HYPOCREACE^ ZEYLANIC^. 87 



It has proved very difficult to discriminate between Berkeley 

 and Broome's Ceylon species of Hyjwcrea, even with the 

 assistance of fresh specimens. Indeed, this genus has given 

 more trouble than any other of the Ceylon fungi whose 

 revision has hitherto been attempted. The colour of the part- 

 spores provides an excellent primary character, but very many 

 of the specimens are so immature that no spores are present, 

 and where hyaline spores are present in the ascus, it is not 

 possible to be certain that that is their final colour, as the 

 development of colour is usually of late occurrence. Com- 

 parison with fresh specimens is rendered difficult by the 

 changes which these species undergo when drying or in the 

 herbarium. Colour, of course, usually fades, or changes to a 

 luiiform red-brown, while one character which can be relied 

 upon when the specimens are fresh, viz., whether the stroma is 

 opaque or subtranslucent, has no definite parallel in the dried 

 specimens. Some subtranslucent species dry to a horny 

 consistency when immature, but become opaque and fleshy, or 

 even friable, if dried in the mature stage, and the texture of 

 the stroma appears quite different in the two stages. The 

 characters which appear to be the most reliable are (1) the 

 manner of attachment of the stroma, whether over the whole 

 base or only centrally ; (2) the position of the perithecia, 

 whether at the same or different depths, and the position of 

 the perithecial layer in the stroma ; and (3) the shape of the 

 perithecia, which, though not, as a rule, constant, varies in a 

 definite direction, viz., from globose to laterally oval, or from 

 globose to vertically oval. 



The genera Cordyceps and Torrubiella have been omitted 

 from the present list. 



Epichloe pulvinulus B. & Br, 



This species was described by Berkeley and Broome in Fungi 

 of Ceylon, No. 981. It was transferred to Hypocrella in 

 Saccardo, Syll. Fungorum, II., p. 581. In Ann. Perad., 

 VI., p. 172, it was re-described as Balansiella pvlvinula (B. & 

 Br.) Fetch. 



6(1)20 (13) 



