Experiments upon the colour-relation, cc. 261 
(f) was a pale (5), pinkish ; almost entire absence of spots. 
(9) 5, light (4), yellowish. 
(h) ,, deep (5), lightish, with typical absence of pigment. 
(2%) 45 » (5), dead; but Mr. Griffiths had noted, ‘‘ Green, 
almost unspotted; thorax dark; bright 
yellow median line.” 
The effects of yellow being so much more characteristic 
and pronounced than those of pink, it is more clear 
than ever that in these results the yellow surroundings 
have had their full influence unaffected by the colours 
which were added later. The strongly pronounced green 
tint of the pup, and the very marked absence of pig- 
ment, entirely agree with those described above, and 
with them confirm Mr. Poulton’s results with yellow and 
orange in a most interesting manner. 
V. Green. 
A. The tissue-paper was of a slightly bluish-green 
colour when held up to strong daylight; otherwise a 
medium green colour. 
Data. — Hight larve, which pupated between Sept. 
Qnd and 18th. One of these was attached to the glass 
covering, although almost surrounded by the green 
background. 
Results.— 
1 pupa was (8), greenish, the patches well-developed. 
2 pupx were light (3), both typical and pinkish. 
1 pupa was dark (4), typical and greenish, very faint colour. 
3 pup were (4), 1 typical and greenish; 2 typical and pinkish. 
1 pupa was pale (5), typical. 
8 
B. In the case of two other larve the surroundings 
were constituted by the green of mignonette-leaves. 
Data.—Two larve come under this head: of these 
one pupated on August 7th amongst mignonette-leaves, 
in a white jar covered with plain glass; the pupa, when 
found, was upon, and surrounded by, the leaves, and 
quite apart from the white surface of the Jar, but without 
any silken attachment. The second larva pupated in an 
earthen jar, concealed amongst mignonette-leaves. 
Results.—Both pupe were deep (5), one being of a 
bright green colour, highly typical, with hardly any 
