MOCC
*s3rc« <cc BSK<fCC3
tc^ ccc rcips ccc SIdi c ccir
ccecce c ccc Lplcicc (CCffl C&MpCC (CCC OKlpX (CC wc^cmjc < c c <ic
Bfcglcc carr RC(v cc ; j ■'<®._CC cccc
o|c||C
5biiC»Ei%^«c' C* «<C CfCCCC (,c«t
|
(C^CCC«CCCj Iwlfll |
|
|
(^Cjc c(Cj® fc^CC CCC^ rl'cC^Ci |
lUCpCI pCCCi |
|
CCCCCC c*33gpg LCCJC3SXm^ffi5 |
|
|
pKCga |
|
|
gycjf Cy« rc?c c((« |
|
|
ntiCjCjc^i |
|
|
rpCC^ |
|
|
CCCjCCC# |
|
|
cfcsjXjti |
|
|
rcwCffl |
|
|
rwm3 |
|
|
^glljQ |
|
|
^~J' if CT |
|
|
1C CCCC |
ggefer |
|
erect |
s c si |
|
^fiT"5Cl^'<®CS |
c< c <Sj |
P b n o, Yi £
ZOOLOGICA
SCIENTIFIC CONTRIBUTIONS OF THE
NEW YORK ZOOLOGICAL SOCIETY
PUBLISHED BY THE SOCIETY THE ZOOLOGICAL PARK, NEW YORK
NEW YORK ZOOLOGICAL SOCIETY
General Office: 630 Fifth Avenue, New York City
OFFICERS
President, Fairfield Osborn First Vice-President, Alfred Ely
Chairman, Executive Committee & Second Vice-President, Laurence S. Rockefeller Treasurer, Cornelius R. Agnew
General Director, Zoological Park and Aquarium, Allyn R. Jennings Assistant General Director, Harry Sweeny, Jr.
SCIENTIFIC STAFF
Zoological Park
Raymond L. Ditmars, Curator of Reptiles and Insects Lee S. Crandall, Curator of Birds Claude W. Leister, Curator of Mammals and Educational Activities Leonard J. Goss, Curator of Health William Bridges, Editor and Curator of Publications
Aquarium
Charles M. Breder, Jr., Director Christopher W. Coates, Aquarist Ross F. Nigrelli, Pathologist G. M. Smith, Research Associate in Pathology Homer W. Smith, Research Associate in Physiology
Department of Tropical Research
William Beebe, Director John Tee-Van, General Associate Gloria Hollister, Research Associate Jocelyn Crane, Technical Associate
Editorial Committee
Fairfield Osborn, Chairman
Allyn R. Jennings Charles M. Breder, Jr.
William Beebe William Bridges
CONTENTS
Part 1. March 18, 1940.
1. The Breeding Behavior of the Common Shiner, Notropis
cornutus (Mitchill). By Edward C. Raney. (Plates I-IV ; Text-figure 1) 1
2. Divergence and Probability in Taxonomy. By ISAAC GlNS-
BURG 15
3. Miscellaneous Notes on the Eggs and Young of Reptiles. By
Roger Conant & Alexander Downs, Jr 33
4. Occlusion of the Venom Duct of Crotalidae by Electro-coagu-
lation: an Innovation in Operative Technique. By Duval B. Jaros. (Text-figure 1) 49
5. Eastern Pacific Expeditions of the New York Zoological So-
ciety. XVII. A Review of the American Fishes of the Family Cirrhitidae. By John Tee-Van. (Plate I; Text-figures 1-4) 53
6. Eastern Pacific Expeditions of the New York Zoological So-
ciety. XVIII. On the Post-embryonic Development of Brachyuran Crabs of the Genus Ocypode. By Jocelyn Crane. (Text-figures 1-8) 65
7. New Species of British Guiana Heterocera. By W. Schaus . . 83
8. A Papillary Cystic Disease Affecting the Barbels of Amei-
umis nebulosus (Le Sueur), Caused by the Myxosporidian Henneguya ameiurensis sp. nov. By R. F. NlGRELLi & G. M. Smith. (Plates I- VIII; Text-figure 1) 89
9. Caudal Skeleton of Bermuda Shallow Water Fishes. IV. Order Cyprinodontes : Cyprinodontidae, Poecilidae. By Gloria Hollister. (Text-figures 1-17) 97
10. The Histology of the Eye of the Cave Characin, Anoptichthys.
By E. B. GRESSER & C. M. Breder, Jr. (Plates I-III) 113
Part 2. July 3, 1940.
11. Plankton of the Bermuda Oceanographic Expeditions. IX.
The Bathypelagic Caridean Crustacea. By Fenner A. Chace, Jr. (Text-figures 1-64) 117
12. Eastern Pacific Expeditions of the New York Zoological So-
ciety. XIX. Actiniaria from the Gulf of California. By Oskar Carlgren. (Text-figures 1-8) 211
13. Morphological and Embryological Studies on Two Species
of Marine Catfish, Bagre marinus and Galeichthys felis.
By Daniel Merriman. (Plates I-V; Text-figures i-9) 221
PAGE
14. Propagation of the Electric Impulse Along the Organs of the
Electric Eel, Electrophorus electricus (Linnaeus) . By C. W. Coates, R. T. Cox, W. A. Rosenblith & M. Vertner Brown. (Plate I; Text-figures 1-3) 249
15. Notes on the Display Forms of Wahne’s Six-plumed Bird of
Paradise. By Lee S. Crandall. (Text-figures 1-3) 257
16. Acute Hemorrhagic Gastro-enteritis in a Giant Panda. By
Leonard J. Goss 261
17. Two New Species of Trematodes from the Deep Sea Scorpion
Fish, Scor])aena maclurensis Cuv. & Val. By Ross F. Nigrelli. (Plate I; Text-figures 1 & 2) 263
18. Report of the Hospital and Laboratory of the New York
Zoological Park, 1939. Mortality Statistics of the Society’s Collection. By Leonard J. Goss 269
Part 3. November 14, 1940.
19. Eastern Pacific Expeditions of the New York Zoological So-
ciety. XX. Medusae of the Templeton Crocker and Eastern Pacific Zaca Expeditions, 1936-1938. By Henry B. Bigelow. (Text-figures 1-20) 281
20. Two New Species of Trematodes ( Apharyngostrigea bilobata :
Strigeidae, and Cathaemasia nycticoracis : Echinostomidae) from Herons, with a Note on the Occurrence of Clinos- tomum companulatum (Rud.). By 0. Wilfred Olsen. (Plate I) 323
21. Nesting of the Sunfish, Lepomis auritus (Linnaeus), in Tidal
Waters. By Neil D. Richmond. (Plate I) 329
22. Eastern Pacific Expeditions of the New York Zoological So-
ciety. XXL Notes on Echinoderms from the West Coast of Central America. By Hubert Lyman Clark. (Plates I &
II; Text-figures 1-4) 331
23. The Nesting Behavior of Eupomotis gibbosus (Linnaeus), In
a Small Pool. By C. M. Breder, Jr. (Plates I & II; Text- figures 1 & 2) 353
24. Reproductive Activities of a Hybrid Minnow, Notropis cor-
nutus X Notropis rubellus. By Edward C. Raney 361
Part 4. December 31, 1940.
25. Eastern Pacific Expeditions of the New York Zoological So-
ciety. XXII. Mollusks from the West Coast of Mexico and Central America. Part I. By Leo George Hertlein & A. M. Strong. (Plates I & II) 369
26. On the Electric Powers and Sex Ratios of Foetal Narcine
brasiliensis (Olfers). By C. M. Breder, Jr., & Stewart Springer 431
VI
PAGE
27. A Study of the Activities of a Pair of Galago senegalensis
moholi in Captivity, Including the Birth and Postnatal Development of Twins. By Florence De L. Lowther. (Plates I-VI) 433
28. Diets for a Zoological Garden: Some Results During a Test
Period of Five Years. By Herbert L. Ratcliffe 463
29. The Biology of the Smoky Shrew ( Sorex fumeus fumeus
Miller). By W. J. Hamilton, Jr. (Plates I-IV ; Text-figure 1) 473
30. Social and Respiratory Behavior of Small Tarpon. By Arthur
Shlaifer & C. M. Breder, Jr. (Plates I & II ; Text-figure 1)493
31. New Observations on the Blood Group Factors in Simiidi
and Cercopithecidae. By P. B. Candela, A. S. Wiener
& L. J. Goss 513
32. Muscle Dystrophy in Tree Kangaroos Associated with Feeding
of Cod Liver Oil and. Its Response to Alpha-Tocopherol.
By Leonard J. Goss 523
33. Mortality Statistics for Specimens in the New York Aquarium,
1939. By Ross F. Nigrelli. (Plates I-III) 525
y- 34. A Comparison of Some Electrical and Anatomical Character- istics of the Electric Eel, Electrophorus electricus (Lin- naeus). By R. T. Cox, W. A. Rosenblith, Janice A. Cut- ler, R. S. Mathews & C. W. Coates. (Text-figures 1-7) . .553
Index to Volume XXV 563
V
ZOOLOGICA
SCIENTIFIC CONTRIBUTIONS OF THE
NEW YORK ZOOLOGICAL SOCIETY
VOLUME XXV Part 1
Numbers 1-10
PUBLISHED BY THE SOCIETY THE ZOOLOGICAL PARK, NEW YORK
March 18, 1940
CONTENTS
PAGE
1. The Breeding Behavior of the Common Shiner, Notropis
cornutus (Mitchill). By Edward C. Raney. (Plates I-IV; Text-figure 1) 1
2. Divergence and Probability in Taxonomy. By Isaac
Ginsburg 15
3. Miscellaneous Notes on the Eggs and Young of Reptiles.
By Roger Con ant & Alexander Downs, Jr. 33
4. Occlusion of the Venom Duct of Crotalidae by Electro-
coagulation: an Innovation in Operative Technique.
By Duval B. Jaros. (Text-figure 1) 49
5. Eastern Pacific Expeditions of the New York Zoological
Society. XVII. A Review of the American Fishes of the Family Cirrhitidae. By John Tee-Van. (Plate I; Text- figures 1-4) 53
6. Eastern Pacific Expeditions of the New York Zoological
Society. XVIII. On the Post-embryonic Development of Brachyuran Crabs of the Genus Ocypode. By Jocelyn Crane. (Text-figures 1-8) 65
7. New Species of British Guiana Heterocera. By W.
Schaus 83
8. A Papillary Cystic Disease Affecting the Barbels of Ameimus nebulosus (Le Sueur), Caused by the Myxo- sporidian Henneguya ameiurensis sp. nov. By R. F. Nigrelli & G. M. Smith. (Plates I- VIII; Text-fig-
ure 1) 89
9. Caudal Skeleton of Bermuda Shallow Water Fishes. IV. Order Cyprinodontes: Cyprinodontidae, Poecilidae.
By Gloria Hollister. (Text-figures 1-17) ... 97
10. The Histology of the Eye of the Cave Characin, Anop- tichthys. By E. B. Gresser & C. M. Breder, Jr. (Plates I-III) 113
ZOOLOGICA
SCIENTIFIC CONTRIBUTIONS OF THE
NEW YORK ZOOLOGICAL SOCIETY
1.
The Breeding Behavior of the Common Shiner, Notropis cornutus (Mitchill) .
Edward C. Raney
Zoology Department, Cornell University,
Ithaca, New York
(Plates I-IV; Text-figure 1).
Although several authors have written concerning the habits of the common shiner, Notropis cornutus (Mitchill), there is little detailed infor- mation available on the breeding behavior of this widely distributed cyprinid. The more pertinent papers giving data on spawning are those by Adams & Hankinson (1928), Greeley (1927 and 1929) and Hankinson (1932). The author has made an intensive study of the breeding habits of the common shiner, Notropis cornutus cornutus, during the past spring (1939) in the region about Ithaca, New York, and these data, together with observations made elsewhere mostly during the previous four years, are presented here. The breeding act has been observed many times at vari- ous places and under different conditions. However, the details were obscured, as they are in many fishes, because of the rapidity of consumation until photographs were taken under nearly ideal conditions in nature (see Plates I-IV). The method used in securing the photographs was similar to that used by some ornithologists. A nest of common shiners was located at a spot where the water was shallow and clear. If the shiners were at a high pitch in their spawning activity they would return after a period rang- ing from one-half to several hours and carry on normally with the camera mounted on a tripod and operated by an observer only 4 to 5 feet away. By utilizing exposures of 1/100 of a second or less most action could be stopped and the spawning process pictured. This method will probably be helpful in studying the habits of other nest building species among our fresh water fishes.
Migration.
An inshore migration in lakes and, at least in some localities, an up- stream movement is made by the common shiner. At the south end of Cayuga Lake, near Ithaca, New York, large schools are found moving toward
2 2 mo
2
Zoologica: New York Zoological Society
[XXV :1
shore during the first two weeks in May when the water has reached a temperature of 55° to 60° F. Their coming is eagerly awaited by bait dealers since, at this time, the common shiner is in demand as bait for the pike, Esox lucius. Nurnberger (1931, p. 215) has noted the presence of large schools of the common shiner appearing along the shore of Big Sandy Lake, Minnesota, on June 1, 1928, when the water temperature was 68° F. In the absence of actual reported observations of spawning in ponds and lakes it is assumed that the common shiner resorts to the gravel riffles of tributary streams, but Hubbs & Cooper (1936, p. 65) have pertinently stated that “its abundance in some of the inland lakes (Michigan) suggests that it may also spawn successfully over the gravel shoals of these lakes.”
In streams there appears to be at least a slight upstream migration from the deeper pools where the adults have wintered. On May 8, 1939, adult common shiners were found moving upstream into the trap of a weir placed across Grout Brook, the inlet of Skaneateles Lake, Cortland County, New York. Constant surveillance of the waters below the weir had failed to note any shiners before this date. The nearest satisfactory wintering place was 300 or more yards away and it is not at all improbable that they wintered in the lake 600 yards distant. Just how far individual shiners may migrate is, in the absence of tagging experiments, unknown.
Sexual Dimorphism.
Sexual dimorphism is pronounced in the common shiner, as in most other nest building or territory guarding cyprinids, as Semotilus, Cam- postoma and Nocomis. Breeding males of the common shiner attain a larger size, have well developed pearl organs or breeding tubercles and are more highly colored than females. It is obvious that the largest females in a stream exceed in size the smallest breeding males (Text-figure 1). Neverthe- less the difference in size is one of the best characters for distinguishing the sexes at a given nest and in many instances of spawning I have yet to see a male spawn with a larger female. It is difficult to generalize regarding the actual sex ratio but with adult shiners on the spawning ground invariably there are more females than males. The maximum size for male common shiners is about 8 inches in total length. A male Notropis cornutus chryso- cephalus in breeding color is figured in Forbes & Richardson (1920: oppo- site page 147) and a male Notropis cornutus frontalis in Greeley (1927: plate number 5).
The following color notes were made one hour after preservation in 10% formalin, from specimens of Notropis cornutus cornutus taken from over a nest in Salmon Creek at Ludlowville, New York, on June 6, 1939. The sides of the breeding males become suffused with red, the amount varying considerably with the individual. The posterior border of all fins tends to have a red band which in turn is edged by a much narrower clear area. There is a pronounced tendency for the pectoral and ventral fins to become suffused throughout with red except for a light posterior border and a milky white anterior border. The rays of the dorsal, caudal and pectoral fins tend to become dusky just anterior to the red band while the base of the anal fin usually remains milky white. The base of the ventral fins shows a slight tendency to become dark in some individuals. The branchiostegals are red and the lower part of the cheek and opercles are at times a dull red. The upper part of the opercles and cheek are a slate blue as well as a narrow band immediately behind the opercle. There is a longitudinal mid-dorsal streak and a streak on each side of the back which vary from a silver to a greenish color but at all times are highlights on a dark background which make excellent field recognition characters (see Plate II).
Breeding females are dusky above and silvery along the sides. In a few cases a slight amount of red has been seen, especially at the posterior border of the caudal and anal fins. The rays of the dorsal, pectoral and caudal fins
1940]
Raney: Breeding Behavior of Notropis cornutus
3
tend to become dusky while the ventral and anal fins are usually white. While females have a faint longitudinal streak down the mid-line of the back and one along each side of the back they are not as pronounced as in the males. The swollen abdominal region of the ripe or near ripe female is also a helpful field character.
The males begin to show color about a month before spawning, a sub- dued red tint appearing first on the lower fins. The red which is to be seen along the sides in the ripe males develops last and appears only a few days before spawning takes place. After breeding the color fades quickly and is usually gone in two weeks, the red on the sides being the first to go. How- ever, an occasional male is taken even in late summer with a slight amount of color in the fins.
The breeding male has well developed pearl organs scattered on the back in front of the dorsal fin, on the nape, on the top of head and on the snout; a few under the anterior part of the eye, in the region of the pre- maxillaries and maxillaries and a single row along the lower jaw. The tubercles on the head are better developed than the more posterior ones. A few minute pearl organs occur on the upper sides, below and in front of the dorsal fin; on the posterior border of the opercle; on a few of the an- terior scales just dorsad of the base of the pectoral and on the belly between the pectoral and ventral fins. An irregular row of fairly large pearl organs is to be seen on the anterior edge of the first ray of the dorsal fin and smaller ones occur on the sides of the first few dorsal rays. The dorsal side of the pectoral fins are well supplied with fairly large tubercles which undoubtedly are of use in enabling the male to hold the female in spawning. Minute tubercles occur on the ventral side of the pectoral and on both sides of the ventral fins near the anterior border. A careful examination of a dozen ripe females disclosed no tubercles. However, Fowler (1912, p. 473) reports that “occasionally a tuberculate female is found.”
As with the color, the tubercles of the male grow gradually, starting about a month before spawning, and are usually gone within two weeks after spawning, although the scars of the larger pearl organs may oftem be seen up until one month after breeding.
Functions of Breeding Tubercles.
The well developed breeding tubercles on the snout, chin, head and ori the back, in front of the dorsal fin, of the males are of unquestionable value in the fighting which takes place in attempting to hold a territory and in driving away predators which attempt to eat the eggs. These tubercles are also a protection to the male in the sporadic digging activities in which, they engage. Probably the only tubercles which are of value in the spawn- ing act are those on the dorsal side of the pectoral fin which functions as a grasping organ and those on the ventral side between the pectoral and ventral fins which come in direct contact with the side of a female (Plate II).
Time of Spawning.
In Big Sandy Lake, Minnesota, Nurnberger (1931, p. 215) reports that spawning occurred in 1929 between June 24 and 29 when the water reached 66°-70° F. In Illinois, according to Forbes & Richardson (1920, p. 148) spawning occurs from May 1 to the last of June. Hubbs & Cooper (1936, p. 65) give the spawning season in Michigan as extending from the latter part of May into June and mention that “spawning probably rarely occurs at water temperatures lower than 60° to 65° F.” In western Pennsylvania spawning takes place in May and June while in New York the season may extend from the last two weeks in May until the middle of July in some of the colder Adirondack streams. About Philadelphia, Pennsylvania, Fowler
4
Zoologica: New York Zoological Society
[XXV :1
(1909, p. 540) reports that the shiner spawns from late April to early summer. Tracy (1910, p. 68) gives spring and early summer as the season in Rhode Island. In the Connecticut Lakes in Maine Kendall & Goldsborough (1908, p. 31) give the breeding time as spring or early summer, depending on the temperature of the water.
The lowest water temperature at which I have actually observed shiners spawn is 64° F. but they very likely begin to spawn at a temperature of 60 F. or thereabouts. However, in certain cold streams spawning probably occurs at a slightly lower temperature. Spawning has personally been ob- served when the water temperature was as high as 78° F. and Greeley (1929, p. 172) saw Notropis cornutus frontalis spawning in Silver Creek, a tribu- tary of Lake Erie, Chautauqua County, New York, as late as July 9 at a water temperature of 83° F. At any one locality spawning is usually over in ten days although ripe males are often seen for longer periods.
No breeding was observed taking place at night and at this time few adult fishes remain in close proximity to the nest. Most of them retire to the deeper pools nearby where they are relatively inactive. The spawning act was seen as early as 9 A.M. on May 27, 1939, in Willseyville Creek at Willseyville, Tioga County, New York. Spawning reached a high point about 12 noon and gradually tapered off after 4 P.M., none being noted later in the day than 6 P.M. Spawning had reached a peak of activity at this par- ticular locality on May 27. Earlier and later in the season there was but little breeding activity at this place except in the afternoon.
Spawning Site.
A very important factor in explaining the wide distribution and the local abundance of the common shiner is its capacity to spawn successfully in many different habitats. Shiners may (1) spawn over gravel beds in running water, (2) excavate small depressions in gravel or sand in running water or, (3) utilize the nests built by other species (Plate I), even when these nests occur in pools as is frequently the case ivith Exoglossum. They appear to prefer the nests of other species as nesting sites when they are available, although in 100 yards of Catatonk Creek, at Candor, New York, on May 31, 1939, shiners were observed utilizing all three of the above-men- tioned nesting places. Their preference for Exoglossum nests, especially in small streams, often finds them spawning over these nests in the relatively quiet waters of shalloAV pools rather than using the gravel riffles which are so common nearby. The depth of the water over a spawning site is usually less than 8 inches except when nests of other fishes are used. Four large male shiners were once observed spawning over an Exoglossum nest situated at the edge of a deep pool in 2 feet of water.
Gravel Beds as a Nesting Site.
Shiners often utilize gravel in or at the head of riffles as a spawning site. These areas can hardly be considered nests although they can easily be distinguished by the cleared nature of the bottom. The clearing away of the silt is accomplished in two ways; first, by the constant moving of the male fish over the gravel during which the lower fins make some contact with the bottom, and, second, by the occasional sporadic digging motions practiced by the male. He inserts his head between two pebbles and, by a quick sidewise motion, dislodges one or both of them. Only rarely do they pick up small stones in their mouths.
Fowler (1909, p. 540) has observed breeding schools of several hundred brilliantly colored males packed close together over clear gravel or sandy shallows in the region about Philadelphia, Pennsylvania. Several splendid opportunities have been had recently for observing spawning (subspecies cornutus ) over gravel in Fall Creek, Ithaca, New York, and in Salmon Creek,
1940]
Raney: Breeding Behavior of Notropis cornutus
5
at Ludlowville, New York. In western Pennsylvania Notropis cornutus chrysocephalus has been observed spawning over gravel in riffles. Of the above-mentioned localities when gravel is utilized many more shiners spawn together over a given area than when the nests of other species are used. At Ludlowville on June 6, 1938, from 80 to 100 males (Text-figure 1) had taken a position in a riffle over a gravel bottom where the water varied in depth from three to five inches. Under such conditions each male attempts to hold a small territory which, of course, is quite limited because of the large number of males in such a restricted area and, at most, consists of a distance of three inches or so to either side. Much fighting occurs among the males but after one retreats from the rush of another male he usually comes back to his original position and the rushes may be repeated many times between spawnings. The offensive activity of a male consists in quickly
50 60 70 80 00 100 110 120 130
STANDARD LENGTH IN MM.
