. 14, gl. Ich.). The main part continues forward and supplies, from time to time, small rami to the skin between the eyes (Plate 5, figs. 12, 13). '^ lb. Ramus nasalis iiia.) . This is the main ophthalmic branch and takes the characteristic course through the orbit (Plates 4, 5, 6, figs. 10-17). It passes into the orbit with nerve III and passes dorsally to the optic nerve and over the ocular face of the dorsal rectus mus- cle (Plate 5, fig. 14). Just before its entrance into the orbit (Plates 2, 3, figs. 4, 6, rx. cil. V.) it gives off the long root of the cihary nerve (see p. 46). With this exception there are no branches or connec- tions within the orbit. It passes out of the orbit on the median side of the Harderian gland (Plate 3, fig. 7; Plate 4, fig. 11) into the space bounded by the anterior median w4ng of the palatine, the frontal, the prefrontal and the origin of the ventral oblique muscle (Plate 4, fig. 10). Here the nerve passes through the "ethmoidal" sympathe- tic ganglion (gn. eth.), wliich will receive especial description. This ganglion lies on a level with the anterior boundary of the bony orbit, and through it a cutaneous branch of coarse fibers is given off to the skin of the anterior angle of the eyelids (Plate 4, fig. 10, na}). Soon after this the main nerve divides into a lateral branch (Plate 4, fig. 9, na. I.) and a median branch {na. m.). These richly supply the skin of the snout. The median one gives oft' in addition a branch to the mucous membrane of the nose; its cutaneous branch passes forward along the median line to be distributed to the end of the snout (Plate 2, fig. 4). According to Watkinson (:06, p. 458), who mentions both the fron- V 54 bulletin: museum of comparative zoology. talis and the nasalis, Varanus agrees with Anohs in so far as no branches are given off from the nasalis between the ciliaris and the posterior extremity of the internal nares, where;, as she states, the " r. comm. cum ramo platinus VII" is "composed of at least two distinct fiber bundles," which communicate with palatine VII; then follow the same divisions as noted for Anolis anterior to this, i. e., "r. latera- lis" and "r. medialis," these having much the same distribution as in Anolis. To the medialis are assigned the following branches, r. pre- maxillaris superior (dorsal) and r. premaxillaris inferior, these inner- vating the skin of the nose and the lips. Ethmoidal ganglion {gn. eth.). This is a ganglion of oval form and one fourth mm. in length, which lies closely applied to the ventral and lateral sides of the nasahs nerve (Plates 2, 3; Plate 4, fig. 10). The ganglionic cells do not entirely surround this nerve, and there is no passage of medullated fibers from nerve to ganglion. The gan- glion is in connection with the median branch of the palatine nerve through communicating bundles of fibers ; these fibers are also accom- panied by sympathetic ganglion cells, which form a sort of column of cells extending from the side of the ganglion opposite the nasalis nerve to the palatine ramus (Plate 4, fig. 10). A dissection of this ganglion and its connections, mounted in balsam, shows some fine, lightly medullated fibers from the stalk passing both caudad and cephalad in the palatine. The almost complete lack of medullated fibers among the cells indicates that the ganglion has to do largely with non-meduUated fibers. A very small bundle of such fibers joins the ganglion on its posterior side close to the nasalis nerve. To what extent this ganglion may be in connection with non-medullated fibers of the nasalis, could not be determined, but such a relation would appear to be the only explanation of the anatomical relations observed. The cells of the ganglion are smaller and less clearly defined than those of the cerebral ganglia, showing in this their sympathetic char- acter (Carpenter, :06). The ethmoidal ganglion occurs regularly in birds, as far as they ha^'e been studied, as a group of ganglion cells on the ophthalmic branch of V. Bonsdorff ('52) gave it the name of "ganglion ethmoidale," and Rochas ('85) "g. orbitonasale." In the goose Cords (:04, p. 59) de- scribes this ganglion as being 1 mm. long and ^ to | mm. broad, and as having the same connections as we find in Anolis, i.e., with the ophthal- mic branch of V and the palatine branch of VII. There can be no doubt of the homology of this ganglion as described for birds with the structure to which the name has been given in zVnolis. willard: cranial nerves of anolis carolinensis. 55 An anastomosis of palatine VII with the ophthahnic branch of V in the anterior orbital region seems to be of wide occurrence in the Saurop- sida and Amphibia. It is described in Amblystoma (Herrick, '94), tadpole of the frog (Strong, '95 [farther forward here]), Spelerpes (Bowers, :00), Amblystoma (Coghill, :02) and Amphiuma (Norris, :08). Of the above mentioned investigators, Coghill finds a ganglion at the point of union, and Norris (p. 540) says "there seems to be a ganglion on the palatine nerve shortly before the anastomosis"; be- yond the anastomosis he finds mixed cutaneous and viscero-sensory rami. With this observation by Norris agrees the statement I have made regarding Anolis ; that this ganglion, while surrounding ophthal- mic V, is really more closely related to palatine VII. Among the reptiles an anastomosis is mentioned by Gaupp ('88), Osawa ('98), and Watkinson (:06). On the other hand Fischer ('52) does not mention it at all. The presence also of a ganglion, at least one discernible by dissection, seems to be less general in the described reptiles than the condition in Anolis would suggest. The only men- tion of it which I have found was in Gaupp's ('88, p. 469) description of the snake Tropidonotus, in which he refers to it as "eine leichte Anschwellung," and further "Vielleicht entspricht dieselbe dem ganglion ethmoidale der Vogel." Watkinson (:06, p. 463) speaks of palatine VII and nasalis V as lying ciuite close together at a point corresponding to the ethmoidal ganglion and having one or two con- necting branches. It is probable that sections would have disclosed a ganglion here. She found no cutaneous branch of the nasalis at this point. Peripheral distribution of the ramus nasalis. The distribution of the cutaneous sensory fibers carried in this nerve begins with the bundle of coarse fibers entering the posterior end of the ethmoidal ganglion. Distal to the ganglion the main nerve divides into the lateral and median branches before mentioned (Plates 2, 3, figs. 4, 6, na. I., and na. m.), which are of about ec^ual size. One of these retains a median position while the other crosses the roof of the nasal capsule to a more lateral one. (a) Ramus lateralis nasi. This ramus in passing cephalad and laterad across the roof of the olfactory chamber occupies a position between the cartilaginous capsule of the chamber and the cranial bones. It is distributed to the skin over the subterminal region of the snout, that is, just posterior to the external nares. Before reaching the integument the lateral ramus passes through the nasal gland,. 56 bulletin: museum of comparative zoology. where it bifurcates. Each branch gives off a small twig (Plate 2, fig. 4, na. l.^ and na. /.-), which passes out through this gland to the skin. The more lateral of the tAvo bundles again divides, and these three terminal divisions of the ramus (Plate 4, fig. 9) then emerge from the anterior side of the gland to innervate the skin overlying that region. The two small twigs proximal to the three terminal divisions (Fig. 4) supply a somatic sensory innervation to the olfactory epithelium. (b) Ramus medialis nasi. After giA'ing off a few fibers to the olfactory epithelium, the median ramus supplies a larger branch {na. m.^) to the skin over the nasal bone; the foramen (for. na. vi.) through which this nerve emerges is clearly indicated in fig. 1, Plate 1. The main ramus then passes cephalad wdthout branching to the extreme tip of the snout to innervate richly the skin anterior to the external nares. Its terminal branches are shown in Plate 2, fig. 4, but not in Plate 3, fig. 6. In its course it keeps close to the internasal septum and the median dorsal part of the premaxillary bone (Plate 4, figs. 8, 9). II. Ramus maxillaris (Plates 2> 3, figs. 4, 6, 7, mx.). The second main branch of the trigeminus is given off from the lateral side of the portio major of the Gasserian ganglion. It passes between mm. pterygoideus and capiti-mandibularis to reach the posterior ventral rim of the bony orbit. It makes a circuit of this ventral rim to the anterior side of the orbit where it passes into the marrow cavity of the maxillary bone and continues its course in the upper jaw. Besides numerous rami distributing somatic sensory components, this nerve becomes involved with the superficial sympathetic rami of the head in what may be called the "lachrxTnal plexus," and with the viscero- sensory-sympathetic in the orbital plexus. It will simplify the ac- count to describe first the somatic sensory rami, as far as possible, as though the plexus did not exist and give an account of the latter separately. The following branches are given off from the ramus maxillaris: (a) the first branch (mx.^) is given off from the dorsal side. It passes around the anterior side of the m. capitis mandibularis and turning back supplies the skin (Plate 2, fig. 4; Plate 3, fig. 6). In its course it crosses the sympathetic rami in the lachrymal region, but its cutane- ous fibers do not mingle. In some dissections it appears to be inde- pendent, although in figure 7 it joins the lachrymal plexus in such a way that its terminal ramus could not be identified, (b) The second branch (mx.-) also passes across the plexus, where it is difficult to folloAV it in sections, but a dissection, upon being cleared and mounted, willard: cranial nerves of anolis carolinensis. o/ showed the continuity of the cutaneous fibers, which is indicated in Plate 3, fig. 6, by dotted Hues. This ramus divides into several branches to suppl}^ the skin at the posterior angle of the eyelids, and also the posterior half of the lower eyelid, (c) Another branch (mx.^), similar to (nix.-), runs forward to supply the anterior half of the lower eyelid. This shows an anastomosis with vix} (Plate 3, figs. 6, 7). (d) A branch (mx.'^) which passes into the cavity of the ventral bony arch of the orbit supplies the skin over it as far forward as the anterior limits of the orbit. The rest of the nerve {if orb.) now passes cephalad across the orbit to enter the infraorbital foramen, an opening bounded by the maxillary, lachrymal and jugal bones (Plate 1, fig. l,for. if orb). Before entering this foramen it gives off cutaneous fibers at two points, which, however, are not free from connection with the palatine rami and are indicated only by the somatic sensory component (yellow) included in these palatine rami. One of these combines with the visceral components of VII to form a good sized branch {pal. I.), which passes cephalad under the fold of epithelium on the median side of the maxillary bone. This fold is rich in tubular glands and taste buds (Plate 4, figs. 9-11). The other passes mesad to meet the inter- mediate ramus of the palatine {yal. i'm.), and the combined nerve comes out to the mucous membrane farther cephalad, where it con- tinues the innervation of the lateral field (Plates 2, 3) . Ramus alveolaris superior {ah. su.). This is the term applied to the intra-maxillary portion of the maxillary nerve. In Anolis its distribution is wholly to the integument at the side of the jaw, which it reaches through numerous foramina that are indicated in the skull (Plate 1, fig. 2), and to the teeth, as described more fully for the lower jaw. Connection of the sympathetic loith the ramus maxillaris trigemini. There is one well defined and constant sympathetic ganglion on the main trunk of the maxillary nerve. This, because of its position at the beginning of the infraorbital course of this nerve, is known as the infraorbital ganglion (Plate 2, 3, figs. 4, 7, gn. if orb.). It is much flattened laterally (Plate 5, fig. 14), the flattening being due, no doubt, to its location. It is connected with the main superficial sympathetic ramus of the head by a strand of non-medullated, or very slightly meduUated, fibers. The medullation is so slight that the connection is very difficult to establish in the sections, the dis- section preparations, however, leave no doubt on this point. It is impossible to tell to what extent recurrent fibers may connect with the trigeminal ganglion. It is certain, however, that the main 58 bulletin: museum of comparative zoology. extension of the sympathetic ramus is cephalad in the infraorbital nerve. This is shown by the character of the cross section of the nerve and by the direction of the fibers in the gangUon. The further anastomoses of nerves V and VII to form the infraorbital plexus offers opportunity for the wide distribution of this system throughout the head region. Practically every point of juncture in this complex is the seat of a larger or smaller group of sNTupathetic ganglion cells (Plate 2, fig. 4, cl. gn.sy.). Except in the case of the infraorbital ganglion, none of those on the maxillary nerve is large enough to have been discovered by gross dissection without the aid of mounted pre- parations of dissected parts, a method which apparently has not been used by my predecessors. It is noticeable that, after the nerve is free from visceral connection, as in the ramus alveolaris superior, no more ganglia are found. III. Ramus mandibularis (nid.). This is the largest of the tri- geminal branches, but, unlike the other two, it is not wholly afferent in its composition. It includes a certain number of motor components, which are given off to the mylo-hyoideus muscle. The mandibular ramus leaves the portio major of the Gasserian ganglion from its ventro-lateral side (Plate 3, figs. 6, 7), being covered dorsally by the ramus maxillaris. Its course to the lower jaw lies between the mm. pterygoideus and pterygo-sphenoidalis posterior. It crosses dorsal (lateral) to the pterygoquadrate process just behind the insertion of the m. pterygo-parietalis, passing through the deeper part of the pterygoideus to reach the median side of the mandible (Plates 5, 6, figs. 15-17). x\s it approaches the mandible two rami are given off together to be distributed to skin and muscle of this region. The continuation of the main nerve is known as the ramus alveolaris inferior {ah. if.). It still contains a few motor fibers for the inner- vation of the anterior part of the m. mylo-hyoideus. Ill, 1. The two branches above mentioned are given off from the main ramus at the point where the latter makes a turn cephalad to enter the mandible. They might be considered one branch which immediately divides into a posterior and an anterior ramus. The posterior ramus (md.^) turns back (Plates 2, 3, figs. 5, 6) and crosses Meckel's cartilage dorsally, reaching the lateral side of the mandible (Plate 6, fig. 16) through a foramen between the dentary and angu- lare bones. It passes through the superficial throat muscles to the mylo-hyoideus muscle, a very small sensory twig passing on to the skin. Most of the fibers are motor, innervating a part of the mylo- hyoideus muscle (Fig. J). willard: cranial nerves of anolis carolinensis. 59 The anterior ramus (lab. if. md.) is much the larger of the two and appears in the sections to be purely cutaneous sensory. It passes into the cavity of the complementare bone (Plate 5, fig. 15) to run cephalad a little distance and then out through a foramen on the dorsal side of this bone. It passes forward along the side of the mandible to innervate the integumentary portion of the lower lip for about two thirds of its length (Plates 4, 5, figs. 10-15), the anterior part of the labial region being cared for by branches from the alveolar ramus (Plates 2, 3, 4, figs. 5, 6, 8, 9, lab. if. ind.). Ill, 2. Ramus alveolaris inferior. This (alv. if.) is the continuation of the mandibular ramus into the lower jaw. It takes a position on the dorsal side of Meckel's cartilage (crt. Mkl.) where this is still exposed (Plate 5, fig. 15), and when the membrane bones of the mandible close around the cartilage they include both this nerve and the chorda tympani. The latter is at first ventral to the alveolaris, but gradually assumes a more median position and finally takes up the medio-dorsal part of the cross section of the combined nerves (Plate 5, figs. 13-15, ah. if. and cd. tym.). The two bundles are easily distinguishable up to the place where the first branches are gi\-en off. The alveolaris gives off several branches at a level with the angle of the mouth. (a) A small branch, not shown in the figures, is given off from the dorsal side of the main ramus at the line of separation between the fine fibers of chorda tympani and the coarser ones of the alveolaris. It contains the coarser, well-medullated fibers, not more than ten or twelve in number. This appears to be a constant structure, but its distribution and its function remain undetermined. There is nothing in the nature of its fibers to indicate that it is viscero-sensory or sympathetic, yet it cannot be followed to any peripheral structure; the fibers separate in the interosseous tissues of the jaw and cannot be traced outside. It is recurrent in its course, passing caudad and dorsad between the outer dentale and the enclosed complementare and is lost on the epitheliod osteoblastic layer betweSn the dentale and coronoideum. Its course continued a little farther caudad would bring it to the lateral side of the coronoideum, to the place of inser- tion of a part of the m. pterygoideus, but such a distribution was not established. Some of the fibers passing out in this nerve appear to be the coarser ones originally carried by the corda tympani. (b) Opposite the point where (a) is given off a ventral mixed branch (mf/.2) leaves the ramus alveolaris. In passing cephalad it circles Meckel's cartilage swinging down the lateral side and up the median 60 bulletin: museum of comparative zoology. side to emerge through the same foramen with another large branch (Ing. V. + cd. tym.), which includes the chorda tympani (Plates 2, 3, and 5, figs. 5, 6, 12). Upon its exit from the jaw it turns ventrally, dividing to send to the skin a branch which also includes a small motor twig {myl-hy.) supplied to the mylo-hyoideus muscle. The main part goes farther cephalad among the interlocking ends of the mylo- hyoideus and cerato-mandibularis muscles to an overlying cutaneous area between mandibles (Plate 4, fig. 9, md?). (c) A short distance cephalad of the place where (a) and (b) are given ofi", the main ramus bears on its dorsal side a sympathetic ganglion (Plates 2, 3, figs. 5, 6, gn. md.), which is composed of small cells forming a group flattened against the sides of the chorda tympani and alveolar portions of the nerve. As the combined lingualis and chorda tympani separate from the intramandibular portion of the nerve, the cells of this ganglion crowd down into the angle thus formed in the shape of a wedge. As in other sympathetic ganglia, there appears to be no admixture of medullated fibers among the cells. The ganglion seems to be more closely associated with nerve V than with nerve VII, and after the division its cells continue forward, ap- pearing in a number of sections within the perineurium of the alveolar part. Gaupp ('88, p. 460) states that in Lacerta there is an "Anschwel- lung" at the point of union of the chorda tympani with the alveolaris. He also refers to the lingualis, which includes the chorda tympani, as given off immediately after this union. This swelling undoubtedly corresponds to the mandibular ganglion {gn. md.) of Anolis. The lingual ramus of V passes out of the jaw in a common sheath with the chorda tympani (Plates 2, 3, 5, figs. 5, 6 and 12, Ing. V + cd. tym.). The combined bundle goes cephalad for some distance without branching, then divides into two rami of about equal size. The dorsal one is composed almost entirely of the fine fibers of the chorda tympani and will be considered as the continuation of that nerve, to be described with the facialis. It includes, however, about a dozen large cutaneous fibers. The other (ventral) division is almost entirely cutaneous, as is shown by comparison with ramus {md.-), which lies next to it. It contains, however, some of the fine fibers characteristic of the chorda tympani. It appears, therefore, that in the final separation of the lingualis and chorda tympani there is a slight interchange of fibers. The lingualis {Ing. V.) now runs cephalad and divides for distribu- tion to the papillae of the tongue. In this peripheral region, however, willard: cranial nerves of anolis carolinensis. 61 its fibers are inextricably mingled with those of XII, which they closely resemble. The difficulties of plotting the mixed rami are here further increased by the retraction of the tongue muscles, which throws the terminal rami into numerous coils. The plotting (Plate 3, fig. 5) attempts to trace the lingualis and XII only so far as to show the union of the entire lingualis by two branches with branches of XII. This takes place after the lingualis has been traced cephalad for some distance parallel with the chorda tympani and then back again parallel with XII close to the mucous membrane at the base of the groove which bounds the tongue laterally (Plate 4, figs. 10, 11). Numerous dissections were made in this region, which added little information as to the exact form of the anastomoses. The fact was established, however, that, after joining XII, some of the fibers of V continue caudad presumably to reach the periphery through more proximal branches. Terminal distribution of the ramus aheolaris inferior. The final distribution of the sensory components, left in the alveolar branch after the separation of the chorda tympani, the lingualis, and a large cutaneous branch, is accomplished through a series of terminal rami, which make their exit from the alveolar canal through a number of foramina on the lateral side of the dentary bone (Plates 2, 3, figs. 5, 6, lab. if. alv.). These are termed by Watkinson (:06, p. 462) the "rami ad glandulae labiales inferiores," the emphasis thus being placed upon their distribution to glands. They probably carry sympathetic fibers to the glands, but if so, these are non-meduUated and are not distinguishable in sections. A conspicuous feature in Anolis is the highly sensitive nature of the lips, as is indicated by the rich supply of typical cutaneous sensory fibers to the skin in this region. The epidermal sense organs are also more numerous here than elsewhere. The strongly medullated sensory fibers are strictly limited to the epidermis at the sides of the lips lateral to the external labial gland (Plate 4, figs. 8-10). The external glands are not very largely developed here, although they occur along the entire length of the lips both upper and lower. The n. alveolaris does not emerge as a whole from the alveolar canal until the end of the mandible is reached. Here it comes out as four or five terminal branches, giving to this region an even richer nerve supply than was accorded the lips along the lateral part. Innervation of the teeth. The teeth are of the typical pleurodont type, being fused directly with the median side of the dentary bone. The pulp cavity of each tooth is in commimication with the cavity 62 bulletin: museltni of comparative zoology. of the dentary bone, which carries the nerve. Through this opening into its cavity the tooth receives its nerve and blood supply. The innervation was not determined for the posterior teeth, the first fibers given off from the alveolar nerve being farther forward, where a small number (3 or 4) of fibers with distinct sheaths come off from the main ramus and run forward within the alveolar cavity for some distance, but eventually pass into the cavities of the teeth (Plate 4, fig. 9, rm. de.). Where these fibers are given off they show a character similar to the cutaneous sensory components, although they either become less heavily myelinated or, what is more likely, their position makes fixation defective, for they can -ft-ith difficulty be traced through the communication into the pulp cavity of the tooth. 'It is not improbable that here they do lose their myelin sheaths. It is notice- able in the mature teeth that the dentinal tubules are particularly large at the apex of the teeth. No nerve fibers could be followed, however, as far as the odontoblasts underlying this region. This alveolar ramus is described b}' Watkinson (:06, p. 462) as innervating the teeth through the "rami dentales." Such rami in Anolis would consist of a few fibers given off at wide intervals, which would be difficult to discover except by microscopic methods, even though they were not entirely \\athin the bony part of the jaw. Xorris ( :08, p. 522) referred the innervation of the teeth in Amphiuma to a combined V and VII nerve which runs along the median side of the mandible. This nerve corresponds to the main part of the chorda tympani in Anolis, which after temporary union with the mandibular nerve has separated from it, reappearing with a mixture of cutaneous components, which are distributed \\dth it. Although many fine fibers pass from this nerve along the base of the teeth to the epithelium of the lingual gums, no connection was established with the. teeth themselves. N. FACIAL NERVE. The roots of the facial nerve. Two roots can be demonstrated for nerve VII, a lateral (motor) one and a dorsal (sensory) one. The sensory passes out directly from the fasciculus solitarius (Plate 6, fig. 18, rx. VII) as a strong bundle to emerge from the brain dorsal to, but in contact with, the root of VIII. It passes around the anterior side of this root, being entirely covered laterally by the large acoustic ganglion. On the ventral side of the root of VIII the sensory root of willard: cranial nerves of anolis carolinensis. 63 VII joins the motor root and the combined roots pass out through the foramen as one (Plate 6, fig. 18). The motor root leaves the brain just ventral to the fibers of nerve VIII where it is joined by the sensory root. It was traced centrally as a distinct bundle near to the ventro-lateral floor of the fourth ventricle. Here the root bundle becomes diffuse and its nucleus of origin was not identified among the several groups of cells occurring in this locality. The motor components of the facial are all included in the hyoman- dibular nerve (hy-md.), -with a distribution posterior to the ear. The viscero-sensory components are distributed anterior to the ganglion, the roof of the mouth being supplied through the palatine ramus {pal. VII), while the tongue and adjacent regions are supplied by the chorda tympani {cd. tym.), which leaves the ganglion in company with the motor elements and includes the sensory components of the hyo- mandibular nerve. The geniculate ganglion (Plate 3, fig. 7, Plate 6, fig. 18, gn. VII.) lies closely applied to the cranial wall just outside the foramen (Plate 1, fig. 2, for. VII), through which the roots make their exit. It is a small ganglion, but one readily demonstrated by dissection. Fischer ('52) found it in all cases, but Watkinson (:06) states that in Varanus VII shows no swelling at the point where the ganglion should appear. (1) Ramus palatinus VII (Plates 2-6, pal.). This is composed wholly of fine lightly meduUated fibers of quite uniform character. This at least is the condition presented by the sections. The pres- ence of some non-medullated fibers cannot be denied, however, in regard to any nerve described. The source of these medullated fibers of the palatine nerve is twofold: (1) from the cells of geniculate ganglion, and (2) from fibers carried in the communicating branch between the IX + X complex and nerve VII. The latter is known in lower groups of vertebrates as Jacobson's anastomosis, where it is homologized with the pretrematic ramus of IX and considered viscero- sensory in character. It was pointed out by Cole ('98, p. 145) that many fishes show this communicating ramus as combined sympathetic and viscero-sensory components derived from IX. This appears to be the condition in Anolis. The connection is by one or two fine strands (comn. i.), which lie close to the artery just outside the ear capsule (Plate 3, fig. 7). They come into the proximal part of the palatine ramus where some, if not all, of the medullated fibers of the communicating ramus join those of the palatine to be distributed with the latter. Other fibers, for the most part non-medullated. 64 bulletin: musel'^ni of comparative zoology. apparently stop here, and a small group of sympathetic ganglion cells is found at the point of juncture of the two ner^•es. These facts sup- port the conclusion that we have here a combined sympathetic and viscero-sensory ramus. The palatine ramus passes cephalad from the ganglion swinging in toward the median line (Plate 3, fig. 7). Its course is along the groove which marks the division between the prootic and the basioccipital (Plate 1, fig. 3). It passes into the sphenoid and crosses the base of its process dorsally, being carried in an imperfectly closed foramen (Plate 6, fig. 17, pal). The palatine passes into a narrow space of the sphenoid and out again on the median side to assume a position just underneath the mucous membrane median to the basiptergoid process of the sphenoid (Plate 5, fig. 15). In its course cephalad it keeps along the dorsal side of the pterygoid near its median border, gradually taking a more lateral position, to the point where it bears the ganglion palatinum (Plate 3, fig. 7, g7i. pal.). Throughout this portion of its extent it gives off no fibers for distribution. It receives (or gives off) the anastomosing branch connecting with V by way of the ramus to the palpebral muscle (Plate 3, fig. 6, dcp. palp, if., see p. 50). This does not change the character of the palatine in any recognizable way, and there are no ganglion cells in connection with this anastomosis. The ganglion palatinum is a s^inpathetic ganglion, which appears on the main palatine ramus proximal to any terminal branches. It lies on the floor of the orbit ventral to the palpebralis muscle and marks the beginning of the infraorbital plexus between nerves V and VII. A small number of medullated fibers are given off laterally, the main bundle dividing again just distal to the ganglion. The ganglion cells are strongly grouped at the first division point and also extend along the lateral branch for a considerable distance (Plates 2, 3, figs. 4- 6). From this point forward to the level of the ethmoidal ganglion the palatine components are carried in rami which anastomose more or less with each other and with the infraorbital ramus of nerve V. The infraorbital plexus consists of a number of anastomosing branches spread out in the orbit on the dorsal side of the pterygoid and palatine bones, these branches connecting rm. palatinus VII with maxillaris V (infra orbital portion). The posterior (proximal) limits of this plexus are marked by the sympathetic ganglia of palatine VII and infraorbital V (Plates 2, 3, figs. 4, 6, 7, gn. pal. and gn if orb.). This plexus was studied both from series of sections and from dissections with a view to determining its constant features. The complete plexus was much better demonstrated by the latter method willard: cranial nerves of anolis carolinensis. 65 (Plate 3, fig. 7), the plotting from sections (Plates 2, 3, figs. 4, 6) being incomplete. Notwithstanding great variation in detail, the scheme of the plexus seemed to be in the main the following : — (1) Taking the main palatine as a point of departure, all the rami have their origin from VII at one point, or nearly so. This is where the palatine ganglion lies, as described by several authors. The palatine here is a ramus of considerable size and immediately gives off two main branches. These come off separately, but close together, in the plotted series (Plate 2, 3, figs. 4, 6). In fig. 7, Plate 3, they both arise from the ganglion. One of these {pal}) swings abruptly outward crossing ventral to the infraorbital ramus of V (Fig. 7). It innervates the mucous membrane and taste buds (gm. gus.) on the way. The other (pair) keeps a more median position until it crosses V to anastomose with the first {pal}) to form the lateral branch of the palatine {pal. L). (2) There is an anastomosis {an'stm. pal. I.) with V, from near this juncture, and sections always show ganglion cells at this point (Fig. 4, cl. gn. sy.). (3) There are always two or three branches (Fig. 7, a, 13, y) which turn back to join V just distal to the infraorbital ganglion, and their appearance indicates a large proportion of non-meduUated fibers. (4) Anterior to the orbit there are anastomoses (a) that connect the intermediate branch of the palatine with the maxillary nerve just before the latter passes into the maxillary bone (Plate 2, 3, figs. 4, 6 and 7, an'stm. pal.i'm), and (b) that connect the, median palatine ramus with ramus nasalis V through the ethmoidal ganglion. These anastomoses account for the mixed character of the lateral and intermediate rami of palatine VII. Terminal twigs given off from the infraorbital plc.vus. Of more importance than the exact form of the anastomoses are the twigs for terminal distribution given off from the branches making up this plexus. In all cases the main course of the principal branches is explained by an examination of the mucous membrane of the roof of the mouth. x\ll the rami carry viscero-sensory fibers for the innerva- tion of the taste buds. These sense organs are distributed within two restricted fields of the mucous membrane underlying the orbit, a small median area and a lateral one. The median group of taste buds (Plate 5, fig. 12, gm. gus. m.) is much more limited than the lateral group. Here the buds are within a sensory-glandular patch little more than a millimeter in length (between sections 732 and 874), which lies on the pterygoid. The position is approximately indicated in the fig- ure of the skull by the foramen in the pterygoid bone (Plate 1, fig. 3, for. pt.). The innervation is from a twig {gm. gus.) of the median 66 bulletin: museum of comparative zoology. palatine ramus (Plate 3, figs. 6, 7, paU), which passes through the foramen along with an artery and vein. The medullated fibers in this twig can not be followed farther in their distribution than the limits of this gustatory patch. Anterior to the orbit the viscero-sensory fibers are found in three rami, a median (pal. m.), an intermediate {pal. Vm.), and a lateral {pal. I.) ramus. All these rami carry a few coarser fibers not charac- teristic of the palatine proximal to its anastomoses. These are inter- preted as cutaneous sensory from V. (a) Ramus medialis. This is a continuation of the principal branch of the palatine {pal.^). At about the level of the ethmoidal ganglion {gn. eth.), or proximal to it, there is a division of the principal ramus {pal.