LASIOCAMPA QUERCUS. 53 



localities, is strikingly characteristic, narrowing much the chestnut 

 outer marginal area of the forewings and often reducing it on the hind- 

 wings to a rather narrow marginal chestnut band (=ab. roboris). In 

 the female the tendency is for the outer marginal area to assume the 

 colour of the band entirely in both fore- and hindwings (a characteristic 

 seen most strongly in the $ in ab. marginata, and also in var. sicula but 

 only in the hindwings). Bacot says: "I speak with some hesitation as 

 to the imagines, as my series are not very long. The differences are 

 most apparent in the males. L. var. sicula differs more markedly from 

 any of the other forms that I know than they do among themselves, 

 and, so far as my experience goes, is very constant. The ground colour 

 of the males is russet-brown, the band on the hindwings is orange-yellow, 

 and extends to the margin, while that on the forewings is narrow, the 

 former feature being characteristic of the females as well. L. var. 

 viburni and L. var. meridionalis I am unable to distinguish. The 

 males have a ground colour of a soft but rich deep red - brown, 

 and the yellow band on the forewings is much straighter than is 

 usual with the British races. Some variation exists in regard to 

 the width of the band, but I have never seen a trace of the 

 epaulette markings that are tolerably frequent in L. var. callunae 

 and not infrequent in English L. quercus, nor have I ever seen a 

 forewing band of the Cannes races so waved as they usually are in 

 the English specimens, nor a hindwing band extended to the 

 margin of the wing as in L. var. sicula. Males of L. var. callunae 

 are often, if not usually, of a deeper ground colour than English L. 

 que reus or L. var. meridionalis. The forewing band is waved and the 

 hindwing band is sharp, not fading into the ground colour towards 

 the margin as is frequently, but not invariably, the case with L. 

 var. viburni and L. var. meridionalis. English L. querctls are the 

 most variable in the imaginal as in the larval state. It seems 

 useless to lay down even the loosest rule. My own short series of 

 nine males shows a wide variation. One taken during July at 

 Lyme Regis would easily pass unnoticed in a series of L. var. 

 callunae. Several have the epaulette marking, while others have no 

 trace, most of the specimens have the forewing band waved, but 

 two have it much straighter than the others, and with one of these 

 it is quite as straight as in the Cannes varieties. Two, in the 

 width of the hindwing band and its tendency to suffuse the ground 

 colour as tar as the margins, approach L. var. sicula as does 

 another specimen in regard to the ground colour, which is of a 

 russet-brown. I am rather inclined to consider that our English 

 race is the nearest to the ancestral stock, and that the variation is 

 largely due to climatic causes, although it is possible that some of 

 the variation towards the southern races may be caused by inroads 

 of fresh blood on our southern and eastern coasts." Among the 

 more or less unclassified notes that we have collected relating to 

 aberrations of this species are the following : A male emerged at 

 Chester with the right lower wing absent (Arkle) ; Selys, Foucart, 

 and Bristowe record females nearly of the colour of the male, and 

 Sumner bred one nearly as dark as a $ in July, 1894, whilst similar 

 ones (referable to var. callunae) are in the " Doubleday " collection ; 

 Turner notes a female from Carlisle with the base suffused with male 

 coloration, and the submarginal area subdiaphanous ; males from 



