Febbtiaet 10, 1911] 



SCIENCE 



221 



both edusa and helice types. He remarks that 

 the fact that edusa ? can come from helice ap- 

 pears to disagree with his hypothetical 

 scheme, but he adds that the typical female 

 differs from the male, and suggests that there 

 may be or have been a possible " a type " of 

 female resembling the male (such a type is 

 well known in the Papilio under discussion). 

 It is interesting to be able to state that the 

 hypothetical " a type " of female C. edusa is 

 actually known, and may be found mentioned 

 in Entomologist, 1889, p. 26. 



I can not help thinking it probable that 

 whether or not the precise Mendelian hjrpoth- 

 esis offered by Professor Punnett is justified 

 by subsequent research, the facts will be found 

 to be very much as he has postulated. I do 

 not think, however, that the theory of mim- 

 icry is thereby contradicted. According to the 

 old view that all organisms are everywhere 

 varying (aside from non-inherited environ- 

 mental effects), and that natural selection is 

 necessarily in continuous operation to keep 

 them constant or to modify them as needed, it 

 must be confessed that some of the observed 

 facts are hard to interpret. According to the 

 newer view that " original variations " happen 

 at relatively rare intervals, and that a stable 

 type once produced may continue indefinitely 

 if not discriminated against, the matter as- 

 sumes a very different aspect. Consider the 

 great antiquity of insect genera, as shown by 

 fossil remains; consider the kaleidoscopic 

 changes in insect-type producing innumerable 

 species often without material advance in the 

 general type; with all this time and change 

 there must have been produced many pairs of 

 more or less unrelated species resembling one 

 another. When this resemblance has been ad- 

 vantageous it has been preserved, while other 

 forms have died out; and hence to-day the 

 proportion of such cases is vastly greater than 

 we could expect from chance coincidence. It 

 is not necessary that everywhere and at all 

 times mimicry should be functional; the evi- 

 dence seems to show that it generally is, and 

 that is sufficient. Indeed, if a type has been 

 preserved because of its ability to " throw " 

 mimetic forms, it is likely enough to continue 



to do this, even in places where this is unnec- 

 essary. 



Those who are confronted by the vast array 

 of insect species rarely think of the unseen 

 gaps in the ranlts. These may perhaps be best 

 appreciated by considering the fauna of the 

 Hawaiian Islands, as elucidated by Sharp, 

 Perkins and Walsingham. Here we have 

 large genera with multitudes of allied species, 

 no doubt the result of the immigration in an- 

 cient times of single types of a few groups. 

 Comparatively free from the stress of com- 

 petition these Hawaiian groups have, as it 

 were, nearly their full membership; on con- 

 tinental areas only remnants usually remain. 



Thus I think that the newer work on hered- 

 ity, read aright, only strengthens the theory 

 of mimicry, by relieving it of a load it was ill- 

 fitted to carry. I do not see any other plaus- 

 ible way of accounting for the facts, unless it 

 is by supposing that similar environments 

 give rise to similar modifications of the germ- 

 plasm. This idea loses support when we re- 

 member the cases (e. g., in butterflies and 

 bees) in which the same superficial appear- 

 ance is due to entirely different structures. 



It may still be debated whether natural 

 selection has had much to do with the pro- 

 duction of mimetic forms, in the sense of 

 bringing about the accumulation of favorable 

 variations. Por my own part, I can not doubt 

 that this cumulative effect of selection is real, 

 and is a necessary cause of the more striking 

 and complex instances of mimetic resemblance. 

 The rarity of original variations, while great 

 enough to relieve selection from the necessity 

 of acting continuously on all characters, is 

 doubtless not so great as to prevent it from 

 bringing about many striking cumulative re- 

 sults, in the manner postulated by Darwinians. 



I have wandered too far away from Mr. 

 Eltingham's book, but I must return to it to 

 mention his remarkable experiments with the 

 larva of a moth, Odontoptera hidentata. 

 Larvae fed on ivy were offered to a lizard, and 

 found extremely distasteful. They were, al- 

 though of cryptic coloration, nearly always re- 

 jected by the reptile. Several larvae were then 

 transferred to apple, and after feeding on this 



