BULLETINS OF THE Zoological Society of San Diego No. 15 A Further Study of Pattern Dimorphism in the California King Snake By L. M. KLAUBER Consulting Curator oF Reptiles^ Zoological Society of San Diego SAN DIEGO, CALIFORNIA SEPTEMBER 29, 1939 Digitized by the Internet Archive in 2017 with funding from IMLS LG-70-15-0138-15 https://archive.org/details/bulletinsofzoolo1519unse '• t . ■Tf. • t ' ■ i } I ■■4 i i / ¥. R ^ > • f 4- ! J Sft 1 >• //* 1 *1 I i X I • i t I BOARD OF DIRECTORS AND COMMITTEES OF ZOOLOGICAL SOCIETY OF SAN DIEGO ROAKl) OF D[RECTORS Harry M. Wegeforth, M.D., President L. 'r. Olmstead, Viec-President Fred Kunzel, Seeretary Frank C. Spalding, Treasurer J. Waldo Malm berg T. M. Russell Mrs. Robert P. Scripps Robert J. Sullivan F. L. Ann ABLE C. L. COTANT W. C. Crandall Gordon CjRay EXECUTIVE STAFF Mrs. Belle J. Benciiley, Exeeutivc Seeretary Louis Gill, C. R. Sciiroeder, D.V.M., AreJiiteet Pathologist L. M. Klauber, a. T. Mercier, Consulting Curator of Reptiles Consulting Engineer HOSPITAL COMMITTEE Dr. R. j. Pickard, Chairman Dr. Howard A. Ball Dr. Eaton M. MacKay Dr. Hall G. Holder Dr. Francis M. Smith Earl P. Warren EDUCATIONAL COMIHITTEE Mrs. Lena P. Crouse, Chairman Robert D. Harwood Earl U. Homuth AQUARIUM COMMITTEE Gerald MacMullen, Chairman Captain R. H. Baker Bates Harper Lieut. Comdr. D. P. Marvin Dr. Charles H. Townsend THE PURPOSE OF THE SOCIETY 1. To advance science and the scien- tific study of nature. 2. To foster and stimnlate interests in the conservation of wild life. 3. To maintain a permanent Zoolog- ical Exhibit in San Diego. 4. To stimulate public interest in the building and the maintenance of a Zoological Hospital. 5. To provide for the delivery of lec- tures, the exhibition of pictures, and the publication of literature dealing with natural history and science. 6. To operate a society for the mu- tual benefit of its members for non-lucrative purposes. BULLETINS OF THE ZOOLOGICAL SOCIETY OF SAN DIEGO No. 15 A FURTHER STUDY OF PATTERN DIMORPHISM IN THE CALIFORNIA KING SNAKE BY L. M, Klauber Consulting Curator of Reptiles Zoological Society of San Diego SAN DIEGO, CALIEORNIA September 29, 1939 4 Bulletin 15 : Zoological Society of San Diego ST A X 1 ) A R 1 ) B AI^TB: R N S B'ig. 1. Ringed Bhase Fig. 2. Striped Bhase Fig. 3. jNIixed Ringed and Striped ABERRANT FATTERNS Fig. 4. Black Ventrum Fig. 5. Broken Stripe Ki .AUBER : Pattern Dimorphism in King Snakes 5 A FURTilER STUDY OF PATTERN DIMORPHISM IN THE CALIFORNIA KING SNAKE Introduction In a paper published in 1936 (Herpetolopica, 1, 18-27), it was shown that the ringed and the longitudinally-striped king snakes of southern California, hitherto considered separate species, LampropcJtis gctulus boylii and L. calif orniac, are in reality only pattern phases of a single species, Lam pro pci fis gctulus calif orniac. This conclusion followed the hatching of two clutches of eggs, one laid by a striped, and one by a ringed mother. Each brood contained representatives of both pattern phases. More recently eight additional broods have been hatched. While these do not change the conclusions previously reached, they do provide some additional statistics on the inheritance and sex-linkage of these patterns, which seem to justify a supple- mental report. However, before presenting the data on the newly available broods, it is necessary to develop certain statistics of the general king-snake population, especially sex- and pattern-ratios, in order to have a standard with which the broods may be compared. This king snake does not possess the power to produce the dual pattern throughout its range, for the pure striped phase is found only in southern California and northern Lower California, in the region of which San Diego is the center, while the ringed phase has a much wider range. In the San Joaquin Valley of south-central California, and in the Cape region of Lower California, there is produced an occasional aberrant pattern, showing some similarity to the perfectly striped snakes of the San Diegan region ; but these individuals are highly variable in nature and evidently represent a comparatively small proportion of the population. Elsewhere throughout the ex- tensive range of the Pacific King Snake in southern Oregon, central and northern California, Nevada, Utah, and Arizona, only the ringed pattern has been found. Eastward to the Atlantic coast, other sub- species of Lampropeltis gctulus occur, but all of these are ringed or spotted ; none has developed an alternate striped phase. Pattern Dimorphism in the General Population There are indications that, even in San Diego County, the power to produce the striped pattern is not present to the same degree in the snakes on the desert side of the mountains. We have the following statistics of the frequenc}^ of occurrence of the two forms in the county, as derived from a 16-year record of the snakes reported from this area ; 6 Bui-letin 15: Zoological Society oe San Diego Area Ringed Striped Per Cent Striped Coast 511 357 41.1 Inland Valleys and Mesas 330 211 39.0 Coastal Foothills 160 137 46.1 Mountains 19 14 42.4 Subtotal 1,020 719 41.4 Desert Foothills 40 5 11.1 Desert 9 0 0.0 'fotal 1,069 724 40.4 llius we find a fairly constant proportion of striped snakes eastward from the coast, in the zones to and including the mountains, with a conspicuous falling off on the desert slope. By the chi-square method of testing