Studies in Entomogenoiis Fungi. T. Fetch. 115 



of some thick- walled hxphae, but the mass of the piil\'mate 

 stroma subsequently formed consists, in general, of thick-walled 

 hyphae, flexuose and irregularly intert\%ined. In some cases, 

 the h\-phae on the surface of the stroma tend to fuse into a 

 continuous, glabrous sheet; this is well-marked in species from 

 the American tropics. In the only a\-ailable species on Lepi- 

 doptera, the h\-phae of the stroma are thin- walled. 



The asci are very long. c\lindric. narrow. \nth a capitate 

 apex. They contain usually eight linear ascospores, more or 

 less spirally arranged, and as long as the ascus. These become 

 septate and subsequently di\'ide into cyhndrical part-spores, 

 which may become rounded at the ends and narrow-oval. This 

 di\'ision into part-spores has been recorded in T. tomentosa, 

 T. rubra, F. sericicola, F. ochracea, F. Lecanii, Barya montana, 

 and Helminthascus arachnophthorus. On the other hand, 

 Boudier did not find part-spores in F. aranicida, nor did Hen- 

 nings in F. rostrata. Zimmermann stated that he did not 

 obser\"e septation or the formation of part-spores in F. luteo- 

 rostrata, and von Keissler recorded the ascospores of F. hrunnea 

 as continuous and apparently not di\iding. 



Among the many unexplained phenomena attending fungi 

 parasitic on scale insects, perhaps the most notable from the 

 svstematic standpoint are the comparative rarity of the 

 ascigerous stage in Hypocrella and Sphaerostilbe, and the rarity 

 of occurrence of fully ripe perithecia when the ascigerous stage 

 is found in Hypocrella and Forrubiella. In the latter two genera 

 the h\-phae of the stromata are sclerotioid and adapted to %\ith- 

 stand periods of drought; it may, perhaps, be suggested that 

 these fungi may be subject to, and survive , a cessation of gro^^th 

 at anv stage, and that the spores are onl}- matured and ejected 

 from the perithecium under certain limited weather conditions. 

 Certainly, the majority of the collections of Hypocrella and those 

 of Forrubiella on scale insects contain only immature asci. Under 

 such circumstances, descriptions of the ascospores based on a 

 single collection can be correct onh' by a lucky chance. 



In the specimens examined by me, septate ascospores have 

 been commonly observed. Loose part-spores have been seen, 

 i.e. not in an ascus, and also part-spores in asci, the greater part 

 of the contents of which appeared to have become disorganised. 

 But I have not been able to find the stage which was expected, 

 that which has been frequently seen in Hypocrella, in which the 

 separate part-spores are arranged in the asci in fines corre- 

 sponding to the original undi\"ided ascospores. It appears, 

 however, to be qmte certain that the ascospores of Forrubiella 

 di\*ide into part-spores \nthin the ascus. 



In Revue Mycologique (1887), p. 15S, Boudier described 



8—2 



