IN THE BRITISH ISLANDS. 119 



genus. The great mass of specimens are coloured with whitish-grey, 

 yellow-ochreous or red, extending in some specimens (principally 

 Scotch ones) to bright reddish-brown, of the same shade as in N. 

 rubi var. quadratum of Hiibner, to which such specimens bear more 

 than a superficial resemblance. Two other (almost purely Scotch) 

 forms occur, one, of a deep grey, the other, of a dark purplish-red or 

 plum-colour, the purplish tint being produced as in certain forms of 

 N. sobrina, N, baia, Agrotis hyperborea and many other species. There 

 is another Scotch form, dull reddish-brown in colour, common in the 

 Aberdeenshire districts, which is much darker than any of our more 

 southern forms. In general appearance, too, there is great difference, 

 some specimens are very mottled, others have a distinct dark quadrate 

 spot between the stigmata and another beyond the orbicular, whilst 

 sometimes the basal area (to the central shade) is very pale (grey, 

 ochreous &c.), the outer area being much darker. When the extreme 

 outer margin, beyond the subterminal is also pale, the insect has a 

 banded form, and sometimes this band is most striking in its develop- 

 ment. The stigmata vary but little ; they are generally pale in colour 

 and well-developed. Only in one specimen of a long series are the 

 two quadrate spots joined by a line under the orbicular, although an 

 occasional specimen shows a tendency that way. There is also con- 

 siderable difference in the development of the transverse lines, but the 

 only one of these that occasionally presents any striking character, is 

 the median shade, which often stands out conspicuously dark on a pale 

 ground colour. In size there is great variation, and our exposed 

 localities, in the north of England and Scotland, produce the small 

 specimens which Newman erroneously called and figured as conflua in 

 his ' British Moths,' p. 349. Hiibner's type is a very rare form, and 

 I am indebted to Mr. Wylie of Perth, for perhaps the best specimen J 

 have ever seen of it. It has the basal area to the central shade of a 

 clear bluish or slaty colour, the outer area being of a bright red. Of 

 this type Guenee writes : " If we only referred to the phrase in the 

 * Wien.-Verz.,' ' Dunkelrothe und perlfarbige ' and to the position 

 of the species among those NOCTU-SB ' pupurfarbig,' with delphinii and 

 purpurina, we should be left in great doubt, but the figure of Hiibner, 

 which was perhaps even made from the Therdsien collection itself, 

 which is in fact half purple and half pearly grey, and which repre- 

 sents well, however, our f estiva, will serve to explain the difficulty " 

 (' Noctuelles,' vol. v., p. 331). This is the most difficult member of 

 the genus to deal with so far as its varieties are concerned, and it is 

 only possible to give the merest rough classification of the forms we 

 get. In coloration (as has been pointed out), the specimens vary from 

 whitish-ochreous to brownish-red and dark grey, and in general 

 appearance, from almost obsolete, to a pale basal half and dark outer 

 half, and from strongly marked quadrate spots between the stigmata 

 and beyond the orbicular to total absence. Some Scotch specimens are 

 pure lilac-grey, as in Agrotis subrosea var. ccerulea. The following is 

 an attempt to classify the various forms : 



A. Dark purplish or reddish-brown. 



1. Without quadrate spots, but having pale base = f estiva,' Tib. 



2. Without quadrate spots (mottled form) = var. congener, Hb. 



(turbida by error). 

 3. With quadrate spots = var. sulrufa, Haw. 



