140 PROCEEDINGS OF THE NATIONAL MUSEUM. vol.52. 



July 17; was still unhatched on the morning of the 19th, but was 

 hatched on the morning of the 20th. It was fed daily on freshly 

 killed May-flies, the remains of the supply of food given on one morn- 

 ing being invariably removed before the fresh supply was given on the 

 succeding day. It completed its growth and began its encasement on 

 July 25. On the 21st another of these marked nests was opened and a 

 larva apparently about 24 hours old was secured. It had been pro- 

 vided with food, but whether on this date or on the 20th could not be 

 determined. This larva was placed in the breeding box, and reared 

 under conditions similar to the preceding. It likewise completed its 

 growth on the 25th but failed to encase. On the 23d another of 

 these marked nests was opened and a larva secured therefrom and 

 placed in the breeding box. This one completed its development on 

 the 26th but failed to encase — due to neglect on my part. In the 

 case of the first two the time from the deposition of the egg to com- 

 pletion of the feeding of the larva was eight days and in the third nine. 

 It is quite probable that the time normally required by the wasp is 

 somewhat greater than this since the character and quantity of food 

 supplied to the larvae in the breeding case was much superior to that 

 usually furnished by the mother wasp. 



When the period of feeding is complete the larva encases itself in 

 an elongated egg-shaped cell or cocoon composed of grains of sand 

 held together by threads of silk that are covered with some adhesive 

 substance, probably furnished by glands in the mouth of the insect. 

 My breeding experiments tend to show that the larvae can not encase 

 unless they are buried in the sand. Many larvae brought to maturity 

 in the vials simply spun a flat web of silk in their cells, and perished 

 without forming cocoons whereas others burrowed down into the 

 sand in the vials and readily encased. Others failing to burrow into 

 the sand were covered loosely with sand by filling up the vials. These 

 readily formed their cocoons. 



The first encased larva found in the field was secured on June 28 

 and it is not probable that many mature earlier than this date. Two 

 weeks later encased larvae were very frequently found when burrows 

 were opened. It is not easy to distinguish a newly constructed co- 

 coon from one formed the year before without opening the cocoon. 

 If the cocoons are very carefully removed the newly constructed 

 ones invariably have adhering to them loose silk fibers resembling 

 the web of a spider; these are never present on a cell a year old. 

 The old cells are also somewhat darker, but this is not a reliable cri- 

 terion. In no instance was a cell found in a brood chamber or among 

 the remains of food. It appears that the larva when it ceases feeding 

 either advances from the brood chamber a short distance into the 

 tunnel where its narrower diameter permits the larva to reach the 

 sand on all sides or, more rarely, makes its way into the sand from the 

 side of the brood chamber. In all these instances where the cell was 



