HARVARD UNIVERSITY Library of the Museum of Comparative Zoology Bulletin of the Museum of Comparative Zoology AT HARVARD COLLEGE. Vol. LXV. No. 1. A NEW FOSSIL CETACEAN. By G. M. Allen. With One F^ate. CAMBRIDGE, MASS., U. S. A. PRINTED FOR THE MUSEUM. August, 192 L No. 1. — A New Fossil Cetacean. By Glover M. Allen. In the course of re\'ising the collection of fossil mammals in the INIuseum, an unlabeled cranium was found, which was so largely em- bedded in a hard fine-grained marl, that its true nature was not at first appreciated. The specimen, after this matrix had been carefully chiseled away, proves to be of unusual interest. It lacks the vertex of the brain-case, the jugals, and most of the rostrum including the tooth-bearing parts of the maxillae and preraaxillae. What remains, however, is fairly well preserved and clearly pertains to a toothed cetacean of a very primitive type, related apparently to the Eocene Agorophius, but differing in certain important details from the only known cranium hitherto referred to that genus. It is therefore doubly unfortmiate that so important a specimen should be quite without record of locality, horizon, discoverer, or donor. It lay by itself in a tray without label or catalogue nmnber, ha^'ing probably been put aside just as received many years ago. The likelihood is that it was sent to Louis Agassiz in the early days of the Musernn, possibly from some locality in the southeastern United States, at the time when he was planning a memoir on "Phocodon" (see Wyman, Amer. journ. sci., 1850, ser. 2, 10, p. 230, footnote). One or two barnacle bases on the upper side indicate that it lay for a time, partly exposed, in the sea. In the hope that there might be characteristic Foraminifera in the marly matrix, a sample from within the brain-ca\dty was submitted to Dr. Joseph A. Cushman, who very kindly examined it and reports that " there are a few Foraminifera contained in it, most of which are not well preserved. A few, however, seem to show that the material is probably Upper Eocene (Jackson) in age, and its general appearance would seem to indicate that it came from the Gulf Coastal Plain of the United States, probably from Alabama." The cranium belonged to a dolphin-like animal, probably some five or six feet long. Obvious peculiarities are its relatively narrow and flattened brain-case, wide mastoid diameter, elongate flattened nasals, parietals forming part of the vertex, the relatively small and promi- nent occipital condyles, and the long and forward-sloping instead of vertical nasal passage with the remnant of a dorsal chamber above the main part of the nasal ca\'ity. These characters, notwithstanding the lack of corroboration from the teeth, are sufficient to indicate its 4 bulletin: museum of comparatr^e zoology. relationship to the Mesoceti as defined by Dames (1894). ^Yhile it possesses several primitive features in common with Prosqualodon, its relationship is perhaps nearer to Agorophius, w-ith both of which it may be associated in Abel's family, Agorophiidae, whose three known members, while perhaps in no case directly ancestral to the more developed Squalodontidae, yet indicate pre^'ious stages in evo- lution. Though quite as primitive in many respects as Agorophius, the new fossil shows so many points of difference that it seems w^orthy of rank as a separate genus. Archaeodelphis, gen. nov. Diagnosis. — A long-beaked dolphin-like cetacean; teeth unknown, but apparently long-rooted, probably resembling those of Agorophius and Prosqualodon; nasals long, narrow, and flattened dorsally; max- illae covering the anterior three fourths of the orbital portion of the frontals; orbit large, with thickened rim and prominent postorbital process; parietals meeting across the vertex of the skull behind the orbits; zygomatic process of squamosal relatively small, \\'ith small and nearly horizontal glenoid fossa; mastoid region thickened and produced obliquely downward and backward to or beyond the pos- terior edge of the condyles which are small and protuberant. Palatals large, expanded anteriorly, separated medially for more than half their length at the back end and by a deep notch at the front end of their combined margin; pterygoids widely sundered, their free margins partly overarching the narial passage. A well-marked nasal chamber is present above the anterior end of the passage, and the vomer forms a cylinder that completeh' encloses the basal end of the mesethmoid cartilage. The genus is based on the specimen here described. Archaeodelphis patrius, sp. nov. Type-specimen. — A cranium, M. C. Z. 15,749 (Cat. Fossil Mamm.) lacking the bones of the vertex, the jugals, the teeth, and all but the basal portion of the rostrum. Locality and horizon. — Probably from Jackson formation of the Upper Eocene of the southeastern United States, possibly Alabama, as suggested by the Foraminifera from the matrLx. ALLEX: A NEW FOSSIL CETACE.VN. O Description . — A striking characteristic of the dorsal aspect is the narrow rectangular outline of the nasals whose inner anterior corners seem to have been slightly produced to form a blunt median point. They completely roof over the front end of the nasal passage so that the anterior nares open forward, a primitive character common also to the Archaeoceti. At either side of the nasals appears the base of an intermaxillary, its width about equal to that of a single nasal, its termination at about five eighths the length of the nasal, where it abuts against an anterior prolongation of the frontal. Laterally the proximal end of the maxilla extends back to the level of the base of the nasals, and overspreads about three fourths of the orbital process of the frontal dorsally, reaching the edge of the orbit about half way down Fig. 1. — - Side view of the cranium, from a photograph. /, frontal; nix. maxillary, part of base (the dotted line shows the limit of its backward extension); n, nasal; p, parietal; pi, palatal, ascending portion; px, premaxillary, basal end; 1, spheno- palatine foramen; 2, optic foramen; S, orbital fissure; 4. foramen rotundnm; 5, foramen ovale. on its anterior rim; below this point it forms the front portion of the orbit. Posterolaterally the frontal is produced to form a tapering supraorbital process, whose decurved point is separated from the zygo- matic process of the squamosal by about one third the length of the temporal fossa. Its median portion at the point of least interorbital width shows a depression on each side of the cranial axis narrowing to a point forward, which probably received corresponding anterior pro- cesses of the parietals. In Agorophius there is also a median prolonga- tion of the parietals fitting into a corresponding depression of the frontals but the projection is simple, not bifurcate. 6 bulletin: MUSELTkl OF COMPARATIVE ZOOLOGY. Of the parietals themselves very Httle remains in the specimen save a portion of the lateral wing of each, (Fig. 1, p), whose lower boundary is faintly traceable on the inner wall of the temporal fossa, whence it extends forward as a narrowing border on the posterior rim of the supraorbital process. In Agorophius the highest point of the dorsal profile is formed by the base of the maxillaries, back of which the summit of the skull extends on a nearly horizontal though very slightly depressed plane, to the vertex of the supraoccipital. In Archaeodelpliis, on the contrary, there was ob\'iously a gradual upward slope of the profile (Fig. 1) which, if the parietals were in place, must have been continued a slight distance to the junction with the supraoccipital, where, as in recent dolphins, the highest point of the profile must have been. This upward slope of the forehead is further indicated by the upward bevel along the edge of the marl matrix filling the brain-cavity, close to the broken edge of the frontoparietal region. The brain-case itself, though relatively narrow as compared with that of modern dolphins, is nevertheless nearly one and a half times as wide as long. The zygomatic process of the squamosal is relatively weak and ends in a blunt conical point 30 mm. behind the supraorbital process, which slightly exceeds it in size. This is in strong contrast to Agorophius, Prosqualodon, and modern toothed cetaceans, in which it is large and thickened, and produced forward so as to be nearly in contact with the supraorbital process (in the figure of Agorophius, it is seen to be broken near the tip in the only known specimen). Correlated with this difference, is the form of the glenoid ca\'ity for the articulation of the jaw. In Archaeodelphis the ca^ity is nearly flat, and faces almost ventrally, though the posterior border, e\'idently forming a distinct postglenoid process, appears to be slightly broken away. Medially the articulating surface extends for a distance nearly equal to its length. In Agorophius, Prosc^ualodon, and Patriocetiis, as in the modern dol- phins, the articulating surface is relatively larger and includes the con- cave ventral (or anterior) face of the zygomatic process. This differ- ence e\'ideritly implies in Archaeodelphis a more precise limitation of the movements of the jaw, to insure a certain amount of shearing action between the opposing sets of teeth, in addition to their seizing function (the main use of teeth in modern cetaceans). Possibly such a cutting action enabled Archaeodelphis to feed upon small armored fishes, such as the young of ganoids. It may be regarded as a primitive feature, inherited from the supposed creodont or carnivorous ancestors. ALLEN : A NEW FOSSIL CETACEAN. 4- Most remarkable is the development of the exoccipitals and their extension backward, outward, and downward, thereby greatly increas- ing the massive aspect of the mastoid region. A somewhat similar appearance is shown by Agorophius and Prosqualodon but in these genera the exoccipitals do not extend so far backward, hardly surpass- ing the base of the condyles, whereas in Archaeodelphis they equal or exceed the protuberant condyles and are produced strongly downward below them. The occipital condyles are very different from those of modern cetaceans. In the Delphinidae their articulating surface is relatively large and almost continuous witii the surrounding bones of the occiput so that the head rests firmly upon the atlas with its correspondingly enlarged and flattened anterior facets. In Archaeodelphis on the con- trary, as well as in Agorophius and Prosqualodon, they are relatively smaller but very much more protuberant and are set off by a distinct neck or constriction. Their greatest axis is not quite vertical though much more nearly so than in most modern cetaceans, as for example, DelphiniLS. An approach to this condition, however, is found in Platanista among the more primitive lixing forms. This much more primitive condition was doubtless correlated with free instead of fused cervical vertebrae, a fact which, taken in connection with the enlarged mastoid region for muscle attachments, indicates a very much greater mobility of the head both up and down, and sidewise, than in modern cetaceans. Probably with the more forward-opening nostrils, the rostrum rather than the vertex of the head was first thrust above water in breathing, or the front of the head merely elevated from the hori- zontal position when near the surface, as a seal might do. Very fortunately the base of the rostriun and most of the lower portion of the cranium were embedded in the matrix, so that it has been possible by clearing this carefully away, to disclose the structure of these important parts. Contrary to the condition shown by the type-specimen of -Agorophius in which the nasals, intermaxillaries, and vomer seem to have been loosely attached, and have become lost, these bones in Archaeodelphis are strongly soldered together. A \ev\ remarkable and interesting development of the vomer and adjacent bones is seen in a front view of the rostrum (Fig. 2) which in the speci- men is broken short off so as to give nearly a vertical section. The dorsal three fourths of the premaxillaries are considerably thickened with outward-fiaring inner faces bounding the sides of the nasal open- ing. Their ventral fourth encloses the vomer whose lateral wings are here expanded to form a cylindrical tube, containing the mesethmoid 8 bulletin: museum of comparative zoology. cartilage. This tube was obviously continued forward with its sup- porting rod of cartilage to give strength to the rostrum, as in the Denticeti. At its base, the tube separates the two intermaxillaries medially for a space of 9 to 14 mm. and is continued dorsally as a thin knife-like partition quite to the under side of the nasals, so as to divide the nasal chamber longitudinally. There appears to be also a vertical wing on each side lining a portion of the outer wall of the nasal open- ing. Ventrally, the vomer is continued as a median keel from the rostral cylinder and appears on the palatal aspect as a narrow line separating the maxillaries. Viewed from the posterior narial opening, Fig. 2. — The cranium in front view, from a photograph, /.frontal; m, maxillary; n, nasal; p, premaxillary ; v, vomer, forming a rostral tube to enclose, i, the meseth- moid cartilage. the backward extension of this tube is seen to become laterally com- pressed, and continuing its course in the plane of the palate, abuts against the wall of the nasal cavity some 30 mm. from the opening of the posterior nares. With the apparent exception of Ceterhinops, no similar rostral tube is known in other cetaceans, for, as in Pro- squalodon (Abel, 1912) it is usually open dorsally at the base and the mesethmoid cartilage, more or less ossified, appears at the base of the rostrum between the intermaxillaries. The posterior part of the narial passage is flattened dorsoventrally, ALLEN: A NEW FOSSIL CETACEAN. 9 mth divergent sides, and is largely enclosed by the arching palatals and the incurved pterygoids, except medially where these bones are separate below. Behind the pterygoids the narial passage viewed from below, is continued as a broad shallow trough with raised and slightly divergent sides, nearly to the foramen magnum, much as in modern dolphins, except that this portion of the narial passage lies nearly in the plane of the palate instead of being bent at an angle with it. This angle is ob\'ious in Agorophius (True, 1907, plate) as well. The palatal region, so important for its diagnostic characters in the Cetacea, is beautifully preserved except for the tooth-bearing parts of the maxillaries. In most extinct cetaceans, however, this aspect of the skull is seldom preserved or figured so that full comparisons are not as yet possible. In the specimen, only the basal portions of the maxillaries between the tooth-rows remain. Here a slight longitudinal groove-like depression is indicated on each side of the mecHan line, corresponding perhaps, with the shallow palatal grooves seen in Delphinapterus. The palatal bones are perfect and lie in a plane very slightly depressed from that of the maxillaries. As usual in Cetacea, as well as in seals, the tooth-rows lie anterior to the front margin of the palatals. Each palatal is expanded at its forward end, where its outline is strongly convex, so that there is a distinct emargi- nation at the median portion of their combined front edges. Together they nearly fill the space between tooth-rows, and are in contact medi- ally for a trifle less than one third their length before diverging evenly at their posterior ends. At the ventral edge of the orbit each sends up a dorsal branch at right angles to the palatal portion. Just above this edge and close to the anterior margin of the ascending wing is a small but distinct sphenopalatine foramen (Fig. 1, 1). The pterygoids are relatively small, their ventral portion incurved so as partly to embrace the opening of the posterior nares. They are widely separate and their posterior margins divergent. Laterally, on either side of the trough that continues the narial passage, is a deep groove with sharply defined boundaries, extending forward as far as the pterygoid bone. About half way on the length of this groove opens the large foramen ovale, (Fig. 1, 5) its course con- tinued laterally as a shallow furrow. The orbit shows three large foramina for nerves. Slightly above and in advance of its center is the optic foramen of relatively small size (Fig. 1, 2). Below and behind this is the very large orbital fissure (foramen lacerum anterius) deeply excavated in the wall of the orbit, while close against it postero- externally, and separated only by a thin bony partition is the fora- 10 bulletin: museum of comparative zoology. men rotundum (for the second division of the fifth nerve) lying in the same deep groove with the orbital fissure (Fig. 1, 3 and 4). What appears to be the opening of the lachrymal canal lies just below and ahead of the optic foramen, where the outline of a small lachrymal bone can be faintly traced, wedged in between the ascending process of the palatine and the base of the orbital portions of frontal and maxillary. The antorbital foramen perforates the latter just exterior to the lachrymal, and appears in the section of the broken rostrum as a large triangular orifice with its point directed downward. The tympanic bullae are lost, and were evidently but loosely attached as is usual in Cetacea. The petrous and mastoid portions of the ear-bones, however, are still present, and as in some of the more primitive existing cetaceans, (Balaena, Plata- nista) are firmly wedged between exoccipital and squamosal. The petrosum is small (17 X 11.5 mm.), roughly egg-shaped, with its long axis directed anteroposteriorly, and lies close against a bony eminence bounding the inner side of the glenoid fossa. The mastoid portion (28 mm, long) extends obliquely outward and backward to the periphery, expanding to a width of 20 mm. where it reaches the outer surface of the cranium. A notch separates it from the post- glenoid process. The nasal cavities are fortunately preserved intact and were, with some difficulty, quite cleared of matrix on one side of the median sep- tum formed by the vomer. The greater part of their vertical diameter is taken by the narial passage itself which extends from the laterally compressed anterior opening, obliquely backward and downward, expanding laterally as it approaches the posterior nares. Directly back of the anterior narial opening and wholly above the air-passage itself, is a pocket extending backward and nearly cut off below by a blunt projection of the outer wall of the cavity, so that a distinct dorsal di^^sion of the nasal chamber (Fig. 3) is formed, a primitive feature of which no vestige remains in modern cetaceans. Stromer (1903, pi. 11, fig. 1-3) has shown sections of the nasal cavity of Zeu- glodon (Basilosaurus) zitteli in which there is a much better developed olfactory chamber, similarly situated, and wholly cut off ventrally from the main air-passage by a lamina terminalis extending inward Fig. 3. - — Diagrammatic cross-section of nasal passage at base of ros- trum, to show the vesti- gial dorsal nasal cham- bers. ALLEN: A NEW FOSSIL CETACEAN. 11 from the outer wall of the cavity. He found also indications of naso- and maxillo-turbinals. It is therefore probable that the blunt pro- jection from the outer wall of the nasal cavity, above referred to, is the remnant of a lamina fenninalis, but there is no indication of tur- binal bones, which probably had atrophied. Measurements. — The following dimensions indicate the size of the cranium : mm. Tip of nasals to end of occipital condyles 180 Front edge of palatal bone to same point 158 Anteroposterior length of temporal fossa 92 Length of right orbit 54 Length of nasals 41 Combined width of nasals 38 Width across front of orbits 145 Mastoid width 180 Least width between temporal fossae 64 Combined width of palatal bones 69 Width across occipital condyles 57 Approximate width across supraorbital processes (twice one half) 190 Height of muzzle at tip of nasals 70 Summary of Relationships. Of primitive cetaceans whose skull characters are sufficiently known to admit of comparison with Archaeodelphis, three genera stand out as bearing a considerable degree of similarity to it, namely, Agoro- phius, Prosciualodon, and Patriocetus. The first of these, with the single species A. pygmaeus, is still known from the type-specimen only — now lost — the history and peculiarities of which have been fully set forth by True (1907). Although the intermaxillaries and nasals as well as most of the inferior side of the cranium of this speci- men were not preserved, still it bears obviously a general superficial resemblance to Archaeodelphis in the somewhat flattened profile, the great anteroposterior extent and the breadth of the temporal fossae, and the resulting narrowness of the region separating the two fossae anteriorly. This narrow isthmus in both genera, is formed dorsally by the parietals which instead of being excluded from the peak of the cranium as in modern cetaceans, meet behind the frontals at the dorsal line. Further points of resemblance are found in the shape of the brain-case and in the great lateral extent of the orbital portion of the frontal with its well-developed and tapering postorbital process. 12 bulletin: museum of comparative zoology. Both species, further, have small and prominent occipital condyles, indicating a considerable mobility of the head. On the other hand, Archaeodelphis differs from Agorophius in many important characters, both primitive and progressive. Thus its basicranial axis is not bent at an angle with the plane of the palate, whereas in Agorophius the fragments of basioccipital and basisphenoid remaining, clearly form a distinct angle with the palate, foreshadowing the considerable angle seen in many modern dolphins; again, the zygomatic process of the squamosal is but weakly developed in Archaeodelphis whereas in Agorophius it is large and well arched for the extensive jaw-articulation, in addition to being much more produced forward. On the other hand, Archaeodelphis is the more progressive in its higher vertex and shows a special development of the mastoid region downward and backward. A comparison of nasals, intermaxillaries, and vomer is not possible, but since these parts are lost in the type-specimen of Agorophius, it may be that they were less solidly fused than in Archaeo- delphis. In the latter, the extraordinary formation of the vomer, completely enclosing the mesethmoid cartilage in a tube and divid- ing the nasal cavity by a thin bony septum is possibly a specialization; while the retention of elongate, narrow nasals well solidified with the surrounding bones and a distinct olfactory chamber dorsal to the main air-passage are primitive characters. From his study of the three known specimens of Prosqualodon, from the Miocene of Patagonia, Abel (1912) has shown, that although possessing many primitive characters, such as the low vertex, narrow brain-case, broad zygomatic processes, parietals meeting at the vertex behind the frontals, and large temporal fossae, it shows nevertheless a great advance over Agorophius in many respects, and though hardly ancestral to Squalodon, yet foreshadows many of its delphinoid char- acters, such as the reduction of the nasals, the greater anteroposterior compression of the cranium, more nearly vertical nasal passages, and relatively smaller temporal fossae. Its teeth x\bel interprets as being more speciahzed than in the squalodonts, and as a further progressive character, the intermaxillaries are toothless. It has a well-marked maxillary notch as in squalodonts and modern dolphins. In comparison with Patriocetus, a new generic term proposed by Abel (1912, p. 69) for Squalodon ehrlichii, Archaeodelphis is at once distinguished by the absence of the pronounced overhanging ledge that partly roofs over the front end of the temporal fossa, somewhat as in the zeuglodonts (Basilosaurus). The zygomatic process of the squamosal is large as in Agorophius and Prosqualodon, and as in the ALLEN: A NEW FOSSIL CETACEAN. 13 former the dorsal profile of the brain-case is nearly flat. The basi- cranial axis seems to be bent slightly to form an angle with the plane of the palate. x\s True (1907) had previously indicated, this ceta- cean seems very different from typical Squalodon, though its characters are still imperfectly known. The recent discovery of a well-preserved example in the upper Oligocene at Linz (Konig, 1911) should help to elucidate its relationships when the promised studies of Dr. Abel on this important specimen are published. There seems to be no close relationship between Archaeodelphis and the zeuglodonts, which, as lately shown by the studies of Dames (1894), Stromer (1903), Fraas (1904), and Andrews (1906), appear to be only remotely connected with the more typical cetaceans (Mesoceti and Denticeti) if not a wholly independent offshoot from a primitive creodont stock. They reached their maximum development in both size and skeletal modification during Eocene times, and then became extinct. Their ancestry, however, seems to be clearly indicated through the discovery by Fraas (1904) of the skull of a small species {Protocetus atavus) from the lower Middle Eocene of Mokattam, near Cairo, Egypt. This was a primitive surviving type, contemporaneous with more evolved types that inhabited the same Eocene seas. Its dentition, however, instead of exliibiting the usual compressed pre- molars and molars with serrate edges, is like that of a typical creodont. So far as can be judged from the specimen here described, Archaeo- delphis stands as a very primitive cetacean, probably nearest related to x\gorophius of known forms, and to be associated tentatively with it in a separate family, Agorophiidae. It represents a dolphin-like animal belonging in a general way to a type ancestral to the Squalo- dontidae and through them to the more modern delphinoids. A word may be added as to Leidy's genus Ceterhinops. This was founded on a fragment of a cranium which included portions of max- illae, premaxillae, vomer, and frontal. The vomer formed at its base, a cylindrical tube, much as in Archaeodelphis, and this was continued dorsally as a thick bony septum quite separating the nasal passages. The figure given by Leidy (1877, pi. 34, fig. 7) indicates, however, a skull of different configuration, perhaps lacking such nasal bones as Archaeodelphis possessed, and having the basal ends of the pre- maxillae tapering to a point between the frontal and the maxillae. Its fragmentary nature renders a further comparison difficult, but indicates a possible relationship. Leidy's specimen came from the Ashley. River phosphate beds of South Carolina. 14 bulletin: museum of comparative zoology. REFERENCES. Abel, Othenio. 1912. Cetaceenstudien. III. Mitteilung: Rekonstruktion des schadels von Prosqualodon australe Lyd. aus dem Miozan Patagoniens. Sitzb. K. akad. wiss. Wien, math.-nat. cl., 121, pt. 1, p. 57-7.5, pi. 1-3. Andrews, C. W. 1906. A descriptive catalogue of the Tertiary Vertebrata of the Fay ''<> !'■' w ^52 83." / -84 85./ 86 ' §7 8^90 9,92 70° Fig. 1.— Grampus Stations 10340-10398, July-August, 1916. U. S. Bureau of Fi.sheries, to whose industry and skill the value of the results is due. The main objective of the Cruise was the solution of certain fisheries problems. But it soon appeared that so far as temperatures were con- cerned 1916 was an aberrant season, hence full hydrographic data and BIGELOW: COASTAL WATER EXPLORATION. 