Text-figure 1.
Frequency distributions of breeding males and females of Notropis cornutus cornutus taken from over one spawning area in Salmon Creek, at Ludlowville, Tompkins County, New York, on June 6 and 9, 1939. Measurements of standard length were taken on 130 males and 232 females preserved in alcohol.
6
Zoologica: New York Zoological Society
[XXV :1
circling either to right or left and attempting to strike with the tubercled snout. Few direct hits are made as opposing males are usually alert and beat a quick retreat at the moment an antagonist starts circling. The water ■“boils” over such groups of males as they break water with their dorsal and caudal fins during this fighting and maneuvering for position. Occasionally “deferred combat” (Reighard, 1910, p. 1129) occurs as two males of nearly the same size (Plate III) tensely swim off parallel to each other, often for a distance of 3 to 5 feet. If they are of equal size they soon return to or close to their original position on the gravel. There was a tendency at times for this large breeding group to break into two, about half the males moving upstream several feet and there repeating the same combat for territory.
Approximately 150 females could be seen on the gravel below the hold- ings of the males. When they were ready to breed the females would come upstream and take a position over a male and spawning would take place in a manner to be described later. This large spawning group was somewhat unusual in that it was located in a shaded spot under a huge American elm Wee and received no direct sun during the afternoon. In such a situation where spawning occurs over coarse gravel the eggs may be seen and re- covered by picking up or scuffing the pebbles. Some of the eggs are free between pebbles, singly or in groups, while others are found adhering to the stones.
At times, especially toward the end of the spawning season, a large depression is made on the bottom by a large group of spawning males. Whether a nest-like excavation is apparent depends somewhat on the type of bottom. At Ludlowville the pebbles were fairly large, averaging about one inch in diameter with few under one-half inch in diameter. Under con- ditions such as this the digging activities of the males have little effect other than to give a clean look to the area. In lower Fall Creek, at Ithaca, New York, a similar breeding group will excavate a shallow depression in the bottom consisting of sand and small gravel, which is often 3 to 4 feet in diameter. Such nests are hardly distinguishable from the type which is discussed next.
Nests Excavated by the Common Shiner.
Both subspecies cornutus and frontalis have been observed digging and guarding their own circular nests, which are usually 8 to 12 inches in diameter. In most cases a pit partly or wholly excavated by Semotilus (Plate I) or Campostoma is taken over by a number of shiners, the number varying usually from 3 to 20 males although occasionally but one male oc- cupies a nest. These nests are most often built at the head of a riffle but are also often seen in fairly fast flowing riffles. They are probably never built in quiet water. The males make the nests by digging activities in which they insert their heads between two stones and by jerking the head to the side dislodge the stones laterally. Several males will engage in this digging at one time although there are frequent interruptions as the males dash at each other, vieing for the position furthest upstream in the nest. This digging usually occurs irregularly between spawning acts.
Utilization of Nests of Other Fishes.
Several authors including Greeley (1929, p. 172), Hankinson (1920, p. 8, and 1932, p. 415), Hubbs & Cooper (1936, p. 65) and Van Duzer (1939, p. 73) have pointed out instances of the common shiner utilizing the nests of other species, such as Nocomis micropogon, Nocomis biguttatus, Semo- tilus a. atromaculatus , Leucosomus corporalis, Exoglossum maxillingua and Campostoma anomalum. Shiners appear to prefer the pebble nests of Nocomis and Exoglossum when available. Their second choice under most
1940]
Raney: Breeding Behavior of Notropis cornutus
7
conditions appears to be the pit and ridge nest dug by Semotilus and the author has observed them taking over and enlarging the circular pits exca- vated by male Campostoma anomalum pullum. The breeding pattern is the same, however, whether the nest of another species is utilized or a nest of its own is excavated.
On several occasions Semotilus and Nocomis have been working on their nests at the same time that shiners were spawning there. The male shiners seldom attacked a male Semotilus and the Semotilus in turn continued re- moving stones from his pit and piling them in a ridge upstream (see Reighard, 1910, p. 1125). At times he would turn on and strike with his tuberculate head a shiner that became too bold. As Hankinson (1932, p. 418) has pointed out a male Nocomis will frequently continue bringing stones to add to his ever-growing nest even though many shiners are present and spawning is occurring. An amazing example of the complacency of a male Nocomis micropogon was noted in Catatonk Creek at Candor, New York, on May 31, 1939. Twenty male and thirty female cornutus were swarming over a Nocomis nest. These were accompanied by a dozen breed- ing Notropis rubellus. A number of specimens of Exoglossum maxillingua and Rhinichthys cataractae were darting in from their position on the periphery to eat what eggs they could secure. During this great activity which made the water “boil” over the nest, the male Nocomis calmly added stones to his nest for a period slightly more than one hour without at any time attacking any of the breeding fishes that were so close. At the same time the male shiners carefully respected the Nocomis. However, female Semotilus or Nocomis seldom appear when shiners swarm the nests, and actual spawning has not been seen under these conditions.
With spawning Exoglossum, as Van Duzer (1939, p. 73) has pointed out, “the mating of the cut-lips was always definitely lessened and sometimes stopped by their ( cornutus ) presence and activity at the nest.” It is quite probable that the male cornutus have learned to respect the sizeable male Semotilus and Nocomis, armed as they are with well developed pearl organs on the head, while they have become equally familiar with the abortive rushes of the non-tuberculate and usually smaller male Exoglossum.
An interesting side light is the considerable number of natural hybrids which occur largely as a result of this habit of spawning over the nests of other species. Two species have often been seen spawning at the same time over a Nocomis nest.
Hybrids of the combination Notropis cornutus X Notropis rubellus are frequently collected and both of these species have been seen spawning over a Nocomis nest at the same time (see Hankinson, 1932, p. 417). The hybrid Notropis cornutus X Nocomis micropogon has been recorded by Greeley (1938, p. 51). Clinostomus elongatus spawns over the nests of both Notropis cornutus and Semotilus atromaculatus atromaculatus and both Greene (1935, p. 89) and Greeley (1938, p. 51-52) have recorded the hybrids Notropis cornutus X Clinostomus elongatus and Semotilus atromaculatus atromacul- atus X Clinostomus elongatus. Greeley (1938, p. 52) has also reported hybrids Notropis cornutus X Semotilus atromaculatus atromaculatus and Notropis cornutus X Leucosomus corporalis which likely result from cornutus spawning over the nests of these species. Thompson (1935, p. 492) has mentioned that in Illinois hybrids of Notropis cornutus are com- mon presumably because cornutus spawns in the nests of other species.
Sex Recognition.
The female that successfully spawns, invariably approaches the nest from the downstream side and assumes a position dorsad and slightly down- stream from the male, who faces upstream (Plate I). Should she continue upstream to a position in front of a male, as occasionally happens, she would be driven away by the rush of a male, usually the male holding the dominant
8
Zoologica: New York Zoological Society
[XXV :1
position in the nest. The male would usually dart at and cause to retreat downstream any female getting into the nest at his body level, that is, near the bottom of the nest. In case a female was chased she would beat a retreat downstream, assume a position from one to several feet below the nest and then gradually come upstream again over a male. The constant turning of his head from side to side enables him to spot any fish which comes above him. The behavior of a male when a female has assumed a position above him is evidence that he recognizes her by the position she takes. The male at this time will, while facing upstream with body nearly straight, incline his body from side to side so that his reddish sides flash alternately (Plate I). While the male may react to the mere presence of a fish body above and behind him, there is a possibility that the male is attracted by the milky white area about the slightly protruding external opening of the oviduct of the female. In the case of one ripe female, but two inches long, this white area was so pronounced it might easily be seen by an observer standing five feet away. It must surely be an attractive mark when viewed from the position of the male. A female will take a position above any male who is in or slightly below the nest. Often a small male on the outside edge of the nest (Plate II) will be successful in spawning more often than the large male dominating the nest since the latter is forced to spend a large part of his time attempting to chase away other fishes. However, one excep- tionally large and vigorous male was once seen holding a Semotilus pit so effectively by his wild dashes at intruding male cornutus that he successfully spawned with a half-dozen females in about one and one-half minutes before any other males seriously challenged and finally displaced him.
There is one bit of evidence that female cornutus react to fish of larger size whether of the same species or not. At one instance a male Exoglossum returned to his nest, and took a position facing upstream. He moved slightly to right and left over the nest much as does a male cornutus, and several female cornutus took a position above and dipped down to take a spawning position beside him. The male Exoglossum reacted by turning and rushing at them and finally chased the female shiners away from the nest.
Spawning Act.
With a female above a male and the male alternately inclining himself to a semi-recumbent position first to one side then to the other, conditions are set for the spawning act (Plate I). The female dips downward beside that side of a male which is inclined toward the bottom or making the more acute angle. The anterior end of the head of the female is just beyond the anterior border of the pectoral fin of the male. The pectoral fin of. the male is inserted under or slightly behind the head of the female, the distal end of his pectoral being curved upward around the female. At the same time his caudal peduncle is swung up over the caudal peduncle of the female and then the caudal peduncle of the male moves downward. As a result of this pressure the female is lying on her side at right angles to the male with her head toward shore and usually with her ventral side upstream (Plate II).
The male continues to bend his body in such a way as to enclose the female within a curve (Plate III). It appears that the eggs are laid at this moment, the eggs being forced out by the downward pressure exerted by the male. The female then straightens out (Plate IV) and shoots head upwards to, or almost to, the surface just as the male begins to straighten himself. When observing a large group of spawning shiners, females may constantly be seen breaking water as they come out of the spawning em- brace. The speed with which this spawning act takes place is a fraction of a second or just slightly longer than the time necessary to snap ones fingers quickly, and the details are about impossible to see clearly with the unaided eye, even when observed at close range. The spawning act is much like that
1940] Raney: Breeding Behavior of Notropis cornutus 9
of Semotilus atromaculatus atromaculatus as described by Reighard (1910, p. 1130).
The male shiner immediately takes his position facing upstream and may spawn again with another female in a few seconds. The female swims slowly downstream a few feet, takes a position below the nest and soon comes back to spawn again, at times with the same but usually with a different male. One female was observed to spawn twice with the same male in a period of one minute. A small male about three inches in total length who at the time was the lone male on the nest, was counted spawning 20 times within 10 minutes. There were about 30 females below and over this nest. Although a male will spawn with the female to either side, many more righthanded matings were noted. An exceptionally vigorous male may hold the dominant position in the nest, and spawn many times, for as long as 20 minutes. Most males, however, drop downstream and rest in less than half this time; thus the spawning individuals over a nest are changing constantly.
Male shiners recognize the spawning act when they observe another male spawn. When a breeding group has been purposely frightened away from a nest it is usually one of the smaller male shiners that returns to the nest first, drives away the egg predators and spawns with the first female to come upstream. On the completion of the spawning act from 5 to 10 males would immediately rush to the area and the never-ending battles for the leading places in the nest would resume.
One group of breeding shiners had become conditioned so that when a daily train rushed by at high speed only 20 yards away, spawning con- tinued without interruption. A horned dace that was present in the nest at the same time, however, beat a hasty retreat toward the deep pool above. These shiners were easily frightened by any quick movement such as the passage of a bird close above the water or the movement over the surface of a ripple caused by a sudden gust of wind. However, they did not move when loud vocal notes were made within six feet of them.
A common water snake, Natrix sipedon sipedon, approximately two feet in length, once quietly moved downstream into a nest. The snake stopped with its head near the surface and its body dragging on the bottom six inches below. Neither the spawning shiners nor other fishes such as Rhin- ichthys a. atratulus and Campostoma anomalum pullum appeared at all disturbed. Several shiners swam leisurely up and casually examined the intruder and one even investigated by touching the snake’s body with its mouth as if testing its edibility.
An Exoglossum nest that was used by a spawning male shiner was completely covered by the author with large flat stones. The male shiner returned shortly and made a few dashes at adult Rhinichthys atratulus atratulus that were eating what eggs they could get in the crevices. The male shiner remained in the immediate vicinity for three minutes and then dropped downstream twenty feet and attempted to drive away another male shiner with a holding over an Exoglossum nest.
No exact observations ai'e available but it is thought that but few eggs are laid at one time, probably not more than fifty. Occasionally a very ripe female will lose her entire supply of eggs when she is handled and it is thus probable that at times many more eggs are laid than at others. As has been mentioned above the eggs fall to the bottom and lodge on and among the pebbles of the nest. Apparently few are washed downstream. The flow of water over a depression such as cornutus may make, or over the pit of a Semotilus nest, is such that eddys are formed, bringing the water in the bottom of the pit to a relatively quiet condition even in fairly fast riffles. Thus very few eggs are washed downstream. When Exoglossum and Nocomis nests are used the eggs are usually driven into the anterior slope of the nest by the current.
10
Zoologica: New York Zoological Society
[XXV :1
Other species of fishes, chiefly darters and other minnows, are quick to dart in to eat the eggs after a spawning act is consumated. Indeed some Rhinichthys atratulus atratulus seem to be conditioned so that they make a mad dash to the spot at the moment spawning occurs. The male shiners are usually successful in keeping these predators at bay.
Although some species of fishes desist from taking food during spawn- ing, both male and female Notropis cornutus cornutus will eat during this period. Ball (as reported by Marshall, 1939, p. 153) “found no cessation of feeding during the breeding season” in Notropis cornutus chrysocephalus. Ripe females lying in the middle layers of water just below the nest will readily rise to the surface to take insects which are floating by. Males, when not too much occupied with their territory guarding or spawning, will rise to eat these insects.
After the spawning period many of the larger males are found dead. Invariably they are badly fungused. However, most have probably reached the end of the normal life span of the species at this time. Numbers of dead specimens of other species of minnows such as Campostoma and Nocomis are often seen just after the spawning season is over.
Eggs and Young.
The demersal eggs become adhesive when water-hardened about two minutes after they are laid. At first they are orange but fade considerably in a short time. The average diameter is 1.5 mm. The eggs drop between and are washed under pebbles and some are subsequently buried by the digging activity of the males to a depth of several millimeters in the sand. Some single eggs are found and numbers up to twenty have also been found in loose clumps.
The incubation period is unknown, as every effort to hatch eggs in aquaria resulted in failure. Little definite data is available on the behavior of the newly hatched larvae, but later, after the loss of the yolk sac, the post-larvae may be seen in small schools at and slightly under the surface. At this time they often school with the young of other species as Catostomus commersonnii commersonnii, Hybognathus regius and Notropis hudsonius hudsonius in the sluggish waters about Ithaca, New York. In smaller streams in the same region the young cornutus are to be found in the com- pany of the young of Rhinichthys atratulus atratulus, Clinostomus elongatus and Catostomus commersonnii commersonnii and other species. After they have reached a length of around 15 mm. they are still to be found in small schools but now they are often seen in fast water as well as in pools.
Throughout the summer the young of the year tend to refrain from associating with juveniles and adults of the same species although by Sep- tember a few mixed collections of young and small adults have been taken. The schools of juveniles and adults, with but few exceptions, resort to the deeper pools during winter. However, the young are often to be found in great numbers, associated with the young of other species, _ in shallow, ice covered backwaters along streams. For a detailed description of the eggs and young of Notropis cornutus chrysocephalus, the reader is referred to Fish (1932, p. 339).
Acknowledgment is due James Kezer who read the manuscript critically and to Ernest A. Lachner and Dwight A. Webster for their assistance in the field.
Summary.
1. In May an inshore migration in lakes and, in some cases at least, an upstream movement is made by Notropis cornutus.
2. Sexual dimorphism is pronounced. Breeding males have well devel-
1940]
Raney: Breeding Behavior of Notropis cornutus
11
oped pearl organs or breeding tubercles, are highly colored and reach a larger size than breeding females. The breeding tubercles on the various parts of the body are of value in the fighting which occurs among the males, in driving away predators and in the breeding act.
3. Spawning occurs from May 1 through the middle of July in the north- eastern states, beginning usually when the water has reached a temperature of 60° to 65° F. At any one locality spawning lasts about ten days and is limited to the daylight hours.
4. The common shiner may (1) spawn over gravel beds in running water, (2) excavate small depressions in gravel or sand in running water or, (3) utilize the nests built by other species such as Nocomis micropogon, Nocomis biguttatus, Leucosomus corporalis, Semotilus atromaculatus atro- maculatus, Exoglossum maxillingua and Campostoma anomalum whether these are built in running or the still waters of shallow pools. They prefer to spawn over the nests of other species when these are available.
5. Hybridization often occurs with Notropis cornutus as one parent largely as a result of their spawning over the nests of other fishes.
6. The number of males that will spawn over one nest varies con- siderably, there being from one hundred or more over a gravel bed to as few as one male over a small depression. More females are present at the nests than males.
7. Males fight continually for the leading position, the position furthest upstream in the nest.
8. A male shiner recognizes a female and takes a semi-recumbent posi- tion alternately from one side to the other when the female approaches from the downstream side of the nest and comes to a position above and slightly downstream from the male.
9. The female reacts to a fish whether it be a male common shiner or another species providing he has taken a position over a nest, head facing upstream and moving slightly from side to side.
10. The female takes the initiative in the breeding act by dipping down- ward, to lie beside the male. The male throws his caudal peduncle over that of the female, curves his body by bringing his head and tail in close proximity with his pectoral fin underneath the head of the female. The eggs are forced from the female at this moment while she lies, usually on her side, her ventral surface facing upstream and her head pointing toward shore. The entire breeding act is over in a fraction of a second and the details cannot be clearly seen with the unaided eye.
11. Probably fewer than fifty eggs are laid at once. The demersal eggs become adhesive after water-hardening in about two minutes after being laid and drop between the pebbles on the bottom of the nest to which they adhere. When first laid they are orange and average 1.5 mm. in diameter.
Bibliography.
Adams, Charles C. & Hankinson, T. L.
1928. The ecology and economics of Oneida Lake Fish. Bull. New York State College of Forestry, Syracuse University, Vol. 1, No. 4a ( Roose- velt Wild Life Annals, Vol. 1, Nos. 3 and 4) : 235-548.
Fish, Marie Poland
1932. Contributions to the early life histories of sixty-two species of fishes from Lake Erie and its tributary waters. Bull. U. S. Bur. of Fisheries, Vol. 47: 293-398.
Forbes, Stephen Alfred & Richardson, Robert Earl
1920. The fishes of Illinois. Second Edition. Nat. Hist. Surv. 111., Vol. 3: pp. 1-cxxxi; 1-357.
12
Zoologica: New York Zoological Society
[XXV :1
Fowler, H. W.
1909. A synopsis of the Cyprinidae of Pennsylvania. Proc. Acad, of Nat. Sci. of Phil, for 1908 (1909) : 517-553.
1912. Some features of ornamentation in fresh water fishes. Am. Naturalist, Vol. 46: 470-476.
Greeley, J. R.
1927. Fishes of the Genesee Region with Annotated List. In: A Biological Survey of the Genesee River System. Suppl. 1 6th Ann. Rept. N. Y. Cons. Dept., 1926: 47-66.
1929. Fishes of the Erie-Niagara Watershed. In: A Biological Survey of Erie-Niagara System. Suppl. 18 th Ann. Rept. N. Y. Cons. Dept., 1928: 150-179.
1931. Fishes of the Area with Annotated List. In: A Biological Survey of the St. Lawrence Watershed. Suppl. 20 th Ann. Rept. N. Y. Cons. Dept., 1930: 44-94.
1932. Fishes of the Area with Annotated List. In: A Biological Survey of the Oswegatchie and Black River Systems. Suppl. '21st Ann. Rept. N. Y. Cons. Dept., 1931: 54-92.
1938. Fishes of the Area with Annotated List. In: A Biological Survey of the Allegheny and Chemung Watersheds. Suppl. 21th Ann. Rept. N. Y. Cons. Dept., 1937 : 48-73.
Greene, C. Willard
1935. The distribution of Wisconsin fishes. State of Wisconsin, Conserva- tion Commission, Madison: 1-235.
Hankinson, T. L.
1908. A biological survey of Walnut Lake, Michigan. Mich. State Board Geol. Surv., Rept. for 1907 (1908) : 157-288.
1920. Report on investigations of the fish of the Galien River, Berrien County, Michigan. Occ. Papers Univ. of Mich., Mus. Zoology, No. 89: 1-14.
1932. Observations on the breeding behavior and habits of fishes in southern Michigan. Papers Mich. Acad. Sci., Arts and Letters, Vol. 14, 1931 (1932) : 411-425.
Hubbs, Carl L. & Cooper, Gerald P.
1936. Minnows of Michigan. Bull, of Cranbrook Inst, of Science, No. 8: 1-95.
Kendall, W. C. & Goldsborough, E. L.
1908. The fishes of the Connecticut Lakes and neighboring waters with notes on the plankton environment. U. S. Bur. Fisheries, Doc. 633: 1-77.
Marshall, Nelson
1939. Annulus formation in scales of the common shiner, Notropis cornutus chrysocephalus (Rafinesque). Copeia, (3): 148-154.
Nurnberger, Patience Kidd
1931. Observations on the spawning temperature of Luxilus cornutus. Trans. Am. Fish. Soc., Vol. 61: 215.
Reighard, J. E.
1910. Methods of studying the habits of fishes with an account of the breeding habits of the horned dace. U. S. Bur. Fisheries Bull., Vol. 28, for 1908 (1910) : 1111-1136.
1915. An ecological reconnaissance of the fishes of Douglas Lake, Cheboygan County, Michigan, in Midsummer. U. S. Bur. Fisheries Bull., Vol. 33: 215-249.
Thompson, David H.
1935. Hybridization and racial differentiation among Illinois fishes. Natural Hist. Surv. 111. in Appendix. Vol. 20: 492-494.
1940]
Raney: Breeding Behavior of Notropis cornutus
13
Tracy, H. C.
1910. Annotated list of the fishes known to inhabit the waters of Rhode Island. Fortieth Ann. Rep. Comm, of Inland Fisheries of Rhode Island: 35-176.
Van Duzer, Evelyn M.
