10 E. J. BUTLER. 
plate II, of Smith’s paper on this point. Microscopical prepara- 
tions taken from cultures on nutrient agar, showed no definite 
character by which the conidial stages of the two fungi could be 
differentiated. 
As the successful infection experiments carried out at Dehra 
Dun were made with conidia obtained from the roots of pigeon-pea 
and not directly from Nectria ascospores, it at once became evident 
that these conidia might have belonged to either of two distinct 
fungi, Nectria or Neocosmospora, and that in this might lie the 
explanation of the irregular results obtained. It appeared possible 
that the successful experiments might have been made with conidia 
belonging to Neocosmospora, but erroneously taken for those of 
Nectria. This became the more probable when it was found that 
Neocosmospora occurs not only at Dehra Dun, but throughout a 
large area of Northern India, where pigeon-pea wilt is common, and 
also that on almost every diseased root examined, Neocosmospora 
perithecia should be induced to form by appropriate treatment. 
Work was therefore continued with the latter fungus. 
Experiments with Neocosmospora vasinfecta Smith on 
Pigeon-pea. 
To test the parasitism of Neocosmospora a series of infection 
experiments was carried out with pure cultures of the fungus. 
Only ascosporal cultures or those known to have originated from 
perithecia were used, since it was evidently useless to work with 
conidial forms, which there was no means of referring to their 
proper genera. 
SERIES V. 
The cultures used were obtained as follows :— 
Some young perithecia of Neocosmospora were washed care- 
fully in several changes of distilled water, and were sown separately 
in decoction of pigeon-pea. They sprouted freely from the peri- 
thecial cells (as most Hypocreaceous fungi will) and produced 
quantities of Cephalosporium spores. 