the significance of the difference between the proportions exemplified by two groups, we find that the ratios in the desert foot- hills and the desert are significantly different from those which obtain on the coastal side of the mountains. Thus, we conclude that the transmontane snakes do not possess the genetic factor which produces the pattern alternation to the same extent as their coastal congeners. The zonal ratios in the four zones, from the coast to the mountains in- clusive, are found by test not significantly different from the general mean or from each other, f fence we consider 41.4 per cent striped (and mixed) snakes, the weighted mean in these four zones, to rep- resent our best estimate of the composition of the coastal population, and this figure will be used in commenting on the ratios in the broods. For the purposes of this tabulation all specimens having a pattern mixture were classified as striped, since the statistics of these aberrant snakes were not kept with sufficient accuracy to permit separate groupings. The characteristics of these aberrant patterns will be subsequently discussed. To give an idea of the consistency of the percentage of the striped phase out of the total cismontane population, we present the figures for the 16 successive years making up the total given in the previous table : 7 KlAUHER: l^ATTERN DIMORPHISM IN KiNG SnAKES Striped Ringed Total Per Cent Year Snakes Snakes Snakes Striped 1923 32 52 84 38.1 1924 29 52 81 35.8 1925 44 59 103 42.7 1926 37 46 83 44.6 1927 55 55 no 50.0 1928 44 50 94 46.8 1929 41 70 111 36.9 1930 80 86 166 48.2 1931 35 57 92 38.0 1932 16 42 58 27.6 1933 33 63 96 34.4 1934 45 68 113 39.8 1935 55 94 149 36.9 1936 30 35 65 46.2 1937 65 78 143 45.4 1938 78 113 191 40.8 Total 719 1,020 1,739 41.4 It will be observed that the annual percentage of striped snakes varies from a inaximinn of 50.0 to a minimum of 27.6. No secular change in the ratio is evident ; as a test, the totals for the first and last 8-year periods were compared and no significant difference was found in the proportions. No attempt should be made to attribute to a greater population the evident increase in the total number of snakes re- ported ; the increase has resulted from certain conditions affecting the collecting and reporting of specimens rather than from any real change in the king snake population. Standard Patterns Although, from the statistics thus far presented, it might be pre- sumed that these king snakes may be readily and consistently divided into two pattern groups, such is not the case ; for there is a consid- erable variability in each of the two patterns, and this is further accentuated in a small proportion which seem to be mixtures or pattern intermediates. In order to clarify the discussion of these inter- mediates. it is necessary, first, to repeat the descriptions of the typical forms which appeared in the prior paper. These descriptions are jiar- ticularlv applicable to the San Diego County specimens, it being understood that the striped phase reaches its highest development in 8 liULLETIN 15: ZoOLOt’.lCAL SOCIETY OF SaN DiEGO this area, both in perfection of pattern and in nninbers relative to the total population. The ringed and striped phases do not differ essentially in bead pattern: they are likewise similar in body color. The light marks in either may vary from white to huff, and the dark from red-brown through chocolate-brown almost to black. The color contrast is usually more accentuated in the young than in adults. The ringed pattern consists of light cross-rings separated by dark, there being from 22 to 40 of these rings on the body and 6 to 10 on the tail (Fig. 1).* The relative widths of the two sets of rings vary considerably ; hut usually the dark rings, where they cross the mid- dorsal line, are 3 or 4 times as wide as the light. Sometimes the light rings engage only a single scale, hut more often they cover about 2 scales, end to end, where they cross the middle of the Ijack. Laterally, as the light rings ai)proach the ventrum, they widen at the expense of the dark, so that as they cross the belly the widths of the light and dark rings are approximately equal, or the light may be wider. The striped phase has a light longitudinal stripe on the dark ground color from head to tail (Fig. 2). This stripe usually engages the mid- dorsal row and half of the row on either side, thus being two rows wide. The lateral edges of the stripe are generally even, but the stripe may he imperfect in a variety of ways ; it may be broken into a series of dashes or spots ; it may vary in width or he laterally offset in sec- tions by a scale-row or two. Sometimes the edges are serrated. Breaks in the stripe on the tail are usual, the perfect stripe from head to tail- tip being rare indeed. The lateral pattern is quite variable, but if we accept the specimens with the more perfect dorsal stripes as being those most truly rep- resentative of the striped pattern, we find the corresponding lateral pattern to consist of light lateral scale rows next the ventrals, with increasing darkness as the dorsum is approached. This gradual change is effected by a dark border around each scale, which l)order expands at the expense of the light center, continuously reducing the light spot. There is considerable variability in the row at which the light centers finally disappear from the scales, hut this usually