SPECIFICITY OF GENE EFFECTS 



251 



tween genes. One may well ask if there is any other manifestation of heterosis 

 in these species hybrids and backcross hybrids. Extensive measurements of 

 eight body characteristics, as over-all length, extent, width of tarsus, width of 

 band, length of wing, beak, middle toe, and tail, were made over a 7)eriod of 

 years under the supervision of L. J. Cole. The differences in the averages of 

 these various characteristics between the parental Pearlneck and Ring dove 

 species, as yet unpublished, were statistically significant, and the averages of 

 the measurements of these characteristics in the species hybrids showed them 

 to be in general intermediate between those of the parental species. Thus 

 there was no evidence of heterosis in any external characteristic of the species 

 hybrids, and no correlation with the hybrid substance of the blood cells. 



CELLULAR CHARACTERS WITHIN A SPECIES 



The finding that one or more genes on each of nine or ten pairs of chromo- 

 somes of Pearlneck had effects on the species specific antigens of the blood 

 cells of this species made plausible the belief that many more genes than 

 commonly believed would have effects within a species making for individual- 

 ity of the cellular patterns. Acting on this assumption, a series of exploratory 

 tests w^ere made in experimental animals, principally in cattle and chickens. 

 For example, following the transfusion of the blood of a young cow into her 

 dam, an antibody was obtained from the serum of the recipient which re- 

 acted (produced lysis of the reacting cells upon the addition of complement 

 to the serum-cell mixture) with the cells of some individuals, but not with 

 those of others. The reactive substance was called A. 



The objective was to be able to detect each antigenic factor separately, ac- 

 cording to the following criterion. The reactive cells from any individual 

 should remove the antibodies from the reagent specific for those cells, when 

 added in excess to the reagent. However, if there were antibodies in the re- 

 agent which recognized two or more distinct blood factors, any such ab- 

 sorption with cells containing only one such substance would remove only a 

 part of the antibodies. Those remaining would still be reactive with all cells 

 containing the substance corresponding to the unabsorbed antibody. 



To this criterion was added that of genetics for a single character, using 

 the gene-frequency method since controlled matings were not possible. A 

 typical example of the analysis is that for substance A, as follows: 