89 plankton hauls were obtained at stations so located as to afford a section from Gloucester across the western side of the Gulf of Maine and Georges Bank, and a general survey of the coastal zone between New York and Chesapeake Bay. The equipment of the Grampus has been described (Bigelow, 1917a, p. 165). Fig. 2.— Grampus Stations 10399-10417, October-November, 1916. OCEANOGRAPHY. The Gulf of Maine. Temperature. Four years' work in the Gulf of ]\Iaine, together with earlier records (Bigelow, 1914a, 1917a), have shown that though the seasonal variation in its temperature, salinity, and plankton is ex- 90 bulletin: museum of comparative zoology. treme, annual fluctuations are usually slight. However, cold sum- mers occur occasionally, for instance, 1882 and 1916. Thus the cen- tral part of Massachusetts Bay was l°-4° colder at all depths on July 22 (Station 10341) and off Gloucester (Station 10340) than previously noted in summer in that general region (Fig. 3), though the localities of 3 4 )° 6 1 Temperature, Centigrade 5 9 10° 11 12 13 M 15° 1 6 1 7 18 V _-H — ' ■ / ^ ^ "^ ^ // ^ B ^ ^ D / '/ / / / ■^ / / f / 1 1 1 7 / i \ A • M«ter 0 10 20 30 40 50 60 70 SO 90 100 no 1^0 130 140 150 160 170 180 ISO 200 Fig. 3. — Temperatures off Gloucester at the mouth of Massachusetts Bay, various years and seasons. A. July 10, 1912, Station 10002. B. July, 19, 1916, Station 10341. C. August 31, 1915, Station 10306. D. Feb. 13, 1913, Station 10053. F. May 4, 1915, Station 10266. record for the several years are so close together that no regional temperature difference was to be expected on geographical grounds. In fact, below 50 meters, the July temperatures for 1916 are hardly higher than the May readings for the previous year; at 80 meters only BIGELOW: COASTAL WATER EXPLORATION. 91 about .5° higher than the winter minimum for 1913 (Bigelow, 1914b). The water of? Cape Cod (Stations 10344, 10345, Fig. 5, 6) was hke- wise decidedly colder in 1916 than in the summers of 1913-1915, (the 20-40 meter temperature about 2°-3° lower than in 1913; 6°-9° lower than in 1914). And in the southwest corner of the Gulf (Station 10346) the surface was actually 10° colder on July 22, 1916 than on July 19, 1914, though that this very great discrepancy was chiefly due to active vertical circulation is clear, from the very small vertical range of temperature at the station in ciuestion. And the vertical warming below 100 meters so characteristic of this side of the Gulf in 1914 and Fig. 4. — Profile of temperature and density crossing Massachusetts Bay just west of Stellwagen Ledge, July 19, 1916. Temperature curves solid. Density curves dotted. The contour of Stellwagen Ledge is shown by the heavy broken curve. 1915 (Bigelow, 1917a), was hardly appreciable in 1916. During the interval July 22- August 29, the mid-layers off northern Cape Cod warmed by about l°-2° (Stations 10344, 10398, Fig. 5) ; even then, how- ever, the temperature did not equal that of 1912 on the same date (Station 10043, August 29) or of 1913 three weeks earlier (Station 10086, August 5). The Grampus did not visit the eastern side of the Gulf in July, 1916, but observations taken in the Bay of Fundy under the auspices of the Biological Board of Canada, summarized below in a letter from Dr. A. G. Huntsman, show that these waters were also unusually cold during that summer: 92 bulletin: museum of comparative zoology. "The temperature of the water in the Fundy region" he writes, "was un- usually low during the summer of 1916. The data given me by Craigie [(1916a, 1916b), Craigie & Chase (1918)], and by Vachon [(1918)] show that in the St. Croix river near St. Andrews and in Passamaquoddy Bay the temperature of the greater part of the water during the first half of August was approxknately one degree (C.) lower in 1916 than in 1914. In the Bay of Fundy off Campobello Island, the water was slightly colder on July 25, 1916 than it had been on July 14, 1915, and nearly two degrees (C.) colder on August 16, 1916, than it had been on August 27, 1914. Also in the Bay of Fundy east of Grand Menan the temperature of the body of the water was nearly one degree (C.) lower on July 24, 1916, than on July 15, 1915, and more than two degrees (C.) lower on August 16, 1916, than on August 27, 1914. This shows that in the Bay of Fundy the water was colder m the summer of 1915 than in that of 1914, and still colder in that of 1916." Meter 0 10 20 30 40 M 60 70 80 90 100 110 12Q no 140 I 4 I ° ( Temperature, Centigrade ) 7 8 9 10° 11 12 13 14 15* U "~~- '^^ . — ^ ^= ==*■ ?' r" ^ ^^^ ^ — / y ^ c^ \ \ / / ^ ^,\ \ / - / / 1 I C 1 \ ^ / \ \ \ / \ B\ \ \ \ \ \ \ A i \ \a 30.4 .6 .8 3) .2 ,4 .6 .8 32 .2 Salinity; %, .6 .8 33 .2 Fig. 5. — Temperatures (solid curves) and salinities (broken curves) off northern Cape Cod.July 19, 1914, Station 10213 (A). July 22, 1916, Station 10344 (B). August 29, 1916, Station 10398 (C). Even at the end of October and the beginning of November both the sink off Cape Ann, and the deep trough north of it were appre- ciably cooler than earlier records (Bigelow, 1914b, 1917a) would have led us to expect (Fig. 9, 10). The general seasonal temperature BIGELOW: COASTAL WATER EXPLORATION. 93 cycle of 1916, however, was as previously observed. Thus, the sur- face cooled by about the same amount (6°±) in 1916 as in 1912, and the deeper layers warmed (3°± at 50 meters, 1.5°± at 80-100 meters), between July and November at the mouth of Massachusetts Bay, Meter 0 10 5° 6 Temperature, Centigrade 7 8 9 10' n 12 13 14 15° 16 17 18 20 30 40 60 60 70 80 90 100 110 120 130 140 150 160 170 180 190 200 — V > I 1 1 A ^ ■ - 7 X ^ 1 1 \ [y / y ^ t \ \ ^ ^ / / y ^ h- V^ «i \ ' / ^ B / ' . 1 '/ / /f / b 1 ^N\ \ / / / \ V / \ \ 'A ' \ \ ^ V I 1 \ \ \ ; 1 B : 1 1 c\ \ \ i \ \ \ \ 1\ \ i \ 'l 1 1 \ \ \ 1 1 \ V \ \ d\ \\ D B I 30.4 ,6 .8 31 .2 .4 .6 .8 32 .2 .4 .6 Salinity; %j .8 33 Fig. 6. — Temperatures (solid curves) and salinities (broken curves) ofif CapeCod, July 8, 1913, Station 10058 (A). July 19, 1914, Station 10214 (B). July 22, 1916, Station 10345 (C). November 8, 1916, Station 10404 (D). with much the same relationship between the temperature curves for the two months off Cape Cod (Fig. 6) except on the surface. ^ The autumn stations illustrate especially, and this is their most interesting feature, the progress of autumnal cooling, and its local variations relative to distance from land. Off the mouth of Penobscot 94 bulletin: museum of comparative zoology. Bay (Station 10402, Fig. 8) this had been so rapid that even as early as November 2 the temperature was practically uniform vertically down to 50 meters, indeed fractionally cooler at the surface than immediately below: near the Isle of Shoals and in the western basin however, (Stations 10400, 10401, Fig. 10) a temperature gradient of about 2° still persisted. The November records for the western basin are further valuable, there being no previous data for the off-shore parts of the Gulf between September 1 and May, as showing that the surface cooling which takes place during the autumn (Fig. 10), is ac- companied there, just as near land, by a rise of temperature in the mid-depths consequent on the increasing freedom of vertical circula- FiG. 7. — Salinity profile crossing Massachusetts Bay just west of Stellwagen Ledge, July 19, 191b. The contour of the Ledge is shown by the heavy broken curve. tion allowed by decreasing vertical stability. Unfortunately none of the November stations went deep enough to reach the bottom layers of the Gulf, so important in their influence on its general hydrography. Salim'fy. The general distribution of salinity in Massachusetts Bay, July, 1916, closely paralleled that of temperature (Fig. 4, 7), its water being freshest when warmest, saltest when coldest, with the curve for 32%o agreeing almost exactly with 5°; 31.5%o with 9°. And the salinity, like the temperature of the western side of the Gulf was then decidedly lower than in any recent summer of record; •5%o-l%o lower for example at all depths oft' Gloucester than in July 1912, with even greater difference between 1916 and 1913, although Temperature, Centigrade 7 8 9 10° 11 12 13 14 15° M-ter 0 \ ■1 1 \ 10 c\ B / 'h ' 1 \ 20 / ^A 1 \ 1 \ \ \ 30 40 / ^ \ \ \B / / \ \ \ V- SO / / A^ [ \ \ s 60 / / \ s \ \ - 70 / \ 80 / V C 90 / \ 100 110 / \ / 120 / 130 . 140 J i 3 1 . i 4 3 salin 8 : ity; 2 . 2 . i 5 .8 a Fig. 8.— Temperatures (solid curves) and salinities (broken curves) off Penobscot Bay, September 16, 1915, Station 10318 (A). October 9, 1915, Station 10329 (B). November 2, 1916, Station 10402 (C). Meter 0 10 20 30 40 50 60 70 80 30 100 no 120 130 140 Temperature, Centigrade 6 7 8 9 10' 11 12 13 \ \ V i /■ ^ / \ / 7 > \ / / / \ w A r / «^ 1 / fe l^ V \ / \ \ \ 1 1 1 .4 6 .8 32 Salinity. /^, Fig. 9. — Temperatures (solid curves) and salinities (broken curves) at the mouth of Massachusetts Bay, Station 10399, October 31, 1916 (A). Mean of Stations 10338 and 10339, October 27, 1915 (B). 96 bulletin: museum of comparative zoology. the station localities for these two years are so close together that none was to be expected on geographic grounds. The Stellwagen Sink (Fig. 7) also was fresher than we have ever found the similar sink off Cape Ann, even when the annual minimum of salinity is reached. As a whole the salinity off northern Cape Cod was about the same Temperature, Centigrade 8 9 10° 11 12 13 14 15° 16 17 18 19 Meter 0 10 20 30 40 50 60 70 80 SO 100 110 120 130 140 150 160 170 180 ISO m Salinity; j>il, Fig. 10. — Temperatures (solid curves) and salinities (broken curves) in the Sink west of Jeffrey Ledge, Station 10400 (A), in the western Basin, Station 10401 (B), on November 1, 1916. In the western Basin, Station 10307, August 31, 1915 (C). in the upper layers of water in July, 1916 as in that month of 1914 (Fig. 5). But from the 25 meter level down to 80 meters or so the water was decidedly fresher in the former summer than in any of previous record, nor did the salinity rise appreciably there between / I 1 ^ r^ y / ' tr / i '/ \ /, y \ t, \ } -4 / \ — / / / \ \ \ // /I 3 \ I \ / 'y '\ \ C / f \. \ // 1 \ \ ^^ \ 1 \ 1 1 1 A \ — \ \ ■T \B \ A — \ \ \ \ — V \ \ \ \ — V \ \ \ .8 3 ,2 .4 .6 .8 3 X .2 .4 .6 .% 3: .2 A .6 Meter 0 10 20 30 40 60 60 70 80 90 100 no 120 130 140 150 160 170 180 190 200 B~^ ^ — s \\ ^ c V ^V \ ^ \ \ ^ s K \\ V \ \ \ \ \ 1 \ B C \ V 30.4 £ .8 A .6 .8 32 Salinity; %, .8 33 Fig 11. — ^[Salinities in the deep basin of the Bay of Fundy, June 10, 1915, Grampus Station''l0282 (A). July 24, 1916, Vachon's Station 3 (B). August 2.5, 1916, Vachon's Station 3 (C). [Meter 0 10 20 30 40 50 60 70 80 00 Temperature, Centigrade 10° n 12 13 14 15° 16 17 h \ /I .^ ^ J^ s lA -V- \ / f \ t 1 1 .8 32 .2 .4 .6 .8 Salinity; /d 33 Fig. 12. — Temperatures (solid curves) and salinities (broken curves) on the north- western part of Georges Bank, July 9, 1913, Station 10059 (A). July 20, 1914, Station 10215 (B). July 23, 1916, Station 10347 (C). 98 bulletin: museum of comparative zoology. July 22 and August 29. A salinity comparison between the summers of 1913, 1914, and 1916 is less satisfactory for the deep southwestern basin of the Gulf, the stations for the several seasons not coinciding so closely in location. Nevertheless, that 1916 was the freshest sum- mer of the three can hardly be disputed, the difference between the Meter 0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 ISO too Temperature, Centigrade 8 10° 11 12 13 M 15° 16 17 18 IS 20 \ \ \ \ y y ^ ^ \ ^ ^ ^ \ ^^ ^ "1 -Tv^ Jb ^ ^/i /^ A y / \ \ X B \ \ I / \ 1 \ \ \ 'Vj \ \ \ \ \ \ \ \ \ \ \ \ b\ W \ \ L \ \ \ \ N \a A .8 32 .2 A 33 .2 .4 .6 Salinity; %, .8 34 .2 .4 .6 Fig. 13. — Temperatures (solid curves) and salinities (broken curves) at the outer edge of the Shelf south of Nantucket Shoals, July 24, 1916, Station 10351 (A). July 10, 1913, Station 10061 (B). curves (Fig. 6) being too great to be accounted for by small dif- ferences in geographic location in the open sea. (Station 10314, in 1914 was situated 15 miles. Station 10058 in 1915, 25 miles east of the 1916 Station 10345). BIGELOW: COASTAL WATER EXPLORATION. 99 The summer of 1916 was also fresh in the Bay of Fundy, to judge from Vachon's (1918) July records as compared w^th our Station in the deep water near its mouth in 1915 (Bigelow, Station 10282, June 10, 1915). True, there is a considerable seasonal interval between the latter and Vachon's earliest for (July 24, 1916), while differences in Meter 0 20 40 60 SO 100 120 140 160 180 200 220 240 260 280 300 320 340 360 380 400 ■ 4 5 ° I " Temperalure, Centigrade 8 9 10' 11 12 13 14 15° 16 17 18 19 20* 2) ^ ^ ^ — r— _ ^ ;;;;: ^ 7^ ':^ ^^ / / B A y ^ -^ / 4 ( \ \ ' y — \ -^N \ \ \ ' / X \ \ "x / \ \ / ; / \ / // /, / / / / A ' \ ^/ / / \ / / 1 / / / / / / / // / 1 V 1 / / 1 Fig. 14. — Temperatures over the Continental Slope south of Georges Bank, July 24, 1916, Station 10352 (A). South of Cape Sable, June 24, 1915, Station 10295 (B). South of Georges Bank, July 21, 1914, Station 10218 (C). South of Marthas Vineyard, August 17, 1889, Libbey's (1891) Station 9, Line G, (D). geographic location may well be associated with varying salinities in a region of tidal currents as strong as those of the Bay of Fundy. But it is unlikely that the whole difference between the two sets of observa- tions can be so accounted for, a conclusion consistent with the close 100 bulletin: museum of comparative zoology. agreement between the Grampus salinities for June, 1915, and Vachon's for September 6, 1916 (Fig. 8). It is not wise to lay much stress on our November salinities for 1916, not only for the reason given below (p. 178), but also because of differ- ences in geographic location between them and the July stations. So far as they go, however, they suggest that the water was about as much Salter than usual then than in Jul v. 5' 6 Temperature, Centigrade 9 10' 11 12 13 14 15° It 17 18 19 Meter 0 10 20 30 40 &0 EO 70 80 50 '"" .8 32 V 4 6 8 33 .2 4 6 .8 34 .2 4 .6 .8 I Salinity; %, Fig. 15. — Temperatures (solid curves) and salinities (broken curves) on the Shelf ofif Nantucket Light Ship, July 25, 1916, Station 10354 (A). July 9, 1913, Station 10062 (D). Also on the Shoals, July 25, 1916, Station 10355 (B). July 9, 1913, Station 10060 (C). ^, k \ T 1^ \ / ■^ ^ 7 ■^ \ B \ \ \ cA A. \ f^ ^^ /\ y ^ /^ V \ \ /" \ / \ ^- \ y f \ I % \ \ iV V A ' 1 Georges Bank, and off Marthas Vineyard. Temperature and Salinihj: — Annual comparisons are less satis- factory for Georges Bank with its violent tidal currents, and corre- sponding local differences in vertical circulation, than for the deeper and quieter waters to the north. But so far as the Grampus records can be relied upon as typical of their respective years, the waters over its northwestern part were several degrees colder in July, 1916 than at the corresponding season in either 1913 or 1914 (Fig. 12). And this was equally true for the salinity of 1916 as compared with 1913, with a maximum difference between the two series of observations of about .6%o (Fig. 12), though there was little discrepancy between 1916 and 1914 in this respect. But it was over the southern slope of BIGELOW: COASTAL WATER EXPLORATION. 101 the Bank that the hydrographic differences between July, 1916, and the summers of 1913 and 1914, were most striking, for while violent annual, even day to day, fluctuations are to be expected there because of the constant conflict between the warm Gulf Stream and the cool coastal and northern currents, nothing in our pre\-ious experience 1»et( r 0 20 'q ^ - -^ — . r^ s. ■^ "c" 40 60 *<= 8. ^^ ^ -^ y 8U ^ \^ \ \ \ / too :^: \, / 120 - D\ ^^ \, / / UQ \\ A / too \ /' 180 ) / 200 / m / 140 2E0 f j 280 / 300 J20 340 / / / 3G0 / 380 / 3Z4 .8 33 .6 .8 34 .4 .6 .8 35 .2 .8 3& Salinity; %, Fig. 16. — Salinities over the Continental Slope south of Georges Bank, July 24, 1916, Station 10352 (A). South of Cape Sable, June 24, 1915, Station 10295 (B). South of Georges Bank, July 21, 1914, Station 10218 (C). South of Cape Sable, July 28, 1914, Station 10233 (D). foreshadowed summer temperatures so low as those of that year. Thus on the southern edge of the Bank, where our stations for 1916 (10349) and 1914 (Bigelow, 1917a, Station 10217) coincide closely in location, and where the surface temperatures were practically the same for the two years (17°-18°), the bottom water was 4° colder for 102 bulletin: museum of comparative zoology. 1916 (6.7°, as against 10.6°). Similarly, some forty miles further west (Station 10351, Fig. 13), both the immediate surface, and the depths below 50 meters, were 3°-4° colder than at the same relative location on the slope off Nantucket Light Ship in July, 1913 (Bigelow, 1915b, Station 10061), the minimum temperature (4.8°), being nearly as low as that of the Gulf of Maine at the same level (Fig. 10). And the less-depths were even colder over the outer part of the slope (Station 10352). This last Station is especially interesting, not only because the actual temperatures were much lower here in 1916 than in 1914 (Fig. 14), with vertical cooling much more rapid, but because the curve reveals a pronounced cold layer of 4°-5° at about 50 meters, with the 44 Meter 0 Fig. 17. — Temperature profile from the Basin of the Gulf of Maine to Georges Bank, July 22-23, 1916. underlying stratum, 200 meters or so thick, about 3° warmer. In fact to again find water as cold as was the 40-60 meter level here it is necessary to descend into the abyss, below 500 meters. Xo such cold layer was encountered anywhere along the slope, from off Nantucket to the eastern channel in 1914 (Bigelow, 1917a, Stations 10218, 10220, 10261). But the distribution of temperature was of this type in 1913, judging from the one station in this region for that year (Bigelow, 1915, Station 10061), though the actual readings were higher (mini- mum at 75 meters about 10.8°). And temperature curves very similar to those of 1916, with the minimum layer at about the same level, though some 2°-5° warmer, result from Libbey's data for 1889 (Fig. 14. Libbey, 1891, p. 454). BIGELOW: COASTAL WATER EXPLORATION. 103 There is, then, nothing unprecedented in a vertical distribution of temperature of this type, for this part of the slope: indeed its repeated occurrence suggests that something of the sort is to be expected except when obscured by encroachments of the warm water of the Gulf Stream. What was surprising for 1916 is that the temperature of this coldest layer was so very low: and that water so cold lay so close to the surface of the open sea at this latitude in midsummer. But to find these conditions reproduced it is only necessary to turn to the temperature curves in the mixture of Cabot Current and ocean water, at the same relative position on the slope off southern Nova Scotia, about 200 miles to the northeast. Met< JDDt^ Fig. 18. — Salinity profile from the Basin of the Gulf of Maine to Georges Bank, July 22-23, 1916. The temperatures over the shelf off Nantucket shoals were almost exactly the same for 1916 (Stations 10354, 10355) as for 1913, from about the 30 meter level down to the bottom in 70 meters, though 4°-5° lower -on the surface (Fig. 15). Generally speaking, salinity, like temperature, was decidedly lower in 1916 than in 1913 or 1914, both over Georges Bank (Fig. 12) and over the Shelf off Marthas Vineyard (Fig. 15), down to about 70 meters, with even wider divergence of the same sort between the stations for 1916 and 1914 on the Continental Slope (Fig. 16). ^ But the salinities of all our deep stations (Fig. 16), like their temperatures, are practically the same (34. 8-35. 2%^) below 250 meters. ^Nlore 1 The salinity curve for Station 10349 is probably much like that for Station 10352. there was some confusion of the water samples, it is omitted here. But as 104 bulletin: museum of comparative zoology. instructive than the local differences, always to be expected along this zone of varying currents, is the fact that in its salinity (Station 10352), practically reproduces the mixed water as found off Nova Scotia in June of 1915 (Fig. 16), especially at the level of minimum temperature. The general distribution of temperature and salinity is further illustrated here by the usual profiles normal to the coast. It is not Meter Fig. 19. — Temperature profile crossing the Continental Shelf off Marthas Vineyard, July 24-26, 1916. possible to construct a complete profile across Georges Bank, data being lacking for the critical regions just inside, and just outside the 200 meter contour (no temperatures were taken at Station 10350, or beyond Station 10349). But partial ones for both salinity and temperature are given to show the relationship between the waters of the basin of the Gulf on the north, and those over its boundary rim BIGELOW: COASTAL WATER EXPLORATION. 105 on the south (Fig. 17, 18). The temperature profile shows that the water was sHghtly warmer over the Bank than in the Basin to the north, for while direct comparison between these two localities is obscured by the vertical uniformity of temperature on the Bank caused by active vertical circulation (p. 126), the mean temperature was higher there (about 10-11°) than for the same stratum of water Fig. 20. — Salinity profile crossing the Continental Shelf ofiF Marthas Vineyard, July 24-26, 1916. in the Gulf (about 6°-7°). The water was likewise Salter over the Bank than in the Gulf (p. 178) with the salinity of the whole column of water (0-50 meters) over the center of the former (Station 10348) about .3%o higher (32..54-32.o7%o) than the mean of the same stratum in the Basin (about 32.2%o). The first complete profile afforded by the Cruise of 1916 is from the 106 bulletin: museum of comparative zoology. neighborhood of Marthas Vineyard out to the Continental Slope (Stations 10351-10357). Along this line (Fig. 19, 20), warmest and freshest at the shore end (Station 10357), as is usually the ease in such shallow regions, the cold band (5°) touched the bottom only between the 90-130 meter contours if at all, with warmer bottom- FiG. 21. — Chart of minimum temperature out to the 500 meter contour, July-August, 1916. Dotted, 5° — ; unshaded, 5°-7°; single hatched, 7°-10°; cross hatched, 10° + . water both on the shelf on the one hand, and deeper down the slope on the other. Passing thence off-shore the axis of the coldest layer rose to within 50 meters of the surface over the ISOO meter contour (Station 10352), which was near its seaward limit. Owing to tlie lack of hydrographic data for Stations 10350 and 10353 it is a cjuestion BIGELOW: COASTAL WATER EXPLORATION. 107 just what relationship the cold band bore to the slope of Georges Bank further east, i.e., whether it passed north or south of the rather warmer water at Station 10349. Nor does the salinity help here, the records for the 30 and 80 meter levels at the latter Station being unreliable (p. 103). Mete Fig. 22. — Temperature profile crossing the Continental Siielf off New York, .\ugust 1-2, 1916. There is nothing novel in the discovery of such a shelf-like intrusion of cold coast into warmer off-shore water in this region, any more than in the existence of a cold band (p. 103). On the contrary, a phenomenon of this sort has long been recognized (Libbey, 1891, 1895; Bigelow, 1915), and was well exemplified in 1913 by the profile off Nantucket (Bigelow, 1915, p. 1&4, fig. 10); the only important 108 bulletin: museum of comparative zoology. difference between that year and 1916 being in its exact extent, and precise temperature. The most important feature of the sahnity profile off Marthas Vineyard (Fig. 20) is its faihire to reach water as salt as 35%o, this demonstating that the inner edge of the Gulf Stream lay some distance outside of, instead of close to the Slope, as in July, 1914 (Bigelow, 1917a, p. 189, fig. 26). Meta Fig. 23.- 1916. • Salinity profile crossing the Continental Shelf off New York, August 1-2, Marthas Vineyard to Chesapeake Bay. Temperature and Salinity. Temperature and Salinity in August: — ■ No observations were taken during the summer of 1916 over the outer part of the Shelf between Nantucket and New York; but an unbroken band of surprisingly BIGELOW: COASTAL WATER EXPLORATION. 109 cold (4°-5°) water with higher temperature (7°-9°) on the bottom deeper down the Slope, was again encountered at 50-100 meters depth all along the outer half of the Shelf from abreast of New York nearly to the latitude of Chesapeake Bay (Fig. 21), though whether or not continuous to the east with the equally cold water off Nantucket is uncertain. *» Fig. 24. — Temperature profile, crossing the Continental Shelf off Delaware Bay, August 11-12, 1916. Three profiles across the Shelf, off New York, off Delaware Bay, and off Chesapeake Bay serve to show the distribution of temperature in some detail. On the first of these (Fig. 22) there was very little horizontal change in the upper 40 meters or so, out to the 100 meter contour, though the vertical temperature gradient was very steep; and the cold band (4°-5°), which lay here on the bottom, was only no bulletin: museltm of comparative zoology. about 25 meters thick; but its influence was evident over the Slope as a shelf-hke projection of water of 8°-10° into warmer water at Sta- tion 10368, where the comparatively high temperature of the upper 50 meters suggests the neighborhood of the inner edge of the Gulf Stream. Attention should also be called to the fact (p. 109) that the Meter 82 83 84 Fig. 25. — Salinity profile, crossing the Continental Shelf off Delaware Bay, August 11-12, 1916. bottom water was about .3° warmer (8°+) between the 120 and 200 meter contours than higher up on the Shelf. The corresponding salinity profile (Fig. 23) illustrates the general land-sea increase in salinity, especially over the Continental Slope, with which we are familiar along our coast. And the steep gradient of the salinity curves over the Slope suggest that 35%o water would have been found at the BIGELOW: COASTAL WATER EXPLORATION. Ill surface 20 or 30 miles further out at sea. The saHnity of the cold band was about 32.8%o, that of the bottom water at 100-300 meters The temperature profile off Delaware Bay (Fig. 24) agrees in its general outlines with the foregoing; but here the cold band was not only much wider than off New York, but thicker as well (about 50 meters at its maximum), the 5° water indenting seaward as a very pronounced Shelf into the warmer water over the Slope, with the result that for a distance of about 20 miles it overlay bottom water some 2°-5° warmer. The two temperature profiles, taken together also suggest a minor warm band on the surface following the outer edge of the Shelf (Stations 10366, 10382), a phenomenon often en- countered along this zone, where the interaction of warm and cool currents is so complex. The corresponding salinity profile (Fig. 25) shows the same general seaward rise in salinity over the Slope as ob- served off New York (Fig. 23). But the influence of the land water from Delaware Bay is evident across the whole breadth of the profile in the upper 30-40 meters, and especially at the oft'-shore end, where the surface layer is fresher than 32%o. On the bottom, at the outer edge of the Shelf the transition between the cold band (salinity 32.69^0-32.89^0) ^^^ the warmer salter water which bounds it off- shore is as distinct in salinity as it is in temperature. No water so cold as 5° was encountered on the profile off Chesa- peake Bay (Fig. 26), though the lowest temperature there was only fractionally higher (5.08° at Station 10392), and water as cold as 6° was here limited to a very narrow zone at the outer edge of the Shelf. At the shore end, the profile reveals the influence of the Bay by high surface temperature, as does the salinity profile (Fig. 27) even more clearly. Its outer end, however, suggests that the Gulf Stream lay near by, and that 35%o water washed the bottom between the 175 and 380 meter contours, though the data do not absolutely establish this. It is also worth noting that the salinity of the coldest water (aoout 32.8) is exactly the same here as off New York (p. 112, Fig. 23). These profiles, together with the data for the individual stations (p. 178) show not only that the cold band, the most interesting dis- coverv of the Cruise, was \\adest off Delaware Bav, but that it was there that water of 5° or cooler lay nearest the land, and nearest the surface (Station 10381, about 28 meters); there, too, the lowest temperature was recorded (4.02°). But its presence would not have been suspected from surface temperatures, which were upwards of 20° for this part of the Shelf as a whole (Fig. 28). Consec^uently the 112 bulletin: museoi of comparative zoology. vertical temperature gradient, from the surface downward, was steeper than pre\-iously found in summer. The sahnity of the cold band was almost as uniform as its tempera- ture, the extreme range, for the layer colder than 5°, being onlv about 32.4%o-33.2%o (usually 32.