October 31, 1907] 



NA TURE 



675 



dark infusion which now occupies, not only the apex, but 

 also a large part of the base of the forewing, and the 

 whole, or nearly so, of the hindwing. After a little study 

 of some of the typical members of each of these geo- 

 graphical groups, it becomes easy to pronounce, with a 

 considerable degree of confidence, upon the local habitation 

 of a species that we may never have met with before. 



If facts of this kind were the only ones with which 

 we had to deal, there might be some justification for 

 adopting the theory of the direct effect of geographical 

 conditions, but it is now incumbent on us to consider 

 whether this hypothesis of common surroundings producing 

 a common aspect will bear further examination. We will 

 take the instance of a group of ant-liUe insects caught 

 by Mr. Guy Marshall in Mashonaland on one day on a 

 single plant. All were to outward appearance ants ; but 

 while the first four were veritable ants, the next two were 

 bugs, and the last was a locustid, belonging, that is to 

 say, to the order of crickets and grasshoppers. If a 

 common environment has of itself produced the ant-likc 

 appearance of the bugs and the locust, why has it not 

 done something towards assimilating the points of struc- 

 ture that do not meet the eye? As a matter of fact there 

 is no such approach. In internal organisation each 

 member of the group preserves the exact characters of its 

 own order. 



There is a certain ant-like locustid, possibly of the 

 same species as that last mentioned, in which the body 

 of an ant is, as it were, painted on that of the locust. 

 The constriction between thorax and abdomen, real in 

 the ant, is in the locustid only apparent. Can the externa! 

 conditions which are supposed to have caused the 

 characteristic shape of the ant actually paint a copy of 

 the ant on the otherwise unaltered body of the locust? 



.\gain, there are cases where the supposed external 

 influence must have acted, if at all, as sculptor instead 

 of painter. In a certain ant-like Mcmbracid (an insect 

 allied to our common " cuckoo-spit ") the body of the 

 insect is concealed beneath a shield, wfiich grows back- 

 ward from the fore part of the thorax. This shield or 

 screen, which is quite separated from the body except 

 along one line of attachment in front, is hewn or moulded, 

 so to speak, into the form of an ant, reproducing even 

 the small swelling in the peduncle which is characteristic 

 of some ants of the region that this insect inhabits. 



Another instance, probably familiar, but so much in 

 point that I cannot refrain from mentioning it, is that of 

 the immature form of a Membracid found by Mr. \V. L. 

 Sclater in Guiana among a number of leaf-cutting ants. 

 The flat green body and brown head and legs of the 

 Membracid make a very fair copy of the ant engaged in 

 its occupation of carrying home the cut leaf, the picture 

 including, not only the ant, but the leaf as well. Ants 

 are avoided by some enemies, though not by all, and in 

 a procession of ants of this kind it is not likely that an 

 enemy, however sharp-sighted, would readily pick out the 

 Membracid from among its leaf-carrying companions. 

 The idea that external conditions can produce in another 

 insect a copy, not only of the ant, but of the leaf which 

 it carries, needs, I think, only to be mentioned to be 

 dismissed. 



Looking at the matter from a slightly different point 

 of view, we may take the instance of the wonderful 

 African butterfly Papilio dardanus, no very distant relative 

 of our English " swallowtail." The male of this insect 

 is non-mimetic, while the female occurs in three or four 

 different forms, each of which is a palpable mimic of a 

 separate model. On the theory of direct external causes we 

 have to explain why these external conditions have brougfit 

 about a resemblance between each form of the female and 

 a separate model of different affinities, while these causes 

 have not been able to prevent individuals of the same 

 species from going off in four or five different directions. 



The facts here have been questioned, but as all the 

 diverse forms have been found among the offspring of one 

 individual, there is no longer any room for doubt that 

 thc'v are all really conspecific. 



We can get more light on the subject if we return for 

 a moment to our assemblage with transparent wings, the 

 assemblage, . that is, which contains the two Dircennas, 

 rhaco and epidero. 