^), which results in a branch passing laterad to help make up the intermediate ramus (Plates 2, 3, figs. 4, 6, 7), but the main branch is continued forward as the ramus medialis {pal. m.). In the dissection (Plate 3; fig. 7) this connection seems to be opposite the ganglion instead of proximal to it. A very small group of sympathetic ganglion cells is found at this junction. Another, shorter, branch con- nects the median ramus with the ethmoidal ganglion, thus forming an anastomosis with the nasalis. It would appear that this anastomosis serves the purpose of bringing the ethmoidal ganglion into connection with all the other rami. This connection of the median ramus "wdth the ethmoidal ganglion has already been described (p. 54). Distal to this point the median ramus continues an uninterrupted course to the tip of the snout to innervate the mucous membrane of the pre- maxillary region. This is a glandular region richly supplied vnXh taste buds. It is produced by the median union of the lateral glandu- lar gustatory folds of the two sides. The course of the ramus medialis is just dorsal to the median edges of the palatine and vomer bones on either side of the interscapular nasal cartilage (Plate 4, figs. 8, 9). A short distance proximal to its distribution this ramus is joined be- neath the cartilage with its fellow of the opposite side by a median group of sympathetic ganglion cells (Plate 2, fig. 4, cl. gn. sy.). Distal to this the united mediales continue forward for some distance as a common median bundle; this bundle, however, splits again near the region of its distribution. (b) Ramibs intermedius {pal. i'm.). This one of the terminal rami of the palatine is formed by the union of two branches lying dorsad of the palatine bone, one from maxillary V, just before its entrance into the infraorbital foramen, the other from the ramus medialis at, or proximal to, the ethmoidal ganglion. At all three of these junction willard: cranial nerves of anolis carolinensis. 67 points s^inpathetic ganglion cells are to be found. When formed, the ramus is a mixed nerve containing some fine medullated viscero- sensory fibers from the palatine VII, some coarser medullated fibers of the cutaneous type from the maxillary V, and likewise non-medul- lated sympathetic fibers. The distribution of the medullated fibers of this ramus (Plate 4, figs. 9-11) is along the median side of the glandular gustatory strip (Plate 4, fig. 9, gm. gus. I.), which begins at the angle of the mouth and extends to the anterior end of the upper jaw. Posterior to the distribution of this ramus the glandular band is narrower and is innervated wholly from the ramus lateralis (Figs. 10, 11) ; here however, it broadens. By the narrowing of the jaw the two rami (r. lateralis and r. intermedins) are brought closer together, until finally their terminal branches mingle (Plates 2, 3, figs. 4, 6; Plate 4, fig. 9). (c) Ramus lateralis {pal. I.). This, like the preceding, is a mixed ramus; but contains a larger proportion of cutaneous fibers. These are drawn off from the maxillary ramus at the anastomoses indicated in the plotting and dissection {an'stm. pal. I.). In the dissection (Plate 3, fig. 7) these anastomoses of the intermediate and lateral palatine branches with maxillary V are effected at the same point. The ramus lateralis passes cephalad along the median side of the maxillary bone and innervates taste buds and general epithelial sur- faces (Plate 4, fig. 10). As described above, it has a terminal area of distribution which is common to it and to the intermediate ramus. The relation between ner\'es V and VII, described here as the infra- orbital plexus, may be taken as representative of the group. About the only constant feature in the complex, however, is the palatine ganglion or the point corresponding to it. It is believed that this junction of V and VII would always disclose a ganglion if examined microscopically, and the apparent importance of this ganglion as a structural feature in Hatteria (Osawa, '98, p. 603, fig. 45) may be due to, the fact that the cells are clustered instead of being scattered along the nerve. The latter condition is probably due to the position of the nerve, compressed as it is against the floor of the orbit. There is seen to be some evidence, from the character of the fibers in these anastomosing branches, to justify the acceptance, in a general way, of Fischer's ('52, p. 138, 139) distinction between a posterior and an anterior anastomosis, the posterior being a "sling" of the sympathetic system, while the anterior is a mixture of fibers from V and VII for distribution. Fischer did not recognize the sympathetic ganglia present in this region, but based his view on the proximity of the 68 bulletin: museum of comparative zoology. connection between the superficial sympathetic trunk of the head and the infraorbital nerve. (2) Ramus hyomandihularis (hy-md.). This nerve leaves the genicu- late ganglion (Plates 2, 3, figs. 4, 6, 7, g7i. VII) at its posterior end. It embraces \'iscero-motor and viscero-sensory components. In cross sections it is to be seen that these two components are entirely sepa- rate throughout their course, the motor fibers having the more dorsal position. The latter include all the motor elements of nerve VII; these pass from the root across the dorsal side of the geniculate gan- glion, not penetrating it, and then turn sharply into the ramus hyo- mandihularis. The motor and sensory elements each take up about half the area of a cross section of the nerve, the sensory components being of course much finer in caliber than the others. The course of this ramus is dorsad and caudad, following the same projecting ledge of the prootic bone (Plate 1) as does the palatine, but in an opposite direction (Plate 3, fig. 7) . This course leads it to a point just ventral to the articulation of the quadrate with the prootic process of the ear capsule (Plate 6, 7, figs. 19-20). Here occur the crossing and anastomoses of the head sympathetic trunk {comn. ex.) with the hyo- mandibular ramus of VII (Plate 3, fig. 7) . The superficial sympathetic trunk {comn. ex.) from this point to the lachrymal plexus and infra- orbital ganglion is called by Watkinson ( :06, p. 464) " ranms recurrens nervi trigemini ad facialem " ; its continuation to nerve IX, the " ramus communicans externus cum glossopharyngeo." In a series of sections of Anolis, in which fixation in this region was excellent, it is shown with certainty that practically all the medul- lated fibers in this sympathetic ramus, which in every way resemble viscero-sensory fibers, pass the facial nerve in continuous course. The facial, however, makes its way between the sympathetic fibers, most of which cross the facial nerve mesally; a few only split away from the others to cross it laterally, and then immediately rejoin the main bundle. This is not a sympathetic center of any importance, although from four to six very small ganglion cells are inclosed l^etween the two nerves at the point of crossing. While there was no evidence that any of the fibers in the part of the sympathetic ramus between VII and V turned into any of the parts of VII at this point of union, there does seem to be a strong indication that some of the fine medul- lated fibers contained in the hyomandibular ramus of VII are continu- ous with a part of those in the posterior section of this sympathetic trunk, i. e., that between VII and IX. Attention is here called to the fact that cutaneous fibers are not willard: cranial nerves of anolis carolinensis. 69 demonstrable in VII and are not generally considered pi-esent in rep- tiles. They are, however, found in the Amphibia (Norris and others), and their distribution there would agree quite closely with the reptiles if it were discovered that a trace of the cutaneous component were carried to the skin in this sympathetic ramus, for the course of this ramus is superficial, and anteriorly it anastomoses freely with the cutaneous fibers of V. The conditions for study here were not such that a failure to observe this would preclude the possibility of a very few cutaneous fibers taking this course, should the central relations indi- cate a connection with the somatic sensory tract. The hyomandibu- lar just distal to the crossing of the sympathetic divides into the motor ramus hyoideus and the sensory chorda tympani (Plate 7, figs. 20, 21, hy. and cd. tym.). (a) Ramus hyoideus \hy.). After parting with its sensory compo- nents, the motor part of the hyomandibularis continues its course to the muscles supplied by it. As this nerve reaches the depressor mandibular (digastric) muscle it divides into two branches for dorsal and ventral distribution (Plate 3, figs. 6, 7, hy., and Plate 7, fig. 23). The final distribution of this nerve is well showni in figure J, which is a drawing from a mounted dissection that had previously been treated with vom Rath's mixtures. This shows that the digas- tric, sphincter colli, and the most posterior part of the mylo-hyoideus are innervated by ramus hyodeus (motor VII). This agrees with what Ruge ('97, p. 331) found in Varanus, although Watkinson (:06) was not able to discover it in her dissection. Ruge ('97), in his extensive monograph on the facial nerve in the vertebrates, considers the mylo-hyoideus muscle in reptiles as belong- ing to the inner\ation field of motor VII, and he finds this demonstra- ble in Hatteria (p. 325), where the ventral ramus of VII is figured as extending almost to the end of the jaws. In the same form the ramus mylo-hyoideus of V lea^•es the jaw in the manner typical of that nerve, but Ruge considers it wholly cutaneous sensory. In the alligator the same muscle is innervated by the motor fibers carried in the ramus mylo-hyoideus of V, and Ruge (p. 381) concludes that V has recei^^ed these " intracranially " from VII. Comparing with the amphibian Amphiuma, as described by Norris (:08), the motor components in ramus hyoideus VII of Anolis are directly homologous with the ramus jugularis of Amphiuma, which innervates the digastric, sphincter colli and posterior part of the mylo-hyoideus muscles, leaving the anterior part of the mylo-hyoideus to be inner\'ated by the ventral division of the main mandibular, which is evidently the ramus mylo-hyoideus of V as described in Anolis (p. 61). 70 bulletin: museum of comparative zoology. (b) Chorda tympani {cd. tym.). This nerve, which draws off all the sensory elements from the hyomandibular ramus (except possibly some which turn back into the sympathetic trunk), passes laterad on the roof of the middle ear chamber to reach the median face of the quadrate bone (Plate 3, fig. 7, qd.). In its course it passes dorsad of the ligament of the columella auris (Plate 3, fig. 7, clml. aur. and Plate 7, fig. 20, lig. tym.) and then follows the quadrate ventrally on the anterior side of the middle ear chamber {aur. m.). It at once enters the cavity of the articulare (Plate 6, fig. 18) through a special foramen. Its course is now cephalad within the jaw; but at some distance forward it passes out through the dorsal side of the articulare (or angulare) to take a position on the dorsal side of Meckel's cartilage, which is here exposed (Plate 5, figs. 14, 15). It keeps this position in its forward course until joined by mandibularis V, which comes down on to it from the dorsal side. It becomes included within the same perineurium with this mandibular ramus, gradually shifting around to the dorso-median side of it (Plate 5, fig. 13). At about the middle of the length of the jaw it is split off with a large somatic-sensory con- tingent and passes out of the foramen as previously (p. 59) described (Plates 2, 3; Plate 5, fig. 12). The question of the homology of the chorda tympani in the different vertebrate groups has been recently discussed by Sheldon (:09) in a very thorough manner and it need only be added here that Anolis offers no obstacle to the conclusions there reached. My studies of Anolis embryos in connection with the present work show the chorda tympani to have a history similar to what it has in mammals (Emmel :04), in that it belongs to a posttrematic ramus which is di-awn across the developing tympanum after it has established connections distally with the lower jaw. Versluys (:03) has recorded for Lacerta a development similar to that of Anolis and shows in an embryo with one open cleft that the chorda tympani passes posterior to the cleft to reach the lower jaw. He interprets the adult condition as due to the fact that the closing membrane of this cleft does not give rise to the adult tympanum, this structure being developed posterior to the chorda tympani nerve. Thus the latter is a^true posttrematic ramus notwithstanding the evi- dence to the contrary wiiich is presented in the adult. In the adult Anolis the chorda tympani can be followed in the main to its terminal branches, and the close correspondence between these and the distribution of taste buds in the lower mouth region adds a new kind of evidence to support the conclusion that the chorda tym- pani is the nerve of taste for the tongue region. willard: cranial nerves of anolis carolinensis. 71 Gaupp ('88) made a comparative study of the innervation of the mouth and nasal glands in vertebrates. Of the saurians he studied Chamaeleo, Platydactylus, and Lacerta. He also described the con- ditions in the other reptilian orders. His results for the lizards, briefly summarized, are as follows. Superior labial glands innervated by maxillar}' V in the maxillary part, by the ophthalmic V (nasalis) in the premaxillary part, possibly by Vn also. Median palatine glands by palatine VH exclusively. Lateral palatine glands by V and VH. Inferior labial glands by the terminal twigs from ramus alveolaris inferior V. Sublingual gland by lingual V and chorda tympani VH. Lingual glands same as sub- lingual. On the basis of innervation he homologizes the palatine glands of the lizard with the "Rachendruse" (Born '76) of Amphibia, and the reptilian sublingual with the mammalian sublingual and sub- maxillary. In the nerve distribution described, Gaupp recognized the essential relations between V an,d VII in both upper and lower mouth wall that have been pointed out in Anolis. He does not take into account other structures associated with the glands, which might account for the presence of certain nerves in proximity to them; for example, the taste buds, which are universally present wherever glands occur within the mouth region (not including the labial glands) . Nor does he recognize important visceral elements in the form of sympathetic fibers which are carried by all these nerves and whose relation to the glands is not well understood. O. GLOSSOPHARYNGEAL NERVE. Nerve IX is connected with the brain by two roots (Plates 2, 3, figs. 6, 7), a dorsal fine-fibered and a more ventral coarse-fibered root. Both roots appear at about the same point in the series of cross sections, which, because of the flexure of the medulla, would indicate a more posterior position for the ventral one. The dorsal root is composed wholly of fine fibers of the viscero-sensory type. The ventral is presumably a motor root because the fibers are similar to the motor components of VII and may be traced directly through the ganglion and to its union with XII along with some of the fine fibers. The motor and sensory roots pass down separately to the closing membrane of the foramen, through which they emerge as a common trunk (Plate 6, fig. 19, rx. IX). The central courses of these two components were partly made out from the series of cross sections. The fasciculus solitarius is a clearly 72 bulletin: museum of comparative zoology. marked tract posterior to the sensory root of nerve VII and, as in the case of nerve X described below, the fine-fibered sensory root of IX passes directly to this tract. What is stated regarding the motor components of IX would apply equally well to nerve X. The combined roots pass out of the cranium by way of the recessus scalae tympani between the ear capsule and the basoccipital bone. Versluy's ('98, p. 180) general statement applies to Anolis on this point: " Bei alien Lacertilia vera tritt demnach der Nervus glosso- pharyngeus nicht durch ein eigenes Loch in der Paukenhohle, sondern durch eine grosse Oeffung, welche die altern Autoren meist Fenestra rotunda, die neuern Foramen jugulare externum genannt haben." The root passes caudad underneath the mucous membrane to its ganglion (Plate 7, figs. 20-23). This, the ganglion petrosumi (gn. IX), as compared to the root ganglion of the vagus, has a much more distal position on its root (Plates 2, 3), and probably represents the trunk ganglion of X, inasmuch as there has been described in other forms a root ganglion in addition to the petrosum. This ganglion occupies the free space between the other organs on a level with the posterior edge of the ear capsule (Plate 7, fig. 23). Its form is that of a uniform o^'al, and it lies on the dorsomedian side of its fiber bundle, which it incompletely surrounds. The coarse motor fibers may be traced directly through this bundle and out into the nerve trunk beyond. They have the most ventral position — that farthest removed from the ganglion cells — in their course through the ganglion. The other fibers seem to be non-medul- lated within the ganglion, but this appearance may be due to absence of impregnation by the osmic acid. Between the ganglion of IX and the union of its main trunk with XII there are connections with X and with the sympathetic system. These vary in their position, as study of several series of sections has shown, although certain relations are quite constant. In the series plotted (Plates 2, 3) the anastomoses are relatively simple. The frontal projection (Plate 2, fig. 4) is the only one that shows them. These will be described first and the variations referred to later. A very small bundle of fibers from the vagus (Plate 3, fig. 7, comn. IX-X) enters the petrosal ganglion on its proximal side; these become entirely mingled with fibers of IX so that the two are not separable beyond the limits of the ganglion. When the bundle emerges from the ganglion as the main trunk of IX, it shows' in cross section six or eight coarse motor fibers, which have been followed from the motor root above described; the rest are very fine fibers, but with sharply willard: cranial nerves of anolis carolinensis. 73 staining medullary sheaths. These fibers are not placed compactly together and there is an appearance as though non-medullated fibers lay between them. The second connection of IX with other nerves occurs at the level of the distal end of the ganglion. Here the sym- pathetic, formed by the union of the "rami communicantes internus et externus IX et VII" {cornn. i. and comn. ex.), joins the bundle of IX on its ventral side and, after contact and some mingling of fibers for the distance of 1/10 mm. becomes free again (Plates 2 and 3, figs. 4, 7). (In dissections, no distinction can be made between the actual mingling of fibers and inclosure within a common sheath). This connection apparently has no relation to the ganglion, for it occurs on the ventral side of the ganglion at a point where the most distal ganglion cells occupy only the dorsal side. After this contact the nerve shows no more connections up to its union with the superior laryngeal X. The variability in these connections is further emphasized when figure K is compared with figs. 4-7, Plates 2, 3. In the former the first named connection is absent and the second is accomplished by means of a s-mall ramus joining the sympathetic trunk. P. VAGUS NERVE. The central connections of the vagus nerve were less fully deter- mined than those of the other nerves, owing to the fact that the roots are extremely small and the few fibers which each contains do not keep together within the brain but separate into even smaller bundles or single fibers. There is also some variation as to the number of roots that could be identified peripherally. There is considerable shifting of central nuclei in the sharp flexure of the hind brain, making it impossible to determine conclusively the origin of efterent fibers from particular cell groups without the aid of the Golgi, or some similar method. Roots of the vagus. In series 30 (Plates 2, 3, figs. 4-6) the vagus enters the jugular foramen as three roots (c/. Fig. K) . Two of these appear in Plate 7, fig. 20 {rx. X) ; to avoid confusion of lines, the plottings show only one root for each component. Of the three, the posterior one contains the deeply staining motor roots; the fibers in the other two do not appear to be of uniform character. The middle root is the smallest. The posterior root has its superficial origin along ^4 bulletin: museum of comparative zoology. the lateral surface of the medulla a few sections anterior to the fora- men. Its fibers enter the brain at a sharp upward angle. The coarser fibers of the posterior root remain together and may be traced as they pass mesad in a broad upward curve to become spread out in the midst of a group of cells lying a little dorsal and lateral to the fasciculus solitarius. The more lightly medullated components of this root could be found making continuous connection with the fasciculus solitarius, into which they abruptly turn. The other two roots enter the brain along the same line as the posterior one. The smaller one could not be followed, but the larger could be traced to the fasciculus solitarius. It also contained several coarse fibers, which take the same direction as those of like character in the posterior root. The three roots remain separate until they enter the jugular ganglion. The ganglion jugulare, or root ganglion of the vagus (Plates 2, 3, figs. 4, 5, 6, gn. rx. X), Plates 3, 7, figs. 7, 22, gn. X), lies closely crowded into the angle formed betw,£^ the otic capsule and the basioccipital where the jugular foramen opens. This foramen appears in the same transverse section as the first occipital foramen of XII (Plate 7, fig. 20). The ganglion which is larger than that of IX or VII, is triangular in transverse sections of the head (Fig. 22, gn. X). This form is the result of the pressure of surrounding structures, the ganglion being crowded against the otic capsule by the spinalis colli muscle (Plate 7, fig. 22, spi. coll.). The form of the ganglion as a whole is notable, owing to the fact that the ganglion cells are so grouped on the mesial side of the fiber bundle that in dissections (Plate 3, fig. 7) the ganglion appears to lie free along the root bundles for a short distance. The fibers entering the ganglion as separate roots emerge on its distal side as one bundle, the coarse motor fibers, about sixteen in number, being grouped in its dorsal portion. This bundle (A') passes caudad parallel with IX and XII (Plates 2 and 3). On its way it shows the small ramus connecting it with IX (Plates 2 and 3, figs. 4, 7, com7i. IX-X). Posterior to the ganglion of IX, the trunk of the vagus divides into a sujicrior laryngeal ramus and a visceral ramus {mc. X.). The superior laryngeal ramus soon joins the pharyngeal ramus of IX to form the phar\Tigo-laryngeal ramus of IX + X (Plate 2, fig. 4, phx-lar.). This then combines with the trunk of XII (fig. 7) to reach the ventral side of the pharynx, {cf. fig. K) . (a) Ramus laryngeus superior {lar. su.). This division of the vagus includes all the coarse fibers of the main trunk and about one half willard: cranial nerves of anolis carolinensis. 75 the fine fibers. It forms a bundle about two thirds the diameter of the main trunk of IX, which it joins to form the pharyngo-laryngeal nerve (fig. K, Plate 2, fig. 4, phx-lar.). From this ramus are given off several small twigs to the constrictor muscle of the jugular vein (p. 44). These are fibers somewhat larger than the viscero-sensory fibers, but with extremely delicate myelin sheaths, and for this reason they were not discovered in the series of sections plotted; but in another series through this region (Fig. 7v), especially fixed, the innervation of these muscle fibers was determined. Of the three twigs shown in the draw- ing only one is given off from the superior laryngeus before its union with the ramus pharyngeus IX. It is important to note that these visceral muscle fibers (Plate 7, fig. 23, co'st. vn. j.i.), although striated, do not draw off any of the coarse fibers from the vagus, but are supplied by nerve fibers which are indis- tinguishable from the other fine fibers when mingled with them, but which nevertheless possess slight differences, as is shown when they are grouped together. We probably have in the innervation of this muscle a case analogous to that of the ciliary muscle, which primarily is non-striated, but in the sauropsida is striated. If the striated muscle fibers surrounding the jugular vein have been differentiated from the smooth muscle cells of the vessel wall, which are believed to be innervated by non-medullated postganglionic neurons, the question 'suggests itself as to what modification of the innervation has accom- panied that of the musculature. As before stated the nerve fibers show a slight medullation indicating to that extent a change from the sympathetic type, but their continued course through the ganglion, suggesting direct central origin, was not shown in the sections although this was clearly demonstrated for the more heavily medullated fibers passing into the pharyngo-laryngeal branch. Onuf and Collins ( :00, p. 174) describe two nuclei for eft'erent neurons of nerves IX and X in the mammals (cat). The dorsal glossopharyngeal and vagus nucleus, is, according to them, the nucleus of origin for the efferent sympathetic fibers carried in the roots of these nerves; the ventral, nucleus ambiguus, gives rise to the nerve fibers innervating muscles ■of visceral origin but of somatic function, derivatives of the striated gill-arch musculature of the fishes. The spinal accessory nerve, when present, is exclusively of the latter type. The innervation of the special jugular vein muscle of Anolis suggests a condition intermediate between the sympathetic and the viscero-motor of the cerebro-spinal type. The slight development of this latter component in nerves IX and X made it impossible to establish this suggestion as a fact through the analysis of the central terminations. 76 bulletin: museum of comparative zoology. Bruner ( :07, p. 47) gives quite a detailed description of the innerva- tion of the striated muscle of the jugular vein in reptiles. He finds this muscle in Lacerta aqilis to be innervated by a number of nerve twigs which are given off from the rami communicantes internus et externus. The latter join the proximal end of the petrosal ganglion, of nerve IX, which, in addition to its root is also joined by a communi- cating ramus from ganglion X. Bruner applied stimulation methods at various points in this nerve complex and thus determined the path of the motor fibers to the "m. constrictor venae jugularis internae" to be from the brain through the roots and root ganglion of nerve X, across to the petrosal ganglion through the communicating ramus between IX and X, and then cephalad along the rami communicantes internus and externus to points where the " nervi tumefactores capitis " are given off to the adjacent muscle surrounding the jugular vein. According to Bruner the function of this musculo-nervous mechanism is to contribute to the swelling of the cephalic veins and sinuses of the head by blocking the return of blood through the internal jugular vein. In Anolis (Fig. K) these "tumefactor" nerves show^ closest anatomical relations to nerve X, the indirect course described by Bruner lieing unnecessary because here the constrictor muscle lies relatively more caudad than in Lacerta. Notwithstanding the fact that these nerves are given off elsewhere, the ramus communicans X et IX occurs quite constantly (see Plates 2, 3, absent Fig. A'). It is possible, then, that this communicating ramus is an efferent sympathetic path not exclusively related to the constrictor muscles of the jugular vein. After the union of the superior laryngeal ramus of X with that of phar^^Tigeal IX the combined nerve (phx-lar.) joins XII in its course to the ventral side of the pharynx. Beyond this point not all the components of IX and X can be followed and accounted for positively in their distribution. It seems certain, however, that the laryngeal branch (phx-lar/), the first given off from XII after it reaches the ven- tral side (figs. 5, 6), represents a large portion of this nerve, although its smaller size establishes the fact that the trunk of XII still carries some of the fine fibers. By using fine needles in dissection this laryn- geal ramus can be split away from the main trunk of XII and thus it may be demonstrated to represent the larger part of ramus pharyngo- laryngeus, whose union with XII is mentioned above. 1. Relations between Nerves IX and X. Preceding an account of the terminal rami of IX and X, a general- ized summary of the relations of roots and main trunks of these two willard: cranial nerves of anolis carolinensis. 77 nerves will be given. It is based on the details of this nerve complex as worked out in six eases, and gives the features common to all. There is much variation in the details of the connection between nerves IX and X. Without reference, for the time being, to the anas- tomoses of uncertain significance, the essential features of the two nerves may be stated as follows: (a) Nerves IX and X arise each by several separate roots, at least one root of each nerve being motor. The component character of each nerve appears to be the same, although X exceeds IX in the number of both its sensory and motor components. The ganglion of IX (Plates 2, 7, figs. 4, 23) lies some distance from the brain, as already stated, and probably is not strictly homologous with the ganglion of X, the jugular, which is just outside the foramen. (b) The coarser motor fibers in each case are readily seen to pass through their respective ganglia. (c) There is a postganglionic division of each nerve which results in each case in two bundles; a bundle of mixed coarse and fine fibers and a bundle composed exclusively of fine fibers. (d) The mixed bunrlles come together (as pharyngo-laryngeal branch) and then join XII for distribution on the floor of the pharynx. (e) The fine fibered bundle of IX is very small and joins the sym- pathetic trunk, from which it may later separate, along with sympa- thetic elements, to reach palatine VII. (f ) The purely sensory bundle of X is a large one, and passes caudad to its trunk ganglion, ganglion nodosum (gn. nd.) ; it then divides (Fig. L) to form the ramus recurrens X and the ramus visceralis dis- tributed to the lungs, heart and alimentary canal. (g) In all cases studied except one IX and X show an anastomosing ramus (Fig. 4, comn. IX-X) which connects a preganglionic point of IX with a postganglionic point of X. The one exception is shown in Figure K, already referred to, where there is no connection whatever between IX and X proximal to the union of the phar^^^lgeal and lar^m- geal rami. 2. A VESTIGIAL DORSAL GaNGLION ON THE RoOTS OF THE VaGUS. There was found in several cases in Anolis a very small group of ganglion cells situated on the dorsal side of one of the roots of nerve X. Sucla gangla were studied in three cases. In two cases they were situ- ated upon the largest, most posterior root, which carries the motor fibers. In the third case the ganglion, being smaller than in the others, consisted of only three or four cells located on the small middle root. 78 bulletin: museum of comparative zoology. In no case observed did there occur more than one group of such ganglion cells upon the several roots of the same vagus nerve. No such ganglia were found upon IX. The size of these cells shows them to be of the cerebro-spinal rather than the sympathetic kind. This fact is brought out by comparing them with cells of the geniculate ganglion, and also wath those of the spinal ganglion, and contrasting these with the sympathetic cells found at the base of the palatine. Embryological studies by various investigators have demonstrated for both reptiles and mammals transitory root ganglia in this region. These are generally interpreted as the remnants of the lost dorsal roots between the first spinal ganglion and the vagus. Van Wijhe ('86) and Chiarugi ('89) considered these ganglia as contributing permanently to the accessorius part of the vagus. Fiirbringer ('97, p. 502) recognizes the existence of these ganglia in Sauropsida, but states that they later disappear entirely and have nothing to do with the vago-accessorius. In mammals such rudimentary ganglia have been noted in the embryo of the pig (Lewis, :03) and in man (Streeter, :04). In the Amphibia IX, X, and sometimes VII, possess cutane- ous components. In mammals these are reduced to a small bundle, which proceeds from the jugular ganglion as the ramus auricularis, and small clumps of cells may remain among the vagus roots even in the adult, an indication of the more extensive existence of this cuta- neous component. In Anolis no cutaneous fibers v/ere discovered in any of the nerves between V and the third spinal; unless these rudimentary vagus ganglia be ascribed to the cutaneous components, all traces of these -components have been lost in these segments. In the whole group of reptiles this absence of cutaneous rami appears to prevail, as no mention of such nerves is made by any author. If this is the case, the reptiles stand alone in the extent to which this component has been lost. The birds, however, need investigation on this as well as on other points. Cords (:04) has described in birds a cutaneous sensory branch of VII going to the lining of the external auditory meatus, to which she gave the name "ramus auricularis," the same term that is applied to a nerve of similar component character in mammals, but derived from X. If microscopic study should verify Cords's observation, we should have in birds the persistence in VII of a component which is absent from this nerve in practically all other forms above fishes; but the same component would in birds be absent from a nerve (X) which possesses it in all vertebrates except the Sauropsida. It is important in this connection to note willard: cranial nerves of anolis carolinensis. 79 that Cords describes a root ganglion, "ganglion jugulare," for IX in addition to the petrosal ganglion. When this appears in a rudi- mentary form it is called a somatic sensory structure. If it is large enough to be discovered by dissection methods, one might look for a cutaneous ramus associated with it. Is it possible that Cords's ramus auricularis VII has any relation to her "ganglion jugulare" IX? The ^•alue of Fischer's ('52) work lies in the range of his material, which justifies certain deductions, the validity of which might be arrived at directly by microscopic study. For comparison with Anolis some of the more important statements made by Fischer may be considered. This author does not find a root ganglion on IX, al- though the petrosal ganglion is to be recognized in practically all cases. This ganglion is united with nerve X and joined by rami of larger or smaller size, the union with the latter is generally on the proximal end of the ganglion through the combined rami communicantes internus et externus IX ad VII. In Platydactylus, however, the external sympathetic ramus does not join the internal; this leaves, then, only the connection to palatine VII known as Jacobson's anastomosis and consequently no apparent connection with the sympathetic system. In another form {Varanus hmgalensis) the reverse is true, the internal communicating ramus passes IX to join the main sympathetic trunk farther distad. A Jacobson's anastomosis in this case would have to go by way of the external ramus. In this form also no ganglion petrosum was discoverable, although it is very large in another species (J^aranus nUotocus) of the same genus. Neither Watkinson ( :06) nor Osawa ('98) discovered with certainty the petrosal ganglion. This would indicate either a scattered condition of the ganglionic cells along the trunk, or a less developed viscero-sensory component in Varanus and Hatteria than exists in the case of Anolis, in which, though a smaller animal, it was demonstrable by dissection methods. The failure to find the ganglion by this method would not indicate its entire absence. The union of IX and X also shows considerable variation, as does likewise the union of these two, combined or separately, with the main trunk of the hypoglossal. In comparing with all the forms hitherto described, Anolis may be put down as typical in the combina- tion of the main pharyngeal branches of IX, X, and XII into a common trunk, which later separates; but as this union has no significance other than as a common path around the pharjVTix, it is modified wherever there is much variation in shape of head and relative posi- tion of parts innervated. In a few cases XII is wholly free, and in 80 bulletin: museum of comparative zoology. others IX is independent of both X and XII; these conditions have significance in determining the probable source of the terminal branches in such a form as Anolis, where these branches cannot be followed back through the combined trunk; they will be referred to again in the account of the terminal distribution of IX and X. Since nerve X is larger than IX in all the forms described, its ganglia have been more regularly found than those of IX. The trunk gan- glion of X {gn. nd.) is a more constant feature than its root ganglion {gn. rx. X, gn. X), having been described for all species hitherto studied. On the other hand, the root ganglion, such as is found in Anolis, has been definitely identified as an independent ganglion in only three forms. 