23 
At this stage it seemed feasible that the fungus was really an 
Ascomycete, although the spore-balls pointed to its being a near 
relation of some Ustilago, such as Urocystts, or possibly 
Geminella. It appeared to grow too readily as a saprophyte to 
belong to such a typically parasitic group as the smuts, and, 
further, endoconidia are characteristically produced by the 
Ascomycetes only, while Eurotium-like and Ascobolus-like 
ascogomia were more than suggestive of the latter. | Never- 
theless, a long series of experiments, carried out on 
all the media and under all the conditions I could 
command, gave no further results, and, finally, I . had 
to fall back on infection experiments with the living 
nuts. For this purpose, what appeared to be sound nuts 
were infected with chlamydospores, or with the spore-balls, after 
carefully sterilizing the outer skin with an alcoholic solution of 
corrosive sublimate. The wounds were then dressed with 
paraffin wax to stop the entry of any other fungi, and the nuts 
were stored in slightly-moistened sand. Under these conditions 
it was possible to obtain a number of pure cultures of the fungus, 
though some of the nuts were attacked by Penicillium and 
Mucor. The chlamydospore infections gave a crop of chlamy- 
dospores only, the spore-balls gave spore-balls and small 
reddish-brown, hard-walled perithecia. The walls of the 
perithecia were smooth and without bristles, and the ostiole was 
small, and flush with the surface, z.e., not raised on a papilla or 
forming a neck. It contained no paraphyses, but numbers of 
broadly club-shaped asci, each with eight elliptical, dark-walled 
spores, slightly pointed at each extremity. The development 
of these perithecia proved very difficult to trace, and all of the 
details have not so far been clearly made out. Thus the origin 
of the ascogenous hyphe is still undetermined with any degree 
of certainty, and I have to assume that they spring from the 
scolecite. In the earliest Stages I have been able to obtain the 
whole of the perithecium was filled with a pseudo-parenchyma 
of hyphe, richly stored with food materials. Midway between 
the apex and the centre this parenchyma broke down, forming 
a spherical cavity lined with delicate hyphe, all pointing in an 
upward direction, which finally formed the lining of the ostiole. 
The whole mass of pseudo-parenchyma_ then became dis- 
came septate at the apex, the penultimate cell of the group giv- 
mg tise to an ascus after the method described by Dangeard. 
The broken-down parenchyma appeared to serve as food- 
material for the developing asci, which matured rapidly, and 
practically filled the whole of the perithecium. Hitherto lack of 
