VARIATION IN THE GENUS EREBIA. 91 



The facts to be deduced from this table are as follows : The whole 

 range of variability in ocellation in the two forms taken together 

 is from 16-0. Taken separately the two forms overlap one 

 another at opposite extremities, the M of ligea being 9j, that of 

 euryale 6^. Now, if the two are to be miited into one species, the 

 following conditions must be fulfilled — (1) the mean combined 

 from the means of the two forms must equal the mean of the 

 whole range of variability ; (2) the combined mean should fall at 

 a point where the two forms confessedly pass over into one 

 another. Now, it will be seen that both these conditions are 

 satisfied by the table, for (1) i (9|- + 6^-) = 8, and (2) 8 falls at 

 the point round which the two forms adyte and euryale type are 

 grouped, and it is admitted that these are transitional forms. 



I cannot give here the evidence derived from clasp-forms, 

 though, to my mind, it is even more convincing than any that 

 can be derived from wing-facies, because it is so unfailing in its 

 application to other species of the genus. But anyone who has 

 worked through a series of them, as I have done, will, I think, 

 agree with the following remark of Dr. Chapman: " The clasp 

 differences are not great enough to render this (identity of the 

 two species) otherwise than likely where they occur together on 

 the same ground." That the various forms do occur together 

 more commonly than is supposed, I can only suggest by referring 

 back to my own remarks and the following of Mr. Elwes : " The 

 larger the number of specimens which are brought together from 

 many localities, the more difficult it becomes to name those 

 varieties ; and I have therefore dropped the names of a few which 

 I had previously adopted, though I have not done so in cases 

 where, as with many Asiatic forms, my knowledge is as yet in- 

 sufficient to justify this course." 



From the foregoing arguments it can easily be gathered that 

 if we treat all forms of ligea and euryale as one species, we have 

 here the kind of evidence that we set out to find. Collectors 

 from different parts of the world bring back such different speci- 

 mens of this species that separate varietal and even specific 

 names are given to them ; as collections from these areas become 

 more complete, intermediate links between these varieties and 

 species are supplied, which show that all are only one species. 

 The explanation I would give of these facts is that a casual 

 collector in some area meets only with the M of that area (M 

 being the mediocre or commonest condition) ; it is only after long 

 collecting in the same area that the other terms of the scheme 

 can be supplied. Now we saw that, according to our hypothesis, 

 the M of different areas would probably be different, and this 

 supposition is confirmed by the number of named varieties that 

 exist whichjwere supposed at first to be peculiar to some par- 

 ticular locality. But as collections become more perfect, it is 

 found that what occurs as an habitual variation in one area 



