214 [September, 



range of variation, but the material that I have is insufficient for any 

 more definite decision. In localities that have been accessible to me 

 of late years, neither A. helvola nor A. varians occur, so that I have 

 been unable to collect rujicnrnis from these hosts. 



Smith in the first edition of his book says that he has found 

 rvficornis, (i.e., rvficornis + jlava) " entering burrows of trimmerana, 

 nitida, varia" (sic! no doubt = varians), "and fidva, its attacks are 

 very general." In his 2nd edition he says, " parasitic on several 

 species of Andrena, frequently so on trimmerana, atriceps, fidva, and 

 nigroaeneay Without very strong evidence in its favour I should 

 consider atriceps an improbable host. The fact that A. fidva has the 

 peculiar A. signata attached to it, and likewise a specially large form 

 of A. ruficornis proper, forms extremely dissimilar, though the host is 

 the same, seems to me a decided reason for considering such forms 

 more than mere varieties and ranking at least with the so-called races 

 of certain Bombi, such as jonelhis, pratorum., and lapponicus. It is 

 certain that in the ruficornis group the specific characters are much 

 more evident in the ? ? than in the f^ (^ , as if the former sex had 

 been the first to acquire distinctive characters, which is what might 

 have been expected in accordance with their habits. Certainly in 

 most cases the different forms that I have tried to distinguish can in 

 the female sex, in the great majority of individuals, be at once 

 separated at a glance with the naked eye, but the males are much 

 more difficult and generally require minute examination, or in rare 

 cases are only dubiously separable, imperfectly studied as they have 

 been. 



On the evidence of observations made by myself in 1890, I long 

 believed that one dark form of ruficornis was parasitic on A. albicans, 

 and even now, improbable as this host seems, I have difficulty in 

 persuading myself that my observations were incorrect. At the time 

 mentioned I had previously captured only fiava, the so-called light- 

 coloured form of ruficornis. In the early spring of the year 1890, 

 however, on the edge of a small copse, where burrows of albicans were 

 scattered amongst the grass, I found a Nomada entering the burrows 

 of this Andrena, as I supposed. Knowing well that N. bifida was the 

 common parasite of albicans, and seeing no resemblance between my 

 new captures and my previous ones of fiava, which Saunders had 

 named ruficornis, I jumped to the conclusion that the foi-mer must be 

 an extraordinary variety of bifida with the mandibles pointed ! In 

 fact I actually recorded these specimens as a variety of bifida in the 



