﻿144 



near the sea, and they emerged in June ; whilst those in the 

 fourth row were bred from larvae taken on sallows, which 

 grew on the upper border of said wood and in the adjacent 

 hedgerows ; these attained the perfect state in July. At the 

 time the bilberry-fed specimens were leaving the pupa in my 

 breeding cages, the insect was to be taken in numbers as they 

 flew, just before dusk, over the bilberry. When the sallow- 

 fed specimens first came out in my cages (July 13th), the 

 bilberry wild examples were getting worn, but the sallows 

 continued to yield fine sordidata up to the end of July. 



" It will be observed that specimens bred from bilberry are 

 much smaller than the examples bred from sallow. I regret 

 to say that I did not carefully compare the respective 

 larvae. 



" As regards aberration of the species, it will be seen that I 

 have divided the North Devon series into twenty-four detach- 

 ments of vars. Placed as I have put them, each form appears 

 to have aberrant character sufficiently well defined to entitle 

 it to a varietal name, but arranged in the way I have a series in 

 my cabinet, there is no clear line of demarcation between the 

 various forms. Two forms have been named in the past, i.e., 

 fusco-undata, Don., and infuscata, Stand. The fore-wings of 

 the former have a reddish ground colour, and are traversed 

 by a black fascia. The last-named variety is more or less 

 suffused with fuscous. Hiibner figures four specimens which 

 appear to be modifications of the type — that is sofdidata, Fab,; 

 these are numbered 224, 382, 384, and 385. His figures 381 

 and 383 represent specimens of the fiisco-imdata forms, the 

 latter tending towards infiiscata, Staud. 



" It is curious to note that the mfiiscata form does not appear 

 to occur among the North Devon bilberry specimens, neither 

 is there, as far as I could find, anything quite similar to the 

 fiisco-undata form among the sallow examples in that place. 

 I examined large numbers of each, but only retained those 

 specimens which served to illustrate the range of variation of 

 the species in that particular district. Among the moorland 

 examples of sordidata from the neighbourhood of Barnsley 

 are examples of the infuscata form, but compared with speci- 

 mens from South England all the examples I have seen from 

 that district are much darker in tint. The specimens from 