&-32.8%o) as follows: — Range of Range of Station SALINITY FOR SALINITY for 4°-o'^ WATER Station 4°-5° water 10365 32.6-32.81 10382 32.6-32.7 10370 32.8-33.2 10383 32.6-33. 10373 32.5-32.6 10392 32.77 103S1 32.4-32.61 Fig. 26. — Temperature profile crossing the Contineiital Shelf off Chesapeake Bay, August 21-22, 1916. BIGELOW: COASTAL WATER EXPLORATION. 113 In short this body of water was practically uniform in its physical characters, over a distance of 200 miles, irrespective of the depth at which it lay, or of its vertical thickness. The hydrographic and biologic importance of a mass of water so cold, at such a tri^-ial depth, so far south, must not obscure the fact that it did not extend out to the edge of the Shelf on the bottom. Meter Fig. 27. — Salinity profile crossing the shelf off Chesapeake Bay, August 21-22, 1916. though it did in the mid-depths. On the contrary, the bottom water (Fig. 29) all along the 150-200 meter zone was several degrees warmer, (8-9°) corresponding to the "warm band" long known off ^larthas Vineyard and Nantucket, though it was slightly cooler than the Grampus found it in 1913 (Bigelow, 1915, p. 164). This band was small in extent, it is true, in comparison with the cold zone, neverthe- 114 bulletin: museum of comparative zoology. less it occupied an area of not less than 1200 square miles between the latitudes of New York and Chesapeake Bay. And if it extended without interruption (though with slowly decreasing temperature) as far east as the profile off Nantucket, which was probably the case in 1916, as in 1913 (Bigelow, 1915, p. 164, fig. 10), its total area can not Fig. 28. — Surface temperature, August, 1916. have been less than 2500 square miles. Its economic importance is as great as its biologic, for it is the chief, if not the exclusive, habitat of the Tilefish (LophaJotilus chamadconticeps Goode & Bean), a species the food value of which is now generally recognized, and which is the object of an important fishery, thanks to the efforts of the U. S. Bureau of Fisheries. BIGELOAV: COASTAL WATER EXPLORATION, 115 Direct station for station comparison between 1916 and 1913 is possible only in a few cases for this part of the coast water. But wherever it can be made, the contrast in temperature is considerable in the deeper layers, though there is but little difference on the Fig. 29. — Bottom temperature, July-Augrust, 1916. single hatched, 7°-10°; cross hatched 10° + . Dotted, 5° — ; unshaded, o°-7°; surface. Thus a few miles off New York, the 40 meter temperature was 3° lower on August 1, 1916, than on July 12, 1913 (Fig. 30), though the difference in season would suggest just the reverse; not till the end of x\ugust did the temperatures for 1916 equal those taken a month earlier in 1913. Similarly, the minimum temperature, over the outer part of the Shelf on this line, was 2° lower in 1916 (Station 116 bulletin: museum of comparative zoology. 10365, 4.8°) than in 1913 (7°). And passing southward "vve find the temperature difference between the two years increasing, owing to the fact that while the temperature of the cold band was uniform through- out its length in 1916, in 1913 the coldest layer warmed gradually from north to south (Fig. 31, 33). The annual difference is further illus- trated bv the charts for bottom temperature (Fig. 29, Bigelow, 1915, fig. 8). \ _ The salinity of the coastal water south of New York was likewise much lower as a whole in August, 1916 than in July-August, 1913. And not only was this true of the surface, which was fresher than 31%o over the whole inner half of the Shelf in 1916, instead of only close to the mouth of Chesapeake Bay as in 1913, but of the whole Meter 0 10 Temperature, Centigr&de 5° 6 7 8 9 10° 11 12 13 14 15° 16 17 13 19 20° 21 --^ ' ^^ "^ " ~ — -- ^ •*». *^, ■* ^ ^- > ' K ^ ^ ^ ^ V N ) / ^ i A 'b 20 30 40 50 ^".6 .8 31 ,2 .4 .6 .8 • a2 .2 .4 .6 .8 33 .2 .4 .6 .8 Salinity; %, Fig. 30. — Temperatures (solid curves) and salinities (broken curves) off New York, August 1, 1916, Station 10364 (A). July 12, 1913, Station 10066 (B). column, surface to bottom, as well. Thus the bottom water off New York was salter than 33%o up to the 15 meter contour in 1913 (Bige- low, 1915, p. 195, fig. 39) whereas in 1916 salinities as high as 33%o were confined to depths greater than 80 meters (Fig. 23). Off Dela- ware Bay (Fig. 25, 32) where the salinity on the bottom was 33-34%^ in 1913, it was only 32-32. 8%o in 1916, with a similar relationship in the upper layers. And off Chesapeake Bay the salinity contrast between the two years is even more striking, for the water over the whole Shelf above the 100 meter contour was fresher than 33%o in 1916, whereas in 1913 its outer half was salter than 34%o> below the tri"vial depth of 10 meters or so. The mixed water over the Continental Slope, like the coastal water proper, was much fresher above 200 meters in 1916 than in 1913, as illustrated by the pairs of stations at corresponding locations (Fig. 34), BIGELOW: COASTAL WATEE EXPLOEATION. 117 3U .8 I 2 .4 .S .8 33 .2 .4 Salinity; %, .8 34 .2 .4 .6 .8 35 Fig. 31. — Temperatures (solid curves) and salinities (broken curves) on the outer part of the Shelf off Cape May, August 3, 1916, Station 10370 (A). July 19, 1913, Sta- tion 10070 (B). 5° 6 7 Meter 0 M 10 40 20 (0 30 U 40 t«0 50 120 60 140 70 160 80 Temperature, Centigrade 9 10' n 12 13 14 15° 16 17 18 19 2tr 21 — ^ L .^ - «.. ^, ^ ^ ^ X\ N, ^x ^ -^ "■^ s ^^ ^ X "" _^ -- 1 \ ,^ \ \ / B \ \ B / '' \ ^ A A .8 31 .2 .4 .6 .8 32 2 4 .6 .8 33 .2 .4 .6 .8 34 i Salinity; %, Fig. 32. — Temperatures (solid curves) and salinities (broken curves) off Delaware Bay, August 4, 1916, Stations 10373 and 10375 combined (A). July 21, 1913, Station 10072 (B). 118 bulletin: museutvi of compabative zoology. with a maximum difference of no less than 38%o. And it was Hkewise cooler (Fig. 35), especially at 40-100 meters, i.e. at the level of the cold band, the effect of which is particularly evident off Delaware and Chesapeake Bays (Stations 10384, 10393), though with little annual difference on the surface. But all these off-shore stations approach each other closely both in salinity and in temperature at 400-500 meters. And the data for 1916 support our previous experience that abyssal temperatures (about 4°), and salinities (about 34.9%o) prevail at, and below 500 meters all along the slope from Chesapeake Bay to Nova Scotia. Meter 0 10 Temperature, Centigrade 3 4 5° 6 7 8 9 10' 11 12 13 14 15' 18 17 18 19 20° 21 22 23 20 50 60 70 80 90 100 ■"* — -- ~T — — - -. . _, L \ B ~ " **^ • -- g- \n \ __— - " — 1 A^ / ' V \ ■^ ^ / ^ V > B ^ "B / \ / 1 1 / 1 1 A A i 31 A . 5 .! i 3 2 . ? .' » .( > .! i 3 3 u ! .i 1 5 5 3 4 .1 ! .< \ ) \ 3 5 Salinity; /^ Fig. 33. — Temperatures (solid curves) and salinities (broken curves) off Chesapeake Bay, August 22, 1916, mean of Stations 10391 and 10392 (A). July 24, 1913, Station 10077 (B). The combined evidence of temperature and salinity together with the plankton (p. 139) prove that in August the inner edge of the Gulf Stream lay further off-shore in 1916 than in 1913. Nor is there am'- thing in the records for 1916 to suggest the salt tongues (35%o) en- countered between Chesapeake and Delaware Bays in 1913. The very considerable differences in absolute temperature and salinity between 1913 and 1916, just outlined, must not mask the fact that the same basic contrast between coast and off-shore water, i.e. the low temperature and salinity of the former, compared with salt- ness and warmth of the latter, obtained in the one year as well as the other. And while the transition from the one type of water to the BIGELOW: COASTAL WATER EXPLORATION. 119 other was not as sudden over the Slope in 1916 as in 1913, the seaward rise in temperature and in salinity revealed by the outermost stations on all the profiles south of New York for that year, together with the off-shore limitation of the cold band, and the presence of Avarmer water below it (p. 113) suggest that it would not have been necessary Meter 0 711 ^ -'^^ V ^^ \ in 84 \ s ■^ .^ \ \ 60 ^ -^ ■^ \ 80 100 \. ■~-^,- ^.^ "^ N \ X X \ i \ 93^ \ \ ua N N i N ^\ 180 200 220 240 9Kn \ \ / \ / 1 / 1 II 280 1 1 1 1 1 / J 320 ^ 1 '/ 340 1 360 / ; 18(1 400 1 J A 31.6 .8 32 2 .8 33 34 .6 .8 35 .2 .4 t Salinity; ^ Fig. 34. — Salinity of the mixed water on the Continental Slope off New York, August 2, 1916, Station 10368 (68). Off Delaware Bay, August 12, 1916, Station 10384 (84). Ofif Chesapeake Bay, August 22, 1916, Station 10393 (93). Off Dela- ware Bay, July 20, 1913, Station 10071 (broken curve). to run the profiles much further seaward to have reached the inner edge of the Gulf Stream. Furthermore, though the actual values differ, the horizontal dis- tribution of surface salinity over the Shelf was of essentially the same type in 1916 (Fig. 36) as in 1913, with the same general alternations. 120 bulletin: museum of comparative zoology. of fresher and Salter water. From northeast to southwest these are: 1, an unmistakable outpouring of land water from the mouth of Long Island Sound, more evident and traceable further east, in 1916 (Fig. 36) than in 1913. S, An eddy off the mouth of the Hudson River, with fresher water on its inner, salter on its outer (northern) side, the r Q 20 5 " e ' i i Temperature, Centigrade ( 10' n 12 13 14 15° 16 17 18 19 20° 21 ■7=^ ^ — ^ ' 40 60 80 too 120 140 160 180 200 ^ — ^ ^ ^^' ^ / ^ ^^ y y- < A ^ u y 93. { ^ \ / / /• \ >v. y t f N (\ / / f / \ / ( / \ / 1 / //' 220 240 //. / > y/ / 1/ y / r 2S0 inn / /( / / / // J / / / W) u / > 340 A/ / / 360 A / 3S0 400 / 1 1 / f/ 1 I Fig. 35. — • Temperature of the mixed water on the Continental Slope off New York, August 2, 1916; Station 10368 (68). Off Delaware Bay, August 12, 1916, Station 10384 (84). Off Chesapeake Bay, August 22, 1916, Station 10393 (93). Off Delaware Bay, July 20, 1913, Station 10071 (broken curve). axis of the outflow, (or of the freshest water, which may be merely reminiscent of a pre-existing current), directed more to the south in 1913, more to the east in 1916. 5, A similar phenomenon off Dela- ware Bay, much the same for the two years except that the actual salinities were lower in 1916 than in 1913. 4, The outflow from BIGELOW: COASTAL WATER EXPLORATION. 121 Chesapeake Bay, which was traceable further off-shore in 1913 than in 1916, though here the surface sahnities for the two years differed but httle. The Hudson, and the DelaM'are eddies, if they may be so christened, are similarly illustrated by salinities at 20 meters (Fig. 36), the curve for 'S2.o%o, at this level, roughly paralleling 32%^ Fig. 36. — Salinity on the surface (broken curves) and at 20 meters (solid curves), July-August, 1916. on the surface. But these e\"idences of land water were confined to this superficial stratum. With a vertical range of temperature as great as obtained in 1916, the slightest vertical circulation by tides, currents or storms, or even by constant off-shore winds, may wholly mask the normal state of the surface by bringing up much colder water from a few fathoms deeper 122 bulletin: museum of comparative zoology. down. And conversely, the surface warms very rapidly along our coast in summer, under the sun's rays, if the weather be calm, and the wind on-shore. Allowing for such disturbing influences, the distribu- tion of surface temperature was much the same, in its main outlines, in 1916 (Fig. 28) as at the same season in 1913 (Bigelow, 1915, p. 157, fig. 2) though the absolute values were slightly lower, the curves for 20° and 22° for 1916, closely paralleling 21° and 24° in 1913 (Bigelow, 1915, fig. 2), with the same north-south rise in temperature, and high surface readings off the mouth of Chesapeake Bay (Bigelow, 1915, p. 155, fig. 1, 2). This w^as fundamentally true of the deeper water-layers as well, although the subsurface temperatures were much lower in 1916 than in 1913 (p. 115). That is to say, the coldest water occupied the same general position, relative to the warmer layers, in 1916 as in 1913, being localized along the outer part of the Shelf, with higher tempera- tures off-shore, as well as nearer land. But if corresponding profiles for the two years be compared, considerable differences appear in detail. Perhaps most important of these is that whereas in 1916 the greatest mass of cold bottom water, with the lowest actual tempera- ture lay off Delaware Bay (p. Ill, Fig. 29), in 1913 the cold band was, to all intents, obliterated a few miles south of the New York profile, its main body lying to the east of the latter (Bigelow, 1915, p. 162, fig. 8). And, correspondingly, the shelf-like indentation of cold water seaward into warm extended as far as Chesapeake Bay in 1916, whereas in 1913 it was limited to the edge of the Shelf from the Barne- gat profile eastward (Bigelow, 1915, p. 166, fig. 11, 12). And no part of the coastal water (within the limits surveyed by the Grampus) was as warm in 1916 as was the region between the Barnegat and Chesapeake profiles in 1913. Temperature and Salinity in November: — No observations were taken on Georges Bank in November. But several stations were occupied during that month between Marthas Vineyard and Chesa- peake Bay, especially welcome because they give us our first \'iew of the subsurface temperature, salinity, and plankton of this section of the coast water, for the autumn. Unfortunately, however, these salinities must be regarded with suspicion (p. 178). The most apparent change from August to November, one which might have been prophesied, was a decided cooling of the surface over the whole area, the actual readings now being from 11.3°-14.17° (the latter in the mouth of Chesapeake Bay) usually about 13°, as against 17°-24° in August (p. Ill, fig. 28). BIGELOW: COASTAL WATER EXPLORATION. 123 Below the surface an equally widespread change took place, for the November temperature curves for every station show either practical uniformity, vertically, or even a slight vertical warming, down to 30-60 meters, instead of the extreme vertical cooling which charac- terized the region as a whole in summer. Furthermore, although this layer of uniform temperature overlay a zone in which some vertical cooling obtained, the bottom water on the Shelf, at 30-80 meters, was as a whole considerably warmer in November than in August, the minimum temperature having risen from 4°-5° to about 8° off New York; from about 5° to about 9°-10° off Chesapeake Bay. A seasonal change of this sort was, of course, to be expected, in the absence of disturbances by extra limital currents, as the first step in the vertical equalization of temperature so characteristic of northern coastal waters in late autumn and winter. The temperature of the warm zone deeper down the slope was like- wise higher in November, when water as warm as 10° bathed the bottom below, say, 150 meters, as far north as the latitude of New York at least, than in August, a fact of biologic importance. The seasonal change in temperature is further illustrated by com- paring the profiles off New York for summer and for autumn. These, it is true, do not precisely coincide in location, except at their outer ends. But the geographic difference is not sufficient to presuppose any appreciable divergence in temperature, though it may for salinity (p. 120). x\long this line, the layer of 10°-12° water was some 60 meters thick over the inner part of the Shelf in November (Fig. 37), instead of only about 10 meters thick, as in August, with the November curve for 10° closely paralleling the August curve for 7°. Perhaps the most in- teresting feature of this November profile is its demonstration of the fact, already recognized, that autumnal cooling in our coast waters proceeds from the land, seaward. As pointed out (p. 132) this process had progressed so far on the inner half of the Shelf as nearly to obliter- ate the pre-existing stability of the wa.ter. But further oft'-shore only the immediate surface had yet been chilled by the cool land winds, consequently the underlying water at 20-50 meters was l°-2° warmer than the surface, so that the curves for 12° and 13° suggest a landward intrusion of off-shore water. This interpretation the salinities (if they are to be trusted) forbid; and in point of fact this apparent tongue is merely reminiscent of the maximum temperature reached at this level during late summer. As pointed out (p. 178) it is a question how much dependence can be placed on the November salinities. However, they probably 124 bulletln: museltvi of compaeatre zoology. represent the actual conditions fairly well as a series, even though no one determination in particular is trustworthy. With this reserva- tion, we find that the immediate effect of the outflows from the several large rivers and bays is no longer discernible at this season, which was to be expected, for the surface was then upwards of 32%o, even off Chesapeake Bay and off Long Island Sound, as against 30-3 1%^ in August (Fig. 36). And since no corresponding increase took place at 30-40 meter IcA-el, November salinity, like November temperature, was practically uniform vertically, down to this depth, at every station but one (10410). Mieter 0 Fig. 37. — Temperature profile crossing the Continental Shelf off Marthas Vineyard, November 10-11, 1916. These records afford no CAadence that any Gulf Stream water had flooded the Shelf since August. At the off-shore ends of the profile off Marthas Vineyard for Novem- ber (Fig. 38) and off New York for August (Fig. 23), where the geo- graphic locations coincide, the November salinities are slightly the lower, 34%o water now being restricted to depths greater than 100 meters, only actually touching the bottom at and below 120 meters, instead of at about 9-100 meters as in summer; nor have the high temperatures at 40-70 meters of Stations 10407 and 10408 any counterpart in salinity (p. 123). The profile deserves a further word because of the low bottom salinity (32.8%o) at Station 10407, which coincides wdth the minimum temperature (about 8°); and is almost BIGELOW: COASTAL WATER EXPLORATION. 125 exactly the same as the saHnity of the coldest water in July, though the latter lay some distance further up the Shelf (Fig. 22). This, of course, taken at face value, suggests that the very cold water moved do^^'n over the bottom during early autumn, with gradually rising temperature, which is not inconsistent with the August densities (p. 130). But it being possible that the 60 meter salinity at Station 10407 is too high, it is idle to discuss it further. The partial profile off Chesapeake Bay is not figured here because the bottom salinity at Station 10415, if correct, would result in a very peculiar course for the curve for 33%o, whereas if it is too high, as is Meter 0 Fig. 3S. — Salinity profile crossing the Continental Shelf ofif Marthas Vineyard, November 10-11, 1916. verv likelv, this would not be the case. Suffice it to say that in November salinity was practically uniform vertically down at least to 50 meters out to the 200 meter contour along this line; and that the boundary between waters fresher than, and Salter than 33%o was practically vertical down to about 75 meters. Below that level there seems to have been a seaward projection of fresher, as there certainly was of slightl}- cooler water. Certainly neither this, nor the preceding profile reached the inner edge of the Gulf Stream. Density. With the progress of modern oceanographic research it becomes increasingly clear that it is the juxtaposition of water masses of differing density which is primarily responsible for the most important 126 bulletin: museum of comparative zoology. ocean currents. And in coastal waters subject, as are ours, to \'iolent fluctuations in temperature and salinity, the vertical distribution of density and its seasonal changes, largely determine the activity of vertical currents, with all that follows in their wake. The subject is extremely complex, and no attempt is made here to touch on more than its merest outlines. For a discussion of the principles involved, their calculation, and application to a definite area I refer to the re- sults of the Canadian Fisheries Expedition (Sandstrom, 1919). Our previous work has shown, as indeed was to be expected, that the waters of the western side of the Gulf of Maine are normally very stable, vertically, in summer. And 1916 was no exception to this rule. But the extreme steepness of the density gradient from the surface downward, and the fact that the rate of vertical increase (i.e., vertical stability) was much less pronounced below 30-40 meters than above that level (Fig. 39) deserves emphasis, for its bearing on vertical circulation. The profile crossing Massachusetts Bay (Fig. 4) shows a close general correspondence between temperature and density, cold (5°) and heavy (25.5) water lying closer the surface in the center than near either the northern or southern shore; with the surface pro- gressively lighter from north to south. For these horizontal inequali- ties the topography of the bottom is largely responsible, the central part of the Bay being partially enclosed below the 30 meter level by Stellwagen Bank, hence retaining the low temperature and high density of winter below that depth, whereas seasonal warming had penetrated in small measure even to the 100 meter level in the waters off the north shore with their strong tidal currents. Thus there was a dynamic tendency toward horizontal circulation even in the deepest layers of the Bay, as well as toward a movement of the surface water from south to north. But how far such currents are effective, as against the strong tides of this region, can only be settled by more intensive study. Contrasted with the deeper parts of the Gulf (Stations 10344, 10345), the waters on Georges Bank show comparatively little vertical range of density (Fig. 39, 40); as they do of temperature (p. 105), and salinity (p. 105), a phenomenon resulting from active vertical circula- tion. And so far as the three Grampus stations in that region go, they suggest that the upper layers were less dense over the southern than over the northern edge of the Bank. Density increases, in general, from the land seaward, over the inner half of the Shelf off Marthas Vineyard (Fig. 41); but is more uniform, horizontally, in the cold band (Stations 10351, 10352). BIGELOW: COASTAL WATER EXPLORATION. 127 Meter 0 10 20 30 40 SO 60 70 80 SO 100 110 120 130 V ;32;3^ — - \ \ B^ -_ I^ A -..^ ^ — . ^ \ c i N D^ \ ^^ ^ \ V i\ \ \\ A^ \ i 2f.5 22 23 .5 24 .5 25 .5 26 Fig. 39. — Density curves, for typical regions, July-August, 1916. A, Gulf of Maine, Station 10344. B, off Delaware Bay, Station 10382 (lower part, probable curve). C, off Chesapeake Bay, Station 10392. D, on Georges Bank, Station 10347. iMeter Fig. 40. — Density profile from the Basin of the Gulf of Maine to Georges Bank, July 22-23, 1916. 128 bulletin: museum of comparative zoology. The vertical distribution of density was of essentially the same type in the coastal water off New York (Stations 103G2-10365, Fig. 42), and off Delaware Bay (Stations 10379-103S2, Fig. 39, 43) as in Massachusetts Bay; i.e., the vertical stability of the upper layers pronounced, that_of the bottom water slight, a phenomenon probably Met. Fig. 41. — Density profile crossing the Continental Shelf off Marthas Vineyard, July 24-26, 1916. chiefly responsible for the low bottom temperature here (p. 171). But within the influence of the outflow from Chesapeake Bay the intermediate water layer alone was notably stable, the density gradient being much less steep both from the surface down to 10-20 meters, and from, say, 40 meters down to the bottom than in the mid-la^'er. Consequently, while the upper and lower layers were insulated from BIGELOW: COASTAL WATER EXPLORATIOX. 129 each other, each was free to respond, with vertical currents, to any disturbing influence. The subsurface densities at the outer ends of the profiles off New York (Fig. 42), off Delaware Bay (Fig. 43), and off Chesapeake Bay (Fig. 44) deserve discussion because of their bearing on the permanence Met Fig. 42. — Density profile crossing the Continental Shelf off New York, August 1-2, 1916. of the warm band which follows the slope at about the 200 meter level (p. 113). Any movement, down the slope, of the considerable mass of much colder water on the Shelf would obliterate this warm band, with far reaching biologic effect ; indeed some such event was probably responsible for the destruction of Tilefish in 18S4. But these profiles show that there was no dynamic tendency of this sort in August, 1916, 130 bulletin: museum of comparative zoology. but just the reverse, the coldest water being considerably lighter (about 26.2='=) than the waters bounding it off-shore. The densities as indicated on the profile off New York (Fig. 42) would demand a movement of the cold bottom water down the Shelf to about the 60-70 meter contour (equal density \ying about 10-20 meters deeper at Station Fig. 43. — Density profile crossing the ContinenteJ Shelf off DelawEire Bay, August 11-12, 1916. 10366 than at Station 10365). But from here seaward, there was some dynamic tendency for the cold water to rise from the bottom, with the warmer, })ut Salter and hence denser off-shore water mo\'ing toward the slope below it, which is entirely consistent with the temper- ature profile (Fig. 22). Seldom, indeed, are the motions of water masses of different physical characters as precisely outlined by the temperature curves as in this case. All this points to a zone of active BIGELOW: COASTAL WATER EXPLORATION. 