3. Terminal Distribution of Nerves IX and X. From the foregoing account it is seen that all the components of IX and X (excepting the rami to the jugular vein) are distributed peripherally from three rami. (1) Jacobson's anastomosis, carrying viscero-sensory fibers to VII; (2) Pharyngo-laryngeal, carrying the superior laryngeal branch of X and the pharyngeal branch of IX, both of which include viscero-sensory and viscero-motor fibers; (3) ramus visceralis, carrying viscero-sensory fibers of X. To what extent efferent sympathetic fibers may be carried in any of these rami could not be determined. (1) Jacobson's anastomosis. This term is here used for the com- municating ramus {comn. i.) between IX and VH, which joins the base of the ramus palatinus. As the term is generally employed it is restricted to a connection between these points, which carries a viscero-sensory palatine branch from IX to be distributed with pala- tine VII, and also innervates the mucous surfaces caudad to VII. Because of the development of the sympathetic system of the head, the main trunk of which takes this course, it is difficult to determine to what extent this is a true viscero-sensory branch. In Anolis two points on VII are connected with IX, usually at the distal end of the petrosal ganglion. This connection is often so slight that it is lost sight of in the more evident fact of the direct passage of this sympa- thetic bundle caudad to its ganglion. It is always large enough, however, to give rise to all the meduUated fibers contained in the internal communicating branch, which is the one having the position of Jacobson's anastomosis. Whether it docs give rise to them cannot be stated. It can be stated that if Jacobson's anastomosis (as re- stricted) exists in Anolis, it contains very few viscero-sensory elements. willard: cranial nerves of anolis carolinensis. 81 A comparison with other reptiles would seem to warrant the broad- ening of this statement to a general one applicable to reptiles as a group. There are found in practically all forms of reptiles these two sympathetic rami named by Fischer ('52, p. 30) "ramus communicans internus rami palatini cum glossopharyngeo " and "ramus communi- cans externus nervi facialis cum glossopharyngeo." The first of these, which would have the position of Jacobson's anastomosis, Fischer (p. 30) refers to as one of the finest nerves in saurian anatomy, which could be identified only with great difficulty; at the same time he adds that it belongs to "den bestandigsten Nerven," and for this reason must be considered essential to the plan of the saurian nervous system. Fischer, however, showed one case, before referred to (p. 79), where this internal ramus makes no connection with IX, which sup- ports the view that it is principally a sympathetic ramus from the deeper part of the head to the common cervical trunk. Bender (:06, p. 388) gives to this connection both a sympathetic and viscero-sensory function in Chelonia, and states that the petrosal ganglion is closely bound up with a ganglion of the sympathetic. In Anolis no sympathetic cells were recognized, nor did the petrosal ganglion show any division. Cords ( :04, p. 79) also specifically states that, in the goose, this anastomosing branch from IX to VII is com- posed of fibers from IX and from the sympathetic ganglion. (2) Ramus pharyngo-laryngeus (phx-lar' .) . This nerve is given off from the main trunk of XII, in which it is temporarily carried, at about the posterior end of the genioglossus muscle. Its course is cephalad and mesad between the genioglossus and the cerato-hyoideus to the trachea posterior to the lar^^-nx. As it leaves XII its composi- tion is almost identical with that of the combined IX and X pharyngeal rami before they join XII, although it is somewhat smaller. There are about 20 coarse fibers mingled with the fine ones. In its course across the m. cerato-hyoideus it loses about half of the larger fibers, so that it is found to contain about eight or ten of these fibers distal to its course across that muscle. The remaining coarse fibers supply the muscles of the larynx, the fine fibres being sensory. When the nerve has reached a position just beneath the mucous membrane (Plate 5, fig. 13), some very fine fibered branches are given off from the main trunk (not shown in the figure). They are so twisted about the blood vessels that their final distribution, whether to the epithelium or elsewhere, was not demonstrable. A small ganglion at the base of the fine fibered ramus indicates that a part of the fibers are of sympathetic character. 82 bulletin: museum of comparative zoology. The main trunk passes fon\-ard, gi\dng off two more branches to the mucous membrane (Plate 2, fig. 5) before turning abruptly across the ventral floor of the larynx to form an H-shaped anastomosis with the same nerve of the opposite side (Plate 2, fig. 5, an'stm. lar.). The chiasma is not a complete one; hence each muscle of the larynx is stimulated through motor fibers from both sides of the brain. The muscles of the larynx consist of an outer longitudinal (dilator) and an inner transverse (constrictor) pair, (Plate 4, fig. 11, lar. Ig. and co'st. lar.). They are innervated exclusively from the coarse fibered elements carried in the pharjugo-laryngeus IX and X. Attention is called to the fact that this motor innervation may be from IX, from X, or from both. There is no possibility of determining in Anolis, except experimentally, which nerve has given up its fibers to the cerato- hyoideus muscle. If dissections may be trusted on this point, the condition as described by Watkinson ( ;06) in Varanus would support the view that the fibers innervating the cerato-hyoideus (hyoglossus) muscle are from IX. In Varanus IX does not anastomose with XII and does not appear to combine with X in a manner corresponding to the condition in Anolis. In the former the nerve described as IX innervates the cerato-hyoideus muscle as it crosses the ventral face of that muscle. There are more proximal anastomoses, however, making possible some combination of the fibers of IX and X as in Anolis; so, in the absence of microscopic observation, any conclusion must be tentative. The larjTigeal anastomosis is quite generally mentioned where the innervation of this region is described in higher vertebrates. The ramus recurrens X is usually described as entering into this "sling" and joining in the motor innervation. In Anolis, however, the ramus recurrens takes no part in the sling, although the terminal ramus of this nerve passes through the longitudinal laryngeal muscle (Plate 2, Fig. 5, lar. Ig.) to reach the dorsal free edge of the laryngeal cartilage, where its fibers may be seen turning in to innervate the epithelium of the lar;>Tix (Plate 2, fig. 5, rcr. X). (3) Ramus visceralis (vsc. X). This is the name given the main trunk of the vagus after the superior lar>Tigeal nerve is given off. It is a fine-fibered bundle, which closely resembles the main sympa- thetic, with which it has a parallel course across the dorsal side of the thymus gland. It bears the large trunk ganglion {gn. nd.) in its course and then gives off to the mucous membrane sensory branches, which correspond to the posterior laryngeal, and also the ramus recurrens X. In the series plotted the sections are not carried will.\rd: cranial nerves of anolis carolinensis. 83 back far enough to reach the trunk ganghon. This was clone in an- other series. The best demonstration however, of the course of the vagus to show the trunk ganglion and the postganglionic branches was furnished by a dissection of this region mounted in balsam. As these structures lie underneath all muscles, it was possible to remove the floor of the pharynx in this region and pin it out on cork and then fix it in Vom Rath's fluid. The preparation, which consisted of the mucous membrane, the blood vessels and the nerve trunk with all the branches to the mucous membrane, was cleared and mounted on a glass slide giving a diagrammatic picture of the distribution of the nerves. A drawing (Fig. L) of a portion of such a preparation^- is given Fig. L. — Camera drawing of a portion of a dissection to show trunis: ganglion of the vagus and the relation of ramus recurrens to the arterial arches. This is part of a preparation made in the same manner as that showing dis- tribution of motor VII (figure J). For vcr. X. read rcr. X. to supplement the plotting. It will be seen that no branches are given off proximal to the ganglion, which lies just clear of the thymus gland at its posterior end. The ganglion itself is a pear-shaped struc- ture through the center of which there is a distinct fiber path. At its distal end two very fine rami are given off, one {phx. X) mesally, to the pharynx wall, the other {sy. X.) to the main sympathetic trunk. The vagus nerve crosses the arterial arches on their ventral side and reaches a position alongside the trachea just cephalad of the bronchial division. As it nears the median line the ramus recurrens laryngis {rcr. X.) is given off. It crosses the arterial arches on their dorsal side thus forming the loop. The posterior ramus visceralis proper was not followed farther than is shown in this preparation. It branches 84 bulletin: museum of comparative zoology. almost immediately and the lungs are very richly supplied. The branches to tlie heart did not appear. Ramus recurrens laryngis X (Plate 2, fig. 5; rcr. X.). The course of this ramus is directly cephalad to the larynx, following along the lateral side of trachea to reach it. Numerous very fine branches are given off in its course (not shown in Plate 2, fig. 5). Reaching the larjTix this nerve breaks up, first dividing into the two main terminal divisions shown in the figure. One of these lies on the dorso-lateral side of the larynx and passes through a portion of the longitudinal muscle to reach a more anterior position, where it innervates the mucous mevibrcme just posterior to the glottis. The other branch, the more iKedian one in the figure, innervates the mucous membrane of the veiifJCr-lateral part of the larynx. The first branch in its pas- sage through the muscle becomes closely involved in the motor com- plex, and is separable from it only through the study of sections. If it contributes motor fibers to this, it is only very slightly, and my belief is that in Anolis the ramus recurrens is wholly sensory. Sec- tion 659 (Plate 4, fig. 11) is anterior to the main branches so that no part of the recurrent ramus appears. Some motor twigs are shown in the muscle. More data are needed to homologize the branches of nerves IX and X with those of Amphibia. In the latter the ramus recurrens X in- nervates the muscles of the larynx (Coghill, :02, p. 245 ; Norris, :08, p. 552). Through comparative anatomy Fischer arrived at conclusions which conform with the facts as stated for Anolis. He found the intimate mingling of the terminal twigs of the ramus recurrens X and the pharyngo-laryngeus, so generally described in other forms, to be absent in two cases, so that the distribution of the two nerves was distinct, and in these cases the recurrent ramus is held to be sensory, not motor. " Im den Fallen, wo der R. recurrens sich nicht mit jenem [phar>^lgo-laryngeus] verbindet (Varanus Bengalensis, Platydactylus guttatus) geht dieser [r. recurrens] nicht in die Muskeln, sondern an die Schleimhaute des Kehlkopfes" (Fischer '52, p. 48). But Watkin- son ( :06, p. 467) for another species of Varanus states that the united fibers of the ramus recurrens X and the terminal branches of IX are distributed to the muscles of the larynx. Her observation was not properly supported, however, in regard to either point, as in her species (1) nerve IX previously received fibers from X, as in Anolis, and (2) a mere union of rami as demonstrated in dissection does not in itself prove similarity of distribution. will.\rd: cranial nerves of anolis carolinensis. 85 By the same comparative method Fischer ('52, p. 49) estabhshed the innervation of the cerato-hyoideus from IX and the laryngeal muscles from X. For in two cases (Euprepes schac and Lacerta ocel- lata) IX was found to be free from X, and in these cases its distri- bution was to the cerato-hyoideus muscle and to the phar^Tix wall anterior to the lar\Tix. Nerve X in these cases is a pure superior laryngeal and goes to muscles and mucous surfaces of the larynx (i. e. is of mixed nature, carrying both motor and sensory fibers). This gives ground for the view that a like condition exists in those forms where it cannot be actually demonstrated. Van Bemmelen ('89) considers Fischer's work open to criticism in this particular connection because he did not establish the homol- ogy of the rami by means of their relation to the aortic arches. Both Van Bemmelen and, more recently, Goppert ('99) contradict Fischer's conclusions regarding the sensory nature of ramus recurrens X. Gop- pert, in an article which deals comparatively with the larynx region in Amphibia and reptiles, concludes with this statement (p. 23): "Bei alien Reptilien haben wir also Berechtigung zu der Annahme, dass der Recurrens bis zum Kehlkopf gelangt, trotz des oft weiten, von ihm zuriickzulegenden Weges. Dass er dann aber iiberall die Kehlkopfmuskulatur ^■ersorgt, wird keinem Zweifel unterliegen konnen, nachdem er sich fiir die Lactertilier direkt erweisen Hess." These differences of opinion cannot be attributed to the study of different reptiles, for Goppert makes use of the same genus (Platydac- tylus) as that employed by Fischer. The present results in Anolis, then, do not clear away the uncer- tainties of the general question of larynx innervation; they tend, however, to suggest the probability that both the sujierior laryngeal and the recurrent branch may carry motor fibers, these showing different proportions in the rami of different forms. Q. SPINAL ACCESSORY NERVE. A spinal accessory nerve was not discovered in Anolis. In all the described reptiles a portion of the vago-glossopharyngeal components are grouped as nerve XI, or spinal accessory. Peripherally there are generally two anatomical conditions which warrant this interpreta- tion : (a) the caudad extension of the vagus series of roots beyond the limits of the cranium, and (b) the distribution of a motor ramus from the vagus (distal to the ganglion) to certain of the muscles of the shoulder girdle. Both these features are absent in Anolis. 86 bulletin: museum of comparative zoology. By examination into the component character and central relations of this nerve, where it exists, it is found to be similar to those viscero- motor components of X and IX which innervate striated muscles (Johnston, :06, p. 203), such as the laryngeal muscles. These all have their cells of origin (in the higher vertebrates at least) in a portion of the viscero-motor column ventrally separated from the rest and known as the nucleus ambiguus. While the motor nuclei of this region of the hind brain have been incompletely identified in Anolis, microscopic study has fully demonstrated that IX and X contain a very limited number of components such as innervate the laryngeal muscles and that these all pass ventrally with XII, which gives off no branches that cannot be directly traced to muscles of the ventral head region. It still remains to be demonstrated whether there is a caudally extended motor nucleus ambiguus which contributes fibers to the spinal nerves innervating the muscles originally supplied by the accessorius. In the absence of such a relation, it would then be a question whether the trapezius muscle in Anolis is homologous to the one so named elsewhere. • The apparent absence in Anolis of anything corresponding to the spinal accessory is an anomolous condition in reptiles, because, even making allowance for many misinterpretations depending on gross dissections, the universal mention of such a nerve in the literature bearing on reptilian anatomy would indicate a greater development of the vago-accessorius group than is shown in Anolis. Fischer ('52, p. 62) finds that the condition first described by Bischoff ('32) is realized in all the forms he studied, viz. that from 5 to 9 fine bundles, generally increasing in strength posteriorly, arise along an irregular line extending from the level of the second cervical nerve to the origin of the vagus. All these root bundles assemble into one trunk, which generally fuses with the vagus. In two species of the genus Salvator, however, this trunk remains separate, although it passes out through the same foramen with the vagus. This independent course of the accessorius was also described by Bendz ('43) for Chdonia my das. What Fischer calls the accessorius includes, in addition to the ramus externus, fibers which have a distribution with the laryngeal branch of the vagus, or with the ramus recurrens vagi, or with both. The ramus externus, which by gross methods is the only portion of the accessorius that can be followed to its distribution, was not found by him in all forms. It was absent in Chamaeleo vulgaris and Agama spinosa. No mention was made as to whether there was a correlated willard: cranial nerves of anolis carolinensis. 87 reduction of the posterior roots; if such a reduction occurred in these two species, they would be in correspondence with Anolis. Fischer demonstrated the ramus externus definitely in nine other species of liz- ards and in the crocodile. Bischoff ('32) and Vogt ('39) had described this ramus externus as supplying a small twig to the cervical muscle, but this could not be expected, and was not verified by Fischer. Fiirbringer ('76, p. 649) in his account of the innervation of the shoulder muscles of saurians refers to the part taken in their innerva- tion by the vago-accessorius as though it w^ere a regular feature to be met with in all forms. The ramus externus is distributed, he says, to the ventral half of the capiti-cleido-episternalis (capiti-dorso- clavicularis), where it as a rule anastomoses with the cervical plexus of the anterior spinal nerves. INIore recently Osawa ('98, p. 616) has described in Hatteria a spinal accessory having the typical superficial origin, which extends caudally as far, as the third spinal nerve. This accessory joins X, and distal to the ganglion there is given off a weak ramus internus and a strong ramus externus, the latter going to innervate the muscle capiti-dorso-clavicularis. Schauinsland's (:03, Taf. VIII, fig. 70) observation on the embryo of the same form practically coincides with Osawa's description. There appears to be some confusion in Watkinson's (:06) account of this nerve in Varanus. She says (p. 467) " the ramus externus vagi innervates the muscle sterno-cleido-hyoideus near its origin from the skull. This muscle also receiving innervation from the third cervical nerve, the end fibers forming an intricate sling with those of the ramus externus." Reference to her Plate XII, figs. 10 and 11, shows that the author means " m. capiti-cleido-episternalis." Another ambiguity rests in the use of the term "m. cucullaris" (Plate XII, fig. VIII) for the most superficial neck muscle. This evidently corresponds to the sphincter colli of x\nolis and, like the latter, is innervated by VII. Notwithstanding the great number of papers that have appeared on the morphology of the nervus accessorius, the subject is far from settled, and, in view of the great range of variation in the anatomical relations of roots and peripheral nerves, the necessity of microscopical analysis is obvious. Lubosch's ('98 and '99) extensive review of the subject in vertebrates from the standpoint of comparative anatomy only tends to emphasize the fragmentary nature of our knowledge regarding this region in amniotes. 88 bulletin: museum of comparative zoology. R. HYPOGLOSSAL NERVE. The twelfth cranial nerve is well developed in Anoli.s, in correlation with the condition of the tongue and especially with its muscular tongue papillae. It arises from three distinct roots and, with the exception of the brief union with the pharyngo-laryngeal ramus of IX and X, takes an independent course to the tongue, where the greater part of its fibers are distributed to the intrinsic musculature. A few small bundles are given off to the cerato-mandibularis group as it crosses these muscles. Not all the hypoglossal nerve is distributed to the ventral region, a part of the last (third) root retaining the dorsal and lateral rami of the spinal nerve from which it phylogenetically has arisen. These go to the cervical muscles. Roots of the hypoglossal. Although nerve XII is a combination of roots which emerge from the cranium through three separate foramina, there is no separation of root bundles at their origin into three groups. They form a continuous series along the somatic motor line, and a comparison shows that they may be differently combined to emerge from the cranium. There is, however, one feature which is constant, viz., the presence of a cervical part on the last root of the series, which is marked off from the hypoglossal part by a difference in the size of its fibers, thus showing a correlation between its functions (as indi- cated by its distribution) and the structure of its fibers. Special series of sections prepared for the purpose of tracing these differences to central nuclei have thus far failed, owing to the difficulty of carrying them through the meninges of the brain. Where the root bundles pass through the foramina the fibers are well preserved and here the two kinds of fibers are distinctly segregated and were easily followed into their respective rami. Distribution of the cervical part (Plates 2, 3, figs. 4-6). The cervi- cal part is made up of larger fibers than those forming the hypo- glossal nerve proper. They equal in caliber the motor components of the first and second spinal nerves, and their distribution is similar. The hypoglossal bundle in the posterior root of XII is in every way like the other roots of this nerve which join it to form the main hypo- glossal trunk. The cervical portion, having a dorsal position in the common root as it emerges from the foramen, divides at once into a lateral {crv. I. XII) and a dorsal {crv. d. XII) ramus. The lateral ramus is the larger and is distributed to the spinalis colli muscle (Plate 7, fig. 23, spi. coll.). The dorsal ramus passes caudad on the willard: cranial nerves of anolis carolinensis. oy dorsal side of this muscle to join the dorsal ramus of the first spinal nerve, with which its terminal fibers become mingled (Plate 3, fig. 6, spi. d. 1). It is anticipated that, upon careful study of the brain with reference to the nucleus of XII, some difference in origin of these two bundles {orv. I. XII and crv. d. XII) as compared with XII proper, will be demonstrable. In its course around the pharynx to reach the ventral side of the neck XII crosses dorsad of the thymus gland (Plate 7, fig. 24, gl. thy.), and also of the visceral ramus of X and the sympathetic trunk, which lie close together on the dorsal side of this gland (Plates 2, 3, figs. 4, 6, and 7, vsc. X and sy.) . XII, turning ventrally, passes between the thymus gland and the jugular vein mesad of all muscles. Section 1595 (Plate 7, fig. 24) falls in a plane just caudad of XII so that no part of it appears; the other structures referred to show well in this section. Section 1480 (Plate 7, fig. 23) shows XII after it has attained its ventral position. This is reached by crossing the end of the first cerato-branchial on its lateral side. Nerve XII then takes a direction cephalad (Plate 7, figs. 22, 23) between the cerato-hyal (ker-hy.) and the first cerato-branchial {ker-brn. I), on the ventral side of the cerato-hyoideus (hyo-glossus) muscle. Its direction from this point onward is cephalad and mesad along the lateral edge of the genio- glossus muscle and the median side of the cerato-hyal. As it passes forward it comes to lie on the ventral side of the genio-glossus (gen-gls.), where it parallels the mandible to a point as far forward as the basi- hyal (Plate 6, figs. 16-18). Here we find a division of the main trunk into two rami of about equal size. The median one immediately divides, so we then have a three-fold division (Plate 5, fig. 12, big. I. XII, lug. i'm. XII, and Ing. m. XII) of the main trunk, the lateral one containing about half the fibers. From this point forward these three main rami diverge and pass up into the tongue musculature for distribution (Plate 2, fig. 5). This course of the main trunk and termi- nal divisions of XII has been described without reference as yet to certain small rami which it gives off. The branches of XII will now be described in more detail: Distribution of the hypoglossal nerve, (a) Ramus pharyiigo-laryngeus {phx-lar'.). Although r. pharyngo-larvTigeus is given off as a branch of XII, the principal, if not the sole, source of its fibers is from nerves IX and X, as described in connection with the account of those nerves. (6). Between (a) and the main divisions of XII there are given off several very small rami, which inner\ate the muscles between which the hypoglossal nerve passes. The first of these (Plate 2, fig. 5, XII ^) supplies the cerato-mandibularis 1 (Fig. 18) in its posterior 90 bulletin: museum of comparative zoology. region. The second {XII ^) goes to the small slip of muscle described as the cerato-mandibularis 3 (Plate 6, figs. 16-19, kcr-md.^) sending also a few fibers to the same muscle that receives the first twig. Some distance cephalad a third ramus (XII ^) is given off, which in- nervates this first muscle, the most of the bundle, however, passing far forward to reach the portion described as cerato-mandibularis 2 (Plate 2, fig. 5; Plates 4, 5, figs. 9-12, ker-md.^). Just anterior to this several small branches (Plate 2, fig. 5, XII ■*) supply the posterior portion of the genioglossus. It is noticeable that all these small rami to the more superficial tongue muscles draw off from XII the largest and most strongly medullated of its fibers, although not exclusively fibers of this kind. There is no further distribution of fibers of XII from the main trunk, which now may be said to supply the tongue proper through the three main divisions referred to above. For convenience in description these will be described as median, intermediate and lateral rami. Ramus lingualis medialis XII (big. m. XII). This ramus crosses the ventral side of the main longitudinal tongue muscle, genioglossus, to reach the median edge of this muscle. Here, on either side of the glossohyal (gls-hy.) and underlying the larynx, begins a mass of muscle composed of short vertical fibers (Plate 4, fig. 11, lug. vrt.). This forms a continuous vertical muscle surrounding the glossohyal as far forward as its anterior end, which is well toward the tip of the tongue. The median ramus of XII runs the length of this muscle supplying it on the way (Plate 2, fig. 5). As the vertical fibers gradually run out, this nerve also dwindles. The fibers innervating this muscle are less heavily medullated than those of the small rami previously described. Ramus Imgiialis intermedius XII (Plate 2, fig. 5, Ing. i'm. XII). This ramus is a little larger than the median one and the fibers are larger and more heavily myelinated. It runs cephalad first on the sur- face of the muscle genioglossus then within that muscle. This ramus appears to supply the genioglossus exclusively. Its course may be followed in the drawings of cross sections (Plate 4, 5, figs. 9-12). Ramus lingualis lateralis XII (Ing. I. XII). This is the largest division of XII and the one which forms the anastomoses with the lingual branch of V (Plate 2, fig. 5, lug. V) ; before this union, however, it divides into two rami of about ecjual size (Ing. I. XII ^ and Ing. I. XII -) giving off just proximal to the division a small ramus (XII *) supplying the longitudinal tongue muscles. Of the two main divisions one is distributed at once in a series of branches to the transverse musculature (Plate 4, figs. 10, 11, lug. t.) on the upper surface of the tongue, the nerve fibers reaching this muscle at its extreme lateral willard: cranial nerves of anolis carolinensis. 91 margin by passing around the genioglossus muscle (Plate 4, fig. 10, Ing. I. XII'). The other division (Ing. I. XI P) of the lateralis XII combines with lingual V and the two mingled components are dis- tributed together (Plate 2, fig. 5). The dorsal musculature of the tongue begins here to take on the crossed arrangement of its fibres (p. 35), and this seems to be related to the presence of papillae on the surface characteristic of the anterior end of the tongue (Plate 4, fig. 9). In the distribution of the three main divisions of XII, it is important to note how each is quite definitely Umited to a particular part of the musculature, and that the part of XII which is mingled with V inner- vates only the strictly intrinsic musculature, for the most part those short muscle fibers which are inserted into the mucous membrane of the papillae-bearing dorsal surface. These papillae deserve special stucly l:)ecause of certain peculiari- ties in their finer structure. These conditions were in part brought out in the unstained material. The muscle fibers, which show a high degree of dift'erentiation between the light and dark transverse bands, extend all the way to the tip of the papillae, and meduUated nerve fibers could also be followed the same distance. Inasmuch as these somatic motor and general cutaneous fillers are approximately of the same size, it is not possible to distinguish between them by this criterion. It is fair to assume, however, that the fibers at the ends of the papillae are sensory, for, from what we know of the innervation of striated muscle, it is not to be expected that the muscle fibers are innervated at their scattered distal ends rather than at a more proximal point, where they are closer together and nearer the source of nerve supply. We know also that nerve fibers of the general cutaneous type have come into this region in the lingual branch of ramus maxil- laris V, and presumably they must reach the surface. No taste buds were found among these papillae, although at the sides of the tongue such buds were found among the tubular glands.^ In a region where these papillae are best developed they are flattened at the end, where the epithelium is of the stratified columnar type. The sides are cov- ered with simple glandular epithelium. The flattened ends show many cells extending out as though protruding individually. The free ends, being knob-like, contain the nuclei, while the base is attenu- ated into a slender column. The result is a sort of tuft of knobbed projections forming the end of each papilla. It is to this flattened terminal portion that the muscle and nerve fibers pass. ' Later study on the histology of the tongue has disclosed a few papillae bearing single taste buds in their flattened ends. 92 bulletin: museum of comparative zoology. S. SUMMARY ON THE DISTRIBUTION OF NERVES IX, X AND XI I. These nerves contain viscero-sensory (red), viscero-motor (dark blue), and somatic motor (pale blue) elements. The viscero-sensory (red) are contained in IX and X, and reach their end organs by two nerves, (a) the pharyngo-laryngeal branch, coming off from XII, and (b) the visceral and recurrent rami of X. The sensory elements of the former appear to be distributed to the mucous membrane of the pharynx lateral and anterior to the glottis; those of the latter (not including those in the posterior visceral ramus) to the pharynx wall posterior to (a) and, through the ramus recurrens X, to the tracheal and laryngeal epithelium itself. The viscero-motor elements (dark blue) are carried in the pharyngo- laryngeal ramus of XII (excepting the fine motor fibers to the m. con- strictor jugulae) and may reasonably be assumed to be the same as those found in the coarse fibered roots of IX and X. They first give off fibers to the cerato-hyoideus muscle, then, after losing the sensory elements to the pharyngeal wall, they form a partial crossing and each has a bilateral distribution to the laryngeal muscles. The somatic motor components (pale blue) comprise the three roots of XII. These differ from the somatic motor of the spinal nerves in presenting collectively a much smaller-fibered nerve. A marked exception exists however, in the third root, which possesses a cervical part and a hypoglossal part, the two standing in contrast to each other in size of fibers. The cervical part is distributed to the dorso-lateral longitudinal neck muscle. The hypoglossal part, combining with all the components of the other two roots to form the main trunk of XII, innervates all the muscles of the ventral longitudinal series except those extending from the girdle to the hyoid apparatus and the cerato-hyoideus. T. ADDITIONAL INNERVATION TO THE VENTRAL HEAD REGION. The other muscles of the ventral head region are supplied by V and VII, thus indicating their visceral origin. Nerve VII innervates the digastric and superficial constrictors, i. c, sphincter colli and mylo- hyoideus in part; V innervates the mandibular series and all but the willard: cranial nerves of anolis carolinensis. 93 posterior part of mylo-hyoideus. The mucous membrane covering the tongue and along the inner side of the lower jaw is innervated by somatic sensory and viscero-sensory components, these being mingled in the same rami, the former by way of the lingual branch of mandibu- lar V, the latter through the chorda tympani from VII. The somatic sensory elements appear to be especially well distributed to the long papillae in the glandular subterminal region of the tongue, while the viscero-sensory are associated with regions bearing taste buds and glands along the sides of the tongue and the jaw. U. SPINAL NERVES. Inasmuch as the anterior spinal nerves have undergone modifica- tion of their typical character due to the same process of cephaliza- tion which has affected the cranial nerves, an account of the first three is included in this paper. The third cervical is the first one that possesses both somatic-motor, and somatic-sensory components typi- cal of a complete spinal nerve. 