in a wire cage and bred the descendants in the same tempera- 

 ture. We obtained in the second generation of these selected 

 ones with the most purplish color, and we got nine females 

 in which the characters were much mixed; as we lost many 

 caterpillars through disease we did not get many butterflies, 

 and it is therefore not possible to give the exact influence 

 of one or the other. 



Three females had the double ocelli, of the male, on the 

 forewings and the purplish color of the female on the same 

 wdng; and the hindwing showed the yellow ring, also a char- 

 acteristic of the male, but smaller and the circle-field little 

 more purple; two had all the characters of the male, includ- 

 ing a double ocellus ; one had a small appendix on the ocellus 

 of the forewing, but showed much purple, two had no appen- 

 dix, and had the yellow ring around the hindwing ocellus 

 nearly lost ; and one had a double ocellus and purple color on 

 the forewing, and in the ocellus of the hindwing the inner 

 circle-field was completely purple in color. We got 12 males, 

 three of these had a small appendix on the ocellus of the fore- 

 wings ; seven of them had the purple ocellus of the hindwings 

 complete, the other five had this ocellus only partly purplish 

 in color. Upon the whole we find in this second generation 

 that the selection of the most purplish colored butterflies for 

 this experiment improved the purple color, and we got a 

 higher percentage of those with that color. We selected the 

 one female which had the ocelli of the hindwings completely 

 purple in color and mated this with similarly colored males ; 

 this female had also a small double ocellus on the forewing. 

 In the third generation we obtained 10 females and all had 

 purple ocelli on the hindwings, three of these had a small 

 appendix not formed to a double ocellus like this selected 

 mother had; the forewing showed the purplish color in some 

 specimens much extended, not only in the dark scales, but 

 had taken the place up in the two orange yellow colored spots 

 near the costal margin, so that no yellow color was left. We 

 obtained 13 males and also these had now all the characters 

 of the purple female. In all those from the mother in this 

 generation had double ocelli of the forewings; this was so 

 much influenced from the characteristic small ocelli of the 

 purple-colored ones, that they showed only in three females 

 and was therefore so much reduced in size, that only a mod- 

 erately sized appendix was left. 



Now we will say what became of the first pair which we 

 found in copula and of which we raised the caterpillars of the 

 first generation in the glasshouse in normal summer tempera- 

 ture to keep them free of the wilt disease. When these cater- 

 pillars were almost full grown we put them in the incubator 

 in about 80° F. and humidity, to keep the parasites from the 

 chrysalides, the glasshouse was full of these little flics, but 



36 