occurs from the 4th to the 8th row al)Ove the ventrals, thus leaving from 6^4 to 24^ solidly dark rows on either side of the dorsal strii)e. Just as the mid-dorsal stripe is highly variable in quality, so also is the coloration of the lateral areas, which, deviating from what While the figures show the light scales edged with dark, and vice versa, this is done to outline the scales. Actually the scales are unicolor. Klauber : Pattern Dimorphism in King Snakes 9 seems to lie the theoretical pattern, may he mottled, harred, or with longitudinal dashes in a variety of arrangements. The helly of the striped phase is generally light, except that the outer edges of the ventrals are usually marked with dark blotches or irregular streaks. But it is highly characteristic of this pattern phase that the underside of the tail is always dark, there being no exception to this rule in more than 100 specimens investigated. The transition from light to dark occurs quite sharply at the vent or slightly posterior thereto. Aberrant Patterns With a variability of this character and extent, particularly in the striped phase, it might he thought impossilile to make any classifica- tion of those aberrant specimens in which the abnormal deviations of both are mixed ; for the intermediates, even in those sections of the body which tend to follow one or the other pattern, often do not incorporate pure or typical patterns, but rather somewhat distorted analogues. For example, in the ringed sections of the intermediates the light rings are often narrower than is customary in the pure ringed phase, and do not widen laterally to the usual extent. And in the striped sections, the light dorsal line is likely to wander considerably from the expected mid-dorsal location. Yet with all this variability we find it possible to divide the aberrants into three general groups, based on the predominance of three kinds of deviations from normal : ( 1 ) Snakes having both ringed and striped sections on the body (Fig. 3). Specimens with short cross-bars, which, however, do not cross the lateral scale rows just above the ventrals, are included in the third, rather than this group. (2) Snakes with dark ventrums (Fig. 4). In its most extreme form this type of aberrant has the belly all dark. A segregation has been made on the proportion of dark venter, specimens with more than half dark being allocated to this class. (3) Snakes with regularly interrupted mid-dorsal light lines (Fig. 5). In spite of the frequent irregularity of the dorsal stripe it is possible to segregate a particular group with this stripe rather consistently divided into a series of round light spots, or short longitudinal or cross dashes. Such an interruption on the tail alone is not considered significant, since nearly all specimens have this irregularity on the tail, even when the body stripe is unbroken. I have arbitrarily placed in this category such specimens as have at least 6 complete breaks in the mid-dorsal stripe on the Iiody. 10 Bulletin 15: Zoolouical Society of San Diego Before proceedin^^ to the statistical data on these alierrants, 1 wish to mention some of their other peculiarities. Mixed aberrants of the first type, which, by their very mixtures seem best to fit our ideas of what a true intermediate should he, have several peculiarities. Usually the striped sections are present at the head and tail, while the mid-body is ringed. Where the ringed part greatly predominates, the anterior striped section is likely to lie sup- pressed, hut the tail stripe persists, d'he light rings are often narrower than is usual in the pure ringed jihase, and fre([uently they do not widen as they approach the ventrum, as in their pure analogues. In this these cross rings resemble the longitudinal stripes of the other pat- tern, turned sitlewise. d'hus these intermediates are generally darker than normal specimens. /Vherrants of the second group with dark ventrums are often heavily [ligmented laterally as well. I have noted one specimen in which the lateral light scales are reduced to a single row above the ventrals. Sometimes the dark ventral color is concentrated on the outer edges of the ventrals, with a light streak down the mid-ventral line; hut in other cases the opposite is true, with the dark streak or hand in the middle. When a longitudinal color difiference is evident, as is frequently the case, the dark pigment is concentrated posteriorly. Aberrants of the third type are characterized by a considerable irregularity and are the least readily segregated of the three. The lilotches, which represent a partly suppressed dorsal stripe, may be reduced to round rots, or thev mav he dashes, often twisted at an angle, or turned exactly crosswise. A lilack dorsum may go with this pattern, that is, a specimen may fall into both the second and third classifications. There is a fourth, somewhat indefinite, group of aberrants which I have not sought to segregate from the striped specimens, since it is too difficult to assign definite limits for their classification. Their irregularities are largely evident laterally, although the mid-dorsal line is likely to he imperfect, with ofifsets and variations in width. On the sides there are light and dark hands of varying width, broken and separated by light and dark cross blotches, which do not, how- ever, cross the dorsum. These specimens are few in number, some 5 or 6 out of three