100 Mr. E. B. Poulton. On a special Colour-Relation [Feb. 10, 



did not produce nearly such strong effects as the gilt surroundings. 

 It is almost unnecessary to point how completely the old theory of 

 •pupal sensitiveness is broken down by the analysis. The experiment 

 ■shows that the more elaborate methods alluded to above could be 

 applied to Stage (iii) with at any rate a fair prospect of success. 



7. The Search for the Sensitive Larval Surface, (a.) The Ocelli. — 

 The most obvious suggestion pointed towards the larval ocelli (six in 

 number on each side of the head) as the possible sense-organs which 

 were acted on by surrounding colours, and formed the beginning of 

 the physiological chain of which the end is seen in the colours of the 

 pupa. In many different experiments the larvae were divided into two 

 sets, with precisely similar conditions of surrounding colours and 

 illumination, the one set of larvae being normal while the ocelli of the 

 other larva? Avere carefully covered with an opaque black varnish, 

 which was renewed more than once if necessary (the larvae being very 

 much irritated by the process and flinging their heads about so as to 

 remove some of the varnish). The material made use of was a 

 quickly- drying, photographic varnish, rendered opaque by the addition 

 of lamp-black. Experiments of this kind were conducted with green, 

 white, and gilt surroundings, but the pupae which were formed from 

 the blinded larvae could never be distinguished as a whole from the 

 others, having been equally acted upon by surrounding colours. 

 Even supposing the conditions of experiment had not been quite 

 perfect, so that the ocelli were not wholly eliminated, we should 

 expect some differences between the resulting pupae, if these organs 

 represent the efficient sensory surface. After repeated experiments 

 with negative results, I subjected two sets of larvae to the influence of 

 black surroundings in darkness, thinking it possible (but highly 

 improbable) that the process of blinding, or the varnish itself, might 

 act as a stimulus to the ocelli, and so produce the light-coloured 

 pupae. Again it was possible that the blinding might assist the 

 influence of black surroundings, although it could not prevent the 

 action of bright colours. Of , the . resulting pupae, the set produced 

 from the blinded larvae were rather lighter than the others, but there 

 was little difference, and hence both suggestions were negatived, for 

 the process obviously did not assist the influence of the surroundings, 

 and the difference between the two sets was so slight as to offer no 

 explanation of the brilliant pupae produced from blinded larvae by gilt 

 or white surroundings, on the hypothesis that the process of blinding 

 itself supplies a stimulus. 



(/3.) The Complex Branching Spines. — It seemed possible that these 

 spines, of which there are seven on most of the segments, might 

 contain some terminal organ which receives impression from coloured 

 surfaces. When the spines are snipped off the bases bleed a little, so 

 it is clear that a subcuticular core is contained within them. The 



