VARIATION AND ABEHRATION IN BUTTERFLIES. 149 



on V. atalanta, Dr. F. A. Dixey discusses the question of phylo- 

 geny. He finds phylogenetic relations between the heat-modi- 

 fication of V. atalanta and V. callirrhoe, F., &c. ; also between 

 the cold-modification of V. atalanta and Grapta c-aureum, L., 

 Pyrameis goncrilla, F., &c. 



And now, in conclusion, a few words as to the influence of 

 moisture on the pupa. 



In my 'Manual,' pp. 124, 125, the fact is recorded that out 

 of a large series of puppe of Exdromis versicolor, L., Saturnia 

 jnivonia, h., and Aglaia tau, L., I had a few emergences without 

 previous hybernation of the pupa. The aspect of the individuals 

 belonging to the first two species was markedly altered. 



The numerous temperature-experiments of the last few years 

 just recorded, which resulted in so remarkably evident a reaction 

 of the species selected for experiment, lead unquestionably to the 

 conclusion that a case of this land is not purely a matter of 

 innate or primary tendency on the part of individuals whose 

 constitution is biologically abnormal, but rather that some 

 external factor must here have been brought into play. 



Since then I have made the following observations. Large 

 numbers of pupa3 of Saturnia (in these cases from 200 to 400 

 specimens were used) were kept very dry for 7 — 10 weeks from 

 June to the end of September. They were then freely and 

 repeatedly moistened, and about 1 per cent, of the moths emerged 

 fiom these pupte 10 — 20 days after the damping. 



The fully-developed moths mostly show a departure from the 

 ordinary form of the species, which may be characterised as 

 follows : the elements of the pattern are not sharply outlined, 

 but more or less washed-out and confused. 



As I Lave frequently repeated this experiment with the same 

 result, it cannot possibly be a matter of mere chance ; and the 

 facts already mentioned in my 'Manual,' and given at length 

 above, may be referred to similar causes which were disregarded 

 at the time. 



Further, on a comparison of these observations it seems 

 probable that similar conditions in nature may have similar 

 results, i. c, that the exceptional appearance at midsummer or 

 in autumn of species which as a rule do not occur at such times 

 in the perfect state, but hybernate as pupas, may have similnr 

 causes ; that is to say, a copious rainfall after a long period of 

 drought. 



Supposing then the numbers of such abnormal individuals of 

 a species (whose descendants must, it is true, accommodate them- 

 selves to widely different conditions of life) to be sufficient for 

 the permanent establishment of the abnormality, then these 

 individuals will form the starting-point for a new series of de- 

 velopments, which in course of time, in consequence of the 



