( civ ) 



this point of view, because at Ibadan both were constant in 

 pattern as compared with the other species. But M. far- 

 quharsoni was extremely variable, and at the same time the 

 commonest of all the species and with the largest number 

 of pairs (see Plate B, where pairs taken in cop. of these 

 three species are represented). Out of the 127 specimens 

 captured, 125 are classified in the table on next page, the 

 patterns being very gradually transitional from light forms 

 in which the black is reduced, to dark ones in which it is 

 increased. The two omitted specimens were both males 

 which combined the patterns of grades 1 and 6. Although 

 the absence or presence of a black humeral stripe forms 

 logically a sharp demarcation between the two sets of grades 

 A and B, it is not a conspicuous feature, and does not 

 constitute an obvious break in the series, and the same is 

 true of the isthmus between the pairs of bars. Asymmetry 

 is common, so that some of the specimens in grades 7-9 

 would be placed in one grade by the pattern of one side, 

 in another by the pattern of the other. Furthermore, the 

 isthmus described in these grades is present, although usually 

 very reduced, in a few individuals of grade 5. The table 

 shows at a glance that there is no sex-linking in the patterns, 

 although the females are especially numerous in grade 5, 

 which nearly represents the average pattern. 



Allowing for the extremely small difference between the 

 grades, a close resemblance is manifest in pairs B, D, G, H, I, 

 J, L, M, N, and ; viz. in two-thirds of those captured. A, C, 

 and K are distinctly different, E and F less so. A, although 

 separated by only 3 grades like F, is a far more discordant 

 union, because of the entire absence of the anterior black bar 

 in the ?. On the whole, the results afford some indication 

 that sexual preference may be influenced by pattern or by 

 some other character correlated with pattern. All 15 pairs 

 are figured on Plate B, but it is difficult to show such minute 

 differences in the patterns of beetles satisfactorily. It is 

 hoped that the form of table here proposed will afford a 

 convenient means of testing this most important hypothesis, 

 which was proposed long ago by H. W. Bates in his historic 

 paper on mimicry. 



E. B. P, 



