182 ROLAND THAXTER ON THE 



to Sept. 5 of the past summer (1887), occurring commonly on the largest Tipulae, which 

 were often found in considerable numbers adhering to wet logs or similar oljjects along 

 the small brooks, so constantly met with in the Carolina mountains. Localities most at- 

 ti'active to the Tipulidae were naturally the most productive of the Entomophthora and, 

 in situations where the brooks were obsti-ucted by fallen trees or roots over which they 

 partially flowed, a dozen or more often occurred side by side. A plentiful supply of 

 moistui-e seems always necessary for the development of the fungus, which is only found 

 on surfaces saturated with water. This circumstance results in a very luxuriant growth 

 of the conidiophores which makes this species the most conspicuous of any form on Dip- 

 tera known to me. Although the abdomen of the host is usually nai-row and not more 

 than a few mm. broad, the conidiophores, after bursting through the integument about 

 the thorax and between the segments, become confluent and spread to some distance on 

 either side of the body forming a clear white mass of considerable dimensions. The ef- 

 fect thus produced is decidedly ghost-like, especially when a number of Tipulidae are 

 seen in this condition beneath a shaded log, and has suggested the specific name. 



The species is remarkable from its formation of true zj'gospores in a fashion similar 

 to that discovered by Nowakowski in his E. ovispora,E. conica and E. curvlspora. I 

 was unable to determine that the zygospores were formed from the same hyphae which 

 bear the conidia, as stated by Nowakowski in the species above mentioned, and it seems 

 more probable from my observations that certain of the hyphal bodies develop conjugat- 

 ing hyphae, while others develop conidiojjhores. I was unable to satisfy myself on this 

 point, since cultivation of the chlamydospores in water produced only conidia; and, in 

 specimens producing zygospores, the hyphae were invariably too much bi"oken up to 

 allow of any definite conclusion in this respect. I found no instances in which zygo- 

 spores alone were formed and usually the number was comparatively small. They were, 

 however, almost invariably present in greater or less abundance in the considei-able num- 

 ber of specimens examined; yet only in a small number of cases did I find them in the 

 early stages of formation. This is apparently due to the great rapidity of the process 

 after it has once begun, since, in several instances, specimens in which no zygosporic 

 formation was visible when first examined, contained ])erfectly formed zygospores after 

 the lapse of a few hours. The singularly large cystidia are very noticeable and readily 

 seen with the naked eye, the drop of moisture which, as a rule, they bear at their apices 

 giving them the appearance of conidiophores with single apical spores of unprecedented 

 dimensions. The species is allied to E. ovisjjora and E. conica as well as E. rJnzosjjora 

 with all of which it agrees exactly in the origin of its zygospores. It is perhaps most 

 nearly allied to E. ovispora from which, however, it differs by its much larger and dif- 

 ferently shaped conidia, which although subject to some variation in size and shape, es- 

 pecially when developed in situations where the supply of moisture is insnflScient, are 

 yet tolerably constant at from 40 to 50,.,. in length. The ease with which the outer wall 

 becomes separated in this species is noticeable in fresh material, and in several instances 

 I observed this phenomenon carried to such an extent that the spore proper floated free 

 within the inflated and spherical mother cell (fig. 321). 



It should be noted that the occurrence of this species although common was decidedly 

 local, and in no instance did I find it in company with allied forms at even a short dis- 

 tance from the water. 



