28 PRINCIPAL HOUSEHOLD INSECTS. 



molting. They either stuck to the crumbs, which were rather greasy, 

 or to the sides of the glass, which had also become somewhat greasy. 

 On the same date the larvjie in the dry vessel, from which these ten 

 had been removed, commenced to spin up. Many were restlessly run- 

 ning about iu search of suitable places for spinning, and some had even 

 reached the top of the blotting joaper. A thin layer of gray cotton 

 was iDlaced between the two blotting ])apers to give them suitable 

 spinning places. The eggs hatched in two days, having been kept dry 

 all the time. The first skin was cast two days after hatchiDg, and the 

 beginning of spinning occurred eight days after hatching. 



By July 19 no more specimens had spun up and many had died. The 

 receptacle seemed to be too dry and too hot", and the blotting paper was 

 somewhat moistened. But oue pupa was found, which was that of a 

 larva which began to spin July 15. July 21 no others had spun up, 

 although they were still very lively. The pupa had become brownish. 

 July 22 the adult flea issued over night and escaped through the cov- 

 ering. From that time until July 29 no more spun cocoons, and many 

 of them died. On July 30 one of the survivors commenced to spin, 

 twenty-four days after hatching. The cocoon is delicate, white in 

 color, and is very well shown at fig. 5. 



On August 2 this larva, which commenced to spin July 21, changed 

 to pupa. On August G it was still white in color, becoming somewhat 

 yellowish on the 7th and quite brown on the 8th. On the 9th the adult 

 flea was found to have issued overnight. The pupa state, therefore, 

 lasted about eight days, and it is to be noted that the pupa remains 

 white until shortly before the emergence of the adult. It was supposed 

 that the pupa stage in this instance was longer than usual, on account 

 of the fact that the larval stage was so very much longer than in the 

 first instance. 



On July 11 another series of experiments was started, in order to 

 gauge the variation in the duration of the stages and settle the ques- 

 tion of the number of larval molts. Eggs collected on this date 

 hatched July 13. Ou July 16, of fifteeu larvae eleven had cast the first 

 skin. On July 18 five specimens cast the second skin. July 19 the 

 weather was extremely warm and a number of the larvse died. July 20 

 the heat continued, and more died. On July 23 seven larvae which had 

 cast the first skin remained; one of them had begun to spin up. There 

 were on the morning of this date three cast skins in the receptacle, so 

 that there are apparently three molts. In this final state the bristles 

 have become longer and the mandibles have two teeth at the apex. 

 The remaining four were carried on until August 8, when the last one 

 died, none of them having succeeded in casting a third skin. Of the 

 entire lot, but one was reared to the pupa state, and this pupa was 

 preserved in alcohol for drawing. The record of this advanced speci- 

 men shows three molts, and that it began to spin eight days after 

 hatching. The average of the others shows that the eggs hatch in 



