19 
cessive boilings in plugged tubes, were infected from it. Under 
these conditions the fungus grew vigorously, forming a hardly 
visible, silky, web-like coating of mycelium, which, in about 
ten days, seemed to be transformed into a dense brown powder, 
owing to the abundance of spores it produced. In old cultures 
this layer was often two or three millemetres in thickness. On 
examining cultures a week old, the mycelium was found to be 
septate, colourless, and strongly vacuolated. Arising from it 
were numbers of upright a-septate hyphe, branching to form 
either a simple or compound raceme. The ultimate branches 
were almost invariably in groups of threes. The apices of these 
branches bent over crozier-fashion, swelled slightly, and then 
coiled into a close spiral of one-and-a-half turns (Fig. 1a). Three 
transverse septa then appeared in this closely coiled portion, and 
frequently a fourth one lower down the hypha. The penulti- 
mate cell of the group, Le. the middle one of the three, then 
snereased in size rapidly, while: the others grew more slowly. 
The result of this unequal growth was to fuse the three cells 
into a group in which the original spiral arrangement was lost 
sight of (Fig. 1b). At this stage the contents of the penultimate 
cell became strongly vacuolated, and its wall gradually thickened, 
became brown in colour, and covered with large warts. Similar 
changes, but far less pronounced, occurred in the other two 
cells. Meanwhile further septa were formed in the fertile 
hyphz, and the group of three cells was cut off as a whole. It 
now consisted of a thick-walled, warted spore, partially invested 
by the cells originally immediately above and below it, though 
now displaced by the unequal swelling (Fig. tc). From its 
mode of development, the central spore was evidently a 
chlamydospore, not a conidium. 
Berkeley, in his original description, makes mention of certain 
abnormalities, and a striking point in the life history of this 
fungus is the frequency with which they occur. Normally, the 
group of three spores has much the appearance of a teleutospore 
of Triphragmium, in which two of the cells are only partially 
developed, but occasionally owing to the development of the 
two upper cells only they resemble those of a Puccznza, while 
more rarely Phragmidium-like groups were met with. The last 
two forms can be produced in abundance in cultures on melon- 
rind. In size, the normal spores were from 25-30u in diameter 
but when grown under water they were frequently as much as 
45, and with thin, translucent, smooth walls. Grown in a very 
dry atmosphere they were only some 15-18 in diameter, while 
the walls were very massive, dark brown, and again wartless 
These marked changes in size, colour and markings, make one 
wonder whether we do not often lay too much stress on spore 
