It) 



SCIENCE 



[N. S. Vol. XXXIV. No. 862 



Admitting that our present knowledge 

 would lead us to answer these various ques- 

 tions in, the negative, Kellogg closes with the 

 inquiry : 



Then why should the order of inheritance in the 

 silkworm moth be different in the generations 

 after these double matings from the order in the 

 generations following a single mating? 



But is it? I think not. And since it 

 would seem to be of doubtful wisdom to seek 

 possible explanations for a fact which is not 

 true, let us first make sure of the fact. 



Kellogg presents his data without any at- 

 tempt at analysis, and this fact I think has 

 led him, as it naturally does the reader, to 

 suppose that the ordinary Mendelian inherit- 

 ance of cocoon color is in these matings much 

 disturbed. Before any critical discussion of 

 the data is possible they must first be classi- 

 fied. Comparison may then be made with the 

 behavior of the same races in single matings, 

 which Professor Kellogg himseK has given 

 us in part in his 1908 paper. 



In agreement with Coutagne (1902) and 

 Toyama (1906), Kellogg finds that the in- 

 heritance of cocoon color in silkworms fol- 

 lows in general the Mendelian laws of domi- 

 nance and segregation. Yellow color usually 

 dominates over white in crosses, but in the 

 case of certain races the dominance is not uni- 

 form. Yellow dominates over white in part 

 of the zygotes only, in the remainder white 

 dominates over yellow. This fact was first 

 observed and clearly recorded by Coutagne 

 (1902), who, though at that time unac- 

 quainted with the Mendelian laws, presented 

 clear and convincing evidence of their appli- 

 cability in the cases studied by him. Discus- 

 sing Coutagne's results in 1905* I pointed out 

 the fact (which Kellogg seems to have over- 

 looked) that in cases where yellow dominates 

 in Fj, there occur in F. three yellows to one 

 white, whereas, when white dominates in Fj, 

 there occur in F, three whites to one yellow. 

 In other words the character which behaves 

 as dominant in F, continues to behave as 

 dominant in F.. Kellogg's experiments show 



^ Carnegie Institution Publication, No. 23, p. 59. 



this same result both in single and in double 

 matings, as I shall presently point out. 



It happens that Kellogg has used in his 

 double mating experiments a white race 

 (Bagdad white) which is sometimes domi- 

 nant,^ sometimes recessive in crosses with yel- 

 low, and this seems to have been the real rea- 

 son why he thought the " inheritance per- 

 turbed " by double mating. As a matter of 

 fact the perturbation is no greater in the 

 double than in the single matings. 



In 1908 Kellogg reported the results of six 

 crosses of Bagdad white with Istrian yellow, 

 which were carried through two generations, 

 as shown in Table I. One of the seven orig- 

 inal matings produced only yellow offspring, 

 two produced only white, and four gave a 

 mixed progeny consisting of 82 whites and 

 60 yellows. Six matings of F^ yellows inter 

 se produced in F^, 117 whites : 350 yellows, 

 or Iw : 3y. Six matings of F^ whites inter se 

 produced in F^, 418 whites : 140 yellows, or 

 3w : ly. In both cases, it will be observed, 

 the character which dominated in Fj was in 

 F, approximately three times as numerous as 

 the other, a consistent Mendelian behavior. 

 Further, when white dominant in F, was 

 mated with yellow dominant in F^, the result 

 was the production of both sorts in numbers 

 approximately equal. Seven such matings 

 produced, in F,, 324 whites and 381 yellows, 

 or Iw : 1.2y. 



^Kellogg expressly recognizes (p. 784) the fre- 

 quent behavior of Bagdad white as a dominant 

 character in crosses with yellow in single matings, 

 yet on page 788 makes the surprising statement: 

 ' ' After a double mating the whites of the Fj gen- 

 eration mated with other whites of the same gen- 

 eration do not always produce whites. They may 

 produce both yellows and whites." [Certainly! 

 If white is dominant, it should behave in just 

 that way.] Kellogg then continues: "Or this 

 latent carrying of the yellow character by these 

 presumably strictly recessive (white) carriers may 

 not be manifest tUl an F3 generation. ' ' Kellogg 

 then proceeds to seek an explanation in the double 

 mating of the mother, having forgotten appar- 

 ently his express statement on page 784 that 

 Bagdad white frequently behaves as a dominant, 

 which fact would explain everything. 



