Studies in Entomogenous Fungi. T. Fetch. 127 



It would seem probable, therefore, that Torrubiella ochracea 

 is identical with. Cordyceps Sphingum, but a final decision on the 

 point must be deferred until a comparison can be made with 

 authentic specimens of the latter. 



Torrubiella flava Fetch, n.sp. This species (Plate II, 

 fig. 9) occurs at Hakgala, Ceylon, on spiders. In the specimens 

 hitherto collected, the insect is attached b}* the mycehum of the 

 fungus to the under surface of U\'ing leaves. The fungus clothes 

 the body of the insect \\"ith pale yeUow mycehum, forming a 

 flattened-pulvinate, tomentose, somewhat spong\- stroma, with 

 a fimbriate margin spreading shghtl\- over the leaf. The h\'phae 

 of the stroma are 2-^-2^^i diameter, usuall}- thin- walled, but 

 sometimes %%ith walls up to 0-5^ thick, especiallv towards the 

 free ends of the h\-phae. The perithecia are seated on the stroma, 

 superficial, ultimately crowded, tomentose \\-ith a rather thick, 

 separable layer of tomentum, pale yellow, with a yeUow-brown 

 translucent apex, ovoid, o-S mm. high, 0-4 mm. diameter. The 

 h^-phae of the tomentiun are 2-2-^ yL diameter, moderately thick- 

 walled, and minutely granular. The asci axe long, cylindric, 

 capitate, 5-6 /x diameter, four or eight-spored. The ascospores 

 are as long as the ascus, and di\"ide into cvlindric part-spores, 

 5-S X 1-5 /Li, the septa in the undiWded spore being exceptionally 

 weU-developed. Paraphyses are present, as long as the asci, 

 thin-waUed, lax, collapsing, not capitate, expanding to 3/i 

 diameter at the apex. 



A specirnen (Plate II, fig. 10) on the same species of spider, 

 gathered \\'ith the foregoing and possessing a stroma of the same 

 character, bears conidial Stilhum-\ik,e clavae. These are paUid 

 yeUow, cyhndric, up to 4 mm. high, 0-15 mm. diameter, tapering 

 shghtly upwards, terete, smooth. The}- are clothed %Wth spheri- 

 cal or subglobose basidia, up to 6 x 5/i, scattered or crowded, 

 each bearing a single apical, slender, rigid sterigma, about 2/x 

 long. The conidia are apical, narrow-ovaJ. ends acute, con- 

 tinuous, hyaline, 4-S -, 2/i. This is a Hirsiitella, which has been 

 described as Hirsutella arachnophila Petch (= TricJwsterigma 

 arachnophilum Petch). The perithecia and the conidial clavae 

 have not been found on the same stroma. 



Hirsutella arachnophila has also been collected at Peradeniya, 

 on spiders, on two occasions. In one of these examples, the 

 clavae were lilac-grey when fresh. It is possible that the latter 

 may belong to a different species, but it does not show any 

 morphological differences. A similar difference in colour occurs 

 in Gibellula elegans P. Henn., in which species both the stroma 

 and the conidiophores may be yellow, or the stroma veUow and 

 the conidial heads lavender. In the latter case, the conidial 

 heads become white in old herbarium specimens. 