131 mixing over the slope; which all our data support. And the conse- quent dissipation of the cold zone at its outer edge, added to the gradual penetration of heat from above, amply explains the warming of the bottom water which took place by November. Only once, and then as a surface stratum not over 150 meters or so Met Fig. 44. — Density profile crossing the Continental Shelf ofiF Chesapeake Bay, August 21-22, 1916. thick, have we encountered undiluted Gulf Stream water on the Grampus cruises, even at the outermost end of our profiles (Bigelow, 1917a, p. 189); there is no actual e\adence that such water even reaches the sea floor on the slope, and densities suggest that the heavier mixed water pre\'ents this. But Gulf Stream water may approach close to land on the surface, before losing its distinctive 132 bulletin: museltm of comparative zoology. characters, after prolonged on-shore winds; and thus exert an im- portant, though sporadic influence on the coast water itself. Unfortunately the unreliability of the November salinities (p. 178) precludes any discussion of the densities for that month further than to point out that thanks to vertical equalization of both salinity and temperature down to 30-50 meters, the upper layers of water then possessed little or no vertical stability. PLANKTON. ZOOPLANKTOX, GuLF OF IMaINE. In spite of the abnormally cool water, the animal plankton of the Gulf was essentially of the same type in July, 1916, as in previous summers, i.e., consisted chiefly the copepod Calanus finmarchicus; with many Pseudocalanus and the other typically boreal organisms which are usually found there. And not only was the Arctic com- ponent, which reaches us in the spring, totally lacking in its western part in July and August, 1916, but forms which, while not purely Arctic, are primarily northern in origin were no more abundant there than we have usually found them in summer. Thus Calanus hyper- borcus was detected at one station only (10345, 100-0 meters, 1 speci- men); and only a single Clione limacina was taken in the Gulf (Sta- tion 10246, surface). On the other hand, none of the oceanic warm water forms, which occasionally appear there (Bigelow, 1917a, p. 246) were taken in July or August, 1916. Among the typically boreal organisms, Sagitta cJcgans was notably abundant in July at 30 meters off Gloucester (Station 10.340) where it formed an important element of the plankton. And in the center of Massachusetts Bay (Station 10341) where the upper layers (30-0 meters) were occupied by a swarm of Calanus, the catch at 80 meters consisted chiefly of Sagitta eJcgans. Large numbers were also taken in the channel between Stellwagen Bank and Cape Cod (Station 10342) though the Sagittae were overshadowed there in faunal im- portance by Calanus and Pseudocalanus. And S. elcgans was like- wise plentiful in the southwest corner of the Gulf (Stations 10344, 10346). As a whole Sagittae were most numerous relatively in the deepest hauls, as is illustrated by Station 10241, where none were taken on the surface, and relatively few in the swarms of Calanus at 40 meters, but where they constituted about half the catch at 80-0 meters. BIGELOW: COASTAL WATER EXPLORATION. 133 Sagitfa serratodentata did not appear at all in our hauls in the Gulf in 1916. Large, noticeable forms, which usually give a distinctive character to the Gulf of !Maine, were as a whole rare in the summer of 1916, apart from the Sagittae. For example, the amphipod Euthemisto compressa did not appear at all in the catches in ^Massachusetts Bay in July of that year (Stations 10340-10342), though then represented by an occasional specimen in the southwest part of the Gulf (Stations 10345, 10346), while its companion species, E. hispinosa was not taken there at all in the summer in question. A few E. compressa were, however, captured in the Bay in August, which reproduces our earlier experiences. Adult copepods other than CaJanus finmarchicus and Pseudocalanus were of so little faunistic importance in the Gulf, that the only species detected in the preliminary examination of the catches were one Calanus hyperhoreus (p. 132), a few Metridia lucens, Anamalocera, and Euchaeta norvegica (Station 10341, 80-0 meters, occasional specimens). And the scarcity of the latter in the deeps of the open Gulf (Station 10345) is especially noteworthy, for it is usually one of the most striking members of the plankton of its deeper layers, occurring in practically every haul from 80 meters, or more (Bigelow, 1917a, p. 292, fig. 88). ' Adult euphausiids were similarly rare in July, being represented by occasional specimens of Thysanoessa raschii (Station 10341), T. inermis (Stations 10341, 10342), and Mcganydiphanes rwrvegica (Station 10341) only. This was also true of the one pteropod, Lima- cina halea (Stations 10342, 10344), which is endemic in any numbers in our Gulf. But the presence of many euphausiid larvae off Cape Cod (Station 10344, 10345) suggests that this group may have been more important later in the season, as was certainly the case in 1915.^ And the scarcity of Limacina halea is probably a seasonal phenomenon of the same sort, for we have never found it common off Massachu- setts Bay or along Cape Cod so early in the season, though it ap- proaches the western and northern coasts of the Gulf as the summer advances. Tomopteris was not taken in any of the July hauls. Considerable numbers of the neritic Hydromedusa Mitrocoma cruciata were present in Massachusetts Bay in July (Station 10340), though it has usually disappeared there by that season (Bigelow, ' In that year Thysanoessa inernis appeared in large numbers in August off Cape Ann, where it was rare earlier in the summer. 134 bulletin: museum of compaeative zoology. 1917a, p. 305). And Aglantha digitalc occurred in some numbers at all the July stations north of Georges Bank, notably in Massachusetts Bay (Stations 10340, 10341, 10343). Melicertwn campanula was encountered at one July Station (10340). The total absence of Stavrophora mertensii in the catches of 1916 deserves mention in ^^e^v of the steady decrease which seems to have taken place in its numbers since 1912 (Bigelow, 1917a, p. 305). No explanation for this change is yet apparent. The Grampus did not \-isit the eastern or northern parts of the Gulf in 1916. But Dr. A. G. Huntsman, of the Biological Board of Canada, has very kindly supplied the following notes on the occurrence of certain plankton forms in the Bay of Fundy during that summer. The relative numbers of distinctly northern forms present in this Bay during any given season are good evidence for, or against, an un- usual influx of northern water, for only in small numbers are any of them {e.g., Paraihcmisto ohlivia) endemic there, it being, as he points out, entirely unsuitable for the development of most pelagic eggs and larvae. Such index animals are Paratheniisto oblhia; Tomopteris helgolandicai Mertcnsia ovum; Limacina helicina: — perhaps CUone limacina (p. 174) and Aglantha digitale. Dr. Huntsman's notes re- garding them are as follows : — ■ " Parathemisto ohlivia was found very occasionally and only as isolated indi- viduals, in 1916. It is an estuarial species occurring in large numbers in the Gulf of St. Lawrence, particularly in the estuaries of the northern part. It is virtually absent from the banks on the south coast of New Foundland, and is, therefore, an indicator of St. La'OTence water as opposed to that of the Labra- dor Current. But its presence in the Gulf of Maine and in the Bay of Fundy is not conclusive evidence of St. Lawrence water as it breeds successfully in the estuary of the Kennebecasis river near St. John, New Brunswick. Tomo- pteris catharina (= helgolandica) was not infrequent in May of 1916, but its numbers decreased during the summer. It was obtained however, in every month until October at least. This species appears never to be abundant south of the Xew Foundland banks, where it breeds. Wright (1907) how- ever, reported the young from Canso (as "T. mariana"). It is rare in the Gulf of St. Lawrence and does not seem to breed there. As its range ex- tends as far south as New York, the adults must survive for a long time in the south-flowing coastal water. Mertensia ovum was not found in 1916. Limacina helicina does not seem to occur in the Bay of Fundy, and was not found in 1916. ^ CUone limacina was not found in 1916, but has occurred in the Bay of Fundy in other years, sometimes in considerable quantities. It is abimdant on the New Foundland banks and occurs regularly in the Gulf of St. 1 For its occurrence in the Gulf of Maine see Bigelow, 1917a, p. 248. BIGELOW: COASTAL WATER EXPLORATION. 135 LawTence and off Nova Scotia. Larvae are found over this entire northern region, but not in the estuaries. Aglantha digitale is likewise verj^ rare in the Bay of Fundy, and scarcely any specimens were taken in 1916. However, it resembles Sagitta clegatis in breeding successfully in large numbers over an ex- tensive area of coastal waters to the north and east." These notes show that there was no unusual immigration of north- ern plankton animals into the Bay of Fundy in 1916, indeed, rather less than usual. Our previous experience had been that the zooplankton of our Gulf is much the same in November as in midsummer (Bigelow, 1914b, p. 403); and this was true in 1916. Thus the deeper November hauls caught chiefly large Calanus finmarchicus and Pseudocalanus, the only other copepods occurring in any numbers below 25 meters at these stations being Mdridia luccns, Euchaeta nonrgica, and an occasional Calanus hypcrborcus (Station 10401). Further in regard to copepods I need only record a swarm of Tcmora longicornis, with occasional Centropages hamatus on the surface off Gloucester (Station 10399) ; of Pseudocalanus and young Calanus on the surface at Sta- tions 10400, and 10401; of Centropages hamatus on the surface at Station 10404; many MetricUa lucens on the surface off Penobscot Bay (Station 10402) ; and Anomalocera at Station 10400. In November, just as in July, the only chaetognath detected in the Gulf was Sagitta elegans, large specimens of which occurred in greater or less numbers in all the deeper hauls, particularly at Stations 10399 and 10400, where they were in great abundance at 60-0, and 90-0 meters. And the fact that small specimens (10-12 mm. long) were taken on the surface at all the November stations, abundantly off Cape Ann (Station 10399), is worth noting. The general thesis that the number of Euthemisto present in the Gulf increases during sum- mer and autumn is borne out by the fact that this hyperiid occurred at all the November stations. As a rule its few representatives were young, or at most medium sized; but at Station 10404 it was repre- sented by large specimens of l)oth species, E. covipressa and E. bis- pinosa, such as often swarm along the Continental Slope. Euthemisto comprcssa was taken at Stations 10400-10404 in the Gulf; E. bispinosa at Stations 10399, 10401, and 10404. A similar increase evidently took place in the case of Limacina balea which was detected at Stations 10399, 60-0 meters; 10400, surface; 10402, surface; once, 10399, in considerable numbers. Whether the same change characterized the euphausiids is not clear, for while many large Myganyctiphanes norvcgica, -sA-ith a few Thysanocssa raschii were taken on the surface 136 bulletin: museum of comparative zoology. at Station 10402, a few jVIeganyctiphanes and T. inermis at lSO-0 meters, Station 10401, and a few Meganyctiphanes at Station 10400, none were detected at Stations 10399, 10403, or 10404. The occurrence of young Aglantha digitale in swarms off Cape Ann (Station 10399, surface, and 60-0 meters) and, in smaller numbers on Stellwagen Ledge (Station 10403), is worth a word, as is a swarm of Beroe (probably B. cucumis) further off shore (Station 10401, SO-0 meters). And, finally, I may add that Tomopteris was wholly lacking in November, just as it was in July (p. 133). Discussion of the pelagic fish eggs, and young fishes taken in 1916 is reserved for a future comnumication. The Qvaniitaiirc Hauls. Thanks to abundant Calanus, the catches made in the Gulf in July with the quantitative nets i were xerx rich as illustrated by the following table : — Large Vol. Vol. COPEPODS CoPEPODS Large Calanus Calanus PER SQ. PER APPROX. PER PER PER Depth METER CUBIC M. PER CUBIC SQ. METER CUBIC Station METERS C.C. C.C. SQ. METER METER METER 10340 4.5 125 2.5 9.3000 2066 38000 844 10341 80 250 3.1 265000 3312 78700 983 10342 55 250 4.5 338000 6145 111000 2018 10344 80 225 2.5 179250 2240 74000 925 10345 150 200 1.3 139600 930 83600 557 10346 62 200 3.2 247200 3987 36800 1390 Average 208 2.8 210340 3113 78600 1119 And the Calanus population being largely confined to the deeper levels, as CAndenced by the po^'erty of the surface catches, the plankton was eWdently much denser, locally, than e^'en the amounts per cubic meter would suggest. For example, the haul at 40 meters, at Station 10344, with the meter net, yielded about 6 liters in 15 minutes, up- wards of two and one half million large Calanus. These are considerably larger amounts of plankton, both per square, and per cubic meter, than were usually present in these waters in the summer or autumn of 1915, the only previous year when the same type 1 Taken with the Michael Sars net, \ meter in diameter, as in 1915 (Bigelow, 1917a, p. 307). The same method of counting was employed as previously (Bigelow, 1914a, p. 128). BIGELOW: COASTAL ^VATER EXPLORATION. 137 of quantitative net was used; or, taking our records at their face value,! 1912, 1913, or 1914, as illustrated by the following table for the southwestern part of the Gulf: — ■ Extreme vols. Extreme vols. Average vol. Average vol. PER SQ. METER PER CLBIC M. PER SQ. meter PER CUBIC M. Year CC. CC. CC. CC. 1912 15-250 105 19131 180 1.4 ISO 1.4 1914 60-210 .42-1.9 102 .94 1915 30-150 .7-1. 103 .74 Average for 4 years 122 1.03 Copepods were also more numerous, as follows: — Extreme nos. Extreme nos. Average no. Average no. PER PER per PER YE.A.R 3QL:.\RE METER CUBIC METER SQUARE meter CUBIC METER 1912 7,500-235,300 45-1012 103,575 723 1913 2 50,500 3SS .50,.500 388 1914 53,500-189,500 306-1722 101,562 712 1915 42,500-112,500 368-1142 65,800 765 Total a vera g e = 80.359 647 1916 93,000-265,000 930-6145 210,340 3113 ZoopLAXKTOx OX Georges Baxk axd OFF Marthas Vixeyard. Special interest attaches to the plankton along the outer (southern) edge of Georges Bank, because of the mixing of Cabot Current, Gulf Stream, and Bank water which takes place there (p. 165). Over the Bank itself (Stations 10347 and 10348) the July plankton, very scanty in amount (Station 10347, total bulk of vertical haul, 15, c.e.), was of the usual boreal type, without any distinct northern component, i.e., chiefly Calanus finmarchicus and Pseudocalanus, Sagiita clegans, Euthcmisto comprcs.sa, together with such neritic forms as Cyanea, the 1 In 1912, 1913, and 1914 the Hensen, in 1915 the Michael Sars nets were used. In neither case has the coefficient of filtration been allowed for in calculating the results of the catches. 2 One haul only. 138 bulletin: museum of comparative zoology. free floating campanularian hydroids which apparently characterize the summer plankton of these shallow waters (Bigelow, 1914b, p. 414; 1915, p. 306); and many crab larvae. And the presence of these neritic organisms so far from land, illustrates how similar to an actual coast line in their effect on the plankton, are these shallow off-shore banks. It was along the outer edge of the Bank, and over the Conti- nental Slope (Stations 10349-10352) where the temperature was very low (p. 101) that the plankton might have been expected to show an x\rctic component, if anywhere. But nothing was taken which could be so described. However, tropical organisms, usually so common here at this season, were poorly represented in 1916, the nets yielding much the same assemblage as on the Bank, and in larger amount; ^ notably C alarms fimnarchicus, and Pseudocalanus, with lesser amounts of Mdridia hicens; occasional Eiicharta uorvcgica (Stations 10349, 10352) and Euchcirilla rostrata (p. 147) with large Euthcmisto, both E. comprcssa and E. bispinosa, in considerable numbers, as is usually the case along the Slope. A few Lwiacina halea were likewise detected, as were Tkysanoessa incrmis and T. raschii (Station 10351), while Ncmatoscclis mcgalops was present in large numbers at the Station furthest off shore (10352). Sagitia scrratodcnfata, not detected in the Gulf in the summer of 1916 (p. 133), was as numerous as S. elegans at Station 10349 and outnumbered it at Station 10351, while at Station 10352, neither of these chaetognaths was found, they being replaced there by occasional specimens of Eukrohnia hamata and Sagitfa ma.vima^ in the deep haul (500-0 meters), which also yielded a beautiful example of the mesoplanktonic medusa, Pcriphylla hyacmthina Steenstrup. The only animals to which a Gulf Stream origin can safely- be- ascribed are a few Salpafusiformis at Station 10349, many at Station 10352; a single Physophara hydrostatica (Station 10352), a large Pyrosoma (Station 10352), a few fragments of Gulf weed (Sargassum sp., Station 10353). This community contrasts, as strongly as did the salinities and temperatures of 1916, with the Gulf Stream plankton encountered at this general locality in 1914 (Bigelow, 1917a, p. 245), but almost exactly reproduces what we have previously found at this season at the same relati^'e position on the slope oft" the northeast face of Georges Bank and off southern Nova Scotia (Bigelow, 1917a, p. 245), rather less tropical than the former, less northern than the latter. 1 At Station 10349 the yield of the quantitative net was at the rate of about 175 c.c. per square meter of sea surface. 2 I follow Huntsman (1919) in regarding the S. maxima of Conant (1895) Ritter-Zahony (1911) and Kramp (1918) as distinct from S. lyra. with which Michael (1911, 1919) has united it. BIGELOW: COASTAL WATER EXPLORATION. 139 The shallow waters off Marthas Vineyard (Stations 10354-10356) were likewise occupied by a typically boreal plankton, as was the case in 1913 also, chiefly the copepods Calanus finmarchicus and Pseudo- calanus, with Sagitta elcgans taking the place of S. serratodcntata. But large Calanus were less numerous close to land (Station 10357), where they were replaced by swarms of small copepods yet to be identified. And in general the various neritic animals played an in- creasingly important role as the Grampus approached land. Of these I need mention only abundant decapod larvae in the 30-0 meter hauls at Stations 10355 and 10356; the Medusa Obelia at Station 10356, the copepod Tcmora longicornis (Stations 10356 and 10357), and Evadne (Station 10357). Zooplankton, Marthas Vineyard to Chesapeake Bay. The vmusually low ocean temperatures which prevailed over the Shelf south of New York during the summer of 1916 add interest to the plankton of those waters. In July, 1913, the only previous summer for which plankton records are available for this region, copepods, which played the major r61e east of New York, were practi- cally negligible further west and south, their place being taken by a combination of immigrants from the inner edge of the Gulf Stream, particularly Salpae, on the one hand, and neritic animals, e.g., Mnemi- opsis, on the other, together with swarms of Pleurobrachia whose exact faunal status is still doubtful. And the warmer and more saline waters over the Continental Slope carried with them a tropical plankton community. In 1916 the summer catches were of quite a different complexion, a rich copepod plankton occupying the whole breadth of the Shelf oft" New York, chiefly Calanus finmarchicus, associated with which were such other boreal forms as Sagitta elegans, and Euthemisto, i.e., much the same as prevails in the Gulf of Maine. The Calanus Comnnmity. This Calanus community, to give it the name of its most important member, reached as far south as Chesapeake Bay (Fig. 45). But south of New York it was lacking over a widening coastal zone (Stations 10377, 10378, 10379, 10387, 10390, 10391), where it gave place to neritic copepods (p. 146), ctenophores (p. 158), and other neritic forms, e.g., crab (Callinectes) larvae. And on the line oft' Chesapeake Bay it was restricted to the extreme outer edge of the Shelf (Station 10392), a fact suggesting this as its extreme southern limit. The importance of Calanus finmarchi- cus in the natural economy of boreal seas being so great, and its 140 bulletin: museltm of comparative zoology. presence in mass totally unexpected south of New York in July and August, the precise conditions under which it occurred there in the summer of 1916 require examination. To begin with, Calanus was by no means equally distributed throughout the area occupied by it (Fig. 45-47) or at all levels in the water, being most plentiful over the 4f IQ' Fig. 45. — Chart showing the relative abundance of Calanus, horizontally, July- August, 1916. Cross hatched, Calanus swarms; single hatched Calanus moderately numerous; dotted, few or no Calanus. outer part of the Shelf in the deep hauls (Station 10364, 40-0 meters. Station 10365, 50-0 meters. Station 10370, 50-0 meters, Station 10373, 60-0 meters. Station 10375, 40-0 meters. Station 10381, 69-0 meters. Station 10386, 45-0 meters, Station 10392, 75-0 meters), where the meter net yielded from 1-3 quarts, chiefly large Calanus. Nearer land fewer were taken (p. 139). And its numbers diminished similarly BIGELOW: COASTAL WATER EXPLORATION. 141 passing seaward from the 100 meter contour, there being very few at any level over the Slope (Station 10368, 500-0 meters, none, Station 10384, Station 500-0 meters, few, Station 10393, 250-0 meters and 500-0 meters none). Our data on the absohite abundance of Calanus in these waters is confined to quantitative hauls at two stations ofP New York, at one of which (10363), the horizontal haul yielded few, at the other (10365) 2^ quarts at 50-0 meters. And the catches of the quantitative nets Fig. 46. — Profile off New York (Stations 10362-10370) showing the relative abun- dance of large Calanus, August, 1916. Hatched (1), rich, 1 liter or more taken; dotted (2), less than 1 liter; and blank, occasional or none. differ correspondingly, the total volume of plankton being about 35 cc, of large Calanus 9,500 per square meter of sea surface at the former; "100 cc, with 27,000 large Calanus at the latter. Assuming that practically the whole catch of Calanus was made below 10 meters depth at these stations, an assumption justified by the fact that very few were taken on the surface, the number per cubic meter of water was at least 430 at Station 10363, 540 at Station 10365. And proba- bly the actual density of the Calanus population at the latter was locally many times greater than this, the richness of the horizontal net catch at 50 meters, just noted, suggesting a localization at and near that level. In estimating the abundance and bathymetric occurrence of Calanus finmarchicus over the rest of what I may call the Calanus zone, we must rely on the results of the quantitative hauls, which must be used wath caution, first, because they fish both on the way down and on the way up, as well as at the level at which the major part of the haul is made; second, because the latter can not be known exactly, for even 142 bulletin: museum of comparative zoology. if a depth recorder be attached to the net it can only reveal the greatest depth reached; third, and most serious, because the amount of water filtered by the net varies, even when the hauls are of the same duration, owing to variations in the speed of the vessel. In hauls deeper than, say, 200 meters, so much time is occupied in lowering and hoisting the nets, and so uncertain is the level at which they work, that these objections become serious enough to vitiate all but the most general quantitative results. But this is not true of our shallow hauls on the Continental Shelf, where the few mj'nutes spent in lowering and hoisting the net are only a tri\'ial part of the total time it is at work. And inasmuch as the catches of vertical hauls in these shallow waters are very scanty, compared with the yiel Is of the horizontal nets at the same stations, it is safe to assume that at least the major part of the latter can be credited to the approximate level at which the net is working during most of the haul. This can not be known absolutely. But when the extreme depth of the haul is only 40 or 50 meters, the total amount of wire outboard is so short that an error of more than 10 meters one way or the other is unlikely. The variations in the length of the column of water fished through prevent the results of such hauls from being strictly comparable one to another. But when all the hauls are of the same duration (30 minutes is the Grampus standard), the nets of the same diameter (or the catches reduced to one standard diameter), and when the speed of the vessel is so regu- lated as to keep the angle of the A\are nearly constant (say 45°-50°) the results have considerable comparative value. At the worst, they tell us which regions or levels are well, which sparsely, populated; for example that the surface was uniformly barren of large Calanus south of New York, irrespective of the time of day, as illustrated by the following data for the surface catches at stations where this cope- pod was plentiful deeper down in the water: — Large Large Station Caianus Time Station Calanus Time 10362 none 9 A.M. 10380 occasional 2 A.M. 10363 11 3-4 P.M. lw;i8l few 8 A.M. 10364 f< w 7-8 P.M. 10382 a noon 1036.") none 7 A.M. 10384 none 9 P.M. 10369 {( 5 A.M. io;,86 occasional 5 P.M. 10370 occasional 4 P.M. 1,)..'.)' a 3-6 A.M. 10372 few 2 A.M. . 1 • none 9 A.M. 10375 occasional S A.M. n 3 P.M. BIGELOW: COASTAL WATER EXPLORATION. 143 Large amounts of Calanus were taken only from depths of 20 or more meters. For example, at Station 10362 very few were taken at 10 meters, many (1 pint) at 20 meters; Station 10363 many at 20 and at 30 meters; at Station 103S1, only a few at 25 meters whereas a swarm (2 quarts) was encountered by the 70 meter net. On the other hand the deeper water (100-120 meters) at the outer edge of the Shelf was less productive in Calanus than the mid-depths, as illustrated by Station 10370, where the haul at 120 meters yielded only 1 pint, whereas there were swarms at .50 meters. At Station 10382 Calanus was apparently comparatively sparse at all levels down to 120 meters (bottom). The largest catches south of New York were made along the outer ^ of the shelf, at from 40-70 meters depth (Stations 10364, 10365, 10370, 10373, 10375, 10381, 10386, 10392). And though no study has yet been made of the proportion in which the different growth stages occur, it is worth noting that all these rich catches were composed in the main of very large individuals, giving the plankton the same monotonous aspect that so often characterizes it in the Gulf of ISIaine. Only at one Station (10362, 10-0 meters) along this part of the Shelf, and once near Long Island (Station 10396) were young Calanus numerous. Never before in our Grampus cruises, has it been possible to corre- late the occurrence of any of the more important plankton animals as closely with the physical state of its environment as can be done for the Calanus stock south of New York in the summer of 1916. The most cursory comparison of the temperature profiles (Fig. 22, 24, 26) with those for the relative abundance of large Calanus at different levels (Fig. 46, 47) shows a very close correspondence between catches of 1 quart or over and water of 4°-7°, the precise temperatures in which the richest hauls were made being as follows: — Depth of Approximate Depth of MINIMUM Minimum CATCH OF HAUL temperatures temperatures Station CALANUS METERS meters 10364 1 swarm 40-0 40 5.5° 1036.5 1 swarm 50-0 40-60 .8° 10370 swarm 50-0 50 4.9° 10370 few 1 18-0 — — 10373 3 quarts 60-0 60 4.5° 10375 l| quarts 40-0 40 6° 10381 few 2.5-0 — — 10.381 2 quarts 70-0 30-65 4.5° 10395 3-4 quarts 38-0 38 9° 1 Taken with net .63 meter in diameter. 