1939. Observations on the breeding habits of the cut-lips minnow, Exoglos- sum maxillingua. Copeia, (2) : 67-75.
14
Zoologica: Neiv York Zoological Society
EXPLANATION OF THE PLATES.
Plate I.
A typical breeding group of Notropis cornutus cornutus over a nest begun by a male Semotilus atromaculatus atromaculatus at Willseyville, New York. The large male at the right, occupying the leading position in the nest, is as- suming a semi-recumbent position in anticipation of spawning with the small female now dipping toward his left side. He is followed by eight males which are easily recognized by the prominent light colored streak along the upper side. Three small females are to be seen to the left and one is downstream from this group of eight males. Another female appears above the third from the last male in the nest. Several other females, that do not appear in the photograph, were scattered further downstream. All fishes are facing upstream.
Plate II.
The beginning of a spawning act in which a small male cornutus, to the right of the larger leading male in the nest, has thrown his body over that of the female who now lies on her side. The breeding male is just starting to bring his caudal peduncle toward the head in the typical breeding curve. The right pectoral fin of the male is under the anterior end of the female. This much of the spawning act had happened so rapidly that the other three males to his left have apparently not noted it. The upstream male will probably turn on the spawning male and drive him off when he finally sees him. The open mouth is typical of a spawning male. Two female cornutus are to the left and one is behind the male on the left side of the spawning group. Two males on the down- stream side of the nest appear at the extreme left of the figure. Note the adult Rhinichthys atratulus atratulus in the right foreground on the periphery of the nest in a search for eggs.
Plate III.
The climax of the breeding act has been reached with the male cornutus in the foreground curved over the female. Although the pectoral fin of the male is usually under the female it is above in this case and the position of the female is somewhat abnormal. The eggs are forced out of the female at this point. Another female may be seen above and to the left of the leading male in the nest. The two males near the center of the figure are in a position males assume before attempting to dash at each other. Note that their heads and caudal fins are apart while their bodies almost touch in the middle. The light colored tubercles of the males stand out against the dark background of the skin. Three males may be seen at the extreme left of the figure where the water breaks over the riffle. An adult Rhvnichthys atratulus atratulus is present in the right foreground.
Plate IV.
The breeding act has been completed. The male cornutus is straightening his body and the female has started to move upward. The large dominating male to his right is turning to rush at the male that has just spawned. Three other males are near the middle of the nest and four more appear at the left of the figure. One small female is present just in back of the largest male near the center of the photograph. Two other females appear in the upper right hand corner. An adult Rhinichthys atratulus atratulus is present in the foreground.
RANEY.
PLATE I.
THE BREEDING BEHAVIOR OF THE COMMON SHINER, NOTROPIS CORNUTUS (mITCHILl).
RANEY.
PLATE II.
THE BREEDING BEHAVIOR OF THE COMMON SHINER, NOTROPIS CORNUTUS (mITCHILl)
RANEY.
PLATE III.
THE BREEDING BEHAVIOR OF THE COMMON SHINER, NOTROPIS CORNUTUS (MITCHILL
RANEY.
PLATE IV.
THE BREEDING BEHAVIOR OF THE COMMON SHINER, NOTROPIS CORNUTUS (mITCHILl).
Ginsburg : Divergence and Probability in Taxonomy
15
2.
Divergence and Probability in Taxonomy.
Isaac Ginsburg
U. S. Bureau of Fisheries, Washington, D. C.
Taxonomists of past generations have generally been content with de- scribing and establishing species based on one or but a very few specimens. The business of distinguishing species by this method is a comparatively easy matter. Using very few specimens as a basis of comparison, closely related species, in their large majority, appear to be sharply differentiated. In occasional instances a sharp distinction on the basis of even a few speci- mens proved troublesome, and such specimens were generally assumed to represent “varieties,” “races,” etc., of the same species.
This easy method proved to be inadequate, as it was bound to. Later investigators found that such distinctions all too frequently did not accord with their material. This is due to the fact that related species often ap- proach closely or even intergrade in their differentiating characters. From a taxonomic point of view we do not know all we should about a species until we know its range and manner of variability, at least in the few crucial characters by which it is distinguished from closely related ones. This is, of course, true of races and other subdivisions of the species. Tax- onomists come more and more to realize this and act accordingly. In dealing with mass data obtained in the study of variability, it is desirable to reduce them, when it is consistent to do so, to single figures, statistical constants, for convenience in comparison, discussion and interpretation. This paper considers one such class of constants, that concerned with the measure of divergence, which is of the utmost importance to taxonomists, as related to another class, that concerned with the expression of probability, which is often used in place of the first.
Probability in its numerical expression is often referred to as the “test of significance.” As it is my hope that this paper will prove to be of interest to taxonomists to whom the latter term is not a household phrase, it may be well to consider briefly here its precise meaning.
When a taxonomist compares the likenesses of and the differences be- tween two closely related populations — be they species, subspecies, races, etc. — he does not study the variability of the entire population, but his comparison is based on a relatively restricted number of specimens ; in other words, on two samples drawn one each from the two populations. The degree of difference or divergence shown by the two samples determines his conclusion regarding the taxonomic rank of the two populations, whether they are to be considered as species or as belonging to a category of the next or second next lower rank. However, we know that different samples drawn even from the same variable population will generally not be the same, but, on the contrary, due solely to chance, will exhibit dif- ferences of greater or lesser degree. The question then comes up, is the difference shown by the two samples compared in taxonomic research a real population difference, or is it due to the fortuities of sampling? It may be assumed that due solely to chance it may happen sometimes that two
16
Zoologica: New York Zoological Society
[XXV :2
samples drawn even from the same population will exhibit a difference as great or greater than that between the two samples of the two populations compared, and in that case it would, of course, be inappropriate to separate the two samples taxonomically. An answer to the above question, in part, as discussed below, is given by certain mathematical formulae developed in connection with the theory of probability. By the use of such formulae — based on the difference between the averages of the two samples, the squares of the deviations of the individual specimens from the averages, and the number of specimens in the samples — it may be determined, for any one given character, how often such a difference, or a greater difference, is likely to be obtained at random, by mere chance, from two samples of the same population. If such an eventuality is likely to occur but rarely, say, two times or less per 100 trials, we may state that the difference is “significant” and that it may be concluded with comparative assurance that the two samples compared in our taxonomic research belong to two distinct populations.
The above is a bare outline of the “test of significance,” but it is hoped that it presents its essential idea. The “test of significance” then results in a number that represents the numerical expression of probability, indi- cating the probable value of a difference determined in biologic research, or the probable reliability of the two samples compared, for the purpose of drawing pertinent conclusions. In the practical application of the formulae an arbitrary limit is postulated and a figure obtained as a result of the test of significance, which equals or is higher than the accepted limit, is taken to denote “significance.” It should be noted in particular that this test merely establishes that a determined difference is “significant.” It does not indicate definitely whether the difference is of specific, subspecific or racial magnitude. The taxonomic rank of the two populations compared is determinable definitely only by some appropriate measure of divergence.
It is very important and can not be too strongly emphasized that it is necessary to draw a sharp distinction of the fundamental difference between the two concepts, measure of divergence and expression of probability, from both a theoretical and a practical standpoint. This fundamental idea has been formulated by Fisher (1936, p. 59) as follows: “It must be stressed that the test of significance calculates a probability; it does not calculate a racial difference.” Although it is, or should be, generally realized that a test of probability is not the same as a measure of divergence, yet, some- how, the two concepts become inextricably mixed in deliberation and discus- sion. Somehow or other there appears to be a lingering idea with some biologists that the greater the numerical value of the figures showing “sig- nificance” obtained by the use of current formulae that express probability, the greater the divergence between the pair of populations compared. Often this is true; but it is only a partial truth, and like all partial truths it is bound to lead us sadly astray. This confusion of concepts appears to be a stumbling block not only with biologists who are not given much to the employment of statistical formulae, but even with some who employ them extensively.
If two separate comparisons be made of two pairs of populations, and the test of significance have a much greater numerical value as between one pair of populations than between the other, it does not always mean that the former pair diverges to a greater extent; although in many cases a greater numerical value for the test of significance does coincide with a greater de- gree of divergence. The real meaning is that for the comparison showing a greater numerical value, one or both samples are too large for our purpose, for that particular pair of populations with their spread and relative regu- larity of the frequency distributions and their difference between the means. Smaller samples would have been sufficient to prove what we set out to dis- cover, if our purpose was the determination of the probable mathematical significance of the difference between the means. More specifically, when
1940]
Ginsburg: Divergence and, Probability in Taxonomy
17
two values of significance obtained in two comparisons are above its accepted limit, but differ widely— say, one is 10, the other 50 — the meaning is that the samples in the latter comparison are larger than necessary for the pur- pose of determining significance. It is evident, therefore, that figures of different magnitudes which express mathematical probability or signifi- cance, cannot consistently be employed for the purpose of expressing rela- tive divergence.
A notable example of a substitution of such constants is furnished by the “coefficient of racial likeness” which is extensively used by some physical anthropologists as a measure of population divergence. Regarding its proper use Morant (1923, pp. 205-207) states as follows:
“It [the coefficient of racial likeness] is not a true measure of absolute divergence, and must not for a moment be considered as such, but neverthe- less we shall speak of it, for convenience, as if it were an absolute measure of racial affinity. When it is said that a low coefficient between two races A and B indicates a closer relationship than a higher coefficient between, say, A and C, what is meant always is that it is more probable that A and B are random samples from the same population than that A and C are.”
This is a lucid statement of the underlying idea. The coefficient of racial likeness is essentially an expression of probability and not a measure of divergence. Only as an expedient make-shift is it used as a measure of divergence. It has been extensively used as such by Morant and others. However, a make-shift should be used only when it does not lead to false conclusions ; but the coefficient of racial likeness often does lead to absurdly inaccurate biological conclusions, as shown by Seltzer (1937). It would seem to be best to abandon altogether the use of this coefficient as a measure of divergence, and if it is still desired to employ it as an expression of prob- ability, to change its misleading designation. (The coefficient of racial like- ness is used primarily to combine two or more characters for the purpose of measuring divergence. For any single character the misleading results ob- tained by using as a measure of divergence a certain formula that funda- mentally expresses probability, is discussed by me in another place (1938, pp. 279-282). The problem of measuring divergence for a multiplicity of characters I have considered in another paper (1939).)
Physical anthropologists of the school of the London Biometric Labora- tory having become inured to the use of the coefficient of racial likeness — which, as stated, is essentially an expression of probability — as a measure of racial divergence, we find a similar substitution of constants employed in still another connection. In a later paper, Morant (1936, p. 32) states as follows: “Different characters will arrange the series in very different
orders, and it is not clear, at first, why more importance should be attached to one of these orders than to any other ... A grading of the characters in order of importance for the purpose in view can be obtained by noting the number of significant differences found for each in a particular set of comparisons.” He then lists the percentage of times, of the total number of comparisons made, in which each one of a number of characters showed a “significant” value for a (alpha is the chief, compound factor in the formula for determining the coefficient of racial likeness; Morant postulates that if a is greater than 10, it shows significance).
Now, what do we understand by an “important” character? Obviously a character is important in distinguishing populations when it manifests a comparatively high degree of divergence, and the opposite is true of an unimportant character. With respect to populations of specific or lower rank, the degree of divergence it shows is the criterion by which the im- portance of a character may be judged. The importance of characters in such populations may be considered from two points of view.
First, often a character may be said to be important in the sense that it may be employed to divide a number of related populations of similar taxonomic rank, such as a number of races within a species or a number
18
Zoologica: New York Zoological Society
[XXV :2
of species within a genus, into two major groups. Its importance then con- stitutes a group divergence. Such a character will show a comparatively high divergence when a population of one group is compared with a popula- tion of the other group, and a relatively low divergence when a pair of populations within either group is compared, although even within the limits of each group it will generally show different degrees of divergence to a certain extent. In such species or genera, when the populations are divided into pairs in all possible combinations and the pairs compared, the char- acter, in general, will appear important in approximately half the number of comparisons and unimportant in the remainder; providing the number of populations in one group approaches equality to that of the other group, as they often do.
Second, more generally, the terms “important” and “unimportant” as applied to any given character is applicable only in connection with a given pair of populations, and they have no universal application. A character that may be important, that is, manifest a comparatively high degree of diver- gence, with respect to one pair of populations may be unimportant with respect to another pair, and vice versa. This is often true of a series of closely related populations. In a species containing many races, or in a genus comprising many species, that are divisible by important group characters into primary, secondary and tertiary groups, certain other important char- acters may crop up independently in some of the groups, and the same im- portant character may appear in groups that are otherwise not immediately related. Often important characters thus appear in such kaleidoscopic fash- ion that they cannot be used consistently for the major division of the species in a genus or the races in a species. In general, therefore, a char- acter may be said to be important only with respect to the comparison of a particular pair of populations.
Morant’s attempt to determine certain characters as of general impor- tance is, therefore, rather irrelevant. However, this is beside the point. What I am after is to point out that here also Morant uses a test of prob- ability to express what is fundamentally a divergence.
In comparative biological research, the essential thing we are after, in general, is to determine a difference or a divergence. This is true of both morphological and physiological comparisons, using the latter term in a broad sense to comprise all life processes including the complicated chain of events connected with the reproductive process. Whether we compare the morph- ology of pairs of related populations in taxonomic work, the comparative yield of milk for a given breed of cattle in feeding experiments, the percept- ible effects of a particular drug on guinea pigs or human beings in phar- macological research as compared with controls, etc., we are trying to deter- mine the precise divergence between two variable quantities or populations (in cases similar to the latter, between treated and untreated individuals, or between the same individuals before and after treatment) . This is our prime object. A secondary consideration is the mathematical determination of the probable reliability of the samples from a study of which the data are drawn that form the basis of our conclusions.
This being so, it is remarkable that hitherto most attention has been directed to the secondary consideration, the determination of probability, while the primary object, the determination of an adequate measure of divergence, has been rather neglected. A measure of divergence that is universally employed is the difference between the means of the two sets of data compared, but this is evidently not always adequate. It is certainly altogether inadequate in taxonomic research. A fundamental defect of measures of divergence in taxonomic research based on such values as the mean, median, or mode, is that they represent denominate numbers which are altogether unlike, their absolute values differing widely, in pairs of populations that differ by widely unlike characters. Consequently, the figures expressing the measures of divergence for different pairs of populations,
1940] Ginsburg : Divergence and Probability in Taxonomy 19
when based on denominate numbers, are not fairly comparable. That the figures are not comparable for characters the measures of which are ex- pressed in different units is self evident; but even when expressed in the same unit they are often not fairly comparable, if the characters are unlike. For instance, if the divergence of a pair of closely related populations of mice be expressed by the difference of the means of the tail length measure- ments, and that of a pair of populations of fishes by the same difference of the head length measurements in the same unit, the relative divergence of the two pairs may not be fairly comparable. Furthermore, measures of divergence based even on the same character expressed in the same unit, are not fairly comparable for different pairs of populations if the spread of their distributions differ widely. An ideal measure of divergence, one that could be used as a universal yardstick, should be an abstract number based on the degree of overlap, positive or negative, of the two frequency distributions, such as the measure employed by me (1938). That measure appears to be fairly adequate for taxonomic work. Whether that measure, or a modification of it, will be found applicable to research problems similar to the other two mentioned above, I am not prepared to discuss. I am here speaking chiefly from the point of view of the taxonomist. (In another paper considering the measure of divergence with respect to a multiplicity of characters, I (1939) concluded that a measure of divergence based on the principal character is fairly adequate, and that if the other characters are to be considered at all in its determination, they are to be afforded minor weights. In any further attempt at the combination of several characters, the figures for the different characters used should be such abstract num- bers that measure the divergence of every one separately, rather than de- nominate numbers that express their absolute values.)
Current formulae that are generally suitable for the determination of the probable reliability of samples investigated in taxonomic research, have as their fundamental bases the difference between the means of the two samples and their probable or standard errors, the size of the samples, and their variability as expressed by the standard deviation. The practical use of this determination in taxonomic research is rather limited. Fisher (1936, p. 59) states succinctly the proper, general application of the determination of significance as follows:
“It will be seen that the test of significance does no more, and attempts no more, than to answer the straightforward question, ‘Could these samples have been drawn at random from the same population?’ It calculates a probability. If the probability is very small the answer is ‘No.’ If it is not so small as to reach the level of significance required, the answer is ‘Yes, they could.’ The answer never is ‘Yes, they must have been.’ ”
To this I may add that in taxonomic practice, in the majority of cases, the actual arithmetical determination is rather unnecessary. If the two frequency distributions are fairly regular (that is, the frequencies in, the successive classes diminish successively at both sides of the mode, even though the distribution be skewed) as they usually are when based, respec- tively, on homogenous material and the sampling is adequate; and further, if the modes are at different even though closely a'djacent classes, as they usually are when the two populations represented by the distribution really differ and the degree of divergence is rather considerable, the arithmetical determination of probability will usually result in a “significant” figure. Most of the cases covered by Fisher’s first contingency may then be judged for practical purposes by a mere inspection of the data arranged in the form of frequency distributions.
In regard to pairs of distributions falling under Fisher’s second con- tingency, that is, distributions showing a divergence of relatively low mag- nitude, and the differences of which do not reach the level of mathematical significance as determined by the samples examined ; this class of examples will no doubt include many in which the differences are biologically signifi-
20
Zoologica: New York Zoological Society
[XXV :2
cant. In nature, differences in taxonomic characters between pairs of pop- ulations form a gradual series from small to large values, with virtually all possible intermediate values. (The series may be visualized as represented by a straight line of the equation, mx—y = 0.) Small values near the extreme of the series must have a biological significance, although mathe- matically their significance appears doubtful. For such populations the arithmetical determination of probability is of no practical value by itself, because it fails to give a definite answer to the question in which we are interested, namely, is the difference real, even though small, or is it due to the vicissitudes of sampling? The mathematical answer to this question, to adapt Fisher’s style in the preceding citation, virtually is “no” or “yes,” which is no direct answer at all. When the test for significance results in a low numerical value, lower than the accepted limit, it may mean either one of two things: (1) The difference is not real. (2) The difference is real but its magnitude is such that the samples are not large enough to prove its reality mathematically. Larger samples are necessary for a mathematical test of significance. The meaning of too low a figure then may be similar but opposite to what was noted above that too high a figure for significance shows that the samples are too large; but when the figure expressing signifi- cance is high the answer is direct and positive, and when it is too low the answer is indirect and limited.
While some of the small but real differences that are not too extreme will show mathematical significance when the size of the samples are greatly increased — and theoretically any real difference, no matter how small, will show significance by taking samples that become infinite as the differences, in a series of pairs of populations compared, approach zero — in actual prac- tice the size of the samples necessarily must be more or less limited. In work- aday biological practice, therefore, it can hardly be doubted that small differ- ences of biological significance will appear mathematically insignificant. In passing, it may be mentioned that instances may occur in which it would be impossible to obtain very large samples. Supposing we compare two populations of which the actual number of living individuals is very lim- ited, and find a small difference which, based on the entire number of living individuals, does not show any mathematical significance. That does not mean that such a small difference does not have a biological significance. In general, in cases coming under Fisher’s second contingency, our conclu- sions must be based on the biological evidence rather than on mathematical deduction.
From the standpoint of the comparative practical unimportance of the determination of probability in taxonomic research, it has received an undue share of attention from certain biologists whose work is essentially taxo- nomic, such as those investigations dealing with population or “racial” differences in various groups of living things. In general, this is also true of some statistical constants now in use in taxonomic work as noted below. From the point of view of the taxonomist at least, a great deal of what is being done along this line may be said to represent mathematical, rather than biologic research, employing biological data for the purpose of solving mathematical problems or formulating mathematical propositions. Of course, mathematics represents one of the important disciplines in the sum total of human culture, and there can be no objection for workers who are inter- ested in mathematical research to illustrate their problems and propositions by the use of biological data, if they wish to do so. But it should be remem- bered that a great part of such research is of little importance in solving taxonomic problems. In taxonomic problems, what we are greatly interested in is to determine divergence as precisely as possible, while the determina- tion of probability is of secondary importance. The thing to be regretted is not so much that a great deal of attention is being paid the latter, but that it is apparently done at the expense of the former. A few examples of published reports will illustrate this idea.
1940J
Ginsburg: Divergence and Probability in Taxonomy
21
An outstanding, valuable and well known taxonomic work to which reference is often made in biological discussions is that by Crampton (1916, 1925, 1932) dealing with the terrestrial gastropods of the genus Partula. The investigation forming the basis of Crampton’s reports is unusual as compared with taxonomic studies in general, by the number of specimens examined and the detail with which they were examined. The information furnished by Crampton makes it evident that Partula is at the present time level in an early and active stage of speciation, and as such, its detailed study is of special importance for an understanding of the process of evolution.
The study of speciation or raciation in Partula does not lend itself altogether readily to statistical treatment, because some of the important distinguishing characters are rather of a qualitative nature and are not readily expressible in terms of definite figures, although it is not altogether impossible to do so. One important character, the direction of the spiral of the shell, dextral or sinistral, can be expressed in terms of exact figures. In some species it is always either sinistral or dextral ; while in other species, or populations of lesser rank, the direction of the spiral varies with the individual. In the latter, Crampton very helpfully gives the precise numbers or proportion of the sinistral and dextral individuals in his samples. For certain other characters that are measurable with more or less precision, Crampton furnishes a wealth of statistical data and constants in tabular form.
However, the tables published by Crampton furnish only a part of the information that his data were in position to furnish, and that not the most important part. For each character he generally gives the range of varia- tion, the mean, the standard deviation and the error of the last two figures. For some species he also gives their coefficient of variation. Now, in the distinction of the species, or the different populations within a given species, and in the interpretation of the relationship of the various populations, of specific or lower rank, of what material difference, in general, is a knowl- edge that the standard deviation, or the coefficient of variation, in one is larger than in another population? Also, at their best such data are only approximate, and of what material difference, in general, is a knowledge of the small value of the error of the mean for the foregoing purposes? These figures are interesting, but they are largely of academic interest. Of course, there can be no objection if an author wishes to furnish such figures. What is regretable is that more pertinent information is omitted. From a taxo- nomic point of view we are intensely interested in how far or to what degree the different populations diverge with respect to the various characters. For that purpose we are presented only with the ranges and the means of the various characters, and these are altogether inadequate. To determine the precise extent of divergence, by some such method employed by me (1938), frequency distributions for the different characters for the sep- arate populations are needed and these are omitted for the characters based on measurements.
