THE LAW OF SUFFUSION IN BUTTERFLIES. 1351 



this way we may perliaps prevent tlie description of such forms as new and 

 independent species, as lias been done in no less tlian four instances with 

 our own buttertiieti. 



A little study of any one of these forms by an adept will quickly determine 

 to what species it belongs, if the sj)ecimen be in sufficiently perfect condition 

 to determine the structure of the minuter parts ; for in doubtful cases, 

 nowhere more than here, is such a careful examination of structural details 

 essential as is uniformly insisted upon in the present work. This deter- 

 mined, comparisons with the typical distribution of the markings may be 

 instituted. It will then be seen that in oreneral the term suffusion is well 

 chosen, though it does not cover all the elements involved. The disguise 

 of suffusion is produced by the blending of certain colors, especially of 

 black, white, or silvery tints, which are normally found at distinct parts of 

 the wing ; usually it occurs where there are two parallel or sub-parallel 

 scries of markings, following in general the same course as the outer margin 

 of the wing : there may be, for example, two pai-allel series of white spots 

 normally crossing the wing in the middle and near the border ; under this 

 disguise of sufftision, the whole intervening area, in the interspaces where 

 the spots occur, is covered or shot with this color, generally more or less- 

 sprinkled with atoms of the normal ground, producing then an impure 

 tint, and sometimes obliterating altogether any intervening markings of 

 another character. 



In the excursus on the origin of the diversity of ornamentation in but- 

 terflies (pp. 510-519) I have suggested that such parallel series of 

 markings originated in a single marginal shade, and if the grounds for such 

 an opinion are sound, we have, possibly, in these cases of suffusion, 

 instances of reversion. This would sufficiently explain an otherwise 

 puzzling fact that when these sports occur they take definite directions and 

 repeat themselves over and over again. Both the strigate and the obso- 

 lescent types of suffusion may thus be brought under a common law, and 

 where the sufftision is incomplete or imperfect, the different forms may 

 readily be seen to be only relatively extended steps in a single direction. 

 So far as I have seen, the suffusion shows no preference for the upper or 

 under surface of the wings, or for one pair of wings. 



It is, perhaps, due to the fact of the generally more complicated diver_ 

 eity of patterns in the highest family that we find this phenomenon relatively 

 commoner among Nymphalidae than elsewhere. I have sought every 

 possible example for description in the present work and run through many 

 cabinets in the search for them, and the collection of instances here brought 

 together is, I believe, the largest ever made anywhere, so that the following 

 figures may not be without significance. We find examples of suffusion in 

 thirteen of the forty Nymphalidae described in the body of this work, or, 

 approximately, in one-third of the species ; the numbers are both actually 