144 bulletin: museum of comparative zoology. This relationship between cold water and abundance of large Calanus is further illustrated by the close correspondence between the charts of minimum temperature on the Shelf (Fig. 21) and of the horizontal limits of the Calanus swarm (Fig. 45a). Station 10395, where the minimum temperature was about 9°, was the one exception to the rule that Calanus swarmed only in water of 6° or colder. But we have occasionally made large catches of Calanus in water warmer than 10° in the Gulf of ]Maine in August (for example, at Station 10027, 1914a, p. 103), though usually, it is true, where the underlying layers were colder. As a whole the salinity of the waters most densely populated by Meter 0 Fig. 47. — Profile off Chesapeake Bay. showing the relative abundance of large CcJanus, August, 1916. Shading as in Fig. 4 and 6. Calanus was likewise decifledly uniform, that of the coldest water on the Shelf south of New York being about 32.6%o ^^1 along this zone from north to south (p. 112). The presence of great numbers of Calanus (the largest catch of all) off Long Island (Station 10395) as late as August 26 shows that no diminution had taken place in its numbers there during the three weeks since the Grampus sailed south from New York (Station 10362, August 1); a constancy recalling that temperature had risen only slightly, salinity not at all (p. 182) during the same period. And even as late in the season as this, large Calanus were taken in some numbers close in to the Long Island Beach (Station 10396). But by mid- November Calanus had decreased greatly in numbers over such parts of the Shelf as were visited by the Grampus on her autumn run from Cape Cod to Chesapeake Bay (Stations 10405-10417), for nowhere BIGELOW: COASTAL WATER EXPLORATION. 145 at that season did the nets yield any such masses of this httle crusta- cean as in midsummer. And the shrinkage was particularly evident on the outer part of the Shelf off Delaware Bay, where the Calanus swarm of August (Station 10370) had dwindled to only a fraction by November 12 (Stations 10409, 10410). Calanus finmarckiciis was, however, taken in every November haul south of Cape Cod, except at the mouth of Delaware Bay (Station 10417), i.e., it had colonized the coastal zone where it was wanting in August. Furthermore, the surface Avas by no means as barren of large Calanus in November as in August, most of the surface hauls then yielding a considerable number of medium sized individuals. This vertical equalization of Calanus corresponds to the vertical equalization of temperature and salinity which takes place in autumn (p. 123). And it suggests that the failure of the southern Calanus swarm to migrate to the surface during the midsummer nights, as it so often does in the Gulf of Maine and elsewhere (p. 143) was due either to the very high surface tempera- ture, or possibly to the very low surface density. With the advance of autumn both these barriers are weakened by surface cooling, until in winter, thanks to the vertical uniformity of the water, the only physical barriers to vertical migration are sunlight and geotropism. ^^'hile (p. 143) no statistical study of the proportions of Calanus of different ages has yet been attempted, the fact that young stages were scarce, or absent, even on the surface, in November, is good evidence that no active multiplication of Calanus was taking place at that time, in contrast to the local swarms of juveniles encountered only shortly prcAious in the Gulf of Maine (p. 135). No quantitative hauls were made during November; hence we have no data as to the actual numbers of Calanus present in the water at that time. A notable difference l^etween the Calanus community south of New York and the boreal plankton of the Gulf of !Maine is the scarcity of the small copepod, Pseudocalanus among the former. In 1913 Pseudo- calanus was not detected at all south of New York (1915, p. 292): in August 1916, apart from the Gulf of Maine, and Georges Bank, (Station 10347) it has been detected at two stations only, both on the line off New York (Stations 10363, 10365). Adults, at least, were equally lacking south of this in November, though it is possible that it may be represented among the yoimg copepods taken at that season.^ Other Copepods. Where Calanus swarmed, few other large adult copepods occurred at the same level; (all the catches contain young 1 November occurrences of Pseudocalanus are Stations 10399-10407. 146 bulletin: museltvi of comparative zoology. and microcopepods, the identification of which must await the spe- ciahst), but over the coastal zone Ccntojxigcs ty pints was taken in numbers ' (Fig. 48), particuhirly in the upper levels; and it was often plentiful on the surface where Calanus swarmed deeper down, for example, at Stations 10375, 10381. How closely this small copepod was confined to the immediate surface is illustrated by the fact that it abounded there at Stations 10375, 10380, 10381, 10383, 10386, though few or none were taken at the corresponding hauls from 40 or more meters; C. typicus was equally widespread in these waters in 1913 (Bigelow, 1915, p. 287). Cenirapages hamatus is more northern, oc- cupying much the same zone in the Gulf of St. Lawrence as does typicns south of Cape Cod (Willey, 1919, p. 200). In 1916 hamatus was detected only twice, off New York (Station 10394) and in Vine- yard Sound (Station 10396). As is well known, many species of copepods are more neritic in habit than is Centropages. And such occur in greater or less numbers in most of our hauls near land south of Xew York. But until the exami- nation of them is complete, only the more striking (on account of abundance, or reafly identification) need be mentioned. Such is Acartia tonsa, which swarmed locally off Delaware Bay in August (Stations 10377, 10378, 10379), within the immediate influence of the outflow from the Bay. Lahidoccra acstiva was similarly limited to the region off the mouth of Chesapeake Bay (Stations 10387, 10389, 10390). But as this genus is abundant at Woods Hole during the summer (Wheeler, 1901), and is recorded by Willey (1919, p. 203) from the Gulf of St. Lawrence, it e^'idently covers a very wide range of latitude along the American coast. Its absence from most of our haids is due to their location in the open sea, some distance from land. Neither Acartia nor Labidocera was taken in November. Less distinctively neritic than the foregoing, though less oceanic than Calanus, is Tcmora longicornis. As noted, this species swarmed off Marthas Vineyard (Stations 10355, 10356, 10357) ; it also occurred off New York (Station 10362), but not further south, which repro- duces our experience in 1913. But by November when the Grampus again encountered it in numbers across the whole breadth of the Shelf off Marthas Vineyard (Stations 10405, 10406, 10408) it had extended its range far to the southward, being then taken both off Delaware Bay (Station 10411) and off Chesapeake Bay (Station 10417). The large, easily recognized copepod Mctridia luccns is rather more oceanic in our waters than Calanus fiumarchicus, though often taken ' Ceniopages typicus has so far been detected at Stations 10362, 10375, 10379, 10380, 10381, 10386, 10387, 10388, 10391, 10394, 10395, 10396. BIGELOW: COASTAL WATER EXPLORATION. W, with it. In August, 1916, M. luceus was taken at the following sta- tions south of Cape Cod; 10349, 10352, 10365, 10366, 10369, 10375, 10382, 10393. As the chart shows, these captures are localized along the outer part of the Shelf, and over the Slope, except for one locality off Delaware Bay (10375). In November M. lucens was not taken south of Cape Cod, though widely distributed then in the Gulf of Maine (Stations 10399-10403). The Arctic M. longa, which occasionally appears in the Gulf of Fig. 48. — Occurrence of certain copepods, July-August, 1916. A, Acarlia tonsa. C, Cenlropages lypicus. E, Eucheirella roslrata. H, Cenlropages hamalus. L, Labido- cera. M, Metridia lucens. N, Eiichaeta norvegica. P, Pleuromamma. T, Temora longicornis. The solid curve marks the general line of demarkation between neritic and oceanic species west and south of Georges Bank. Maine (Bigelow, 1917a), and is widespread in more northern waters on this side of the Atlantic (Willey, 1919), has not been detected in the Grampus hauls of 1916. Xo copepods to which a true tropical origin can be ascribed, were taken in 1916. But the list does include two large conspicuous forms which are typically oceanic in the North Atlantic, and which have usually been found, on our cruises, in the zone of mixed water along the Continental Slope, i.e., Pleuromamma and Eucheirella rostrata. The former occurred off the Slope of Georges Bank only (Station 148 bulletin: museum of comparative zoology. 10352, 500-0 meters), the latter over the Slope and on the outer edge of the Shelf (Stations 10349, 10351, 10352, 10368, 10370, 10382, 10384). Eucheirella was limited to the deeper levels, as exemplified by Stations 10370 and 10382, where it occurred in the nets from 120 meters, but not in the 50 meter hauls. ^ Eucheirella is similarlv dis- tributed off Nova Scotia OYilley, 1919, p. 189, fig. 9). It occasionally penetrates within the Gulf of ]\[aine (Bigelow, 1917a, p. 246). South of the Gulf of Maine, Euchacta norvegica is strictly limited to depths of 100 meters or more, outside the Continental Shelf. In 1916 it was taken at Stations 10352, 500-0 meters, 10368, 450-0 meters, 10384, 500-0 meters (Fig. 48). A single large Cahinus hyprrborcus at Station 10381, 70-0 meters, deserves special mention, as being the most southerly known occur- rence of this Arctic species, so common from Xova Scotia northward (^Yilley, 1919). Other conspicuous copepods so far identified are Eucalanus (Station 10389), and the large blue Anomaloccra paitcrsoni (Stations 10342, 10392, 10393, 10395, 10396, 10398). Neither of these were detected in the tows south of Cape Cod in November. Enlhemisfo. The hyperiid amphipod Euthemisto occurred gen- erally throughout the Calanus zone west of Cape Cod and south of New York in August, the relative abundance of the two species, E. compressa and E. hispinosa, at that time being as follows^ : — ■ Species Species Species Species Station PRESENT PREDOMINANT Station PRESENT PREDOMINANT 10362 C. B. 10380 C. B. C.B. 10363 C. B. B. 10381 C. B. B. 10364 C. B. B. 10382 C.B. C.B. 10365 C..B. C. B. 10384 0 10368 C. B. C. 103S6 C. B. C.B. 10369 C. B. C. B. 10387 10370 C. B. B. 10389 10373 C. B. B. 10390 10375 C. B. C. B. 10.392 C.B. 10377 10393 C. C. 10378 10394 C.B. 10379 10395 C.B. B. 1 The depths of the other captures of Eucheirella are. Station 10.349, 130-0 meters. Station 10351, 160-0 meters. Station 10352, 500-0 meters. Station 1036S, 450-0 meters, and Station 10384, 500-0 meters. ' C. = compressa. B. = hispinosa. C.B. = about equal numbers of the two species. Pre- dominant species named only for stations where one is several times as numerous as the other. BIGELOW: COASTAL WATER EXPLORATION. 149 The limits of E. compressa and E. hispinosa correspond almost exactly ^^•^th those of Calanus finmarchicus. The apparent absence of E. bispiiiosa oA'er the Continental Slope off Chesapeake Bay (Station 10393), where compressa occurred in small numbers, is an exception to our previous experience that of the two species hispinosa is the more oceanic, as is the fact that it was along the middle of the Shelf that hispinosa predominated over compressa. But the relative abundance of these two species fluctuates much from haul to haul. Euthemisto, like large Calanus, only occasionally appeared in the surface hauls in mid-summer even where numerous deeper down, and then (as at Station 10395) usually as small specimens too young for specific identification. But on the outer edge of the Shelf off New York (Station 10369) young stages of both species formed the bulk of the rather scanty surface catch.i Although Euthemisto appeared so regularly in the deeper tows on the Shelf side by side with Calanus, it was, quantitatively, far less important than the latter, only once swarming in August (Station 10395). Other stations south of New York, where it formed a considerable part of the August plankton were 10362, 10363, 10370, 10375, 103S0. At all these locahties small, or medium sized specimens alone were taken, as has usually been our experience on the inner part of the Continental Shelf. Both species occurred at practically all the November stations ^ south of Marthas Vineyard, except at the mouths of Delaware and Chesapeake Bays (Stations 10411, 10417), where none were detected, hispiiiosa being then the predominant member of the pair over the outer part of the Shelf south of Delaware Bay (Stations 10412-10416), and over the Slope oft' New York (Station 10408), while compressa was decidedly the more numerous of the two close to Marthas Vine- yard (Station 10405). Elsewhere the two were roughly equal, or, as at Station 10407, compressa predominated at one level (in this case, surface) hispinosa at another (75-0 meters). Euthemisto was rela- tively a more important factor in the plankton of these waters in November than in summer, thanks to the diminution in the numbers of Calanus. Notably large amounts were taken at Stations 10415, 90-0 meters, and 10416, 40-0 meters (about \ liter in each case): juveniles swarmed on the surface at Station 10406 and 10412. But only occasionally (Station 10407, compressa) were any very large specimens noted at this season. Like Calanus, it was less strictly ' Euthemisto has been detected in the surface hauls at Stations 10364, 10368, 10369, 10373, 10392. 10396. ' E. compressa, detected at Stations 10405-10407, 10409, 10410, 10413-10416 as well as in the Gulf of Maine (p. 135); E. hispinosa at Stations 10406-10410, 10412-10416. 150 bulletin: museum of comparative zoology. confined to the deeper water-layers in November than in August, occurring in the surface catches for Stations 10405-10410, 10412 and 10413, twace in large numbers (Stations 10406, 10412). Euphansiids. Only a preliminary examination of the euphausiids has yet been made; hence some species may have been overlooked. But inasmuch as the group was practically non existent in the coastal water south of New York in the summer of 1913, its occurrence there in 1916 is worth discussion. Thysanoessa inermis occurred among the Calanus, off Marthas Fig. 49. — Occurrence oi Euthemisto compressa (o). and E. bispinosa X, or both species ®, July- August, 1916. •= Larvae. Vineyard in July, 1916 (Station 10354); likewise (usually in compara- tively small numbers and represented by small individuals) at most of our stations in the Calanus zone south of Xew York in August. But its area of distribution was less extensive than that of Calanus, neither approaching as close to the land, on the one hand, or extending out- side the Continental Shelf on the other, while Chesapeake Bay marked its extreme southern limit. Thysanoessa hngicaudata was taken side by side with T. inermis off Delaware Bay at this time; displaced it over the outer edge of the Shelf off Xew York; and was the only BIGELOW: COASTAL WATER EXPLORATION. 151 Thysanoessa detected in the hauls made in deep water over the Conti- nental Slope. The occurrences of the two south of Xew York as so far established, are as follows: — T. incrmis, Stations 10363, 10364, 10370, 10373, 10380, 10381, 10386, 10394, 10395; T. lomjkaudata, Stations 10365, 10369, 10370, 10373, 10380, 10384. The southern species, T. gregaria, has not been identified from any of the hauls in 1916; hence if was certainly less common, proportionately to the other species of the genus, than we have usually found it even in the Gulf of ^Nlaine in summer. . ]Meganyctiphanes was not detected at all at the August stations south of New York, though it is possible that it may be represented among the very young euphausiids, nor was the boreal- arctic T. raschii. Although Thysanoessa was so widespread in the Calanus zone at this time, it was not an important factor in the plankton, being en- tirely overshadowed by the swarm of Calanus with which it was asso- ciated. Two oceanic euphausiids, Euphausia krohnii (Zimmer, 1909) and Neviatoscelis megalops, which usually occur, sometimes in swarms, in the mixed water along the Slope east of Cape Cod, formed a con- siderable part of the catches made in August at this relative position off Xew York (Station 10368) and off Chesapeake Bay (Station 10393, 500-0 meters) ; w-ith X'ematoscelis (but not Euphausia) off Delaware Bay (Station 10384, 500-0 meters). In their bath\-metric relationship to each other these catches reproduce our earlier experience, for at the only station where Euphausia was numerous (10368), it occurred in the surface haul, Xematoscelis in the haul from 450 meters just as was the case off Shelburne, Xova Scotia in 1914 (Bigelow, 1917a, p. 283). For the occurrence of Xematoscelis off Georges Bank, see page 138. Except for Euphausia as noted above, no euphausiids were detected in the surface tows in these waters. And so far as the data go, they suggest the same limitation of Thysanoessa to the cold water layers as obtained for Calanus (p. 143). In Xovember Thysanoessa, if not wholly lacking south of Cape Cod, was at least so rare there that none have been detected in the pre- liminary examination of the plankton. Small numbers of Euphausia and Xematoscelis were, however, taken at the outer edge of the Shelf (Station 10408, 150-0 meters) and of Euphausia at Station 10409, 135-0 meters. I may also mention swarms of small mysids, not yet identified, off the mouths of Delaware and Chesapeake Bays (Stations 10411, 10412, 10417), Xovember 16 and 17. 152 bulletin: museum of comparative zoology. Sagittae. Sagittae, as pointed out (p. 139), occurred in greater or less numbers associated with Calanus fininarchicus throughout the Calanus zone; though only locally (Stations 10362, 10373, 10382, 10386) did they form any considerable part of the catch south of New York in summer. Their relative importance in the plankton was considerably greater in November, especially at the outer edge of the Shelf (Stations 10409, 10410, 10414); at the mouth of Delaware Bay (Station 10411) and at Station 10413. But at none of these localities did they swarm as sometimes in the Gulf of Maine; the largest catch of Sagittae, even with the 1 meter horizontal net, being only about \ liter. During the summer of 1913 the prevalent chaetognath in this part of the coastal water was ° +.4° -3.6° 0 11.4 6.1 6.4 23.8 10.3 2.5 3.3 .01 ' No temperatures were recorded for January or February of those years. 2 Compiled from the U. S. Weather Review. 170 bulletin: museum of comparative zoology. Temperature (Fahrenheit) departure from normal Sno\vfall, inches 19131 Jan. Feb. Mar. Apr. Jan. Feb. Mar. Apr. New York , + 9.8° j + .2° Cape May ' + 9.1° ' +1.5' Norfolk +10.8° +1.2° +6.5° +4.2° +2.6° +2.9° + .5° + 1.4° 3 0 0 2.4 .8 1.2 1 0 0 trace 0 0 The annual ^'ariation in snowfall deserves emphasis, for not only was there about six times as much snow at New York in 1916 as in 1913, but the heaviest fall took place at just the season (March) when spring warming ordinarily commences in the waters along shore. It is during just such a spring as this, when the melting of the snow in the mountains back from the coast is not completed until late in the season, that the cold discharges from the Hudson and Connecticut Rivers, and from Delaware and Chesapeake Bays, may be expected to retard the normal spring warming of the waters off their mouths. And while they probably are not as important a factor in this respect as Tizzard (1907) supposed, temperatures taken by the U. S. Bureau of Fisheries in Chesapeake Bay during INIarch, 1916, show that they do so function to some extent, for the whole column of water, in its deeper parts, was still only 3°-4° on the 6th and 7th of the month. Consequently, whether or not the minimum winter temperature of the coastal water was lower in 1916 than usual, we have good reason to assume that spring warming was abnormally delayed in the water next the land b}' climatic conditions during that spring, quite apart from any possible influence by northern currents. Unfortunately we haxe no data on the rate with which spring warming, so rapid on the surface, normally penetrates downward into the deeper le^■els off this part of our coast with the advance of the season. This, since the downward propagation of solar heat by radiation is very slow (except in the uppermost layer) in sea water at rest (Krummel, 1907-1911; Murray & Hjort, 1912), by conduction negligible, depends chiefly on how actively the water is stirred by vertical currents, which in turn is go\'erned by tides, currents, and the degree of stability of the water. And the weakness of the former (the rise and fall of the tide being only 3-4 feet), suggests that in the coastal waters south of New York it is principally the latter which governs, 1 Compiled from the U. S. Weather Review. BIGELOW: COASTAL WATER EXPLORATION. 171 and which therefore deserves chief consideration here. Now the stabihty of any cohimn of sea water depends, of course, on the vertical distribution of density, and in winter, owing to uniformity of saUnity and temperature, the density of our boreal coastal waters is practically uniform from the surface downward, i.e., the water has no inherent stability, but is free to respond ^vith vertical movements to any local disturbance. But as the surface warms, and is freshened by water from the rivers in spring, it becomes lighter than the underlying layers, a change continuing until summer, when the water is so stable that vertical currents are hindered, if not prevented, and the deeper layers insulated from temperature influences from above. The effi- cacy of this process depends on the precise interrelation between tidal and other currents on the one hand, and the vernal warming and freshening to which the surface is subject on the other. Wherever the former are weak, the latter rapid and considerable, surface density decreases so rapidly in spring that the column of water soon becomes extremely stable. And this is especially true if vernal warming be sudden, for if more gradual, stability may at no time be suflficient to hinder vertical mixing seriously, even though a greater amount of heat be absorbed by the water during the season. Hence, para- doxical though it appear, the more rapidly the surface is warmed by the sun with the advance of spring, the less can this solar warming penetrate downward. This generalization is admirably illustrated in parts of the Gulf of Maine; even better in the southern part of the Gulf of St. Lawrence where though the surface warms to 18° by midsummer, the tempera- ture only some 50 meters down hardly rises at all (Dawson, 1913, Krummel, 1907-1911, Bjerken, 1919). The extreme vertical stability prevailing in summer in the coastal zone south of New York may similarly be expected to preserve the temperature of winter below 40-50 meters, until late in the season, unless warmed by an influx from the inner edge of the Gulf Stream, which never lies far outside the Continental Slope. And cAadence that this insulation of the deeper layers is actually efi'ective there is afl'orded by the fact that the temperature rises only very slowly at such a tri\aal depth as 30-40 meters, even as late as August, unless influenced by the Stream. For example there was no appreciable change in bottom temperature (8.3°-8.5° at 35-40 meters) from July 12 to August 1 in 1913, oft" New York (where the unchanged salinity showed that no important change had occurred in the composition of the water), though the surface was then very warm (19.4°-23.3°). And the 20-30 172 bulletin: museum of comparative zoology. meter temperature was correspondingly constant at the same general locality, in 1916, warming only by about 1° from August 1 (Station 10363) to August 26 (Station 10394), though the surface temperature was then at about its annual maximum. And if the course of events be disturbed by the Gulf Stream, the event is at once betrayed by a rise in salinity, as well as in temperature. Conseciuently, in any year when the Stream does not encroach on the Continental Shelf until August or later, the vertical distribution of summer temperature in the coast water would naturally be of just the type which prevailed in July and August, 1916, i.e., a very warm surface, sudden cooling down to 30-50 meters, with little if any further cooling from that level down to the bottom in 60-70 meters. Inasmuch as the temperature of the cold water on the Shelf south of New York was not so low, in the summer of 1916, as the probable minimum of the preceding winter (p. 169), the difference between that summer and 1913 is not too great to be accounted for as simply the result of a cold winter, with heavy snowfall and tardy spring, coupled with a failure of the off-shore water to encroach on the Shelf as it did in 1913. Further support of this contention is afforded by the comparative narrowness of the cold band from New York southward, i.e., its limitation to a strictly coastal zone, which did not even reach the outer edge of the Shelf on the bottom (p. 113). And while the possibility that northern water was partly responsible for its low temperature is open, the observed phenomena do not demand it. It is clear that the precise salinity of this part of the coastal zone depends on the relative amounts of off-shore water on the one hand, and of river water on the other, which annually, or periodically mix there. But the low salinity of the water along shore south of New York for 1916, as contrasted with 1913, does not necessarily mean that the influx of land water was greater during the one spring than during the other, for it is not unlikely that the salinities at any particular time are the result of conditions prevailing during the preceding year, or even earlier. In waters subject to such diverse influences, fluctua- tions in salinity are to be expected. iVnd until we have a more nearly continuous record of them it is idle to do more than record them, and to point out the major causes on which they depend. We have yet to consider whether the plankton catches substantiate the general conclusion, reached above, that both in the Gulf of Maine, and in the coastal water south of New York, the low temperatures of 1916 were mainly due to local causes, whereas the cold band off Georges Bank was a product of the Cabot Current. BIGELOW: COASTAL WATER EXPLORATION. 