THE COLORATION PROBLEM. 



153 



swallow that, but I can see the logic and harmony that at once reigns 

 when the Batesian and Miillerian theories are brought to bear upon 

 that wonderful assemblage of startlingly decorated models and their 

 faithful copies. [Professor Poulton has pointed out that precisely the 

 same difficulty lies in the way of explanation by External Causes in 

 the Nearctic Region (1908), Trans. Knt. Soc. Load., p. 449 fi seq., and 

 Rev. K. St. Aubyn Rogers has shown the same difficulty exists in B. 

 East Africa (1912), Proc. Ent. Soc. Lond., p. 496, and Professor 

 Poalton has called attention to the same difficulty in the Planema 

 alcino'e, P. iiiacarista and Acraea alciope combination, Bedrock loc. cit., 

 pp. 57 et sei].] . 



Moreover the ochreous-orange of H. chri/fiippiis and H. diocippus is 

 not a prevalent colour amongst blue butterflies in my experience, 

 which I admit is limited. The browns associated Avith blues, except 

 the little lunules on the wings, are the dull Satyrid browns and not 

 the warm browns of the Danaidae. 



Besides all this, if temperature be the explanation, why have the males 

 retained their ancestral and conspicuous dress? One would have expected 

 them to have varied likewise under the same stress. Again, p. 456, Lieut. - 

 Col. Manders argues on the basis that variations are necessarily small. 

 They often are, and natural selection is supposed to act upon such small 

 variations, though many of the aberrations and variations that have 

 come under my notice in nature have been rather marked departures 

 than gradual. W. Bateson, in Materials for the Stiidi/ of Variation, 

 has collected a mass of evidence showing that variation often takes 

 place by quite perceptibly sudden steps. When the variation has 

 appeared, the principal of Mendelian heredity would tend to cause it to 

 reappear, and if an advantageous variation, natural selection would 

 gradually stereotype the variation to an extent that might turn the 

 particular set of peculiarities which make the variation into an 

 ascendant form, {confer Prof. Poulton, Bedrock, loc. cit., p. 63, for 

 example of such a variation in Acraea alciope). Once again, Lieut. -Col. 

 Manders argues that it is difficult to see how a small variation could 

 be selected by a bird or deceive a bird, but it may well be that the 

 detail in which we see synaposematic coloration to-day is the result of 

 the increasing discrimination of birds, as the discriminative faculty 

 would be enhanced in each generation by the fact that the bird who 

 successfully discriminated, would have a better area of food supply, 

 because it would be sure of palatable insects all the time, instead of 

 chancing unpalatable insects, which judging by Mr. Marshall's 

 experiments on the Mantis, seem to be positively harmful if sufficient 

 of them be consumed. [Confer, G. A. K. Marshall (1908) Trans. Ent. 

 Soc. Lond.. p. 96.] I find I have expressed myself almost in Mr. 

 Marshall's words, but I had not read his paper since it appeared, and 

 only referred to it after. On the whole Mr. Marshall has expressed 

 what I mean much more forcibly and clearly than I have done. 

 Professor Poulton remarks tome in litt. January 2lst, 1913, "In 

 synaposematic approach we do not need that a bird be mistaken, 

 provided it is reminded of a previous experience and approaches the 

 Miillerian mimic cautiously, so as to have every opportunity of 

 appreciating its defence — that might often be sufficient. Or again, if 

 the bird found both model and mimic unpleasant it could more easily 

 remember the experience if they were alike, even though it did not 



