INTRODUCTION. IX 



peculiarity becomes common again in Tceniocampa. The development 

 of dots in certain of the Leucania although not normally present in 

 such species is noticeable. The claviform stigma is indefinitely modi- 

 fied in certain genera, being entirely absent and developed into a 

 longitudinal black streak in different specimens even of the same 

 species. This is particularly noticeable in certain species of the genera 

 Apamea and Miana as well as parts of Agrotis and Hadena. When the 

 claviform is thus developed into a black streak it unites the two trans- 

 verse lines, immediately before and beyond the stigmata, and presents 

 a remarkable banded form in its extreme character. Perhaps Apamea 

 gemma var. remissa is one of the best examples we have, although some 

 of the banded forms of Apamea didyma and Miana strigilis are almost as 

 remarkable. The development of this claviform stigma also, by joining 

 the transverse lines shows up another common line of variation in the 

 same genera, viz. the difference in distance between these two lines as 

 they reach the inner margin. The width of these varies indefinitely in 

 Apamea gemina, A. connexa, A. didyma, Miana strigilis, M. fasciuncula, 

 M. literosa, and many species in other genera, such as Agrotis and 

 Noctua. The variation in the space between these lines has been pre- 

 viously referred to as existing in Asphalia flavicornis and other Cymato- 

 phoridce. Some species, such as Charceas graminis, Heliophobus hispidus and 

 Helotropha fibrosa, offer very great variation in the amount of sup- 

 pression or development of certain special markings, as also do certain 

 species of Tceniocampa and Dianthcecia ; but in this direction certain 

 species of Agrotis, viz: tritlci and cursoria, and to a lesser extent ripce 

 and nigricans offer the most striking development. These species have 

 each two well-developed forms, one of which has the whole of the 

 wing of a certain ground colour, with the stigmata more or less well- 

 developed and with a varying number, generally three or four, trans- 

 verse lines crossing from the costa to the inner margin, but practically 

 without the slightest trace of longitudinal markings. The second form 

 has the same ground colour as the first, and the stigmata equally well- 

 developed, but the transverse lines are reduced to a minimum, and a 

 pale (variable in colour, from white to ochreous) streak, varying in 

 width and intensity, runs along the costa from the base to the apex, 

 while another pale streak runs along the central nervure and the bases 

 of its branches. In my previous writings on these forms, I have always 

 called the group of varieties which have this pale costa the pale-costa 

 group, whilst the former which do not possess a pale costa, I have 

 called the non-pale-costa group ; but it is really remarkable, that in 

 every gradation of colour from pale whitish-grey, ochreous, reddish 

 and black to which these species attain, every colour variety has these 

 two parallel forms distinctly developed. The relation of these to allied 

 species is suggestive, the normally pale-costa part of the genus including 

 Agrotis valligera, A. obelisca and A. ravida, whilst the normally non- 

 pale-costa part of the genus includes Agrotis segetum, A. lunigera, A. 

 exclamationis, A. corticea, and A. cinerea, although we frequently find 

 traces of a pale costa in exclamationis, and there is a non-pale-costa 

 form of obelisca (var. ruris) on the Continent of Europe. The presence 

 or absence of the row of cuneiform spots which appear to be developed 

 by the filling in of the spaces between the nervures in direct con- 

 nection with the subterminal line, appears to give a great deal of 