173 The total absence, in the Gulf, of the easily recognized Arctic faunal component which the Cabot Current carries with it, the presence or absence of which, in the eastern side of the Gulf and off Nova Scotia, coincides very closely with the expansion and contraction of the Cur- rent, does not suggest that there was more northern water in 1916 than usual. True, several of the more important Arctic "index" species, e.g., Mertensia, Limacina hclicina, and Oikoplcura vanhoffeni, are so delicate that their absence at any particular time does not necessarily have any bearing on conditions some months earlier. But this objection does not apply to the copepods Calanus hyper- boreus, and Metridia longa, so abundant further north (\^'illey, 1919). Nor were the microscopic members of the Gulf of Maine plankton any more Arctic in 1916 than usual. Though our previous studies of the zooplankton of the Gulf have not been sufficiently intensive to allow an exact seasonal comparison from year to year, they suggest that in 1916 the plankton retained up till midsummer features usually characteristic of May and Jime; just what might be expected if the low temperature of that year was simply due to the cold preceding winter and spring, the occurrence of the medusa Mitrocoma cruciata pointing, in particular, in this direction, for in 1915 it was comparatively abundant in May and June, disap- pearing in summer (Bigelow, 1917a, p. 305), whereas in 1916 it was taken in some numbers in July (Station 10340, p. 133). Furthermore, the microplankton of July, 1916 closely paralleled the June hauls of 1915 in the abundance of Ceratium longipes, scarcity of C. tripos, and total absence of C. arctica. The plankton of the cold water off Georges Bank contained no Arctic indicators (p. 137). But apart from a few tropical forms, evidently stragglers from the Gulf Stream, it closely paralleled, as did its temper- ature and salinity (p. 103, 104) catches made in the mixed water off Cape Sable in July, 1914, the latter being no more Arctic than the former at that season, though several conspicuous Arctic forms have been found there in June (Bigelow, 1917a, fig. 81). And the facts that the microplankton of this cold band was not Arctic, as it was not tropical; and that while the diatoms on Georges Bank were more northern in their affinities in 1916 {Thalassiothrix longissima) than in 1914 (Guinardia) they were combined with characteristic temperate species (Rhizosolenia sfyliformis, R. shnihesolei), are equally strong evidence that the cold band was not a direct, but an indirect, expan- sion of the Cabot Current. If my explanation of the low temperature south of New York as 174 bulletin: museum of comparative zoology. reminiscent of seasonal cooling during the preceding winter be correct, the plankton might also be expected to reproduce conditions prevailing some months earlier in warmer years. And that such was the case is indicated by the fact that Rathbun (1889) long ago recorded a typical boreal community, in fact much what we find in the Gulf of Maine, at many localities on the outer part of the Shelf between the latitudes of New York and of Chesapeake Bay, in April and May of 1887. For the details of the catches I refer to the original account, merely empha- sizing here the abundance of Calanus, Centropages, Temora; the frequent captures of Euthemisto, euphausiids, Limacina, and Clione limacma. Similarly boreal in composition are two plankton samples collected by Capt. John McFarland of the fishing schooner Victor, May 3 and May 9, 1913, at 38° 45' N. Lat.; 73° 52' W. Long., and 38 49' N. Lat., 73° 38' W. Long, respectively (z'.c, near the location of Station 10373). And Fowler's (1912) report of Calanus abounding along the New Jersey coast in June, 1911 and early July, 1912 shows that it is usually an important factor in the plankton community of this section of the coastal water until well into the summer. The fact that the Calanus swarm of 1916 was composed of large adults, hence was the result of a wave of reproduction which took place earlier in the season, likewise points to the conclusion that our xVugust catches in 1916 were just what might be expected in a tardy season; they do not suggest that the pelagic community was immigrant from the colder seas to the north. A possible exception to this generalization is the presence of Clione limacina off Chesapeake Ba^-, it being usually considered an Arctic form. But although Clione is widely distributed in New Foundland waters and in Arctic seas generally (Damas and Kofoid, 1907, Murray & Hjort, 1912), it is not a safe indicator of Arctic water, being at times abundant in Atlantic water, e.g., south of Ireland (Massy, 1909, Murray and Hjort, 1912). While essentially it is a northern form on our Atlantic coast, usually occurring but sparsely even in the Gulf of Maine, seldom further south, it has occasionally been encountered there in swarms (Wood, 1869) ; and it has already been recorded as far south as the coast of Virginia (Rathbun, 1889, Dall, 1889). In all proba- bility such local occurrences as that of 1916 are the result of temporary breeding activity, under rarely favorable conditions, of such few specimens as from time to time stray southward past Cape Cod, not of a direct immigration, via any Arctic Current. And support is lent to this conclusion by the fact that Clione larvae were taken at Station 10386, side by side with the adults. BIGELOW: COASTAL WATER EXPLORATION. 175 The neritic plankton along the coast affords some further evidence of a tardy season, the limitation of the swarms of Pleurobrachia and Mnemiopsis (p. 158) to small areas, and the juvenile state of the Beroe off Chesapeake Bay, being most readily explained on this basis. The scanty microplankton of the coastal water south of New York is easily reconciled with this, for it not only lacked Arctic indicators, but closely reproduced what we found here in 1913, if the tropical elements so conspicuous in that year be excluded. The limits of Gulf Stream water off our coasts are usually as clearly outlined by its inhabitants as by its temperature and salinity; some- times, indeed, more so, the former often surviving after its charac- teristic physical features have been obliterated by mixture with other waters. The scarcity of tropical plankton organisms in the summer of 1916 (p. 156), contrasted with their abundance in 1913, shows as clearly as do the low temperatures and salinities of that year, that the inner edge of the Stream lay some distance outside the Continental Slope. 176 bulletin: museum of compaeative zoology. TABLE OF STATIONS. Depth Depth of plankton Station Lat. N. Long. W. Date METERS hauls, METERS 10340 42° 32' 78° 38' July 19 50 0, 30-0, 45-0. 10341 42° 18' 70° 27' U 19 83 0, 30-0, 80-0. 83-0. 10342 42° 07' 70° 17' u 19 60 0, 55-0, 60-0. 10343 Provincetown Harb or u 20 30 0, 1.5-0, 30-0. 10344 42° 07' 69° 59' u 22 88 0, 40-0, 80-fl, 88-0. 10345 41° 52' 69° 40' u 22 150 0, "100-0, 150-0. 10346 41° 27' 69° 22' it 22 62 0, 30-0, 62-0 10347 4r 06' 68° 51' ii 23 78 0, 50-0, 60-0. 10348 40° 49' 68° 21' a 23 51 0, 30-0, 4.5-0, 50-0. 10349 40° 15' 68° 05' a 24 190 0, 130-0, 180-0. 10350 40° 23' 68° 15' u 24 118+ None. 10351 40° 06' 68° 57' u 24 182 0, 160-0. 10352 40° 00' 68° 44' u 24 1000+ 0, 500-0. 10353 40° 14' 69° 08' a 25 110 0. 10354 40° 26' 69° 24' a 25 76 0, 60-0, 70-0. 10355 40° 43' 69° 53' u 25 36 0, 16-0. 10356 40° 57' 70° 18' u 26 32 0, 10-0, 30-0. 10357 41° 11' 70^ 44' a 26 29 0, 20-0. 10359 41° IS' 71° 20' a 28 0, 5-0. 10361 41° 04' 73° 02' ii 29 27 0, 20-0. 10362 40° 22' 73° 38' Aug 1 25 0, 10-0. 2.5-0. 10363 40° 13' 73° 21' U 1 32 0, 20-0, 30-0, 32-0. 10364 40° 01' 72° 56' a 1 45 0, 40-0, 4.5-0. 10365 39° 41' 72° 39' ii o 61 0, 50-0, 60-0. 10366 39° 40' 72° 23' it 2 91 None. . 10367 . 39° 34' 72° 01' ii 2 239 None. 10368 39° 33' 71° 53' a 2 500+ 0, 450-0. 10369 39° 21' 72° 29' ii 3 137 0, 137-0. 10370 38° 55' 72° 54' a 3 lis 0, .50-0, 118-0. 10371 38° 56' 73° 06' a 3 84 0. 10372 38° 57' 73° 20' a 3 75 None. 10373 38° 57' 73° 35' a 4 62 0, 60-0. 10375 38° 59' 74° 08' a 4 41 0, 40-0. 10376 Cape Maj Harbor ii 7 10 0, 5-0. 10377 38° 54' 74° 44' a 10 16 0, 16-0. 10378 38° 48' 74° 53' a 11 14 0, 14-0. 10379 38° 46' 74° 35' a 11 27 0, 27-0. BIGELOW: COASTAL WATER EXPLOEATION. 177 Depth Depth of plankton Station Lat. N. Long. W. Date METERS HAULS, METERS 103S0 38° 27' 74" 25' Aug. 12 42 0, 3.5-0. 10381 38° 09' 74° 12' u 12 69 0, 25-0, 69-0. 10382 37° 55' 74° 05' (( 12 120 0, 50-0, 120-0. 10383 37° 52' 74° 04' (1 12 175 None. 10384 37° 46' 74^ (1 12 500+ 0, 250-0, 500-0. 10385 37° 28' 74° 25' a 13 140 None. 10386 37° 03' 74° 58' u 13 62 0 45-0. 10387 1 36' 52' 75° 48' u 21 13 0, 12-0. 10388 36° 51' 75° 38' a 21 None. 10389 36° 50' 75° 27' a 21 19 0, 18-0. 10390 36° 49' 75° 14' a 21 26 0, 26-0. 10391 36° 48' 74° 55' a oo 40 0, 40-0. 10392 36° 47' 74° 37' u 22 82 0, 75-0. 10393 36° 46' 74° 29' ti 22 500+ 0, 250-0, 500-0. 10394 40° 15' 73° 08' ii 26 42 0, 42-0. 10395 40° 32' 72° 44' u 26 38 0, 38-0. 10396 40° 50' 72° 07' It 26 29 0, 29-0. 10398 42° 10' 70° 09' It 29 45 0, 35-0 10399 42° 30' 70° 21' Oct. 31 137 0, 60-0. 10400 42° 58' 70° 14' Nov . 1 142 0, 90-0. 10401 42° 37' 69° 46' a 1 219 0, 90-0, 180-0. 10402 43° 37' 69° 15' a 2 144 0, 50-0. 10403 42° 16' 70° 12' a 8 32 0, 25-0. 10404 41° 53' 69° 37' u 8 183 0, 1.50-0. 10405 41° 17' 71° 03' a 10 36 0, 25-0. 10406 40° 37' 71° 19' a 11 69 0, 50-0. 10407 40° 03' 71° 43' a 11 93 0, 7.5-0. 10408 39° 52' 71° 47' a 11 183 0, 75-0, 150-0. • 10409 39° 10' 72° 41' u 12 140 0, 80-0, 13.5-0. 10410 39° 10' 73° 01' a 12 76 0, 76-0. 10411 OilD elaware Capes a 16 16 0, 1.3-0. 10412 38° 14' 74° 39' a 10 34 0, 34-0. 10413 37° 44' 74° 35' a 17 56 0, 56-0. 10414 37° 15' 74° 30' a 17 146 0, 140-0. 10415 37° 12' 74° 42' u 17 91 0, 91-0. 10416 37° 06' 75° a 17 41 0, 41-0. 10417 36° 59' 75° 41' u 17 0. ^ Approximate position. 178 bulletin: museum of comparative zoology. TABLE OF TEMPERATURES. SALINITIES AND DENSITIES. The salinities for the July-August Cruise were determined during the following autumn, and are, I believe, reliable within the usual limits of the titration process. But owing to the interruption of all oceanographic work by the war, the water samples from the Novem- ber Cruise were not titrated until the summer of 1919. For this reason, and also because I believe, from experiment, that the standard water with which the titrations were performed had undergone some slight change due to erosion of the glass tubes in which it was con- tained, the salinities for that Cruise can not be depended upon except within wide limits. They, and the resulting densities are therefore useful only in a general, and comparative way. Temperatures (centigrade) ; salinities (%o) and densities (at the temperature in situ). The densities are corrected for pressure, by Ekman's (1910) table 4 alone, this being sufficiently accurate for the small depths involved. •July-August Cruise. Depth Temperature Density Station METERS CENTIGRADE Salinity in silu 10340 0 11.95° 31 . IS 23.66 25 6.49° 31.87 25.17 50 5.19° 32.00 25.54 10341 0 16.39° 30.48 22.22 25 5.08° 32.03 25.46 50 3.9° 32.2 25.83 SO 3.67° 10342 0 17.22° 30.61 22.13 30 - -oO t . (6 31. 5S 24.79 60 6.14° 31.87 25.39 10344 0 15.83° 30.75 22.55 25 4.91° 32.10 25.. 53 50 4.07° 32.20 25.81 80 4.19 32.38 26.08 10345 0 10.° 31.53 24.27 50 4.17° 32.25 25.84 100 3.85° 32.66 26.45 150 4.06° 32.86 26.81 10346 0 7.22° 32.03 25.07 30 6.41° 32.07 25.35 60 4.47° 32.38 25.96 BIGELOW: COASTAL WATER EXPLORATION. 179 Depth Temperature Density Station METERS CENTIGRADE Salinity in situ 10347 0 30 11.39= 10.91= 32.54 24.81 60 9.61° 32.14 25.31 10348 0 25 11.67° 11.34= 32.54 24.75 50 11.26° 32.57 25.10 10349 0 17.5° 32.47 23.49 30 10.36 33.86 Y SO 7.16° 32.47 J 130 6.75° 34.42 27.64 ISO 6.72 34.83 28.23 10351 0 30 15.56° 10.66° 32.47 23.93 80 4.82° 33.42 26.85 130 5.88° 34.20 27.56 ISO 7.13° 34.72 28.06 10352 0 16.95° 32.47 23.61 50 4.85° 33.08 26.42 100 7.65° 34.36 27.32 200 7.65° 34.92 28.25 300 5 . 75° 34.87 28.92 400 5.15° 34.87 29.48 500 4.1° 34.96 30.14 10353 0 15.° 10354 0 13.61° 32.27 24.18 30 8.71° 32.63 25.48 70 6.07° 32.86 26.21 10355 0 11.95° 31.73 24.08 30 10.97° 32.14 24.71 10356 0 16.11° 31.78 23.29 30 12.14° 32.14 24.50 10357 0 17.78= 30.90 22.19 25 14.28= 31.58 23.62 10359 0 16.67° 31.09 22.58 10361 0 20= 26.24 18.14 10362 0 21.1° 30.57 21.13 22 11.9° 31.73 24.19 10363 0 20.5° 31.17 21.74 30 8 . 15° 32.41 25.38 These two water samples were probably transposed. 180 bulletin: museum of comparative zoology. Depth Temperature Density Station METERS CENTIGRADE Salinity in situ 10364 0 21.8° 30.73 21.06 40 5.54° 32.72 26.02 10365 0 20.3° 31.26 21.86 20 13.91° 32.23 24.18 40 4.94° 32.72 26.09 60 4.82 32.81 26.27 10366 0 19.4° 32.61 23.12 30 15.58° 32.68 24.22 60 5.89° 32.84 26.18 90 5.8° 33.30 26.67 10367 0 19.72 50 8.08° 33.68 26.48 100 10.92° 34.92 27.23 150 10.65 35.25 27.77 200 7.89° 35.07 28.32 10368 0 20° 32.48 22.87 100 11.21° 35.10 27.32 200 8.58° 35.07 28.21 400 5.18° 34.99 29.58 j:oo 4.68° 34.92 20.04 10369 0 90 21.7' 7.99° 10370 0 21.7° 30.99 21.27 25 8.94° 32.66 25.42 50 4.98° 32.86 26.24 100 9.47° 34.54 27.18 10371 0 22.2° 10373 0 21.1° 31.09 21.46 60 4.52° 32.63 26.17 10375 0 21.9° 30.62 20.96 40 6.18° 32.39 24.90 10376 0 24.4° 31. 20.52 10377 0 21.9° 30.86 21.13 15 15.77° 31.29 23.03 10378 0 21.1° 14 21.04° 30.93 21.48 10379 0 23.6° 30.9 20.63 25 10.88° 32. 24.59 10380 0 23.6° 30.57 20.44 20 6.73° 32.38 25.51 40 6.78° 32.41 25.64 BIGELOW: COASTAL WATER EXPLORATION. 181 Depth Temperature Density Station METERS CENTIGRADE Salinity in silu 10381 0 23.6° 31.11 20.85 20 17° 32 .00 23.32 30 4.12° 40 4.02° 32.61 26.10 65 4.06° 32.61 26.21 10382 0 24.4° 31.11 20.62 40 7.02° 32.61 25.75 80 4.71° 32.72 26.31 120 6.75° 10383 0 24.7° 31.09 20.51 25 13.45° 32.25 24.30 50 4.88° 100 5.46° 34.20 27.48 150 11.32° 35.23 27.63 175 9.76 35.19 27.99 10384 0 24.7° 31.46 20 77 50 9.52° 33.39 26.53 100 10.29° 34.81 27.27 150 11.01° 35.03 27.54 250 9.29° 35 . 10 28.36 350 6.22° 35.16 29.34 500 4.73° 34.97 30.10 10385 0 23.4° 10387 0 23.3° 23.73 15.38 12 23 . 3° 31.08 20.95 10389 0 23.9° 30.75 20.49 10 23.31° 30.79 ■ 20.72 18 22.26° 31.02 21.23 10390 0 23.9° 30.75 20.49 12 23.41° 30.75 20.67 25 16.54° 31.92 23.40 10391 0 22.8° 32.18 21.88 20 oo ■[° 32.14 22.12 40 10.84° 32.50 25.08 10392 0 •->■> v)° 32.50 22.29 20 21.9° 32 . 59 22.54 40 6.6° 32.72 25.89 60 5.85° 32.77 26.12 80 5.08° 32.77 26.30 10393 0 22.2° 32.56 22.33 50 11.86° 33.10 25 39 182 bulletin: museum of comparative zoology. Depth Temperature Density Station METERS CENTIGRADE Salinity in situ 10393 100 8.73° 34.11 26.97 200 11.23 35.21 27.88 300 6.26° 35.03 28.99 400 6.79° 34.92 29.32 500 5.2° 34.97 30.03 10394 0 20.8° 31.15 21 64 20 12.25° 32.25 24.53 40 7.35° 32.41 25.54 10395 0 21.1° 31.11 21.54 17 17.72° 31.29 22.62 35 9.02° 32.21 25.13 10396 0 19.2° 29.65 28 12.13° 31 98 10398 0 16.95° 31.27 22.70 20 7.89° 31.89 24.97 43 4.91° 32.05 25.57 October-November Cruise. Depth Temperature Density Station meters CENTIGRADE Salinity in silu 10399 0 10° 31.71 ■24 Al ■ 30 9.18° 31.91 24.86 60 6.43" 32.41 25.76 90 5.43° 32.56 26.14 120 5.23° 32.59 26.33 10400 0 9.72° 32.03 24.71 30 8.21° 32.09 25.12 60 7.07"^ 32.45 25.61 90 4.84° 32.57 26.23 130 4.41= 32.81 26.65 10401 0 10.56° 31.98 24.54 50 . S.90° 32.30 25.28 100 4.24' 32.65 26.40 150 4.53° 32.99 26.89 200 4.99° 33.53 27.53 10402 0 8 . 33° 32.36 25.18 25 8.89° 32.34 25.18 55 8.19° 32.56 25 . 62 BIGELOW: COASTAL "WATER EXPLORATION. 183 Depth Temperature Density St.^tion METERS CENTIGR.'lDE S.u:,rNiTY in situ 10402 85 6.59° 32.78 26.15 135 4.97° 32.90 26.70 10403 0 9.17° 31.87 24.66 30 8.39° 32.07 25.08 10404 0 10.28° 32.01 24.60 50 8.04° 32.18 25.32 100 4.85° 32.88 26.52 175 4.78° 33.15 27.09 10405 0 11.95^ 32.05 24.31 30 12.52° 32.06 24.36 10406 0 11.67° 32.23 24.52 30 ll.?5° 32.36 24.72 60 9.98° 32.54 25.35 10407 0 11.28° 32.54 24.83 30 11.85° 32.56 24.88 60 13.08° 33.15 25.24 90 7.72° 32.88 10408 0 11.39° 32.59 24.86 25 12.06° 32.63 24.83 50 14.0° 33.71 25.46 100 9.26° 34.01 26.80 180 10.26° 35.00 27.79 10409 0 12.78 30 13.22 32.83 24.90 60 13.34 33 62 25.58 90 9.07 33.68 26.35 135 10.28 34.78 27.39 10410 0 13.05 32.59 24.56 25 13.62 32.83 24.74 50 13.82 33.10 25.02 75 11.18 33.12 25.65 10411 0 12.22 31.50 23.87 15 12.49 31.51 23.89 10412 0 13.33 32.58 24.48 30 13.66 32.58 24.58 10413 0 13.33 32.86 24.62 25 13.55 32.90 24.74 50 13.56 32.90 24.86 10414 0 13.05 33.12 24.96 25 13.42 33.10 24.96 55 13.11 33.17 25 . 24 184 bulletin: museum of comparative zoology. Depth Temperature Density Station METERS centigrade Salinity in situ 10414 85 9.34 32.86 25.80 135 9.49 33.73 26.70 10415 0 12 . 78 . 32.85 24.80 30 13.35 32.91 24.86 60 13.36 32.99 25.08 90 11.91 33.18 25.65 10416 0 13.78 32.92 24.64 20 13.14 32.94 40 14.20 10417 0 14.17 32.52 TEMPERATURE AND SALINITY IN CHESAPEAKE BAY, MARCH, 1916. Depth Station Date L.\titude Longitude METERS Temperature Salinity 8460 March 6 36° 57' N 76° W 0 s 15 3.7° 3.8° 3.5" 3.6° 28.15 29.72 30.53 30.55 8464 ]March 7 37° 15' N 76° 5' W 0 4 9 18 20 27 30 40 3.2° 3.3° 3.3° 3.2° 3.3° 3.3° 3.3° 3.4° BIGELOW: COASTAL WATER EXPLORATION. 185 BIBLIOGRAPHY. BiGELOW, H. B. 1909. Reports on the scientific results of the expedition to the Eastern Tropical Pacific, 1904-1905. XVI. The Medusae. Mem. M. C. Z., 37, 243 pp., 48 pis. 1914a. Explorations in the Gulf of Maine, July and August, 1912, by the U. S. fisheries schooner Grampus. Oceanography and notes on the plankton. Bull. M. C. Z., 58, p. 29-148, 9 pis. 1914b. Oceanography and plankton of Massachusetts Bay and adjacent waters, November, 1912-May, 1913. Bull. M. C. Z., 58, p. 383-420, Ipl. 1915. Exploration of the coast water between Nova Scotia and Chesa- peake Bay, July and August, 1913, by the U. S. fisheries schooner Grampus. Oceanography and plankton. Bull. M. C. Z., 59, p. 149- 360, 2 pis. 1917a. Explorations of the coast water between Cape Cod and Halifax in 1914 and 1915, by the U. S. fisheries schooner Grampus. Ocean- ography and plankton. Bull. M. C. Z., 61, p. 161-358, 2 pis. 1917b. Explorations of the United States coast and geodetic survey steamer Bache in the Western Atlantic, January-March, 1914, Oceanography. App. 5, Rept. U. S. comm. fisher, for 1915, 62 pp., 1 chart. Bjerkan, Paul. 1919. Results of the hydrographical observations made by Dr. Johan Hjort in the Canadian Atlantic waters during the year 1915. Canad. fisher, exped., 1914-1915. Dept. naval service, p. 347-404, 3 pis. Conant, F. S. 1895. Description of two new chaetognaths. Johns Hopkins univ. circ, 14, p. 77-78, [pi. 1], fig. 6, 7. CONSEIL permanent INTERNATIONAL POUR l'eXPLORATION DE LA MER. 1919. Variations de la temperature de I'eau de surface dans certains carres choisis de I'Atlantique pendant les annees 1900-1913. Bull, hydrogr., 37 pp. 7 pis. Craigie, E. H. 1916a. Hydrographic investigations in the St. Croix River and Passama- quoddy Bay in 1914. Suppl. 5th. Ann. rept. Dept. naval service [Canada], Fisher, branch, p. 151-161. 1916b. A hydrographic section of the Bay of Fundy in 1914. Suppl. 5th. Ann. rept. Dept. naval service [Canada], Fisher, branch, p. 163-167. Craigie, E. H., and Chase, W. H. 1918. Further hydrographic investigations in the Bay of Fundy. Suppl, 7th. Ann. rept. Dept. naval service [Canada], Fisher, branch, p. 127-148. 186 bulletin: musei-^i of comparative zoology. Dall, W. H. 1889. A preliminary catalogue of the shell-bearing marine mollusks and brachiopods of the southeastern coast of the United States. Bull. 3 U. S. nat. mus., 221 pp., 74 pis. Damas, D. et Koefoid, E. 1907. Le plankton de la mer du Gronland. Due d'Orleans. Croisiere oceanog. mer Gronland, p. 345-454. Dawson, W. B. 1913. The currents in the Gulf of St. La^Tence. Dept. naval service [Canada], 461 pp., 1 chart. Dickson, H. X. 1901. The circulation of the surface waters of the North Atlantic Ocean. Philos. trans. Roy. soc, ser. A, 196, p. 61-203, pi. 1-4. Ekman, V. W. 1910. Tables for sea-water under pressure. Publ. circ. 49, Cons, inter, expl. de la mer, no. 49, 48 pp. Fowler, H. W. 1912. The Crustacea of New Jersey. Ann. rept. Xew Jersey state mus. for 1911, p. 29-650, 150 pis. Gran, H. H. 1908. Diatomeen. Xordisches plankton. Botan. theil, 146 pp. Lief. 3, 19. Huntsman, A. G. 1919. Some quantitative and qualitative plankton studies of the eastern Canadian plankton. Canad. fisheries exped., 1914-1915, Dept. naval service, p. 405-486. Jee, E. C. 1919. Review of the physical and chemical properties of the surface waters, and the variations of these properties during the period 1902- 17, inclusive, in the area of the Atlantic Ocean centred on 50° N.: 20° W. Fisher, invest. Bd. agric. and fisher. Great Britain, ser. 3, Hydrogr. 3, Atlantic Ocean, pt. 1, 23 pp. Knudsen, Martin. 1901. Hydrographical tables. Copenhagen, 63 pp. 1909. Resume de I'hydrographie des mers explorees par le conseil. bull, trimestr. Cons, inter, expl. de la mer, 1906-1907. SuppL, 78 pp., 23 pis. Kramp, p. L. 1918. Chaetognatha collected by the Tj.^lfe expedition to the west coast of Greenland in 1908 and 1909. Vid. medd. Dansk. naturh. foren. 69, p. 17-55. Kriimmel, Otto. 1907-1911. Handbuch der ozeanographie. 2 vols. Leipzig and Stuttgart. Kuntz, Albert and Radcliffe, Lewis. 1917. Notes on the embryology and larval development of twelve teleostean fishes. Bull. L". S. bur. fisher., 35, p. 87-134. BIGELOW: COASTAL WATER EXPLORATION. 187 LiBBEY, William and others. 1891. Report upon a physical investigation of the waters off the south- ern coast of New England, made during the summer of 1889 by the U. S. fish commission schooner Grampus. Bull. U.S. fish comm. for 1889, 9, p. 391-4.59, pi. 124-158. 1895. The relations of the Gulf Stream and the Labrador Current. Rept. Sixth int. geogr. congress, p. 461—474, 1 chart. Massy, A. L. 1909. The Pteropoda and Heteropoda of the coasts of Ireland. Fish- eries. Ireland Sci. invest., 1907, no. 2, 52 pp., 1 pi. Matthews, D. J. 1907. The surface waters of the North Atlantic Ocean south of 60° N. latitude, September, 1904-December, 1905. North Sea fisher, invest, comm., 2nd [British] rept. (southern area), 1904-1905, p. 269-348, 3 pis. Mayer, A. G. 1912. Ctenophores of the Atlantic coast of North America. Carnegie inst., Publ. no. 162, 58 pp., 17 pis. MiCH.^L, E. L. 1911. Classification and vertical distribution of the Chaetognatha of the San Diego region. Univ. Cal. publ. ZooL, 8, p. 21-186, pi. 1-8. 1919. Report on the Chaetognatha collected by the United States fisheries steamer Albatross during the Philippine expedition, 1907- 1910. Bull. 100 U. S. nat. mus., 1, pt. 4, p. 235-278, pi. 34-38. MuRR.\Y, John and Hjort, Joh.an'. 1912. The depths of the ocean. London, 20, 821 pp., ills. OSTEXFELD, C. H. 1913. Noctiluca and Globigerina. Bull, trimestr. Cons, inter, expl. de la mer. Resume observ. plankton, 3, p. 291-297. Paulsen, Ove. 1908. Peridiniales. Nordisches plankton. Botan. theil, 124 pp. Lief. 8, 18. Rathbun, Richard. 1887. Ocean temperatures of the eastern coast of the United States, from observations made at light-houses and twenty-four light-ships. The fisheries and fisher}' industries of the V. S., sect. 3, p. 157-177, 32 charts. 1889. Notice of the small surface organisms taken in the tow-nets, and of the contents of IVIackerel stomachs. Bull. LT. S. Fish comm. for 1887, 7, p. 259-267. Ritter-Zahony, Rudolf von. 1911. Revision der chaetognathen. Deutsche Siidpolar. exped., 13, Zool., 5, heft 1, p. 1-72. Sandstrom, W. J. 1919. The hydrodynamics of Canadian Atlantic waters. Canad. fisher, exped. Dept. naval service, p. 4, 221-343, 15 pis. 188 bulletin: museum of comparative zoology. TiZARD, T. H. 1907. Dr. Otto Pettersson on the influence of ice-melting on oceanic circulation. Geogr. journ., 30, p. 339-344. Vachon, Alexander. 1918. Hydrography in Passamaquoddy Bay and vicinity. Suppl. 7th Ann. rept. Dept. naval service [Canada], New Brunswick. Fisher, branch, p. 295-328. Wheeler, W. M. 1901. The free swimming copepods of the Woods Hole region. Bull. U. S. fish comm. for 1899, 19, p. 157-192. WiLLEY, Arthur. 1919. Report on the Copepoda obtained in the Gulf of St. Lawrence and adjacent waters, 1915. Canad. fisheries exped., 1914-1915, Dept. naval service, p. 173-220. Wood, W. 1869. The Clio borealis on the coast of Maine. Proc. Portland soc. nat. hist., 1, p. 185-188. Wright, R. Ramsay. 1907. The plankton of eastern Nova Scotia waters. [Suppl.] 39th ann. rept. marine and fisheries Canada, p. 1-19, 7 pis. ZiMMER, Carl. 1909. Die nordischen schizopoden. Nordisches plankton. Lief. 12, 6, 178 pp. Bulletin of the Museum of Comparative Zoology ATHARVARDCOLLEGE. Vol. LXV. No. 6. BIRDS FROM DARIEN. By Outram Bangs and Thomas Barbour. CAMBRIDGE, MASS., U. S. A.: PRINTED FOR THE MUSEUM. September, 1922. No. 6. — Birds from Daricii. By OuTRAM Bangs and Thomas Barbour. The latter author, with Mr. W. S. Brooks, spent several months this spring (1922) in Panama. Part of the time was occupied in a journey to the Sapo Highlands and to the heavy forests of the Sambu \'alley, previously unAisited by a zoologist and one of the least known areas of all tropical America. The original plan was to collect mammals and reptiles but having the opportunity to engage the services of IMr. C. F. Underwood while in Costa Rica, it was decided to attempt a collection of birds. Un- favorable weather conditions and the illness of Mr. Underwood made it undesirable to prolong this work. Nevertheless it is only proper to record that, working under conditions very different from the luxurious bird collecting of Costa Rica, with a veteran corps of trained assistants, and camping here, under the most difficult conditions, Mr. Under- wood preserved every single birfl worth saving. We worked under high pressure and had some good Chocoano Indian shooters, but the num- ber of days devoted to ornithology was only about seventeen. Since nearly seven hundred birds were collected it will readily be understood that some of the days were long and well occupied. In one day eighty-seven birds were shot and skinned. One hundred and fifty- nine species are represented in the collection. It would have been easy to increase this list very considerably. The clearings about the village of Garachine and about the Indian plantations on the Sambii swarmed with many species of birds characteristic of the open country. Many forms were so familiar to us all and so well represented in col- lections that our stays about the clearings were devoted to other acti\ities. The collection really represents an attempt to obtain only those species unrepresented in the M. C. Z., so far as our knowledge went. Inevitably, however, many common species were preserved. Field notes, in many cases of no great value, are appended. The only excuse for so doing is the paucity of data concerning the region and the fact that so many collections of birds are reported upon by persons who have not had the good fortune to share in securing the material. Collections of mammals, reptiles, amphibians, and fishes were also obtained and reports upon these series, together with a general ac- 192 bulletin: museum of comparative zoology. count of the country and its inhabitants, have been or will be published elsewhere. Several occasions stand out as specially memorable. The chance to see two great flights of hawks (probably Swainson's) was very pleasing. One afternoon (16 April) coming down the Sambu in dugouts we noticed what we first believed to be an enormous gathering of turkey buzzards but in a short time the whirling or wheeling cloud passed westward near enough for us to see that what we had mistaken for buzzards was in reality an enormous crowd of hawks. They were very high, the day happened to be finely clear, and several thousand were visible. A few days later another migration was observed from a foothill of Mt. Sapo but the number of hawks was distinctly less. On another afternoon, when resting after a steep climb, at the foot of a gigantic espave tree, we heard a noise overhead and looking up were surprised to see the great crested head of a Harpy Eagle looking down at us. Hastily pushing in 2's I (Barbour) fired both barrels but the bird was very high and only a wing was broken. It had just sufficient power of flight to launch out from the tree top and in a second it was away, down the slope of an almost perpendicular valley which we searched for hours without avail. Although we had several fine views of flying Harpys we never got another shot, and losing this bird was a bitter disappointment. In this paper we follow the order used by Ridgway in the Birds of North and Middle America (Bull. 50, U. S. N. M.). It is also to be noted that whenever Chapman is referred to, with no specific reference, we allude to The Distribution of Bird-Life in Colombia (Bull. Amer. mus. nat. hist., 1917, 36), and similarly with the use of Hellmayr's name, to A Contribution to the Ornithology of Western Colombia (P. Z. S. London, 1911). In conclusion it gives us great pleasure to thank President Porras and the Sen ores Alfaro and Morales of his Cabinet in Panama; Dr. R. P. Strong and Major Bocock, Superintendent of the Hospital Santo Tomas and to many other officials and friends both here and in Central America who gave us help and adv^e. Messrs. Nelson, Chapman, Todd, and Penard have also aided us by loaning specimens which have been invaluable, in connection with our material in Cambridge, in arriving at the taxonomic conclusions which we have reached. Finally it may be well to add a word regarding the location of the collecting stations named. Esnape was a camp on a stream of that name situated several hours march northeast of the junction of the BANGS AND BARBOUR: BIRDS FROM DARIEN. 