16 CROP PEST COMMISSION OF 



efforts, as they can swim much faster than they can crawl when 

 on dry land. They are, no doubt, carried for long distances 

 during periods of high water, either unprotected or in old logs., 

 pieces of rotten wood or in fact in any plant remains of ca- 

 pacity enough to harbor them. Their appearance in places dis- 

 tant from water may often be explained by their transportation 

 during overflows, and it is safe to suppose that the wide dis- 

 tribution of the species is in a measure due to its distribution 

 J?y the waters of rivers and creeks. It seems probable to me that 

 many larvae reach maturity miles from where they were 

 hatched, and that a mass of eggs placed on a leaf overhanging 

 the waters of such a. river as the Mississippi might produce 

 adults for several different States. 



Larvae in the breeding cages stop eating in late fall, even 

 though they are kept at a temperature as high as is normal for 

 the summer months. This may be taken as an indication that 

 they hibernate during the winter months when compelled to do 

 so. I have spent some time in trying to gather data on this point, 

 and have found the larvae of the species under consideration in 

 more or less exposed places repeatedly, at the beginning of win- 

 ter. I have not succeeded in finding them in a frozen condition, 

 ay I have larvae of other insects, but there is no doubt that they 

 are often caught by winter frosts in northern climates, and in 

 case they are there smns to be no reason why they should be 

 injured thereby any more than the larvae of the moths and butter- 

 flies. So far as my observation goes, all species of horseflies pass 

 the winter in the larval stage; therefore, the effect of cold on 

 this stage is a subject that should be thoroughly investigated. 



FIG. 10. American horsefly (Tabann-K atncri.ca.nufi), 

 female, natural size. 



