CONCEALMENT IN COCOONS 159 



in the middle of the exposed surface and not at one end, 

 as it is when the moth emerges. Besides, the cocoon 

 had been opened and cracked by a blow from some hard 

 object such as a bird's beak, and the sharp irregular 

 margins were quite different from those of the natural 

 opening made by the moth, doubtless by means of a 

 corrosive fluid, as in the allied species, Dicranura vinula 

 (the " Puss Moth"), which Mr. O. H. Latter has recently 

 shown to secrete caustic potash for this purpose. Further- 

 more, the moth emerges far later in the year, and, had 

 it emerged at an exceptional time, the empty pupal skin 

 would have been left behind in the cocoon. We may 

 therefore safely assume that the opening was the work 

 of an enemy, and, as the cocoon was five feet from the 

 ground, it was probably due to some tree-creeping, bark- 

 exploring species of bird. ... It is probable that the 

 attention of the enemy is directed to any cocoon-like 

 object by the sense of sight and that the object is then 

 tapped, and, if found to be hollow, opened and the pupa 

 devoured.' 1 



The cocoons of bifida are spun in the autumn, but the 

 attack did not take place for several months. The 

 example is probably typical in this respect. The pro- 

 cryptic preparation of the autumn is the adaptation by 

 which the average numbers of the species are kept up in 

 spite of ceaseless bark-hunting during the months when 

 the trees are leafless and food is scarce. The Lamarckian 

 interpretation fails to account for the cocoon-making 

 instinct for two very sufficient reasons : first, a chrysalis 

 is incapable of learning by experience how to improve 

 anything, — even more obviously incapable of learning 

 concerning a structure which it never makes. Secondly, 

 however intelligent a chrysalis may be, the experience 

 itself is of such a nature that its stores of learning cannot 

 be handed down to posterity. 2 



1 The enemies of Lepidopterous pupae enclosed in bark-formed cocoons. 

 — Science, xxiii, 1894, p. 62. The date of the observation is erroneously 

 given as the year of issue instead of 1893. Some of the later sentences 

 of the same communication are also quoted with slight modification on 

 the present occasion. 



3 This argument also is briefly stated in the Proc. Boston Soc. Nal. 