193 Sambu and Sabalo Rivers, Lat. 8° 02' 13" and Long. 78° 12' 00". The Quebrada Esnape is said to be tributary to the Rio Taimiti. Jesusito refers to camps at several points in the lowland forest on the Rio Jesusito which rises in the Sapo Mountains. It is one of a series of small streams which have their source in the highlands and flow toward the Sambu Valley, but which apparently, in this case, is not tributary to the Sambu but probably empties in a marsh drained by the Rio Celorio. Mt. Sapo refers to stations on the headwaters of the Rio San Antonio, a stream which enters the sea near the town of Garachine. The river drains the western slopes of Mt. Sapo itself. The collections are all representative of the fauna of the lower Tropical Zone, the stations being all in the rain forest below 1500'. ARDEIDAE. 1. Canchroma zeledoni Ridgway. One skeleton, Jesusito. Only a single Boat-billed Heron was seen during the entire trip and this was shot while night hunting with an acetylene lamp. It stared stupidly at the light and made no effort to escape. 2. Pilherodias pileatus (Boddaert). One adult female, Jesusito, 6 April, 1922. This appears to be the first record for Panama. It has already been recorded from western Colombia by Chapman. A pair of these lovely night herons was observed several times near the lower camp on the Jesusito River. They were rather shy and ap- peared to be fairly active by day. Only one of the pair was killed. CATHARTIDAE. 3. Sarcoril\mphus papa (Linne). One adult male, Pacora, April, 1922. The chief of the Canal Zone Fire Department, Captain Brown, found this splendid vulture sitting on an open plain while hunting near Pacora. He approached the bird and found it to be perfectly blind, although its eyes gave no appearance of being abnormal. He kept it alive for some time and finally gave it to the Museum party. King vultures were seen several times flying low over the tree tops in the Sambu Valley but none were shot. 194 bulletin: museltm of comparative zoology. AQUILIDAE. 4. Ibycter americanus americanus (Boddaert). One adult female, Jesiisito, 13 April, 1922. This skin affords a wing-length of 355 mm., which places it as the small form representative of tropical South America, and which Swann has already recorded as being found northward to Panama. This remarkable hawk was a constant source of surprise. Small parties of four or five were often encountered in the deep woods. They flapped sluggishly for short distances when disturbed uttering the most unearthly shrieks. They were feeding in the fruit trees and reminded one far more of macaws, in voice and actions, than of hawks. The Panamanians call them \^-itches (brujos) and do not consider them at all as being in the catagory of eagles or other birds of prey. About the ]Mt. Sapo camp small flocks of brujos appeared every morning and evening flying up and down the steep hillsides just above the forest, screaming like macaws all the while. 5. Rupornis magnirostris ruficauda (Sclater and Salvin). One adult male, Jesusito, 7 April, 1922. 6. Falco albigularis Daudin. Two adults, male and female, Jesusito, April, 1922. Swann, reluctantly to be sure, recognizes three races of the White- throated Bat Falcon. We, however, detect no geographical variation in the species and cannot allow Chubb's two forms. Occasionally, just at dusk, one of these hawks would dart past the lower Jesusito camp which was on a rough pebbly beach across the stream from an old abandoned Indian banana plantation. The pair secured were shot as thev flashed bv, to the intense excitement of our Chocoano companions. TINAMIDAE. 7. Tinamus major castaneiceps Salvadori. Two adult females, Jesusito, April, 1922. Many of the large tinamous were killed for food and were delicious. The Indians had small difficulty in securing specimens but neverthe- BANGS AND BARBOUR: BIRDS FROM DARIEN. 195 less they were very shy, having been persistently hunted. At dawn and eve their protracted and musical cry, oft repeated in crescendo, was one of the forest sounds which serve the Indians as most exact time-markers. 8. Crypturus soli panamensis Carriker. One adult female, Jesusito, 9 April, 1922. The little tinamou was far rarer and far less shy than its larger ally. The Indians brought in a number, shot at such short range that they were impossible to preserve. CRACIDAE. 9. Crax globicera Linne. One adult female, Jesusito, 13 April, 1922. Miller and Griscom (American museum novitates), no. 25, p. 7, 1921, discredit Crax panamensis Ogilvie Grant, saying that in series all characters of that supposed species break down, and that it is not separable from C. globicera of Mexico to Honduras. The pavones were rare, very rare, and it was only by making long excursions from the upper Jesusito camp far into the Sapo Hills that an occasional pavon could be secured for food. 10. Penelope cristata (Linne). One adult, male, Jesusito, 15 April, 1922. The same remarks apply to this species. The Indians prefer the flesh of these birds to any other and have hunted them persistently for generations. In the region of the Rio Pavarondo near the Colombian frontier, where there are fewer Chocoanos, all game is said to be much more common. ODONTOPHORIDAE. 11. Odontophorus guianensis panamensis Chapman. Four specimens, adults both sexes, Mt. Sapo and Jesusito, April, 1922. The Panamanians call this bird the " Mulatto Dog" (Perro Mulato), 196 bulletin: museum of comparative zoology. though the connection is not easy to see. About equally abundant with the following species, they were singularly tame and were often seen, in pairs, walking about camp in the most unconcerned fashion. 12. Rhynchortyx clntctus cinctus (Salvin). Seven specimens, adults of both sexes, Mt. Sapo, April, 1922. Two downy young were also preserved. When camp was quiet, as it sometimes was, these little quail often appeared, walking in pairs, and when disturbed ran away to some thicket. Neither this species, nor the preceding, were ever seen to take flight. RALLIDAE. 13. Ar amides cayanea cajanea (P. L. S. Miiller). One adult female, Jesusito, 12 April, 1922. We are unable, with a very large series of specimens, to verify a single character of the so-called Araviides cajanea salmoni Chubb, of northern Colombia and Panama, and cannot satisfactorily separate birds from that general region from those from Guiana. A recogniz- able, insular form, however, A. c. latens Bangs and Penard, is found on Pearl Island, in the Bay of Panama. Along the Sambii wood rails were often heard calling at dusk and during the early night. Away from the main stream of the river they were far more rare, indeed were but very seldom heard and the one female, killed by an Indian, was the only one which was brought to camp. COLUMBIDAE. 14. Leptoptila cassini cassini Lawrence. Six adults, both sexes, Mt. Sapo, Jesusito, and Rio Esnape, April, 1922. Pigeons of all sorts were scarce in the high forest although passing up and down the Sambu, especially through the mangroves, numbers of individuals of certainly three or more species were very conspicuous. BANGS AND BARBOUR: BIRDS FROM DARIEN. 197 PSITTACIDAE. 15. EuciNETUS HAEMATOTis cocciNEicoLLARis (Lawrence). One adult, female, Jesusito, 9 April, 1922. The little red collared parrots were not very abundant. However, in common with the other species, they fed in such gigantic trees that they were not easily observed and identified. 16. PiONUS MENSTRUUS (Linne). One adult (sex not determined), Jesusito, April, 1922. Flocks of this species passed every day over the camp on the lower Jesusito at night-fall and about dawn. They whirred overhead, band after band, until the whole forest resounded with their screams and ordinary conversation was next to impossible. ^Vhence they came or where they went we never knew, and these blue parrots seldom stopped to rest where we could observe them. 17. Amazona farinosa inornata (Salvadori). One adult, female, Jesusito, 10 April, 1922. In contrast to the preceding species this great plain looking green parrot roosted in myriads near the lower Jesusito camp. They stayed about for some time in the morning to feed, shrieking and chattering the while, but almost always far out of range of our twelve bore, so high were the high tree-tops. About eight every morning they flew toward the hills, returning late in the afternoon, occasionally accompanied by a few pairs of macaws none of which were ever shot. CUCULIDAE. 18. PlAYA CAYANA THERMOPHILA Sclatcr. One adult female, Mt. Sapo, 24 April, 1922. This specimen is not chfferent from the average of skins from Costa Rica and the Canal Zone of Panama, and does not seem in any way to approach P. c. nigricrissa (Cabanis) of western Ecuador and \yest Colombia. This cuccoo is rare in the deep forest. The only one shot chanced to light in some creepers at noon one day while watch was being kept under the "Cotinga Tree" to be mentioned later. 198 bulletin: museum of comparative zoology. MOMOTIDAE. 19. MoMOTUs subrufescens reconditus Nelson. Three adult males, Rio Esnape, April, 1922. The first of these to be secured was caught in a steel trap set at the mouth of its enormous nesting burrow. 20. Urospatha martii semirufa (Selater). Three adults, male and two females, ISIt. Sapo, April, 1922. This and the preceding species of motmots were common and were heard daily calling during the hot still noon hours. Much larger series could easily- have been secured. 21. Electron platyrhynchus minor (Hartert). One adult, female, Jesusito, 9 April^l922. This specimen, with a culmen of 39 mm. differs in no way from (eight examples) birds from Costa Rica, and western Panama. Chap- man has already pointed out that E. p. suholcs Nelson, was either based on a slightly abnormal individual, or has a very restricted dis- tribution which seems impossible as it is a species of the Tropical Zone. This curious, broad-billed motmot was very rare and no other in- dividual was secured or seen. 22. Hylomanes momotula obscurus Nelson. One adult male, Mt. Sapo, 24 April, 1922. An aged negro, armed with an ancient French muzzle loader, brought this motmot to camp one day. He said that he had killed it high on the slopes of the Cerro de Sapo itself. ALCEDINIDAE. 23. Chloroceryle amazona (Latham). One adult, female, Jesusito, 12 April, 1922. This, and the two following species, were all abundant and many could have been secured. In general kingfishers were confined to the • swift clear woodland torrents and were seldom seen about the sluggish Sambii or the lower Jesus, both streams having water the consistency and color of pea-soup. BANGS AND BARBOUR: BIRDS FROM DARIEN. 199 24. Chloroceryle americana isthmica (Goldman). Two adults, male and female, Mt. Sapo and Jesusito, April, 1922. 25. Chloroceryle inda (Linne). Four adults, both sexes, Jesusito, April, 1922. BUCCOXIDAE. 26. XoTHARCUs PECTORALis (Gray). One adult, female, Rio Esnape, 3 April, 1922. A single specimen brought to camp by an Indian hunter. 27. Malacoptila panamensis panamensis Lafresnaye. Four adults, both sexes, Mt. Sapo, Rio Esnape, and Jesusito, April, 1922. This fine bucco was met with at most of the camps, but was rare and all of the indi\'iduals seen bv the members of the party were secured. 28. Monasa pallescens pallescens Cassin. Three adults, male and two females, Jesusito, April, 1922. These vary among themselves as to the shade of gray on the wing- coverts and also as to the grayness or blackness of the throats and chests. We are unable to distinguish them, either by color or size, from skins from Antioquia,'identified by Chapman as paUcsce7is. In the deep woods these beautiful birds were far from common but about the Indian habitations they were very abundant. At the house of Churima, where several nights were passed, they might be seen at any time dusting in the little pathway which led to the canoe landing or perched about the house almost as tame as the little paraquets which crawled about playing with the Indian children. 29. Nonnula frontalis pallescens Todd. Eight adults, both sexes, Jesusito and Rio Esnape, April, 1922. This series agrees with skins already in M. C. Z. from the Canal Zone. The series presents some \ariation in color but on the whole is 200 bulletin: museum of comparative zoology. much paler than old "Bogota" skins. Some time ago Bangs and Penard compared Canal Zone skins with the specimens sent them by Todd, upon which paUescens (type-locality, Fundacion, Santa ISIarta, Colombia), was founded, and found them to agree essentially. There is a doubt in our minds as to Todd's having selected the right form for his new name; provided, however, there actually are two forms. The little Nonnulas were never observed except in the darkest and dampest of lowland woods. GALBULIDAE. 30. Jacameeops aurea penardi, subsp. nov. One adult, male, Rio Esnape, 3 April, 1922. Type.— M. C. Z. 116,609 adult 9 , Costa Rica: Carrillo, 19 Novem- ber, 1898. C. F. Underwood. Characters. — Similar to ./. a. aurca (P. L. S. Miiller) of Surinam, but with a much more slender bill ; colors about the same except that in the northern form the lower surface of the tail is perhaps rather more purplish, less greenish. Measurements. No. Sex LocAXiTY Wing /. aurea aurea (P. L. S. Muller). CULMEN TO Base of FOREHE.U) Width of bill, at Nostril 47,855 9 Brazil: "Amazons" 113 49 13 Penard Coll. 9 Surinam: Paraku Saramacca 114 51 13 (( (1 c^ 111 J. aurea penardi Bangs and Barbour. 52 13 116,608 o^ Costa Rica, Carrillo 111 51 12 116,609 46,445 9 9 " " " 107 Panama, Line of R.R. 108 49 47 11 11 87,472 c? Rio Esndpe 112 49 11.5 Remarks. — We ha\-e named this northern form of J. aurea, which ranges from the Caribbean slope of Costa Rica south at least to eastern Panama, in honor of our colleague Thomas E. Penard, who has lent us two fine topotypes of the typical form, and has kindly helped in the preparation of this paper with the loan of other specimens and with many suggestions on points of nomenclature, etc. BANGS AND BARBOUR: BIRDS FROM DARIEN. 201 Having, by chance, previously concluded that this form should be separated, we turned to Ridgway's Birds of North and Middle Amer- ica and found that, in reality, he had already called attention to the very slender bill of the northern form. On comparison the bill in the two forms is seen at once to be so obviously broad in the southern and so slender in the northern that we consider that they must be separated on this character alone. Measurements, unfortunately, do not show these differences very well. The splendid giant jacama was killed among the very fii'st birds taken and although afterward constantly sought, no other example was ever seen. 31. Brachy GALEA SALMONi Sclater and Salvin. One somewhat immature male, Jesusito, 7 April, 1922. Directly across the stream and right opposite the lower Jesusito camp was a rather low dead tree and on this, every day, perched some small jacamas, all obviously immature. We shot them all, but in spite of many sharp Indian eyes only one was ever found, so thick were the nettles and thorny creepers into which they fell. The adults were never seen, although watched for daily. RAMPHASTIDAE. 32. Pteroglossus torquatus torquatus (Gmelin). Two adults, male and female, Jesusito, 9 April, 1922. These gaudy little toucans were abundant and as they were stupid and fed in rather low trees many were killed to eat. 33. Selenidera spectabilis Cassin. Four adults, both sexes, Mt. Sapo, April, 1922. This toucan was much rarer than the preceding, but equally tame. 34. Ramphastos piscivorus brevicarinatus Gould. One skeleton, Jesusito. The big vellow-breasted toucans were common everywhere. Thev sounded like amphibians, not birds, and looked like tiny bow heavy airplanes as they flapped and then sailed, often for long distances, high 202 bulletin: musetjm of comparative zoology. over the forest. They were great favorites for food, and though in reahty not ver}' shy they flew so high they were by no means easy to secure. CAPITONIDAE. 35. Capito maculicoronatus pirrensis Nelson. Three adult males, Mt. Sapo, April, 1922. xAnother beautiful species, never seen except when it visited the "Cotinga Tree." {Cf. sub Cotinga natcrrcri). PICIDAE. 36. Melanerpes pucherani pucherani (]\Ialherbe). Two specimens, adult male and immature male, Mt. Sapo and Jesusito, April, 1922. Small woodpeckers were often heard and seen but they were usually so high that it was quite impossible to kill them. 37. Celeus loricatus mentalis Cassin. One adult female, Mt. Sapo, 21 April, 1922. 38. Ceophloeus lineatus mesorhynchus Cabanis and Heine. One adult male, Rio Esnape, 3 April, 1922. This big woodpecker seems to have been rarer than the following species. They were not differentiated in the field, howcA'er, and large woodpeckers in general were much more abundant than the smaller species. 39. Campephilus malherbei Gray. Five adults, both sexes, Rio Esnape, and Jesusito, April, 1922. TROGOXIDAE. 40. CURUCUJUS MELANURUS MACROURUS (Gould). Four adults, two males, two females, Jesusito, April, 1922. Four species of trogons were collected, they all seemed about equally abundant and the usual difficulty was encountered in getting decent BANGS AND BARBOUR: BIRDS FROM DARIEN. 203 specimens. The Curcujus and true Trogon were only found in the lowhxnd forest. The Chrysotrogon only at some distance up in the Sapo Hills. The latter species sheds it plumes when shot, even more readily than the others. 41. Trogon strigilatus chionurus Sclater and Salvin. One adult female, Jesusito, 12 April, 1922. 42. Trogonurus curcucui tenellus (Cabanis). Three adults, male and two females, Mt. Sapo and Rio Esnape, April, 1922. These appear to be quite like birds from the Canal Zone and from Costa Rica and show no approach to T. c. cupreicauda Chapman, of western Colombia. 43. Chrysotrogon caligatus caligatus (Gould). Two specimens, male and female, Mt. Sapo, April, 1922. TROCHILIDAE. 44. EUTOXERES AQUILA SALVINI Gould. One adult, female, Mt. Sapo, 20 April, 1922. W. S. Brooks. The only specimen seen, and the same site was visited again and again, was taken while feeding on the red flowers of one of the small species of banana-like plants. 45. Phaethornis guyi coruscus Bangs. Three females, Mt. Sapo, April, 1922. This was one of the solitary hermits of the dark damp groves of "tagua" or ivory nut-palms. The two following species were found in similar situations and while a good many were seen, they were so active and made such long erratic flights that they were by no means easy to collect. They visited none of the flowering trees and shrubs which were regularly watched for hummingbirds. The Threnetes was similar in habits. 46. Phaethornis longirostris cassini Lawrence. Three adults, both sexes, Rio Esnape and Jesusito, April, 1922. Simon, the latest authority on the Hummingbirds (Historic natu- 204 bulletin: museum of compakative zoology. relle des Trochilidae, 1921), sinks P. longirostris ccphalus (Bourcier and Mulsant) type-locality Rio San Juan, Nicaragua, into the syn- onymy of P. longirostris longirostris (Lesson and Delattre), type-locality Guatemala, and recognizes as a distinct species P. cassini Lawrence, type-locality Turbo, giving as its range northern Colombia to Ver- agua. It -is, of course, true that specimens from extreme eastern Panama are much more different from true P. longirostris than are birds from Nicaragua and Costa Rica, in fact many specimens from the latter countries are barely distinguishable from P. longirostris longirostris of Guatemala. We cannot accept cassini as a distinct species, though recognizing it as a subspecies and allowing cephalus to stand as a rather poorly characterized intermediate form. 47. Phaethornis adolphi nelsoni, nom. no v. Phacthornis adolphi fraterculus Nelson, Smithsonian misc. coll., 27 September, 1912, 60, no. 3, p. 9. Type-locality Cana, Panama. Not Phacthornis fraterculus Gould, Monogr. Troch., 1S61, 1, p. 18. Two adults, female and sex not determined, Jesusito, April, 1922. 48. Threnetes ruckeri darienensis, subsp. nov. Two adults, male, Mt. Sapo, 23 April, 1922. Type.— M. C. Z. 87,511 adult cf , E. Panama: Mt. Sapo, 23 April, 1922. Barbour, Brooks, and Underwood. Characters. — Similar to T. r. fraseri (Gould) of Ecuador and western Colombia, and with similar dark green upper parts, but with the cinnamon throat-patch much larger and brighter and with the belly much paler and clearer gray. Similar also to T. ruckeri ruckeri (Bourcier), (Nicaragua to western Panama) but upper parts dark green instead of bronzy green; and belly gray, not buffy. Chapman mentions the characters of this well-marked form and although he appears to have had plenty of material he did not name it. 49. Chalybura buffoni micans, subsp. nov. Eighteen adults, both sexes, Mt. Sapo and Jesusito, April, 1922. Type.— M. C. Z. 87,514 adult cf , E. Panama: Mt. Sapo, 25 April, 1922. Barbour, Brooks, and Underwood. Characters. — Similar to C. b. huffonii (Lesson) of the Magdalena Valley, (common in Bogota collections) but adult male with the tail. BANGS AND BARBOUR: BIRDS FROM DARIEN. 205 including middle rectrices, blue-black (a few skins only in the large series before us from eastern Panama and western Colombia have some slight greenish at base of middle rectrices and along the extreme outer edges of other rectrices, ciuite different from the dull, bronzy middle rectrices and the other extensively bronze rectrices, in C. b. hiiffoni) and underparts paler and much more bluish green. From C. h. aeneicauda Lawrence of Venezuela and the Santa Marta Region of Colombia the new form differs at once in its blue-black instead of green middle rectrices. Size similar to that in the other race. This is another well-marked subspecies to the characters of which Chapman has already called attention but which he did not name. It seems to us too different from the other subspecies ever to be confused with any of them. Its range extends from the Canal Zone in Panama to western Colombia. This big, mealy green hummer was a common visitor to low flower- ing shrubs near the two camps on the Jesusito and on Mt. Sapo. It was very abundant, though far more males were seen than females. 50. Klais guimeti (Bourcier and Mulsant). One immature male, Mt. Sapo, 21 April, 1922. 51. Damophila panamensis Berlepsch. Five adults, both sexes, Jesusito, April, 1922. This species and the following were taken at flowering shrubs which they visited at all hours of the day in common with the Chalybura. o2. POLYERATA AMABILIS (Gould). Eighteen adults, both sexes, Jesusito, April, 1922. 53. Thalurania fannyi fannyi (Delattre and Bourcier). Three adults, two males and a female, Mt. Sapo and Esnape, April, 1922. The two males are exceptionally fine ones, with very long tails, long bills and of very dark rich coloration. At first we thought they might represent a new form, but afterwards decided to consider them hand- some old adults of true fannyi. Measurements. CULMEN No. Sex. Locality Wing Tail to base 87,556 cJ' ad. Mt. Sapo 53 45 25. 87,555 cf ad. Rio Esndpe 53 44 23.5 206 bulletin: museum of comparative zoology, 54. Heliothrix barroti (Bourcier and Mulsant). Two adults, females, Mt. Sapo and Rio Esnape, April, 1922. These lovely hummers were seen very occasionally and then onh^ when they came to bathe in the streams near camp. They hovered motionless over the brook for a few seconds and then dipped hurriedly up and down into the cool, clear water perhaps half a dozen times and then disappeared. By watching the pools during the late afternoon the two specimens, both females, were secured. FORMICARIIDAE. 55. Cymbilanius lineatus fasciatus Ridgway. One adult male, Jesusito, 9 April, 1922. Nineteen species of ant-thrushes were taken. There is nothing special to remark as to their habits. All are found walking with characteristic tread on the floor of the deep, dark forest. Many were shot by the Indians who were especially useful in getting these birds as they made no noise while walking and one and all had keen eyesight. There was great rivalry to see who could secure the largest day's bag, tobacco being the reward. 56. Thamnistes anabatinus coronatus Nelson. Two adults, male and female, Mt. Sapo, 20, 22 April, 1922. 57. Thamnophilus punctatus atrinucha Salvin and Godman. Thirteen adults, both sexes, Mt. Sapo, Rio Esnape, and Jesusito, April, 1922. 58. Dasyithamnus puncticeps puncticeps Salvin. Four adults, one male, three females, Mt. Sapo, April, 1922. 59. Myrmotherula brachyura (Hermann). One adult male, Jesusito, 7 April, 1922. 60. Myrmotherula surinamensis pacifica Hellmayr. Three specimens, adult male and female and immature male, Mt. Sapo and Jesusito, April, 1922. BANGS AND BARBOUR: BIRDS FROM DARIEN. 207 61. Myrmopagis AXILLARIS ALBiGULA (Lawrcnce). Twenty-eight specimens, adults of both sexes and immature males. Mt. Sapo, Rio Esnape, and Jesusito, April, 1922. This and the succeeding were two of the very commonest forest species. A very large series could easily have been collected. 62. Myrmopagis fulviventris (Lawrence). Sixteen adults, both sexes, Mt. Sapo, Rio Esnape, and Jesusito, April, 1922. 63. MiCRORHOPiAS BoucARDi CONSOBRINA (Sclater). Four adults, three males and a female, Mt. Sapo and Jesusito, April, 1922. 64. Cercomacra tyrannina tyrannina (Sclater). Five adults, three males, two females, Rio Esnape and Jesusito. 65. Myrmeciza laemosticta palliata Todd. Twelve adults, both sexes, Mt. Sapo, April, 1922. Making due allowance for the difference in season, — Todd's origi- nal series, from Santander, Colombia, was taken from August to December, ours in April — this series is scarcely separable from the well-marked Colombian form, palliata. The character, of our skins, on the whole, is very slightly intermediate, for instance, the flanks and sides in our males are slightly darker and more reddish brown than in Todd's. This, however, is not so in the female kindly lent us by Todd. In one of our males the black of the throat ends abruptly at the upper chest, in the others it is more or less extended irregularly over the chest and upper breast. 66. My^rmeciza maculifer cassini (Ridgway). Thirteen adults, both sexes, Rio Esnape and Jesusito, April, 1922. 67. FoRMiCARius ANALis PANAMENSis Ridgway. Six adults, both sexes, Mt. Sapo, Rio Esnape, and Jesusito, April, 1922. One adult male gives a wing-length of 93 mm., the other three run from 85 to 88. In the two females it is 87 and 90 respectively. 208 bulletin: museum of compakative zoology. This ant-bird called like a little tinaraou. Its long clear whistle repeated in crescendo six or seven times was easily imitated and several times the bird was brought fairly near and a shot thus secured. In general, however, the species was much more shy than the others of the family. 