For three characters Crampton does give frequency distributions, namely, the direction of the spiral of the shell, the degree of tooth develop- ment, and the color pattern. Crampton’s presentation of the data for the last two characters is especially interesting, because they are rather quali- tative in their nature. As such, their determination in terms of definite figures is only approximate and dependent to some extent on a subjective estimate. Qualitative characters are generally described by authors in adjectival words or phrases that necessarily must be indefinite to a certain extent, and not in terms of definite figures. Crampton shows that such characters also can be expressed, approximately, in the form of frequency distributions. Similarly qualitative characters in other groups as well may be expressed in figures, and although such figures necessarily must be only approximate at their best, they should yet prove to be of importance in
9°
Zoologica: New York Zoological Society
[XXV :2
determining divergence between closely related populations. An interesting example of this kind is furnished also by the work of Sumner on mice of the genus Peromyscus. Characters based on color differences, in general, are qualitative, yet Sumner (1929) has found it possible to express quantita- tively seven such characters, presenting his results for two of them in the form of histograms (p. Ill), and giving the averages for the other five.
I have used Crampton’s reports as an example because they constitute a work of unusual value and interest as compared with the ordinary run of taxonomic papers, but the foregoing statements apply to many other pub- lished papers m which taxonomists employ statistical methods. I may here cite three recent papers in my own specialty, in fishes, that happened to come to my attention, namely, by Schultz (1937), by Matsubara (1938) and by Storey (1938). These papers are much less extensive in scope than Crampton’s reports, in that they deal with much fewer populations. They also differ more or less in the manner of the statistical presentation of the data; but they illustrate in different ways some of the points raised above.
Schultz compares the Pacific with the Atlantic population of the cape- lin. He compares a larger number of characters than usual in such cases, but for each character he publishes only the range of variation and the mean with its error. These figures are altogether inadequate for determining the precise divergence between the two populations, the thing in which Schultz as well as other taxonomists are chiefly interested. Had Schultz’s data been published in the form of frequency distributions, they would constitute a valuable example showing the differing degrees of intergradation of the several characters in two closely related populations that differ by more than one character, in addition to forming a basis for the determination of the precise divergence between the two populations. (Schultz’s method of combining several characters for the purpose of determining divergence I consider in another paper (1939).)
Matsubara, working with Japanese lizardfishes, does not employ statis- tical formulae or constants and does not calculate probabilities. Neverthe- less, his method is essentially statistical in its nature, as it properly should be in a problem such as the author was confronted with. However, the data for the variability of the characters that are employed in comparing and distinguishing his populations (which happen to be of specific rank), are presented in graphic form and are not altogether suitable for the purpose of calculating the extent of divergence in terms of precise figures. Of course, the frequency distributions of the several characters may be approxi- mately determined from the graphs, but it is very difficult or impossible to get the exact figures. For the precise determination of divergence, it is im- portant to have the actual frequency distributions obtained during the investigation.
Storey, repoi’ting on an investigation of the Atlantic populations of Harengula, also presents her data in graphic form and the same remarks apply to hers as well as to Matsubara’s method of presentation. Further- more, her data for characters having a continuous variation, namely, pro- portional measurements, are presented in the form of curves “smoothed by threes three times.” “Smoothing” has the slight advantage of producing somewhat more regular curves which are rather more pleasing to the eye, but it has an important disadvantage in that the curves tend to mask hetero- geneity in the material studied. That the material of Harengula pensacolae, for instance, possibly was heterogeneous is shown by her comparison (p. 35) of the specimens from Sanibel with those from other localities. Storey sug- gests that the differences in the measurements may be due to the different preservative used, formaldehyde instead of alcohol. This may be so to a certain extent, but part of the differences quite possibly represent a popula- tion divergence. The difference in the gill raker count of the Sanibel speci- mens would certainly seem to represent a population divergence. However, in clupeid species in general, the gill raker count differs greatly with the
1940]
Ginsburg : Divergence and Probability in Taxonomy
23
size of the specimens and Storey does not appear to have segregated her data in sufficiently restricted size groups to reveal any possible intraspecific population differences in this character.
Detailed studies of other clupeid species have shown that they tend to diversification into distinct, statistically measureable, local populations of lesser rank, subspecies or races. It is highly probable that this is also true of the four Atlantic species distinguished by Storey. In view of the close approach or even general intergradation between these four species in the characters determined, it is quite possible that if such a detailed study be made, the relationship of the various populations will receive a modified interpretation than that obtained by the data available to the author. The size of the samples studied by the author were rather restricted (Storey, 1938, pp. 16-17), and in order to apply to them current statistical formulae, the grouping of the data adopted necessarily had to be comprehensive. A study of larger samples and of the same characters determined by the author, measurements, gill raker count and ventral scute count, with the data segregated by locality, and those of the measurements and gill raker count by smaller size groups, would possibly present a somewhat different picture of the relationship of the various populations, than that obtained by the grouping of the data as adopted by the author. It is evident that not only is it important— in order to determine precise divergence, distin- guish properly the different populations and determine their relationship — to have frequency distribution tables published, but to subdivide the data where necessary by size, sometimes also by sex, and also by locality where heterogeneity is suspected.
Instead of presenting detailed frequency distribution tables, Storey gives derivatives of her data (table 3, pp. 16-17) in the form of certain constants, the most important of which are: the standard deviation, the mean, the difference between the means of the two populations compared and its standard error. These are not of much value in determining diver- gence as stated above. She also gives the relative deviate and the value of p, which express probabilities, as they are intended by the author to do, but are not suitable to determine divergence.
A paper based on the study of populations of flies that are of much in- terest in connection with some phases of the species problem, was very re- cently published by Mather & Dobzhansky (1939). It deals with the two well known “races” of Drosophila pseudoobscura, generally designated in the literature, following Lancefield’s suggestion in his original report (1929) announcing their distinction, as race A and race B. The two populations occupy different but overlapping geographic ranges; they are also incom- pletely segregated ecologically (Dobzhansky 1937a, pp. 406-408).
The apparent principal character proving that the two populations are distinct is a physiological one and refers to the sterility of hybrid offspring when they are crossed. The sterility is partial, being confined to the males. Hybrid females are fertile, at least in part. A backcross of Fi females to males of either parent population gives rise to both sterile and fertile males. Besides this principal character, Mather & Dobzhansky review and enumer- ate other, minor diverging physiological characters, and differences based on gene arrangement in the chromosones.
Morphological differences between the two populations that have so far been discovered show certain degrees of intergradation. The object of the paper mentioned is to deal with the morphological characters, and it takes up five such characters, namely, the number of teeth in both the proximal and distal sex comb on the leg of the male, the length and width of the wing and the length of the tibia. The former two are sex characters. The latter three characters were determined for males and females separately and they were found to differ by sex as well as by population. All five characters differed also according to minor populations or “strains” within each one of the two major populations.
24
Zoologica: New York Zoological Society
[XXV :2
Now, this problem is fundamentally taxonomic in its nature. Our con- cern is the determination of the relative rank of the taxonomic category in which the populations are to be placed. This determination, in its turn, must have as its basis a determination of the relative divergence of the populations. Given the known facts regarding the populations, let us con- sider this particular case from the taxonomist’s viewpoint. This case may not be as remarkable as it appears. Every careful taxonomist of wide ex- perience no doubt can cite similar instances in which distinct populations show relatively low and varying degrees of divergence with respect to mor- phological differences. Its apparent remarkableness rests on the partial sterility of the hybrid offspring correlated with a relatively low morpho- logical divergence, and very likely is due to the fact that relative sterility and fertility of hybrids has been definitely determined only in a rather negligible number of very closely related populations.
In appraising the case under consideration taxonomically, it is well to consider the relative importance of physiological and morphological criteria in classification. There is no fundamental reason why the former should not be used for this purpose the same as the latter. Morphological criteria are generally used in taxonomy because they are determinable more readily and with greater precision. In the relatively few known cases in which a physiological character shows a greater divergence than any known mor- phological character, the former may be employed as the principal character in determining the taxonomic rank of the pair of populations compared.
In the case considered, the sterility criterion is evidently important, and we may confine ourselves to a consideration of this physiological cri- terion. The precise value of this criterion in classfieation in general cannot be said to be as yet firmly established, and it cannot well be appraised, be- cause it is known for relatively few populations as compared with their untold multitude. However, in general, it is evident that this criterion is not absolute, but, on the contrary, it is fully expressible in terms of degrees of magnitude only. Even in regard to the classical example of hybrid sterility, the mule, one now and then finds in the literature apparently au- thentic records of fertile individuals. Possibly, if sterility in the mule be investigated extensively and systematically, the percentage of fertile indi- viduals may be found to be greater than such haphazard observations would seem to indicate. At any rate, judged by what we already know in regard to hybrid sterility in general, it is evident that this criterion shows all de- grees of differences, from perfect or almost perfect sterility through dif- ferent degrees of partial sterility to comparatively unhampered fertility, depending on the populations crossed.
There is an incomplete correlation between relative sterility and the relative degree of divergence of morphological characters. When an at- tempt is made to cross two closely related populations that have reached a sufficiently high degree of divergence, as determined by morphological cri- teria, to be generally regarded as species, one of several things may happen: (1) They may not be crossable. (2) They may produce zygotes showing various degrees of inviability, that is, they die at various stages of develop- ment, depending on the populations.' (3) When viable offspring are pro- duced, they may be sterile or show infertility of varying and rather pro- nounced degrees. On the other hand, when a cross is made between two populations the divergence of which is of such a rather low degree, as de- termined by morphological criteria, that they are generally regarded as of a taxonomic rank below that of species, fertile offspring generally seem to be produced. However, even with our present rather meager knowledge re- garding sterility of hybrid offspring, it is evident that there are frequent exceptions to the above generalizations. A pair of closely related popula- tions, which, judged by morphological criteria, are generally regarded as species sometimes, perhaps often, on being crossed give rise to a progeny that is fertile.
1940]
Ginsburg: Divergence and Probability in Taxonomy
25
Since the magnitude of sterility is relative, being merely a matter of degree, it follows that if it be used as a criterion for grading populations into taxonomic categories, it would be necessary to draw arbitrary lines between the species and various categories of lower rank, the same as when morpho- logical characters, especially quantitative ones, are used for that purpose. It is then necessary to devise a measure for expressing the degrees of sterility, and the most obvious measure that suggests itself is the relative percentage of sterile and fertile individuals in the hybrid progeny of the pair of popula- tions compared. For instance, we may decide arbitrarily that if a pair of closely related populations on being crossed produce, on the average, a progeny 90% or more of the individuals of which are sterile, the populations are to be designated as species; they are to be designated as subspecies when the percentage of sterile individuals is 75 — 85%, other things being equal; they are to be designated as races when the same percentage is 60 — 70. These are the tentative arbitrary lines which I (1938) suggested to draw for morphological characters. For the sterility criterion even less data are extant than for morphological criteria, to enable us to draw the most pertinent arbitrary lines ; but wherever drawn the lines evidently must be arbitrary. It may perhaps be found desirable to draw arbitrary lines for the sterility criterion that differ in numerical value from those employed for morphological criteria.
There being no absolute correlation between morphological divergence and relative sterility, the sterility criterion, if employed in taxonomy, evi- dently is to be used on a par with morphological criteria and coordinate with them. Whichever is the most divergent, it is to be used as the principal character for determining the taxonomic rank of the pair of populations compared. If any morphological character shows a divergence of specific magnitude the two populations are to be designated as species even though a cross between them produces offspring that are 100% fertile. Conversely, if the degree of sterility is greater than any morphological character that has been discovered, the former is the chief factor to be used in determining the taxonomic rank of the pair of populations compared.
Bearing the preceding propositions in mind, let us turn to the question of the taxonomic rank of the two major populations of Drosophila pseu- doobscura. It is evident that for a pertinent decision of the question we need to know the degree of divergence of the various characters, morphological as well as physiological. For this purpose the paper by Mather & Dobzhan- sky furnishes only the averages of the morphological characters, which are entirely inadequate for measuring divergence, as discussed above. From a taxonomist’s viewpoint, the data presented educe further questions. The authors very properly subdivide each one of the two major populations, races A and B, into minor populations which they designate as “strains.” In the morphological characters the differences between the extreme strains within each race is nearly as great or greater than the differences between the major, composite populations. The question then is, when is a population to be designated as a “strain” and when is it to be designated as a “race”? More specifically, from the data presented it is evident that we have two (or more) “strains,” one in each “race,” that morphologically are approxi- mately alike. What criterion then do they use for placing one strain in one race, and the other, morphologically similar strain, in the other race? Al- though I searched the paper for a definite statement in answer to this ques- tion, it could not be found. Apparently their basic criterion is sterility, for Dobzhansky (1937b, p. 285) states: “. . . the Fi hybrid males from crosses between race A and race B are always sterile . . .” But other questions present themselves: On how extensive a body of data is the above quoted statement based? Especially, on how many crosses between different strains is it based? Is sterility of Fi hybrid males absolute also between strains that are alike morphologically?, since we have seen that there is a rough corre- lation between morphological divergence and sterility. I did not deepi it
26
Zoologica: New York Zoological Society
[XXV :2
necessary for the present purpose to enter into a complete analysis and re- view of the recorded investigations that have a bearing on an answer, if any, to the preceeding questions. Even should sterility of the hybrid male eventually prove not to be absolute between all the strains, it nevertheless seems evident from the above quoted statement that its degree, in general, is comparatively high.
Let us see now to what taxonomic conclusion we may come on the basis of the data that have been recorded.
A consideration of the condensed data presented by Mather & Dobzhan- sky makes it apparent that the morphological character showing the greatest divergence between the two major populations, taken in their entirety, re- fers to the number of teeth in the proximal sex comb of the male (with that based on the number in the distal comb a close second). While the precise degree of divergence between the two primary populations cannot be de- termined from the condensed data, it seems apparent that divergence is rather considerable but intergradation must also be pronounced. Arrang- ing the averages in the authors’ table 2 in their order of magnitude, five of the extreme minor populations, “strains,” of race B have averages of 5.88, 5.92, 5.96 6.00 and 6.08 respectively, nearly the same as three extreme minor populations of race A with averages of 5.92, 5.96, and 6.00. The total number of minor populations compared are 20 of race B and 19 of race A; those having the averages nearly alike are, consequently, 25% of the total “strains” of B and 16% of A, or an average of about 20%. Had the authors given their data in the form of frequency distributions, the individuals com- prised in the samples enumerated would apparently be seen to represent intergrades to a large extent. In addition, it seems apparent that a con- siderable number of individuals of the other minor populations, especially of those populations the averages of which are next in magnitude of those enumerated above, would also prove to be intergrades. It may be reasonably expected then that the two major populations taken in their entirety would show an intergradation of 25% or more in the morphological character of greatest divergence, according to the measure of divergence suggested by me (1938). According to the suggested arbitrary lines in the foregoing paper, this represents a divergence of racial or nearly subspecific magnitude. Judged by this criterion, on the basis of the incomplete data, the two popu- lations perhaps are to be taxonomically designated as races, or they may possibly be near the borderline between the subspecies and the race.
The other important matter to consider is sterility. At stated, for the present it is difficult to form a judgment regarding the precise treatment of this criterion in taxonomy in general. We must also bear in mind the fact that in this particular case sterility is confined to one sex, the male. While by the use of morphological criteria species are sometimes based on char- acters of one sex, and properly so, the question remains whether the same course is to be followed with respect to the sterility criterion. Nevertheless, the fact of hybrid male sterility appears to be of tremendous biological significance, especially if we assume that it is 100%, or nearly so, as the above quotation from Dobzhansky would seem to indicate. Therefore, tak- ing into consideration the sterility criterion, the rather considerable diver- gence in two morphological characters, and the other differences mentioned by Mather & Dobzhansky, the divergence would seem to be approximately of subspecific magnitude. A taxonomist’s best and most reasonable judgment based on the extant incomplete evidence, would then seem to dictate the course, at least tentatively, of recognizing the two major populations as subspecies. According to a common taxonomic practice, it may be desirable to formally distinguish them by name, as, for instance, to designate them Drosophila pseudoobscura2 and Drosophila lancefieldi2, employing the nota- tion for subspecies as suggested in my (1938) paper.
The course here suggested will give taxonomic expression to the relative divergence of some of the major populations of the genus Drosophila with
1940]
Ginsburg : Divergence and Probability in Taxonomy
27
respect to morphological criteria and the sterility criterion. Some of the other major populations diverge in varying but pronounced degree morpho- logically, and when crossed produce inviable or altogether sterile offspring; they are, therefore, recognized as species. The two populations under con- sideration diverge morphologically in lesser degree and produce partly fer- tile offspring, and are consequently designated as subspecies.
The sterility criterion may be broadened to include the various graded results that occur when a cross is made between two populations. The grades are: incrossability, inviability of zygotes graded, in its turn, accord- ing to stage of development at which they die, offspring viable but sterile (even such populations may be only partly crossable), progeny partly fertile. It is interesting to note that in the one genus Drosophila there appears to be a correlation, at least partial, between hybridization results and morpho- logy. Species that are readily separable morphologically appear to be either not crossable or to produce inviable zygotes. D. melanogaster and D. simu- lans, a cross of which results in viable but sterile offspring (Sturtevant 1929), also diverge morphologically in a lesser degree than some other species, so much so that they were not distinguished until comparatively recently (see Sturtevant, 1921, pp. 91-92). Finally, the two major popula- tions of D. pseudoobscura1 show a comparatively low morphological diver- gence correlated with partial fertility of hybrids.
The differences between D. miranda and D. pseudoobscura1 ( Dobzhan- sky, 1935) seem to be rather intermediate between that of the latter two pairs of populations mentioned. They produce viable but altogether sterile offspring the same as the cross between D. simulans and D. melanogaster. But morphologically the difference between them is evidently not greater than that between “races” A and B of pseudoobscura, whereas D. simulans and D. melanogaster are more greatly divergent morphologically, showing one apparently discontinuous difference, that relating to the structure of the male genitalia (compare Sturtevant, 1921, especially his figures 13-14. p. 34, with Dobzhansky, 1935).
Dobzhansky (1935) and Dobzhansky & Tan (1937) describe important differences in the chromosome structure between D. miranda and D. pseu- doobscura; but, except in the peculiar distribution and number of sex chromosomes, such differences are evidently also nothing more than a matter of degree, since similar differences appear to exist not only between the two major populations of pseudoobscura, but also between the “strains” or minor populations (see Dobzhansky, 1937b, pp. 92-95). In general, hardly the sur- face has been scratched in elucidation of differences in chromosome structure between very closely related populations, and it cannot be used at present as a criterion in classification with any degree of assurance.
D. azteca and D. athabascae evidently constitute another example of two closely related populations of Drosphila, “. . . that are very similar ex- ternally but that produce sterile Ft hybrids . . .” (Dobzhansky, 1937b, p. 113).
At least some of the major populations of the genus Drosophila have evidently reached at the present time level a fairly advanced stage of spe- ciation, so that they are rather easily separable and are designated as species. Yet, even in this genus, when all major populations are considered, there is a certain gradual transition in degrees of divergence as determined by both morphological criteria, and physiological criteria based on hybridiz- tion. As for the minor, intraspecific populations, it may well be expected that various degrees of divergence will be discovered when the several species are subjected to such taxonomic analysis as was carried out by Mather & Dobzhansky on D. pseudoobscura x. The same state of affairs very likely will be found to exist when we know more about the variability of some of the numerous poorly known species, or what are now recognized as species.
28
Zoologica: Netv York Zoological Society
[XXV :2
The taxonomic course suggested above indicates how a taxonomist would or should form a judgment in this particular case. A decision could be made with greater assurance, had the authors presented frequency dis- tributions of at least the morphological character showing the greatest di- vergence. That would have given us a basis for a determination of the precise degrees of divergence between the minor populations and of the divergence of the two major populations by combining the data of each one of the two groups of minor populations, by some such method as was sug- gested by me (1938). Instead, the authors present only derivatives of their data in the form of means and squares of deviations for each character, and for a combination of characters, “a score,” the latter by a method developed by Fisher. The main results of these figures is that they lead to other, final derivative figures expressing mathematical tests of significance of the dif- ferences between averages by standard methods. Now, just what are the values of the final figures in taxonomy? Do they convey any special mean- ing or ideas to the biologist, which would help him to come to a better decision, one made with greater assurance, in regard to the taxonomic status of the two populations, than that based on their morphological divergence and on the sterility criterion? The test of significance is interesting, but it is only of minor interest in taxonomy. In this instance especially, the definite determination of mathematical significance would seem to be of no more than academic interest. Arranging the averages given by the authors in their tables 2 and 4 in order of magnitude, one with some experience in applying current statistical formulae may see at a glance, without going through the actual arithmetical calculation, that the figures for the averages would result in values for the test of significance that would reach its ac- cepted limits for the principal character. Any difference in the magnitude of the values beyond that limit does not have any special taxonomic meaning, as noted above. To an experienced and careful taxonomist, the figures for the averages, even condensed summaries though they are, speak much more eloquently, they constitute a much better basis for the formation of con- structive decisions than the figures resulting from the test of significance. In sum, what the paper virtually accomplishes is to determine the figures for the test of significance. But this is only a secondary part of the problem. The main thing that we need to determine is the precise degree of diver- gence. This question is very inadequately answered.
While the greater part of the paper deals with the test of significance which is a matter of but minor interest in taxonomy, the subject of our primary concern, the determination of precise divergence to serve as a basis for forming a decision with assurance is incidentally considered, only insofar as the given averages form a very inadequate measure for such a determination. The valuable data determined by the authors in their in- vestigation is not presented in such manner that the precise degree of di- vergence could be determined. Here then is an example of an investigation that bears the earmarks of care and reliability in its execution and carried out by reputable investigators, but the report of which fails to furnish the data in such form as will be of most help in deciding the question of our chief concern. To use a favorite expression of biological statisticians, the authors failed to extract all the information which their data were capable of furnishing, and the most important part is omitted. They, fail to give even the ranges of variability of the different characters. This investigation is especially interesting in that, in a sense, it represents a study in ex- perimental taxonomy. The groups of individuals, designated by the authors as “strains,” upon which the data were determined, were bred in the labora- tory from parents of known origin. It would be very interesting to compare divergence in such populations with that of the same populations as they occur wild ; but the data as presented permit only an inadequate comparison.
I wish to emphasize here that I am not criticizing the paper as such. What I am after is to discuss its value in taxonomy, and the problem with which the paper deals is primarily a taxonomic one.
