Experiments upon the colour-relation, ce. 253 
hoped that future investigators will remember that the 
early part of Stage IIT. is the time which most of all 
requires defining in various species; for it may be 
assumed that Stage II. is the chief time of suscepti- 
bility in all species. 
It will be seen that the larve attached themselves, in 
nearly every instance, to the coloured-paper linings, in 
preference to pupating upon the mignonette leaf in the 
jar. A single larva, however, which was surrounded by 
tinfoil, fixed itself to the leaf, but subsequently died. It 
appears at first singular that the larve should have so 
generally avoided the leaf of their food-plant, but a 
probable explanation may, I think, be found in the fact 
that the pup hybernate, whilst nearly all their food- 
plants are annuals, or die down in the winter. 
Conflicting colour experiments were also attempted ; 
but as Mr. Griffiths added the second colour, and in 
many cases both colours, after the beginning of Stage 
III., it is quite clear that no real test was imposed. 
Those cases in which one colour was allowed to act 
during the whole period before pupation, and a second 
added later beneath part of the body at a probably non- 
susceptible time, may be considered as practically single- 
colour experiments. The conflicting colours were the 
same as those made use of in the single-colour experi- 
ments, with the addition of red and gold. 
3. Tue Nature or THE CoLourRs EMPLOYED. —The 
coloured tissue-paper made use of was in some respects 
unfortunate, the paper in most cases being so transparent 
as to admit much white light, greatly diminishing the 
colour-influence.. On the other hand, the adoption of 
colours, somewhat different from those used by Mr. 
Poulton, has proved to be the reverse of detrimental to 
the experiments, since a fuller opportunity of comparing 
the results of near tints has been afforded. The results, 
as will be seen, are equally satisfactory with those of 
Mr. Poulton, and are entirely confirmatory of those 
obtained by him with stronger and generally opaque 
colours employed as a back-ground for the pupe. 
Further, the adoption of so much paler tints has even 
proved a distinct advantage, in that we can more fully 
TRANS. BENT. SOC. LOND. 1888. PARTI. (JUNE.) T 