68. Hylophylax naevioides naevioides (Lafresnaye). Thirty-eight adults, both sexes, Mt. Sapo, Rio Esnape, and Jesusito, April, 1922. In the Birds of North and Middle America, Ridgway did not recognize the Costa Rican Hylophylax naevioides capnitis (Bangs). Now we have a very large series of each race and believe that the Costa Rican form can easily be distinguished from true naevioides of Panama. The characters originally claimed for the northern bird are slight, and perhaps only average characters, but besides these, the female of capnitis is much grayer, less buffy below and the spots on the chest much darker and more distinct, and in both sexes the subterminal dark band on the tail is narrower and the tip is darker more cinna- momeous, less buffy or whitish. This was another of the excessively common woodland species of which an enormous series might easily have been secured. 69. Anoplops bicolor bicolor (Lawrence). Five adults, both sexes, Mt. Sapo and Jesusito, April, 1922. 70. Phaenostictus macleannani chocoanus, subsp. nov. Three adults, two males and a female, JVIt. Sapo and Rio Esnape. April, 1922. Type.— M. C. Z. 87,352 adult cf, E. Panama: Mt. Sapo, 20 April, 1922. Barbour, Brooks, and Underwood. Characters. — Similar to P. m. macleannani (Lawrence) of the Canal Zone, but much paler throughout; pileum much grayer, less brownish; front paler still, whitish gray; chest-band pale cinnamon- rufous and margins to feathers of back and lower underparts very pale buff. Measurements. CULMEN No. Sex Locality Wing Tail Tarsus TO BASE 87,352 o^ Mt. Sapo 94 89 33 21 87,353 9 U u 88 85 31 23 87,354 cf Rio Esndpe 93 87 32 22 BANGS AND BARBOUK: BIRDS FROM DARIEN. 209 71. Rhopoterpe stictoptera Salvin. Eight adults, both sexes, Mt. Sapo, Rio Esnape, and Jesusito, April, 1922. So far as we know this species has not been recorded from anywhere south of Nicaragua, yet we are unable to separate our eastern Panama skins from Xicaraguan specimens w4th which we have compared them.- 72. Pittasoma michleri michleri Cassin. One adult female, Mt. Sapo, 21 April, 1922. Although special effort was made to secure the large Grallaria-like birds, this was the only individual secured. 73. Hylopezus perspicillata perspicillata (Lawrence). Two adults, male and female, Rio Esnape and Jesusito, April, 1922. On geographical grounds these two skins must be referred to true perspicillata. Judged by characters alone one should be referred here and one to H. perspicillata lizanoi (Cherrie). In running through a large series of the latter we find all its characters to be very variable and the form seems hardly worthy of recognition. FURXARIIDAE. 74. Sclerurus mexicanus axomalus, subsp. nov. Two adults, male and female, Mt. Sapo, April, 1922. Type.— M. C. Z. 87,367 adult 9 , E. Panama: Mt. Sapo, 25 April, 1922. Barbour, Brooks, and Underwood. Characters. — Similar in size and proportions to S. m. pullus Bangs of Costa Rica and Chiric^ui but coloration much paler; the upperparts and belly warm sepia or Vandyke-brown ; rump and upper tail-coverts chestnut. Similar also to S. m. mexicanus Sclater but bill much shorter and stouter and throat and chest paler (not far from Sanford's-brown in the new form, and from chestnut or tawny chest- nut in mexicanus) and more sharply defined against the color of the lower underparts. Measurements. No. Sex Locality Wing Tail Tarsus CULMEN 87,367 9 Panama, Mt. Sapo 78 48 22. 23. 87,366 d" u a u 80 50 21. 107,336 cT " Loma del Leon 79 52 22. 23.5 107,337 c? U U ti <1 82 54 22.5 24. 210 bulletin: museum of comparative zoology. Remarks. — This form, which usually has been referred to S. in. mexicanus in spite of the fact that a much darker subspecies, C. m. pullus Bangs, occurred in Costa Rica and Chiriqui cutting the range in two, appears to be quite distinct. It differs from true mexicanus conspicuously, the latter form having a much longer and more slender bill (culmen to base, 26 to 27.5 mm.) and in the color of the throat and chest which are paler and brighter in the Panama bird. In western Colombia another dark form C. »?. ohscurior Hartert occurs, even darker than pullus and dusky rather than brownish on the lower under- parts. With the few Guatemalan specimens seen by us, we are unable to separate S. m. certus Chubb of Guatemala from true mexicanus. We have not seen any skins from Amazonia. As we know the species the northern races stand thus : — 1. Sclerurus mexicajius mexicanus Sclater. Southeastern ^Mexico to Honduras. 2. Sclerurus mexicanus pullus Bangs. Costa Rica and Chiriqui. 3. ISclerurus mexicanus anomalus Bangs and Barbour. Panama, Canal Zone to extreme eastern Panama. 4. Sclerurus mexicanus ohscurior Hartert. Northwestern Equador and western Colombia. In the field we were not able to observe any great chfferences in habits of some of the furnariids from the dendrocolaptids. 75. Sclerurus guatemalensis guatemalensis (Hartlaub). Six adults, both sexes, ^Nlt. Sapo, Rio Esnape, and Jesusito, April, 1922. These skins are somewhat darker, more olive-brown, less rufous brown than specimens from Guatemala and British Honduras; they can, however, be matched by certain examples from southwestern Costa Rica and western Panama. They are probably somewhat intermediate tending a little toward S. g. salvini Salvador! and Festa of western Ecuador, which, however, we have not seen. 76. Xenops gexibarbis ridgwayi Hartert and Goodson. Three specimens, two males and a female, Mt. Sapo and Jesusito, April, 1922. BANGS AND BARBOUK: BIRDS FROM DARIEN. 211 77. Philydor fuscipennis Salvin. One somewhat immature female of this very rare species was se- cured at Rio Esnape, 3 April, 1922. 78. AuTOMOLUS PALLiDiGULARis PALLiDiGULARis Lawrence. Four adult males, Mt. Sapo and Rio Esnape, April, 1922. DENDROCOLAPTIDAE. 79. Dendrocolaptes sancti-thomae sancti-thomae (Lafresnaye). Two adult females, Jesusito, April, 1922. These specimens are wholly referable to true sancti-thomae from which thev differ onh' in sli";htl\- darker chestnut tails, a difference perhaps due to being freshly made skins as compared with old material. The range of sancli-thomac therefore extends through eastern Panama to western Colombia (see Chapman, Bull. Amer. mus. nat. hist., 1917, 36, p. 427) and north probably along the Caribbean slope of Panama and Costa Rica, just to southeastern Mexico. The well-defined D. s. hesperius Bangs, occupies a narrow belt in western Panama, western Costa Rica, and western Nicaragua. The six forms of dendrocolaptids collected all had more or less simi- lar habits. They have, as is well kno\\"n, the habit of creepers but often drill vigorously with their bills, making quite as much noise as a small woodpecker. Although we saw several columns of army-ants we did not find these birds coming to the ground to feed upon them or the insects which they stir about as the formicariids naturally did. 80. XiPHORHYXCHUs LACHRYMOsus LACHRYMOsus (Lawrencc). One adult female, Mt. Sapo, 24 April, 1922. 81. XiPHORHYNCHUS NANUS NANUS (Lawrence). Two adult males, Jesusito, April, 1922. 82. Glyphorhynchus cuneatus pfctoralis Sclater and Salvin. Eleven adults, t)otii sexes, Mt. Sapo, Rio Esnape, and Jesusito, April, 1922. 212 bulletin: museum of comparative zoology. 83. Dendrocincla lafresnayei ridgwayi Oberholser. Six adults, both sexes, Mt. Sapo, and Jesusito, April, 1922. 84. Dendrocincla homochroa ruficeps Selater and Salvin. One adult female, Mt. Sapo, 21 April, 1922. COTIXGIDAE. 85. CoTiNGA NATTERi (Boissoneau). Ten specimens, adults of both sexes and immature males, Mt. Sapo, April, 1922. This splendid bird was only taken at the "Cotinga Tree" near the Mt. Sapo camp. Shortly after this camp was located two long trails were cut leading up two high hog-back spurs of the main range. One of these led to the ridge called La Jarcia where trees were cut down until a fine look-out was cleared. From this look-out a maze of steep crests and valleys was disclosed. We were always looking for white Cotingas and several were seen from our clearing evidently visiting a feeding tree but it stood in an absolutely inaccessible spot. After this, search for feeding trees was redoubled and before many days a tree with ripe fruit, a small fig evidently, was found on another ridge only a mile or so from the camp. After this tree was located some one or other of the party kept a constant watch there every day from dawn until dark. Xo wliite Cotingas came to the tree but ten blues were collected. The buccos and Cotingas at large visited the tree at three very definite times of day. A few in the early morning but most about 8.30 a.m., then there was a lull and the tree had only an occasional visitor, usually some migrant warbler, until noon when for about an hour there was abundant visitation, thereafter little appeared until about 4 p.m. This last feeding period was distinctly the least im- portant of the three. The birds came to the tree in perfect silence simply appearing from out of the vast immensity of the surrounding forest, remaining a few moments and then leaving as they came. Several other feeding trees were seen but they grew on such precipitous slopes that retrieving the fallen birds would have been impossible, had one been able to find a place to stand and shoot once the tree was reached. It is sometimes easy to locate a feeding tree from a distance but identifying the same tree when one is close to it in the forest is often next to impossible. BANGS AND BARBOUR: BIRDS FROM DARIEN. 213 86. Attila citreopyga citreopyga (Bonaparte). One adult male, Mt. Sapo, 20 April, 1922. A single example of this species appeared one morning and no other was seen. 87. LiPAUGUS HOLERYTHRUS HOLERYTHRUS Sclater and Salvin. One adult male, Mt. Sapo, 24 April, 1922. This skin is slightly richer and more rusty in color than northern specimens (Guatemala to Costa Rica) and thus approaches L. liolery- thrus rosenbcrgi Hartert, type-locality Rio Dagua, western Colombia. These reddish cotingas came to the tree quite often and frequently in company with other species, so they were not always collected. This species and the following are beyond doubt really congeneric. 88. Lathria uxirufa castaneotixcta Hartert. Six adults, both sexes, Mt. Sapo, April, 1922. Chapman says he cannot distinguish L. u. clara of Panama from L. u. castaneotinda of western Colombia and northeast Ecuador. The present series bears out what he has said, though Panama birds are on the whole a trifle, either paler or duller than those from western Colombia. The larger " red cotinga" was the one species of the family that was sometimes identified and shot in the forest away from a feeding tree. 89. Pachyrhamphus cinnamomeus Lawrence. Three adults, two males and a female, Jesusito, April, 1922. This species was only killed in a feeding tree on the summit of a steep hill near our upper Jesusito camp. 90. TiTYRA SEMIFASCIATA COLUMBIANA RidgWaV. & ' One adult male, Mt. Sapo, 24 April, 1922. This specimen is rather small, the wing-length being only 116 mm. The tail-pattern is typically that of Columbian a. This bird appeared once, almost at dusk, and the flash of white made us think that at last the pajaro del EspiritM Santo had appeared. 214 bulletin: museum of comparative zoology. PIPRIDAE. 91. Manacus vitellinus vitellinus (Gould). Six adults, both sexes, Rio Esnape and Jesusito, April, 1922. A common species in thickets along the stream-courses. 92. PiPRA ERYTHROCEPHALA ACTINOSA, subsp. UOV. Thirty specimens, adults of both sexes, and young males, Mt. Sapo, Rio Esnape, and Jesusito, April, 1922. Type.— M. C. Z. 87,170 adult d", E. Panama, 21 April, 1922. Barbour, Brooks, and Underwood. Characters. — Similar to P. e. erythrocephcda (Linne) but larger. The adult male, similar in color; adult female slightly paler olive- green above and slightly paler and grayer, less yellowish olive below. \Ying in adult male, 58.01 (57-59); fifteen specimens. Remarks. — After recognizing the slightly different form we have just named, Pipra erythrocephala is made up of four geographic races, as follows: — 1. Pipra erythrocephala erythrocephala (Linne). Surinam, Cayenne, Venezuela, Trinidad, etc. Small wing in adult male, 54.55 (53-5(3); head intense orange-yellow. 2. Pipra eryihrocephala actinosa Bangs and Barbour. Eastern Panama to the Santa JNIarta Region of Colombia. Larger wing in adult male, 58.01 (57-59), male similar in color, female slightly paler. 3. Pipra erythrocephala flammiceps Todd. Santander, Colombia. Small wing in adult male, 53-54; head darker, more reddish orange. 4. Pipra erythrocephala berlepschi Ridgway. Bogota Region of Colombia to Peru. Large wing in adult male 61., (60-62); head pale yellow. In all the forms the adult female is a little larger than the male, about 2 mm. on the average, in the length of the wing. These little manikins were most amusing birds to watch. ^^ hile on a long tramp one experience with these lively little feathered imps was most pleasing. It was late afternoon when attracted by a most sur- prising whirring and snapping of wings; and searching about for a moment a low tree was found literally swarming with yellow-headed BANGS AND BARBOUR: BIRDS FROM DARIEN. 215 manikins. They had evidently met by design and were strutting, bowing, raising and lowering their wings and generally going through such antics as one only expects from Birds of Paradise. No females were anywhere about and the observation was never repeated though the tree was revisited on various occasions. 93. PiPRA VELUTINA MINUSCULA Todd. Ten adults, both sexes, Mt. Sapo and Jesusito, April, 1922. Our skins are wholly referable to this well-marked form, lately de- scribed by Todd; its characters were, however, long before dwelt on at length by Hellmayr, who refrained from giving it a name only be- cause he had no skins from Veragua, the type-locality of P. vclufina velutina Berlepsch. The little Blue-headed Manikin frequently came to the " Cotinga tree" but many specimens were taken elsewhere as well. 94. CoRAPiPO ALTERA ALTERA Hellmayr. Four adult males, Mt. Sapo, April, 1922. We agree with Carriker that the short first primary is a character that separates C. altera altera Hellmayr and C. altera heteroleuca Hellmayr specifically from C. leucorrhoa (Sclater). For some reason only four specimens of this very conspicuous bird were found, all in a small low coppet near camp. 95. ScoTOTHORUS TURDiNUS STENORHYNCHUS (Sclater and Salvin). Five adults, both sexes, Mt. Sapo, Rio Esnape, and Jesusito, April, 1922. 96. Sapayoa aenigma Hartert. Eight adults, both sexes, Mt. Sapo, April, 1922. This series of full plumaged spring birds, shows that the male (be- sides having a semiconcealed yellow crest which the female does not possess) is paler and much more yellowish olive-green than the female, especially on the underparts, the throat and chest being often quite yellow. This species was abundant and confined to the stream-bottoms. 97. Laniocera rufescens (Sclater). One adult female, Mt. Sapo, 22 April, 1922. This bird was taken at the "Cotinga Tree" in company with the 216 bulletin: museum of comparative zoology. various species of what our outfit came to call "red cotingas." It is almost impossible to believe tliat the bird does not really belong in that family. TYRANNIDAE. 98. Onychorhynchus mexicanus fraterculus Bangs. Nine adults, both sexes, Mt. Sapo, Rio Esnape, and Jesusito, April, 1922. The Paradise flycatchers were abundant and widespread. One observation at the time seemed very striking. Two birds were cap- tured only slightly wounded, each acted in exactly the same way. The mouth was opened wide, the great crest fully expanded and then the head was slowly waved from side to side. Exactly the same mo- tion was often enacted by a large rose-crested cockatoo which was shot and wounded near Wahaai, Ceram. A captive cockatoo subsequently repeated the minatory gestures whenever it was shown a snake. It is curious that birds having wide transverse crests but so far separated in the system should use such a similar method of attempting to terrify an enemy. 99. Oncostoma olivaceum (Lawrence). Two adults, male and sex undetermined, Rio Esnape and J esusito April, 1922. It is surprising that this was found to be a rare bird in eastern Panama, it is so common throughout most of its range. 100. Placostomus coronatus superciliaris (Lawrence). Eight adults, both sexes, Mt. Sapo and Rio Esnape, April, 1922. Nests made of a filamentous lichen, great pendulous purses quite without form, were observed hanging from limbs above all the streams. None were occupied but the Indians said that they were made by this boat-billed flycatcher, which was rather abundant. 101. Craspedoprion olivaceus bardus, subsp. nov.' Eight adults, seven males and a female, Mt. Sapo, Rio Esnape, and Jesusito, April, 1922. Ti-PE.— M. C. Z. 87,029 adult &, E. Panama: Mt. Sapo, 20 April, 1922. Barbour, Brooks, and Underwood. 1 Bardus, a um. Inactive, dull, sluggish, etc. BANGS AND BAKBOUR: BIRDS FROM DARIEN. 217 Characters. — Similar to Craspedoprion olimccus aequinoctialis (Sclciter) of eastern Ecuador and eastern Colombia, but larger; upper- parts brighter, yellower, less dusky olive-green; chest and breast paler, yellower; lower underparts much brighter yellow. Similar also to C. o. flaws Chapman of the Santa ]\Iarta region of Colombia and of about the same size; the upperparts similar, but underparts very much deeper and richer yellow, (lemon-yellow), on the belly; the olivaceus flammulations of chest and breast darker, less grayish. Wing in seven males, 74-77; in one female 71. Remarks. — In 1914 when Chapman named his C. o. flavus from Santa Marta he gave its range as extending north to Panama, explain- ing that lack of proper material had caused the Panama bird to be called C. aequinoctialis. At that time Chapman had two Panama skins only. Since then the American Museum has received a series from eastern Panama collected by Richardson, these have kindly been lent to us and agree exactly with ours, and differ much from a good series from Santa Marta lent by the Carnegie Museum (^Y. E. C. Todd). The color of the underparts in flavus of Santa Marta is much nearer to the color of the underparts in C. olivaceus guianensis McConnell of the Guianas and Venezuela (the forms of course differ in other respects) than it is to the rich, bright yellow of the lowerparts in the new bird from eastern Panama. 102. Tyranniscus vilissimus parvus (Lawrence). One adult male, Mt. Sapo, 23 April, 1922. This species may well be more abundant than the appearance of only a single specimen in the collection would indicate. Such ex- cessivelv small birds are reallv onh' found bv chance. 103. Myiozetetes cayanensis harterti Bangs and Penard. Three adults, two males and a female, Jesusito, April, 1922. Though somewhat intermediate, these skins are nearer to harterti than to M. c. hellmayi, Hartert and Goodson, of v>'estern Colombia and northwest Ecuador. This is one of a number of species which were only found about the lower Jesusito camp where there was more or less of a clearing near by. It was not a bird of the forest at all. 218 bulletin: museum of comparative zoology. 104. Myiozetetes granadensis Lawrence. Two adults, male and female, Jesusito, April, 1922. These are slightly darker above and perhaps a little deeper yellow below than Canal Zone birds, in this respect being like skins from western Colombia. This difference haA'ing already been pointed out by Chapman. It seems to us, as it did to Chapman, that the form is not worthy of recognition. This is another species, with king bird-like habits to which the same remarks apply as the preceding. 105. PiPROMORPHA oleaginea parca Bangs. Five adults, both sexes, Mt. Sapo, April, 1922, Rio Esnape, and Jesusito. 106. MiONECTEs OLiVACEus HEDERACEus Bangs. Four adults, both sexes, Mt. Sapo, April, 1922. These specimens are much nearer hedcraccus of western Colombia than they are to M. o. olivaceus Lawrence of Costa Rica and western Panama. 107. Cnipodectes subbrunxeus subbrunneus (Sclater). Four males, Rio Esnape, April, 1922. All large birds, (wing, 94-97 mm.). Ridgway recognized a large and small species occurring together. Hellmayr claims that the small examples are immature, but retains the name C. s. minor Sclater for the form ranging from eastern Peru to western Brazil, which he says differs slightly in color. This was an inhabitant of the deepest woods, whereas very few other members of the family were found there. 108. Myiobius atricaudus atricaudus LawTence. Two adults, male and female, Jesusito, April, 1922. A common bird of the open clearings, this and many other tyrant birds abound near the Indian's field, but as we confined nearly all of our work to the forest, few appear in the collection. 109. Myiodius sulphureipygius aureatus Bangs. Thirteen adults, both sexes, Mt. Sapo, Rio Esnape, and Jesusito, April, 1922. BANGS And BARBOUR : BIRDS FROM DARIEN. 219 This species occurred not only about the open clearings but where there were small sun-lit openings along the streams as well. 110. Terenotriccus erythrurus fulvigularis (Salvin and (Godman). Five adults, both sexes, Mt. Sapo, Rio Esnape, and Jesusito, April, 1922. 111. XuTTALLORNis BOREALis MAJORiNUS Bangs and Penard. One adult female, Mt. Sapo, 24 April, 1922. One afternoon a small, dull colored bird came to the "Cotinga Tree," fed upon the fruit and was promptly shot. When picked up it proved to be this familiar compatriot. The date seems rather late. 112. Myiochanes virens (Linne). Three adults, male and two females, Mt. Sapo and Jesusito, April, 1922. 113. Myiarchus crinitus crinitus (Linne). One adult female, Jesusito, 9 April, 1922. 114. Myiarchus ferox panamensis (Lawrence). One adult male, Mt. Sapo, 26 April, 1922. 115. Myiarchus tuberculifer nigriceps Sclater. Four adult females, Rio Esnape and Jesusito, April, 1922. Another of the species most characteristic of the clearings. 116. Myiodynastes luteiventris Sclater. Three adults, male and two females, Mt. Sapo and Jesusito, April, 1922. These birds were shot from a feeding tree which was often visited from the upper Jesusito camp. The tree was very high and many birds escaped. This species was a frequent visitor and fed greedily on the fruit. The specimens shot staining freely about the vent from the fruit-juices after the manner of some Cotingas. It is hard to believe that this species is not in very truth a Cotinga and not a tyrannid at all. This, indeed, seems about certain. 220 bulletin: museum of comparative zoology. 117. Myiodynastes maculatus nobilis (Sclater). One adult female, Mt. Sapo, 24 April, 1922. This example agrees in all of the distinguishing characters, longer wing (109 mm.); olive not cinnamomeus pileum; much paler and on primaries much narrower, wing-edgings, and much more narrowly streaked undertail coverts, with M. m. iiwoley^s Ridgway of southeast Mexico. If that form were migratory, which we believe it is not, this skin would of course be referred to it. Probably our bird is in reality an aberrant example of 7wbilis which approaches very close to insolcns through indi\-idual variation. We have often seen such instances before, sometimes with island forms. This is another species from the " Cotinga Tree" which may be mis- placed in the system. 118. Tyr.anxus tyrannus (Linne). Three adults, two males and a female, Jesusito, April, 1922. This and the following were common yellow-billed king birds of the clearings. 119. Tyrannus melancholicus chloronotus Berlepsch. Three adults, two males and a female, Jesusito, April, 1922. OXYRUNCIDAE. 120. Oxyruncus brooksi, sp. nov. Three adults, male and two females, INIt. Sapo, April, 1922. Type.— M. C. Z. 87,199 adult &, E. Panama: Mt. Sapo, 25 April, 1922. Barbour, Brooks, and Underwood. Characters. — Similar to Oxyruncus f rater (Sclater and Salvin) of Costa Rica and western Panama, but at once distinguished by having white, not yellow, underparts; similar also to 0. hypoglaucus (Salvin and Godman) of British Guiana and with white underparts as in that form, but with upperparts, paler and brighter, more yellowish, olive- green; larger wing-coverts broadly margined with light yellow; secondaries also widely margined terminally with light yellow; under- parts with smaller and fewer blackish spots (than in any form), which became very sparse and indistinct on flanks and sides; flanks and sides washed with dull pale yellow; tail shorter. BANGS AND BARBOITK: BIRDS FROM DARIEN. 221 No. Sex Measurements. Wing Tail Tarsus Exposed CULMEN 87,199 ■ d" 90 52 19 15.5 87,198 9 88 50 19 16. 87,200 9 91 51 19 17. Remarks. — It gives us great pleasure to name this fine new form in honor of W. Sprague Brooks, who did so much to make the trip a success. To find in eastern Panama a very distinct and white-beUied sharp- bill was indeed a surprise. We are, however, rather inclined to be- lie\'e, that the new form, in spite of its white underparts is more nearly related to the yellow-bellied f rater than it is to the only other white- bellied form, hi/poghtucus of Mt. Roraima. We have not followed the custom of modern ornithologists in allow- ing subspeeific rank only, to the various forms of the sharp-bill. The discontinuous distribution of the species in tropical America, together with the excellent characters shown, may well, we think, be used as an argument for considering them all full species. The birds have the habits of Cotingas and these specimens came to the feeding tree in company with several species of Cotingidae. It is not improbable that the Oxyruncidae is scarcely separable from the Cotingidae. TURDIDAE. 121. Hylocichla ustulata swainsonii (Cabanis). Four adults, both sexes, Mt. Sapo, Rio Esnape, and Jesusito, April, 1922. 122. TuRDUs TRiSTis DAGUAE Berlepsch. Five specimens, three adults, two males and a female and two spotted young, Mt. Sapo, April, 1'922. A deep wood thrush with the actions and appearance«of our familiar robin but naturally in a very strange setting. TROGLODYTIDAE. 123. Pheugopedius fasciato-ventris albigularis (Sclater). One adult, male, Jesusito, 8 April, 1922. Most of the tropical wrens encountered were birds of the scattered 222 bulletin: museum of comparative zoology. patches of thick coppet and of the tangles of AJnes and creepers in the deep lowland forest. Some species, liowever, were decidedly ter- restriid with the habits of ant-thnishes. 124. Henrorhina leucosticta darienensis Hellmayr. Kio-ht adults, both sexes, Rio Esnape and Jesusito, April, 1922. 125. Tmryophilt's nigricapillus schottii (Baird). h'onr ;i(hilts, hotli sexes, Jesusito, April, 1922. 12(). Thryophilt^s galbraithii galbraithii (LawTence). Two adults, male and female, Mt. Sapo, 25 April, 1922. Though these skins just match some individuals from the Canal Zone, they are, howe\er, as Chapman has already pointed out, slightly darker and less rufescent than the average from that region; where- fore we agree that the dift'erentiation is too trifling to be considered of subspecific \alue. 127. MiCROCERCT'LTTS PHILOMELA LT^SCINIA vSalvin. Tw^o specimens, male and female, Mt. Sapo, April, 1922. One of these, an adult female, 86,985, has the feathers of the chest and breast, gray with W-shaped markings of grayish white at their tips, below which is a dusky spot. It is somewhat closely similar to a specimen (121,307 M. C. Z.) from El General, Costa Rica. The other, a male, perhaps somewhat immature, though its back is plain ^'andyke-brown, without dusky bars, has the chest and breast dark brownish gray, the whole underparts, crossed by fine, faint dusky bars. This skin also is very much like some birds from Costa Rica. ^^e follow ( 'arriker, (Birds of Costa Rica, p. 753) in allowing but one form to Costa Rica and Panama, which for the present we consider a sul)species of phiJomrJa of Guatemala. \N e huxe had for comparison one specimen only, an immature individual from Guatemala, which matches closelV some of the Costa Rican immatures. This curious little creature has the mouse-like habits of the tiny rails only it was a denison of the deepest, darkest, high and rather open woo