302 



WILLIAM LAWRENCE TOWER. 



proportion of yl. 83 :wh. 154, and in the third stage the white 

 larvae spHt into 



or approximately 4:2:2:1. Each of these groups then gave 

 rise to three classes of-'adults, precisely similar to those found to 

 arise in the second hybrid generation in Exp. No. H 701, that is, 

 from each of the four larval groups there arose forms like the 

 female parent, others like the male parent, and an intermediate 

 form ; and these three categories of adults are clearly distinguish- 

 able one from the other without intermediates between them. The 

 general results and behavior of this experiment are shown in 

 Plate VII. 



L. undecimlineata 9 X d L. signaticollis. 

 Exp. No. H 701 B. 



A female L. undecimlineata, from Exp. No. 722, g. IV (E), 

 and a male L. signaticollis from Exp. No. 419, g. XIII (CAAB), 

 were crossed under the following conditions: 



Food : Normal — uniform. 



Max. ± 98° F. 

 Min. zb 59° F. 

 Av. 80.2° F. 



R. H. 



Max. d= 95 per cent. 

 Min. =t 40 per cent. 

 Av. 68 per cent. 



and gave rise in the Fj generation to larvae which were all alike 

 in the first stage, but which separated into yellow and white in the 

 second stage, in a proportion of yl. 21 : wh. 28 ; the white larvae re- 

 mained white larvae without spots in the third stage and the yellow 

 larvae remained yellow larvae without spots in the third stage. 



From the white larvae without spots there came a type exactly 

 like the female parent (2 c?, 2?), which, when inbred, con- 

 tinued to breed true; and from the yellow without spots came 

 two classes of adults, a type exactly like the male parent (3 c?, 

 2 9), and a mid type (3 cf, 4 9); of the three types appearing 

 in the F^ generation, which are sharply marked one from another 

 without any trace of intermediates, the observed numbers were : 



