MKXDl^ILTAN EXPRIIIMEXTS WITH FOWLS. 255 



foi- tlie females 3 slightly pignienteil, 3 fully piginenteil, and 

 2 uupigmented. According to my hypothesis the results would 

 be as follows : — 



F, WP J X WP 9. 



Gametes W + P W + P. 



F, WW + 2WP + PP. 



That is 1 unpigmented, 2 slightly pigmented, and 1 fully 

 pigmented in both males and females. 



Bateson and Punnett point out that according to their formulae 

 no fully pigmented males would be expected from this mating, 

 while as a matter of fact a considerable proportion of fully 

 pigmented males actually occurred, namely 12 out of a total 

 number of 75. They regard these as of the constitution ffPPII 

 or ffPPIi, and suppose that in the presence of a double dose of the 

 ]»igmentation factor the efiects of the inhibitor are in considerable 

 measure overcome, in corroboration of which view they found 

 that these specimens became less fully pigmented when adult. 

 But here they have made a mistake in reading their own formulae, 

 for the combination PP occurs only once among the males 

 (namely ffPPli, fig. 4, p. 190 of the paper), and therefore the 

 expecta.tion on their own interpretation would be 1 in 8, not 1 in 

 4 as they assume. The actual numbers are : — 



Males. Females. 



It will be seen that the numbers obtained by experiment of 

 fidly pigmented males are too large for Bateson and Punnett's 

 hypothesis and too small for mine, while it must be admitted that 

 the numbers of females agiee more closely with their assumptions 

 than with mine. If we add together the numbei-s for the two 

 sexes we have : — 



Full. Some. None. 



40 70 34, 



which is a close approximation to the expectation on my hypothesis, 



36 72 36. 



The difi'erences, therefore, between the actual results and the 