1. The third spinal nerve, as might be inferred from the last sentence, is the most anterior spinal nerve to have both ventral and dorsal roots. The two roots and the spinal ganglion lie in nearly the same transverse plane (Plate 7, fig. 24). The dorsal root enters the spinal cord on its dorso-lateral side as one compact strand, coming, within the vertebral canal, from the spinal ganglion, which lies in the large intervertebral foramen at the level of the ventral side of the spinal cord. This foramen is between the second and third cervical verte- brae. It is impossible to analyze these roots further than to indicate the position of their somatic components. The sections do not show visceral components in the distribution of peripheral branches, but since the muscles tend to obstruct the fixation and blackening of the nerve fibers, it is possible that some have escaped observation. The sensory bundles emerge from the ganglion as parts of dorsal and ventral rami, the ventral ramus being about double the diameter of the dorsal. The ventral root, composed of somatic-motor fibers, arises as a number of rootlets passing out from the ventral horn of the cord. This root, while on the median side of the ganglion, splits into dorsal and ventral divisions (Plates 2, 3, figs. 4, 6). The ventral immediately joins the ventral sensory ramus, while the dorsal again divides, dorsal to the ganglion, into a part which joins the dorsal sejisory ramus and a 94 bulletin: museltvi of comparative zoology. lateral part {spi. I. 3), which divides at once into several small rami, innervating the muscles immediately adjacent and caudad to it. The dorsal ramus {spi. d. 3) divides into two branches; the one, wholly motor (Plate 2, fig. 6), passes dorso-mesad to supply the muscle lying against the neural arch; the other, a mixed sensory-motor, re- ceives a communicating branch from the dorsal ramus of the second spinal nerve and innervates the most dorsal portion of the longissimus muscle anterior to the general position of the nerve as a whole (Plates 2, 3, figs. 4, 6). The cutaneous part passes through this muscle to the skin, where it divides to be distributed both dorsally and laterally; but like the motor part its field of distribution is chiefly anterior to the nerve. The ventral ramus is a combination of the ventral divisions of the motor and sensory components. The mixed branch thus formed passes ventro-laterad between the m. longissimus colli and spinalis colli (Plate 7, fig. 24, spi. v.S). It crosses on the dorsal side of ramus visceralis X and the sympathetic trunk (Plate 2, fig. 4). Between the muscles above mentioned it receives a reinforcement of motor fibers from the ventral ramus of the second spinal nerve (Plate 7, fig. 24, spi. v. 2), about half of whose fibers join the third spinal for distribution, the rest continuing caudad. This combined ^•entral ramus now passes laterad into m. capiti-cleido-episternalis. Here the motor elements leave the main ramus to supply this muscle in both caudal and rostral directions. (The motor fibers appearing in m. depressor mandibulae adjacent to this region are found by dissection to come from ramus hyoideus VII). The cutaneous components also divide into two branches (Plate 2, fig. 4). One passes ventrad between mm. cucullaris and depressor mandibulae, the other passes directly through the latter muscle to a position just lateral to the main trunk of nerve XII, where it turns cephalad to be distributed to the integument along the ventro-lateral region of the neck and throat (Plates 3, 7, figs. 6, 23, spi. v. 3). A variation in the ganglion of the third spinal nerve deserves men- tion in this connection. This is indicated in the plotting and consists of a small group of ganglion cells on the dorsal sensory ramus not far from the main ganglion. In the labelling a dotted line runs to this as well as to the main ganglion from the letters {gn. spi. 3). In other series of sections of Anolis this small ganglion did not appear. The size of its cells and its nerve connections do not suggest for it a sym- pathetic function, the cells being in every way similar to spinal gan- glion cells, their position probably being due to a migration of some of these cells from the main ganglion. willard: cranial nerves of anolis carolinensis. 95 2. Second spinal nerve. The ventral root of the second spinal, which constitutes the entire nerve, is similar to that of the 3rd spinal, being composed of coarse somatic-motor fibers arising from several bundles that have their origin in the ventral-horn cells. The 2nd spinal is somewhat smaller than the 1st. The rootlets in each case combine to form a compact bundle, which passes out through the intervertebral foramen and immediately divides into a large dorsal ramus (Plates 2, 3, spi.d.2) and a smaller ventral ramus {spi.v.2). A lateral ramus {spi. 1.2) is given off as a branch of the dorsal ramus. The ventral ramus passes ventrad between the neck muscles, where it turns abruptly caudad to cross the ventral ramus of the 3rd spinal on its median side, giving up a portion of its fibers to that nerve, as already mentioned. The rest continues caudad to be distributed to the ventral neck muscles in the region of the 4th spinal nerve. The dorsal ramus of the 2nd spinal is a large branch, which supplies the dorsal and lateral neck musculature. 3. First spinal nerve. This nerve (Plate 7, fig. 23) has the largest number of somatic motor fibers of any yet described. It supplies the dorsal muscles with a large dorsal ramus, as does the 2nd spinal, and, in addition, sends a good sized branch to the ventral side to innervate the omo-hyoideus and sterno-hyoideus muscles. The large root trunk passes out between the cranium and the first vertebra. Just outside the foramen it divides into ventral and dorsal rami. The ventral ramus (Plates 2, 3, figs. 4, 6, spi. v.l) passes ventrally along the body of the vertebra and gives off one branch supplying the ventral cervical musculature, and another that passes between the longus colli and more superficial muscles. The latter then turns slightly dorsad to pass on the dorsal side of X and the sympathetic nerve, attaining a position alongside the third root of XII, with which it is sometimes loosely bound. From here it passes ventrally (Plates 2, 3, figs. 5, 6, omo-hy. and stn-hy.) between the omo-hyoideus and sterno-hyoideus muscles, both of which it richly supplies. The dorsal ramus {spi. d.l). This splits into three branches (Plate 3, fig. 6), the lateral (Plate 7, fig. 23, spi. l.l) is distributed immediately to the adjacent muscle, another to the ventral part of the longissimus muscle, and the largest to the dorsal neck muscle, a portion being dis- tributed at once to the muscle lying alongside the vertebra (Plate 7, fig. 23), while the rest passes into the most dorsal division of the longissimus, where it divides into four small terminal branches, two passing cephalad and two caudad. 4. Connections of first three spinal nerves with sympathetic. While 96 bulletin: museum of comparative zoology. dissections indicated a connection with the median sympathetic trunk, the study of sections shows this to be only a connective-tissue union, the sympathetic bundle passing directly across without any break in its sheath that would indicate the passage of nerve fibers. V. GENERAL CONSIDERATIONS ON NERVE XII AND THE SPINAL NERVES. From a comparative standpoint the hypoglossal nerve in Anolis presents two questions for consideration: First, as to the number of spinal nerves that enter into its formation and their position in the series of spino-occipital nerves, and, secondly, as to the differentiation of its component neurons from the typical somatic-motor type from which they come. The first question appears to be answered in part by the obvious facts presented by Anolis itself, through the persistence of three distinct occipital foramina, indicating three separate seg- mental nerves. Only rarely have three cranial roots for XII been described in the adult saurian. Among all those described by Fischer ('52, p. 66), three roots are mentioned for only Platydactylus; but in Anolis not only are there three roots, but they emerge through separate foramina. In two species he finds only one root, and in seven he finds two. In all cases XII either unites with the first spinal or receives branches from it. Fischer states that, as a rule, the first two spinal nerves are without dorsal roots and ganglia, although some- times there occurs a weak dorsal root on the second spinal nerve. As in Anolis, the third is a nerve well developed in both its motor and sensory components. The relation of XII to the spinal nerves varies according to the strength and number of its roots. This fact points to the correctness of Fischer's view that the cranial part of XII does not represent the same number of spinal nerves in all lizards. Fiir- bringer ('97, p. 501) arrives at the constant number of three roots for all sauropsida, although the first and second emerge through a common foramen in most reptiles. Reference to Fiirbringer's table (p. 546) shows his conclusion regarding the homologies of these nerves. He designates them, the first three, as occipito-spinal nerves, their position being fixed through discovery in the embryo of the older occipital nerves (anterior to these), which have a transitory existence. There are two features in the twelfth nerve of Anolis which seem to show that it is less completely incorporated into the head than in other rep- willard: cranial nerves of anolis carolinensis. 97 tiles: (1) the persistence of three occipital foramina, and (2) the mixed spinal and hypoglossal character of its third root. The latter condi- tion is not described for any other lizard. There is practically no union of the hypoglossal part of the last root of XII with the first spinal nerve, a condition which is described by Fischer as general. If we imagine the cephalization process to progress further in Anolis, we should expect roots one and two of XII to merge with each other, the spinal or cervical part of root three to disappear and the first spinal nerve to be drawn more into the field occupied by it. This is practi- cally what is represented (Fischer) in those forms where but two roots have been described. Evidence from the embryological side (Van Wijhe, '86, Van Bemmelen, '89, Hoffmann, '79-90) supports Fiir- liringer's ('97) generalization that the hypoglossal of reptiles repre- sents three ventral spinal roots. The cervical plexus is represented in Anolis by the combination of XII with the first spinal and the commissure between the second and third spinal nerves. The ventral ramus of the first spinal is but loosely associated, sometimes not at all, with XII. The only constant connection, then, is that of the dorsal ramus of XII (cervicalis dorsalis XII) with the same ramus of the first spinal (Fig. 6, crv. d. XII). This anastomosis may be considered a remnant of the closer rela- tion of these nerves ^A-hich existed before the rise of the tongue mus- culature. Anolis agrees with the typical condition of reptiles in the absence of a sensory component in the first and second spinal nerves. Fischer's reference to the exceptional occurrence of a dorsal root for the second spinal is not carried farther in his descriptions. Rabl-Riickhard ('78, p. 342) states that in the alligator the third is the first of the spinal nerves to possess a sensory part, but in contrast to the lizards (Anolis) the third and fourth spinal nerves also have (Fischer) greatly reduced dorsal roots, indicating a less sensitive integument in the alligator. It will be seen that in Anolis the field innervated by the sensory com- ponents of the third spinal nerve extends far cephalad both on the dorsal and ventral sides, thus demanding a strong dorsal root. In regard to the second point in the comparison, i.e., the differentia- tion of the nerve itself, we have to deal with a histological problem which cannot be profitably discussed without a complete knowledge of the histological elements involved. The fixation of my material is not uniform enough in all parts to admit of a detailed comparison of the caliber of the medullated fibers as found in different nerves. How- ever, in Anolis the difference in the fibers of XII, as compared with 98 bulletin: museum of comparative zoology. those of the ventral roots of the spinal nerves, is further emphasized in the third hypoglossal root and points to a very direct correlation between the size of the fiber and its function. In this case there is no doubt that the larger fibers run the shorter distance, thus contradicting Schwalbe's ('82) law and confirming Dunn's ( :02, p. 323) results in the spinal nerves of the frog. There may be, and probably are, other factors entering into this particular case. Koch ('88) and Brandis ('93) speak of a differentiation in the central nucleus of XII in the medulla of birds, where a " dorsal" portion is separated from the rest; and the suggestion is made by them that this may be correlated with the syringeal musculature. In Anolis any relation of XII to central nuclei bearing on peripheral differences must await a detailed study of the brain. W. GENERAL SUM^vIARY. 1. Anolis possesses the cranial nerves typical of the amniote verte- brate with one exception; there was not discoverable any representa- tive of the spinal accessory nerve described in other reptiles, and the muscles innervated by this nerve in other forms seemed to be supplied in Anolis wholly from spinal nerves posterior to the second cervical. 2. The ganglia of cranial nerves V, YII, IX, and X are distinct from one another and the roots of all issue from the cranium through independent foramina. The ophthalmic ganglion also shows no fusion with the other portion of the Gasserian ganglion. 3. There is a wide distribution of sympathetic ganglion cells along the afferent rami of the cranial nerves. These form definite ganglia on palatine VII (palatine ganglion) on palatine VII and nasalis V (ethmoidal ganglion), on maxillaris V (infraorbital ganglion), and on mandibular V (mandibular ganglion). The topographical facts would lead one to associate the development of these ganglia with specialization of the glands of the head. No medullated nerve fibers were found passing through the connective tissue surrounding these glands. The presence of smooth muscle fibers in the head region might also affect the development of the sympathetic. The sympa- thetic system of the head in the matter of the arrangement of rami and ganglia (as worked out incidentally to the study of the cranial nerves), when compared with other described forms of reptiles, points to the existence of a typical sauropsidan type of quite constant character. 4. The nerve components (excepting the sympathetic) reach their willard: cranial nerves of anolis carolinensis. 99 end organs, or peripheral terminations, through the following nerve trunks; Somatic sensory (yellow), via nerve V, over ophthalmic (rmm. frontalis and nasalis), maxillary and manbibular rami. Somatic motor (light blue), via nerves III, IV, VI, and XII. Viscerosensory (red), via nerve VII over the palatine ramus and the chorda tympani; via nerve IX over the pharyngeal ramus and probably Jacobson's anastomosis; via nerve X over the superior laryngeal and recurrent rami. Viscero-motor (dark blue), via nerve V by a number of inde- pendent rami and over the mandibular ramus; via nerve VII over hyomandibular division and ramus hyoideus; via nerve IX over the pharyngeal ramus; and via nerve X over the superior laryngeal ramus. (a) This shows a greater reduction of the somatic sensory (as indicated by peripheral paths) in iVnolis than is found in the described forms of other groups, such components not being found in nerves IX or X of Anolis although their presence in the same nerves has been reported in each of the other classes of vertebrates. (b) Vestigial ganglia exist in a variable manner on the intra- cranial roots of X, which may be somatic sensory in their origin. 5. The morphological character of the fibers of different com- ponents is sufficiently differentiated to form types peculiar to each component. But the distinction in character appeared to be less than that described for the lower groups of vertebrates. However, there was considerable individual variation in the size of fibers. Nerve XII shows a marked difference in the size of the fibers going to neck muscles and those going to tongue muscles. In this case the smaller fibers have much the longer course. In at least three instances striated muscle fibers of visceral origin are innervated by nerve fibers of smaller caliber and lighter myelin sheaths than is characteristic of the other viscero-motor components of V, VII, IX, and X. These are the ciliary muscle, the protrusor oculi, and the constrictor of the jugular vein, all of which are more closely associated with visceral functions than the other striated visceral muscles. 6. The skin is well-supplied with special tactile organs, which are more abundant along the jaws than elsewhere. These organs are quite generally, if not always, covered by a thinned plate of the horny * layer of the epidermis, which bears in its center a tapering "hair." The innervation of these hairs was not determined beyond the fact of the proximity of the strongly myelinated cutaneous fibers in the dermis beneath. 7. The distribution of taste buds is such as to preclude their innervation (save a very limited number in the laryngeal region) by 100 bulletin: museltm of comparative zoology. anything except the chorda tympani and palatine VII. A large proportion of the fibers carried by these rami are for such sense organs, their innervation fields being covered for general sensory purposes by the somatic sensory of V. 8. Anolis presents a well-balanced form for the study of the rep- tilian nervous system. It is an active, responsive animal with well- differentiated muscles and sense organs, yet presenting no excessively specialized features. It is small enough readily to be sectioned and large enough for experimental operations, and it is suggested that degeneration and stimulation experiments on this form would ad- vance our knowledge of the reptilian nervous organs even more than similar anatomical work on other forms. The anatomical work already done, however, should be supplemented by the proper tech- nique to determine the final nerve terminations. willard: cranial nerves of anolis carolinensis. 101 BIBLIOGRAPHY. Aliis, Jun., E. P. '98. The Homologies of the occipital and first spinal Nerves of Amia and Teleosts. Zool. Bull., 2, p. 83-97. Andersson, O. A. '92. Zur Kenntniss des sympathischen Nervensystems der urodelen Amphibien. Zool. Jahrb., Abt. f. Anat., 5, p. 184-210, taf. 15-18. Bayliss, W. M. :00. The Presence of efferent Vaso-dilator Fibres in posterior Roots. Journ. Physiol, 25 (Proc. Physiol. Soc), p. xiii-xiv. Bender, O. :06. Die Schleimhautnerven des Facialis, Glosspharyngeus und Vagus. Semon's Reiseberichte, 4, p. 341-454, taf. 1-9, 22 text fig. .4^50 Jena. Denkschr., 7. Bendz, H. '43. Bidrag til den Sammenlignende Anatomie af Nervus glossopharyn- geus, vagus accessorius Willisii og hyoglossus hos Reptiherne. K. Danske Viden.sk. Selsk. Naturv. og Math., Afl. Deel 10, p. 113-152, tab. 1-10. Beraneck, E. '84. Recherches sur le developpement des nerfs craniens chez les Lezards. Recueil Zool. Suisse, ser. 1, 1, p. 519-603, pi. 27-30. Also Diss., Geneve, 1884. BischoflE, T. L. W. '32. Nervi .accessorii WilUsii anatomia et physiologia. 4 maj., 6 pp., Darmstadii, Commentatio. Acced. tabb. VI. 1832. Bojanus, L. H. '19. Anatome testudinis europeae. Vilnae, 1819-21, vi + [viii] + 179 pp., 31 tav. Bonsdorff, E. J. '52. Symbolae ad anatomiam comparatum nervorum animalium verte- bratorium. I. Nervi cerebrales Corvi cornicis (Linn.). Acta. Soc. Scient. Fennicae, 3, fasc. 2, p. 505-569, pi. 6, 6a, 7, 7a. Born, G. '76. Ueber die Nasenhohlen und den Tlu-anennasengang der Amphibien, Morph. Jahi-b., 2, heft 4, p. 577-646, taf. 39-41. Bowers, Mary A. :00. Peripheral Distribution of the cranial Nerves of Spelerpes bilinea- tus. Proc. Amer. Acad. Arts Sci., 36, p. 179-193, 2 pis. 102 bulletin: museum of comparative zoology. Bradley, O. C. :03. The Muscles of Mastication and the Movements of the Skull in Lacertilia. Zool. Jahrb., Abt. f . Anat., 18, p. 475-488, taf. 44. Brandis, F. '93. Untersuchungen iiber das Gehirn der Vogel. II. Ursprung der Nerven der Medulla oblongata. Arch. f. mikr. Anat., 41, p. 623-649, taf. 35. Bruner, H. L. :07. On the cephalic Veins and Sinuses of Reptiles, with Description of a Mechanism for raising the venous Blood-Pressure in the Head. Amer. Journ. Anat., 7, p. 4-118, 3 pis. Busch, C. H. '98, Beitrag zur Kenntniss der Gaumenbildung bei den Reptilien. Zool. Jahrb., Abt. f. Anat., 11, p. 441-500, taf. 34-40. Carpenter, F. W. :06. The Development of the Oculomotor Nerve, the ciliary Ganglion, and the abducent Nerve in the Chick. Bull. M. C. Z., 48, p. 139-230, 7 pis. Cartier, O. '72. Studien iiber den feineren Bau der Haut bei den Reptilien. I. Abth. Die Epidermis der Geckoiden. Verhandl. d. phys-med. Gesellsch. Wiirzburg, n. f., 3, p. 281-295, 2 taf. Cartier, O. '74. Same. II Abth. Ueber die Wachsthumserscheinungen der Ober- haut von Schlangen und Eidechsen bei der Hautung. Ibid., 5, p. 192- 211, 1 taf. Also in Arb. a. d. Zool.-Zoot. Inst., Wiirzburg, 1, p. 83-96; p. 239-258, taf. 3, 4, 12, 1874. Chiarugi, G. '89. Lo sviluppo dei nervi vago, accessorio, ipoglosso e primi cervical! nei Sauropsidi e nei Mammiferi. Atti Soc. Toscana d. Sc. Nat., Mem. 10, p. 149-245, tav. 11, 12. Coghill, G. E. :02. The cranial Nerves of the Amblystoma tigrinum. Journ. Comp. Neur., 12, p. 205-289, pis. 15-16. Cole, F. J. '98. Observations on the Structure and Morphology of the cranial Nerves and lateral Sense Organs of Fishes; with special Reference to the Genus Gadus. Trans. Linn. Soc. London, ser. 2, Zool., 7, p. 115- 221, pis. 21-23. Cope, E. D. :00. The CrocodiUans, Lizards, and Snakes of North America. Rept. U. S. Nat. Mus., 1898, p. 153-1270, 36 pi., 347 text fig. Cords, Elisabeth. :04. Beitrage zur Lehre vom Kopfnervensystem der Vogel. Anat. Hefte, 26, p. 49-100, taf. 1-4. willard: cranial nerves of anolis carolinensis. 103 Corning, H. K. '95. Uber die Entwickelung der Zungenmusculatur bei Reptilien. Verhandl. d. Anat. Gesellsch., Versamml. 9, in Basel, p. 165-175. Dunn, Elizabeth H. :02. On the Number and on the Relation between Diameter and Dis- tribution of the Nerve Fibers innervating the Leg of the Frog, Rana virescens brachycephala, Cope. Journ. Comp. Neur., 12, p. 297-328. Emmel, V. E. :04. The Relation of the Chorda Tympani to the visceral Arches in Microtus. Journ. Comp. Neurol, and Psychol., 14, p. 411-417. Fischer, J. G. '52. Die Gehirnnerven der Saurier anatomisch untersucht. Abhandl. Naturw. Verein in Hamburg, Abt. 2, p. 109-212, 3 taf. Froriep, A., und Beck, W. '95. Ueber das ^"orkommen dorsaler Hypoglossuswurzeln mit Ganglion, in der Reihe der Siiugetiere. Anat. Anz., 10, p. 688-696. Fiirbringer, M. '76. Zur vergleichenden Anatomic der Schultermuskeln. III. Morph. Jahrb., 1, p. 636-816, taf. 23-27. Fiirbringer, M. '97. Ueber die spino-occipitalen Nerven der Selachier und Holocephalen und ihre vergleichende Morphologie. Festschr. Gegenbaur, 3, p. 349- 788, 8 taf. Gaskell, W. H. '86. On the Structure, Distribution and Function of the Nerves which innervate the visceral and vascular Systems. Journ. Physiol., 7, p. 1-80, pis. 1-4. Gaskell, W. H. '89. On the Relation between the Structure, Function, Distribution and Origin of the cranial Nerves; together with a Theory of the Origin of the nervous System of Vertebrata. Journ. Physiol., 10, p. 153-211, pis. 16-20, 3 text fig. Gaupp, E. '88. Anatomische Untersuchungen iiber die Nervenversorgung der Mund- und Nasenhohlendriisen der Wirbelthiere. Morph. Jahrb., 14, p. 436- 489, taf. 19. Goppert, E. '99. Der Kehlkopf der Am phibien und Reptilien. Theil. II. Reptihen. Morph. Jahrb., 28, p. 1-27, taf. 1, 2. Giinther, A. '67. Contributions to the Anatomy of Hatteria (Rhynchocephalus Owen). Phil. Trans. Roy. Soc. Lond., 157, p. 595-629, pis. 26-28. Herrick, C. J. '94. The cranial Nerves of Amblystoma punctatum. Journ. Comp. Neurol., 4, p. 193-207, pis. 19, 20. 104 bulletin: museuini of comparative zoology. Herrick, C. J. '99. The cranial and first spinal Nerves of Menidia; A Contribution upon the Nerve Components of bony Fishes. Journ. Comp. Neurol., 9, p. 153-455, pis. 14-20. Hoffmann, C. K. '79-90. In Bronn's Klassen und Ordnungen des Thier-Reichs, 6, Abt. 3, ReptiUen. 2089 pp., 170 taf. Hoffmann, C. K. '89. tjber die Metamerie des Nachhii-ns und Hraterhirns, und ihre Beziehung zu den segmentalen Kopfnerven bei Reptilienembryonen. Zool. Anz., 12, p. 337-339. Holl, M. '76. Beobachtungen iiber die Anastomosen des Nervus hypoglossus. Zeitschr. f. Anat. u. Entw., 2, p. 82-97, taf. 6. Holl, M. '78. tJber den Nervus accessorius Willisii. Arch. f. Anat. u. Physiol., Anat. Abth., p. 491-518, 4 fig. auf taf. 18. Huber, G. C. '99. A Contribution on the Minute Anatomy of the sympathetic Ganglia of the different Classes of Vertebrates. Journ. Morph., 16, p. 27-90, 3 figs., pis. 3-5. Johnston, J. B. '98. Hind Brain and cranial Nerves of Acipenser. Anat. Anz., 14, p. 580-602. Johnston, J. B. :06. The nervous System of Vertebrates. Philadelphia, 1906, xx + 370 pp. Kathariner, L. :00. Die Mechanik des Bisses der solenoglyphen Giftschlangen. Biol. Centralbl., 20, p. 45-53. Klein, E. '71. On the peripheral Distribution of non-meduUated Nerve-fibres. Quart. Journ. Micr. Sci., 11, p. 405-416, pis. 19, 20. Koch, P. D. '88. Untersuchungen uber den Ursprung und die Verbindungen des Nervus hypoglossus in der Medulla oblongata. Arch. f. miki-. Anat., 31, p. 54-71, taf. 6. Langley, J. N. '98. On the Union of cranial autonomic (visceral) Fibres with the Nerve Cells of the superior cervical GangUon. Journ. Physiol., 23, p. 240-270, figs. 1-6. Lewis, F. T. :03. The gross Anatomy of a 12-MM Pig. Amer. Journ. Anat., 2, p 211-226, 4 pis. willard: cranial nerves of anolis carolinensis. 105 Leydig, F. '72. Zur Kenntniss der Sinnesorgane der Schlangen. Arch. f. mikr. Anat., 8, p. 317-357, taf. 15, 16. Leydig, F. '73. Uber dis ausseren Bedeckungen der Reptilien und Amphibien. Arch. f. mikr. Anat., 9, p. 753-794, taf. 32. Lubosch, W. '98. Die vergleichende Anatomie des Accessoriusui'spruages, Diss. Berlin, 1898, 31 pp. Lubosch, W. '99. Vergleichend-anatomische Untersuchungen iiber den Ursprung und die Phylogenese des N. accessorius Willisii. Arch. f. mikr. Anat., 54, p. 514-602, taf. 17. Maurer, F. '95. Die Epidermis und ihre Abkommlinge. Leipzig, 1895, i.x + 352 p., 9 taf., 28 text fig. Merkel, F. '80. tjljer die Endigungen der sensiblen Nerven in der Haut der Wirbel- thiere. Rostock, vii + 214 p., 15 taf. McMurrich, J. P. '85. The cranial Muscles of Amia calva (L), with a Consideration of the Relations of the postoccipital and hypoglossal Nerves in the various Vertebrate Groups. Studies Biol. Lab. Johns Hopkins Univ., 3, p. 121-153, pi. 10. Minot, C. S. :03. A Laboratory Text Book of Embryology. Philadelphia, 1903, xvii + 380 pp. Mivart, St. G. '67. Notes on the Myology of Iguana tuberculata. Proc. Zool. Soc. Lond., p. 766-797, 18 text fig. Mivart, St. G. '70. On the Myology of Chamaeleon parsonii. Proc. Zool. Soc. Lond., p. 850-890, 29 text fig. Miiller, J. '40, Vergleichende Neurologie der Myxinoiden. Abh. Akad. Wiss. Berlin, 1838, pp. 171-252, 4 taf. Also sep. 1840, 83 pp., 4 taf. Noack, [T.]? :07. Tiber die Entwickelung des Mittelolires von Emys europaea nebst Bemerkungen zur Neurologie dieser Schildki-ote. Arch. f. miki". Anat., 69, p. 457-490, taf. 18, 6 text fig. Noon, L. :00. Some Observations on the Nerve Cell Connection of the efferent Vagus Fibres in the Tortoise. Journ. Physiol., 26, pp. v-vii. 106 bulletin: museum of comparative zoology. Norris, H. W. :08. The cranial Nerves of Amphiuma means. Journ. Comp. Neur. and Psych., 18, p. 527-568, pis. 4-8. Onuf, B., and Collins, J. :00. Experimental Researches on the central Localization of the Sym- pathetic with a critical Review of its Anatomy and Physiology. Arch. Neurol, and Psychopathol., 3, p. 3-252, 9 pis. Osawa, G. '98. Beitrage zur Anatomie der Hatteria punctata. Arch. f. mikr. Anat., 51, p. 481-691, 53 text fig. Parker, W. K. '79. On the Development of the Skull and its Nerves in the Green Turtle (Chelone midas), with Remarks on the Segmentation seen in the Skull of various Types. Proc. Roy. Soc. London, 28, p. 329-346. Penzo, R. '93. tjber das Ganglion geniculi und die mit demselben zusammen- hangenden Nerven. Anat. Anz., 8, p. 738-744, 1 fig. Ploschko, A. '97. Die Nervenendigungen und Ganglien der Respirationsorgane. Anat. Anz., 13, p. 12-22, 10 fig. Rabl, K. '87. tJber das Gebiet des Nervus facilais. Anat. Anz., 2, p. 219-227. Rabl-Riickhard, [H.] '78. Das Centralnervensystem des Alligators. Zeitschr. f. wiss. Zool., 30, p. 336-373, taf. 19, 20. Reinhardt, J. '60. Herpetologiske Medelelser. I. Om nogle smaa Gruber, hvormed Skjseell'cne hos mange Slanger ere forsynede. Videnskabehge Meddelel- ser fra den naturhistoriske Forening i Kjobnhavn, no. 10-14, pp. 209- 228. Reinhardt, J. '61. tJber einige kleine Gruben an den Schuppen mancher Schlangen. Arch. Naturg., Jahrg. 27, 1, p. 127-147. [Cf. Reinhardt, J., '60.] Retzius, G. '84. Das Gehororgan du Wirbelthiere. II. Das Gehororgan der Rep- tilien, der Vogel und der Saugethiere, Stockholm, 1884, viii4- 368 pp., 39 taf. Rochas, F. '85. Des nerfs qui ont ete appelles vidiens chez les Oiseaux. Compt. Rend. Acad. Paris, 101, p. 573-575. Ruge, G. '97. t)ber das peripherische Gebeit des Nervus Facialis bei Wirbel- thieren. Festschr. Gegenbaur, 3, p. 193-348, 76 text fig. Sanders, A. '70. Notes on the Myology of Platydactylus japonicus. Proc. Zool. Soc. Lond., p. 413-426. willard: cranial nerves of anolis carolinensis. 107 Sanders, A. '72. Notes on the Myology of Liolepis belli. Proc. Zool. Soc. Lond., p. 154-183, 16 text fig. Sanders, A. '74. Notes on the Myology of the Phrynosoma cornutum. Proc. Zool. Soc. Lond., p. 71-89, 6 text fig. Schauinsland, H. :03. Beitrage zur Entwicklungsgeschichte und Anatomic der Wirbel- tiere. I, II, III. Zoologica, Heft 39, 16. (Stuttgart), p. 1-168, 56 taf. Schwalbe, G. '68. tJber den Bau der SpinalgangHen nebst Bemerkungen uber die sympathischen Ganglienzellen. Arch. f. mikr. Anat., 4, p. 45-72, taf. 4. Schwalbe, C. '82. Ueber die Kaliberverhaltnisse der Nervenfasern, Leipzig, iv + 51 pp. Shore, T. W. '88. The Morphology of the Vagus Nerve. Joum. Anat. and Phj-siol., 22, p. 372-390. Sheldon, R. £. :09. The Phylogeny of the facial Nerve and Chorda Tj^mpani. Anat. Rec, 3, p. 593-617, 6 figs. Staderini, R. '89. Sopra la distribuzione dei nervi glosso-faringeo, vago e ipoglosso in alcuni lettih ed ucceUi. Atti. Accad. Fisiocritici Siena, ser. 4, 1, p. 585-599. Stannius, H. '56. Zootomie der Amphibien. In Handbuch der Zootomie von v. Siebold und Stannius. Theil 2, Die Wirbelthiere, (Aufl. 2), Buch 2, Berlin, 1856, 270 pp. Sternberg, M. '98. tJber den ausseren Ast des Nervus accessorius Willisii. Arch. f. d. ges. Physiol., 71, p. 158-175, 1 text fig. Stirling, W. '76. Note on the Effects of Division of the sympathetic Nerve of the Neck in young Animals. Journ. Anat. and Physiol., 10, p. 511-512. Streeter, G. L. :04. The Development of the cranial and spinal Nerves in the Occipital Region of the Human Embrj^o. Amer. Journ. Anat., 4, p. 83-116, 14 figs., 4 pis. Strong, O. S. '95. The cranial Nerves of Amphibia. A Contribution to the Mor- phology of the Vertebrate nervous System. Journ. Morph., 10, p. 101-230, pis. 7-12. Van Bemmelen, J. F. '88. Beitrage zur Kenntniss der Halsgegend bei Reptilien. I. Anato- mischer Theil. Bijdragen tot de Dierkunde. Aflevering 16, Leiden, p. 99-146, 2 pi. 108 bulletin: museum of comparative zoology. Van Bemmelen, J. F. '89. Ueber der Herkunft der Extremitaten- und Zungenmuskulatur bei Eidechsen. Anat. Anz., 4, p. 240-255. Van Wijhe, J. W. '86. ITber Somiten und Nerven im Kopfe von Vogel- und Reptilien- embryonen. Zool. Anz., 9, p. 657-660. Versluys, J. '98. Die mittlere und aussere Ohrsphare der Lacertilia und Rhjmchd'- cephalia. Zool. Jahrb., Abt. f. Anat., 12, p. 161-406, taf. 10-17. Versluys, Jr., J. :03. Entwicklung der Columella auris bei den Lacertiliern. Ein Beitrag zur Kenntniss der SchalUeitenden Apparate und des Zungen- beinbogens bei den Sauropsiden. Zool. Jahrb., Abt. f. Anat., 19, p. 107-186, taf. 8-11, text figs. A-K. Vogt, C. '39. Zur Neurologic von Python tigris. Arch. f. Anat. Physiol, und wiss. Medicin, Jahrg. 1839, p. 39-58, fig. 1-4 auf taf. 3. Vogt, C. '40. Beitrage zur Neurologie der Reptilien. Denkschr. allg. schweiz. Gesellsch. Naturw., 4, p. ii + 60, 4 taf. Watkinson, Grace B. :06. The cranial Nerves of Varanus bivittatus. Morph. Jahrb., 35, p. 450-472, taf. 11-13. Weber, M. '77. Ueber die Nebenorgane des Auges der Reptilien. Arch. f. Natur- gesch., 43, Bd. 1, p. 261-342, taf. 17-19. willard: cranial nerves of anolis carolinensis. 109 EXPLANATION OF PLATES. All figures are from A7iolis carolinensis. The projection drawings, figures 4, 5, and 6, and the drawings of transverse sections are from the same series of sections (designated "Anolis trans. 30")- In figures 4, 5, and 6 the outlines of the head, the brain and the eye are indicated. The places of passage of the nerves through bony foramina are indicated by cu-cles. The drawing from a dissection, figure 7, is made to conform as nearly as possible to the size of the projection drawings, although it was taken from a somewhat larger specimen. Note. — Since the nerves and muscles are readily distinguishable in the figures, the abbreviations mu. (for muscles), n. (for nerve) and rm. (for ramus) are omitted in text and plates. The colors used have the following meanings : — On Plate 2 and 3 On Plates 4-7 yellow general cutaneous bones (cartilages) . red viscero-sensory muscles. light blue .... somato-motor brain and retina. dark blue .... viscero-motor pale gray . . . .brain integument and mucous membrane. black contours of nerves, nerves in section. ciliary roots and nerves, and sym- pathetic rami LIST OF ABBREVIATIONS. o, /?, 7 See end of list of abbreviations. a anterior. alv. if ramus alveolaris inferior V. alv. su ramus alveolaris superior V. ang os angulare. an'strn. lar anastomosis between motor laryngeal nerves of right and left sides. an'strn. pal. I anastomosis through which the lateral divi- sion of ramus palatinus VII receives its general cutaneous fibers. an'strn. pal. i'm anastomosis through which the intermediate division of ramus palatinus VII receives its general cutaneous fibers. ao aorta. ate os.articulare. aur. ex external ear. avr. m middle ear. ha'hy basihyal. 110 bulletin: museum of comparative zoology. ha'occ basioccipital. ba'sph basisphenoid. brs bursalis muscle. can. p posterior vertical semicircular canal. cap. crv capiti cervicalis muscle. cap. d'clav capiti-dorso-clavicularis muscle, (m. trape- zius, m. cucuUaris) ; nerve ramus of same. cap. md capiti-mandibularis muscle. cap. 7nd.^ nerve ramus to dorsal part of m. capiti- mandibularis. cap. md.- nerve ramus to anterior part of m. capiti- mandibularis. cat common carotid artery. cat. ex external carotid artery. cat. i internal carotid artery. cd. tym chorda tympani. cil ciliary nerve. clav clavicle. cl. gn. sij cells of small sympathetic ganglia at differ- ent places along viscero-sensory rami. clml. aur columella auris. cmpl OS complementare. cnch. na concha nasalis. comn ramus connecting the ramus to m. depressor palpebrae inferioris with rm. palatinus VII. comn. ex communicating ramus between lachrymal plexus and nerve IX uniting with rm. hyomandibularis. comn. i communicating ramus between nerve IX and rm. palatinus VII. comn. IX-X communicating ramus between radices IX et X. cor OS coronoideum. co'st. lar constrictor muscle of the larynx. co'st. vn.j.i constrictor muscle of the internal jugular vein. crt. Mkl Meckel's cartilage. crv. d. XII ramus cervicalis dorsalis XII. crv. I. XII ramus cervicalis lateralis XII. de OS dentale ; ramuli to teeth from ramus alveo- laris inferior. dep. md depressor mandibularis (digastric) muscle. dep. md.