1940J
Ginsburg: Divergence and Probability in Taxonomy
29
Physical anthropologists have been very assiduous in determining series of measurements of their material in many characters, as may be gathered by going through the volumes of Biometrica, for instance. Now, although they apparently have different standards than other systematists by which they determine the taxonomic rank of the populations studied by them, physical anthropology is nothing more than a highly specialized branch of taxonomy, dealing chiefly with one »genus, Homo, and the same methods are applicable to this specialized branch as to taxonomy in general. In going through the published reports on physical anthropology in Biometrica, one finds that, omitting correlation studies, the figures on which the discus- sions and conclusions are based generally consist of the mean, the standai'd deviation, the coefficient of racial likeness, and their errors; sometimes the coefficient of variation is presented. However, what we are chiefly interested in, is the precise divergence between the populations, and those figures are not of much value in determining that, as stated.
The published reports that are used to illustrate the foregoing dis- cussion, represent taxonomic investigations carried out by the use of appro- priate modern methods, statistical methods. No fault can be found with the methods of investigation adopted by the authors. These are the methods that are to be recommended in taxonomic research. One or another of the investigations reported may be incomplete in one or another direction, but the methods adopted are correct as far as they go. Nevertheless they fail to determine the thing that is most important taxonomically, namely, the precise degree of divergence. This is generally true of published reports of taxonomic investigations in which modern statistical methods are employed. It seems that, in general, workers labor under the spell of statistical for- mulae developed through generations, and having to do chiefly with the theory of probability, something which is generally of but secondary im- portance in taxonomy.
What is to be especially regretted is that the class of reports under con- sideration, with frequent fortunate exceptions, omit the necessary data, complete frequency distribution tables, by which precise divergence could be determined by anybody who is interested. In general, in any investiga- tion it is the data determined that are of primary importance. The manner of treatment of the data, their interpretation and the conclusions drawn from them, as given by the investigator, may not be the only possible ones. It is possible that some biologist at a later date, considering the subject from another point of view, may wish to treat the data in a different man- ner, and such treatment may even necessitate a different interpretation and lead to different conclusions. However, this would be impossible unless frequency distribution tables of the original data are presented. As it is, valuable data forming the quintessence of any investigation, representing a great deal of time and patient and concentrated effort, is thus practically lost to future workers. For instance, I (1938) employed a certain measure for determining precise divergence. This measure seems adequate in taxo- nomic work. However, some future worker may propose a measure that is more adequate, better expressive of the essential facts. But as most reports are presented now, it is not possible to determine divergence by the method employed by me, and will probably prove to be indeterminable by any future method that may be proposed. For the method mentioned at least, frequency distribution tables of the data are necessary, and this will likely prove to be so in any case.
In this connection a few words may not be amiss in regard to the economic aspect of the subject, having the editor’s point of view in mind. Authors are sometimes confronted with the editor’s desire to abbreviate manuscript reports by eliminating parts that appear not very essential. (It is unfortunate that this happens even with reports of outstanding merit.) In such cases, if elimination of some parts becomes necessary, and the ques- tion comes up whether to eliminate frequency distribution tables or graphic
30 Zoologica: New York Zoological Society [XXV:2
representations of such tables, it is best to dispense with the latter. For, graphic representations merely constitute a device to “catch the eye” and clinch the author’s conclusions, and it is usually possible to represent data graphically in more than one way, while the data on which the report is based are quintessential, as stated. If absolutely necessary, frequency dis- tributions should be included even at the expense of parts of the discussion. In some cases, a report may be abbreviated even without urging from the editor, and yet be more informative than in the form in which it finally appears. For instance, some reports include tables giving detailed meas- urements of individual specimens, and yet they fail to include frequency distribution tables. But in mass data individual measurements are not of much consequence. (They may be of some interest in the case of very ex- treme or palpably aberrant specimens.) Such tables are rather superfluous and not likely to be carefully perused even by specialists directly concerned. In mass data what we are chiefly interested in, and what our conclusions are likely to be based on, is the frequency distribution of a population with respect to a given character under consideration.
It is fortunate for the cause of science that taxonomists are more and more abandoning the idea that taxonomy consists chiefly of the publication of local lists and catalogs, and descriptions of new species. This was all right for taxonomy in its pioneering stages, and if carefully and skillfully done, such papers still serve a useful purpose. However, gradually it is com- ing to be realized that the backbone of taxonomy is to be found in the careful and adequate comparison of related populations, whether they be of specific, subspecific or racial rank, by statistical methods, for the purpose of deter- mining the intrapopulational variability, and interpopulationally, their precise degree of intergradation, or divergence, with each other. This forms a proper basis for an understanding of the relationship of groups of closely related populations. Our chief interest is to determine the precise divergence be- tween pairs of closely related populations, or concomittantly their precise intergradation, these two values being complementary. The determination of probability is of but secondary importance in taxonomy, although most attention has hitherto been given to it. In issuing reports of taxonomic in- vestigations in which statistical methods are employed, it is essential to in- clude frequency distribution tables of the data, so that precise divergence, or intergradation, may be determined, by existing methods or by methods that may be discovered in the future.
Literature Cited.
Crampton, Henry E.
1916. Studies on the variation, distribution and evolution of the genus Partula. The species inhabiting Tahiti. Publ. Carnegie Inst. Wash- ington, no. 228, 313 pp.
1925. The species of the Mariana Islands, Guam and Saipan. Ibid., no. 228A, 116 pp.
1932. The species inhabiting Moorea. Ibid., no. 410, 335 pp.
Dobzhansky, Theodosius
1935. Drosophila miranda, a new species. Genetics, vol. 20, pp. 377-391.
1937a. Genetic nature of species differences. Amer. Nat., vol. 71, pp. 404-420.
1937b. Genetics and the origin of species. Columbia University Press, 364 pp.
Dobzhansky, T. & Tan, C. C.
1937. Studies on hybrid sterility. III. A comparison of the gene arrange- ment in two species, Drosophila pseudoobscura and Drosophila mir- anda. Zeitsch. Induk. Abstam. Vererbl., bd. 72, pp. 88-114.
1940]
Ginsburg: Divergence and Probability in Taxonomy
31
Fisher, R. A.
1936. “The coefficient of racial likeness” and the futui’e of craniometry. Jour. R. Anthrop. Inst. Great Britain and Ireland, vol. 66, pp. 57-63.
Ginsburg, Isaac
1938. Arithmetical definition of the species, subspecies and race concept, with a proposal for a modified nomenclature. Zoologica, vol. 23, pp. 253-286.
1939. The measure of population divergence and multiplicity of characters. Jour. Washington Acad. Sc.i., vol. 29, pp. 317-330.
Lancefield, D. E.
1929. A genetic study of crosses of two races or physiological species of Drosophila obscura. Zeitsch. Induk. Abstam. Vererbl. bd. 52, pp. 287-317.
Mather, K. & Dobzhansky, T.
1939. Morphological differences between the “races” of Drosophila pseudo- obscura. Amer. Nat., vol. 73, pp. 5-25.
Matsubara, Kiyomatsu
1938. A review of the lizard-fishes of the genus Synodus found in Japan. Jour. Imp. Fish. Inst. Tokyo, vol. 33, no. 1, pp. 1-36.
Morant, G. A.
1923. A first study of the Tibetan skull. Biometrica, vol. 14, pp. 193-260.
1936. A contribution to the physical anthropology of the Swat and Hunza Valleys based on records collected by Sir Aurel Stein. Jour. 'R. Anthrop. Inst. Great Britain and Ireland, vol. 66, pp. 19-42.
Schultz, Leonard P.
1937. Redescription of the capelin Mallotus catervarius (Pennant) of the North Pacific. Proc. U. S. Nat. Mus., vol. 85, pp. 13-20.
Seltzer, Carl C.
1937. A critique of the coefficient of racial likeness. Amer. Jour. Phys. Anthrop., vol. 23, pp. 101-109
Storey, Margaret
1938. West Indian clupeid fishes of the genus Harengula. Stanford Ichthy. Bull., vol. 1, pp. 3-56.
Sturtevant, A. H.
1921. The North American species of Drosophila. Carnegie Inst., Wash- ington, Publ. no. 301, 150 pp.
1929. The genetics of Drosophila simulans. Ibid., no. 399, pp. 1-62. Sumner, Francis B.
1929. The analysis of a concrete case of intergradation between two sub- species. Proc. Nat. Acad. Sci. U. S., vol. 15, pp. 110-120, and 481-493.
J
'
■
(
\
Conant & Doums: Miscellaneous Notes on Reptiles
33
3.
Miscellaneous Notes on the Eggs and Young of Reptiles.
Roger Conant & Alexander Downs, Jr.
Philadelphia Zoological Garden.
Everyone who is in close daily contact with an extensive collection of living reptiles soon accumulates an assortment of notes, which, while inter- esting in themselves, fail to fit the pattern of his contemplated papers. Thus we have assembled miscellaneous data on some 38 clutches of eggs and broods of young which, except as noted, were laid or hatched in the reptile house in the Philadelphia Zoological Garden. These data, while fragmentary in spots, we append below with the hope that like Klauber’s (1938) “odd items of general interest,” they may prove of use. At least we are fortified by the thought that with startlingly few exceptions the life histories of most reptiles are so poorly known that even scraps of information about them are welcome additions to herpetological knowledge.
The data available in the literature on this subject have come chiefly from captive specimens. The chances of observing egg-laying or hatching in the field are slight, and far more can be learned by maintaining gravid females in captivity under conditions as closely approximating natural ones as possible. Professional curators have contributed much information but here is a big field for the amateur herpetologist to make useful observa- tions. All he needs is an ability to record data accurately, an inexpensive meter stick, and a friendly corner druggist who does not mind having his scales used in the interests of science.
Our weights and measurements were recorded as soon after laying or hatching as was convenient — usually on the same or the following day. Eggs were incubated in aquaria or glass jars nearly filled with moist, decayed wood. A flat piece of glass placed on top of each vessel served as a lid and permitted control of the moisture content within each one. When the under- surface of the glass “sweated” appreciably the lid was moved slightly to one side to allow evaporation. Conversely, when the hatching media became too dry it was sprinkled with water and the lid was replaced. Ordinary room temperatures prevailed during the incubation periods.
The weights and measurements of all female specimens were recorded after their eggs were laid. All weights are recorded in grams and all measurements in millimeters.
The letters “H” and “L” indicate the highest and lowest numbers, re- spectively, in each column of figures, thus representing the extremes of variation in each group. Unusually abnormal specimens are excepted. In cases where the highest and lowest figures appear more than once, only the first is designated by a letter.
For help in taking weights and measurements and in securing other data we are indebted to Mark Mooney, Jr., and to Edmond Malnate, succes- sive superintendents of the reptile collection in the Philadelphia Zoological Garden. Several members of the Junior Zoological Society of Philadelphia also have given valuable assistance.
[XXV :3
'34 ’ Zoological: NeiO York Zoological Society
Alligator mississipiensis (Daudin).
A number of alligators, apparently recently hatched, were discovered along the swampy margins of a large pool on St. Simon’s Island, Georgia, about August 20, 1937, by Robert S. Ingersoll, Jr. There were about 36 specimens in the group, but not all of them were captured. Measurements and weights were recorded when the alligators arrived at the Philadelphia Zoological Garden a few days later.
|
No. |
Length |
Weight |
No. |
Length Weight |
|
|
1 |
241 |
53.2 |
11 |
243 |
46.5 |
|
2 |
233 |
50.9 |
12 |
244 |
52.6 |
|
3 |
249 |
51.9 |
13 |
251 |
52.5 |
|
4 |
252 |
56.4— H |
14 |
247 |
51.7 |
|
5 |
231— L |
42.9 |
15 |
257— H |
52.6 |
|
6 |
247 |
53.2 |
16 |
247 |
52.5 |
|
7 |
247 |
47.0 |
17 |
244 |
51.2 |
|
8 |
249 |
53.2 |
18 |
236 |
42.7 — L |
|
9 |
237 |
48.7 |
19 |
232 |
47.6 |
|
10 |
241 |
43.0 |
Average |
243.58 |
50.02 |
|
Basiliscus vittatus (Gray). |
|||||
|
Eggs found scattered about on the floor of the |
cage on June 29, 1937; |
||||
|
all separate from one another. Shells buff, ovoid and rather soft. |
Female |
||||
|
and male |
in captivity since March 20, 1935. Both |
were confined |
in same |
||
|
cage but |
eggs were infertile. The shell of egg No. 10 was badly dented. |
||||
|
No. |
Length |
Width |
Weight |
||
|
1 |
18.7— H |
10.9 |
1.40 |
||
|
2 |
17.7 |
11.1 |
1.31 |
||
|
3 |
18.3 |
11.7 |
1.60— H |
||
|
4 |
17.4 |
12.6— H |
1.50 |
||
|
5 |
18.3 |
10.5 |
1.25 |
||
|
6 |
16.9 |
10.8 |
1.40 |
||
|
7 |
17.0 |
10.4 |
1.09 |
||
|
8 |
16.2 |
11.3 |
1.05 |
||
|
9 |
17.8 |
10.5 |
1.25 |
||
|
10 |
15.0 |
10.1 |
.65 |
||
|
11 |
15.6— L |
11.5 |
1.10 |
||
|
12 |
16.2 |
10.0— L |
.91 |
||
|
13 |
16.4 |
12.0 |
1.10 |
||
|
14 |
15.9 |
11.1 |
.90— L |
||
|
Average |
16.96 |
11.04 |
1.18 |
Sceloporus undulatus fasciatus (Green).
Group of eggs brought to the Philadelphia Zoo about August 1, 1938. Donor could not furnish the locality where they were found, but stated it was in southern New Jersey.
Total Length
|
No. |
Length |
( snout to vent) |
Weight |
|
1 |
45— L |
22— L |
0.4 |
|
2 |
46 |
22 |
0.4 |
|
3 |
49— H |
23— H |
0.3 |
|
4 |
45 |
22 |
0.3 |
|
Average |
46.25 |
22.25 |
0.35 |
Egg No. 1 hatched on August 8, the others on August 11.
1940]
Coimnt & Downs: Miscellaneous Notes on Reptiles
35
Ophisaurus apoclus (Pallas).
Eggs found on July 20, 1937, in a cage occupied by several large, adult specimens received on June 9, 1937, from the Zoological Society of London. Probably the two eggs did not constitute the entire complement from any one female, but no others were laid.
|
No. |
Length |
Width |
Weight |
|
1 |
33.4 |
18.7 |
6.45 |
|
2 |
33.2 |
18.4 |
6.00 |
|
Average |
33.3 |
18.55 |
6.23 |
Farancia abacura abacura (Holbrook),
A large female of this subspecies was found about August 20, 1937, coiled around her eggs under a large log on St. Simon’s Island, Georgia, by Robert S. Ingersoll, Jr. There were approximately 15 or 20 eggs but un- fortunately some of them were destroyed in transit to Philadelphia and the exact number could not be determined. Two of the eggs were hatched on
|
October 9, 1937. |
, . |
|
|
No. |
Length |
Weight |
|
1 |
204 ( |
4.8 |
|
2 |
231 |
6.2 |
|
Average |
217.5 ! |
5.5 |
Specimen No. 1 was deformed, its tail being kinked iii an S-shape.
Opheodrys aestivus (Linnaeus).
A female specimen. (collected at Royal Oak, Talbot County, Maryland, on July 2, 1938, was kept, alive at Treasure Island Boy Scout Camp, where it deposited 9 eggs on July 13. The , eggs subsequently were given to the Philadelphia Zoological Garden, where they began hatching on September 9, 1938.
|
A To. |
Length |
Weight ' |
|
1 |
183 v |
1,3 |
|
■2 |
201 — II |
L4 — H |
|
'3 |
190 |
1.2 |
|
,■4 |
182 = |
,, .. 1.1— L |
|
5 |
191 |
1.3 |
|
6 |
180— L |
1.4 |
|
7 |
192 |
1.3 |
|
8 |
187 |
1.2 |
|
9 |
190 |
1.2 |
|
Average |
188.4 |
1.27 |
When the eggs first were discovered to be hatching one snake had completely escaped from the shell, another had its head half-way out of the egg and a third had the tip of its nose exposed. Six young had escaped by the morning of September 10; the numbers of slits made by these snakes in their respective egg shells were 5, 3, 2, 2, 2 and 1. All the young were very timid, drawing their heads back into the shells when an observer approached, or hiding in the hatching media after they had left the eggs. The last snakes escaped from their eggs on September 11. On this date the eyes of all were overcast, preparatory to shedding. The young were uniformly greenish-gray in coloration at first, but changed to uniform light green upon shedding their skins.
36
[XXV :3
Zoologica: New York Zoological Society
Coluber constrictor constrictor Linnaeus.
A clutch of eggs of this species was found buried in the earth by H. Leschke at Haddonfield, N. J., on June 19, 1936. They were stained brown, and salt-like protuberances on the shells made them rough to the touch.
|
No. |
Length |
Width |
Weight |
|
1 |
28.0 |
22.4 |
8.1 |
|
2 |
26.8— L |
22.9 |
7.8 |
|
O O |
32.3 |
20.4— L |
7.4 |
|
4 |
31.0 |
22.0 |
8.3 |
|
5 |
29.8 |
22.1 |
8.0 |
|
6 |
27.3 |
23.2 |
7.8 |
|
7 |
27.5 |
21.7 |
7.1— L |
|
8 |
29.5 |
21.9 |
8.3 |
|
9 |
27.0 |
23.7— H |
8.3 |
|
10 |
29.5 |
22.0 |
7 8 |
|
11 |
29.0 |
22.4 |
8.1 |
|
12 |
28.2 |
21.8 |
7.3 |
|
13 |
29.7 |
20.8 |
7.9 |
|
14 |
31.0 |
21.7 |
8.4 |
|
15 |
27.7 |
22.8 |
7.2 |
|
16 |
28.5 |
22.4 |
7.7 |
|
17 |
33.7— H |
21.4 |
8.5 |
|
18 |
29.8 |
20.5 |
7.3 |
|
19 |
31.9 |
22.0 |
8.7— H |
|
20 |
30.0 |
21.5 |
7.6 |
|
Average |
29.41 |
21.98 |
7.88 |
Twelve of these eggs hatched between August 7 and August 12, 1936.
No.
1
2
3
4
5
6
7
8 9
10
11
12
Length
278 288 275 272
279 282 256
253— L 265
305— H
294
284
Weight
5.2
5.2
6.1
4.6
5.8
6.0
4.5— L
4.5 5.0
62— H
5.6 5.8
277.5
5.38
Average
1940]
Conant & Downs: Miscellaneous Notes on Reptiles
37
A group of 15 eggs of the same species was laid on June 6, 1938, in a c,age containing several females, all caught a short time before near Phila- delphia. Determination of which adult laid the eggs was impossible, but it was presumed that the entire lot was deposited by one female. All the eggs were white, the shells had salt-like protuberances on them and all except No. 11 were approximately the same size.
|
No. |
Length |
Width |
Weight |
|
1 |
35.6 |
18.2 |
7.5 |
|
2 |
36.3 |
18.6 |
8.0 |
|
3 |
32.1 |
20.0— H |
7.9 |
|
4 |
33.7 |
19.4 |
7.7 |
|
5 |
36.7 |
19.2 |
7.8 |
|
6 |
33.6 |
19.6 |
8.3 |
|
7 |
34.1 |
19.8 |
7.8 |
|
8 |
37.6 |
18.4 |
7.3— L |
|
9 |
31.7— L |
19.4 |
7.5 |
|
10 |
34.9 |
18.6 |
8.0 |
|
11 |
46.5— H |
17.4— L |
8.3 |
|
12 |
35.0 |
19.4 |
8.1 |
|
13 |
35.5 |
19.7 |
8.6— H |
|
14 |
37.4 |
18.9 |
7.9 |
|
15 |
33.6 |
19.5 |
— |
|
Average |
35.62 |
19.07 |
7.9 (of 14) |
Eggs No. 14 and No. 15 each had one end prolonged into a tip, and No. 15 was broken and leaking when found.
Four of the above eggs hatched in August, 1938, three on the 13th and the fourth on the 15th.
|
No. |
Length |
Weight |
|
1 |
309 |
6.3 |
|
2 |
294— L |
5.95— L |
|
3 |
336— H |
6.8— H |
|
4 |
303 |
6.4 |
|
Average |
310.5 |
6.36 |
Elaphe obsoleta obsoleta (.Say).
A female, probably fx-om near Philadelphia, deposited 11 eggs in cap- tivity on July 24, 1937. Five of these were separate but the others were adherent in pairs. Very probably the snake was disturbed during ovoposi- tion ; several specimens of other species were confined in the same cage with it.
|
No. |
Length |
Width |
Weight |
|
1 |
50.8 |
20.2 |
13.7 |
|
2 |
55.5— H |
19.7 |
14.95 |
|
3 |
43.8 |
22.4 |
12.23 |
|
4 |
48.0 |
18.2 |
10.15 |
|
5 |
44.9 |
17.4— L |
9.15 |
|
6 |
47.2)1 |
21.7 ) 1 |
27.8 |
|
7 |
46.3) |
23.7)— H |
|
|
8 |
47.7 ( |
22.5 \ |
26.7 |
|
9 |
45.5) |
22.3) |
|
|
10 |
45.6 1 |
22.5 ( |
25.8 |
|
11 |
43.5)— L |
22.6) |
|
|
Average |
47.16 |
21.2 |
12.77 |
Egg No. 4 had a marked constriction around its middle.
i Eggs which were adherent to each other. The measurements of eggs in such condition are grouped by brackets and their combined weights are expressed by one number, a policy followed throughout the text.
