INHERITANCE OF WHITE WING COLOR, A SEX-LIMITED (SEX-CONTROLLED) VARIATION IN YELLOW PIERID BUTTERFLIES JOHN H . GEROULD Dartmouth College. Hanover. New Hampshire Received May 24. 1923 TABLE OF CONTENTS PAGE INTRODUCTION................................................................... 495 Terminology .................................................................. 497 Acknowledgments .......................................................... 498 Previous investigations............................................................ 499 Inbreeding heterozygotes; narrative of experiments sincc 1910 ......................... 500 Homozygotes fawhite ........................................................... 502 Homozygotes inbred and later developments ........................................ 503 Types of mating................................................................. 505 Discussion of the regular ratios in tables 1 to 6 ................................... 505 Unexpected ratios ............................................................. 508 R6sum6 .................................................................... 519 A7:9ratio ................................................................. 519 Stability and proportions of the white variety in the general population ................. 520 Breeding experiments withother species of Colias .................................... 526 Colias christim Edw ........................................................... 526 Colias myrmidone Esp., ........................................................ 528 White wing color and climate ..................................................... 529 Comparisons with Papilio ........................................................ 532 Sex-controlled characters in Argynnis paphig and Brtcchus quadrimaczrlatzu ............... 537 Some facts and problems of sex physiology. Comparative velocity of development ...... 541 Summary .................................................................... 544 SU~ARYANDCONCLUSIONS...................................................... 547 LITERATURECITED ............................................................... 550 INTRODUCTION The study of polymorphism in butterflies opens up problems of great interest to the student of evolution and heredity . The different varieties of a polymorphic species that are differentiated by genetic rather than by environmental factors may represent, as DARWINlong ago pointed out, possible incipient species . Such variations are now generally believed to have arisen. not by a long process of natural selection of minute changes. but by a comparatively simple and direct process of mutation involving very few genetic changes. GENETICS8: 495 N 1923 496 JOHN H. GEROULD These variations in butterflies are in many cases sex-limited, or, to use GOLDSCHMIDT’Sterm, sex-controlled. They dependupon hereditary factors the effects of which are realized in only one sex, which in the cases thus far studied in Lepidoptera is the female, though dimorphism in the male is known to occur (Lycaenidae, Arctiidae). Studiesupon sex-limited characters show thatthe differentiating genes, though determining the coloration of the female only, are trans- mitted, not by sex-linkage in (C criss-cross” fashion,. but equally by both sexes, like anyother non-sex-linked character.The phenotypically uniformmale carries and transmits to both sexes of his offspring the determiners of one or another of the different types of female. Cases of sex-limited dimorphism in the female are not infrequent among Pierid butterflies, and, so far as I have investigated them, they corre- spond genetically inall essential respects to the polymorphism of the tropical Papiljos with three types of female. In addition to the ordinary yellow (or orange) female in several species of Colias, a white variety occurs. This is true not only of Colias philodice and C. eurytheme,which I have bred extensively, but also of C. interior and C. christiaa,in America, C. edusa, C. myrmidone, C. hyale and other palearctic species. It is true also of the closely related Terias Zisa of our southern states. In certain species of arctic or sub-arctic distribution, such asC. interior, the white variety is said to be more common than the yellow, while in other species (C. scudderi and C.pelidne) all females are white. In C. philodice and C. eurytheme, however, the white variety is much the rarer, though in restricted localities in New England and New York the two forms of the female philodice are said to be about equally abundant. Nearly every summer and sometimes a part of the following winter, for thirteen years, I have bred the white female variety of C. philodice and that of C.eurytheme. Colias philodice, the common yellow clover but- terfly of the eastern and central northern states, furnished the material for study during the first and last seasons of the study (1908-1911, 1919- 1922); C. eurytheme, the orange alfalfa butterfly of the western states in 1913, 1914 and 1916; and hybrids between the two species in 1912, 1913 and 1914. The same genetic phenomena are exhibited in the inheritance of this dominant female variation, whitewing color,in bothof these species and in hybrids between them. The white variety may be introduced into the species cross from either species, andits inheritance may be studied simultaneously with that of orange Dersus yellow. In 1912 I introduced WHITE WING COLOR IN YELLOW PIERID BUTTERFLTES 497 white into a species cross through the yellow male of C. philodice, in 1914 through the orange male of C. eurytheme. Both were sons of white females, but their mates came from strainsof the other species that were free from the hereditary factor for the white female coloration. The term albino sometimes applied to thewhite female is objectionable because the variation is not dueto thelack of pigment, nor is it, like true albinism in mammals and birds, a recessive unrelated to sex. It is rather a dominant sex-limited color variation in the pigment of the scales that furnish the ground color of the wing. This pigment derived from uric- acid products that accumulate in the blood of the pupa (HOPKINS1906) is normally yellow or orange. The special features of the female wing pattern(broad, yellow-spotted blackborder, contrasted with the nar- row, unspottedborder of the male pattern)are unaffected, only the ground color of the dorsal surface of the wings being atypical. Although white female coloration in C. philodice and C. eurytheme is a definite and pronounced discontinuous variation, it is suffused in many and indeed in most individuals with a slight, fluctuating tinge of yellow due to secondary factors. The recessive non-sex-limited variation in larvalblood color, blue-green, which appeared recently in my cultures of C. philodice (GEROULD1921), which preventsthe appearance of a yellow elementcharacteristic of normallarval skin color, pupalcuticula, the eye of the adult and the egg, is wholly independent of the yellow pigment of the wing scales. The genes determining white wing color and that producing the succession of changes in the life cycle just mentioned, react upon two different yellow pigments both of which are present in the blood of the normal larva and PUP". Terminology The term secondary sexual characters has been used in a broad sense to include not only those characters peculiar to individuals of either sex that are secondarily connected with reproduction, like clasping-organs, milk glands, brood-pouches and the like, but also many traits that serve no conceivable functionconnected with reproduction, like the color patterns and ground color of unisexually dimorphic butterflies. It has been suggested that the termsecondary sexual character should be restricted to those traits and organs obviously connected secondarily with reproduction, like sex differences in genitalia and instincts. If that is done, sex differences not obviously connected in any way with repro- duction would naturally be called tertiary sexual characters. GENETICS8: N 1923 498 JOHN H. GEROULD In the present study, however, I have used the term secondary sexual character in the broader sense, as including these tertiary sexual, or sex-limited, characters. Unfortunately the term sex-limited has had a chequered history, having been applied loosely to several different things. Sex-linked characters, subject to ((criss-cross” inheritance, shuffled from one sex to another in two successive generations, were for a time, especially in Europe, called sex-limited, a confusion which MORGAN(1914) long since corrected. A character that shows sex-reversal of dominance, namely, horns in sheep, has been treated by WENTWORTH(1912,1916) and others, as a typical sex- limited character. GOLDSCHMIDTand FISCHER(1922) have recently proposed the word “sex-controlled” as applying to the varieties of unisexually polymorphic butterflies, the factors for which are transmitted equally by either sex, though realized phenotypically only in one.While this term is not specifically descriptive, for sex-reversed dominance in sheep is of course also sex-controlled,it has the advantageof being free from confusing conno- tations andcapable of serving as atechnical term for this type of heredity. The genes for sex-controlled characters do not express themselves in the male in the cases thus farstudied. Their development and hence their reaiization,is forestalled and inhibited by the development of the characters of that sex. They are latent in the male, patent in the female, sex-con-