^ depressor mandibularis muscle, most ven- trally inserted fibers. dep. path, if depressor palpebrae inferioris muscle; nerve ramus innervating it. willard: cranial nerves of anolis carolinensis. Ill dt. Ich ductus lachrymalis. e'crac epicoracoid. e'pt epipterygoid (columella). e'stn-clei-ynast episterno-cleido-mastoideus muscle (of Vers- luys); nerve ramus supplying it. ex'clml extracolumella. / OS frontale; ramus frontalis V. fen. ovl fenestra ovalis. for. if orb infraorbital foramen. for. Ich , foramen for the lachrymal duct. for. na. I " in the maxilla for passage of na. I. for.na.m " " os nasale " " " na. m. for. pal foramen in the pterygoid bone for the first branch of the palatine nerve to the taste buds and mucous membrane. for. par parietal foramen. for. pt foramen in the pterygoid bone for passage of nerve to taste buds. for. V notch in prootic bone in which rests the Gas- serian ganglion. for VII foramen for radix VII. for. XII foramina for radices XII. gen-gls genioglossus muscle. gl. cat carotid gland. gl. Hard Harderian gland. gl. lab. ex external labial glands. gl. lab. i internal " " gl. Ich lachrymal gland. gl. rba nasal gland. gl. sb'lng. I sublingual gland (lateral). gl. sb'lng. m sublingual gland (median). gls-hy glossohyal. git glottis. gl. thm thymus gland gl. thy thyroid gland. gm. gus gustatory bud in region of larynx; nerve twigs to taste buds given off from infra- orbital plexus. g77i. gus. I lateral field of gustatory buds (both dorsal and ventral) . gm. gus. m median field of gustatory buds (dorsal). gn. cil ciliary ganglion. gn. crv first cervical ganglion of the sympathetic. gn. eth ethmoidal ganglion. gn. if 'orb infraorbital ganglion. 112 bulletin: museum of comparative zoology. gn. md mandibular ganglion. gn. nd ganglion nodosum (trunk ganglion of X) . ign. opth ophthalmic ganglion. gn. pal palatine ganglion. gn. rx. X root, or jugular, ganglion of X. , gn. spi. S spinal ganglion of third spinal nerve. gn. sy. VII sympathetic ganglion cells of the geniculate ganglion. gn.V Gasserian ganglion of V. gn. VII geniculate ganglion of VII. gn. VIII ganglion of VIII. gn. IX petrosal ganghon of IX. gn. X ganglion of X. (This is the same as gn.rx.X). h'hy hypohyal. hy hyoideus muscle; ramus hyoideus VII. hy-gls hyoglossus muscle (m. cerato-hyoideus) . hy-md ramus hyo-mandibularis VII. if orb ramus infraorbitalis V. j OS jugale. ker-brn. I cerato-branchial I. ker-brn. II cerato-branchial II. ker-hy cerato-hyal cartilage; ramus to cerato- hyoid ( = hyoglossus) muscle, a branch of the pharyngolaryngeal. ker-md. ^> ^' ^ cerato-mandibularis muscle and its subdivi- sions; rami to cerato-mandibularis muscle. lab. if. all' rami labiales inferiores given off from ramus alveolaris inferior V. lab. if. md ramus labialis inferior from main ramus mandibulai'is V. lar. Ig longitudinal (dilator) muscle of the larynx. lar. su ramus laryngeus superior X. lax. tym laxator tympani muscle. Ich OS lachrymale. Igs. coll longissimus colli muscle. lig. stp stapedial ligament. lig. tym tympanic ligament. Ing. i'm. XII ramus lingualis intermedius XII. Ing. I. XII ramus lingualis lateralis XII. Ing. I. XII. ^ lateral division of rm. lateralis lingualis XII. Ing. I. XII. ^ median division of rm. lateralis Ungualis XII. Ing. Ig longitudinal tongue muscle superficial to gen-gls. Ing. Ig.^ longitudinal tongue muscle with deep inser- tion. willard: cranial nerves of anolis carolinensis. 113 Ing. Ig? longitudinal tongue muscle with superficial insertion on the mandible. Ing. m. XII ramus lingualis medialis XII. Ing. t transverse tongue muscle. Ing. vrt vertical tongue muscle. Ing. V ramus lingualis of V. Ing. XII + V combined lingual rami of XII and V. md ramus mandibularis V. md.i first motor-sensory branch of rm. mandibu- laris V, supplying skin and m. mylo- hyoideus. md.^ second motor-sensory branch of rm. man- dibularis V, supplying skin and anterior part of m. mylo-hyoideus. mx maxillary bone; ramus maxillaris V. 7tix.^ temporal cutaneous branch of rm. maxil- laris V. mx.^ branch of rm. maxillaris to part of lower lid and its posterior angle. mx.^ branch to anterior part of lower lid. mx.* branch to suborbital region. myl-hy mylo-hyoideus muscle; branches of nerve V supplying it. na OS nasale; ramus of nasalis V through eth- moidal ganglion to skin. na.^ cutaneous branch of rm. nasalis through the ethmoidal ganglion. na. I ramus nasalis lateralis V. na. l.^ ) branches to the skin given off through na. l."^ J external nasal gland. na. m ramus nasalis medialis V. na.m.^, cutaneous branch of rm. nasalis medialis V passing through nasal bone. oh.d dorsal oblique muscle; branch of nerve VI supplying same. oh.v ventral oblique muscle; ramus of III supply- ing it. occ. crv occipito-cervicalis muscle. occ. crv. m occipito-cervicalis medialis. onio-hy omohyoideus muscle; nerve ramus supplying it. omo-hy. prf deeper part of omohyoideus muscle. omo + stn-hy branches of ventral ramus of first spinal nerve to omohyoid and stemohyoidmuscles. opt optic nerve. or.lch opening of the duct of the lachrymal gland. 114 bulletin: museum of comparative zoology. ot otic capsule. p posterior. pal OS palatinum; ramus palatinus VII. ■pal}, pal.-, pal? three divisions of rm. palatinus VII at be- ginning of infraorbital plexus. pal. i'm ramus palatinus intermedius VII. pal. I ramus palatinus lateralis VII. pal. m ramus palatinus medialis VII. pa'occ paroccipital. par OS parietale. pa'sph parasphenoid. p'f _. postfrontal. phx-lar ramus pharygo-laryngeus IX + X. phx-lar.^ first branch given off by the combined XII and IX + X, containing most of the nerve fibres of rm. phx-lar. phx. X small pharyngeal branch given off on distal side of trunk ganglion X . plx. Ich lachrymal plexus. p'orb postorbital bone. pre. pa'ot parotic process. pre. pt-qd pterygo-quadrate process. pref. prefrontal. premx premaxillary . presph presphenoid. pro'ot prootic. protru. oc protrusor oculi muscle; nerve twigs supply- ing it. prt. pt-qd error (fig. 18) for pre. pt-qd. pt pterygoid bone, pterygoid muscle, or nerve ramus to latter. pt.^ deeper part of m. pterygoideus; nerve ramus to same. pt-md pterygo-mandibularis muscle ; nerve ramus to m. pterygo-mandibularis. pt-par pterygo-parietalisjnuscle; nerve ramus to m. pterygo-parietalis. pt-sph. p pterygo-sphenoidalis posterior muscle; nerve ramus to m. pterygo-sphenoidalis poste- rior. qd quadrate. rcr. X ramus recurrens laryngis X. rm.de nerve fibers entering pulp cavity of tooth . rm.tr ramus trachealis. rm. vn.j ramiof IX and X to jugular vein. willard: cranial nerves of anolis carolinensis. 115 rt. a rectus anterior muscle ; ramus of III supply- ing it. rt. d rectus dorsalis muscle; ramus of III supply- ing it. rt. p rectus posterior muscle. rtr. oc retractor oculi muscle. rt. V rectus ventralis muscle; ramus of III supply- ing it. rx. Ill, V, VII, IX, X, Xin 2 3. .roots of nerves III, V, VII, IX, X and XII. rx. cil. Ill radix brevis of cilary nerve. rx. cil. V radix longa of cilary nerve. rx. d. spi. 3 . dorsal root of third spinal nerve. rx. V. spi. 1,2,3 ventral roots of first, second and third spinal nerves. rx. XI P 2 ' first, second and third roots of nerve XII. sa'ang supra angulare. sac. en'lym ■ . . .saccus endolymphaticus. sa'tmp supratemporal. sph OS sphenoidale. spht. coll sphincter colli muscle. spi. coll spinalis colli muscle. spi. d. 1, 2,3 main dorsal ramus of first, second and third spinal nerves. spi. I. 1, 2, 3 main lateral ramus of first, second and third spinal nerves. spi. V. 1,2, 3 main ventral ramus of first, second and third spinal nerves. sq squamosal. stn-hy sterno-hyoideus muscle; nerve ramus sup- plying it. sy sympathetic trunk posterior to union of com- municating rami (internus and externus). sy. X communicating ramus between cervical sympathetic trunk and ramus visceralis X. t OS transversum. lis. tnd tendinous tissue. tnd tendon. tnd. mh. nic tendon of nictitating membrane. ti ' trachea. tym tympanum. vcr. X error in fig. L for rcr. X. vn. j.i internal jugular vein. vom vomer. vsc. X ramus visceralis X. V. spi. 1,2, 3 ramus ventralis of spinal nerves 1, 2 and 3. 116 bulletin: museum of comparative zoology. I-XII cranial nerves I-XII XII. ^ nerve branch to posterior part of m. cerato- mandibularis 1. XII. ■* nerve branch to m. cerato-mandibularis 3 and to anterior part of m. cerato-mandibularis 1. XII. ^ nerve branch to m. cerato-mandibularis 2 (a few fibers to cerato-mandibularis 1). XII. ^ nerve branch to posterior part of genioglos- sus muscle. XII. * nerve branch of Ing. I. XII to longitudinal tongue muscles. a, ^, y posterior communicating rami of the infra- orbital plexus (interpreted as sympathetic). PLATE 1. V Wii.lard: — Cranial Nerves of Auolis carolinensis. ' PLATE 1. Fig. 1-3. Skull. Fig. 1. Dorsal aspect. •Fig. 2. Left lateral aspect. Fig. 3. Ventral aspect. All were drawn from a specimen that had been macerated in water and then dried. The cartilaginous parts are not preserved in such a preparation. $ULL, MUS. COMP. ZOOL. WiLLARD.— Cranial Nerves Anolis, Plate 1 orb. xatrnp. . qd. f^ e^' /or.Vii for.pt. W. A. V/. DEL PLATE 2. Cranial Nerves of Anolis carollnensis. PLATE 2. Fig. 4. Cranial and first three spinal nerves of the left side as projected upon a horizontal plane, dorsal aspect. For clearness, the nerves distributed to the ventral side of the head are shown in a separate figure (fig. 5). The connections of the viscero-sensory components with the fasciculus solitarius are shown by dotted lines. The terminal branches of some of the rami are not carried out in both the figures (4 and 6) , being included only in the figure which shows them with least confusion. X 21 . Fig. 5. Continuation of the nerves showTi in figure 4, showing their course and distribution in the jaw, tongue, pharjoix and larynx. The terminal dis- tribution of the combined nerves V and XII is not included in this figure. Same view and magnification as figure 4. BULL. MUS. COMP. 200L. WiLLARD. -Cranial Nerves Anolis, Plate 2 PLATE 3. Cranial Nerves of Anolis carolinensis. PLATE 3. Fig. 6. Cranial and first three spinal nerves of the left side as projected upon a sagittal plane and seen from the median side. Nerves III, IV and VI with ciliary ganglion and its nerves are omitted; also the terminal rami of V and XII are omitted, because they fall so nearly in the same frontal plane. Fig. 7. Ventral aspect of a dissection of the right side of the head showing about the same group of nerves as figiu-e 4. The figure represents the dorsal portion of the head, the ventral portion, including the bony roof of the mouth, having been cut away to expose the nerves. The plane of division is carried through the tympanum, the quadrate bone and the posterior bony arch of the orbit. The whole mandibular group and ventral cervical muscles are removed, while all those related to the orbit are retained. The shaded background against which the maxillary bone is seen, represents the skin of the dorso- lateral side of the head, the cut edge of which forms the margin of the figure. The two superficial sympathetic rami connecting nerve VII with the lachrymal plexus pass in part through the m. capiti mandibularis, which has been dis- sected away. The structures of the orbit are those immediately exposed upon the removal of the pterygoid and palatine bones. The portion of the maxil- lary bone forming the ventral rim of the bony orbit is cut away exposing the deeper surface of the lower eyelid. The hind brain is exposed, in the region where the last three cranial nerves show their superficial origin, by the removal of parts of the basisphenoid and basioccipital bones. The finer rootlets of nerves IX and X were nearly transparent and practically invisible under the dissecting microscope; the drawing, therefore, does not show their exact number nor connection with the lateral side of the medulla. The part of the medulla exposed is bent sharply away from the observer, making it difficult to represent the linear arrangement of the roots of nerve XII and the occipi- tal foramina through which they emerge. The representation of the details of the muscles of the cervical region is not attempted. The m. spinalis colli and part of the m. longissimus colli are removed, but the roots of the spinal nerves are hidden by the remaining part of the m. longissimus colli. The distribution of the small nerve branch given off from X at the point where the communicating ramus (comn. IX-X.) joins IX, was not determined. It is represented too large in the figure. BULL. MUS. COMP. ZOOL. WiLLARO, -Cranial Nerves Anolis, Plate 3 PLATE 4. Cranial Nerves of Anolis carolinensis. PLATE 4. The sections represented in Plates 4-7 were selected at intervals through the series from v\:hich the projection drawings (Plates 1 and 2, figs. 4-6) were made. The scale included in these drawings (Figs. 4, 6) shows the position of every hundredth section in the series. The planes of the sections shown in figures 8-24 may be determined by referring their numbers to this scale. Branches of minor importance which axe not included in the plottings may appear in the sections. No importance can be placed on a close comparison of the relative diameters of the nerves. These have suffered shrinkage to a different degree in different parts of the head and are represented approximately as they appear in the sections, while they are for the most part enlarged and made uniform in the plottings in order to demonstrate their components. The sections were drawn as projected by the camera and are accurate in respect to those struc- tures which are considered. Fig. 8. Transverse section (No. 130) anterior to external nares. The space between the nasal cartilage and the premaxiUary bone is filled with pericapsular vascular tissue. The median palatine rami are here united. The sublingual gland opens by three ducts (one median, two lateral) anterior to this section. Labial glands open by numerous ducts. The mucous membrane in the floor of the mouth has numerous taste buds located between the gland openings and in the epithelium just median to the upper and lower jaws. Fig. 9. Transverse section (No. 319) tkrough nasal organ, showing external and internal nasal chamber and external nasal gland. The tongue is cut through the region of the greatest development of the lingual papillae, the out lines of which are represented somewhat diagrammatically. The muscle fibers extend to the ends of these, which are provided with long protruding epithelial cells. The taste buds are seen along both upper and lower lingual gum at points indicated {gm. gus. I.). No taste buds were observed on the papillae. A single taste bud was found near the median line in the roof of the mouth, while on the floor of the mouth they appear to be limited to the lingual gum. Fig. 10. Transverse section (No. 579), through ethmoidal ganglion and internal nares, just anterior to the orbit and the larynx. In addition to the series of taste buds found in section No. 319, they appear along the lateral border of the tongue in the stratified epithelium between the openings of the tubular glands. Fig. 11. Transverse section (No. 659) cutting into, the anterior waU of the eyeball and through the Harderian gland. The muscles at the root of the tongue are diverging to each side of the larjoix, which is here cut through the glottis. The two laryngeal muscles appeal-. BULL. MUS. COMP. ZOOL. WiLLARD. -Cranial Nerves Anolis, Plate 4 ym.y-'ao.ex- ■ ^"9 -fa-' rt. ofJbAi 8 130 pre/. ^-^ , 2 md.' , W. A.W. DEL. 579 mv^'^v- 11 659 PLATE 5. Cranial Nerves of Anolls carolinensis. PLATE 5. Fig. 12. Transverse section (No. 767) through the mid-orbit region. The lingual and chorda tympani' nerves are leaving the alveolar canal. No taste buds are found on the floor of the mouth, but they appear in the roof of the mouth among the median palatine glands, and to a less extent along the lateral glandular area. Different parts of the infraorbital plexus appear in the floor of the orbit. Fig. 13. Transverse section (No. 909) through orbit, showing the connec- tion of the optic nerve with the retina. The chorda tympani is closely applied to the median side of the ramus alveolaris inferior. The muscle of the lower lid {dep. palb. if.) shows some of its fibers originating from the connective tissue near the median part of the roof of the mouth. Fig. 14. Transverse section (No. 988) through the anterior part of the optic chiasma (blue tint omitted), showing entrance of ciliary nerves into the eye- ball. The retina is cut tangentially; the infraorbital ganglion, the anterior part of the lachrymal gland, and the mechanism of the bursalis muscle in its relation to the ligament of the nictitating membrane, also fall in the plane of the section. Fig. 15. Transverse section (No. 1064), posterior to the orbit, through the posterior part of the optic chiasma (not lettered). The ciliary ganglion and the lachrymal plexus around a large post-lachrymal blood sinus are shown. The columella (epipterygoid) is cut where it articulates with the pterygoid bone. BULL. MUS. COMP. ZOOL. WiLLARD.— Cranial Nerves Anolis, Plate 5 lab.lJ.viA yf-ftt-. u.f,^md.i„..^ ,1 jcf ,<-i)iH, ,^gr-(ii,(J lah.if-m-d 12 767 -^'^ myt-iiiv 13 909 W. A. V^. DEL, PLATE 6. Willard: — Cranial Nerves of Anolis carolinensis PLATE 6. Fig. 16. Transverse section (No. 1175) through the pituitary body (not lettered) and origin of the retractor oculi and bursaUs muscles. The roots of nerves III and IV here have an intracranial course. Ramus frontalis V is shown (/.) mesad to the origin of the protrusor oculi muscle. The insertion of the m. pterygoparietalis on the pterygoid bone is shown. Fig. 17. Transverse section (No. 1226). This falls in a plane where few main nerve rami are cut, as reference to the plotting will show. The ophthalmic ganglion is shown as completely separate from the mandibular-maxillary part of the Gasserian ganglion. Nerve IV is labelled V by mistake. Fig. 18. Transverse section (No. 1320) through the geniculate ganglion and also showing a portion of the ganglion of nerve VIII. The otic capsule and the tympanic chamber are just appearing. The insertion of m. pterygo- sphenoidalis posterior on the pterygo-quadrate process {prt. pt-qd.) is showTi. Fig. 19. Transverse section (No. 1399) through the inner, middle and outer ear, showing the root of nerve IX about to emerge from the cranium. The chorda tympani, which is shown in this and preceding sections occupying a position in the lower jaw, is here cut at a second place, where it is still in com- bination with the motor part (hy-rnd.) of nerve VII, although its components are distinctly segregated on the lateral side of the. hyomandibular ramus. WiLLARD. -Cranial Nerves Anolis, Plate 6 PLATE 7. Cranial Nerves of Anolis carolinensis. PLATE 7. Figs. 20 and 21. Transverse sections (Nos. 1435 and 1441) through the posterior part of the middle ear cavity, showing the relation of the chorda tympani to the ligament of the extra-columella {lig. tym.). Fig. 20 shows also foramina for nerve X and the second root of nerve XII. Fig. 21 shows the laxator tympani muscle continuing caudad from the place of insertion of the ligament. Fig. 22. Transverse section (No. 1453) through the ganglion of nerve X and the three roots of nerve XII before their union, and immediately anterior to the juncture of Jacobson's anastomosis with the main sympathetic trunk. Fig. 23. Transverse section (No. 1480) through the anterior end of the petrosal ganglion. The skeletal attachment of the constrictor of the internal jugular vein is showTi, also the root and distribution, in part, of the first spinal nerve. Fig. 24. Transverse section (No. 1595) showing ganglion of the third spinal nerve and the position of the sympathetic trunk and ramus visceralis X on the dorsal side of the thjrmus gland. BULL. MUS. COMP. ZOOL. WiLLARD. -Cranial Nerves Anolis, Plate 7 The following Publications of the Museum of Comparative Zoology are in preparation; — LOUIS CABOT. Immature State of the Odonata, Part IV. E. L. MARK. Studies on Lepidosteus, continued. E. L. MARK. On Arachnactis. H. L. CLARK. The "Albatross" Hawaiian Echini. Reports on the Results of Dredging Operations in 1877, 1878, 1879, and 1880, in charge of Alexander Agassiz, by the U. S. Coast Survey Steamer "Blake," as follows: — A. MILNE EDWARDS and E. L. BOUVIER. The Crustacea of the "Blake." A. E. VERRILL. The Alcyonaria of the "Blake." Reports on the Results of the E.xpedition of 1891 of the U. S. Fish Commission Steamer "Albatross," Lieutenant Commander Z. L. Tanner, U. S. N., Commanding, in charge of Alexander Agassiz, as follows: — K. BRANDT. The Sagittae. K. BRANDT. The Thalassicolae. O. CARLGREN. The Actinariahs. R. V. CHAMBERLIN. The Annelids. W. R. COE. The Nemerteans. REINHARD DOHRN. The Eyes of Deep-Sea Crustacea. H. J. HANSEN. The Cirripeds. H. J. HANSEN. The Scliizopods. HAROLD HEATH. Solenogaster. W. A. HERDMAN. The Ascidians. S. J. HICKSON. The Antipathids. E. L. MARK. Branchiocerianthus. JOHN MURRAY. The Bottom Speci- mens. P. SCHIEMENZ. The Pteropods and Heteropods. THEO. STUDER. The Alcyonarians. The Salpidae and Doliolidae. H. B. WARD. The Sipunculids. Reports on the Scientific Results of the Expedition to the Tropical Pacific, in charge of Alexander Agassiz, on the U. S. Fish Commission Steamer "Albatross," from August, 1899, to March, 1900, Commander Jefferson F. Moser, U. S. N., Com- manding, as follows: — R. V. CHAMBERLIN. The Annelids. MARY J. RATHBUN. The Crustacea H. L. CLARK. The Holothurians. H. L. CLARK. The Ophiurans. The Volcanic Rocks. ■ The Coralliferous Limestones. S. HENSHAW. The Insects. G. W. MULLER. The Ostracods. Decapoda. G. O. SARS. The Copepods. L. STEJNEGER. The Reptiles. C. H. TOWNSEND. The Mammals, Birds, and Fishes. T. W. VAUGHAN. The Corals, Recent and Fossil. PUBLICATIONS OF THE MUSEUM OF COMPARATIVE ZOOLOGY AT HARVARD COLLEGE. There have been pubHshed of the Bulletin Vols. I. to LIV. and Vol. LVIIL; of the Memoirs, Vols. L to XXIV., and also Vols. XXVL to XXIX., XXXI. to XXXIV., XXXVI. to XXXVIIL, XL. to XLII., and XLIV. Vols. LV.to LVIL, and LIX. of the Bulletin, and Vols. XXV., XXX., XXXV., XXXIX., XLIIL, XLV. to XLIX. of the Memoirs, are now in course of publication. The Bulletin and Memoirs are devoted to the publication of original work by the Officers of the Museum, of investigations carried on by students and others in the different Laboratories of Natural History, and of work by specialists based upon the Museura Collections and Explorations. The following publications are in preparation : — Reports on the Results of Dredging Operations from 1877 tq 1880, in charge of Alexander Agassiz, by the U. S. Coast Survey Steamer "Blake," Lieut. Commander C. D. Sigsbee, U. S. N., and Commander J. R. Bartlett, U. S. N., Commanding. Reports on the Results of the Expedition of 1891 of the U. S. Fisl Commission Steamer "Albatross," Lieut. Commander Z. L. Tanner, U. S. N., Com- manding, in charge of Alexander Agassiz. Reports on the Scientific Results of the Expedition to the Tropical Pacific, in charge of Alexander Agassiz, on the U. S. Fish Commission Steamer "Albatross," from August, 1899, to March, 1900, Commander Jefferson F. Moser, U. S. N., Commanding. Reports on the Scientific Results of the Expedition to the Eastern Tropical Pacific, in charge of Alexander Agassiz, on the U. S. Fish Commission Steamer "Albatross," from October, 1904, to April, 1905, Lieut. Com- mander L. M. Garrett, U. S. N., Commanding. Contributions from the Zoological Laboratory, Professor E. L. Mark, Director. Contributions from the Geological Laboratory, Professor R. A. Daly, in charge. These publications are issued in numbers at irregular intervals. Each number of the Bulletin and of the Memoirs is sold separately. A price list of the publications of the Museum will be sent on appli- cation to the Director of the Museum of Comparative Zoology, Cambridge, Mass. Bulletin of the Museum of (f^jifijp/irpLt^ive Zottlogy AT HARVAkD COlLIiU^C^E. Vol. LIX. NbJ ^.'' ' RELICS OF PEALE'S MUSEUM. By Walter Faxon. CAMBRIDGE, MASS., U. S. A.: PRINTED FOR THE MUSEUM. July, 1915. IIeports on the Scientific Results of the Expedition to the East- ern Tropical Pacific, in charge of Alexander Agassiz, by the U. S. Fish Commission Steamer "Albatross," from October, 1904, to March, 1905, Lieutenant Commander L. M. Garrett, U. S. N., Commanding; F^'^ii&HED or in preparation: — A. AGASSIZ. V.« Geileral Report on the Expedition. A. AGASSIZ. 1. 1 Three Letters to Geo. M. Bowers, U. S. Fish Com. A. AGASSIZ and H. L. CLARK. The Echini. H. B. BIGELOW. XVI." The Medusae. H. B. BIGELOW. XXIII.ss The Sipho- B. BIGELOW. XXVI." The Cteno- phores. P. BIGELOW. The Stomatopods. CARLGREN. The Actinarla. V. CHAMBERLIN. The Annelids. L. CLARK. The Holothurians. L. CLARK. The Starfishes. L. CLARK. The Ophlurans. F. CLARKE. VIII.8 The Hydrolds. . R. COE. The Nemerteans. J. COLE. XIX.i» The Pycnogonida. . H. DALL. XIV." The MoUusks. R. EASTMAN. VII.' The Sharks' Teeth. GARMAN. XII." The Reptiles. J. HANSEN. The Cirrlpeds. J. HANSEN. XXVII." The Schl- zopods. HENSHAW. The Insects. . E. HOYLE. The Cephalopods. . C. KENDALL and L. RADCLIFFE. XXV." The Fishes. A. KOFOID. III.' IX.e XX.20 The Protozoa. C. A. KOFOID and J. R. MICHENER. XXII." The Protozoa. C. A. KOFOID and E. J. RIGDEN. XXIV." The Protozoa. P. KRUMBACH. The Saglttae. R. VON LENDENFELD. XXI." The Siliceous Sponges. R. VON LENDENFELD. XXIX.^s Hexactinellida. G. W. MUlLER. The Ostracods. JOHN MURRAY and G. V. LEE. XVII." The Bottom Specimens. MARY J. RATHBUN. X." The Crus- tacea Decapoda. HARRIET RICHARDSON. II.» The Isopods. W. E. RITTER. IV.« The Tunicates. B. L. ROBINSON. The Plants. G. O. SARS. The Copepods. F. E. SCHULZE. XI.» The Xenophyo- phoras. HARRIET R. SEARLE. XXVIII.'S Isopods. H. R. SIMROTH. Pteropods, Hetero- pods. E. C. STARKS. XIII." Atelaxia TH. STUDER. The Alcyonaria. JH. THIELE. XV.15 Bathysciadium. T. W. VAUGHAN. VI.» The Corals. R. WOLTERECK. XVIII." The Am- phlpods. >Bull. M. C. Z., « Bull. M. C. Z.. « Bull. M. C. Z.. * Bull. M. C. Z.. 'Mem M. C. Z., 'Bull. M. C. Z., 'Bull. M. C. Z.. 8 Mem M. C. Z. •Bull. M. C. Z.. "> Mem M. C. Z. " Bull. M. C. Z. " Bull. M. C. Z. " Bull. M. C. Z. " Bull. M. C. Z. " Bull. M. C. Z. >' Mem M. C. Z. "Mem M. C. Z. >8 Bull. M. C. Z. i» Bull. M. C. Z. »o Bull. M. C. Z. "Mem M. C. Z. 2» Bull. M. C. Z. 2> Mem M. C. Z. » Bull. M. C. Z. "Mem M. C. Z. " Bull. M. C. Z. "Mem M. C. Z. 28 Bull. M. C. Z. MMem M. C. Z. Vol. XLVI.. No. 4, April. 1905, 22 pp. Vol. XLVI.. No. 6. July. 1905, 4 pp.. 1 pi. Vol. XLVI.. No. 9. September, 1905. 5 pp., 1 pi. Vol. XLVI., No. 13, January, 1906, 22 pp., 3 pis. Vol. XXXIII., January, 1906, 90 pp., 96 pis. Vol. L., No. 3, August, 1906, 14 pp., 10 pis. Vol. L., No. 4. November, 1906, 26 pp., 4 pis. Vol. XXXV., No. 1, February, 1907, 20 pp., 15 pis. Vol. L., No. 6. February. 1907. 48 pp., 18 pis. Vol. XXXV. No. 2, August, 1907, 56 pp., 9 pis. Vol. LI., No. 6, November, 1907, 22 pp., 1 pi. Vol. LII., No. 1, June, 1908, 14 pp., 1 pi. Vol. LII., No. 2. July, 1908, 8 pp., 5 pis. Vol. XLIII.. No. 6, October, 1908, 285 pp., 22 pis. Vol. LII., No. 5, October, 1908, 11 pp., 2 pis. Vol. XXXVII., February, 1909, 243 pp., 48 pis. Vol. XXXVIII., No. 1, June, 1909, 172 pp., 5 pis., 3 map: Vol. LII.. No. 9, June, 1909, 26 pp., 8 pis. Vol. LII., No. 11, August, 1909, 10 pp., 3 pis. Vol. LII., No. 13, September, 1900, 48 pp., 4 pis. Vol. XLI., August. September, 1910, 323 pp., 66 pis. Vol. LIV.. No. 7. August, 1911, 38 pp. , Vol. XXXVIII.. No. 2, December, 1911, 232 pp.. 32 pis. , Vol. LIV., No. 10, February, 1912, 16 pp., 2 pis, , Vol. XXXV., No. 3, April, 1912, 98 pp., 8 pis. , Vol. LIV, No. 12, April. 1912, 38 pp., 2 pis. , Vol. XXXV., No. 4, July. 1912. 124 pp.. 12 pis. Vol. LVIII.. No. 8. August. 1914. 14 pp. , Vol. XLIL, June, 1915, 397 pp., 109 pis. Bulletin of the Museum of Comparative ZoSlogy AT HARVARD COLLEGE. Vol. LIX. No. 3. RELICS OF PEALE'S MUSEUM. By Walter Faxon. S«« corrections and annotations in separatt copjy , MCZ-F CAMBRIDGE, MASS., U. S. A.: PRINTED FOR THE MUSEUM July, 1915. TReports on the Scientific Results of the Expedition to the East- ern Tropical Pacific, in charge of Alexander Aqassiz, by the U. S. Fish Commission Steamer "Albatross," from October, 1904, TO March, 1905, Lieutenant Commander L. M. Garrett, U. S. N., Commanding, PUBi-i&HED ou in preparation: — A. AGASSra. V.« Geileral Report on the Expedition. A. AGASSIZ. I.i Three Letters to Geo. M. Bowers, U. S. Fish Com. A. AGASSIZ and H. L. CLARK. The Echini. H. B.BIGELOW. XVI." The Medusae. H. B. BIGELOW. XXIII." The Sipho- nophores. H. B. BIGELOW. XXVI." The Cteno- phores. R. P. BIGELOW. The Stomatopods. O. CARLGREN. The Actlnarla. R. V. CHAMBERLIN. The Annelids. H. L. CLARK. The Holothurlans. H. L. CLARK. The Starfishes. H. L. CLARK. The Ophlurans. S. F. CLARKE. VIII.' The Hydrolds. W. R. COE. The Nemerteans. L. J. COLE. XIX.i» The Pycnogonlda. W. H. DALL. XIV." The MoUusks. C. R. EASTMAN. VII.' The Sharks' Teeth. S. GARMAN. XII." The Reptiles. H. J. HANSEN. The Clrrlpeds. H, J. HANSEN. XXVII." The Schl- zopods. S. HENP"*'"= ™- W. E. H W. C. K] XX O. A. KC Prol C. A. KOFOID and J. R. MICHENER. XXII." The Protozoa. C. A. KOFOID and E. J. RIGDEN. XXIV." The Protozoa. P. KRUMBACH. The Saglttae. R. VON LENDENFELD. XXI." The SlUceous Sponges. R. VON LENDENFELD. XXIX." Hexactlnellida. G. W. MtJLLER. The Ostracods. JOHN MURRAY and G. V. LEE. XVII." The Bottom Specimens. MARY J. RATHBUN. X." The Crus- tacea Decapoda. HARRIET RICHARDSON. II.« The Isopods. W. E. RITTER. IV.« The Tunicates. B. L. ROBINSON. The Plants. G. O. SAR8. The Copepods. F. E. SCHULZE. XI." The Xenophyo- phoras. HARRIET R. SEARLE. XXVIII. ^s Isopods. H. R. SIMROTH. Pteropods. Hetero- pods. >Bi «Bi >Bi *B\ «M •Bi 'Bi >M »Bi 10 M " Bull. M. C. Z.. Vol. LII., No. 1, June, 1908, 14 pp.. 1 pi. " Bull. M. C. Z., Vol. LII., No. 2. July, 1908, 8 pp., 5 pis. " Bull. M. C. Z., Vol. XLIIL. No. 6. October, 1908, 285 pp., 22 pis. " Bull. M. C. Z., Vol. LII., No. 5, October, 1908, 11 pp.. 2 pis. » Mem. M. C. Z., Vol. XXXVII., February, 1909. 243 pp., 48 pis. 1' Mem. M. C. Z., Vol. XXXVIIL, No. 1, June, 1909, 172 pp., 5 pis.. 3 maps. ■8 Bull. M. C. Z.. Vol. LII., No. 9, June, 1909, 26 pp., 8 pis. " Bull. M. C. Z., Vol. LII., No. 11, August, 1909, 10 pp.. 3 pis. '0 Bull. M. C. Z., Vol. LII., No. 13, September, 1909, 48 pp., 4 pis. " Mem. M. C. Z., Vol. XLI., August. September. 1910. 323 pp., 56 pis. 2» Bull. M. C. Z.. Vol. LIV., No. 7, August, 1911, 38 pp. " Mem. M. C. Z., Vol. XXXVIIL. No. 2, December, 1911, 232 pp.. 32 pis. " Bull. M. C. Z.. Vol. LIV., No. 10, February, 1912, 16 pp., 2 pis. " Mem. M. C. Z.. Vol. XXXV., No. 3, April, 1912, 98 pp., 8 pis. " Bull. M. C. Z., Vol. LIV, No. 12, April. 1912, 38 pp., 2 pis. Bulletin of the Museum of Comparative Zodlogy AT HARVARD COLLEGE. Vol. LIX. No. 3. RELICS OF PE ALE'S MUSEUM. By Walter Faxon. CAMBRIDGE, MASS., U. S. A.: PRINTED FOR THE MUSEUM July, 1915. No. 3. — Relics of Peak's Museum. By Walter Faxon. The zoological collection of the Boston Museum, which had been given to the Boston Society of Natural History in 1893 and 1899, was transferred in 1914 to the Museum of Comparative Zoology at Cam- bridge. It is generally understood that this collection consisted chiefly of the moiety of the old Philadelphia or Peale Museum prop- erty purchased by Moses Kimball in 1850, after the dissolution of that institution in 1846. Since the Peale Museum was the repository of a large number of the types of species described by C. L. Bonaparte, Richard Harlan, George Ord, Thomas Say, and Alexander Wilson, it would seem that a careful study of this material might reveal some of these much-desired types. I have lately examined the North American birds in the collection, with a view to the recovery of speci- mens described and figured by Wilson in his "American Ornithology," 1808-1814. On account of the total loss of the original Peale Museum labels, the task has been a difficult one, and the results achieved are meagTe and often vitiated by uncertainty ; but such as they are, they are placed on record in this paper. Since Wilson had no training as an artist, he found it expedient to draw a bird after it had assumed a definite form and attitude by being stuffed and mounted, often by his own hands. We know this not only by tradition but also by the written testimony of the artist Charles R. Leslie. In his "Autobiographical RepoUections,"^ 1, p. 245 (Amer. ed., p. 163), Leslie says: — "Wilson was engaged by Mr. Bradford as tutor to his sons, and as editor of the American edition of 'Ree's Cj'clopsedia'; while at the same time he was advancing his Ornithology for publication. I assisted him to colour some of its first plates. We worked from birds which he had shot and stuffed, and I well remember the extreme accuracy of his drawings, and how carefully he had counted the number of scales on the tiny legs and feet of his subject." Wilson usuall}' refers to a Peale Museum specimen by its catalogue number at the beginning of his description of each species. Seventy- one out of about eighty-five of the birds described by him under new I Autobiographical Recollections. By the late Charles Robert Leslie, R. A. Eclitei . . . .by Tom Taylor. In two volumes. London: John Murray, 1860. American Edition in one vol., Boston: Ticknor and Fields, ISGO. 120 bulletin: mfseum of comparative zoology. names ^ are thus definitely associated with specimens in the Peale Museum. That the specimens thus referred to by number were in all cases the subjects that he drew and described cannot be affirmed, but is highly probable because even in the case of an extremely abun- dant bird, like the Robin, of which there must have been several speci- mens in the Peale Museum, he cites but one specimen or rarely two. Wilson's custom of drawing from a mounted" specimen affords a clew in some instances for tracing the subject of his pencil. In the absence of original data, we are forced, like Polonius, to find directions out by indirections. When Wilson drew his bird in a simple, conven- tional attitude, agreement between his drawing and a mounted specimen in the old collection will obviously be of slight significance, — an accidental co-incidence, perhaps, since the same correspondence will in all likelihood hold true of birds in divers museums, the work of divers taxidermists. When, on the other hand, he depicts a bird in an unusual or singular posture, conformity of specimen and figure may fairly be taken as evidence that they bear to each other the relation of subject and copy; the weight of the evidence being in strict proportion to the preciseness of the similitude. In a few cases, individual peculiarities of plumage may assist in identifying the origi- nals of Wilson's figures. It does not follow conversely that specimens which do not conform to Wilson's figures are not the subjects that he drew: the exigencies arising from the arrangement of the figures might often compel him to adapt the posture of the bird to the requirements of the plate. With the disappearance of the original Peale labels the chance of recovering most of Wilson's types vanished for ever. Can types ascertained by intrinsic evidence, in the absence of lawful credentials in the form of original labels, be of any practical use? I think they can. Let us suppose, in the way of argument, that it is found convenient to separate a species, Mniotilta varia for example, into two races or subspecies, differing from each other in size or plum- age. In the synonymy of this species stands Certhia maculata Wilson. Will the name maculata be usable for the supposed new subspecies? In order to decide this, it seems to me that it would be incumbent upon the reviser of the species to examine, in the absence of a more authentic type, a specimen in the Boston Museum collection which exactly agrees with Wilson's drawing of Certhia maculata; for we know it was a habit of Wilson's to draw from a mounted specimen; that his 1 Names borrowed by Wilson from William Bartram (Travels through North and South Carolina, etc., 1791) are allotted to Wilson, since their validity depends on his adoption of them. FAXON: RELICS OF PEALE's MUSEUM. 