24 
material has prevented my following out the germination of these 
ascospores in detail, though I hope to be able to do so before 
long. To obtain them free from any adhering spore-balls the 
perithecia were well washed in_ sterilized water, and then 
crushed in beer-wort gelatine, from which drops were removed 
and mounted in cells for microscopic observation. In the 
course of twenty-four hours (16 C) the spores germinated and 
quickly formed a vigorous branching mycelium, the aerial por- 
tions of which produced numbers of spore-balls, and not, as | 
had expected, the chlamydospore form first described. This js 
the only set of germinations I have, so far, had the opportunity 
of making, and whilst it proves that the ascogenous stage really 
belongs to Acrospezra, it still leaves open the question as to how 
the presumed cycle of ascospore to chlamydospore is brought 
about. It seems probable, though, that further experiments 
under varying conditions of food supply will clear up this point 
also. 
Berkeley's Acrospeira mirabilis thus turns out to be one 
stage in the life history of a Sphaeria, a group of fungi which, 
from the biological point of view, very little is known at present. 
A careful study of the life histories of some of the Dematiee 
would, in all probability, show that many more of them are 
merely stages in the life-history of species belonging to this 
group. This work will, however, have to be carried out along 
the lines laid down once for all by de Bary—the story must be 
followed out in cultures from spore form to spore form, for the 
mere juxta-position of two different spore forms, even when 
repeatedly met with, cannot be taken as certain evidence that 
the two are stages in the life history of one and the same 
fungus. Thus in the case before us the mere fact that the 
perithecia occur with the spore balls is valueless unless the ger- 
mination of the ascospore is taken into account, and although 
the spore-balls and simpler chlamydospores do not occur 
together in nature, their connection is better proved than that 
of the former spore forms, for it rests on identity of results on 
germinating under a variety of conditions. 
EXPLANATION OF FIGURES. 
1. (a) Earliest stage in the development of chlamydospores ; 
(b) the curled apex is septate and the penultimate cell s 1s 
beginning to outgrow the others; (c) the ripe spore. 
2. First stages in the development of the spore-ball from : 
culture on sterilized chestnut, showing the formation 0 
the envelope layers. 
3. A spore-ball, in section, from a chestnut culture, with a 
many-layered envelope and many spores. 
