ENTOMOPIrTIIOREAE OK THE UNITED STATES. 



177 



shape. The conidiophores are white in the mass, often tinged with yellowish or flesh 

 color from the coloring matter of the host, which usually assumes a pale brick-red tint ;il 

 or just before death. This change of color is, however, common to most aphides at I acked 

 by Empusaeand cannot be considered distinctive of any species. In one instance I have 

 found a form, apparently this species, on a large bug (one of the Corisiaef) ; but, unless 

 the species subsequently described as E. dipterigena proves to be the same fungus, there 

 is little variation from the usual host. 



Empusa (Entomophthora) dipterigena nov. sp. 

 PI. 18, figs. 241-250. 



Conidia variable in shape, ovoid to oblong or subfusiform, with a papillate base, often 

 bent to one side, and containing numerous large oil globules; average measurements 

 11x22,,, maximum 15X30//. Conidiopliores digitate coalescing over the body of the host 

 in a clear white, very rarely bright pea green mass. Cystidia slender, tapering toward 

 the apex. Secondary conidia like the primary or broad-ovoid. Resting spores (zygo- 

 spores?) produced externally in grape-like clusters; spherical, hyaline, 20-40 /Jt in diameter. 

 Host attached to substratum by rhizoids ; few in number, large, with a disc-like terminal 

 expansion. 



Hosts. Diptera: small Tipulac; other small flies or gnats belonging especially to the 

 Mycetophilidae. 



Habitat. Maine, New Hampshire, Massachusetts, North Carolina. 



Before discovering specimens of this species producing both the conidia and resting 

 spores simultaneously, I was inclined to regard it as a mere variety of E. Aphidis; but 

 the peculiar external production of the resting spores, taken with its generally smaller 

 measurements and quite different host, serves sufficiently to distinguish it. It is nearly 

 allied to E. ovispora and E. eeliinospora, but separable at once from the first by its slender 

 tapering cystidia, and from the second by its smooth resting spores. From E. Ameri- 

 cana it is separable by its smaller, often subfusiform conidia, as well as by the presence 

 of cystidia, its general habit and peculiar rhizoids. 



None of the specimens found contained resting spores at a sufficiently early stage of 

 their development to show the nature of their origin. The youngest examples, in which 

 the spore contents were still granular, indicated a tendency to produce distinctly grape- 

 like clusters, and in all cases the spores w T ere external. The species is not uncommon, 

 occurring only on the under side of leaves in woods or thickets, and was first noticed 

 on Mt. Washington in late August, at the head of Tuckerman's ravine, and subsequently 

 collected in swamps about Kittery, Maine, in small numbers. In North Carolina, it was 

 more common in similar situations, both forms of spores occurring not infrequently. 



I have placed the E. rimosa of Schroeter 1 as a doubtful synonym of this form, with 

 which his description coincides with little variation. Schroeter's species is certainly not 

 the E. rimosa of Sorokin, the spores of which belong to a very different type. 



'I.e. 