121 type was in the Peale Museum ; and that one half of the Peale Museum came to the Boston Museum. It is generally believed that little was added to the Boston Museum collection of birds after the accession of the Peale collection in 1S50. It is equally probable that the nucleus of the Boston Museum collec- tion antedating ISoO was, very slight and unimportant. At least four primitive shows of the early nineteenth or late eighteenth centuries were the springs which fed the first exhibition of the Boston Museum in 1S41. Oldest of these was the Columbian Museum, a collection first exhibited in the American Coffee-House in State Street, Boston, in 1791, by Daniel Bowen.^ Removed afterward to the corner of Bromfield and Tremont Streets, where in 1795 it assumed the name of the Columbian Museum, it was destroyed by fii-e in 1803, but was afterward revived at the corner of Milk and Oliver Street, and in 1806 the exhibition, under the management of Bowen and ^Y. M. S. Doyle, was moved to a new five-storey building on Tremont Street, near King's Chapel; this building again was burned in 1807 and rebuilt as the "Columbian Hall" during the same year. The Columbian Museum collections were sold Jan. 1, 1825, to the proprietors of the New England Museum for about $5000. Woods's Boston Museum, also known as the ^Market Museum, was opened by Philip Woods in 1804 in Market (Faneuil Hall) Square, Boston. This museum, like the Columbian, was sold at auction in 1822 to the proprietors of the New England ^Museum. The New England Museum, E. A. Greenwood, manager, was chartered by the Massachusetts Legislature and opened July 4, 1818, in the block of buildings on Court Street, Boston, occupying the space between Brattle Street and Cornhill. It commenced with the col- lection of Edward Savage called the New York Museum, which was opened in 1812 in Boylston Hall, over the Boylston Market. J. Mix's New Haven Museum was added in 1821, and, as we have seen above, Woods's or the Market in 1822, and the Columbian in 1825. In 1839 Moses Kimball became the proprietor of the New England Museum, and in 1841 it passed into the Boston Museum, located at first on the corner of Tremont and Bromfield Streets, afterward, in 1846, further down Tremont toward Court Street. In 1850 Mr. Kimball bought one half of the collections of the Peale or Philadelphia Museum, the 1 For facts relating to the old museums of Boston, the reader is referred to Old Landmarks and Historic Personages of Boston, p. 41, 42, 132. By Samuel Adams Drake. Boston: James R. Osgood and Company, 1873. Woods' Boston ISIuseum. By Arthur W. Brayley.< T/ie Bosionian. 2, No. 2, May, 1895, p. 125-130. Boston Museum: The Passing of an Historic Playhouse. By John Bouve Clapp. < Boston Evening Transcript. April 25, June 13, 1903. 122 bulletin: museum of comparative zoology. other half going to P. T. Barnum's American Museum, New York, where it Avas consumed by fire on the thirteenth of July, 1865. That little of value in the shape of natural history specimens accrued to the Boston Museum up to the time of the Peale Museum purchase scarcely admits of a doubt. The stuff received before that time was contributed by museums that partook partly of a dime museum, partly of a vaudeville show.^ Among the announcements made by the proprietor of the Columbian Museum in the Boston Ccniinel news- paper in 1797 I find the two following, which will serve to show the character of those primitive places of amusement whose property went to form the nucleus of the Boston Museum collection : — [Nov. 29, 1797.] LATE ADDITIONS TO THE Columbian Museum, Head of the Mall, Boston. Mr. Bowen mforms the Public, That, he has purchased Mr. Raff's much admired Exhibition of CONCERT CLOCKS, which are placed at the head of the Museum Hall, as a valuable and pleasing addition to that very extensive Repository of Curiosities. 1. Canary Bird, which sings a variety of beautiful Songs, Minuets, Marches, &c. as natural as life. 2. A company of Automaton Figures, which dance to the music of a Harpsichord. 3. Three figures which play the Organ and Clarinet, in Concert. 4. Three figures which play the Harpsichord and Hautboys in concert. 5. King Herod beheading John the Baptist, and his daughter holding a charger to receive the head. 6. A Chimney Sweep, and his boy, on the top of a chimney. 7. Three figures which strike the hours and quarters. 8. A butcher kiUing an Ox. The above CONCERT CLOCKS have been exhibited in New-York, with universal applause, and are well worthy the attention of the Citizens of Boston, and the pubUc in general. The Museum also contains the most extensive Collection of ELEGANT PAINTINGS, that ever was exhibited in the United States, some of which are 10 by 12 feet, elegantly framed, and valued from 500 to 1000 Dollars each. Also, a collection of upwards of 50 elegant Figures of W A X-W O R K, large as life, among which are the following (the most interesting) viz. 1 The lack of appreciation of natural history by the American public diu-ing the early part of the last centtiry appears in Scudder's avowal to Wilson that the ""Witch of Endor" and "Potiphar's Wife" brought ten dollars to his museum where the natural history brought one. Scudder was the founder of the old American Museum of New York (Dunlap's History of the Rise and Progress of the Arts of Design in the United States, 1834, 2, p. 199). FAXON: RELICS OF PEALE's MUSEUM. 123 The late King of France, taking an affectionate leave of his family, just before he suffered under the guillotine: The Queen appears in a rage of dis- traction: The King's sister deeply affected. The young Princess is fainting: And the Dauphin is embracing his unhappy father: The Queen's Maid of honor also appears in great distress : A guard of soldiers are waiting to conduct him to the place of execution. This is an affecting scene, which appears natural as life, and is the most interesting group of WAX FIGURES that ever was exhibited in the United States. GEORGE WASHINGTON, late President of the United States, is ele- gantly situated in the centre of the Museum Hall, surrounded by four beautiful Wax Figures, representing LIBERTY, with the staff and cap; — JUSTICE, with the sword and balance; — PEACE, with the olive branch extended; — And PLENTY, with a cornucopia, or horn of Plenty, loaded with fruit. The President is drest in an elegant suit of black, and his figure is 6 feet and one inch in height, with exact proportion, and the likeness is universally allowed to be the most perfect of any ever offered to the public view. With a great vari- ety of Natural and Artificial Curiosities, — Among which are, a variety of Birds, Beasts, Reptiles, Serpents, one of which is a Rattle Snake, 9 years of age, and 4 feet in length. Insects, Diamond Beetle, Glass Frigates, two feet in length, completely rigged and mounted with Glass Guns, Live Owls, &c. The Museum has lately received the principal Additions for this season. — It is opended every Day (except Sundays) and illuminated every Tuesday, Thursday, and Friday Evenings. Tickets, Half a Dollar; 25 Cents for Children. Nov. 29. [July 8, 1797.] Oyi Tuesday next, II th inst. At BOWEN's COLUMBIAN MUSEUM, ■ Messrs. Chalmers and Williamson, respectfully inform the Ladies and Gentlemen of Boston and its vicinity, that they intend giving their entertain- ment of JUST IN TIME, With Mr. Barrett's EVENING LOUNGE. being an antidote for the Spleen; such things have been; Such things are; such things may be. Mirth, Song and Sentiment, Consisting of Readings, Lectures, Recitations and Songs, as performed at Dibden's, Vauxhall, and the Theatres in Europe: — Being a Comic, Satirical, Whimsical, Humourous, Moral, Illustrative Dissertation and display of Heads, Hearts, Passions, Humours, Whims, Oddities and Characters. — To "Catch 124 bulletin: museum of comparative zoology. the Hiring maners as they rise" Has ever been held, not only allowable, but meri- torious; so it is to be hoped the same wish to please and entertain, will in the present case, be viewed at least with candor. Mr. Philip Woods of the older Boston or Market Museum adver- tised, March 24, 1805:1 — A Monstrous Crocodile, which measured when alive 12 feet in length and 4 feet around the body — was killed in Egypt, when devouring a black boy; which is naturally represented with Mimgo in his mouth. On June 29, 1S05, he designates among the attractions added to his show : 1 — The Philadelphia, New York, and Salem beauties and a number of other figures, also a number of natural curiosities, among which is the skin of the sea-elephant in natural preservation, which measured eighty feet in length and six feet around the body. The Boston Museum thus derived by direct inheritance the unique position which it afterward held among American theatres. Even after its dramatic company came to be one of the best in the United States, it still offered to its patrons its side-shows of picture-galleries, stuffed animals, and chambers of horror in wax-work. Many an old Bostonian remembers his Saturday afternoons as a child at the "Museum," — afternoons ending with ice-creams at Copeland's or oysters at Higgins's, and followed by a restless night perturbed by strange dreams of wax images, boa constrictors, and -"Aladdin," or "The Forty Thieves," fused into one composite horror such as never was on sea or land. Perhaps some may recall the taxidermic " artist" who stood ready to set up a pet canary-bird or kitten " as natural as life," while its owner was assuaging his grief for his lost pet by " seeing Warren." And all this, except the supper and the kitten, for fifteen cents! I doubt if children of the present time can get so much for their mothers' money. Peale's Museum was an institution of a very different kind from its Boston contemporaries, — at least during its earlier period under the management of its founder. Charles Willson Peale^ — artist, soldier, 1 Brayley, I. c. 2 See Biograpliical Sketch of Charles Willson Peale. [By Rembrandt Peale].< Doughty's Cabinet of Natural History, 1, p. i-vii, portr., Philadelphia: 1830. Lieber's Encyclopedia Americana, 9, p. 571-572, Philadelphia: 1832. Dunlap's History of the Rise and Progress of the Arts of Design In the United States, 1, p. 136-142, New York: 1834. Peale's Museum. By Harold Sellers Colton. i newspaper of April 13, 1900: — "At the rooms of the Appalachian Mountain Club this afternoon and even- ing, Walter R. Davis exhibits his collection of .... birds. . . . Two of the most valuable birds in the collection are an English Skylark, from the old Charles Willson Peale collection made in 1784 in Philadelphia, and a Golden Pheasant presented to Mr. Peale by George Washington. These specimens have recently been discovered by C. J. Maynard of Newtonville, after having been lost for over fifty years. When the Peale Museum was sold, a portion was bought by P. T. Barnum, much of the remainder was purchased by Moses Kimball of the Boston Museum, and its identity became lost. When this museum was broken up a few years since, the collection was given to the Boston Natural History Society, who sold the birds to Mr. Maynard, not knowing their origin. Many of Alexander Wilson's types are in the collection." What is said in the above-quoted passage about the Boston Society's ignorance of the origin of the Boston Museum collection is not true: in his report at the annual meeting of the Boston Society of Natural History, May 2, 1894,^ the Curator, Alpheus Hyatt, narrated at some length the history of this collection and its connection with the historic Peale Museum. After Mr. Maynard bought the collection it was sent to his residence in Newtonville, Mass., and stored for a while in his barn. It was subsequently redeemed by the Boston Society, all of the specimens, or nearly all of them, being recovered. Then the birds, with excep- tion of a few of the larger kinds, were wi-enched from their stands and packed into tin cases, to the great detriment of their legs and plumage. This should have been done, if at all, only after the collection had been submitted to a careful study. Mr. J. D. Sornborger was then employed by the Society to examine 1 Proc. Boston Soc. Nat. Hist., 26, p. 275-276. FAXON: RELICS OF PEALE's MUSEUM. 127 the collection and report upon it; but nothing resulted from this endeavour except the deplorable loss of such of the original Peale Museum labels as still remained, pinned to or tucked under the wings of some of the birds. These labels were removed from the specimens and put into a paper envelope which was afterward lost. The tickets which now accompany the specimens were probably printed after the collection passed into the custody of the Boston Museum; they consist merely of the name^and habitat of the animals, without any data. I believe the only surviving labels which probably go back to the Peale Museum are two wooden ones belonging to a pair of Golden Pheasants presented to Charles Willson Peale by George Washington.^ I now proceed to my notes made during an examination of the North American birds in this collection, premising that heretofore the origi- nals of only three of Wilson's birds have been located; viz.: — Sylvia maritima Wils., Type. 6, 1812, p. 99, pi. 54, fig. 3. [ =Dendroica tigrina (Gmel.), 1788]. In coll. Vassar College, Poughkeepsie, N. Y. See Orton, Amer. Nat., 4, 1871, p. 714. Falco mississippiensis Wils., Type. 3, 1811, p. SO, pi. 25, fig. 1. In coll. Academy of Natural Sciences of Philadelphia. See Stone, Proc. Acad. Nat. Sci. Phila., 1899, p. 11; Auk, 16, 1899, p. 169. Falco pennsylvanicus Wils., Type. 6, 1812, p. 92, pi. 54, fig. 1. [Nom. praeoc. = Sparvius platijpterus Vieill., 1823; Falco wilsonii Bonap., 1824; Falco latissimus Ord, 18241.- In coll. Acad. Nat. Sci. Phila. See Stone, id supra. 1 Mr. Maynard tells me that there were also two groups of momited iJjirds arranged in two glass cases, presented by Washington to the Peale Museum; these were trans- ferred with the rest from the Boston Museum to th^ rooms of the Natural History Society, but they had been disposed of before Mr. Maynard purchased the collection. 2 See Faxon, Auk, 18, April, 1891, p. 217. 128 bulletin: museum of comparative zoology. Notes on some of the North American Birds in the Boston Museum Collection, now in the Museum of Comparative Zoology. Alle alle (L <; "^ ^ :::^ ^ ^ 71 66 -- ^ c:^ -- ^ [y in 6 ■^ / IS ( / -in \, •, ^ / "^ --- — 65 / 'in / s^ / fin / / Tn / 7S / / 85 90 95 / 110 115 120 125 130 U5 140 I / / / / / i5n k 4- L ... Fig. 4. — Temperature sections on the continental shelf ofif New York and Long Island (Stations 10063, 10065, 10066, 10067, 10082) and at the edge of the Gulf Stream, Lat. 39° 55' (Station 10064). 68°, no doubt influenced by the violent tide. Two days later the surface water was 72° from Woods Hole to the east end of the Sound. But it was much colder (61°) off Monomoy; and only 50° on Pollock Rip, this last being evidence, of course, of thorough vertical mixing BIGELOW: COAST WATER EXPLORATION OF 1913. 159 by the tidal currents. When this dangerous channel was left, the surface temperature rose to 63°, the normal figure for the southern half of the Gulf of Maine at this time of year. Late in August, when the Grampus came southward again (p. 154) the temperature was practically unchanged off Block Island and over the shelf south of ^n < 7 48 49 50» 61 52 5 3 54 55 56 57 58 59 60° «1 2 63 64 B5 66 67 68 69 }Vn r 7 7 1 75 76 s — ' ^ ^ y ^ ^ 10 . " ^ ■-' ri ^ ^ 15 ^- " ■ _ ^ ^ Z= 77 ^ 69 = = J ^ -■ 25 ^ V ^ 72 / ^ IS / _, / / TO / 50 / / 60 / / / 85 / 95 /' / / / 71 135 145 / isn y Fig. 5. — Temperature sections on the continental slielf south of New York (Stations 10069, 10070, 10072, 10073) and at the edge of the Gulf Stream in Lat. 38° 56' (Station 10071). Marthas Vineyard; but near shore south of New York, the water had cooled to 71°-72°; immediately off Cape May to 74° (Fig. 2). On the other hand, the surface south of Nantucket Shoals was several degrees warmer than it was in July, the temperature having risen from 160 bulletin: museum of comparative zoology. 61° to about 67.5° at Nantucket light-ship; and the curves for TO'* and 72° reveal a tongue of warm water extending from the outer edge of the shelf south of Long Island northeastward toward Nan- tucket. Probably it was Gulf Stream water driven northward over 5 10 15 49 50° 51 52 53 54 55 56 67 58 59 60° 61 62 63 64 65 66 67 68 69 70° 71 72 73 74 75 76 77 _ _ _ _ _ _ _ _ _ P _ ^ - ;^ ^ i — • 2 ^ ^ ^ — — — — ~ _j iiS 75 ^ - — —J H ^ ^ =s — —J — — — — — — "l^ -— ^ 77 - ;^ -- ^ ^ ' ^ / 74 / / / / / 1 / 1 / 1 1 j / / / '' .,, Fig. 6. — Temperature sections on the continental shelf south of Delaware Bay (Stations 10074, 10075, 10077), at the mouth of Chesapeake Bay (Station 10078) and at the edge of the Gulf Stream opposite Chesapeake Bay (Station 10076). the shelf by the southerly gale of August 23; but no salinities were taken. For a list of the surface temperatures taken by Mr. Welsh, see p. 350. Temperature sections (Table, p. 344) . In general there was a rapid fall in temperature from the surface downward, all over the conti- nental shelf, from Cape Cod to Chesapeake Bay; and the sections BIGELOW: COAST WATER EXPLORATION OF 1913. 161 show that depth for depth the temperature was lowest in the north- west corner of the broad bight formed by the coast Une, off New York; warmest, as might be expected, along the edge of the con- tinental slope, next the Gulf Stream. Over Nantucket Shoals as a whole, there was probably very little difference between bottom and surface water, the surface, in July, often being as cold as 55° ; and this rather cold water apparently showed its effect as far westerly as Station 10062 (Fig. 3), which was 1-3° colder at all depths down to 25 fathoms than the next station to the westward (Station 10063). Over the outer part of the continental shelf south of Long Island, the temperature was comparatively uniform, station for station, down to 30 fathoms (Fig. 4) cooling rapidly from the surface downward. But the curves for Stations 10061 and 10065 reveal a warm layer of water on the bottom. The water was very much colder close to the shore near New York than it was further off shore (p. 156), and the same was true along the New Jersey coast, for though by the time we came north, the surface had warmed to about 75°, a rise of about 7°, the bottom water in ten fathoms was still only about 52.6°. Off Barnegat the temperatures increase regularly at all depths from the coast eastward (Fig. 5). The ten fathom temperatures for these stations are successively 52°, 58.5°, 70°, 71°; while the fact that at twenty fathoms there was a difference of 17° between Stations 70 and 71 (50° and 67°) only fifteen miles apart, and that the latter, lying 46 47 48 49 50° 51 52 53 54 55 5G 57 61 62 63 64 65 66 67 70° 71 72 73 74 75 76 77 ^-"^-'^"-J-5^^^E==^-"-- ^^t i-^ ^ 84 Fig. 7.— Temperature sections close to the land, south of New York (Sta- tions 10068, 10079, 10080) and off Long Island (Stations 10083, 10084). over the 500 fathom curve, is much warmer than any of the stations on the continental shelf, shows how sudden the temperature transi- tion between coast and ocean water was. Our only station abreast of the mouth of Delaware Bay (Station 10073, Fig. 5) was consider- ably warmer above twenty fathoms than the station next north of it (10072) ; and several degrees warmer, at all depths, except for the sur- face laver of five fathoms or so, than the water south of it (Station 162 bulletin: museum of comparative zoology. 1C079) . And as the high surface temperature of the latter was almost certainly due to the seasonal warming which had taken place during the interval between our two visits, it is safe to say that at Station 10073 a mass of water warmer than the water either north or south of it was crossed. South of Delaware Bay the water was also found coldest next the coast (the warm surface at Station 10078 was the result of the unusually hot weather of the preceding three or four days) . And the curves show, further, that the two stations abreast of Chesapeake Bay (10078 and 10077) were from 1.5° to 3.5° colder, depth for depth, at the lower levels than the two stations immediately north of them,, a fact of interest in connection with the salinity of the region. Fig. 8. — Chart of bottom temperature on the continental shelf for July; and in the Gulf of Maine for August. Temperature below 41°, cross hatched; 41°-45°, single hatched. The dotted line is the 100 fathom curve. BIGELOW: COAST WATER EXPLORATION OF 1913. 163 Station 10071 was considerably the warmest at all depths above 150 fathoms of the three stations outside the continental shelf (Figs. 4, 5, 6) and presented a fairly typical Atlantic curve; the temperature falling rapidly at first from 76° at the surface to 58.8° at fifty fathoms; then more and more slowly until at the lowest level, 250 fathoms, a reading of 43.6 was obtained. Station 10064 was some 6° colder at the •surface, the diiference gradually decreasing downward; but even at Fig. 9. — Temperatiire profile from the southern end of the basin of the Gulf of Maine (Station 10058) across Nantucket Shoals to the continental slope south of Nantucket (Station 10061). 250 fathoms it was 2° colder (41.6°). Station 10076 was the most southerly of the three, and might, therefore, have been expected to be the warmest, as it lay at about the same relative position on the slope. But as a matter of fact the temperature (49.3°) at 150 fathoms (the deepest reading) was about the same as that of Station 10071 : and above this level, Station 10076 was considerably the colder of the two.. 164 bulletin: museum of comparative zoology, Bottom temperature. The chart of bottom temperature (Fig. 8) illus- trates the localization of cold bottom water on the mid-zone of the continental shelf south of Long Island and Marthas Vineyard in July, the southern boundary of which must have been somewhere between the latitude of New York and the line off Barnegat. Shore- ward as well as seaward, the bottom water was warmer than 45°. That this should have been the case nearer land was to be expected, because of the steady shoaling of the water. But the fact that the bottom water was warmer (50°-51°) between 50 and 125 fathoms Fig. 10. — Temperature profile across the outer part of the continental slope southwest of Nantucket (Stations 10063, 10062, 10061). than at 35-50 fathoms, would have been a surprise had not a similar phenomenon been encountered by Libbey (1891) south of Marthas Vineyard in 1889 (p. 241). As pointed out (p. 165) this cold bottom water was not continuous with the cold water in the Gulf of Maine, being interrupted on Nantucket shoals, where the bottom tempera- ture is raised, and the surface correspondingly chilled, by vertical tidal mixing. But no doubt, in winter, the cold water is continuous across the shoals. On the continental slope the temperature was 45° at about 200 fathoms. BIGELOW: COAST WATER EXPLORATION OF 1913. 165 The bottom temperature was higher south than north of Delaware Bay, and instead of being coldest over the mid-zone of the continental shelf, decreased from the shore seaward, with increasing depth. Temperature profiles. The lines were planned to afford three com- plete profiles across the continental shelf, one abreast of Montauk, one off Barnegat, and one opposite the mouth of Chesapeake Bay respectively, besides several incomplete ones in intermediate posi- tions, and a complete profile from the deep basin of the Gulf of Maine to the Gulf Stream via Georges and Nantucket Shoals. The latter (Fig. 9) shows that there was a marked temperature contrast between the waters on either side of the Shoals which form the southern boundary of the basin of the Gulf. On the north, the deep basin, below twenty-five fathoms, was filled with water of 42° or colder, with a rapid rise in the upper twenty fathoms to the surface temperature of 62°-63°. On the southern side, the coldest water was about 47°, at sixty fathoms, while the surface temperature was some 6° warmer at the off shore end of the profile (Station 10061) than in the Gulf (68°). Over the Shoals in the centre of the profile there are local regions of complete vertical mixing by the tidal currents, as for instance on the southwest side of George's Bank (Station 10059) where the temperature was practically uniform from surface to bot- tom (54.7°). On outer edge of the continental shelf the coldest water (47.3°) was not on the bottom, but at fifty fathoms, with warmer water (51.5°) below it. And as Gulf Stream water was to be expected only a few miles further off shore, it is fair to assume that this water colder than 50° indented the warmer ocean water like a tongue, as represented by the curve for 50°. The fact that there was no water on this line colder than 47° shows that the cold bottom water (45°) west of Nantucket Shoals (Fig. 10) was not continuous with the still colder water of the Gulf of Maine. The next profile (Fig. 11), running from the neighborhood of New York to the 500 fathom curve in Lat. 39° 55', shows the cold water on the shelf at 20^0 fathoms, indenting into the warmer water over the slope. The temperature was much higher, depth for depth, outside the edge of the shelf, than over the latter, as is shown by the sharp seaward dip of all the curves. And at the shore end of the profile the same was the case, the curves rising as the land is ap- proached, with equal temperatures about five fathoms nearer the surface at Station 10067 than at Station 10066. In the central part of the profile (Station 10065 to 10066) there was little horizontal change in temperature from east to west. 166 bulletin: museltvi of comparative zoology. Fig. 11. — Temperatiire profile across the continental shelf from New York to the edge of the Giilf Stream in Lat. 39° 55' (Stations 10067. 10066, 10065, 10064). BIGELOW: COAST WATER EXPLORATION OF 1913. 167 Fig. 12. — Temperatxire profile across the continental shelf abreast of Barne- gat (Station 10069), to the GuLf Stream in Lat. 38° 56'. The immediate shore end of the profile is reconstructed from the temperature section a few miles further north (Station 10068). 168 bulletin: museum of comparative zoology. In the fourth profile, off Barnegat (Fig. 12), the water on the shelf was warmer, its minimum being 48° on the bottom at 40-50 fathoms. But on the slope it was only below 150 fathoms that the water was as Fig. 13. Fig. 13. — Temperature profile across the continental shelf 45 miles south of Delaware Bay (Stations 10079 and 10074). Fig. 14. — Temperature profile across the continental shelf to the edge of the Gulf Stream abreast of Chesapeake Bay (Stations 10078, 10077. 10076). Fig. 14. cold as this. And probably there was a belt of bottom water of 50°- 55° at about 100 fathoms, to judge from the other profiles. But there is no bottom reading at this level. In the upper part of the profile ' III 170 bulletin: museum of comparative zoology. the temperature rises, depth for depth, from the land seaward, as in the preceding one. Two partial profiles, one just north, the other just south of Delaware Bay, connect the Chesapeake Bay profile with the one just described. The stations composing the first of these (Stations 10080 and 10072) were, unfortunately, occupied at an interval of two weeks; but other observations have shown that it is only the inter- mediate surface layer which had warmed up appreciably in the interval. At the outer of the two stations the bottom temperature w^as 47.8° at twenty-five fathoms; and corresponding to the steepness of the shelf, this cool water was found nearer shore, though at about the same depth, than further north. Just south of Delaware Bay (Fig. 13) there was no water colder than 50° on the shelf; the lowest temperature being 50.8° at thirty fathoms (Station 10074). But the curves show the progressive warming, depth for depth, from land to sea, which characterize the preceding profiles; the reading (52.5°) being the same at fifteen fathoms at the shore end as at .twenty-seven fathoms at the ofi^shore end of the profile. Off Chesapeake Bay (Fig. 14) the slope was bathed with water of 50°-52° from twenty-five fathoms down to 130 fathoms. There the surface water .cooled from the shore seaward instead of warming as it does further north (p. 165). But though the temperature above five fathoms was highest at the shoreward end of the profile, the ten fathom (bottom) temperature was lower (57.6°) there than further off shore. The general rise in temperature on the shelf from north to south is illustrated by a profile parallel with the coast at about the forty fathom curve (Fig. 15). Below twenty -five fathoms the curves are distorted by the intrusion of warm water (51°) on the bottom near Station 10065, resulting in the extension of cold water southward over warm. The lowest temperature is at the northerly end at forty-five fathoms. Temperature in the Gulf of Maine. Surface temperature. The distribution of surface temperature in the Gulf of Maine was the same in general as in 1912, the north- eastern part being coldest, the southwestern warmest. The surface water (Fig. 1) abreast of Massachusetts Bay, along shore from Cape Cod to Cape Elizabeth, and eastward nearly to German Bank was 60° or warmer, usually 60°-62°; and although the surface was consider- ably warmer (64°-66°) northeast of Cape Cod and in the neighborhood BIGELOW: COAST WATER EXPLORATION OF 1913. 171 of Cashes Bank, this was the result of solar warming, not of Gulf Stream water, as proved by the low salinity (p. 200) . The northern, western, and eastern limits of this warm region can be defined with some accuracy from the hourly temperatures ; but how far it extended to the south is doubtful. It is not likely, however, that it was directly continuous with the warm surface water south of Georges Bank, for the surface temperature on the latter is lowered by the violent tidal currents (p. 155). At the eastern side of the Gulf a sudden transition from the high temperature of the basin to cold surface water on German Bank was noted, the temperature dropping from 60° to 48°, the coldest surface reading of the cruise. Off the Nova Scotia coast the surface tempera- ture was 52°-53°, rising to 54°-56° abreast of the mouth of the Bay of Fundv. Off Mt. Desert Rock Station 10100 showed that the zone f^ n 40° 41 42 43 44 45 46 47 48 49 50° 51 52 53 54 55 56 57 58 jgjg^ '^ ^ ^ Fig. 16. — Temperature sections off Cape Cod in JUly (Station 10057) and in August (Stations 10085, 10086) and In Massactiusetts Bay in August (Station 10106). of 54°-56° water was of considerable breadth. Near the northeast coast of Maine the surface temperature was 50°-52° ; rising to 54°-56° off Mt. Desert Island. Temperature sections. The temperature curves off Cape Cod (Station 10057, Fig. 16; Station 10058, Fig. 3) and off Cape Ann (Station 10087, Fig. 17); near Piatt's Bank (Station 10089, Fig. 18) and near Cashes Ledge (Station 10090, Fig. 18) show a very rapid cooling from the surface down to about thirty fathoms, followed by a layer, reaching down to the bottom, in which the temperature was almost uniform. In 1912, the temperature of the uniform bottom water was 40.3° at all the stations off Cape Ann and Massachusetts Bay; in 1913 it was 43.9° near Cashes Ledge, 41.3° near Piatt's Bank; 40.3° in the southern part of the trough between Jeffrey's Ledge and 172 bulletin: museum of comparative zoology. the mainland. In the northern part of the trough (Station 10104), it was 39.8° at eighty fathoms. In the summer of 1912, the water of the Gulf was invariably cold- est at the bottom; but in 1913 the western basin and two stations in the eastern basin (Stations 10092, 10093, Fig. 17) were coldest in 40' 41 42 43 44 45 46 47 48 49 5r 5 52 53 54 55 5 5 58 59 60° 61 62 6 64 65 66 67 ta.^ / ^ ■;^ r:^ -7" E= -- -^ / _-; ^ ■^ ^ -- ■^ -p 15 20 25 ~ ^ .^ "^ — ^ -— ^ ■^ ^ ^ ^ ■-' ^ ^/ ^ / y* ^ / ' y / ^ Y / / i^ 40 45 50 55 60 65 70 75 / // ^ / 1 / V - / fl / f / / \ ' H 37 90 no 93 92: 145 5? _ Fig. 17. — Temperature sections In the Gulf of Maine from Massachusetts Bay to German Bank (Stations 10087, 10088, 10090, 10092, 10093, 10094). the intermediate depths; the minimum in the former being 41.3° at fifty fathoms, i. e., precisely the same as the uniform bottom water nearer shore, rising to 43.3° at 100 fathoms, below which level it was uniform dowTi to the bottom (150 fathoms). On the western BIGELOW: COAST WATER EXPLORATION OF 1913. 173 side of the eastern basin (Station 10092) the upper layers were colder, but the minimum was warmer (42.2° at forty fathoms), with about 43° at 100 fathoms, below which it was practically uniform to the bottom (130 fathoms). At the eastern side of the eastern basin (Station 10093) the minimum (41.1° at fifty fathoms) was about the same as in the western basin, though the upper layers, and the bottom water (41.6° at 115 fathoms) were both colder than the latter. All these temperature curves are characterized by a sudden change Fig. 18. — Temperature sections in the Gulf of Maine near Piatt's Bank (Station 10089) and north of Cape Ann (Stations 10102, 10103, 10104, 10105). in direction at about the 30-40 fathom level, corresponding to the point at which the fall of temperature ceases to be rapid. And in 1912 this was true of the trough west of Jeffrey's Ledge. But in 1913 the temperature sections at the two Stations in the latter (Stations 10104 and 10105, Fig. 18) show a steadily decreasing rate of cooling from the surface downward. x\nd this is true in general of the Sta- tions off the coast of Maine (Stations 10098, 10099, 10101, 10102, Fig. 19, and 10103) and of the northern end of the eastern basin (Stations 10097, 10100, Fig. 20). The water next the coast was, progressively, 174 bulletin: museum of comparative zoology. colder on the surface, warmer on the bottom, from Cape Ann toward the Bay of Fundy, for example the surface and fifty fathom tempera- tures were 64° and 41.05° at Station 10105; 61° and 44° at Station 10103; 54° and 47.5° at Station 10101. And though this change was interrupted off Mt. Desert (Station 10099), the difference between surface (50.5°) and bottom (48.3°) off the Grand Manan Channel (Sta- tion 10098) was only 2°. At Stations 10097 and 10100 the temperature agreed -at the surface (55°) and at 100 fathoms (43.2°) ; but from about ten fathoms down to about fifty fathoms. Station 10100 was the colder of the two, with a difference of 3° at twenty fathoms, a fact probably due to an upwelling of cold water from below. On the Nova Scotia slope, off Lurcher Shoal (Station 10096), the temperature curve (Fig. 19) agrees very closely with that for Station Fn n 40 ° 4 42 43 44 45^46 47 48 49 50° . \^ 2 3 54. 55 56 57 58 59 60" 61 62 9 i / ^8 y /^ / n f ^ / ^ y 16 20 / / , -— -^ 1 A ^ -^ ' 7 in / 1 35 / 40 / 1 1 / 50 / "^ / 91 / 60 / ^ ^ 70 96 75 i'lG. 19. — Temperature sections in the Gulf of Maine, on Jeffrey's Banli (Station 10091); off Matinicus Island (Station 10101); off the coast of Maine near the Grand Manan Channel (Station 10098) ; near Lurcher Shoal (Station 10096); and on German Bank (Station 10095). 10097 from the surface down to fifty fathoms, cooling from 54° to about 47°, and although the seventy fathom reading (43°) was colder than the water at the corresponding level in the northern part of the basin, it was almost precisely the same as the bottom water there (Stations 10097 and 10100). The temperature was practically uniform from the surface downward, on German Bank; and even over the seventy fathom curve on its western slope (Station 10094, Fig. 17) the difference between surface and bottom was only about 3° (48°-44.9°). BIGELOW: COAST WATER EXPLORATION OF 1913. 