or four hundred. Thev resemble in some wavs the »• ^ aberrants which are found in the San Joaquin Valley; also, they are of freijiient occurrence in the vicinity of Laguna Beach, Orange County. To some extent they also resemble the form from the Cajie Region of Lower California which Van Denburgh described as L. nitida, hut which I now consider a iiattern-aherrant of the ringed Klauber: Pattern Dimorphism in King Snakes 11 king snake of that region. In nitida there is much irregularity in both dorsal and side stripes and the ventrum is always dark. There is less contrast between light and dark areas than is the case in San Diego County ; this characteristic is also observed in the aherrants of the San Joaquin Valley. Thus these snakes are dark mixtures of the other types of aherrants. Proportion of Arerrants In preserving specimens for study, 1 made it a practice for a num- lier of years to save all of those with iieculiar patterns, while many of the more normal specimens were exchanged or exhibited at the zoo. This was a justifiable procedure from the standpoint of securing a maximum of aberrant material for study, hut has had a distorting efifect on the statistics gleaned from the composition of the collection. However, this artificial selection of aherrants was not effective in the case of the juveniles and hence the available young specimens may be taken as a sanqile of the normal distribution of patterns in this county. Using such a segregation we have the following statistics on the specimens under 500 mm. in length, and presumably representing an unbiased sample of the San Diego County population: Classification Number Per Cent of Total Ringed . . . . 76 54.3 Striped . . . . 49 35.0 Aberrant Mixed Striped and Ringed. . . 3 2.1 Black Ventrums 7 5.0 Broken Stripes .... 5 3.6 140 100.0 It will he observed that the total percentage of aberrant specimens (10.7) considerably exceeds that given in the previous paper, wherein it was stated that the aherrants comprised about 3 per cent of the poimlation. However, in the former paper only the mixed specimens having both patterns were recognized as aberrant. It will he noted that in this sample the striped plus aberrant snakes comprise 45.7 per cent of the population as compared with the figure of 41.4 jireviously given for the population as reported from all sources. By test the difference is found to be without significance; 12 Bulletin 15: Zoological Society oe San Diego that is, there is no evidence that this is not a fair sample of the same population as that comprising the census. Sexual Dimorphism One other point remains to lie discussed before applying these statistics to the hroods, this being sexual dimorphism, and particularly whether there is any linkage lietween pattern and sex. First as to the sex ratio in the population ; all specimens available to me, not including the hroods, have the following sex distribution: Per Cent Males Females Total Males Ringed Phase 88 90 178 49.4 Striped Phase* 76 61 137 55.5 Total 164 151 315 52.0 Testing these distributions liy the usual chi-square method with the correction for continuity, we find that neither deviation is significantly different from a theoretical equality of the sexes ; nor is there a sig- nificant difference between the sex ratios of the two patterns. We are therefore justified in concluding that the sexes occur in equal num- bers among the king snakes, and that there is no sex-linkage with pattern, insofar as the two typical patterns are concerned. But with the aberrant specimens we find a different situation, for a sex-linkage with pattern is evident. In this case we employ all avail- aide San Diego County specimens except the hroods ; since there was no conscious selection of these aberrant specimens with respect to sex, we are not restricted to specimens less than 500 mm. long. We have the following data : Per Cent Classification Males Females Total Males Mixed Pattern 16 1 17 94.1 Black Ventrums 2 12 14 14.3 Broken Stripes 5 11 16 31.3 Total 23 24 47 49.0 Testing these distributions against an assumed equal division of the sexes we find that the first two jiroportions are significantly different from a 50-50 sex ratio ; that is, the observed discrepancies are not at- tributable to the chances of sampling. Broken stripes are, however. Including aberrants. Klauber : Pattern Dimorphism in King Snakes 13 not definitely proven to lie a sex-linked character ; in drawing sixteen individuals from a large population having a balanced sex distribution, there would be 21 chances in 100 of drawing 5 or less of one sex and 11 or more of the other. Noting so few of one sex in two of the classes, one might suspect these particular individuals have been misplaced through inaccurate sexing. However, such is not the case; while the sexing of juveniles is difficult, two of the individuals which break down what seemed at first to be a perfect correlation are adults, where no mistake could be made. For example, the single female in the first group, which had an unusually fine intermediate pattern, contained eggs. It will lie observed that if we take all of the aberrant specimens together, there seems to he no sexual unbalance. However, this is too small a sample to permit a dogmatic statement on this point. Brood Data Having derived from the available specimens such information as we may with respect to pattern- and sex-ratios, we now proceed to apply these statistics to the ten broods lately hatched, to see what may be learned concerning the inheritance of the patterns. Unfortunately we know nothing concerning the fathers of these broods, since all the pairs had mated in the wild before the mothers were caught. The pat- tern data on the broods are as follows : Pattern of Brood Aberrant Normal Brood Number Mother Ringed Striped Mixed Dark Ventrums broken Stripes Total 1935-2* Ringed 2 1 , . % . . 