174 EOLAND THAXTER ON THE 



or after pupation and were found to contain irregular bodies resembling hyphal bodies, 

 but often without apparent walls and associated with a general fatty -degeneration of the 

 whole contents of the pupa. At this date (Sept. 17) it was only possible to procure very 

 poor material for infecting a fresh set of Pieris lai-vae, and although much younger larvae 

 were selected than in the previous experiment, no results were obtained j the caterpillars 

 upon reaching maturity either dying from insect parasites or pupating. 



Although these experiments by themselves are not by auy means convincing, the in- 

 fection of even a single Pieris larva from a host so totally different, and under circum- 

 stances such as I have described, when taken in connection with the structural similarity 

 existing between the American and European forms, furnishes in my opinion sufficient 

 gi'ound for considering them as belonging to a single species. 



The occurrence of this species on a butterfly {Colias philodice) , fig. 200, is also an in- 

 teresting fact; the specimen in question, which was found on the ground in a pasture 

 at Kittery, being, I believe, the first recoi'ded instance of this nature. This is also I 

 think, the only species that has been found to attack hymenopterous imagos, on which 

 it very commonly occurs; a fact that might be of interest to bee-raisers, should the fun- 

 gus develop a local fondness for bees. In this connection it may be noted, however, that 

 several hives of bees placed luider the trees in the apple orchards above mentioned, 

 showed no signs of any entomophthorous disease. 



The species probably has a very wide geographical range, and although in IN^orth Caro- 

 lina I have found it less common than in ]S"ew England, it occurred in some quantities on a 

 fly, somewhat smaller than Musca domestica, which I found fixed to the under sides of the 

 leaves in Rhododendron thickets on the summit of Roan Mountain. On Mt. Washington, 

 N. H., I have found it infesting flies, ichneumons, and a minute Thrips which it destroyed 

 in great numbers. This occurrence upon Thrips is interesting as illnstrating the omnivo- 

 rous character of the parasite. Krasilstchik, in his list of insect hosts attacked by fungi, 

 gives Thrips as the host of an undetermined species of Entomophthora which is, I believe, 

 the only other recorded instance of this insect as an JEmpusa host. The infested larvae, 

 pupae and imagines of this insect, as they occurred on Mt. Washington, were hardly vis- 

 ible without a hand lens. Both conidia and resting spores occurred in this instance; the 

 former somewhat smaller and more irregular than in the case of hosts whose size admits 

 of a more luxuriant growth of hyphae, yet differing but little from the usual type. 



The general appearance of the fungus in this, as well as in the succeeding species, 

 when it is well developed on its smaller hosts, is almost always sufficiently peculiar to 

 distinguish it from other Empusae. The whole mass of conidiophores coalesces over the 

 body often leaving only the middle of the thorax uncovered; but instead of presenting 

 the rounded form usnal in other Empusae, the upper surface has a flattened appearance, 

 and is the only portion from which conidia are dischai-ged. This is due to an oblique 

 tendency observable in the growth of the main hyphae, as is indicated in fig. 202, the left 

 hand portion in the figure corresponding to the outer part of the whole mass. The ba- 

 sidia thus have a general tendency to point upwards instead of in all directions as is 

 usually the case. The occurrence of specimens in which the mass of conidiophores is 

 colored green or greenish is not common, and the color fades on drying; but in two 

 cases in the epidemic among leaf hoppers just mentioned, I found specimens in which it 