38
Zoologica: New York Zoological Society
[XXV :3
Two of the above eggs hatched on October 9, 1937.
|
No. Length |
Weight |
|
1 361 |
10.3 |
|
2 360 |
11.2 |
|
Average 360.5 |
10.75 |
|
Elaphe vulpina (Baird & |
Girard) |
|
Three large females collected by Robert H. Mattlin in May, 1938, at |
|||
|
Little Cedar Point, Lucas |
County, |
Ohio, all deposited eggs during July, |
|
|
1938. In two instances, many of the eggs were |
adherent, making it im- |
||
|
possible to measure them and necessitating the weighing of them in groups |
|||
|
as indicated in the tables below. In only one case |
was it possible to deter- |
||
|
mine which female laid which eggs. |
Her weight and length are recorded in |
||
|
their proper place. |
|||
|
I. Eggs laid on July 20, 1938. |
|||
|
No. |
Length |
Width |
Weight |
|
1 |
431 |
241 |
|
|
2 |
43| |
251 |
|
|
3 |
39 1 |
25 1 |
|
|
4 |
42 1' |
25 1 |
|
|
5 |
42 1 |
| |
|
|
6 |
43) |
l |
161.7 |
|
7 |
39| |
— [ |
|
|
8 |
46 1 |
24 1 |
|
|
9 |
40 |
1 |
|
|
10 |
41 i |
241 |
|
|
11 |
42 J |
24 J |
|
|
12 |
36— L |
24 |
13.8 |
|
13 |
44 |
24 |
16.1 |
|
14 |
47— H |
27 |
14.9 |
|
Average |
41.93 |
24.6 (of 10) 14.75 |
|
|
II. Eggs laid on July |
21, 1938. |
Length of female, 1229 mm.; weight, |
|
|
409.5 g. |
|||
|
No. |
Length |
Width |
Weight |
|
1 |
47.3 |
25.7 |
18.9 |
|
2 |
47.0 |
24.1 |
17.3 |
|
3 |
48.2 |
24.0 |
18.7 |
|
4 |
50.4 |
24.5 |
18.5 |
|
5 |
48.0 |
25.8— H |
19.3 |
|
6 |
42.9 |
23.7 |
14.8 |
|
7 |
38.11— L |
22.31— L |
|
|
8 |
48.9| |
' 24.01 |
|
|
9 |
51.2) |
24.1) |
84.8 |
|
10 |
44.81 |
24.2| |
|
|
11 |
58.5 1— H |
23.3 J |
|
|
Average |
47.75 |
24.15 |
17.48 |
Five of the eggs in this group hatched in September— three on the 1st and two on the 5th.
No. Length Weight
1 306 14.4
2 334— H 15.4— H
3 333 14.0
4 300— L 11.9— L
5 331 14.3
Average
320.8
14.0
1940]
Conant & Downs: Miscellaneous Notes on Reptiles
39
III. Eggs laid on July 23, 1938. All were separate, probably because the female moved about while depositing them. The moist substance with which they were covered when laid was sufficiently adhesive to make pebbles from the cage floor stick to several of the eggs.
|
No. |
Length |
Width |
Weight |
|
1 |
39.9 |
26.9 |
16.65 |
|
2 |
40.7 |
24.5 |
14.4 |
|
3 |
40.7 |
26.1 |
15.4 |
|
4 |
37.8 |
27.3— |
-H 16.2 |
|
5 |
44.4 |
23.5 |
14.4 |
|
6 |
29.1— L |
25.0 |
14.7 |
|
7 |
45.0 |
22.9— |
-L 15.0 |
|
8 |
47.2— H |
25.0 |
16.5 |
|
9 |
41.8 |
24.4 |
14.8 ■ i |
|
10 |
40.4 |
24.4 |
14.3 |
|
11 |
40.6 |
24.6 |
14.9 |
|
12 |
38.4 |
24.3 |
• 14.1 |
|
13 |
42.9 |
27.3 |
18.5 |
|
14 |
45.7 |
26.9 |
19.8— H |
|
15 |
42.8 |
24.3 |
14.5 |
|
16 |
39.9 |
26.1 |
15.8 |
|
17 |
43.1 |
27.2 |
13.9 |
|
18 |
41.9 |
26.3 |
18.1 |
|
19 |
35.7 |
25.5 |
13.6 |
|
20 |
38.1 |
24.6 |
13.4— L |
|
21 |
38.6 |
26.3 |
15.1 |
|
Average 40.7 |
25.38 |
15.43 |
|
|
Pituophis melanoleucus melanoleucus (Daudin) . |
|||
|
Two females from undetermined localities |
in southern New Jersey laid |
||
|
eggs in July, 1938. |
Many of the eggs in each |
group were adherent to one |
|
|
another, preventing complete measurements and separate weighings. |
|||
|
I. Eggs laid on |
July 12, 1938. |
Length of female, 1514 mm.; weight, |
|
|
952.3 g. |
|||
|
No. |
Length |
Width |
Weight |
|
1 |
531 |
341 |
|
|
2 |
49 1 |
35 1 |
|
|
3 |
— 1 |
33 1 |
|
|
4 5 |
48| 49| |
35 1 34 r |
285.85 |
|
6 |
53| |
||
|
7 |
-1 |
34 1 |
|
|
8 |
61 1 |
321 |
|
|
9 |
49 |
33 |
32.3 |
|
Average 51.71 (of 7) 33.75 (of 8) 35.35 |
Two of the above eggs hatched on September 19, 1938.
|
No. |
Length |
W eight |
|
1 |
412 |
35.8 |
|
2 |
376 |
26.8 |
|
Average |
394.0 |
31.3 |
40
Zoologica: New York Zoological Society
[XXV :3
II. Eggs laid on July 13, 1938.
|
No. |
Length |
Width |
Weight |
|
1 |
571 |
331 |
|
|
2 3 |
— l 58 f |
321 33 f |
155.0 |
|
4 |
59) |
341— H |
|
|
5 |
60) |
32) |
76.4 |
|
6 |
57) |
33) |
|
|
7 |
66 |
31— L |
39.9 |
|
Average |
59.5 (of 6) |
32.57 |
38.76 |
The eggs in this group started to hatch on September 18, and the last of the 7 emerged on the 20th. There were as many as 5 and 6 slits in some of the eggs and in one or two of them 2 or more slits crossed, to form loose flaps.
|
No. |
Length |
Weight |
|
1 |
433 |
36.5— H |
|
2 |
452— H |
36.5 |
|
3 |
416— L |
34.5— L |
|
4 |
433 |
35.4 |
|
5 |
416 |
35.0 |
|
6 |
429 |
35.0 |
|
7 |
437 |
36.1 |
|
Average |
430.86 |
35.57 |
The young snakes might be described as follows:
Ground color (above) yellowish-cream, becoming orange or brownish- cream on the sides. Dorsal blotches very dark brown (almost black), the most anterior dorsal ones enclosing two rather small, light areas. The orange or brownish coloration on the sides is caused by the anterior parts of the scales being of these colors. Some scales also have a longitudinal streak of the same colors on them.
Top of head olive-gray, marked with a few dark spots. A dark line across in front of the eyes, situated on the anterior edges of the frontal and supraoculars and the posterior edges of the prefrontals. A short, dark dash, obliquely downward from the back of the eye; another straight down from the eye. Some of the sutures between the upper labials dark. Belly whitish, washed with orange-pink.
Lampropeltis getulus florklana Blanchard.
Female collected 20 miles west of Miami, Florida, and sent to the Phila- delphia Zoological Garden through the courtesy of Dr. Thomas Barbour. Eggs were deposited on May 19, 1936. Length of female, 1047 mm.; weight,
282.2 g.
|
No. |
Length |
Width |
Weight |
|
1 |
47.7 |
22.0— L |
13.4 |
|
2 |
50.0— H |
22.6 |
13.8— H |
|
3 |
36.8 |
23.8— H |
11.2 |
|
4 |
38.9 |
22.9 |
11.0 |
|
5 |
43.2 |
23.0 |
13.2 |
|
6 |
33.0— L |
23.4 |
9.8— L |
|
7 |
38.9 |
23.2 |
12.3 |
|
8 |
40.4 |
22.9 |
12.4 |
|
9 |
38.6 |
23.6 |
12.0 |
|
10 |
47.0 |
23.0 |
13.6 |
|
11 |
44.3 |
22.8 |
13.4 |
|
Average |
41.70 |
23.01 |
12.37 |
1940]
Conant & Downs: Miscellaneous Notes on Reptiles
41
Egg No. 2 had a constricted nipple at one end and No. 6 had a flesh- colored area covering about one-third of the egg. All other eggs were plain cream white, with smooth, parchment-like shells. Numbers 7 to 11, inclu- sive, were adherent in a single cluster but wei’e separated for weighing and measuring. Several young had escaped from their shells when the clutch was examined on July 26, 1936, and the last emerged on July 27. Two eggs
|
were not fertile. No. |
Length |
Weight |
|
1 |
290 |
11.1 |
|
2 |
285 |
9.4— |
|
3 |
289 |
10.5 |
|
4 |
277— L |
10.8 |
|
5 |
300— H |
11.1 |
|
6 |
282 |
11.0 |
|
7 |
295 |
12.3— |
|
Average (of 7) |
288.2 |
10.89 |
|
8 |
232 |
7.8 |
|
9 |
216 |
5.5 |
|
Average (of 9) |
274 |
9.93 |
The last two specimens obviously were runts. In addition their sides were somewhat shrunken except in the region of the belly, which was noticeably swollen. They lived only a few days. The other specimens began eating DeKay’s snakes, Storeria dekayi, and one another almost at once; eventually only one specimen remained.
Rhinocheilus lecontei Baird & Girard.
Eggs were found in the cage on July 1, 1935. The female was purchased from a dealer and its origin is unknown.
|
No. |
Length |
Width |
Weight |
|
1 |
37.0 |
15.9 |
7.1 |
|
2 |
36.2 |
16.0— H |
6.7 |
|
3 |
37.2 |
16.0 |
6.6 |
|
4 |
41.0— H |
15.4 |
7.2— |
|
5 |
35.6 |
15.8 |
6.5 |
|
6 |
30.1— L |
15.3— L |
5.3—' |
|
Average |
36.18 |
15.73 |
6.57 |
42
Zoologica: New York Zoological Society
[XXV :3
Natrix cyclop-ion floridana Goff.
Two specimens from Marion County, Florida, bore young in captivity. I. Young born on July 27, 1936.
|
No. |
Length |
Weight |
|
1 |
249— H |
8.3— H |
|
2 |
248 |
7.0 |
|
3 |
229 |
6.3 |
|
4 |
230 |
5.9— L |
|
5 |
240 |
7.5 |
|
6 |
229 |
6.2 |
|
7 |
245 |
7.0 |
|
8 |
248 |
7.0 |
|
9 |
234 |
6.9 |
|
10 |
233 |
7.1 |
|
11 |
225 |
6.6 |
|
12 |
243 |
8.1 |
|
13 |
244 |
7.9 |
|
14 |
222— L |
6.0 |
|
Average |
237.07 |
6.99 |
This litter probably consisted of more than 14 specimens since some had escaped from the cage and were found on the floor. Number 4 had its hemipenes everted; No. 14 was dead. The pattern of the young snakes was more conspicuous than that of adults; it consisted of black reticulations on an olive-brown ground. Their bellies were plain yellow except for the antero- lateral edges of the ventrals, which were black.
II. Young born on August 7, 1936. Length of female, 1389 mm.; weight, 686.6 g.
|
No. |
Length |
Weight |
|
1 |
245 |
9.0 |
|
2 |
259— H |
9.2 |
|
3 |
241 |
7.6— L |
|
4 |
257 |
9.1 |
|
5 |
250 |
8.9 |
|
6 |
244 |
8.4 |
|
7 |
245 |
8.0 |
|
8 |
240— L |
8.7 |
|
9 |
246 |
8.7 |
|
10 |
251 |
9.6— H |
|
Average |
247.8 |
8.72 |
1940]
Conant & Downs: Miscellaneous Notes on Reptiles
48
Matrix erythrog aster transversa (Hallowell) .
A litter was born on September 5, 1933, to a female collected in Kansas and presented to the Toledo Zoological Park by Dr. Reeve M. Bailey. The parent measured 1071 mm. and weighed 335.1 g.
|
No. |
Length • |
Weight |
|
1 |
255 |
6.0 |
|
2 |
267 |
5.9 |
|
3 |
268 |
6.4— B |
|
4 |
256 |
S.l |
|
5 |
250— L |
5.6 |
|
6 |
267 |
6.0 |
|
7 |
262 |
8.1 |
|
8 |
268 |
6.4 |
|
9 |
257 |
6.2 |
|
10 |
2 70 |
6.2 |
|
11 |
264 |
6.1 |
|
12 |
261 |
6.1 |
|
13 |
266 |
6.3 |
|
14 |
267 |
6.3 |
|
15 |
258 |
5.8 |
|
16 |
255 |
6.0 |
|
17 |
265 |
5.9 |
|
18 |
271— H |
6.3 |
|
19 |
271 |
5.7 |
|
20 |
265 |
5.5— L |
|
Average |
263.15 |
6.05 |
An additional specimen (not recorded) was born dead. The living young might be described as follows :
Pattern (above) consists of a dorsal and lateral series of blotches alternating clear forward to the head. In the center of each light area, on the sides between the black blotches, is a large area of rose-brown, narrowly edged with white. Belly uniform white except for the antero-lateral edges of the ventrals, which are black.
These small snakes began eating chopped fish almost at once. As they grew they were measured and weighed at intervals. On December 9, 1933, they averaged 259.3 mm. and 6.61 g. ; on April 5, 1934, 19 of them (one having died) averaged 306.8 mm. and 10.90 g. ; on June 13, 1934, 331.6 mm. and 13.33 g. All but 3 had died by May 2, 1935, when the survivors aver- aged 420 mm. and 26.70 g. These snakes did not hibernate; they were kept warm and active over both winters.
Several young of this subspecies were born at the Toledo Zoological Park during September, 1930, to a female from New Braunfels, Texas. They grew rapidly in captivity on a diet of fish, but as a result of cannibalism the group eventually was reduced to one. This, a female, passed a large, red, in- fertile ovum on September 2, 1933, and two others on September 4, 1933. Thus it would appear that the snake was sexually mature at the age of three years. At that time it weighed 429.7 g. and measured 960 mm. in length. While it was still light in color, its pattern had become rather indistinct and the contrast between blotches and ground color was considerably less- ened. The light crossbands between the dorsal blotches were the most prominent feature of the pattern.
44
Zoologica: New York Zoological Society
[XXV: 3
Natrix septemvittata (Say).
Two females of this species from Detroit, Michigan, gave birth to young on August 23, 1938. Since both females were in the same cage it was im- possible to determine how many each had borne.
|
No. |
Length |
Weight |
|
1 |
212 |
2.8 |
|
2 |
221 |
2.6 |
|
3 |
222 |
3.3— H |
|
4 |
215 |
2.6 |
|
5 |
222 |
3.2 |
|
6 |
223 |
3.3 |
|
7 |
227 |
3.1 |
|
8 |
226 |
3.2 |
|
9 |
225 |
2.8 |
|
10 |
218 |
2.7 |
|
11 |
216 |
2.6 |
|
12 |
206— L |
2.3— L |
|
13 |
230— H |
3.1 |
|
Average |
220.2 |
2.89 |
Another female from Delaware County, Pennsylvania, bore 2 young and passed 2 dead embryos and 2 infertile ova on August 22, 1938. The female weighed 52.9 g. and measured 688 mm. Data on the 2 living young are:
No. Length Weight
1 168 1.4
2 206 2.5
Average 187 1.95
No. 1 was deformed.
1940]
Conant & Downs: Miscellaneous Notes on Reptiles
45
Natrix sipedon pictiventris Cope.
Young were born on August 2, 1936, to a specimen from near Palmetto, Florida, collected by the late C. C. Goff. The female weighed 368.9 g. and measured 771 mm. in length. Most of its tail was missing.
|
No. |
Length |
Weight |
|
1 |
211 |
3.3 |
|
2 |
213 |
3.4 |
|
3 |
194 |
3.2 |
|
4 |
215 |
3.5 |
|
5 |
206 |
3.5 |
|
6 |
208 |
4.0 |
|
7 |
217 |
3.9 |
|
8 |
213 |
4.2 |
|
9 |
215 |
3.7 |
|
10 |
180— L |
2.9— L |
|
11 |
210 |
4.3 |
|
12 |
211 |
4.0 |
|
13 |
199 |
3.7 |
|
14 |
205 |
3.4 |
|
15 |
223— H |
3.8 |
|
16 |
217 |
4.2 |
|
17 |
201 |
4.0 |
|
18 |
200 |
3.9 |
|
19 |
206 |
4.4— H |
|
20 |
199 |
3.7 |
|
21 |
221 |
3.7 |
|
22 |
192 |
3.2 |
|
23 |
206 |
3.7 |
|
24 |
204 |
3.5 |
|
25 |
218 |
4.1 |
|
Average |
207.36 |
3.73 |
No. 24 and No. 25 were dead.
[XXV: 3
46 Zoologica: New York Zoological Society
Thamnophis sirtalis sirt'alis (Linnaeus).
A specimen from Philadelphia, Pennsylvania. Young born on August 4, 1936. Length of female. 706 mm.; weight, 92.2 g.
|
No. |
Length |
Weight |
|
1 |
143 |
1.1 |
|
2 |
142 |
0 9—1. |
|
3 |
156 |
1.1 |
|
4 |
155 |
1.0 |
|
5 |
156 |
1.2— H |
|
6 |
155 |
1.0 |
|
7 |
162 |
1.2 |
|
8 |
164 |
1.2 |
|
9 |
155 |
1.1 |
|
10 |
157 |
0.9 |
|
11 |
147 |
1.1 |
|
12 |
149 |
1.0 |
|
13 |
162 |
1.1 |
|
14 |
153 |
1.1 |
|
15 |
164 |
1.2 |
|
16 |
158 |
1.0 |
|
17 |
150 |
1.1 |
|
18 |
165 |
1.2 |
|
19 |
161 |
1.1 |
|
20 |
160 |
1.1 |
|
21 |
168 |
1.0 |
|
22 |
170— H |
1.1 |
|
23 |
158 |
1.0 |
|
24 |
166 |
1.1 |
|
25 |
152 |
1.0 |
|
26 |
160 |
1.2 |
|
27 |
158 |
1.1 |
|
28 |
157 |
1.1 |
|
29 |
149 |
1.0 |
|
30 |
155 |
1.0 |
|
31 |
140— L |
0.9 |
|
Average |
156.35 |
1.07 |
No. 12 had a deformed back. No. 31 was dead.
Crotalus viridis viridis Rafinesque.
Young born to two females received from South Dakota, collected by A. M. Jackley in August, 1938.
I. Born on September 7, 1938. Parent weighed 297.4 g. and measured
|
mm. in length, to the base of |
the rattle. |
|
|
No. |
Length |
Weight |
|
1 |
281 |
14.9 |
|
2 |
282 |
13.8— L |
|
3 |
286 |
15.1 |
|
4 |
290— H |
15.2 |
|
5 |
273 |
14.2 |
|
6 |
274 |
14.3 |
|
X |
251— L |
14.1 |
|
8 |
259 |
14.2 |
|
9 |
270 |
14.8 |
|
10 |
279 |
15.6— H |
|
11 |
277 |
15.5 |
|
Average |
274.73 |
14.7 |
The eyes of all were overcast, indicative of an approaching moult.
1940] Conani & Downs: Miscellaneous Notes on Reptiles 47
II. Young- born on September 10, 1938; two dead embryos also passed. Female undertermined; several others in cage.
|
No. |
Length |
Weight |
|
1 |
256 |
14.8 |
|
2 |
271— H |
15.6— H |
|
9 O |
237— L |
11.6— L |
|
4 |
249 |
12.8 |
|
5 |
253 |
12.8 |
|
Average |
253.2 |
13.52 |
The “button” was included in measuring the young of both litters.
Sternotherus odoratus (Latreille).
A turtle of this species was collected early in June, 1936, by Byron Gardner, Jr., who found it in South Carolina directly across the Savannah River from Augusta, Georgia. It deposited 2 eggs on June 22.
|
No. |
Length |
Width |
Weight |
|
1 |
25 |
15.1 |
3.6 |
|
2 |
23.2 |
15.4 |
3.5 |
|
Average |
24.1 |
15.25 |
3.55 |
On October 2, 1936, a baby was found walking about in the hatching media. That it may have emerged at some time previous is indicated by the fact that the other egg also had hatched, but the young had died and was badly dessicated. Data on the living one are as follows:
■'} r Carapace Depth of Plastron
Length Width Shell Length Weight
22 18.9 12 15.8 2.2
Chelydra serpentina serpentina (Linnaeus).
A large snapper collected 3 miles north of New Castle, Delaware, on April 24, 1936, deposited 3 eggs in its tank on June 22, 1936.
No. Diameter Weight
1 27.2 10.8— H
2 26.8 — L i. 9.7— L
3 29.5— H 10.7
Average 27.83 10.4
Since the nature of the tank prevented this turtle from nesting, the 3 eggs doubtless represent only a fraction of the entire complement. The specimen was transferred to a spacious, out-door enclosure but whether other eggs were laid was not determined.
48
Zoolog tea: New York Zoological Society
Terrapene Carolina Carolina (Linnaeus).
A set of eggs probably found near Philadelphia, was given to the Phila- delphia Zoological Garden by an anonymous donor. Young hatched on Sep- tember 1 and 2, 1938.
Carapace
|
No. |
Length |
Width |
Depth of Shell |
Weight |
|
1 |
34.6 |
33.3 |
16.6—1, |
7.9 |
|
2 |
33.3— L |
32.1— L |
16.7 |
7.8 |
|
3 |
34.9— H |
33.5 |
17.0— H |
8.4—1 |
|
4 |
33.7 |
33. 7— H |
16.8 |
7.6— |
|
Average |
34.13 |
33.15 |
16.78 |
7.93 |
Pseudemys floridana ssp.
A female from central Florida laid 2 eggs in its tank on June 7, 1936.
|
No. |
Length |
Width |
Weigh t |
|
1 |
33.0 |
24.3 |
9.2 |
|
2 |
35.7 |
24.5 |
8.7 |
|
Average |
34.35 |
24.4 |
8.95 |
As with the specimen of Chelydra s. serpentina above, conditions were not optimum for laying and the 2 eggs doubtless are but a part of the number the turtle might have been expected to lay.
Testudo tornieri Siebenrock.
Four of these tortoises were secured from a dealer on August 7, 1935. One of them, weighing 375.9 g., carapace length 144 mm., carapace width 106 mm., and depth of shell 38 mm., deposited an egg on January 9, 1937. This weighed 33.5 g. and measured 48 mm. in length by 28 mm. in width. As Loveridge (1928, p. '51 ) has noted, this remarkable turtle has a com- plement of only a single egg.
References.
Klauber, L. M.
1938. Notes from a Herpetological Diary, I. Copeia, No. 4, pp. 191-7. Loveridge, Arthur
1928. Field Notes on Vertebrates Collected by the Smithsonian-Chrysler East African Expedition of 1926. Proc. U. S. Nat. Mus., Vol. 73, Art. 17, pp. 1-69, pis. 1-4.
