398 Mr. G. A. K. Marshall on 



. . . Their terror of the insect was most amusing, and 

 was an eloquent testimony to the great value of this form 

 of colouring to so bulky a larva. I do not think any one 

 could now argue that the theory of terrifying coloration is 

 far-fetched, as I have heard contended. The snake-like 

 appearance seems capable of deceiving more intelligent 

 animals than baboons, for it is not long since I received 

 a box containing a mutilated specimen of this caterpillar 

 accompanied by a note inquiring, " Is this a snake ? " 



[This evidence recently obtained by Mr. Marshall, 

 added to that already published by Professor Weismann, Lady 

 Verney, and the present writer (" Colours of Animals," 

 London, 1890, pp. 260, 2G1), leaves no doubt that the 

 conspicuous eye-spots of Gkcerocampa and other large 

 larvse are really terrifying and do actually alarm their 

 enemies. The results observed are consistent with the 

 production of a feeling of terror rather than of distaste or 

 repugnance such as Portschinski supposes to result from 

 the sight of an ocellated spot. In his remarkable papers 

 on " Coloration marquante et Taches ocellees " (St. Peters- 

 burg), this acute and imaginative naturalist states his 

 belief that ocellated spots represent the appearance of a 

 drop of warning liquid. He develops this hypothesis 

 with the greatest ingenuity, and describes and illustrates a 

 large number of such spots in insects of many kinds. In 

 some oceUated spots he sees represented the reflection of 

 the sky in a drop of warning liquid ; in others, the dis- 

 torting effect of gravity upon a drop resting on a 

 vertical surface : in the sounds made by certain irritated 

 Mantides, as they display the spots on their raptorial legs, 

 he believes he hears a rejiresentation of the rushing sound 

 of a war)iing liquid forced through a fine aperture. My 

 kind friend Professor W. R. MorfiU has given me the 

 opportunity of learning the remarkable and highly-imagin- 

 ative views of the distinguished Russian naturalist. On 

 some future occasion I hope to be able to lay them before 

 English-speaking naturalists in much greater detail. For 

 the present I desire to point out that the results obtained 

 by experiment do not support his conclusions, but suggest 

 in the most convincing manner that terror, such as is 

 caused by the appearance of a serpent, is produced by the 

 display of eye-like marks on a large caterpillar. Terror 

 may be similarly caused by the display of large 

 ocellated spots on the wings of imagines, while in other 