1 4 1937-2 Ringed 4 . . . . 4 1938-5 Ringed 7 . « • • 7 1938-6 Ringed 9 , , . , . 9 1935-1* Striped , . 1 1 3 1 6 1938-1 Striped 2 5 • . 1 8 1938-3 Striped 1 1 . . It 3 1938-4 Striped . . 4 . . 2 6 1938-2 Dark Ventrum. , , 1 3 4 1937-1 Broken Stripes. 1 1 4 1 7 Total 26 14 5 11 2 58 Reported in the previous paper. f Also has broken stripes. 14 Bulletin 15: Zoological Society oe San Diego The conclusions to lie drawn from this talile are: (1) A mother of either phase can [iroduce pure patterned young of both phases, but (2) A mother tends to produce young of her own phase. (3) d'he proiHirtion of intermediates and aberrant specimens, amongst these broods hatched in captivity, is greater than in the gen- eral king snake population of the county. Conclusion ( 1 ) is self-evident from the statistics. No less than four of the broods contain individuals with jiure patterns of both phases. One of the mothers was ringed, one an aberrant, and two were striped. However, conclusion ( 1 ) does not l)y any means justify the corollary that the pattern of the progeny is independent of that of the mother. On the contrary, taking only the pure-patterned mothers and progeny, we have the following contingency table : Brood Mother Striped Ringed Total Striped 11 3 14 Ringed 1 22 23 Total 12 25 37 It requires no calculation to demonstrate that this distribution in- dicates a high correlation between the pattern of the mother and that of the progeny; there is a highly significant difiference between the progeny of the two classes of mothers. Another way of looking at the same situation is the following. It will be observed that three out of the ten broods have pure progeny of the same pattern as the mother. The possibility of this occurring by chance is exceedingly small. For example, take Brood 1938-5 (seven ringed young from a ringed mother) and assume a population ratio of 41.4 per cent striped and 58.6 per cent ringed. If pattern were purely a matter of chance, we apply a binominal distrilnition and find that only one brood in 42 having seven young would come out all ringed. Similarly, with nine young, only one brood in 123 would be all ringed. Yet we have both of these contingencies occurring in ten liroods, besides another uniform brood of four snakes. Conclusion (2) is therefore inevitable: The patterns of the young are related to that of the mother — a mother tends to produce young of her own phase. It is found from the above table that, neglecting aherrants (both mothers and ofifspring) striped mothers produce 78.6 per cent striped young, and 21.4 per cent ringed; while the ringed mothers produce 4.3 per cent striped young and 95.7 per cent ringed. These ratios are Klauber; Pattern Dimorphism in King Snakes 15 not consistent with a general population proportion of 41.4 per cent striped, for, starting with these ratios, the proportion in the next gen- eration would be only 35.1 per cent striped (78.6X41.4+4.3X58.6). We may therefore presume that if the two phases are equally pro- lific as regards sizes of broods, a greater nnmher of broods than has been availalile to us would show a slightly higher production of strijied young by either or both phases of mothers. For example, if striped mothers produced 80 per cent, and ringed mothers 14 per cent striped young, a population lialance similar to that noted would be obtained. These ratios could he properly worked out if we had a larger series of broods and knew the patterns of the fathers. The third conclusion, that the broods have a higher percentage of aberrant individuals, is evident from the following table: Per Cent Normal Aberrant Total Aberrant Available specimens under 500 mm 125 15 140 10.7 Broods 40 18 58 31.0 The higher jiroportion of aherrants in the broods is found to be statistically significant. However, this may be explained by the pres- ence of the two aberrant mothers. It is olivious that they have pro- duced more than their shares of the aherrants, since their broods con- tained only 3 normal snakes to 8 aherrants, or 73 per cent abnormal, compared to a population average of 10.7 per cent. Two aberrant mothers out of a total of 10 is a larger than average percentage of aberrant mothers than would be evident if more broods were avail- able ; one in nine would be similar to the proportion in the general population. Taking half of the young snakes in the two broods of the aberrant mothers, plus all broods from the normal mothers, we have the following contingency table : Typical Aberrant Total Broods 38+ 14 52+ General Population 125 15 140 Total 163+ 29 192+ Using the chi-square method with correction for continuity, we find that there is only a little more than one chance in a hundred that this difTerence in proportionalities could occur by chance. We are therefore justified in the conclusion that these broods, hatched in 16 Bulletin 15: Zoological Society oe San Diego captivity have a higher proportion of aherrants than the general