Jaros: Occlusion of Venom Duct
49
4,
Occlusion of the Venom Duct of Crotalidae by Electrocoagulation: an Innovation in Operative Technique.1
Duval B. Jaros
Riverside, Illinois.
(Text-figure 1).
An article by Tait2 on the surgical removal of the venom glands of rattlesnakes prompted Mr. Emil Rokosky of the Chicago Zoological Park to suggest the use of electrocoagulation on the venom gland. It was thought, however, that occlusion of the duct would be a less drastic procedure than removal of the gland. In addition, conservation of the gland would permit interesting experiments and histologic studies. The immediate problem pre- sented was the development of the technical details of a rapid operation which would occlude the duct, preserve the gland and result in minimum injury to the snake. In the procedure here outlined, electrocoagulation is employed.
Electrocoagulation produces destruction of tissue by heat, and corres- ponds to the solidifying of the white of an egg during boiling. The heat of electrocoagulation, however, is not conducted but is induced within the tissue itself, which insures coagulation without even surface carbonization. The coagulum is sterile and simulates a scab, nature’s own wound covering. Sterility, firmness and pliability of the coagulum permit healing without scarring.
Essentials of an electrocoagulating device are two electrodes and a high frequency current. The electrodes may be in the form of a single duo- terminal instrument, a body-plate and a point, or as in the present experi- ment, two separate uniterminal instruments with one serving also as the probe for securing the duct.
The high frequency current alternates at about one million cycles per second. Suitable currents can be obtained from simple spark-gap or ther- mionic tube apparatuses sold by all physiotherapy equipment dealers. The coagulating current of 800 kilocycles, 150 milliamperes and 2000 volts, seals the severed ends of the duct, and is therefore preferable to the cutting current.
In the operation, several innovations employed made general anaesthesia unnecessary and facilitated single-handed performance with relative safety to the operator and the snake. The first device, an operating board (Text- fig. 1) was designed to hold the snake firmly and prevent it from thrashing. The board was fitted with a neck pad, strap and a thin rubber restraining flap for the body. When in use, this flap was drawn over the snake and at- tached along the opposite edge of the board by a zipper running its entire
1 Presented at a meeting of the Amateur Herpetologist Group of the Chicago Academy of Sciences, June 15, 1939, and at the annual meeting of the American Society of Ichthyologists and Herpetologists, Chicago, September 14, 1939.
2 Copeia, 1938:1.
50
Zoologica: New York Zoological Society
[XXV :4
length. Pressed down by the restraining sheet which yields to its struggles, the snake cannot crawl. This operative convenience can be augmented by having the surface of the board on which the snake rests very smooth or covered with yielding rubber similar to the overlying restraining sheet.
A second aid was the surgical skin clip or Michel clip. These clips, small strips of metal clawed at each end and preferably bent into a V shape, when pressed point to point, firmly grasp any interposed tissue. Applied to the snake’s lips, they hold the snake’s mouth shut and make handling rela- tively safe without the use of general anaesthesia.
Operating board designed to hold a snake firmly while the venom duct is occluded.
Another innovation was an improvised scalpel which proved, for my purpose, to be superior to any other obtainable. These scalpels were made from slivers of the cutting edges of Gillette “Thin Blades,” broken off with pliers and clamped in ordinary artery forceps. With the forceps serving as handles, these tiny bayonettes afforded a generous supply of extremely sharp, delicate and inexpensive knives.
The Operation.
The snake, grasped behind the head, was fastened to the operating board, and Michel clips were clamped on the lips ; one on each side below the pit. Care was taken not to include the mandible.
At this time, if desired, a local anaesthetic may be injected at the site of the incision. A 2% procain borate solution was used. Almost imme- diately after the injection a quarter-inch incision was made just below the eye and on a level with the pit. Through this small incision the duct was secured and drawn out with a blunt curved probe. When clear of the sur- rounding tissue, it was coagulated sufficiently to sever it. The severed ends were dropped back into place and the wound was touched with tincture of merthiolate. The wound was so small, and there was so little trauma, that sutures were unnecessary and healing was remarkably rapid. In many cases it was difficult to locate the site of incision after one week.
As the snake is not deprived of glands or fangs, the outward appearance of the head remains unaltered. In the specimens under observation no change in disposition was noticed; feeding remained regular in those which ate regularly before treatment. Each specimen was tested weekly by allow- ing it to bite a rabbit. In every case the result was negative. A cotton- mouth moccasin, ( Agkistrodon piscivorus ) in good health forty-five weeks after treatment and eating well, was nevertheless unable to inflict other than mechanical injuries when biting. Deaths resulted from causes other than the operation — from mouth rot and intestinal disorder.
1940]
Jar os: Occlusion of Venom, Duct
51
The obstructed glands taken from snakes three, seven and eight weeks, and ten months, after treatment were sectioned for microscopic examination. The sections were studied by Dr. George J. Rukstinat, Associate Professor of Pathology at Rush Medical College. He found large acinar spaces filled with secretion, and lined by cells which evidently became cuboidal as a result of the retained venom; but even the compressed cells showed little signs of atrophy. In portions of the sections normal columnar cells survived. Nuclei and cytoplasm were demarcated clearly. Confluence of acini in some regions seemed apparent from the remnants of acinar walls which projected into the lumens of the cyst-like spaces. Many cells contained clear or slightly stippled vacuoles, the character of which is being investigated.
The histological evidences of cell viability were corroborated by the sub-cutaneous injection into mice of secretion from the glands of one snake which was killed ten months after occlusion of the duct. Minute quantities (0.02 cc.) of this secretion proved fatal to mice in four hours, and induced the usual changes of a general haemolysis and of local necrosis.
The potentialities for study of the venom gland through this method are numerous. Perhaps some internal absorption of secretion occurs; or perhaps some specialized cells elaborate components of venom as do specialized cells in organs such as the pancreas.
Summary.
A method of rendering venomous serpents harmless is presented. Elec- trocoagulation is used to prevent, by the destruction of a portion of the duct, the escape of venom from the gland and its passage to the fang. The operator can manage, unassisted and with safety, the entire procedure. The snake is but slightly injured and is not disfigured. The desired results are apparently permanent. The habits of none of the thirty-six snakes treated seemed in any way effected by the operation, even after forty-five weeks. The operation was successfully performed on crotalids ranging in size from eight inches to four feet.
In the performance of the operation three innovations were introduced: a practical device for holding the snake, a cheap and efficient operating knife, and the use of Michel surgical clips to eliminate all danger of being bitten.
I wish to express my grateful acknowledgement to Dr. George J. Ruks- tinat, to Mr. Walter L. Necker of the Chicago Academy of Sciences, and to Mr. Emil J. Rokosky, for advice and encouragement; and to my father, Dr. Joseph F. Jaros, for his help with the technical aspects of the work.
Tee-Van: Review of American Cirrhitidae
53
5.
Eastern Pacific Expeditions of the New York Zoological Society. XVII.
A Review of the American Fishes of the Family Cirrhitidae.'
John Tee-Van
Department of Tropical Research, New York Zoological Society.
(Plate I; Text-figures 1-4).
[This is the seventeenth of a series of papers dealing with the collections of the Eastern Pacific Expeditions of the New York Zoological Society made under the direction of Dr. William Beebe. The present paper is concerned with specimens taken on the Templeton Crocker Expedition (1936), the Eastern Pacific Zaca (1937-1938) and the Arcturus Oceanographic (1925) Expeditions. For data on localities, dates, dredges, etc., of these expeditions, refer to Zoologica, Vol. VIII, No. 1, pp. 1-45 ( Arcturus ) ; Zoologica, Vol. XXII, No. 2, pp. 33-46 (Templeton Crocker) ; and Zoologica, Vol. XXIII, No. 14, pp. 278-298 (Eastern Pacific Zaca ).]
Contents.
Page
Introduction 53
Key for the differentiation of the American genera (Family Cirrhitidae) 54
Cirrhitus 54
Page
Cirrhitus rivulatus (Valenciennes) 54
Cirrhitichthys 58
Cirrhitichthys corallicola new species.... 58
Pseudocirrhites 61
Pseudocirrhites pinos Mowbray 61
Introduction.
For many years the only representative of the family Cirrhitidae known from the Americas was the west coast Cirrhitus rivulatus (Valenciennes). In 1927, however, Mowbray described a species from the Isle of Pines, Cuba, establishing a new genus, Pseudocirrhites, for his fish, and marking the first West Indian record for the family. During the 1937-1938 expedition of the Department of Tropical Research of the New York Zoological Society along the west coast of Mexico and Central America, still another form was found; it is described as a new species in this paper. Specimens of all three species have been examined and are reviewed and described herewith.
In the light of the recent discovery of a cirrhitoid fish in West Indian waters, the finding of a closely related form on the west coast of America tends to explain the apparent isolation of the Atlantic fish and provides an- other link in the chain of evidence that demonstrates the close relationship of the fishes of the Atlantic and Pacific sides of tropical American waters.
1 Contribution No. 590, Department of Tropical Research, New York Zoological Society.
54
Zoologica: Neiv York Zoological Society
[XXV :5
For an explanation of the position of localities mentioned in this paper, see Text-fig. 4.
I am indebted to Dr. Leonard P. Schultz of the U. S. National Museum and Prof. Albert E. Parr of the Peabody Museum, Yale University, for the loan of specimens, and to Miss Janet B. Wilson, who made the drawings used in this paper.
Family Cirrhitidae.
Key for the differentiation of the American genera.
la. Scales on the cheek very small, 16 to 20 rows from eye to edge of pre-
opercle (Pacific) Cirrhitus.
lb. Scales on the cheek large, 4 to 5 rows from eye to edge of preopercle. 2a. Scales absent on nape and interorbital space ; a small patch of scales
behind and between the anterior nosti’ils (Pacific) ... Cirrhitichthys.
2b. Scales present on nape and interorbital space. (Atlantic)
Pseudocirrhites.
Cirrhitus Lacepede, 1803.
Genotype by monotypy, Cirrhitus maculatus Lacepede, Lacepede, 1803, Hist. Nat. des Poiss,, 5: 2.
Cirrhitus rivulatus (Valenciennes).
(Plate I, Text-fig. 1).
Proportions in Percentage of the Standard Length, and Counts: Depth 34-36 (2.4-3) 2 ; head 38-40 (2.46-2.75); eye 7.7 in fish of 119 to 182 mm., 9.2 in 65 mm. fish, 11-11.4 in 30 to 35 mm. fish (3-5.7) ; snout 11-14.7 (2. 3-3.1) ; maxillary 13.6-17.6 (2. 1-2. 8) ; interorbital space 5-6.5 (5.4-7.66) ; pectoral length 25.4-28.6 in fish of 119 to 182 mm., 32.2-34 in 31 to 65 mm. fish; pelvic length 20-23; snout to 1st dorsal fin 38-41; snout to 2nd dorsal fin 64-68; snout to anal fin 68-73; snout to pectoral fin 34-38; snout to pelvic fin 45-51; 1st dorsal height 11.7-14.6; 2nd dorsal height 12.2-15.4; anal height 17.4-20.6. Dorsal fin X, 11 or 12; anal fin III, 6 or 7 ; pectoral fin with 14 rays, the uppermost and the lower 7 (rarely 8) simple; scales: 5% to 6V2 from origin of 1st dorsal fin to lateral line, 46 to 49 in a longitudinal series, 12 to 14 scales from lateral line to origin of anal fin, 7 to 8 predorsal scales; gill-rakers. 5 to 0 on upper limb and 10 to 11 on the lower limb of the first arch.
Body compressed, sturdily and heavily built, the caudal peduncle deep. Anterior profile strongly convex, especially in larger individuals, the eye entering the profile very slightly. Scales present on the body, absent on the head except for the preopercle, which is very finely scaled, and the opercle which has scales on its anterior part similar to those of the body; somewhat smaller scales on the posterior flap of the opercle. Lateral line continuous, the openings of the canals tilted upward.
Interorbital space concave; a low longitudinal crest on the nape; pre- opercle broadly rounded, its upper portion finely serrate, the serrations most conspicuous in small specimens; opercle with an obtuse flap extending back- ward over the pectoral base. Branchiostegal rays 6, the innermost one very small; the branchiostegal membranes broadly united but free from the isthmus; mouth low, the lips thick in large specimens; maxillary extending backward to the vertical of the posterior border of the orbit in large fish and to the center of the orbit in small fish. Upper jaw with small, rather widely spaced canines, the posterior tooth and three or four near the symphysis en-
2 Figures in parentheses are proportions stated in terms of times in the head or standard length.
1940]
Tee-Van: Review of American Cirrhitidae
55
larged; inside of these teeth is a band of smaller teeth, the band widest anteriorly, becoming narrower laterally and vanishing posteriorly. Lower jaw with similar teeth but with two or three of the outer canine-like teeth near the center of each side of the jaw enlarged. Tongue large, wide and free anteriorly. Gill-rakers short, covered with fine asperities.
Spinous dorsal fin rather low, the spines increasing in size to the 5th and 6th, then decreasing, the ultimate spine longer than the penultimate; soft dorsal fin higher than the spinous ; caudal fin truncate; or slightly con- cave ; anal fin with the second spine largest and heaviest, the rays long, twice the length of the spines; pectoral fin with its lower rays considerably swollen, especially in large individuals, the 5th and 6th from the bottom longest, posterior edge of the branched rays oblique, the uppermost simple ray shortest of all; pelvic fins originating under the center of the adpressed pectoral fin, the tip of the pelvics reaching to the vent or slightly beyond.
Color: During growth there is considerable change in the coloration of this species, the principal alterations being caused, first, by the breaking up of the simple vertical bands of the young into groups of spots bordered with darker and lighter, the spots often acquiring a different basal color; second, the shift from the pearly-gray ground color of the very young to the browns or greenish-browns of the adults; third, the loss of scarlet on the upper anterior surface markings of the head and the disappearance of the scarlet, black and gray spinous dorsal fin.
The original description of the young, (nominal Cirrhitus betaurus), given by Gill in 1862, agrees excellently with our similar-sized specimens (35 mm.). Gill’s description is appended here: “The color is ivhitish on the body, blackish on the shoulders and from the dorsal fin to the eyes, and with four complete, oblique, blackish bands; the first under the middle of the spinous dorsal; the second under the last spine; the third under the middle of the soft dorsal, and the fourth encircling the caudal peduncle. The head has three lateral bands, one on the preorbital region, a second on the cheek, and a third on the posterior margin of the preoperculum. The operculum has a longitudinal oblong spot. The chin has four spots forming the angles of a rhomb, and there is another one behind, on the branchiostegal membrane near the margin. The spinous dorsal is margined with blackish, and the two bands beneath more or less ascend on it; anal blackish. The caudal has a blackish B-shaped mark and a band at its base divided by the lateral line. The pectoral is dusky, with a black spot at its base nearly surrounded by a clear area, and separated from a spot in front of the base. The ventrals are blackish, with nearly transparent sides and margin.”
A color plate made in the field from a living 35 mm. specimen shows the following slight differences from this color description : The ground color is pearly gray, the spinous dorsal fin is scarlet with a blackish narrow upper border, the vertical body bands being continued upward onto the fin. All of the dark markings on the head and nape have a central core of reddish- brown. Jordan (Fishes of Sinaloa, Proc. Cal. Acad. Sciences, (2) 5: 473) states of similar-sized small individuals: “First dorsal fin bright orange red in life; second reddish; cross bands on body black.”
In a 65 mm. standard length fish the ground color is much duller and the bands on the sides are starting to break up into spots. This is especially noticeable on the dorsal surface of the body near the dorsal fin.
In our series the next largest specimen (119 mm.) has the pattern and coloration of all of the remaining larger specimens.
The adults may be described as follows, the description being combined from other descriptions and our own observations. Color of body dark green- ish- or yellowish-brown, occasionally more ochraceous brown on the lower head and belly (dark gray: Fowler). Dark brown, spots and bands, occasion- ally becoming yellowish-brown, usually bordered first with darker brown or
56
Zoologica: New York Zoological Society
[XXV :5
Text-figure 1.
Cirrhitus rivulatus Valenciennes. Specimens of 35, 65, 142 and 182 mm. standard length, showing the changes of pattern with age.
black and then with pale blue, distributed as follows: three narrow bands crossing the interorbital space; two to three bands from the eye to the maxillary or premaxillary; a saddle-like transverse band on the nape pos- terior to the eye, in the center of which, posteriorly, is a small circular spot; on the nape are three longitudinal short bars, one on the mid-dorsal line and one on either side of this; a band from eye to supra-scapular ; one from eye to upper border of the preopercle; a vertical band on the preopercle anteriorly and another on the posterior border ; an upwardly ascending band from the upper edge of the preopercle toward the upper tip of the opercle; a short vertical band on the base of the pectoral fin and another similar one
1940]
Tee-Van: Review of American Cirrhitidae
57
on the pre-pectoral region. Body with five upright oblique bands of similarly colored and bordered broken spots, the bands extending onto the spinous and soft dorsal fins; in addition there is a similar band on the posterior end of the caudal peduncle. Caudal fin with sub-circular brown spots bordered with darker brown and light blue, the spots roughly forming vertical and semicircular vertical lines; this arrangement is not always apparent, in large specimens. In some descriptions the tail is described as being dark with a network of pale blue reticulated lines. Dorsal fin dark brown, somewhat mottled with continuations of the bars of the sides of the body; pectoral fins dusky; anal fin dark brown, sometimes green basally, with two to five prom- inent spots similar to those on the body on its posterior portion; pelvic fins dusky, especially toward the tip, sometimes olive basally, gray-black termin- ally.
Iris in life (327 mm. fish) olive green above and below with a broad silvery zone extending longitudinally, within which are two rounded spots in front and two more behind the pupil. A narrow green area immediately surrounding the pupil.
Range: Pacific mainland from Lower California and the Gulf of Cali- fornia southward to Panama (Mexico: Cape San Lucas, Gulf of California, Mazatlan, Sihuatanejo, Acapulco; Nicaragua: Corinto; Costa Rica: Piedra Blanca Bay, Uvita Bay; Panama: Panama) ; Revillagigedo Islands (Clarion Is., Socorro Is.) ; Galapagos Islands (Hood Is., James Is., Tower Is.) ; Malpelo Island.
Local Distribution : A rocky reef and tidepool species, hiding in crevices and darting out for prey.
Method of Capture: Hook and line baited with bait or with a shiny piece of metal, traps, poisoning in tidepools.
Size: Grows to ITV2 inches. A 327 mm. (I2V2 inches) fish weighed 3 pounds and a 450 mm. ( 171/2 inches) fish weighed 5 pounds.
Shidy Materials: 8 specimens from 31 to 450 millimeters, from the following localities: Nicaragua: Corinto; Costa Rica: Piedra Blanca Bay, Uvita Bay; Galapagos Islands: Tower Island. In addition, water-glass sight records were made of this species at the following places: Mexico: Sihuatanejo, Acapulco; Clarion Island.
References: Cirrhites rivulatus Valenciennes, Voyage autour du Monde, sur la fregatte “La Venus,” tome 5, Ichthyologie, 1855: 309, plate 3, fig. 1 (Descrip- tion, color, poor plate: type locality, Galapagos Islands .1 ; Gunther, Catalogue of the Acanthopterygian Fishes in the Collection of the British Museum, 2, 1860: 519 (Short description) ; Jordan, D. S. and Gilbert, C. H., List of Fishes col- lected at Mazatlan, Mexico, by Charles H. Gilbert, Proc. U. S. Nat. Mus., 2, 1882 (1883) : 108 (Check-list, name only) ; Jordan, D. S., A list of the fishes known from the Pacific coast of tropical America, from the Tropic of Cancer to Panama, Proc. U. S. Nat. Mus., 8, 1885 (1886) : 381 (Check-list; Cape San Lucas, Galapagos Islands) ; Jordan, D. S. and Evermann, B. W., The Fishes of North and Middle America, Bull. U. S. Nat. Mus., 47 (2), 1898: 1491 (Description, color, range, short synonymy) ; Jordan, D. S. and McGregor, R. C., List of fishes collected at the Revillagigedo Archipelago and neighbouring islands, Rept. U. S. Fish Comm, for 1898 (1899) : 283 (Clarion and Socorro Islands) ; Jordan, D. S. and Evermann, B. W., The Fishes of North and Middle America, Bull. U. S. Nat. Mus., 47 (4), 1900: plate 227, fig. 576 (figure); Pellegrin, J., Poissons recueillis par M. Leon Diguet dans le Golfe de Californie, Bull. Mus. Hist. Nat. (Paris), 7, 1901: 163 (Gulf of California); Gilbert, C. H. and Starks, E. C., The Fishes of Panama Bay, Mem. Calif. Acad. ScL, 4, 1904: 139 (Panama; restates Gunther’s 1868 record) ; Beebe, W., Galapagos, World’s End, G. P. Putnam’s Sons, New York and London, 1924: plate 5 (colored figure) ; Ulrey, A. B., A check -list of the fishes of Southern California and Lower California, Journ. Pan-Pacific Res. hist., 4 (4), 1929: 18 (Check-list only; Cape San Lucas) ; Terron, C. C., Lista de los peces de los costas de la Baja California, Ann. Inst. Biol., Univ. Nac. Auton. Mexico, 3, 1932: 79 (Check-list, name only, Cape San
58
Zoological New York Zoological Society
[XXV :5
Lucas) ; Breder, C. M., Jr., Heterosomata to Pediculata from Panama to Lower California, Bull. Bingham Oceanogr. Coll., 2 (3), 1936: 37 (Unknown locality).
Cirrhitus rivulatus : Gill, T., Synopsis of the family of Cirrhitoids, Proc. Acad. Nat. Sci. Phila., 1862: 107 (Name, synonymy, range); Gill, T., Catalogue of the fishes of Lower California, in the Smithsonion Institution, collected by Mr. John Xantus, Proc. Acad. Nat. Sci. Phila., 1862: 259 (Cape San Lucas, LoWer California) ; Snodgrass, R. E. and Heller, E., Shore fishes of the Revil- lagigedo, Clipperton, Cocos and Galapagos Islands, Proc. Wash. Acad. Sci., 6, 1905: 385 (Galapagos Islands, range, few proportions) ; Beebe, W., The Arcturus Adventure, G. P. Putnam’s Sons, New York and London, 1926: 150 (Hood Island, Galapagos; method of capture). See p. 434 for specific determination; Jordan, D. S., Evermann, B. W. and Clark, H. W., Check list of the fishes and fishlike vertebrates of North and Middle America north of the northern boundary of Venezuela and Colombia, Rep. U. S. Comm, of Fish, for 1928 (1930) : 358 (Check list, range) ; Fowler, H. W., The Fishes of the George Vanderbilt South Pacific Expedition, 1937, Acad. Nat. Sci. Phila., Monograph No. 2, 1938: 15 (Description, color; Malpelo Island), 53, (Note on color, proportions; James Island, Galapagos), 257, (check list).
