ENTOMOPIITIIOREAE OF THE UNITED STATES. 



181 



species. Having discovered the peculiarity of its resting spores, I looked for it with some 

 care in North Carolina during the succeeding summer and was rewarded by finding the 

 same host repeatedly infested by the same fungus. In Carolina it occurred almost in- 

 variably on the under side of leaves especially of Impatiens growing beside brooks in 

 the woods. The conidial form was extremely rare, but of the resting condition I pro- 

 cured some dozens of specimens. In very many cases the resting spores are produced 

 wholly, or more commonly partially, outside the body of the host, having originated from 

 hyphae (fig. 297) which emerge alter the manner of conidiophores, and subsequently con- 

 jugate as shown in figures 298-302. Certain of tbese hyphae are seen to project some 

 distance beyond the rest (fig. 297) and ultimately form a delicate web which serves to 

 hold the mass of spores to some extent. None of my material was sufficiently good 

 clearly to show conjugation within the body of the host, and the hyphae were so evan- 

 escent and filled witb such large globules of fat that points of conjugation could hardly 

 be made out in the confused mass. The spines develop more rapidly within the body 

 than without and as a result of their formation the abdomen becomes greatly inflated. 

 This inflated appearance together with the peculiarity of the host are sufficient to dis- 

 tinguish the species in collecting, and it is a singular fact that this particular species of 

 fly (Sajjromyza), even in situations where E. dipterigena was common and E. Americana 

 not infrequent, never appeared to be attacked by any other Empusa. The spinose char- 

 acter of the spore is due to elevations of the rather thin-walled epispore only, and is thus 

 readily obliterated by long exposure or transformed into a roughness of the surface. 

 The epispore may also be easily removed by pressure of the cover-glass upon it, disclos- 

 ing within a spherical resting spore of the usual type. The species is very nearly allied 

 to E. dipterigena, the conidia of the two being hardly separable in some cases. The pa- 

 pilla is, as a rule, broader in E. echinospora and the spores seldom tend to become fusi- 

 form. The type also represented in figures 243-244, which is characteristic of groups 

 of conidia in the former species, is wholly wanting in the latter as far as I have observed. 



Empusa (Entomophthora) sepulchralis no v. sp. 



PI. 19, figs. 306-326. 



Conidia long-ovoid to long-elliptical or subfusiform, rounded at the apex and with a 

 papillate base commonly bent to one side; hyaline, with numerous large oil globules; 

 measurements 35—48 X 10-15//, maximum 15 X 55//. Conidiophores digitate, arising from 

 large (60// diam.) spherical hyphal bodies and coalescing over the body of the host in a 

 clear white mass. Cystidia very large (70-90// in diameter), slightly expanded or com- 

 monly becoming branched at their apices. Secondary conidia like the primary or short- 

 ovoid. Resting spores, zygospores, spherical, hyaline, 35-50//. in diameter ; formed as the 

 result of a Sj)irogyra-\ike conjugation between two hyphae and arising by budding from 

 the gametes. Host attached to substratum by numerous rhizoids. 



Hosts. Diptera: imagines of Tipulidae. 



Habitat. North Carolina, E. Tennessee. 



This fine species was collected at Cullowhee, Cranberry and Burbank from June 20 



MEMOIRS BOSTON SOC. NAT. HIST., VOL IV. 26 



