6 



wholly unexplored, and that even in the case of those portions of 

 whose fauna we are least in ignorance, we frequently owe such 

 knowledge as we possess to the passing visits of entomologists who 

 have only been able to give a very short time to any one locality, 

 and that when, or if, these regions are ever explored, the result may 

 be on the one hand to solve some at least of the puzzles with which 

 we are at present confronted, or on the other to upset any conclu- 

 sions at which, " in the present state of our ignorance," we may 

 arrive. 



If the supposition be correct (and it is probably the least un- 

 certain of our conclusions) that we must look in the far north for 

 the earliest home of the genus, it follows almost of necessity that 

 the aiiriuia-gvoui:) is the oldest of the three represented in the palfe- 

 arctic region, since the only species distributed throughout that 

 range belongs to this group, ciz., iduna, which is found from the 

 extreme north of Lapland to the north-east of iSiberia ; the only 

 district where this species has been found further south is the Altai 

 range, together with the Tavgabatai range just below, so that it 

 would seem at present probable that the original home of the group 

 was somewhere to the north of this, say roughly between 80" and 

 100° E. long., and that it spread in early times both east and west, 

 extending largely into North America, almost all the American 

 Melitfeids belonging to this group. The affinities of the North 

 American species are however in most cases more directly with 

 viaturna, so that either the migration must have taken place after 

 the appearance of some near ancestor of this species, or the evolu- 

 tion of the genus must have proceeded on closely parallel lines. 

 The former hypothesis however does not refer directly to matiirna 

 as we know it in most European localities, but rather to the form 

 known as var. molejisis, whose distribution covers the mountain 

 ranges of Asia from the Urals to Amurland, and which with its very 

 pale, sometimes almost white, bands on the upperside is much 

 nearer to idnna. The next oldest form is no doubt irolfciisbenieri, 

 the variety which we know in some of the higher parts of the Alps, 

 but which is also found in the Altai range. This is much duller 

 than typical iiiattona. but has the thin scaling, especially on the 

 underside, which seems particularly associated with high arctic, and 

 to some extent high alpine, forms, whilst typical iiiaturna is remark- 

 able for the density of its scaling, especially in its southern localities, 

 such as South Germany, where the colouring of the darker bands 

 of the underside hindwing often looks as if it had been plastered on 

 with a paint brush. Thus, while from the point of view of nomen- 

 clature Vv'e have to speak of the other two forms as varieties of 

 viationa, phylogenetically the latter is a much more recently evolved 

 lowland form of n-olfenyibcnicri, which in its turn seems to have 

 arisen in Central Asia as a mountain form of the older and more 

 northern uraleusix. Those species in which a white colouring is 

 developed would appear to be rather mountain than arctic forms, 



