THEODORE H. FRISOX 143 



interior measurements of the cells average sixteen millimeters 

 long by nine millimeters wide. Cells freed from all soil not 

 actually entering into the formation of the same average twenty- 

 five millimeters long and have a diameter of ten millimeters at 

 the anterior end and twelve millimeters at the posterior ex- 

 tremity. Several times I have opened the over-wintering 

 cells, removed the larvae within and then filled the cells full of 

 water. On such occasions it was evident that the substance 

 which held the cells together and gave them a glazed appear- 

 ance on the inside, also served to make the cells impervious to 

 water. Such a protection against soil moisture would seem to 

 be a necessity in the case of bees like these which do not spin a 

 silken cocoon within their surrounding cell. 



I am sure that the adult bee when making her brood cells 

 waterproofs them with some secretion. If she did not do so 

 the semi-fluid mass of pollen and nectar on which she lays her 

 eggs would be partially aljsorbed by the cells. Nininger speak- 

 ing of AnthopJiora stanfordiana says that this bee in making 

 her cells used "a thin layer of waterproofing which seemed to 

 be a salivary secretion." Whether the larvae of any or all 

 instars further improve the cell resistance by secretions or 

 exudates I do not know. The larvae do not spin cocoons as 

 do those of MeUssodes himacidata and many other bees not far 

 removed phylogenetically or morphologically. If the larvae of 

 Anthophora ahrupta do improve their cells by a salivary secre- 

 tion, it is possible that the basic substance of the cementing 

 fluid used by this species, may be of a chemical composition 

 quite similar to that of the silk spun by the larvae of MeUssodes 

 himaculata. Another bee, Emphor fuscojubatus, differs bio- 

 logically from Anthophora ahrupta because of its cocoon spin- 

 ning propensities, thus pointing to its closer biological relation- 

 ship with MeUssodes himaculata. 



E. Pupal Stage 



The pupal stage of development is entered into under natural 

 conditions about the middle of May, or at least such was the 

 case in 1919 and 1920. This is evidenced by the fact that on 

 May 8, 1920, when I visited the bee-})ank at Oakwood the cells 

 contained larvae. A similar visit and examination on May 31, 



TRANS. AM. ENT. SOC, XLVIII. 



