ENTOMOPHTHOREAE OF THE UNITED STATES. 151 



oped upon the common web worm {^Ilnpliantria textor) after pupation; but this occur- 

 rence is certainly luiusual. Among Lepido])tera, I have been surprised to find the 

 imagines attacked in numerous instances, a fact, I believe, not hitherto observed; and 

 one species (E. geometralis) seems peculiar to them. Geometridae, Noctuidae and Ti- 

 neidae may be affected, and I have even found the common sulphur butterfly {^Colias 

 jyhilodice) thus attacked. The species or family of the host has hitherto been generally 

 considered a means of determining the species of Einpusa, in most cases, with some 

 accuracy; yet my observations have shown that this is by no means the case and that 

 specific distinctions, based largely upon the character of the host, are of little value. 

 The variety of hosts attacked by a single species is sometimes very great, perhaps in no 

 instance more so than in E. sphaerosperina. I have specimens of this foi-m upon the larva 

 of Pieris, on the imago of Gollas j^^Modice, Diptera of several families and genera, 

 Phytoiiomtis larvae (Coleoptera) , the common rose-leaf hopper ( Typldocyh(i) and Aphis 

 (Hemiptera), on ichneumons of several genera and a small bee (Ilymenoptera), and on 

 a species of Thrlps (Thripidae), while in Europe it is also recorded on LimnopMlus 

 (jS^europtera) . "With such a diversity possible in the hosts, it is obviously unsafe to de- 

 scribe, as new species, forms without peculiar characters of their own, merely because 

 they occur in a new host, or to give any considerable specific weight to the character of 

 the host in support of slight variations in the shape of the conidia or the size of the 

 resting spores. 



In cei'tain instances, where several Empusae were found together, I have noticed two 

 species developed iipon a single host; for example, E. Aplndis and E. Fresenii, as well 

 as E. conica and E. papillata; E. gracilis and E. variahilis; E. lageniformis and E. 

 occideidalis. It is, therefore, not always safe, where such proximity exists, to refer the 

 resting spores that may occur in connection with conidia to the same species of Em- 

 pusa. 



The hahitat of Emjrusae is various, one of the most productive localities being the 

 margins of brooks in shady woods. Certain species are found only in such situations, ad- 

 hering to wet substances, such as moss, logs, stones, etc., in the water or along its mar- 

 gin; a constant supply of moisture being apparently necessary for their development. 

 In other cases dryer situations are preferred, and the fungus readily withstands the al- 

 ternate dryness and moistiire consequent upon the variations of weather, producing its 

 conidia repeatedly, whenever the atmosphere is sufficiently moist, imtil the conidiophores 

 are exhausted. Many hosts, before death, seek conspicuous positions by crawling up- 

 wards on grass, or other substances, whence the conidia are discharged over a consider- 

 able area. Perhaps the favorite position assumed by hosts before death is upon the 

 under side of leaves in shady situations in woods or about houses, Avhere a careful search 

 during wet weather seldom fails to disclose numerous specimens. I have noticed only 

 one species which occurs on flowers attractive to insects, namely E. Muscae; which, al- 

 though common in all parts of houses, I have only seen in nature on the flowers of 8ol- 

 idago and contain certain Umbelliferae. 



Miscdlaneoiis notes. — In collecting Empusae I have found that, as a i-ule, foggy 

 weather is the most favorable for the purpose, since they are more conspicuous when 

 distended by moisture. A shallow tin box, partly filled with moist Sphagmim, is a con- 



