ANDRENA. 205 



this subfamily its collective designation, making the 

 other genera thus converge to it as to a centre. He 

 took its elliptical form as typical. Indeed, it is remark- 

 able how very judiciously this was done, for it is a form 

 not apparent among the normal bees excepting in two 

 exceptional cases, the one upon the frontiers of this 

 subfamily, in almost debatable land, where the last of 

 the Andrenida and the first of the Apidce seem almost 

 to melt into one another ; and in the other case, in the 

 parasitical Nomada, whose parasitism is in every in- 

 stance, but one only, restricted to the first subfamily. 

 A different type of form prevails amongst the Apida, 

 upon which I shall have subsequently occasion to speak. 

 These insects are not distinguished for any elaborate 

 economy. Varying in the species, some prefer vertical 

 banks, others sloping undulations, and again others ho- 

 rizontal flat ground or hard down-trodden pathways. 

 Some burrow singly, and others are gregarious, col- 

 lected in great numbers upon one spot. They are, 

 perhaps, the most inartificial burrowers of all the bees. 

 Their tunnels vary from five to nine or ten inches in 

 depth, and in some species they are formed with other 

 small tunnels slanting off from the main cylinder. The 

 sides and bottom are merely smoothed, without either 

 drapery or polish. The little cells thus formed are then 

 supplied with the usual mixture of pollen and honey 

 kneaded together, which in the larger species forms a 

 mass of about the size of a moderate red currant, its 

 instinct teaching it the quantity necessary for the nur- 

 ture of the young which shall proceed from the egg 

 that it then deposits upon this collected mass of food. 

 The aperture of each little tunnel is closed with par- 

 ticles of the earth or sand wherein the insect burrows, 