175 At the one Station in Massachusetts Bay (Station 10106, Fig. 16) the upper part of the temperature curve agreed almost exactly with the water off Cape Ann, (Station 10087) and near Piatt's Bank (Sta- tion 10089), cooling from 61° at the surface to 48.5° at fifteen fathoms. But at thirty-five fathoms (bottom) it was 2° warmer (44.1°) than either of these. All the temperatures described so far for the Gulf were taken in August, during a week's period: and hence directly comparable /J 40°4I42_43 45 46 47 48 49 50° 51 52 53 54 55 56 57 58 f, " — 1 - ■p 7 in ,^ / / IS / / ?n / / ?5 / / in f / ■ :\h / i\h / in) V 5*1 1 / II Ij / 105 1 / 97 1?0 L _ _ _ _ _ _ _ _ _ _ _ Fig. 20. — Temperature sections in the northeastern part of the basin of the Gulf of Maine (Stations 10097, 10100). with one another. But three Stations, one off Cape Cod (Station 10057, Fig. 16) at the same location as Station 10086, one over the southern part of the basin (Station 10058, Fig. 3), the third on George's Bank (Station 10059, Fig. 3), were made a month earlier. The surface temperature of the first (61°) is exactly the same as it was at the same locality in August (Station 10086); but below the surface the July section is colder at all depths, the greatest difference being on the 176 bulletin: museum of compakative zoology. bottom (forty fathoms), where the water was 41.2° as against 43.2° a month later. Station 10058 was about 2.5° warmer than the Cape Cod stations down to twenty fathoms; but at thirty fathoms it was about .5° colder (41.1°); with a minimum of 40.6° at sixty fathoms, below which it warmed slightly; and its curve is almost exactly parallel with that of the nearest August Station (10085), though about 3° warmer at all depths. The water on the southwestern part of George's Bank (Station 10059) was nearly uniform from the surface downward, in tempera- ture as well as in salinity (p. 188). Mean temperature. If all the temperature curves in the Gulf were parallel, a direct comparison between them would show which regions were potentially warmest, which coldest. But they are so distorted by greater or less active vertical circulation, that it is only by calcu- lating the mean temperatures for each station that light can be obtained on this subject. The mean temperatures for the zone be- tween the surface and the fifty fathom level are given in the following table : — Station Mean tem. Station Mean tem, 10087 46.3° 10096 50.2° 10088 49.4° 10097 50.3° 10089 47° 10100 48.4° 10090 47.5° 10101 49° 10091 50° 10102 48.5° 10092 46.5° • 10103 48.4° 10093 50° 10104 47.2° 10094 47.1° 10105 47° The mean between the surface and forty fathoms, was 46° at Station 10057; 48.5° at Station 10086; 48.8° at Station 10106; the thirty fathom mean was 47.6° at Station 10095 ; 55° at Station 10059. Thus the upper fifty fathoms was coldest, as a whole, on the western side of the Gulf. Passing northeastward along the coast, the mean tem- perature rose from 46.3° near Cape iVnn to 48.4° off Cape Elizabeth, 49° off Penobscot Bay, 48.4° off Mt. Desert Rock and 50.3° over the northern end of the basin. In the centre of the Gulf it was generally 49°-50°, except for one cold Station (10092) . Off the mouth of the Bay of Fundy the mean (50.3° at Station 10096) was as high as anywhere in the Gulf. But the upper fifty fathoms over the slope of German Bank (Station 10094), and the whole column of water on the Bank itself BIGELOW: COAST WATER EXPLORATION OF 1913. 177 Station 10095), was distinctly colder (47.1°-47.5°) than the correspond- ing layer of water either west, north or northwest of it (Stations 10093, 10096, 10097, 10100). Consequently vertical mixing of the upper fifty fathoms of water immediately surrounding the Bank could not repro- duce the temperature observed on the latter; there must have been either an influx of cold water from elsewhere, or some upwelling. The mean temperature of the layer of water between 50 and 100 fathoms was: — Station Mean tern. Station Mean tern, 10088 42.4° 10093 42.4° 10089 41.3° 10097 44.5° 10090 43.7° 10100 44.5° 10092 42.6° At Station 10087, 50-70 fathoms, the mean was 41.2°; at Station 10104, 50-85 fathoms, 40.5°. Thus the bottom water of the deeper parts of the Gulf, like the upper layers was warmest in the northern part of the eastern basin (Stations 10097, 10100); coldest, next the western shore (Stations 10087, 10104). In the preceding sentences the differences in mean temperature have been emphasized; but in reality the striking result of the calcu- lation is the uniformity of the Gulf, the extreme divergence of the mean of the upper fifty fathoms being only about 4°, that of the mean between fifty and 100 fathoms about the same, over an area of about fourteen thousand square miles. The mean temperature of the upper 15 fathoms, i. e., of the zone most subject to solar warming, shows a much greater range (about 11.2°), as illustrated in the following table: — Station Mean tem. Station Mean tem, 10087 54.5° 10096 52.3° 10088 58.5° 10097 53.5° 10089 55.1° 10098 49.3° 10090 55.5° 10100 52° 10091 55.1° 10101 51.5° 10092 53.7° 10103 55.2° 10093 58° 10104 53.2° 10094 47.3° 10106 55° 10095 47.7° 178 bulletin: museum of comparative zoology. The distribution of the fifteen fathom mean, highest in the central part of the Gulf (Stations 10088 and 10093), falling to about 54°-55'' over the western half of the Gulf generally, and lowest in its northeast corner and on German Bank, corresponds with the distribution of surface temperature, and with the proportional strength of the tidal currents, just as might be expected, solar warming being most effective where vertical circulation is least active. Temperature jjrofiles. The general distribution of temperature across the Gulf, from east to west, is illustrated by a profile from Massachusetts Bay to German Bank (Fig. 21, Stations 10106, 10087, 10088, 10090, 10092, 10093, 10094, 10095), its most interesting feature being its illustration of the fact (p. 172) that in the central part of the Gulf the water was coldest at about fifty fathoms, not on the bottom. Water of 41°-43° filled the sink at the mouth of Massachusetts Bay, rising there to within twenty-five fathoms of the surface; and projected eastward, like a shelf, over the western basin, without any rise in temperature at fifty fathoms as far east as Station 10088; warming to 43.5° in the middle of the Gulf (Station 10090) . In the eastern half of the profile, the coldest water extended from shore, westward into the centre of the Gulf. But on this side there was no water colder than 42°, the low^est reading being 42.°, and the cold mass of water was not horizontal but oblique, rising from a depth of 80-100 fathoms on the shore slope, to 40-60 fathoms at its western end, with the coldest water (42°) Hmited to a very thin layer 40-50 fathoms. The cold layer was interrupted in the middle of the Gulf (Station 10090) by water l°-2° warmer at the fifty fathom level. The temperature of the water underlying the cold zone ranged from 43° to 43.9°, coldest at the eastern side of the Gulf, depth for depth, warmest in the centre (Station 10090), i. c, just the reverse of the temperature at fifty fathoms. Above thirty fathoms the water was warmest at Station 10088, coldest on German Bank and off the mouth of Massachusetts Bay (Station 10087), where the temperature was below 43° at a depth of only twenty-five fathoms. The profile shows the spreading of the curves over German Bank (Station 10095) which characterized that region in 1912 (1914a, p. 56) ; caused by vertical mixing by the tides. And there is a similar phenomenon in Massachusetts Bay (Station 10106) ; limited in this case to depths below ten fathoms. A profile running northeast from the mouth of Massachusetts Bay to Station 10089 (Fig. 22) shows that water colder than 42° extended unbroken across the northern end of the western basin, to the south- tg r— CM CO ^ SoOTCS'—CNJCO-^t-LO CO 180 bulletin: museum of comparative zoology. western slope of Jeffrey's Bank. But it gradually receded from the surface, passing toward the northeast, the curve of 42° dipping from thirty fathoms at Station 10087 to forty fathoms at Station 10089, and to the bottom, in about sixty fathoms, on the slope of the Bank. And there was no water as cold as 42° on the northeast side of the bank. Whether the 42° water was underlaid by warmer water in the Fig. 22. — Temperature profile running northeastward from off Cape Ann (Station 10087) toward Piatt's Bank to Station 10089. northern end of the western basin, as it was further south (Station 10088), is uncertain; but this was probably the case. Profiles running off shore from the western side of the Gulf further delimit the 42° water. The first of these, from the trough between Jeffrey's Ledge and the mainland (Station 10104) to the centre of the FatL Fig. 23. — Temperature profile from the trough between Jeffrey's Ledge and the coast (Station 10104) toward the centre of the Gulf (Station 10090) da StationJlOOSO. Fig. 24. — Temperature profile from the neighborhood of Cape Elizabeth (Station 10103) to Jeffrey's Bank (Station 10091) via Station 10102. 182 bulletin: museum of comparative zoology. Gulf (Station 10090, Fig. 23), shows that below forty fathoms the trough was filled with water colder than 42° : and this was also true as far off shore as the ridge which is crowned by Cashes Ledge. But, as already pointed out, 42° water did not extend to Station 10090. And the fact that at the latter the lowest temperature (43.5°) was at fifty fathoms, not on the bottom, suggests a slight shelf -like projection of the 42° water. It is safe to say that Jeffrey's Ledge rises above the coldest water locally, for in places it is covered by less than thirty fathoms. And tidal currents may be expected to cause temperature Fig. 25. — Tempera tiire profile lengthwise of the trough between Jeffrey's Ledge and the mainland (Stations 10104, 10102). disturbances over it. Between the thirty fathom level and the surface the temperature was nearly uniform, depth for depth, from one end of the profile to the other. A profile (Fig. 24) parallel to the last, but some twenty-five miles further north, from Cape Elizabeth (Station 10103) to Jeffrey's Bank (Station 10091) is warmer at all depths, except the immediate surface, than the preceding one, with water colder than 43° limited to depths greater than seventy fathoms, and a minimum of 42.6° at eighty fath- oms. Between five and fifteen fathoms the difference between the BIGELOW: COAST WATER EXPLORATION OF 1913. 183 two profiles is slight; but below that level it grows progressively greater and greater, as shown by the following table : — Depths Temperature Depths A B 5 fathoms 55° 5 fathoms 10 50° 12-15 18 47° about 30 22 45° 40-50 35 43° 70 (.4 is the profile across the Ledge, and B the profile off Cape Elizabeth). The temperature was almost precisely the same, depth for depth, off Cape Elizabeth (Station 10103) as on Jeffrey's Bank (Station 10091). 43.9' Fig. 26. — Temperature profile from the neigliborhood of Matinicus Island (Station 10101) across Jeflfrey's Bank (Station 10091) toward the centre of the Gulf (Station 10090). 184 bulletin: museum of comparative zoology. But in the middle of the profile (Station 10102) there is a pronounced spreading of the curves between ten and fifty fathoms, which, however, is limited to the mid-depths; it is probably an evidence of local dis- turbances. A profile (Fig. 25) running parallel to the coast (Stations 10104-10102) connecting the preceding two, shows that the 42° water can hardly have extended beyond the northern end of Jeffrey's Ledge, Fig. 27. — Temperature profile from the eastern basin of the Gulf (Station 10093) toward the mouth of the Bay of Fundy (Station 10096). while water colder than 40° (p. 178) was confined to the deeper parts of the trough. A profile (Fig. 26) running off shore from the neigh- borhood of Matinicus Island to the centre of the Gulf shows that the bottom water was distinctly warmer on Jeffrey's Bank (Sta- tion 10091) than in the centre of the Gulf (Station 10090). And a profile from Station 10102, off Monhegan, to Station 10089, would show an even greater temperature-difference between the two ends. BIGELOW: COAST WATER EXPLORATION OF 1913. 185 Evidently then, the immediate coast water from Cape Ehzabeth to and across the mouth of Penobscot Bay was distinctly warmer than the coast water further south, or than the water off shore, water colder than 42° being limited, on the northeast by the slope of Jeffrey's Bank. We have no means of knowing how far south water colder than 42° may have extended in August. But the fact that in early Fig. 28. — Temperature profile across the mouth of the Bay of Fundy, from the coast of Maine (Station 10098) to German Bank (Station 10095) cross- ing the northern end of the basin (Station 10097). July it filled the basin off Cape Cod, from thirty fathoms to bottom, suggests that it reached the northwestern side of Georges Bank, though probabl}" underlaid by warmer water in the southern part of the basin, just as at Station 10088. A profile from the basin toward the Bay of Fundy (Fig. 27) shows that vertical tidal mixing was effective from German Bank to Lurcher Shoal, diminishing toward the north, to reappear again off the coast of Maine (Fig. 28). 186 bulletin: museum of compakative zoology. Salinity, Cape Cod to Chesapeake Bay. 1. Surface salinity. The surface salinity (Plate 2) from Cape Cod to the southern edge of Nantucket Shoals was 32.2%o-32.6%o. And the record of 32.3%o at the eastern end of Vineyard Sound agrees so well with Sumner, Osburn, and Cole's (1913) records for the surface waters of that region in August, 1906 (32.2%o to 32.3%o) that we can Fig. 29. — Temperature profile lengthwise of the northeastern part of the basin from south to north (Stations 10093, 10097). For 42° read 42.5°. assume^that value as normal for summer. The few stations in this region suggest that the curve of 32.3%o swings westward toward the mouth of Vineyard Sound, which agrees with their statement (1913, p. 36) that there is a dominant westerly movement of the water through the Sound of about two knots per day. From Nantucket light-ship out to the edge of the continental shelf there was a steady, and fairly uniform rise in salinity to about 33.4%o BIGELOW: COAST WATER EXPLORATION OF 1913. 187 over the seventy-five fathom curve; and there is every reason to as- sume that by a run of a very few miles further to the south Gulf Stream water of 35%o would have been found. Close to the shore of Long Island the salinity was only about 31.2%o, with an expansion of water fresher than 32.2%o olT its eastern end. And there was a second tongue of comparatively low salinity abreast of Barnegat. On the other hand Gulf Stream water (35%o) was encountered on the surface at the outer edge of the continental slope off New Jersey, with a rise of salinity from 32.4%o to 35.25%o in a distance of only twenty miles (Station 10070 to Station 10071). Close to the New Jersey coast the salinity rose, north to south, from 31.2%o near New York to 32.2%o off Cape May. And the importance of Delaware Bay, like that of the Connecticut and Hudson Rivers, as a source of land water, was shown by the pronounced off shore swing of the curve of 32.2%o abreast of its mouth. At the time of our visit its influence was evident for at least fifty miles from Cape May (Station 10072). The curves show a tongue of comparatively salt water approaching the shore north of Delaware Bay; and a much more pronounced one just south of it, where the curve of 33.5%o lies only thirty miles from land, good evidence that the Delaware water had but little effect either south or north of the Bay in July. The approach of water of high salinity toward the coast south of New York is further illustrated by the fact that off Cape Henlopen the curve of 33%o was within thirty-five miles of land instead of at a dis- tance of eighty miles, as was the case abreast of Long Island. And while this phenomenon is in part a concomitant of the steadily decreas- ing breadth of the continental shelf, the water was Salter over the twenty-five fathom curve off Cape Henlopen than over the 100 fathom curve off Long Island. The freshening effect of Chesapeake Bay on the surface is unmis- takable; the water fifteen miles off its mouth being the freshest (29.25%o) water encountered during the cruise. And the surface salin- ity was only 32.2%o over the 100 fathom curve, though 33.5%o water occupied this relative position on the shelf only thirty miles further north. But the water from the Bay had little effect further seaward, for in the next fifteen miles the salinity rose to 33.5%o, i. e., to practically the same saltness as at the same relative position off Barnegat. The work south of Cape Cod occupied only about three weeks time; hence it is hardly to be expected that any considerable change in salinity would have taken place. x\nd as a matter of fact the stations on the way north show no clear evidence of any. But water samples 188 bulletin: museum of comparative zoology. collected by Mr. Welsh on August 22, near the sixty fathom curve off Block Island (Station 10112) proved to be very much salter (surface salinity 34%o) than the water in this region during the first of July; Salter, in fact, than any water on the shelf at that time, showing that an indraught of ocean water took place in August. During the spring of 1913, Captain McFarland, of the schooner Vic- tor, collected water samples at nine localities between Nantucket and Delaware Bay, seven at the surface, four from 15-25 fathoms, which show that early in June the surface salinity was 32.9%o thirty miles south of Marthas Vineyard, 32.6%o over the southwest slope of Nan- tucket shoals twenty miles west of Nantucket light-ship; and that it was practically unchanged at the latter locality on June 21 (p. 351). Thus the water was salter in June than in July; but while the difference was considerable off Marthas Vineyard (32.9%o as against 32.2%o) it was very slight over Nantucket Shoals (June 6, 32.65%o; June 21, 32.68%o, July 9, 32.5%o). Off Cape May, a few miles south of the location of our Station 10072, Capt. McFarland encountered water of 34.18%o on the surface, and near the bottom at twenty-five fathoms, on May 3 and May 9, which is much Salter than it was there in July (about 32.4%o on the surface). But as the curves show (Plate 2), 34% water would have been reached only fifteen miles further off shore at that season. Apparently, then, the coast water, from Cape Cod to Chesapeake Bay, is freshest in July; and hence, since the outrush of river water is at its maximum in May, seaward expansion must be a slow process. After July, ocean water once more has the upper hand. Salinity sections. The water is usually freshest on the surface, salt- est on the bottom, over the continental shelf south and west of Cape Cod, as, indeed, is the general rule in coastal waters in summer. But at three Stations, 10073, 10074, and 10077, all south of Delaware Bay between the 20 and 30 fathom curves, the intermediate layers, were saltest (Fig. 31, 34). The remaining, more normal, sections fall into several distinct classes. There is, to begin with, one Station (10059) with only a very slight rise in salinity from the surface downward (surface 33.06%o; 30 fathoms, 33.1%o), a type familiar in the northeast part of the Gulf of Maine in regions of strong tidal currents ; its loca- tion on George's Bank, where the currents are proverbially violent, and where temperature like salinity was practically uniform at all depths, shows that it is a similar example of vertical circulation. Judging from the tidal currents, it is probable that more or less similar conditions obtain locally on Nantucket Shoals; but on their BIGELOW: COAST WATER EXPLORATION OF 1913. 189 32 .1 ,2 .3 .4 .5 .6 .7 .8 .9 33 .1 .2 .3 .4 .5 .6 .7 .8 .9 34 .1 .2 .3 .4. -•-- -_. c Y^ \ ^ r~ T~ ~~' "~ ~~ \, y^ \ \ \ ,\ "N \ \ \ i ft: \ \ \ 60 ^ \ " ■Av ^■ \ \ ■^^^ 6^ \ 65 V \ N N \ ^s K 61 1 Fig. 30. — Salinity sections on the continental shelf south of Nantucket andC Long Island (Stations 10060, 10061, 10062, 10063. 10065). f^C ^= ^ ^ ~ °= S ^ — ^ — ""] 1 n ~~ ~-C ■~-" L 1 — r^ \ ) ?n " ^ r^ y ^ / ?■) \ 73 / ^ ?n ■ \ s 74 40 45 50 \ 70 _ _ _ _ _ i Fig. 31. — Salinity sections on the continental shelf south of New York (Sta- tions 10070, 10072, 10073, 10074). 31.2 .3 4 .5 fi ./ .8 .9 32 I .2 .3 .4 .5 .6 .7 .8 .9 33 .1 .2 .3 .4 Fa.O 5 10 Fig. 32. — Salinity sections close to land off New York (Stations 10067, 10068, 10082) and Long Island (Stations 10083, 10084). ^ rr :tz — ^ -_ ... - -^ \ ' J!: ■~~- ~1 ^ 83 ~~ - - ^ ^ 5-- ~i \ ^, 67 '^ ^ 84 6 8 190 bulletin: museum of comparative zoology. southern slope the vertical range of salmity was greater (Station 10060, Fig. 30). A considerable vertical range in salinity, with more or less regular increase from the surface downward, characterized Stations 10062, 10065, (Fig. 30), 10066, 10070, 10072, (Fig. 31), 10075 (Fig. 33), and probably 10067, and 10068 (Fig. 32). And though ra.o 31.8 .9 32 .1 .2 ,3 .4 .5 .6 .7 .8 .9 33 .1 .2 .3 .4 .5 .6 .7 .8 .9 34 = - s ^S - ^ ^ ^ S r= = t- 3 a - _ - ~ 81 --I. 1 75 1 !0 79 69| 1 Fig. 33. — Salinity sections close to land. New York to Chesapeake Bay (Stations 10069, 10075, 10079, 10080, 10081). there was a good deal of variation from station to station in the precise rapidity of increase, as a whole the difference between surface and bottom increased from northeast to southwest. At Stations 10063 (Fig. 30), 10066, 10069 (Fig. 33) and 10082 (Fig. 32), there was a rapid rise immediately below the surface, followed by a bottom zone of uniform salinity, 10-20 fathoms thick. The curves for Stations 10081, 10083, 10084, 10060, 10061, are the reverse, the surface layer being nearly uniform with a rapid rise below. As a whole the water was freshest near shore, saltest over the outer part of the continental _ 29.2 .4 .6 .8 30 .2 .4 .6 .8 31 .2 .4 .6 .8 32 .2 .4 .G .8 33 .2 .4 .6 .8 34 .2 .4 .6 .8 35 "a.U n ~ — — -^ """ — -71 — -J __ n ■" - — — — — — ■~ -_ 1") 78 ; ?n / ?'i / in 77 35 , Fig. 34. — Salinity sections on the continental shelf abreast of Chesapeake Bay (Stations 10077, 10078). shelf, with a progressive rise in salinity from northeast to southwest at stations occupying the same relative positions on the shelf. The salinity sections at the three Stations outside the 100 fathom curve (10064, 10071, 10076, Fig. 35) are all of one type, fresh at the surface, saltest in the intermediate layers, and growing slowly BIGELOW: COAST WATER EXPLORATION OF 1913. 191 fresher once more below 100 fathoms or so. Station 10064 is the fresh- est of the three, with 10071 the saltest, 10076 is intermediate between these two. And they approach one another so closely below 150 fathoms as to suggest that they would have been all alike below that depth, had the stations been located a few miles further off shore. 33 .1 .2 .3 A .5 .6 .7 .9 34 ,1 .2 .3 .4 .5 .6 .7 .8 .9 35 .1 .2 .3 .4 .5 .6 .7 ^■==:— ___^— — — — r; ^"^■~~ — "^^"—-^ \ "^^~^ "~~~-~ V^ ~^^ "^^ ^ ""^"'^ i A I 4 t ^ A 4 it t 41 t #- t U 41 ^ P 444 4lt ftt ^ t-. 44 I r ^ ^ -i ..I . t _j 64 Fig. 35. — Salinity sections at the ledge of the Gulf Stream at Lat. 39° 55' (Station 10064); Lat. 38° 56' (Station 10071) and abreast of Chesapeake Bay (Station 10076). Of the three, Station 10071 most nearly approaches a typical oceanic section; but even here the effect of coast water is evidenced by the fact that the surface salinity is lower than that of the intermediate layers, while Stations 10064 and 10076 both gi^•e a similar result though to a greater degree. Salinity on the bottom. The salinity on the bottom of the shelf 192 bulletin: museum of comparative zoology. (Fig. 36) is of comparatively little importance in oceanography, because so largely dependent on depth; but it can not be neglected because of the part it plays in the biology of the bottom fauna. South and west of Cape Cod the bottom salinity (leaving out of consideration the zone between the shore line and the fifteen fathom contour), ranged from about 32.6%o to 35%o, lowest along the south shore of Fig. 36. — Chart of bottom salinity on the continental shelf between Cape Cod and Chesapeake Bay in July, and in the Gulf of Maine for August. Long Island, and off Block Island, highest, as might be expected, along the outer edge of the shelf. In a general way, it corresponded to depth; but there was also an unmistakable increase, independent of depth, from northeast to southwest. Thus a bottom salinity of 34%o was found at about the seventv fathom curve south of Nantucket, BIGELOW: COAST WATER EXPLORATION OF 1913. 193 at about the forty-five fathom curve off Cape May, and at about the eighteen fathom curve off Chesapeake Bay; and 33. 5%o water at the forty, thirty, and ten fathom contours at the same locaUties. Bottom water fresher than 33%o was restricted to a narrow coastal zone north of Delaware Bay; and the curves for this value and for 33.5%o show evidence of water from the Bay, by swinging seaward off its mouth. But the outflow from Chesapeake Bay has no apparent effect on the curves, although it probably does reduce the bottom below what would otherwise obtain. The chart (Fig. 36) represents July condi- tions only; earlier as well as later in the season, the bottom water was much salter at the few localities where water-samples were taken (34.18%o off Cape May, May 9; 35.17%o in 60 fathoms, southwest of Nantucket August 22). Salinity profiles. A profile (Fig. 37) running from the southern edge of the basin of the Gulf of Maine (Station 10058) across Nan- FiG. 37. — Salinity profile from the southern part of the basin of the Gulf of Maine (Station 10058) across Nantucket Shoals to the outer edge of the continental shelf south of Nantucket (Station 10061). 194 bulletin: museltm of comparative zoology. tucket Shoals (Station 10060) to the edge of the continental slope (Station 10061), shows that the water was much Salter south than north of the Shoals, early in July. In the southern part of the Gulf there was comparatively little increase in salinity with depth below thirty fath- oms, and the bottom salinity was about the same on the Shoals as at , the same depth further north; but the surface shows the influence of the Salter southern water by a steady, though slight, rise in salinity from Station 10058 to Station 10060, as well as in the fact that the Fig. 38. — Salinity profile from the neighborhood of Montaiik Point across the continental shelf (Stations 10063, 10062, 10061) to the edge of the shelf south of Nantucket. average salinity for the upper ten fathoms was higher at Station 10060 (32.65%o) than at Station lOOoS (32.5%o). South of the Shoals there was a rapid rise in salinity, depth for depth, from north to south across the continental shelf. But the Shoals are an effective barrier to any active mixing of water on the two sides below about thirty fathoms. The next profile (Fig. 38) runs across the continental shelf from Montauk Point (between Station 10083 and Station 10087) to the continental slope south of Nantucket Shoals (Station 10061). Its BIGELOW: COAST WATER EXPLORATION OF 1913. 195 ■Fa.Q 10 20 30 40 50 60 70 80 .90 100 no 120 130. 140 150 160 170 180 190 20o' ^ "m ^m. ■y//A :?5^ Fig. 39. — Salinity profile across the continental shelf from New York to the edge of the Gulf Stream in Lat. 39° 55' (Stations 10067. 10066, 10065, 10064). 196 bulletin: museum of comparative zoology. most striking feature, apart from the separation into comparatively fresh water on the continental shelf, and much Salter oceanic water on the slope, is a succession of zones of comparatively uniform salinity alternating with zones in which there is a rapid change in salinity both vertical and horizontal. Next the shore there is first a mass of bottom water of 33.2%o, fifteen fathoms thick (Station 10063), separated by a zone of rapid transition from a much fresher though hardly less uniform surface zone of about 33%o (Station 10062), some twenty-five fathoms thick. This, in turn, gives place to much salter water, over the edge of the shelf (Station 10061), where salinity increases only by .2%o (33.41%o-33.62%o) from the surface down to fifty fathoms; below which there is a sudden rise. Since some of these masses of uniform water reappear in other profiles, it is convenient to designate them from the shore seaward, as A, B, and C. On the profile from the neighborhood of New York to the slope, in about latitude 40° (Fig. 39), the salt ocean w^ater is much more in evidence than it is further east, water of 35%o bathing the slope nearly to the fifty fathom curve, although the surface water at the shore end is about the same salinity as in the last profile (Station 10067, 31 .2%o) . Two of the bands, which were noted in the preceding profile, reappear here, i. e., A. and B, with about the same salinities which characterized them further east. Band A is as well defined as in the preceding profile, occupies the same relative position on the shelf; and has the same salinity (33.2%o)- But in the present profile the transition to the fresher water near shore is less sudden than it was further east. Band B is less clearly defined than in the preceding profile, and its salinity is less uniform, both vertically and transverse to the continental shelf, though of the same general value (about 33%o) ; nor does it so nearly reach to bottom, but overlies a layer of much Salter water. Nevertheless the band is distinctly more uniform than the water immediately below, or on either side of it; hence its indi- viduality still deserves recognition. But the third band, C, which characterized the outer part of the preceding profile, can not be dis- tinguished in this one. As a whole the surface is fresher along this profile than the preceding; and this is true even of its off shore end, although the bottom water near the edge of the shelf is much salter than further east. And not only is water salter than 33.2%o nearer the surface over the middle of the shelf, but water with salinity of 33%o and higher washes the bottom to the fifteen fathom, instead of only to the twenty-five fathom curve. All this shows that off New York shore water was more in evidence on the surface, Atlantic water on BIGELOW: COAST \V.\TER EXPLORATION OF 1913, 197 Fa,Q Fig. 40. — Salinity profile across the continental shelf abreast of Barnegat to the edge of the Gulf Stream in Lat. 38° 56' (Stations 10081, 10069. 10070, 10071). 198 bulletin: museum of comparative zoology. the bottom, than off Montauk or south of Nantucket Shoals; and that the transition between the two waters was very sudden. In the profile from off Barnegat to the continental slope in about latitude 39° (Fig. 40), water of 35%o washes the slope below about sixty fathoms, and the curve of 35%o, which may be taken as an arbitrary division between coast and Gulf Stream water, is almost vertical. Generally speaking, too, the surface water was Salter along this whole profile than in the preceding one, except at Station 10070; an exception ex- plained by the fact that this part of the profile cut the southerly tongue of surface water fresher than 32.4%o, noted above (p. 187, Plate 2). Neither band B nor C can be traced as far south as this profile. But Band A is still evident, with precisely tlie same salinity (33.2%o) as in the two preceding profiles, washing the bottom rather nearer shore than was the case further north, and gradually merging into the Gulf Stream water of 35%o on its off shore side, instead of being limited sea- ward by a sudden transition zone. On the other hand there is a great difference in salinity between it and the surface water over it, and also between it and the zone of water closer to shore. The partial profile off Cape May is instructive chiefly be- cause it shows no sign of band A ; hence it is safe to conclude that the latter comes to an end north of Delaware Bay. The profile is otherwise so much like the preceding one, that I have not thought it necessary to reproduce it here. But the next one (Fig. 41), which is south of Dela- ware Bay, reveals an entirely new phenomenon, namely, a tongue of salt off shore water with salinity of 35%o or more, intruding into the intermediate depths over the continental shelf, with fresher water both above and below it. Its landward end lies about over the thirty fathom curve, where the bot- tom water has a salinity of about 34.3%o, with 33.24%o on the sur- face. Apart from the salt tongue, the salinity as a whole is higher Fig. 41. — Salinity profile across the continen- tal shelf south of Delaware Bay (Stations 10079, 10074). BIGELOW: COAST WATER EXPLORATION OF 1913. 199 than in the preceding profile, the bottom sahnity in fifteen fathoms being 33.86%o as against 33.14%o, with 34%o as against 33.5%o on the bottom at the twenty-five fathom curve. The shallower layers, too, are salter, depth for depth, than north of Delaware Bay. A similar shoreward intrusion of 35%o water into the intermediate iG. 42. — Salinity profile across the continental shelf to the edge of the Gulf Stream, abreast of Chesapeake Bay (Stations 10078, 10077, 10076). 200 bulletin: museum of comparative zoology. depths over the shelf is also to be seen in the profile abreast of Chesa- peake Bay (Fig. 42) ; and it has about the same extent and conforma- tion there as further north, the curve of 35%o rising from the sea floor at about the fifty fathom curve, with fresher water underneath it. But here the water near shore was much fresher dowm to five fathoms than in the preceding profile ; the immediate surface layer fresher than any water we encountered further north, as might be expected from the volume of river water which debouches from the Bay in spring. And though this layer was very thin, the salinity rising from 29.25%o on the surface to 33.5%o on the bottom in ten fathoms at the shore end of the profile, its influence is unmistakable out to the edge of the continental shelf. At the outer end of the profile (Station 10076) the water was saltest at 50-100 fathoms (about 35.4%o), just as at the other deep water stations; below that level salinity decreased very slowly, as it does over the north Atlantic as a whole. The change in salinity from north to south over the shelf north of Delaware Bay is illustrated by a profile following the forty fathom contour from Nantucket Shoals (Station 10060) to Station 10070 (Fig. 43). Below about ten fathoms there is a general increase in salinity, depth for depth, from northeast to southwest. But the surface water is freshest at the southern end of the profile (32.2%o), saltest at Station 10062 (32.86%o), and fresher once more (32.63%o) over the slope of Nantucket Shoals. Salinity in the Gulf of Maine. Surface Salinity. Early in July the surface salinity (Plate 2) of Massachusetts Bay, immediately off Gloucester, was about 31.56%o, arise of about .5 since the middle of May (1914b, p. 393), and it was 31.9%o off Cape Cod (Station 10057, p. 205) with 32.4%o over the southern part of the basin (Station 10058), and 33%o on the southwest side of George's Bank (Station 10059). When we returned to the Gulf of Maine a month later, the water was slightly salter along the eastern shore of Cape Cod (32.05%o, Station 10085; 32.09%o, Station 10086), while a greater increase of salinity had taken place off Glouces- ter (to 32.03%o). And by the 25th of August it had risen to 32.16%o in the mouth of Massachusetts Bay (Station 10106). The water immediately abreast of the Bay and along Cape Cod (Plate 2) was 32-32. 