6 INHERITANCE, FERTILITY, AND SEX IN PIGEONS. 



We are undoubtedly justified in assuming that some principle underlies such 

 definite results of hybridization as Mendel discovered in peas. But that principle, 

 if it be dominance, as supposed by Mendel,' can only be a principle of as man}^ 

 degrees and exceptions as dominance itself, from which it is clear that dominance 

 can be no more than an accident, and an accident that as yet we can discover 

 only by the resulting segregation phenomena. 



Segregation is thus bound up with dominance, following it in all its degrees and 

 exceptions. A "law " must level all contradictions between dominance and blended 

 inheritance. At present the latter seems to be the more general phenomenon. 



Dominance, strange to say, seems to disappear in blend between widely distinct 

 species, and to come more into view as the species are more closely related." 

 De Vries goes so far as to make Mendel's laws a test for species, forms subject to 

 them being thus shown to be of one species (1902, p. 141). This seems strange if 

 dominance means antagonism of characters, since the more highly differentiated 

 characters might be expected to stand farther apart and to l>e more and more 

 difficult to blend. What, then, can be the nature of dominance? Then, too, the 

 less dominance we see in the first generation of hj'brids, the more segregation we 

 see in the second. 



In such questions as we are dealing with — dominance, segregation, ancestral 

 and parental influence, etc. — it would be an advantage to know as intimately as 

 possible just what the specific characters are, how they are related phylogenetically, 

 whether they represent ancient or recent acquisitions. This history of characters 

 will presumably assist in understanding the results of crossing.' 



It is desirable also that the characters for study be of a definable kind — localized 

 differentiations rather than whole characters, such as "self" or "whole" colors. 

 The latter are of interest, but do not tell us so much as the regional character.'" 



The infertility of crosses is not to be confounded with sterility; it stands rather 

 for various degrees of specific incompatibility between the germ-cells of the cross- 

 mated birds. We do not know what this incompatibility means, but it is certain 

 that it is in no way accounted for by external causes, such as difficulties in copula- 

 tion or in the conjugation of the sperm and ovum. In many cases the development 

 of the egg is carried far enough — to the formation of a blood-circle — to show that 

 the sperm has entered the egg and fertilized it. The development may halt at this 

 point or earlier, or be carried on to any later stage, even to hatching. The 3'oung 

 bird may die within a few hours, or live on, ajiparently doing well, for several days, 

 a week, or more, and then tlrop off as if life were a time-fuse calculated to end at a 

 definite moment. 



'See Mendel, p. 27, "If furthermore," etc. 



» The blond ring-dove and the white ring-dove may be taken as two most nearly related forms, and the Japanese 

 turtle-dove and the domestic dove as two wide-apart species, belonging to two different families according to 

 systcmatists. 



" The data on this subject will be found in Volume I. — Editor. 



'" See Weldon, Biometrika, i, 2, Jan. 1902, p. 228. Weldon here gives (p. 24.5) the history of the Telephone pea, 

 and says it exhibits "a gradual series of transitional colors from a deep green to an orange yellow." "Peas of all the 

 transitional forms figured (6) were numerous." They show also "every conceivable condition" between smooth and 

 wrinkled forms (p. 246). According to Mendel, a hybrid pea, like Telephone, should after at least 25 generations 

 contain only round peas, if "round" is dominant (p. 246). "The law of segregation, like the Law of dominance, 

 appears therefore to hold only for races of particular ancestry" (p. 251). De Vries and von Tschermak have offered 

 formulse for other plants, but they have just as little prospect of proving vahd generally as Mendel's formula (p. 252). 



