328 
MR. E. B. POULTON ON THE COLOUR-RELATION BETWEEN EXPOSED 
Stage II., in which the larva rests motionless upon the selected surface and towards 
the end of the stage spins the boss of silk for its subsequent suspension. 
Stage III., in which the larva hangs suspended by its posterior claspers from a boss 
of silk. 
It is to be noted that Stage I. must be of very indefinite length, depending chiefly 
upon the varying proximity of places suitable for pupation. It will be shown that if 
such suitable surfaces are not reached the larva finally makes the best of anything 
which happens to be near, or often pupates in a horizontal position without suspending 
itself. Under ordinary circumstances the larva is exposed to the effective colour 
influence during Stages II. and III. only, for in the previous stage it is wandering over 
surfaces of various colours ; hence the pupal colour must, as a rifle, be determined in 
the two later stages. Further experiments will show that these important stages 
are of more constant length than the first stage. 
(2.) The experiment shows the great power of the white surface in producing light 
and gilded varieties. 
(3.) It shows the great influence of closely adjacent, but comparatively small, dark 
objects in modifying the effect which would have followed from a white surface. Thus, 
of the 10 larvae, arranged in a small group so that the dark-skinned component 
individuals were exposed to mutual influence, 9 were (3) and 1 was (4), while of the 5 
larvae isolated upon the roof and floor 3 were (4) and 2 were (5). No importance is 
to be attributed to the removal of the four first-formed pupae on August 17 (after¬ 
noon), because any effect produced by them on their neighbours must have been 
wrought before the transference took place. 
B. Another set of larvae were blinded, and (August 16, evening) placed in a cylinder 
of almost the same size (2TG decimetres in internal diameter, and 1’02 decimetres in 
height), with a similar roof and floor and amount of food, and the same conditions of 
light. It is, however, to be noted as very important that these larvae had not been 
previously exposed to the influence of a white roof, but w r ere taken from a cylinder 
covered with a sheet of green glass, the whole being placed upon the floor in the 
shadow of the table. To compensate for this, the least mature larvae were selected, 
i.e.y those which, as far as possible, were feeding, or had not passed beyond Stage I. of 
the period before pupation. Nevertheless, the comparison with A. is unsatisfactory, 
because in selecting the least advanced larvae out of very large numbers the results 
are liable to be influenced by the fact that such larvae are often less healthy than the 
others, and frequently do not ultimately attain the average size. On August 21 the 
following 1 results were obtained :— 
o 