population. It is evident that there is a greater tendency for the striped mothers to produce aberrant young than the ringed mothers, as indicated by the following table : Per Cent Normal Aberrant Total Aberrant Striped Mothers 14 9 23 39.1 Ringed Mothers 23 1 24 4.2 Total 37 10 47 21.3 This difference is found to he significant. Also, the difference be- tween the proportion of aberrant young produced liy the striped mothers proves to be significantly higher than the proportion found in the general population, as represented by the young under 500 mm. But the small number produced liy the ringed mothers is not sig- nificantly lower than that evident in the general population. The fact that aberrant mothers are fertile is also of interest. Returning to the first conclusion respecting the high proportion of intermediates and aberrant specimens amongst broods hatched in captivity, we observe that if this ratio is greater than in the general population, such a peculiar situation might result from one of the three following conditions : (a) Pattern aberrance in juveniles may be correlated with other characters of a lethal nature, so that many of these queer specimens do not long survive. (b) The pattern, particularly the dark dorsum, may change with age. (c) The captivity of the mothers may affect the progeny. I think that h is ruled out by the fact that there is no indication of color changes in king snakes with age, except for a gradual lightening of the dark brown or black areas in the juveniles, to chocolate or medium brown in the adults. But at all times the pigmented areas may be easily distinguished from the unpigmented, and there is no evi- dence that either changes ontogenetically at the expense of the other. Full grown specimens with dark ventrums are occasionally found and it must be remembered that the underside of the tail is dark in all striped specimens. This would seem to leave the choice between a and c, with a, the suppression of aherrants by the presence of lethal or harmful characters, as much the more likely. I have previously mentioned a similar high percentage of aberrant and defective young Klauber ; Pattern Dimorphism in King Snakes 17 amongst rattlers born in captivity. It would be rather fantastic to suppose that captivity could affect the patterns of the young snakes. As nothing is known concerning the fathers of these broods, we can make no useful deductions with regard to a law of pattern in- heritance. The percentage of striped snakes is found to be significantly above the ratio for a simple mendelian recessive. Sex-Pattern Linkage in Broods The sex distribution amongst the broods was as follows : Aberrant Normal /\Derrant Dark Broken Ringed Striped Mixed Ventrums Stripes Total Males 15 8 5 2 2 32 Females 11 6 0 9 0 26 Total 26 14 5 11 2 58 Percent klales. . . . 57.7 57.1 100 18.2 100 55.1 The decided sexual unbalance amongst the aberrant specimens con tinues in evidence. Combining these broods with our previous tabula tion of the aberrants from the g eneral population. we have : Per cent Classification Males Females Total Males Mixed pattern . . . 21 1 22 95.5 Black ventrums . . 4 21 25 16.0 Broken stripes . . . 7 11 18 38.9 Total 32 33 65 49.2 The mixed pattern and lilack ventrums are significantly sex-linked ; this is not true of the broken striiie jiattern. Evidently, alierrance in males is usually accompanied l)y a mixed or intermediate dorsum, while in females it is evidenced by a dark ventrum ; no doubt these two patterns are sexually complementary. Independence of Pattern and Lepidosis In the previous paper (p. 21) it was shown that there are no scale- count differences which are correlated with the pattern dimorphism, that is to say the ringed and striped phases have scale counts differing insignificantly in means and dispersions. With the added specimens now available this conclusion remains unchanged ; we find no signifi- 18 [■Bulletin 15: Zoological Society of San Diego cant differences in scale rows, ventrals, caudals, snpralaliials, or in- fralabials. These conclusions are based on 167 of the ringed and 103 of the pure striped phase, all from San Diego County, with the broods excluded. We do find, rather unexpectedly, that there is sexual diinor- l)hisin in dorsal scale rows, the female snakes averaging slightly higher than the males. In onr snhseiinent calculations we have treated the sexes separatelv in dealing with scale rows, just as we do with ventrals and candals, in which sexual dimorphism is so obvious as to require no calculation for its demonstration. Having found no differences in lepidosis between the ringed and striped specimens, we pool their scale counts to secure a larger sample of the parent population. With this as a basis of comparison we make two investigations to determine : ( 1 ) whether specimens aberrant in pattern are in any way peculiar in scale counts as well ; and (2) whether the broods differ significantly in scale counts from the general popu- lation. The answer to the first question is in the negative : Specimens ab- errant in pattern are not significantly different in any character of lepidosis from the rest of the population. Thus this queer manifesta- tion of pattern dimorphism carries with it no differences in scalation, either with respect to the two pure pattern phases, or the aherrants which are mixtures of the two. In the case of the broods there are certain