TRANSACTIONS OF SECTION D. 693 



sensitive to the colour of surrounding surfaces. Such an explanation appears to be 

 merely a metaphor borrowed from photography, and it is furthermore unsupported 

 by any proofs. There is, in fact, much a priori exception to be talien to it, inasmuch 

 as it implies that all those pupas which throw off the larval skin on a dark night 

 must lose the advantages of this form of protective resemblance. It is much more 

 probable that the effect is produced by the action of surrounding colours upon the 

 larva during the time (long enough to include many liours of daylight) in which 

 the latter rests upon the surface where pupation will take place. In the first place, 

 I made many experiments in order to test the accuracy of Mr. "Wood's observations, 

 resulting, as I had anticipated, in the most complete confirmation. Among the 

 pupae experimented upon was that of the common tor'toiseshell butterfly ( Vanessa 

 urticce). It was found that by causing pupation to take place upon a white or 

 black screen very different results could be produced. The pupse upon the white 

 screen were often brilliantly golden, and quite unlike all the ordinary forms assumed 

 by this species, and well known to entomologists. (Juriously enough, however, the 

 pupse of this and other species, which are full of parasitic larvre of ichneumons, 

 and which can never produce butterflies, are often as brightly gilded as those upon 

 ■which I experimented. But my pupte were perfectly healthy and produced normal 

 butterflies. Considering the effects of ray first experiments, it appeared probable 

 that an artificially gilded surface would produce even stronger results than a white 

 screen, and experiment soon confirmed this prediction. Such a result seems to 

 imply that the metallic lustre of many e.xposed pupse harmonised with equally 

 brilliant objects among the vegetal oV, more probably, the mineral surroundings. 



The next point was to ascertain the period during which the larva was sensitive, 

 and incidentally to confirm in the mOist complete way my suggestion that these 

 effects are due to the larva itself, and^ not to the freshly formed pupa. These 

 objects were achieved by transplanting the larva at various periods before pupation 

 from one surface to another, which wasXknown to produce an opposite effect. 

 It was thus found that the larvae are sensitive for many hours — even more than a 

 day — before pupation takes place. It was then necessary to ascertain if possible 

 the nature of the larval sensory surface whichis affected by surrounding colours ; 

 and, first of all, the ocelli were eliminated by covering them with an opaque varnish 

 (renewed if necessary), but this treatment did no\ afliect the result. Equally in- 

 effectual was the result of snipping oft' the larval Ijristles, which it was thought 

 might possibly contain the desired sense-organ. Then another method was adopted : 

 as soon as the larvae suspended themselves head downwards (many hours before 

 the change takes place), they were surrounded bj' tubes so constructed that the 

 head and anterior part of the larva were contained in a gilt chamber, while the 

 rest of the body was surrounded by black walls, tending to produce an opposite 

 eflect, the two compartments being separated by a perforated disc which just 

 allowed the larval body to pass through. The head in the lower chamber was 

 always turned on one side, so that the colour of the upper compartment could not 

 possibly affect it. In other cases the colours of the compartments were reversed. 

 When I first looked at the pupse in these tubes I fully believed that the colours 

 followed those of the chamber in which the larval bead had been, and I was much 

 puzzled by this, inasmuch as the only likely sense-organ — the ocelli — had been 

 already eliminated by previous experiments. But when the pupjp were taken out 

 of the tubes, and placed side by side upon white paper, I found that the efiects I 

 have described were entirelj^ tnisleadiug, being due to reflection from the walls of 

 the lower chamber when gilt, and to a dark appearance due to the surrounding 

 black surface in the other cases. The negative result obtained seems to indicate 

 that the sense organ exists in the skiu, or that possibly the light acts in a more 

 direct way upon the larval skin without the intervention of the nervous system. 

 The full results of the investigation will not, however, be obtained until an 

 immense amount of work has been bestowed upon the notes made during these 

 experiments, in which many hundreds of individuals have been employed.^ 



' See Proc. Soy. Soc. No. 237, 1885, and No. 243, 1886, and Trans. Ent. Soo. Lon- 

 don, Pt. I. 1884, Pt. II. 1885, and Pt. II. 1886 for the experiments and observations 

 upon larval colours alluded to above. 



