81 



some larvoe change their habitation three times, but whether all larvse 

 have the same habit, or whether they change of tener than three times, 

 is uncertain. However that may be, it will be seen that the damage 

 inflicted by the larvfe is greatly increased by these migrations, for 

 every leaf attacked is doomed to destruction. 



The duration of the larva state appears to be quite variable. In 

 captivity no larva was actually carried through from the time of 

 hatching to pupation, but from observations in the field it is pretty 

 clear that this period, even in the height of the season, is never less 

 than two weeks, and probably lasts on the average three weeks, being 

 somewhat longer in the beginning of the season than in midsummer. 



The pupa. — The pupa state is assumed within the mine and lasts from 

 six and a half to ten daj^s, after which the perfect beetle breaks its way 

 through the brittle epidermis of the leaf. 



At 11 a. m. of August 5 a larva taken from its mine was found to 

 have transformed to pupa within two or three minutes, the change 

 having taken place while it was being taken from one room to another 

 and returned. August 12 the imago was found fully colored early in 

 the morning, the pupal period having been passed in about six and a 

 half days. Temperature, 75°-80^ F. 



The mmiber of generations annually. — This is a point in the natural 

 history of the insect which has not hitherto been clearly made out. 

 Dr. Harris (Ins. Inj. to Veg., Flint ed., p. 121) is evidently in favor 

 of a single generation, and we have no doubt that this will hold true^ 

 for the more northern States; perhaps elsewhere. In the latitude of 

 Washington the insect, as has been stated, appears in May, one month 

 earlier than in Massachusetts, and is then to be found in all stages 

 throughout the summer, until the beginning of August, when <i^^ 

 laying ceases. It is therefore evident that individuals produced from 

 eggs deposited early in the season developed to the mature insect early 

 enough to produce a second generation in the same manner, but no 

 second generation has as yet been segregated. 



NATURAL ENEMIES. 



Of nonparasitic enemies we have observed at Washington the wheel 

 bug {Arilus^ cristatus)., which appears to be particularly attracted by 

 the presence of the Odontota.^ The young larvaj of the wheel bug 

 may frequently be observed slowly walking over the leaves until they 

 have found a nnne inhabited by the Odontota. This found, the wheel 

 bug deliberately pierces with its beak the Odontota larva through the 



" Formerly Prionidus. 



'' The abundance of the wheel bug on the Agricultural grounds on the trees infested 

 by the Odontota is in striking contrast to its comparative scarcity on the elm trees 

 infested with the imported elm leaf-beetle ( Galerucellu luteola) and which are in close 

 proximity to the locust trees. 

 3622— No. 38—04 6 



