Experiments upon the colour-relation, ée. 265 
6. MiscELLANEOUS HixPpERIMENTS.—A. Conflicting colours. 
—T wo larve were accidentally treated so as to be exposed 
to conflicting colours for nearly the whole of the period 
before pupation. These larve, while wandering in 
Stage I., worked their way behind the tissue-paper and 
fixed themselves to the surface of the white jars, so that 
- the ventral surface was exposed to white and the dorsal 
surface to the coloured tissue-paper. 
Data.—One larva (a) was exposed dorsally to Blue 
tissue-paper ; one larva (b) was exposed dorsally to Pink: 
tissue-paper. 
Results.— 
(a) was (4), typical and pinkish. 
(b) ,, (A), typical and strongly pinkish. 
It appears that these results were caused by the white 
surface, aided, or, at all events, not interfered with, by 
the pale colours. It is noteworthy that these pale 
tissue-papers seemed, in the experiments previously 
described, to produce light pupe rather than those with 
darker pigment, such as might have been expected from 
the colours themselves. 
B. Attempted conflicting colours. — A larva haying 
entered Stage III. on the plain deal side of a glass- 
covered wooden box, fied was added anteriorly and 
Green posteriorly. During the pupal ecdysis the girdle 
was broken. 
Result.—The pupa is described in Mr. Griffiths’ notes 
as “pinkish at first, with a well-defined yellow stripe, 
but it faded to greyish green, with darker wing-covers.”’ 
When examined with the others it was a (4) greenish and 
typical. It is probable that this pupa assumed its 
“normal” colour, perhaps slightly directed towards a 
light form by the colour of the deal surface. 
C. Two larve pupated on clear glass jars (containing 
leaves of the food-plant). 
Results.— 
1 pupa ona smooth surface was (4) pinkish, typical. 
1 pupa on a fluted surface was deep (5), very deep green. 
Tin-foil was used as a stimulus in three cases, but the 
commonest grey pup were produced, and this environ- 
ment was probably without effect upon them. 
No very satisfactory conclusions can be drawn from 
these few miscellaneous experiments, the data being 
