128 THK KNTOMOLOOISt's KKDOKD. 



the nervulav and of the transverse patterns produces two principal 

 series of forms in those grades of the variation of the species in which 

 the nervuhir markings are highly developed. In one series the trans- 

 verse pattern becomes less and less conspicuous as the nervular one 

 gets more extensive, in the other series the two patterns exist together 

 and their extent varies about equally in the different individual forms. 

 I notice that in the true hii/oiiiiw, O., of the Alps one only meets, as a 

 rule, with the latter case, so that the better characterised individuals of 

 the male sex, with broad nervular streaks on underside, always have 

 an extensive black crescent and marginal streaks above, and the 

 corresponding females, with diffused bands on nervures above, always 

 have a prominent apical patch and large discal spots. Instead, in the 

 arctic region, forms of the first series mentioned are more abundant 

 than the latter : males with prominent streaks underneath may be 

 entirely white above and females may be nearly entirely darkened by 

 the diffused nervular bands and yet scarcely show any signs of the 

 apical patch or of spots (see tig. 30 mentioned below). These dark 

 arctic forms I have named I'Ufitilohri/oniat' in Ulio/>. I\il., p. 146 

 (January, 1908), taking as " type " my figs. 86 and 37 of PI. XXXII. ; 

 as this name is not mentioned in the explanation of the plate, 

 Fnihstorfer overlooked it in the text and renamed my figure 37 

 adalwimla in the Kiit. Zeitr. Guben, III,, p. 88 (1909) ; it is quite out 

 of tbe question to base the differences between the races of Alaska and 

 of Finmark on these two figures as suggested by Fruhstorfer, Ruber's 

 radiata is the very rare culminating form in the arctic variation 

 direction, described from a specimen found near Vienna, with no trace 

 of true pattern left. Another important point to be noticed, if one 

 wishes to classify the female forms of napi with prominent nervular 

 pattern, is that, when it is reduced in extent, it follows two different 

 lines. It either does so uniformly on tbe whole of the neuration, so 

 that the streaks end up by being very thin, but stand out sharply in 

 an even network, or else the streaks dwindle iiwiiy in the basal part of 

 the wing whilst on the outer part of it tbey remain quite broad. The 

 first of these lines of variation is obviously that followed in nearly the 

 totality of cases by uajn' of both sexes on the underside of hindwings 

 and here it is usually the outer part of the wing in which the nervular 

 streaks get thinnest earlier ; on the upperside of the female it pre- 

 dominates in the arctic races, both pabvarctic and nearctic (see tig. 35 

 on my PI. XXXII. mentioned above and LXVIL, tig. 16 and 17) ; it 

 is also the commonest line in the wonderfully variable race of Rliidling, 

 near Vienna, and virtually in most races of the plain, though in these 

 it is not striking on account of the inconspicuous streaks. The second 

 of the two lines of variation just mentioned is the only one, to my 

 knowledge, followed by the true bn/ouidt' race of the Alps on upperside 

 of female ; the basal part of the wings may be nearly devoid of nervular 

 streaks, whilst the outer part exhibits broad triangles tapering inwardly 

 to sharp points (form emibryoniae mihi). This line of variation is 

 carried on by the summer forms of tbe species in general and it is 

 particularly obvious in races of damp warm localities, where the females 

 have a comparatively very extensive dark pattern on upperside. whilst 

 their underside follows the other line of variation and maj' not exhibit 

 any streaks at all or very reduced ones near the base, just like some 

 females of bri/o7iiae. If one notes the two sorts of variation described 