2%o, the curve for the latter value swinging eastward from the mouth of Vineyard Sound, and then northerly, toward Penobscot Bay. CO i6 202 bulletin: museum of comparative zoology. Water fresher than 32%o was restricted to a narrow zone close to shore, extending from just north of Cape Ann to Monhegan Island, broadest (twenty-five miles) off Cape Elizabeth. In general there was a rise of surface salinity from west to east across the Gulf, the water being 32.5%o some sixty miles off Cape Cod; n, n 3 . . . .A .1 . u'' .8 .9 3 ! .1 .2 .i .5 .6 .7 .8 .9 3 i . . I . .i ij c s. ^ ^ in ^ \ ^:^ N S s\ ^ ^^^ 0(1 n\ w \ \ \ \ \ \ \ \ \ \N ^ s\ 106 \ \ ■* N \ \ \ CO N \ \^ \ S 70 w \ \ 8 7 V s \ \ \ \ \ \ \ N,^ \ \ \\ \\ ,9: \\ U ) IJU \^ \ Fig. 44. — SaUnity sections in the Gulf of Maine, from Massachusetts Bay to the eastern basin (Stations 10087, 10088, 10090, 10092, 10093, 10106). 32.6%o in the centre of the Gulf, and 32.7%o near the Nova Scotia coast bank. But the increase was far from uniform, the course of the curves being distorted by an outrush of comparatively fresh water (32.2%o to 32.5%o) off the west mouth of Penobscot Bay, and by a BIGELOW: COAST WATER EXPLORATION OF 1913. 203 band of water of the same low salinity extending thence along the coast of Maine to the Grand Manan Channel. The salinity was 32.5%o or less over the coast bank west of Nova Scotia ; and it is probable that the surface of the Bay of Fundy was even fresher than this. The curve for 34.4%o shows that the direct effect of Penobscot water did not extend further south than Jeffrey's Bank (Station 10091), south of which it runs in an S, roughly parallel with the coast, crossing the southern end of the basin, and thence westward across Nantucket Fr^ 31.8 . 9 32 . 2 . 3 . » . B . 7 . 3 . 3 33 .1 2 3 .4 .5 6 .7 5 ^ =^ \ s. S \ ■* s tn ^ ;^ s \ \ \ 15 ^ \ \ \^ ?n V > \ N^ ^ ?5 \^ ^ Ni \ 30 V V k \\ 35 \ \; N^ \ \ s _,.... ^^^^_^_„.___._-:r::rrrr..=^ ^^^^^^^^^^r---^^ ^^^^^^^^^^^^ ^^ Trrr ^^^^^^^^^^^^^^^ '' ''- •• -'Wf^^^^^^^,^ ''--^^^^^^^^^ ^7 ''':W^^^ ''^^^. Fig. 57. — Density profile across the continental shelf southwest i Nantucket (Stations 10063, 10062, 10061) July 10-11. (Fig. 59), shows a similar distribution of density, except that the sur- face, as well as the deeper water was densest at the seaward end, the dip of the curves being especially pronounced in the upper fifteen fathoms or so, and again at 40-50 fathoms over the continental slope. A profile running from Station 10079 to Station 10074 (Fig. 60). shows that just south of Delaware Bay where the surface water was lightest next the coast, the reverse was true below about twelve fathoms, the bottom water being heaviest, depth for depth, next the land, while the seaward dip of the curve of 1.026, suggests a seaward 218 bulletin: museum of comparative zoology. Fig. 58. — Density profile from New York to the edge of the Gulf Stream in Lat. 39° 55' (Stations 10067, 10066, 10065, 10064) July 11-13. BIGELOW: COAST WATER EXPLORATION OF 1913. 219 Fig. 59. — Density profile across the continental shelf abreast of Barnegat to the edge of the Gulf Stream in Lat. 38° 56' (Stations 10069, 10070, 10071) July 19-20. 220 bulletin: museum of comparative zoology. flow over the bottom. And the level at which density is uniform, horizontally (twelve fathoms) exactly coincides with the salt tongue (p. 198). The profile abreast of Chesapeake Bay (Fig. 61) shows a similar distribution of density over the inner part of the continental Fig. 60. Fig. 60. — Density profile across the con- tinental shelf south of Delaware Bay (Stations 10079-10074) July 22-30. Fig. 61. — Density profile across the con- tinental shelf abreast of Chesapeake Bay (Stations 10078, 10077, 10076) July 24-29. Fig. 61. shelf. But the seaward rise of density on the bottom is less rapid than it is further north; and density is uniform, horizontally, below twenty-five fathoms. BIGELOW: COAST WATER EXPLORATION OF 1913. 221 Density in the Gulf of Maine. In the Gulf, in August (Fig. 55) the surface water was lightest close to shore north of Cape Ann (1.0231), off Cape Cod (Station 100S5, 1.0231), and, in an isolated region, over the western basin (Station lOOSS, 1.0229) ; the latter was a local phenomenon, due to high surface temperature. Surface density was highest on German Bank (1.0254) and along the northern part of the coast of Maine (1.025), i. e., in those regions where tidal currents cause the most effective vertical mixing of the water. And the surface was only slightly less dense off Lurcher Shoal, owing to its low surface temperature. We likewise encountered surface water of high density off Matinicus (Station 10101, 1.0248); And no doubt many other anomalies of this kind might be found in the Gulf, caused by local surface cooling by tide rips and vertical currents. The surface density of most of the Gulf was 1.0236-1.0248, increasing from southwest to northeast; i. e., considerably higher than over the continental shelf south of Cape Cod a month earlier; had the observa- tions been taken simultaneously the discrepancy would have no doubt been greater, it being only reasonable to assume that the surface of the Gulf would have been cooler early in July than early in August, but with nearly the same salinity (1914a). The table of density (p. 344) shows that the water was lightest at the surface, heaviest on the bottom, /. e., was in stable equilibrium, everywhere in the Gulf. Where vertical and tidal circulation is active, as on German Bank, the stability was so slight as to offer little resist- ance to vertical overturning of the water. But where tides are weak, as for example off Massachusetts Bay, over the western basin, and in the trough west of Jeffrey's Ledge, the difference between surface and bottom density, and hence the vertical stability, is great. In the western parts of the Gulf in general there was a very rapid rise of density from the surface down to about 20-30 fathoms, corresponding to the rapid rise of salinity and fall in temperature in this zone; fol- lowed by a very much slower, though continuous increase, down to the bottom. But the density curves, like those for temperature are pro- gressively straighter and straighter, passing across the Gulf from south- west to northeast. And in the northern end of the eastern basin, as well as on the Nova Scotian and Maine banks, the rise in density, whether great or little, was nearly uniform in rate, from surface to bottom ; most nearly so where the stability of the water was slightest {i. c, German Bank). T— CM CO 1^ 0003^»— CNJCO'^ BIGELOW: COAST WATER EXPLORATION OF 1913. 223 The density profile (Fig. 62) crossing the Gulf from Massachusetts Eay (Station 10106) to German Bank (Station 10095) shows that the water was nearly uniform horizontally, depth for depth, below seventy fathoms. In the mid-depths the water was densest at Station 10092. Over German Bank there is a distinct spreading of the curves reminis- cent of, and due to the same cause, as the spreading of the tempera- ture and the salinity curves in that region. And the same condition Fig. 63. — Density profile from the mouth of Penobscot Bay (Station 10102) to the centre of the Gulf of Maine (Station 10090) crossing Jeflfrey's Bank (Station 10091) August 10-14. prevails below twenty fathoms in Massachusetts Bay, just as described for salinity (p. 208). A profile from Station 10102, near Penobscot Bay, across Jeffrey's Bank to the centre of the Gulf (Fig. 63) shows a slight rise in density passing off shore, the difference being greatest in the mid-depths. But a parallel profile further east would be exactly the reverse, the surface density being higher at Stations 10101, 10100, and 1009S than at either Station 10092 or 10093. 224 bulletin: museum of comparative zoology. Color of the sea. The observations on color, tabulated below, are interesting chiefly because there is very little precise information as to the color of the water over the continental shelf south of Cape Cod. Color, in % of yellow, according to the Forel scale {Steuer, 1910). station Color Station Color Station Color Station Color 100.57 27 10070 5 10083 20 10096 20 10058 9 10071 2 10084 27 10097 — 10059 20 10072 9 10085 27 1009S 20 10060 5 10073 2 10086 27 10099 27 10061 2 10074 5 10087 14 10100 27 10062 9 10075 20 10088 — 10101 35 10063 20 10076 2 10089 — 10102 20 10064 2 10077 9 10090 9 10103 — 10065 5 10078 14 10091 20 10104 20 10066 — 10079 14 10092 9 10105 20 10067 54 10080 14-20 ' 10093 — 10106 — 1006S 54 10081 9 10094 27 10069 27 10082 — 10095 27 The water was very green (27% yellow) along Cape Cod both in July and in August, and this was also the case on the western side of George's Bank (20%). But it was distinctly bluer (9% yellow) over the southern end of the basin of the Gulf and after crossing Nantucket Shoals the water grew visibly blue to the eye, being almost pure blue (2% yellow) at the 80 fathom curve south of Nantucket (Station 10061). In general the water was greenest near land, bluest off shore, as might be expected, the water being greenest of all near New York (Stations 10067, 10068). The color wj^is 20-27%, yellow, along the coast of New Jersey; that of the coast water south of Delaware Bay 14-20% yellow. The water was nearly pure blue (2% yellow) at all the stations outside the edge of the continental shelf. The water of the Gulf of Maine was considerably greener, most so along Cape Cod (27% yellow), over German Bank (27% yellow), and along the coast of Maine between Mt. Desert and Penobscot Bay (27-35% yellow). The water was considerably bluer (9%) over the BIGELOW: COAST WATER EXPLORATION OF 1913. 225 deep basins; but nowhere in the Gulf did we find the beautiful ultra- marine water which washes the continental slope. South of Cape Cod the general rule is that the water is bluest where saltest, greenest where freshest; though this does not exactly cover the case, because the water was bluer close off Chesapeake Bay than off New York, although the salinity was lower. But in the Gulf of Maine this rule did not hold either in 1912 (1914a) or in 1913, the greenest water being intermediate in salinity, while the saltest water was not the bluest. Current measurements. Measurements of surface and bottom currents with the Ekman Current meter (Ekman, 1905b) were taken at three stations between Cape Cod and Chesapeake Bay, with hourly readings for six hours at each station. The directions are the compass bearings (magnetic) toward which the current flows. Velocity in knots per hour is to the nearest tenth of a knot. I. Station 10065, July 12. High water at Fire Island Inlet at 2.05 p.m. Hour Depth Duration C. C. per sec. Direction Knots per hour 9 a.m. 4' 58" 19.1 WNW. .4 9 40 5' 10" 10.2 NW. by N. 2 10 5' 9.3 WNW. .2 10 40 2' 20" 22.1 W. by N. .4 11 5' IS" 10.7 W. by N. .2 11 40 5' 10" 3.2 NW. Trace 12.30 P.M. 5' 5" 26.9 NNW. .5 12.30 " 40 5' 45" 27.9 S.byE. .5 2 " 4' 24. NE. .4 2 40 4' 45" 24.9 S.byE. .5 2.45 " 5' 33.3 NE. .6 2.45 " 40 5' 5" 15.1 S. .3 226 bulletin: museum of comparative zoology. II. Station 10072, July 21. Low water Barnegat Inlet at 4 a.m. 1.46 A.M. 24 5' 10. S. by E. .2 2.15 « 4 5' 12.4 S. by W. 2 2.30 " 5' 7.3 ssw. .14 3 24 5' 7.2 wsw. .14 3.15 " 4 5' 7.3 S. by W. .14 3.20 " 5' 7.2 SSW. .14 4 24 5' 7.3 NNW. .14 4.15 " 4 5' 7.7 w. .15 4.20 " 5' 27.3 ssw. .5 5 24 5' 25.3 S. by W. .5 5.15 " 4 5' 34.3 W. by S. .7 5.25 " 5^ 38.2 S. by W. .74 6 24 5' 17.3 N. by W. .3 6.15 " 4 5' 36.8 W. .7 7 24 5' 14.1 N.byE. .3 7.15 " 4 5' 36.3 W. by N. .7 7.30 " 5' 36.3 W. by S. .7 8 24 5' 13.2 NE. by N. .3 8.15 " 4 5' . 28. WNW. .54 III. Station 10074, July 22. High water Cape May 11 a.m. High water Barnegat 10.35 a.m. 7.45 A.M. 30 5' 5.9 S. by E. .1 8 5' 30. W. .6 8.45 " 30 5' 2.8 ? Trace 9 5' 28. 8 W. .55 9.45 " 30 5' 9.7 SSE. .2 10 5' 20. NW. by W. .4 10.45 " 30 5' 9.7 S.byE. .2 11 5' 9.7 NNW. 2 11.45 " 30 5' 16.3 SSE. .3 12 5' 10.5 N. by W. 2 1.10 P.M. 30 5' 7.7 S|W. .15 1.20 " 5' 4. NNE. .1 2 " 30 5' 9.9 S. .2 2.10 « 5' 18.3 ENE .35 3 30 5' 5.6 SSE. .1 3.10 « 5' 14.3 E^N. .3 BIGELOW: COAST WATER EXPLORATION OF 1913. 227 At Station 10065, over the 45 fathom curve, fifty miles south of Long Island, the first reading was taken about five hours before high water at Fire Island Inlet, the nearest shore station for which tidal data is available. The surface current ran northwesterly for the first three hours; and then veered to the north and northeast, in which direction it flowed, till the end of the set. Of course the observation does not show conclusively whether or not there was a dominant drift in any direction, because it did not cover the last half of the ebb; but it goes far enough to show that the flood current ran about northwest; the first half of the ebb to the northeast, the strength of the flood being .2-.6 knots, of the ebb .4-.7 knots per hour (Fig. 64). The total drift for the part of the tide covered by the set is about 1.8 knots north. x\nd it seems hardly probable that the last few hours of the ebb would wholly nullify this, the general trend of the coast in this region being such that it is safe to assume that the last part of the ebb flows about east, the first part of the flood westerly. And even if the late ebb ran southeast with a velocity of .5 knots, there would still remain a net northerly drift of nearly .5 knots. It is therefore fair to conclude that there was a slight dominant northerly movement of the surface water over this part of the continental shelf. The bottom current turned an hour earlier than the surface current. Dur- ing the last three hours of the flood the flow on the bottom was toward the northwest, with a velocity dimin- ishing from .4 knot to zero. It then veered to the south by east, and south, running in that direction for three hours with the considerable velocity of .35-.5 knot per hour. The total set showed a net move- ment of water of about 1.4 knot toward the south-southwest; l:)ut it is a c^uestion whether there was any dominant flow on the bottom, for if the current veered to the southeast and east during the last of the ebb, with a northwest current throughout the flood, as is not unlikely, the net drift would be neutralized. Fig. 64. — Surface current , and bottom current at Station 10065; hourly from 9 A.M. to 3 P.M., July 12. The distance between dots (.) shows the drift for each hour; 2.25 cm. = 1 sea mile. The compass arrows are trxie and magnetic. 228 bulletin: museum of comparative zoology. Fifty miles off Cape May (Station 10072) readings were taken at zero, four, and twenty-four fathoms, from 1-46 a.m. to 8-15 a.m., the time of low water being 4 a.m. at Barnegat Inlet (Fig. 65). The surface current ran southwest during the entire set, veering toward the west (S. S. W. to W. by S.) with velocities ranging from .15 knot at the beginning to .7 knot at the end, showing that the tide started to flood shortly before we began work. The total drift was about Fig. 65. — ■ Surface current , 4 fathom current — • — •, and bot- tom current at Station 10072; hourly from 2 a.m. to 8 a.m., July 21. The siirface current of Station 10074,. . . . , is combined with Station 10072 to show total drift for an entire tide. 2.25 cm. = 1 sea mile. 3 knots southwest. At four fathoms the current veered from S. by W. through west, to W. W., N. the velocity ranging from .14 knot to .7 knot, the net drift 3 knots west, i. e., toward shore. The bottom current at twenty-four fathoms veered irregularly from S. by E., through S. W. west, and northwest to northeast, with velocities BIGELOW: COAST WATER EXPLORATION OF 1913. 229 ranging from .14 knot to .5 knot, greatest when the flow was south- westerly and northwesterly. The total drift was about 1 knot to- ward the northwest. These three sets were planned to cover the last half of the ebb, and the first half of the flood. But the observations show that the flood current had begun to run one to two hours earlier than the time of low tide at Barnegat. Hence, the set must have been confined to the flood, and therefore can not show whether there was any dominant drift. To remedy this defect it would have been necessary to continue the set for six hours more, but this was impracti- cable, owing to a sudden squall. Consequently a third set of current measurements was made the next dav at Station 10074, so timed as to 3 P.,Vi. ■■■,3 P.M. Fig. 66. — Surface ciorrent , and bottom current. .... at Station 10074, for each hour from 8 a.m. to 3 p.m., July 22. 2.25 cm. = 1 sea mile. cover the last of the flood and most of the ebb (Fig. 66). The surface current at Station 10074 set westerly for the first two hours, i. e., during the last of the flood. It then veered gradually through north- west, north, and northeast to east, in which direction it was running with a velocity of .3 knot at the end of the set. The velocities were .6 knot for the first two hours; .1 knot to .4 knot after that. The total net drift was about 1.5 knot to the northwest. The bottom readings were less satisfactory than those on the surface, because of the weakness of the current. In general the flow was toward the south and south-southeast, varying irregularly between 230 bulletin: museum of comparative zoology. these two bearings, the total drift being about 1.2 knot toward the south by east. On combining the stations, omitting the first hour of 10074 to compensate for the advance of the tide during twenty-four hours, a southwesterly surface drift of 2.2 knots and a southwesterly bottom drift of about 1 knot results. The last two hours of the ebb are still to be accounted for; the regu- lar veering of the surface current suggests that it continues to swing toward the east and southeast, and general knowledge of similar tidal currents suggests a diminishing velocity. These two stations, then, taken together, indicate a dominant southwesterly current with a velocity on the surface, of two to three knots for an entire tide, i. c, four to six knots in twenty-four hours. Of course the validity of this conclusion depends on whether a combination of these two sets of observations, as though they had been made at one station, is justified, and there is no apparent objection to so doing, either in the contour of the bottom, the course of the shore line, or in the amplitude of the tide at the two stations. Nor was there anything in the weather conditions to suggest that the surface current was a wind drift at either, because Station 10074 was occupied during a calm, and after a calm night; Station 10072 likewise after a calm night, and in a moder- ate breeze. And so far as the observations go, the velocity of the tidal currents was apparently about the same at the two stations, being about .7 knots per hour for the fifth hour of the flood at Station 10072, .6 knot at Station 10074. The bottom currents likewise sug- gest a slight southwesterly drift. Circulation over the continental shelf, July, 1913. Our current measurements, salinities, and densities allow a tentative reconstruction of the movements of the water over the continental shelf at the time of our visit. During the spring there must have been off shore surface currents opposite four main sources of fresh land water, i. e., Long Island Sound, the Hudson River, Delaware Bay, and Chesapeake Bay, to produce the tongues of low salinity which we encountered there (Plate 2). These currents must have been at their height at least a month earlier, i. c, at the time of the greatest river freshets; the Delaware current reaching its maximum after the middle of May, because the salinity was higher ofT the Bay on May 9 (p. 188) than we found it (p. 198). The drift, as indicated by salinity, was easterly off the mouth of Long Island Sound, and there must have BIGELOW: COAST WATER EXPLORATION OF 1913. 231 been a similar, but more pronounced off shore current opposite Chesa- peake Bay, much as it is represented on the current chart of the North Atlantic (Soley, 1911), and surface density suggests that the fresh water from the Bay spreads out, fan-like, to the north, as well as over the heavier ocean water. The Salter water which alternates with these comparatively fresh tongues is in part a contrast phenomenon; but the salinity curves immediately south of Delaware Bay can only be explained as due to an actual shoreward drift of water of high salinity (p. 187). And the current data at Station 10074 suggest, th'ough they do not prove, that this salt tongue was swinging, eddy-like, toward the southwest. Just north of Delaware Bay, there seems to have been a similar eddy-like movement which, added to the southerly flow of coast Avater, produced the strong southwest current which was found at Station 10072. Surface salinities, like the current measure- ments at Station 10064 suggest traces of a northerly movement, or "banking up" of the ocean water south of Long Island, a process which had progressed so far by the end of August as to raise the surface salinity from about 32.8%o (Station 10062) to about 34%o (Station 10112). Surface density, being practically the same off Cape Cod as over the outer part of the continental shelf south of Nantucket, does not indicate any general flow across Nantucket Shoals into the Gulf of Maine in July, or vice versa; nor does surface salinity afford any un- mistakable evidence of a dominant current in that region, though the curve of 33% suggests a possible southeasterly drift. Salinities show that there must have been an indraught of ocean water into the eastern side of the Gulf, which is consistent with the fact that the surface density of the northern and eastern parts of the Gulf was very much high^ than that of the ocean water outside George's Bank. To compensate for this tongue of ocean water, there was an outflow of land water off Penobscot Bay; and the salinity curves suggest a general southward drift of surface water along the western coast of the Gulf (Plate 2). The salinity curves, and our actual current measurements, agree very well with the earlier data, as summarized in the U. S. Coast Pilot (1912). According to the latter the prevailing drift over Nantucket Shoals is easterly, which agrees so well with our salinity curves as to make it a fair assumption that there is actually a dominant easterly current in this region in summer. The few current measurements which have yet been made on George's Bank (U. S. Coast Pilot, 1912, Mitchell, 1881) indicate a similar easterly drift, veering northward 232 bulletin: museum of comparative zoology. near the eastern edge of the Bank. And although the observations are insufficient for any definite mapping of currents in a region where the tides are so strong, it is certainly suggestive that this northerly trend near the eastern end of the Bank corresponds with the salt tongues which were found in the eastern side of the Gulf in both 1912 and 1913. But an easterly and northeasterly movement of water on the Shoals and over George's Bank, does not mean that there is a general easterly long-shore current, both because there is no dominant drift at Nan- tucket light-ship (U. S. Coast Pilot, 1912, p. 10), and because the various records agree in crediting the coast waters south of Marthas Vineyard as a whole with a westerly, southwesterly, or northwesterly drift. In short, present indications point to the conclusion that the movements of surface water are tidal there, in the form of an irregular, perhaps intermittent eddy, which receives greater or less accessions of Gulf water on its northern side, and of ocean water along its south- ern and southeastern edge. The latter is an important factor in summer when it must influence hydrographic conditions on the banks profoundly, just as it does over the continental shelf further west (p. 198). And it exerts an unmistakable influence on the oceanog- raphy and plankton of the Gulf of Maine as well. The outrush of comparatively fresh water from Long Island Sound, shown by the salinity curves, is substantiated by current records; and the northwesterly current over the forty fathom curve south of Block Island, represented ' on the current chart in the Coast Pilot, corresponds with our current records over the same part of the shelf a few miles further west. But the changes which take place in the surface salinity of this region at different seasons show that it is by no means a permanent phenomenon, probably being reversed in spring by the outrush of shore water. The combined evidence of the various records of ocean currents, our own included, points to the conclusion that the dominant drift over the continental shelf, south of New York, is to the southwest; and this is certainly the prevalent opinion of practical navigators and hydrographers. But it does not necessarily follow that this drift is a simple, long-shore current, as has so often been suggested. On the contrary, surface salinity shows that it is interrupted by outpourings of comparatively fresh water off the rivers and bays, at least in spring and summer, and, conversely, by shoreward movements of salt ocean water. Furthermore little evidence was found of any appreciable southerly flow on the bottom, even in water as shallow as twenty-four fathoms, though there was an unmistakable southwesterly current on BIGELOW: COAST WATER EXPLORATION OF 1913. 233 the surface. The correct explanation is that the movement of the surface waters over the shelf is chiefly a series of great eddies, receiving water, on the one hand from the Gulf Stream off shore, on the other, from the land. The accompanying chart (Plate 2) shows an attempt to reconstruct the surface currents, for the summer months; but so intricate is the problem, and so scanty the reliable information yet at hand, that it is only tentative. It is even more difficult to reconstruct the movements of the sub- surface water, because we must rely almost wholly on the Grampus observations. These current measurements do not prove any domi- nant flow on the bottom north or south of Delaware Bay (p. 230), and it is questionable whether any general flow can be deduced from them south of Long Island. But salinity, density, and temperature show that the bottom and intermediate waters over the shelf are far from being stagnant, though their movements, other than tidal currents, are probably slow as compared with the surface currents. The density profile across Nantucket Shoals does not suggest any flow into, or out of the Gulf of Maine in this region at any depth; nor does the density of the bottom water of the Gulf suggest any influx of ocean water from the zone between fifty and 130 fathoms, via the Eastern Channel. The seaward dip of the density curves south of Nantucket together with the cold tongue (p. 165) shows that the bottom water was flowing seaward down the shelf from the fifty fathom curve, indenting into and mixing with the ocean water over the slope (Fig. 10) ; and this agrees with the salinity curves. But south of Long Island, the fact that the density curves are just the reverse, together with the sudden rise of salinity immediately below the cold tongue, suggests tha,t here the ocean water was sinking, obliquely, toward the land below the cold, fresh coast water. And to judge from the densities, a similar movement of water must have been taking place over the outer part of the shelf off Barnegat also. The salt tongue which indents the fresher coast water in the mid- depths over the continental shelf between Delaware Bay and Chesa- peake Bay (p. 198) is as interesting as the cold tongue off Long Island. Just south of Delaware Bay, there seems to have been an actual move- ment of surface water toward the coast (Fig. 60), gradually mixing with and sinking below the much fresher, hence lighter coast water. At twelve fathoms, i. e., the axis of the salt tongue, the density was uni- form, east and west; below twelve fathoms, the density gradient dipped from land to sea. Thus ocean water must have been coasting, 234 bulletin: museum of comparative zoology. as it were, down the density gradient, from near the surface over the 100 fathom contour to about twelve fathoms over the thirty-five fathom contour, with the heavier, though fresher, bottom water of the shelf moving seaward below it. Density points to a similar type of circulation off Chesapeake Bay. But this phenomenon must be transitory, because as the coast water grows warmer with the advance of the season its density on the bottom must fall as low as that of the Salter water off shore. The band of uniform salinity which we traced from Station 10063 to Station 10069 (p. 194) was not the result of vertical mixing; had it been temperature like salinity would have been equalized. Its origin is obscure. Neither density, salinity, nor temperature indicates any general longshore movement of the bottom waters on the shelf. Previous records of temperature and salinity Cape Cod to Chesapeake Bay. The existence of a band of cold water between the Gulf Stream and the coast has been recognized since the days of the early voyages to these shores. By 1850 its general geographic limits were well under- stood (Maury, 1855), since which time a vast body of surface tempera- ture readings has been taken over the continental shelf by vessels entering the ports of New York, Philadelphia, and Chesapeake Bay, as well as by various expeditions and government services. But most of these have never been published; and since, in any e\'ent, the general range of summer temperature is now well known, I need refer here to only a few of the more important sets of observations. The data obtained by the U. S. Fish Commission south of Marthas Vineyard between 1880 and 1882, (Tanner, 1884a, 1884b; Verrill, 1880-1884b), show the general rise of temperature passing off shore from the southern coast of New England. And records have con- stantly been kept at .Woods Hole since that time, so that there are very satisfactory data of the temperature close to shore in that region. The more recent of these are summarized by Sumner, Osburn, and Cole (1913), who find that the monthly surface mean for a five year period, at the Woods Hole Station, is 31° in February, 43.9° in April, 68.8° in July, 69.7° in August, 48.2° in November. In Vineyard Sound the mean surface temperature, August, 1907, was 64.7°, Novem- ber, 1907, 50.9°; March, 1908, 36.6°; June, 1908, 56.5°. The surface BIGELOW: COAST WATER EXPLORATION OF 1913. 235 temperatures of the water close to the coast south of New York are likewise well known for all months in the year (Rathbun, 1887) owing to the extensive series of temperatures taken at various light-houses and light-ships, notably " Winter quarter Shoal," and " Five fathom bank " oflF Cape May from 1881-1885. At the former, on July 25, i. e., about the time the Grampus passed there, the temperature was 74° in 1881, 72° in 1882 and 1883, 69° in 1884, and 74° in 1885. Our records, a few miles away, July 21, were 74°-75°. But by July 30, a surface temperature of 76° close to the light-ship was noted. At "Five fathom bank" the temperature, on July 25, 1881, was 71°; 73° in 1882; 71° in 1883; 74° in 1885: on July 21, 1913, it was 73°-74°, a few miles to the east, rising to 77° close to the light-ship on July 31. Off Sandy Hook, July 19, the surface temperature was 71° in 1881 and 1882;' 74° in 1883; 65.5° in 1884; 69° in 1885. On July 17, 1913, it was 68°-69°. On Nantucket Shoals, 40° 54' N., 69° 49' W., i. e., some seventeen miles north of the present location of the light-ship, the surface temperature, July 10, ranged from 55° to 60° for the five-year period. On July 9, 1913, it was probably about 56°, i. e., about the same; but the Grampus did not visit this exact spot; and the surface temperature varies with the greater or less violent tides over the shoals. The general summer temperature of the water over the outer part of the continental shelf is now well known for the region south of Martha's Vineyard, thanks to Verrill (1880-1884b) and Libbey (1891, 1895). In July and August, 1881, the surface temperature south of Marthas Vineyard was slightly cooler than in 1913, from 63° over the forty fathom curve, to 66° over the fifty fathom curve, and 72° over the 100 fathom curve (Verrill, 1881, 1884b), whereas on July 11, 1913, a few miles further east, it ranged from 65°-67° between the forty and the seventy-five fathom curves. Over the 100 fathom curve, on the other hand, the 1913 temperatures are a little the lower (69°-70°, as against 72°). Unfortunately Verrill's data for 1882 are not directly compar- able, because taken in August. But during that month the surface water outside the sixty fathom curve south of Marthas Vineyard was constantly warmer than 70°, /. e., about as much warmer than the year before for that month, as 1913 was warmer than 1881 in July. In 1889 Libbey (1891) took an extensive series of surface temperatures south of Block Island and Marthas Vineyard, affording the most complete temperature survey of a limited locality yet attempted off the American coast. Any analysis of these records would require a 236 bulletin: museum of comparative zoology. study of diurnal warming and nocturnal cooling, to make them com- parable with one another. But this is not necessary here, because, after all, they are not strictly comparable with our observations, having been taken from three to six weeks later in the season, and hence may be expected to be higher. They suggest that the surface water in that year may have been rather cooler than we found it, for Libbey (1891) found much the same temperature at the end of July that was observed in the first half of the month; i. e., July 24, 1881, 62.8° at the 25 fathom curve south of Nantucket; 66°-67° over the outer half of the shelf; 68° at the 100 fathom curve. A large number of surface temperatures have been collected by Dickson (1901) for the years 1896, 1897. In July 1896, according to his charts, the surface temperature from Marthas Vineyard to New York was between 60° and 68°, above 68° off New Jersey. By August it had risen to 68°, the greater number of the records having been taken, no doubt, along the direct steamship line from Nantucket light- ship to Fire Island light-ship. In 1897 the water was warmer, being upwards of 68° from Nantucket to New York. According to the British Meteorological office (Sumner, Osburn, Cole, 1913, p. 438) the mean surface temperature, some thirty miles south of Marthas Vine- yard, is 67° in July, 69° in August: 66° and 70° respectively in Long Island Sound; 71° and 73° in the mouth of Delaware Bay; 75° for both months over the 100 fathom curve off Chesapeake Bay. Hautreux (1911) gives the average surface temperature, for a five- year period, off Fire Island light-ship, as 66.2° for both July and August. How closely the temperatures obtained in July, 1913, agree with Sumner's averages is illustrated by the fact that we had pre- cisely the same reading off Fire Island light-ship; off Marthas Vine- yard (Station 10063, 67°) ; off the edge of the continental slope abreast of Cape May (73°) ; and the difference off Chesapeake Bay was only l°-2°. And they lie within the range for 1896, as given by Dickson, but are colder than his records for 1897. On the other hand the water was warmer off New York in 1913 than the five-year average given by Hautreux, 69° instead of 66.2°. The summer temperatures outlined above are enough to show that 1913 may be considered a perfectly normal year; 1881, 1884, and 1889 were cooler, and 1897 warmer. And in view of the fact that in summer the surface temperature over the continental shelf depends largely on the wind, it is doubtful whether the very slight differences between these years have any general significance. The general range of surface temperature over the continental shelf BIGELOW: COAST WATER EXPLORATION OF 1913. 237 between Cape Cod and Chesapeake Bay, so far as known, may be sum- marized as follows : — in February, the coldest season of the year, the temperature of the water is very low indeed close to the coast and in the bays and sounds, 31°-36° near Woods Hole, rising, toward the southwest, to about 35° near New York (Rathbun, 1887); about 36° off Cape May, 37°-38° at Winter Quarter Shoal; and even south of New York, freezing temperatures may occur near shore during very cold weather. But such low temperatures are limited to a very narrow belt,