32 JouRN/\L New Yokk Entomological Society, [VoL ix, 



which the female abdomen is rather more than 5 mm. The egg 

 slightly exceeds 3 mm. in length. 



The interval from the beginning of the burrow to the time when 

 the first egg is laid seems, therefore, to be from 18 to 20 days, of 

 which 5 were rainy. Three of the 5 bad days came while the insects 

 were storing food and probably delayed the completion of the work. 

 When the cell is filled and an egg is laid the upper end is closed by a 

 flat disc of the same parchment-like material as that constituting the 

 rest of the cell and set in a little inside the ragged upper edge. When 

 completed one of these cells is not unlike in appearance to a short 32 

 cartridge, and Mr. Brakeley and myself fell into the habit of referring 

 to them as "cartridges" loaded or empty, as they were or were not 

 filled with food. The material of which the pouch is composed is 

 probably altogether salivary. There is no fibrous structure apparent 

 and no mixture of extraneous materials. In casting these burrows the 

 plaster ran to the disc closing the cell, adhered to it and to the slightly 

 projecting rim, so vve obtained the perfect cell whenever there was one 

 in place. 



The early days of April were unpleasant, rain and snow closing 

 out insect work and leveling the surface so that the location of bur- 

 rows was not readily visible. However, on the 7th, two casts made 

 showed one loaded and one empty cartridge. 



Field work was discontinued until April 2 2d and now for the first 

 time we ran against a burrow that seemed to have been filled up. It 

 was assumed that this was due to accident, and no further examination 

 was made. It became probable, in the light of later developments, 

 that this was really a completed burrow in which the bee had made 

 all the brood cells that were considered desirable and which had then 

 filled up normally to the top. 



April 23d, a number of holes were filled with plaster in a locality 

 marked in March, and these required an unexpectedly small amount 

 of plaster. When the casts were dug out they were unusually short, 

 and some of them had a heel of variable length below the lateral, as 

 if the bee had decided to continue its perpendicular. This was what 

 we expected would happen from published accounts, hence it attracted 

 no special attention. Continuing to dig, however, a second, loaded 

 cartridge, was found below, though not in line with the first, and the 

 burrow leading to it had become filled with sand, which, from its 

 color, had evidently been derived from an upper level. This led to 



