164 ROLAND THAXTER ON THE 



spherical hyphal bodies. Secondary conidia like the primary. Resting spores, azygo- 

 spores or zygospores (?) , formed laterally or terminally from hyphae, spherical; hyaline, 

 30-^5/t. Host attached to substratum by long and conspicuous rhizoids, few in number 

 and terminating in an irregular, disc-like expansion. 



Hosts. Lepidoptera: larva oi HypJiantria textor, imagines of Tortrix sp.. Deltoid sp., 

 Petrophora sp. (geometrid) . Diptera: numerous genera of small flies or gnats. He- 

 miptera : imago of a species of leaf hopper ( Typlilocyha) . 



Habitat. Maine and North Carolina. 



Var. major. 

 PI. 15, figs. 71-73. 



Conidia more nearly spherical ; 38 X 45/^ - 55 X GOy. ; basal papilla smaller in propor- 

 tion to the body of the conidium than in the typical form. 



Host. Coleoptera: imago of Ptilodactyla serricollis {fide Henshaw). 

 HaMtat. CuUowhee, ]S^orth Carolina. 



It is with some reluctance that I have given a new name to this form, although by the 

 presence of strong rhizoids and by its apiculate papilla it seems a well-marked species. 

 The existence of three allied species, E. Jassi, E. PlancTioniana and E. conglomerata, 

 concerning neither of which is any information procurable beyond the insufficient data 

 already published, makes the description of a related form necessarily dangerous, so that 

 the present name is in a measure provisional until more complete descriptions or figures 

 of the three forms mentioned are published. The case is rendered still more confusing 

 by the occurrence on aphides of the species described below and placed provisionally un- 

 der E. Planchoniana. This may prove to be the same with the form under consideration, 

 and both may be identical with E. Planchoniana; yet, what the latter species is, it is 

 wholly impossible to determine from published data. 



In the present species we have the first recorded instance of an Empusa growing up- 

 on a lepidopterous imago as its host, although, as will be subsequently seen, this is not an 

 unique case. The geometrid moth, above mentioned as one of its hosts, was taken on 

 the wing, flying slowly with an imnatural fluttering motion and being placed in a col- 

 lecting box it was soon fastened to a leaf by five or six long and powerful rhizoids. 

 These on examination were found to be made up of several hyphae each ending in an 

 irregular expansion (fig. 75) around which a hyaline substance seemed to have been se- 

 creted, converting it into a disc-like sucker. Where several of these terminations had 

 touched the leaf at adjacent points the several expansions usually coalesced to form one 

 large " sucker," with a continuous even outline formed by the secretion just mentioned. 

 This moth contained both conidia and resting spores, the latter forming in a manner ex- 

 actly resembling that subsequently mentioned under E. spJiaerosperma. The appear- 

 ance of a resting spore having its oi-igin in connection with a cross partition, as in fig. 

 74, occurred occasionally and seemed wholly analogous to the similar appearance in E. 

 sphaerosperma (figs. 214-216). 



The conidiophores, in many instances, arise directly from nearly spherical or elliptical 

 hyphal bodies (fig. 64) ; but in some cases I have observed a condition where the body 