NO. 1983, VOL. 76] 



Now if the effect of transparence, which is common to 

 the entire group, had been the direct result of an external 

 cause, we should expect it to have been brought about in 

 all cases by the same means ; but whereas in the Itho- 

 miines the transparence is due to an alteration in shape 

 and diminution in size of the minute scales which normally 

 clothe the wing, in the Pierines the same effect is pro- 

 duced by a mere diminution in size, the shape remaining 

 unaltered. The Danaines of the group owe their trans- 

 parence to a reduction in the uiiinbcr of the scales, not 

 la any alteration in shape or in size ; while in the 

 associated moths the effect results, not from any change 

 in size, shape, or number of the scales, but from the fact 

 that the individual scales themselves become transparent, 

 and are sometimes set up vertically, so as to let the light 

 pass between them. 



In view of these facts, the investigation of which we 

 owe to Prof. Poulton, it is difficult, if not impossible, to 

 imagine any direct agency which will produce the same 

 visual effect by all these different means. The likeness 

 is superficial ; the real difference is profound. The 

 common features, if we may so express it, are only meant 

 to be looked at. They must stand in relation to vision 

 of some sort ; and to whose vision, we may well ask, 

 if not to the vision of would-be enemies? Natural selec- 

 tion will attain the desired end by any means that come 

 to hand, and these observations of Poulton seem to put 

 every other explanation in this case out of court. If we 

 may be allowed to use, without prejudice, teleological 

 language, we may say that these resemblances have been 

 brought about by natural selection for a mimetic pur- 

 pose. Any variation, whether in size, shape, number, 

 transparence, or position of scales, which leads in the 

 required direction, will be preserved; and the final result, 

 though to ordinary vision identical in all cases, will bear 

 evidence, on close examination, of the manner, different 

 in each individual case, in which it has been effected. 



But, it may be said, many of your instances are simply 

 cases of Batesian mimicry, and for them we can allow 

 the sufficiency of natural selection ; it is the other cases 

 which want explaining. This is quite true, but a great 

 point is gained if we have shown that, in many of these 

 cases, neither aflinity nor the direct agency of external 

 conditions will account for the facts, while natural selec- 

 tion will do so if only we can find out why it should 

 be an advantage for these distasteful types to form them- 

 selves into groups. If we can bring both kinds of 

 mimicry under one cause, we are bound to do so. The 

 old logical canon, the " razor of Occam," applies here. 

 " Entia non sunt multiplicanda praeter necessitatem " ; 

 in other words, having found an adequate cause for one 

 case of a given phenomenon, we are not at liberty to go 

 out of our way to seek another cause for a second case 

 of the same phenomenon. We must first try if the cause 

 already established will not meet the requirements of the 

 situation. 



What we have to do, then, is to prove, if we can, why 

 it should benefit these distasteful forms of various affinities 

 to fall into homocochromatic groups, groups, that is, 

 essentially similar in outward aspect. For a long time the 

 key to the puzzle eluded discovery ; it was at last found 

 by Fritz MiiUer. 



This admirable naturalist, working, like Bates, in South 

 America, put forward in the year 1879 a suggestion which, 

 when developed into its full consequences, has revolu- 

 tionised our conception of the whole subject. 



His suggestion rested on the assumption (since shown, 

 mainly by Lloyd Morgan, to be correct) that birds have 

 no instinctive knowledge of what forms would be suit- 

 able for food and what should be avoided, so that each 

 bird has to gain its knowledge by experience. Hence a 

 certain number of distasteful forms must be sacrificed by 

 each generation of birds until these enemies have learned 

 to leave such forms alone. In other words, each dis- 

 tasteful form has to fav a tax for its immunity. 



Now if two distasteful species resemble each other so 

 closely that birds or other enemies do not distinguish 

 between them, the disagreeable experience gained by 

 tasting an individual of one species will be applied to the 

 benefit of the other, and so each of the two species will 

 only need to contribute a portion of the tax, instead of 