differences which do ap- proach the level of significance, hut it would probably he well to re- serve a final opinion on this point until a greater number of broods becomes available. There are indications both of the inheritance of atypical scale counts and arrangements ; and also of the presence amongst the broods of variants from the mode so unusual as to lead to one of the hypotheses previously mentioned — either that these abnormalities are the result of captivity on the mother, or, much more probably, they are linked with lethal characters, so that the young which carry them do not survive. As examples of the inheritance of peculiarities the following may he cited: One mother has the fourth infralahials on both sides of the head divided in a peculiar way ; two out of her brood of eight show the same peculiarity, hut on the left side of the head only. Another mother with ten infralahials has a higher than normal number of young with this number of infralahials. Yet this type of inheritance is not always evident, since a mother with a divided mental did not produce any young with this characteristic. As to the other type of peculiarity, we find amongst the broods a .surprisingly high proportion without loreals, the loreal being fused Klauber: Pattern Dimorphism in King Snakes 19 either to a prefrontal, postnasal, or preocular. Also a number have an upper postocular fused to the supraocular. One specimen (a fe- male) has only 41 subcaudals, the lowest of 155 sjrecimens. This is 4 less than the next lowest, and is 3.7 standard deviations below the mean (52). This is similar to some of the radical deviations observed amongst rattlesnake broods. Period of Incubation The eggs were laid between June 23rd and August 2nd. and hatched between Seiitemlier 5th and October 10th. The period of incubation varied between 71 and 86 days, the average being 76. Batches of eggs were kept on slightly moistened paper in small individual crocks with a glass plate as a cover. This method, devised by C. B. Perkins, is proving highly successful in hatching the eggs of many species. Limitations of Pattern Dimorphism Viewed broadly the most interesting feature of this case of pat- tern dimorjihism is its territorial limitation, compared to the range of the subspecies as a whole ; and the peculiar modifications which occur on the periphery of the affected area. These modifications en- tail a reduction in the frequency of occurrence of the atypical form and coincidentally a change in the manifestation of that form. First we have the situation in western San Diego County. Here, in a small area hardly greater than 50 by 50 miles, the longitudinally striped phase, so different from the ringed snakes which characterize the genus, number about 32 per cent of the population. An additional 10 per cent are intermediates between the ringed and the striped forms, or are modifications more closely resembling the latter. Thus 42 per cent of the population are atypical. As we go out from this center we find a decreasing proportion of atypical specimens. Unfortunately the statistics from outside the county are meagre. From Chino, San Bernardino County, California, out of some fifteen specimens only one was aberrant. In Los Angeles County, the striped form is present, but reduced in relative numbers. Eastward from San Diego, on the desert slope, I have already shown (p. 6) that the production of the atypical form is significantly lower in proportion than on the coast. Likewise to the south the striped form is rarer, although it is known to occur at least as far south as Ensenada. At San Jose (Lat. 31°) in Lower California seven king snakes were collected and all were ringed ; persons living in that vicinity when shown striped specimens said they had never seen such a snake thereabouts. 20 Bulletin 15; Zoological Society of San Diego Coincidentally while the striped phase decreases in numbers with the distance from San Diego, the deviations of the atypical siiecimens from the pure striped pattern increase. Thus such atypical snakes as are found more nearly resemble those which have been termed aber- rants in San Diego County. Although I have seen pure striped speci- mens from ten miles north of Ensenada, Lower California, the aberrants comprise a majority of those which are not ringed; in fact, the prevalence of mixed, rather than pure striped specimens, is al- ready noticeable at Tijuana on the U. S. -Lower California border. To the north pure striped snakes have been seen at Temecula, and are said to occur in the vicinity of Hemet, Riverside County. Along the coast I have no definite records of the pure striped pattern from any point north of San Diego County. Although atypical specimens are not uncommon at San Clemente and Laguna Beach, in Orange County, and at various points in Los Angeles County, they are all of patterns, which, in San Diego County, would be termed aberrants — none has yet been collected with a clear and even mid-dorsal stripe. Lurther to the north, in the San Joaquin Valley, the atypical speci- mens are still rarer, and the patterns more broken and imperfect. The same is true at the southern limit of pattern dimorphism, in the Cape region of Lower California, where the form described by Van Denburgh as L. nitida may, by analogy, be assumed to be an aberrant phase of the ringed snakes of that vicinity. Some of the San Joaquin Valley snakes closely resemble these