Cirrhitus betaurus : Gill, T. N., Catalogue of the fishes of Lower California in the Smithsonian Institution, collected by Mr. John Xantus, Proc. Acad. Nat. Sci. Phila., 1862: 259 (Original description, color; type locality: Cape San Lucas, Lower California) ; Jordan, D. S. and Gilbert, C. H., Catalogue of the fishes collected by Mr. John Xantus at Cape San Lucas, which are now in the United States National Museum, with descriptions of eight new species, Proc. U. S. Nat. Mus., 5, 1882 (1883): 371 (Synonymized with Cirrhitus rivulatus).
Cirrhitichthys rivulatus: Gunther, A., An account of the Fishes of the States of Central America, based on collections made by Captain J. M. Dow, F. Godman, Esq., and 0. Salvin, Esq., Trans. Zool. Soc. London, 6 (7), 1868: 387 (Check list); Galapagos Islands, Panama), 421, plate 86, fig. 4 (Panama, description, figure).
Cirrhites betaurus: Jordan, D. S., The fishes of Sinaloa, Proc. Calif. Acad. Sci., (2) 5, 1895: 472 (Relationship of betaurus and rivulatus discussed, color, Mazatlan) ; Jordan, D. S. and Evermann, B. W., The Fishes of North and Middle America, Bull. U. S. Nat. Mus., 47, 1898: 1492 (Description, color, range); Ulrey, A. B., A check-list of the fishes of Southern California and Lower California, Journ. Pan-Pacific Res. Inst., 4 (4), 1929: 18 (Check list only. Cape San Lucas) ; Terron, C. C., Lista de los peces de las costas de la Baja Cali- fornia, Ann. Inst. Biol., Univ. Nac. Anton. Mexico, 3. 1932: 79 (Check list,, name only; Cape San Lucas).
Cirrhitichthys Bleeker, 1856.
Genotype by original designation Cirrhites graphidopterus— Cirrhites aprinus Cuv. and Val. ; Bleeker, Naturk. Tijdschr. N ederl. -Indie, Deel X (new series, Deel VII) 1856: 474. The generic description was published in 1857 by Bleeker, Vischfauna van Ambonia, Acta Soc. Sci. Indo-Nederl., 2. 1857: 39.
Cirrhitichthys corallicola sp. nov.
(Text-figure 2) .
Type : Holotype, No. 28,710a, Eastern Pacific Zaca Expedition of the Department of Tropical Research, New York Zoological Society; Gorgona Island, off the Pacific coast of Colombia, South America (Lat. 2° 58' N., Long. 78° 11' W.), in coral, March 30, 1938; standard length 58 mm. Para- types: 32 specimens, No. 28,710, same locality and date as the holotype, 22 to 59 mm. standard length. Types in the collections of the Department of Tropical Research, New York Zoological Society.
Measurements in Percentage of the Standard Length, and Counts: Measurements of the holotype: Depth 36.6 (2.7) ; head 34.5 (2.9) ; eye 8.4
1940 j
Tee-Van: Review of American Cirrhitidae
59
(4.1 in head;; snout 11.2 (3.1 in head); interorbital space 5.7 (6.1 in head); maxillary 13.3 (2.6); pectoral fin length 33.5 pelvic fin length 24; snout to origin of 1st dorsal fin 33.5; snout to 2nd dorsal fin 64; snout to origin of anal fin 68; snout to pectoral fin 32; snout to pelvic fin 45; 1st dorsal height 17.6; 2nd dorsal fin height 26.8; anal fin spine height 21.6.
Measurements of 10 individuals, 34 to 58 mm. standard length, including those of the holotype: Depth 33-39.5 (2. 6-2. 9) ; head 34.2-39 (2. 5-2.9) ; eye 7.8-10.9 (4.1-4.5); snout 10.2-12.4 (3. 1-3.5) ; maxillary 12.3-14.2 (2.6-2.9) ; interorbital space 4.4-6. 1 (6.1) ; pectoral fin length 31-37; pelvic fin length 23.7-26.3; snout to origin 1st dorsal fin 31.5-37; snout to 2nd dorsal fin 62-65; snout to origin of anal fin 65-72; snout to pectoral fin 32-36; snout to pelvic fin 42-49; 1st dorsal height 16-19; 2nd dorsal height 23-29; anal fin height (spine) 20-22. Counts: dorsal fin X, 12; anal fin III, 6; pectoral fin with 14 days, the uppermost ray and the lowermost 6 or 7 simple ; scales 43-45 in a lateral series, 4 from origin of dorsal fin to lateral line, 9 or 10 from lateral line to origin of anal fin; gill-rakers 3 to 5 on upper limb, 9 to 11 on the lower limb of the first gill-arch, the lowermost two rudimentary; verte- brae 10 plus 15 plus 1.
Body compressed, the head obtusely pointed, the greatest depth at % the length of the pectoral, depth of caudal peduncle 13.3% of the length. Anterior profile at a 45 degree with the axis of the body, with a slight notch just above the orbit, the latter entering the profile. Upper profile from origin of dorsal fin to the 7th spine straight, the profile then gently curving toward the peduncle. Lower profile from snout to beneath base of pelvic fins at a considerable angle to the line of the abdomen.
Scales cycloid, absent on the head with the exception of the opercles and a small patch between the posterior nostrils; lips naked. Four oblique rows of scales on the preopercle from the eye to the rounded angle of the pre- opercle. Scales present on the opercles and on the branchiostegal membranes immediately beneath the isthmus; the latter scales are very small. Scales extending on the bases of the spinous dorsal fin and on the membranes of the soft dorsal and anal fins. Lateral line continuous, the tubes short and opening upward ; nape with a large number of tubes.
Head obtusely pointed, the eye entering dorsal profile ; interorbital space concave, the supraorbital ridges prominent, especially posteriorly; anterior nostril circular, with a short tube, nearer eye than snout and with a 5-fin- gered fleshy tentacle on its posterior border; posterior nostril with a raised border internally and anteriorly, placed close to the orbit; preopercle broadly rounded with 17 to 18 short but strong serrae, the uppermost slightly longer than the lower; opercle ending in an obtuse flap and with two flat spines, the lower much larger and more evident than the upper; branchio- stegal rays 6, the innermost one on each side very small; branchiostegal membranes broadly connected but free from the isthmus; mouth small, the lips fleshy; maxillary extending to below the center of the orbit, the max- illary almost completely hidden beneath the suborbital. Teeth of the upper jaw with an outer row (20 to 21 on each side) of small, recurved canines, two on each side of the isthmus considerably larger and stronger than the others, the posterior teeth of the jaw slightly larger than their fellows; behind this outer row is a villiform band of small teeth, widest anteriorly, becoming narrower as it progresses backward. Lower teeth similar to those of the upper jaw, except that on the middle of the side of each jaw there is a group of enlarged canines; anterior to these the teeth are similar in size to those of the upper jaw, posterior to the enlarged group the teeth are quite small. The interna] band of villiform teeth on the lower jaw extends back- ward only as far as the anterior tooth of the lateral enlarged canines. Vomer and palatines with villiform teeth, the teeth of the vomer in a broadly arched band. Gill-rakers small and short.
Dorsal fin with the 3rd, 4th and 5th spines highest, the upper anterior
60
Zoologica: New York Zoological Society
[XXV: 5
edge of each interspinal membrane with a series of 8 to 10 dermal tentacles; soft dorsal high, the anterior two rays elongate and forming a slight lobe. Anal fin high, the second spine longer and heavier than the others, its length 1.5 in the head, length of the 1st spine 1.9 in the second, length of the third spine 1.3 in the second. Caudal fin truncate. Pelvic fins in close juxta- position, their origin beneath the anterior 2/5ths of the pectoral fin, the tips of the pelvics extending to the vent. Pectoral fin with 14 rays, the upper- most one unbranched, the six lowermost (rarely the 7th) simple and un- branched, slightly swollen and with their tips free from membrane; 5th from bottom ray longest (1.15 in the head in the type), the 4th and 6th next longest.
Text-figure 2.
Cirrhitichthys corallicola Tee- Van. Drawing of the type, 58 mm. standard length.
Color: In life the general body color is orange yellow with a pinkish or lavender cast, heavily covered with dark orange or red spots placed somewhat in alternate oblique upright rows of larger and smaller spots, the spots ex- tending onto the lower portion of the dorsal fin; similarly colored spots on the head, snout and base of the pectorals are much smaller than those on the body. Dorsally and posteriorly the spots become brighter red in color. Snout reddish. Dorsal fin transparent yellow, mottled with brownish, the tips of the spines and the tentacles on the interspinous membranes scarlet; caudal fin yellow mottled with dusky; pectoral, pelvic and anal fins trans- parent yellow, the latter somewhat dusky. Iris golden red.
In the Gorgona Island specimens the spots are exceptionally well-marked and clean cut. In many of the specimens from more northerly localities, the spots tend to merge and thus produce irregular vertical bands. This condition is especially noticeable in a 30 mm. fish from Acapulco, less so in some others.
In preservative our specimens have become yellow with brownish and gray spots, the pattern of the body being retained. In two specimens from the Pearl Islands, Panama (Vanderbilt Collection) there remain only vague traces of the spots on the body.
Range: Found by us from Sihuatanejo, Mexico, southward along the coast to Gorgona Island, Colombia. (Mexico: Sihuatanejo Bay, Acapulco; Costa Rica: Port Parker, Port Culebra; Panama: Bahia Honda, Pearl
1940]
Tee-Van: Review of American Cirrhitidae
61
Islands (Vanderbilt Alva 1938 Expedition); Colombia: Gorgona Island). The majority of the specimens seen and taken by us were found in the interstices of coral. The fishes were rarely seen from above, but they were exceptionally conspicuous when viewed from a diving helmet while sub- merged.
Remarks : This species is blenny-like in the rapidity of its movements and general habits and also resembles in this respect, some of the smaller serranids. Our field names of “lavender blenny” or “lavender serranid,” given before we had been able to catch an example, are evidence of the ap- pearance of the fish in life.
Specimens Examined : 45, including the types and paratypes, from the localities mentioned above under Range.
Pseudocirrhites Mowbray, 1927.
Genotype by monotypy, Pseudocirrhites pinos Mowbray, in Breder, Bull. Bingham Oceanogr. Coll., 1 (1), 1927: 48.
This genus is close to the Pacific Cirrhitichthys, differing principally in the possession of scales on the nape and interorbital space. In Cirrhitich- thys, as represented by C. corallicola and a Japanese specimen of C. aureus, scales are absent on the regions mentioned above, but present in a small subcircular patch between and posterior to the anterior nostrils. I have not checked on other species of Cirrhitichthys to determine whether this condi- tion is true of all of the species of the genus.
Considering the geographical isolation of the Atlantic species plus the difference in scalation, Pseudocirrhites is maintained as a valid genus, as opposed to synonymizing it with Cirrhitichthys, to which it is closely related.
In the original description of Pseudocirrhites , Mowbray mentions the following as the principal character of his new genus: “The above new genus is based on the broadly united gill-membranes.” This conception of the new genus was probably based on a misstatement in Jordan and Ever- mann, Bull. U. S. Nat. Mus., 47 : 1490, which gained emphasis by being re- peated in Jordan and Evermann, The Shore Fishes of the Hawaiian Islands (Bull. U. S. Fish Comm., 23 (1), 1905: 446), and in Jordan and Herre’s A Review of the Cirrhitoid Fishes of Japan ( Proc . U. S. Nat. Mus., 33, 1907: 157). This error described the gill-membranes in Cirrhitidae as separate and free from the isthmus. Unfortunately this is not true, as the possession of broadly united gill-membranes is characteristic of all members of the family that have been examined and is so mentioned in other descriptions beyond those listed here. (See also: Regan, On the Cirrhitoid Percoids, Ann. Mag. Nat. Hist., (8) 7: 259-262).
The definition of Pseudocirrhites may be rewritten as follows: Cirr- hitoid fishes with rather large cycloid scales, approximately 42 rows in a lateral series; 4 to 5 rows of scales on the cheek; nape and interorbital space as far forward as the anterior nostrils fully scaled. Teeth present on the vomer and palatines; jaws with an outer row of small canines inside of which are villiform bands of smaller teeth, a few strong, backwardly-turned canines on center of side of each lower jaw, upper jaw with a few slightly enlarged teeth near the symphysis and toward the posterior end of each jaw. Preopercle serrate. Branchiostegals 6.
Pseudocirrhites pinos Mowbray.
(Text-figure 3).
Proportions in Percentage of the Standard Length, and Counts : Depth 33-38 (2.6-3) ; head 37 (2.7) ; eye 9.2-9.6 (3.84-4) ; snout 11.3-11.7 (3.1-3.3) ;
62
Zoologica: New York Zoological Society
[XXV :5
maxillary 13-13.4 (2.74-2.86) ; interorbital space 4.8-5.76 (4.15-6.4) ; pectoral fin length 36.5-39; pelvic fin length 21.6-25; snout to 1st dorsal fin 36; snout to 2nd dorsal fin 62-64.5; snout to anal fin 70; snout to pectoral fin 34-36; snout to pelvic fin 45-49. 1st dorsal fin height 15.7 ; 2nd dorsal fin height 18.5-21.5; anal fin height 20.8. Counts : dorsal fin X,ll; anal fin III, 6; pec- toral fin, 14 rays, the top 1, and lower 5 simple; pelvic fin I, 5; scales: 6 from origin of dorsal fin to lateral line, 42 in a lateral series, 8 from the lateral line to the origin of the anal fin; gill-rakers 6 on upper limb of first arch, 8 to 9 on the lower limb of the first arch.
Text-figure 3.
Pseudocirrhites pinos Mowbray. Drawing of the type, 54 mm. standard length.
Body oblong, compressed; depth of the caudal peduncle 2.8 in the head. Body and head covered with large cycloid scales, with the exception of the snout and preorbital, and the maxillary and premaxillary; branchiostegal membranes scaled. Lateral line continuous, nearly straight, extending onto the base of the caudal ; pores turned obliquely upward. A series of small dermal tentacles on the anterior upper edge of each interspinal membrane of the dorsal fin.
Eye placed high and slightly entering the dorsal profile; interorbital space concave; anterior nostril with a fringed tentacle; preopercle rounded, finely serrate; opercle with a single flat spine; branchiostegal membranes united (accidentally bi'oken during re-examination of the type) and free from the isthmus, covered with small scales; mouth small, placed low; maxillary narrow, about % entirely covered by the preorbital; premaxillaries little protractile; villiform teeth in bands in each jaw, on the outer border of which are a band of small canines, 3 or 4 slightly enlarged canines near the symphysis of the upper jaw, a few enlarged teeth at the posterior end of the jaw; a few strong canines on the middle of the sides of the lower jaw; vomer and palatines with small teeth. Caudal fin truncate, its rays about equal in length to the second anal spine, the first spine of the anal fin about 2 in the second, the second spine longest and strongest, the third intermediate in length and strength to the other two; pectoral fin with 14 rays, the upper- most 1, and lowermost 5 simple, the remainder branched; lowermost simple rays considerably longer than the other rays; tip of pectoral extending to the 1st soft ray of the anal fin; pelvic fins with their tips extending to the 1st anal spine.
1940]
Tee-Van: Review of American Cirrhitidae
63
A 26 mm. fish differs from the type in being less deep, and in having the two upper rays of the pectoral fin simple on the right side, and the upper three simple on the left side.
Coloration : Head and body light brown; three vertical light-colored bands on body, broadest interiorly, in the center of each band is a narrow brown line ; a broad band of dark brown covering the entire caudal peduncle ; a brown spot larger than the eye, two-thirds of which is on the body, at the base of the posterior dorsal rays. The bands of the body, both dark and light, extend onto the dorsal fin. Pectoral, ventral and caudal fins pale. Head and nape and dorsal fin with bright red spots.
In a color plate made from life of the 26 mm. fish, the narrow brown lines in the center of the light colored bands are faint; the spot at the base of the soft dorsal and the band on the caudal peduncle are black ; the pectoral and anal rays and the basal portions of the upper rays of the caudal fin are reddish.
Text-figure 4.
Map of the Eastern Pacific Zaca (1937-1938) Expedition of the Department of Tropical Research of the New York Zoological Society, showing the localities mentioned in the text. The Isle of Pines, Cuba, the type locality of Pseudo- cirrhites pinos Mowbray, is the small island immediately south of the western end of Cuba.
Range : Known from the Isle of Pines, Cuba, and from Saba Bank, 6 miles S. W. of Saba Island, West Indies (17° 35' N., 63° 21' W.)
Local Distribution: The two known specimens of this species were taken in coral, in shallow water (type) and at 25 fathoms.
64
Zoologica: New York Zoological Society
Study Materials: The type and a specimen taken by the Arcturus Oceanographic Expedition on Saba Bank, West Indies, March 15, 1925 (Dept. Trop. Res., N. Y. Zoological Society, No. 5061, color plate A. 822.)
References: Pseudocirrhites pinos. Mowbray, in Breder, Bull. Bingham Oceanogr. Coll. 1 (1), 1927; 48, fig. 23 (Description, color; figure; type locality. Isle of Pines, Cuba; Type: No. 382, Bingham Oceanogr. Coll., Peabody Museum, Yale University).
Note on Cirrhitus rivulatus (Valenciennes).
A 505 mm. fish of this species from the Galapagos Islands (U. S. Nat. Mus., 38,302) examined after this paper was in page proof, has the color bars and bands, especially those of the head, much wider than in smaller specimens and occupying relatively a greater amount of space when compared with the interspaces. The teeth of the upper jaw in this speci- men show a few enlarged canines. In the lower jaw the enlarged canines of the middle of the side of each jaw (3 on the right side and 5 on the left) are very large and conspicuous and are wider apart than in smaller fish. Anterior to these there are two or three small canines and posterior to the enlarged group on each side are a group of smaller canines. The villiform teeth in the lower jaw occupy a small patch in the front of each jaw anterior to the large lateral teeth.
EXPLANATION OF THE PLATE.
Plate I.
Cirrhitus rivulatus (Valenciennes). Seven individuals from Corinto, Nicaragua, 65 mm. to 240 mm. standard length, showing the alteration in pat- tern correlated with growth and size.
TEE-VAN.
PLATE I.
A REVIEW OF THE AMERICAN FISHES OF THE FAMILY CIRRHITIDAE.
Crane: Post-embryonic Development of Ocypode
65
6.
Eastern Pacific Expeditions of the New York Zoological Society. XVIII.
On the Post-embryonic Development of Brachyuran Crabs of the Genus Ocypode.1
Jocelyn Crane
Technical Associate, Department of Tropical Research, New York Zoological Society.
(Text-figures 1-8).
[This is the nineteenth of a series of papers dealing with the collections of the Eastern Pacific Expeditions of the New York Zoological Society made under the direction of William Beebe. The present paper is concerned principally with specimens taken on the Eastern Pacific Zaca (1937-1938) and Arcturus Oceanographic (1925) Expeditions. For data on localities, dates, dredges, etc., of these expeditions, refer to Zoologica, Vol. VIII, No. 1, pp. 1-45 {Arcturus), and Zoologica, Vol. XXIII, No. 14, pp. 287-298 (Eastern Pacific Zaca) f]
Contents.
Page
I. Introduction 65
II. Summary of Important Points 66
III. Previous Knowledge of Ocypodid Devel- opment 66
A. Zoea 66
B. Megalops 66
IV. The First Zoea of Ocypode gaudichaudii 67
A. Material and Methods 67
B. Diagnosis 67
C. Description 67
D. Comparison 69
Page
V. The Megalops of Ocypode 69
A. Material and Methods 69
B. Taxonomy and Identification 70
C. Comparison of Uca and Ocypode. . 72
D. Generic Characters 72
Diagnosis 72
Description 72
E. Specific Characters 80
F. Ecology 81
VI. Bibliography 81
I. Introduction.
The present paper is the first of a series dealing with the brachyuran crabs of the Eastern Pacific Zaca Expedition. It concerns the first zoea of Ocypode gaudichaudii and the megalopa of the two Pacific and single Atlantic species of the genus.
Previous reports on the Brachyura of the various Eastern Pacific Ex- peditions of the New York Zoological Society are the following: Rathbun, 1924, “Brachyuran Crabs Collected by the Williams Galapagos Expedition, 1923” ( Zoologica , Vol. V, No. 14) ; Boone, 1927, “Galapagos Brachyura” {Zoologica, Vol. VIII, No. 4) ; Glassell, 1936, “Templeton Crocker Expedi- tion. I. Six New Brachyuran Crabs from the Gulf of California” {Zoo-
1 Contribution No. 591, Department of Tropical Research, New York Zoological Society.
66
Zoologica: Netv York Zoological Society
[XXV :6
logica, Vol. XXI, No. 17) ; Crane, 1937, “The Templeton Crocker Expedition. III. Brachygnathous Crabs from the Gulf of California and the West Coast of Lower California” ( Zoologica , Vol. XXII, No. 3) ; Crane, 1937, “The Templeton Crocker Expedition. VI. Oxystomatous and Dromiaceous Crabs from the Gulf of California and the West Coast of Lower California” ( Zoologica , Vol. XXII, No. 7).
The drawings in the present paper are the work of Miss Harriet Bennett.
To Mr. Templeton Crocker I wish to express my appreciation for the opportunity of collecting material on a cruise of his yacht Zaca; to Dr. William Beebe, Director of the Department of Tropical Research, for his supervision and advice in the preparation of this paper; and to Dr. Waldo L. Schmitt of the United States National Museum for the loan of speci- mens, and for his generous permission to dissect an Atlantic megalops and include the results of the study in this report.
II. Summary of Important Points.
1. The zoea of a species of Ocypode (O. gaudichaudii from the eastern Pacific) is described for the first time. It differs from that of the most closely related genus, Uca, principally in having lateral spines.
2. The megalopa of both eastern Pacific species of the genus, O. gaudichaudii and O. occidentalis, are described for the first time. They differ only in small details (such as the number of setae on the last pleopods) from the western Atlantic megalops, O. albicans, but are totally distinct from the known megalops of Uca.
3. The megalops described by Rathbun (1924, p. 155) as being perhaps referrable to O. gaudichaudii is shown to belong instead to the genus Plagusia, probably to P. depressa tuberculata.
4.