14 THE ENTOMOLOGIST. 



already stated, is both fallible and to be applied with discrimi- 

 nation. 



We have now to apply this hypothesis (somewhat digressively 

 explained) to the facts presented in our tables. 



First of all we will examine instances from the red point of 

 view, and then others from the yellow. 



I pass by the species P. apollo, S. ocellatus, and Z. fili- 

 pendulce — since I know nothing as to any yellow relatives of 

 these species — with the remark that if any yellow Parnassius, 

 Smerinthus, or Zygcena exist, the yellow would probably be 

 unaltered by any reagents. But in the Arctiidse and in the 

 genus Catocala we have two remarkably interesting sets of phe- 

 nomena presented, to the which I invite the reader's careful 

 attention. 



First, as to Arctiidse : having found, by so far invariable 

 experience, that a yellow produced by acting on red could be 

 moved no farther, and not having, at that time, excogitated in full 

 detail, any conclusions as to the progressive stability of yellow, 

 as set forth in preceding paragraphs, I was considerably at a loss 

 to bridge over the gulf between the metamorphosed-red yellow 

 and the readily soluble yellow. Now the red of Arctia caia and 

 Callimorpha hsra could — like all other reds — be modified no 

 further back than to yellow ; but it occurred to me that if one took 

 a yelloiv variety of either of these to start with, that yellow might 

 possibly be accessible. I therefore experimented, and with con- 

 siderable interest, on C. hera lutescens but, as the table shows, 

 in vain ; lutescens was utterly unaffected. Then it occurred to 

 me that, by starting still further back, I might yet be successful. 

 I therefore experimented on Arctia villica, — that is, on a normally 

 yellow species, — hoping that the yellow here might be less ob- 

 durate than in a yellow variety of a red species. But here again 

 my efforts were useless, for villica was practically unaffected. 

 I was therefore forced to abandon my hopes of getting the con- 

 nected stages from red to pure white within this group of the 

 Arctiidse ; and from the point of view which I have now attained 

 this is quite as it should be. 



Next, we will take the case of the Catocalce. The crimson of 

 C. niipta is changed to the usual yellow, and of course no farther. 

 Now, among the specimens kindly sent me by Mr. Jones, from 

 America, were two yellow species of Catocala, — viz., C. palceogama 

 with orange hind wings, and C. cerogama with yellow. Thinking 

 it very interesting to have the opportunity of examining, in a 

 second group, the reaction of yellow species related to a known 

 red species, I experimented on C. palceogama. The result was 

 just what I expected in accordance with the theory already 

 explained, and b}^ that time familiar to me. The orange of C. 

 palceogama was quite wwaffected. And here the matter might well 

 have rested; but somehow — I don't exactly know why — it oc- 



