404 MR. E. B. POULTON ON THE COLOUR-RELATION BETWEEN EXPOSED 
In order to test my suggestions as far as possible, I commuuicated with different 
naturalists in various parts of the world. 
Mr. Roland Trimen kindly sent me the following letter from the Cape :— 
“ As regards gilded butterfly pupae, I am unfavourably placed here for observation 
of them. 
“We have only P. carclui and Danais chrysippus at this end of the country. The 
former varies much in the amount of gilding on the back and wing-covers, but I think 
that the three rows of dorsal tubercles are always brightly gilded. This larva does 
not seem particular as to site of suspension, apparently hanging itself indiscriminate! v 
to plants, walls, fences, etc. ; but I have found it more frequently on walls, especially 
under the coping. 
“ The latter (D. chrysippus) has only a half-girdle (dorso-abdominal) of contiguous 
golden dots—set off by an immediately preceding black tuberculated ridge—and eight 
scattered golden spots about the head and thorax. This pupa is either green or 
pinkish, and sometimes of a tint combining both those colours. It is usually 
suspended to its food-plant. 
“ ‘ I have not found the variable colouring of this pupa to accord with its immediate 
environment, though I have allowed the larvae in confinement free choice of various 
convenient surfaces for pupation, with the view of ascertaining whether there was 
any relation between the green or reddish tint and the colouring of adjacent objects. 
It seems not improbable that this brilliant pupa stands in no need of special 
protection, but, like the imago (and apparently the larva also), is avoided by insecti¬ 
vorous animals.’—Extract from my forthcoming work, ‘ South African Butterflies.’ 
“ Other Danaine pupae appear to be exceedingly brilliant. Thus Thwaites (Moore’s 
‘ Lepidoptera of Ceylon,’ p. 2) says : ‘ The suspended chrysalids ’ (of the Danaime) ‘ are 
brilliantly metallic in colouring’; and Boisduval (‘ Faune Ent. de Madag.,’ etc., 
pp. 36, 37) describes thus the pupa of Euplcea goudotii: ‘ La chrysalide ressemble a 
une bulle d’or extremement brillante ’; and thus that of Danais phcedone : ‘ La 
chrysalide est . . . d’un vert dore brillant.’ 
“ I should incline to the view that in this protected (distasteful) group of the Danaime 
the conspicuously brilliant colouring of the pupae is a warning signal of ‘Not fit to eat!’ 
“ Is it possible that the gilding in pupae of other (not distasteful) Butterflies may 
originally have been acquired as protective mimicry of the brilliant Danaine ones ? 
“ The only other South-African gilded Butterfly pupae that I know of are those of 
the genus Atella (nearly allied to Argynnis). In these two species (which I only know 
by drawings and descriptions) the gilding (which in one is rather silvery than golden) 
is confined to spots on head and thorax, narrow borders of wing-covers, and dorso- 
abdominal raised spots bearing either pointed tubercles or thin spines. 1 have no 
information as to the objects to which these pupae are usually attached. 
“ The pupae of the Acraeinae (which, as you know, are a protected group of Butter¬ 
flies), though (as far as I know) never gilded, are yet exceedingly conspicuous, their 
