no Transactions British Mycological Society. 



biella, but if the strands should happen to become erect it would 

 be Cordyceps. He also described and figured a species, which 

 had been named Cordyceps Moelleri by Hennings, in which the 

 clavae were composed of parallel hyphae, and the perithecia 

 were not sunk in the clava. 



Support of Moller's objection is afforded by a species, Torru- 

 hiella ochracea Pat., which occurs on an undetermined Noctuid 

 in Ceylon. In three examples of this species, collected on different 

 occasions, there are no clavae; the perithecia are situated on a 

 weft of hyphae which covers parts of the insect, or on places 

 where no external stroma is visible. In another example, 

 however, sixteen processes arise from the insect. These are 

 up to 5 mm. high, usually laterally compressed, up to i mm. 

 broad, with dense clusters of perithecia at their apices. The 

 processes are composed of more or less parallel hyphae, firmly 

 adherent to one another, and the perithecia are superficial. In 

 other respects, including the conidial stage, the specimens are 

 identical, and there does not appear to be any room for doubt 

 that they are the same species (Plate II, figs, ii, 12). 



From Moller's figures, Cordyceps cristata Moller and C. Moel- 

 leri P. Henn. would appear to offer a parallel case. Both are on 

 a Noctuid and have yellow perithecia. The perithecia differ in 

 size, those of C. cristata being 300-500^ high, according to 

 Moller, and those of C. Moelleri up to 700 /a high, but this varia- 

 tion is not greater than that in the Torruhiella forms of T. ochracea. 

 In C. cristata the perithecia are situated on a loose stroma over- 

 running the body of the insect; in C. Moelleri they are grouped 

 on erect clavae, up to 1-5 cm. long, composed of parallel hyphae. 

 In Moller's figure 80 (C. Moelleri) the perithecia are clustered 

 and appear free, but the description states that they are more 

 or less embedded in the clava, up to one-third their height. 



Tranzschel, in Hedwigia (1899), p. (11), instituted the genus 

 Helminthascus for a fungus found on a spider in Russia. The 

 perithecia were totally immersed in a flattened pulvinate stroma. 

 Saccardo, in Sylloge Fungorum, xvi (1902), p. 616, wrote that 

 the genus "fere absque dubio cum gen. Torrubiella collidit": 

 that would appear to depend upon the character of the structure 

 described as a stroma (cf. Torruhiella barda and T. sublintea). 



Barya montana Rac. in Bull. Acad. Sci. Cracovie (1907), 

 p. 909, would appear from the description to be Torrubiella. 

 Raciborski placed it in Barya, in preference to Torrubiella, 

 because it had no paraphyses. It occurred on a spider in Java, 

 and was said to have a Stilbum conidial stage. The common 

 conidial fungus on spiders in Ceylon is Gibellula, but it has not 

 been possible to connect that with any Torrubiella. On the 

 other hand, Torruhiella flava, which occurs on spiders, has 



