STUDY OF VARIATION IN THE RACES OF ZYGAENA FILIPENDULAE, L. 107 



easily see that filipendulae simply follows the fundamental law of 

 variation of the wing-markings of the Lepidoptera, according to which 

 in the successive subdivisions of each group the same variations tend 

 to reproduce themselves on a minor scale, except that character or 

 those characters, which become more or less invariable in each case and 

 distinguish that group. The arctica branch of filipendulae corresponds 

 to the purpilralis group of species, the stoechadis branch to the fansta, 

 carniolica, etc., group. It must also be noted that these two types of 

 markings are so generalised in the Lepidoptera, that, under the same 

 conditions of environment, they are actually produced in as distant a 

 Family as, for instance, the Pieridae in P. napi races arctica and 

 bryoniae, up to the extreme form radiata, and the summer broods of 

 race meridioridlis, Riihl., up to form dubiosa, Rob., whilst in the genus 

 Pieris in general the napi group of species corresponds to the 

 purpuralis group of the Zyyaenae and the rapae group corresponds to 

 the fau sta, carniolica, etc., group. 



Another distinctive character between filipendulae and stoechadis, 

 which can be explained by the fact that the true primary pattern is 

 more developed in the latter than in the former, consists in the frequent 

 obliteration of spot 6th in stoechadis, whereas it never disappears even 

 in the darkest forms of filipendulae. To follow up the cause of this we 

 must begin by finding out the origin of the 6th spot. It will be noticed 

 that in a great many, if not in nearly all the Lepidoptera, there exists, 

 just about the third or hindermost terminal branch of the median 

 nervure, a zone of wing in which the pattern is reduced in extent, 

 evidently because that branch is itself atrophic ; it should consist in 

 branches 3 and 4 in which the hind median nervure should fork, as 

 does the fore median nervure in branches 1 and 2 ; the four branches 

 all exist in early stages of the wing in the chrysalis ; the atrophy of 

 pattern is generally more conspicuous on the underside of the hindwing. 

 In the Grypocera there exists a fold in these wings, which stretches 

 across them in front of that nervure and which no doubt is connected 

 with the phenomenon I am describing in the pattern. It is in this 

 intervural space that, for instance, the eye-spot oi the Satyridae is 

 smaller than in the other spaces, or even absent. It is in the next 

 internervular space, just at- the back of the third median nervure, that 

 the true secondary pattern (the grey or fulvous, so-called, ground- 

 colour) on the underside of the hindwings of the Lycaenidi is often 

 cancelled and leaves a triangular space of the real, white, ground- 

 colour uncovered. This important part of the wing I propose calling 

 the atrophied zone, because it is obvious that the development of 

 the wing-markings is impeded there and kept some grades back as 

 compared to the grades reached on the greater part of the wing- surface. 

 In filipendulae and other Zygaena the sixth spot is placed precisely 

 astride of the third median nervure of the forewing and the dark 

 marginal band of the hindwing is narrower on or just at the back of 

 the same nervure, when its inner outline is waved and it varies in 

 breadth. This seems to show obviously enough that when on the 

 forewing the dark marginal band becomes so broad on most of its 

 length as to join and blend with the preceding band (that which 

 crosses the wing beyond the end of the cell and separates spots 5th 

 and 6th from each other) it is in some cases kept back in the atrophied 

 zone and, remains narrower ; the consequence is it does not 



