170 ROLAND THAXTER ON THE 



14 X 30/i-20 X 37/i, average 15 X 35,a. Gonidlophores digitate (?). Secondary conidla 

 like the primary or more commonly long-cylindrical, rounded at either end, and borne ver- 

 tically on capillai-y conidiophores. Resting spores unknown. Host attached to leaves, 

 etc., by its mandibles. 



Hosts. Coleoptera : imago of Chauliognathus Pensylvanicus. 



Habitat. Cullowhee, N. C. 



This interesting species was unfortunately not observed by me until a day or so before 

 leaving Cullowhee, so that my material is limited to two specimens in fair condition, yet 

 not sufficiently good to determine definitely the charaetei- of the conidiophores which, I 

 have little doubt, are typically digitate. The specimens were allowed to discharge their 

 conidia in a damp chamber so that the conidiophores were apparently exhausted by the 

 operation; and those remaining were of the character represented in fig. IGl. The sec- 

 ondary conidia of the second type are very peculiar, and separate the species at once from 

 all others by their vertical position upon the conidiophore, as well as by their peculiar 

 shape which is exactly that of a watermelon. 



In the absence of more definite knowledge concerning the structure of the conidio- 

 phores, as well as from the lack of material containing resting spores, the aifinities of the 

 species are somewhat doubtful. It seems allied both to Triplosjiorimn and to the sjiliae- 

 rosperma group of Entomoplitliora; but the type of development in its resting spores 

 must be observed before this can be determined. 



The melon-shaped secondary conidia are produced in great abundance, even in a fairly 

 moist atmosphere, and in only one or two instances I observed a second form like the 

 primary conidium in process of formation. In many cases, the more common secondary 

 conidia were larger than the spoi-es from which they were produced, and this, together 

 with the apparent thinness of the spore wall, seems to indicate that they are not as re- 

 sistent as is usually the case with secondary conidia when borne on capillary conidio- 

 phores. It will also be noticed that the secondary conidiophores are much less thread- 

 like than is usually the case. 



The affected beetles (fig. 172) were found firmly attached by their mandibles to grass 

 or leaves in open fields; and, to judge from the number of heads that I observed in this 

 position from which the bodies had been broken away, I should infer that the species 

 had been not uncommon in the locality mentioned. The same beetle was abundant at 

 Burbank, E. Tennessee, yet here I observed no specimens of the Hntomopldliora. 



Empusa (Entomophthora) geometralis nov. sp. 

 PI. 17, figs. 173-178. 



Conidia short-elliptical to ovoid; 15-22/^x10-12/.; contents finely granular, with a 

 hyaline nuclear body. Conidiophores digitate, coalescing. Gystidia not observed. Sec- 

 ondary conidia like the ])rimary, or long almond-shaped and borne obliquely on capillary 

 conidiophores. liesting spores borne laterally or terminally on short hyphae as in H. 

 spliaerosperma ; spherical, colorless, average diameter 30,a, maximum, 35/.. Host attached 

 to substi-atum by numerous rhizoids issuing from the abdomen at nearly the same point, 

 and for the most pai't coalescing. 



