20 WASP STUDIES AFIELD 



was begun on July 14, 1915. When we excavated it on 

 July 26, twelve days later, it contained a larva three or four 

 days old, together with four flies. Since the cell at the 

 bottom must have been completed before the egg was de- 

 posited, this leaves eight or nine days for the construction 

 of the nest and the incubation of the egg. This seems in- 

 deed a long time for so rapid a worker as B. nubilipennis. 

 We have occasionally found her resting or loitering in her 

 incomplete or newly-finished burrow, and we know that 

 she does not work on cloudy or rainy days; so when we 

 have made ample allowance of time for these indulgences, 

 this may after all be a fair estimate of the normal time 

 devoted to the making of a nest. 



The normal burrow, three-eighths inch in diameter at 

 the top, pierces the ground obliquely and continues down- 

 ward, at an angle of about 30 to 45 with the surface, 

 for five or six inches to its horizontal oval terminal pocket, 

 which is one inch long by one-half inch in diameter. This 

 cell is hollowed out more from the roof than from the bot- 

 tom of the burrow. The tunnels are usually approximately 

 straight, and only occasionally swerve to the right or the 

 left; the slope of the tunnel rarely varies. 



Most of the nests are normal, as the first two illustrations 

 in fig. 3, but there are occasional exceptions. One nest 

 went down at an angle of 60 ; the tunnel was about twelve 

 inches long, with the usual pocket one inch by one-half 

 inch; another nest had a normal gallery, but a cell an inch 

 and a quarter long and one inch wide. Another was normal 

 at the entrance, but soon curved straight downward and 

 then back under itself (lower illustration in fig. 3). But 

 the most novel form of nest which we have yet discovered 

 was one which started down and eastward, at an angle of 

 30, for two inches; then turned sharply to the south and 

 downward, at an angle of 45, for four inches; then con- 



