3S [February, 



of tlie tliivo species with certainty, provided tlie examples examined 

 are clean and in reasonabl}'' good condition. But it must be confessed 

 that in the species named one does occasionall}'' find examples of the 

 male sex that at any rate need the most careful scrutiny before they 

 can be correctly placed. Most of the examples can be determined 

 off-hand, by anyone who has once mastered the slight distinctions 

 between them. The difficulty is certainly caused by the variability 

 of the double-brooded ovatula, this species being more unstable in its 

 characters than its allies. These closely allied species have always 

 been of particular interest to me, and it may be worth while to mention 

 some of the reasons which confirm their specific distinction. 



(<?) Andrena ovatula {afzeliella^ is in the South generally double- 

 brooded ; the other two species are single-brooded, even in 

 abnormal seasons, when they may appear as early in the Spring 

 as ovatula. Usuall}'' the latter appears considerably earlier in. 

 the season, though later all will be found together. 



(i) A. ovatula is more variable and has in both broods a black-legged 

 form of ^ , fuscata K., Sm. The hind tibiae of the other two 

 are always pale, like the afzeliella form of ovatula. 



(c) A. ovatula is rarely attacked by Stylops, even in most localities 

 in Avhich it abounds, whereas A. wilhella is frequently found 

 stylopized, where it is abundant, and examples changed by 

 this parasite are the convexiuscula of Kirby and Smith. In 

 Smith's collection the series of convexiuscula is made up almost 

 wholly of stylopized loilhella, not of ovatula. Of similis Sm. 

 I have never met with a stylopized example, though I have 

 read that such are common in Germany. 



(f7) A. wilhella {xantJiura^ is the natural host of Nomada oclirostoma, 

 Avhich often abounds at pure colonies of this Andrena. I have 

 entirely failed to find it at pure colonies of either ovatula or 

 similis. Where all three species of Andrena occur together, 

 but ivilhella much less numerously than the others, I have 

 found the Nomada, though it was not then abundant, as one 

 would have expected it to be, had it also attacked the other two 

 species, which w^ere extremely numerous, especially ovatula. 



(<?) A. wilhella is extreme]}' jmrtial to gardens, meadowlands, and well- 

 cultivated districts. " It frequents hedgerows and woodsides in 

 preference to open heaths and sandy situations " — words used by 

 Smith, not of the Andrena, but of its parasite, N. ochrosioma. 

 i^ot that it is necessarily absent from such places, but both. 



