12 



[Jarmary, 



a sprinj? brood, A. teutnnica, Alfk. ; (3) A. i^pinigera, K., with a 

 summer brood, A. anglicn, Alfk. 



I have studied the variation in many examples of A. trimmerana, 

 A spinifjera, and A. amjlica, witli rei^^ard to some characters which 

 have been used for the separation of the races by Saunders and Alfken. 



In North Wiltshire where trimmerana only occm-red, in many 

 hundreds examined two examples of the ^ were found with well- 

 developed <>^enal spine, the spines being qiiite visible to the naked eye 

 and as long as in many true !<i)ini(iera. None of the other characters 

 of spinujera were shown by these examples, which otherwise were 

 typical friininerana. One of the most constant distinctive characters 

 of frimmtrana ? is to lie found in the scopa of the hind tibia?, 

 which, beneath, is of almost silvery whiteness. Very rarely indeed is 

 it of an ochreous colour, exactly as in the ? of amjUca. The 8th 

 ventral segment of trimmerana was deeply emarginate in all the 

 examples examined, and the pygidial area of the $ always distinctly 

 densely punctate. 



In spinicjera ^^ 10 ptu- cent, had the 8th ventral segment mut-h 

 less deeply emarginate than normal ; one ? had the sculpture of the 

 pygidial area obsolete, and in two it was sul^obsolete. All tlie ^ (^ 

 had distinct geual spines visible to the naked eye, though varying 

 somewliat in length. 



In anglica, fourteen i^ ^ taken at random were examined. Of 

 these eight had the Stli ventral segment deeply emarginate, two very 

 slightly or not at all, and four were intermediate between these. Nine 

 of these had no genal spine, tive had a distinct one visi])le to the naked 

 eye, but usually of from \-\ the length of that of fully developed s^xVit- 

 gera. None of these showed any trace of the dark facial clothing of 

 the spring brood. Of 19 females, also selected at random, two had the 

 pygidial area as strongly sculptured as in trimnwrana, two rather less 

 strongly, in four the sculpture was faint, while in eleven it was very 

 faint. One female (not included in the above) wliich was sent to 

 Saunders many years ago, entirely without red colouring and with 

 strongly punctured pygidial area, was returned by him as trimmerana . 

 It, however, had the ochreous or yellow hairs on the scopa characteristic 

 of anglica, but as mentioned above, very unusual in trimmerana. 

 This specimen was taken otf flowers of BuhiiK in company with normal 

 anglica. 



In the spring trimmerana, .ijiitiiyera, and tentonica may all occur 

 at the same time and place (as they do on Dartmoor), or either of them 