Cape specimens. In talking over this situation. Dr. Jean Linsdale has suggested that the power to produce striped and aberrant pattern phases might well be recognized as a subspecific character. Lrom a nomenclatorial standpoint, there is much to recommend this, since there is now no way to distinguish between the snakes of Utah (for example) where all are ringed and those of southern California. Under such a scheme, if we include in calif orniae all snakes which have the power to pro- duce an atypical pattern, even though it may not be the pure striped phase, we would place in the subspecies Lampropeltis getulus cali- f orniae, the king snakes inhabiting the southern San Joaquin Valley, southern California, and Lower California. Elsewhere in California, and in Oregon, Nevada, Utah, and western Arizona, the snakes of this form would be referred to as Lainpropcltis getulus boylii, a more desirable and appropriate name, since this was the name by which they were long known, prior to the discovery that calif orniae was only a pattern phase of the ringed snake, rather than a separate species. This allocation of names would involve no violation of the Code, provided the power of pattern dimorphism be accepted as a Klauber: Pattern Dimorphism in King Snakes 21 valid subspecific character. The only objection is the necessity for classifying newly acquired ringed specimens on purely geographical criteria, depending on whether striped or aberrant snakes are known to occur in the same territory. In the southwest there are three other pairs of snakes, now recog- nized as separate species, which may later prove to be pattern phases. They are : Coluber flagellum frenafum and C. piceus Phyllorhynchus deciirtatus dccurtatus and P. browni Sonora, semiannulata and S', inimata. The first is more a case of color, than pattern dimorphism — a color difference somewhat like melanism, yet not true melanism, since the ventral surfaces are unaffected. Intermediates are found and it is thought that broods will eventually prove these forms identical. One of the peculiarities of the king snake situation is in evidence, namely, C. piceus is found in only a small part of the range covered by C. f. frenafum. Of the leaf-nosed snakes, no intermediates between P. d. decurtatus and P. browni have yet come to light. The only indication that they may be pattern phases, rather than separate species, is a parallel ter- ritorial trend in lepidosis. The ground snakes also are in an uncertain category because of lack of adequate material. They show certain parallel tendencies which may indicate a case of pattern dimorphism ; for example, both forms from Boulder City, Nevada, are peculiarly spotted on the sides. It is to be hoped that the availability of broods may in the future solve these problems as they have that of the king snakes. Acknowledgments I have been assisted in this investigation by Mrs. Elizabeth Leslie, who made many of the statistical computations; and by L. H. Cook, Charles Shaw, James Deuel, and Morris Bloomfield, who made scale counts. The first two broods were hatched at the San Diego Zoo by Robert S. Hoard, the others by C. B. Perkins. The figures were pre- pared by Norman Bilderback. Harold Woodall has kindly supplied data on Los Angeles County specimens. To all I am grateful. Summary and Conclusions A further investigation of the peculiar pattern dimorphism evident amongst the king snakes of the San Diegan region of southern Cali- 22 Bulletin 15: Zoolouical Society of San Diego fornia belonging to the subspecies Lampropcltis yctuhis calif orniae indicates that five pattern types are in evidence. Two of these comprise about ^.X) per cent of the population. The other three are intermediates, two of them being sex-linked. A mother of any phase can produce young of any other phase, but tends to produce a high proixirtion of young of her own phase. Broods hatched in captivity are found to have a higher percentage of aberrants than the general population, indicating a proliable high mortality amongst aberrants. Pattern phases are not correlated with differences in lepidosis. The power to produce the major pattern variant is restricted to a relatively small area within the range of the subspecies. On the periphery of this area the atypical specimens are reduced in number and are imperfect in the nature of their deviations, somewhat re- sembling a small proportion of mixed specimens in the central area. Beyond this periphery there is a much larger area inhabited by the same subspecies but in which there seems to be no power to produce any of the atypical patterns. Note. — While this paper was in press the 1939 broods hatched. Us- ing the classification set forth in tabular form on p. 13 the results were as follows : Pattern of Brood Aberrant Normal Brood Dark Broken Number Mother Ringed Striped Mixed Ventrums Stripes Total 1939-1 Ringed 2 1 .. 1 .. 4 1939-2 Striped 2 5 7 1939-3 Striped 2 7 .. 1 .. 10 Total 6 13 0 2 0 21 These additional broods do not modify any of the conclusions previously reached. Once more mothers of each phase produced per- fect young of both phases. Klauber: Pattern Dimorphism in King Snakes 23 Fig. 6. Contrast of Dorsal Patterns Ringed phase at the left; Striped phase at the right. Fig. 7. Contrast of Ventral Patterns Ringed phase at the left ; hlack ventruni at the right. ( Photographs of anesthetized specimens from San Diego County, California)