8 PROCEEDINGS OF THE NATIONAL MUSEUM. vol. G^. 



crawls about over the ground to a greater extent than does CoUetes, 

 if the parasites can reach the female of Andrena by their own 

 efforts there is no very good reason for believing that some of them 

 can not reach the female of CoUetes in the same way. 



It may be well to report at this time some other observations made 

 in connection with the study of the life history of this beetle. An- 

 other solitary bee, Andrena perplexa Smith, also nests in gi-eat num- 

 bers on this same sunny slope and the nests of the two species are 

 intermingled promiscously over the nesting site. Andirna emerges 

 earlier in the season than Colletes and the nesting operations of the 

 former are in full swing when the latter first appears on the scene, 

 but for a considerable length of time the nesting activities of the two 

 species go on side by side. If our conclusion is correct, that the para- 

 sites get down into the nests of Colletes by obtaining a hold upon the 

 host and riding down to the brood cell in this position, then there 

 can be no doubt tliat a great many go down into the nests of An- 

 drena; for this bee alights usually at a short distance from her bur- 

 row and crawls over the ground to the entrance, whereas Colletes 

 ti lights directly in the mouth of her burrow, which is always left 

 open, and disappears within immediately. 



This beetle can not parisitize Atidrena^ however, and the explana- 

 tion lies in the nesting habits of this species. Andrena perpUxa 

 Smith constructs as a brood chamber, at the end of a vertical tunnel, 

 a cavity whose walls are smoothed and made waterproof by means 

 of a waxy substance. At the bottom of this chamber the bee places 

 a mass of pollen upon which she deposits an egg (fig. 39). The ^gg 

 is placed on end and never in contact with the wall of the brood 

 chamber. Then upon this pollen-mass, upon which the egg rests, 

 she places a quantity of thin, watery honey that completely sur- 

 rounds and almost submerges the ^gg. Hence, the cell when com- 

 pleted is so arranged that the parasite can not reach the egg without 

 first falling into this watery honey. A large number of nests of 

 Andrena were brought into the laboratory in the course of our inves- 

 tigations and placed in breeding receptacles into which first instars 

 of the beetle were introduced. In every case, without a single ex- 

 ception, the parasites got into the honey and perished. We repeat- 

 edly placed the larva directly upon the Qgg, but in every instance the 

 larval beetle sooner or later got into the honey and perished. A larva 

 may become mired in the honey furnished by CoUetes and later 

 struggle out and survive, but once it becomes mired in the honej^ 

 supplied by Andrena it never gets out and perishes in a short time. 



In the case of Colletes rufithorax Smith, the bee, in constructing a 

 brood chamber, excavates, at the end of a tunnel, a cavity, and within 

 it, closely applied to its walls, she constructs a cell of thin, tough, 



