312 NEW YORK STATE MUSEUM 



If the place from wliich a cocoon has been removed is ex- 

 amined, there are found two slits near one end [pl.30, fig-.6css] 

 and a circular excavation near the center which penetrates 

 through the cocoon into the tissues of the plant [pl.30, figMh]. 

 These slits w^ere undoubtedly made by the caudal setae, as the 

 openings are of the same size as the setae and the same dis- 

 tance apart and I believe it is the place where the larva was 

 attached for the purpose of respiration while it was spinning 

 its cocoon. Though numerous underground stems containing 

 larvae were examined, yet 1 have never found a larva in the act 

 of spinning its cocoon nor opened a cocoon with a larva with 

 its caudal setae inserted in the slits. In every cocoon examined 

 containing larvae except one. the caudal setae were in the oppo- 

 site end of the cocoon from the slits, and in those cocoons con- 

 taining pupae, the cast larval skins were in the opposite end 

 of the cocoon from the slits. The onlv other wav in which the 

 slits could have i)Ossibly been made was by the mandibles, but, 

 when they are studied and their weak structure and their short- 

 ness are noted, the impossibility of this explanation is aiipar- 

 ent, and one is forced to the conclusion that there is no way 

 that the slits could have been made other than by the caudal 

 spines, and that the larvae must turn around in their cocoons 

 after its completion and before shedding their last larval skin. 

 In the case of the single exception noted the larva had prob- 

 ably completed its cocoon and had not yet turned around. The 

 cocoons are usually larger than the inclosed larva, so that there 

 would be room for it to turn around if it so desired. In the 

 cocoons ofDonacia aequalis and Haemonia nig- 

 r i c o r n i s the circular excavation near the center of the bot- 

 tom of the cocoon is wanting, and the pupa has to depend en- 

 tirely on the air that can reach it through the paired slits. 



Not only is the attachment of the larvae at the time they are 

 spinning their cocoons important for furnishing an adequate 

 supply of air for the larvae while they are undergoing the gi'eat- 

 est muscular activity of their life, but an abundant supply of 

 air is also undoubtedly important in freeing their cocoons of 

 water. If, when the cocoon was completed, it were filled with 



