166 BRITISH BUTTERFLIES. 



stand below the costal margin in cellules 7 and 6, one (a double spot) 

 between nervules 4 and 6 opposite the middle cell, and also one in 

 cellules 3, 2, and 1, which last usually has an appendage turned 

 towards the inner angle. These spots form two rows, which meet at an 

 angle of from 65° to 90°, in the vertex of which stands the spot 

 between nervules 4 and 6 ; the three upper spots always, and the 

 three lower ones usually, forming a straight row with the spot standing 

 in the vertex of the angle ; sometimes these lower spots stand some- 

 what out of line and farther from the spot in the vertex ; the size of 

 the spots varies in individual specimens very considerably, sometimes 

 they become so small that they stand widely 'separated from each 

 other, sometimes so large that they entirely meet ; rarely one of the 

 spots is wanting (that in cellule 1 or in cellule 7). In the American 

 specimens, on the other hand, the greatest variation in the form, size, 

 number, and arrangement of these spots is presented (even in such as 

 Scudder includes in the same species Colorado), and not one of these 

 shows the form and arrangement of the spots as described in typical 

 comma. Even the two sexes in these forms seem to differ much more 

 strongly than in comma, which shows scarcely any recognisable 

 difference between $ and 2 , except that in the latter the spots are 

 usually larger than in the $ .* A second noteworthy difference 

 between European comma and its American congeners is that, in the 

 former, the fringes on the underside are always spotted with fuscous, 

 at least (in secondaries) on their lower half, while in the American 

 forms the fringes are, as a rule, unspotted. Yet this distinction is not 

 invariable, for two of the specimens submitted (a $■ Colorado and a $ 

 manitoba) have spotted fringes. There exists then, so far as I can 

 discover, only the difference drawn from the underside of secondaries, 

 which, if it were constant, would suffice to separate the American 

 forms from comma, but that it is constant seems to be somewhat 

 improbable, on account of the very great variability which is shown 

 in the shape, number, and arrangement of the square spots, and juba 

 gives a direct proof that we cannot rely upon this feature. A second 

 proof is furnished by Scudder's figures of manitoba, one of which 

 (fig. 10) does not differ in any respect from many forms of European 

 comma in the character of the rows of spots, while the remaining 

 figures (and still more decidedly the two specimens I have) deviate 

 therefrom. We also conclude from Scudder's descriptions that, in 

 this point, manitoba can scarcely, if at all, be separated from comma 

 by any constant difference." Two specimens of sylvanoides, Scudder, 

 submitted to Speyer, were determined by him as not being sylvan- 

 oides, Bdv.,f and he considered the <j> simply an unimportant 

 aberration of comma, and the male with no differences sufficiently 

 important to make him consider sylvanoides, Scudd., as anything more 

 than a local form of comma. The result of these comparisons were 

 summarised by Speyer as follows: — "That among the examples of 

 Scudder's species submitted to me, not one is found which agrees 



* This shows distinctly that Speyer must have been very poorly supplied with 

 European material. The difference noted in the fringes, too, in the succeeding 

 paragraph is very untrustworthy. 



f Edwards says (op. cit. p. 148) that sylvanoides, Scudd., is the same as 

 Columbia, Scudd., the latter for a time thinking that his Columbia was sylvan- 

 oides, Bdv., which latter species he called sonora, a quite distinct species. 



