224 
A black mass of gum that had been collected and allowed to 
nd 
described above had germinated, and ee innumerable very 
minute hyaline sporules, many of which were reproducing 
mea by a process of budding, Torula- or SAN Om gos: 
dier this discovery another mass of black gum was collected 
and allowed to become perfectly dry and horny. After remain- 
ing in this dry condition for several weeks, a portion of the 
aterial was examined, and the same process of germination and 
reproduction of sporules by gemmation was seen to have occurred. 
After another interval of some weeks, during which the material 
remained perfectly dry, a fragment was placed in a hanging drop 
uh sterilized water, and reproduction of sporules by gemmation 
8 Soon as active as € T process had never been interrupted, 
the matrix of gum presumably serving as nutrient material. 
Germination of the tgo brown cells, and continued repro- 
duction of the sporules by gemmation in a dense matrix of gum 
comparatively devoid of air, suggested the idea that under certain 
conditions the fungus could exist as an anaerobic IE or 
urpose of testing the validity of this idea sks of 
nutritive solution, consisting of thoroughly iBEE colourless 
masses of the gum e ien ed ing the early stage of the disease 
and dissolved in ee P; e prepared according to Kitasato’s 
method, which practi RHY Sori in excluding the air by 
a layer of parafin poured on the surface of the nutrient 
on. 
Seven days after inoculation the contents of the flasks were 
plge: and microscopic examination showed this turbidity to be 
due to the presence of myriads of sporules, mostly arranged in 
liiis of two to four cells. 
When removed to hanging drops of the same nutrient solution 
those anaerobic sporules refused to grow, and eb gs of the 
host plant with them produced no sign of the dis 
Grown in hanging drops or in flasks as OM organisms, the 
large brown tes E origin to a very stout, hyaline mycelium 
composed of t o four cells, constricted at the septa. When 
full grown, vids vale cells give origin from every portion of 
their surface, but most Mittdanty near the septa, to numerous 
small elliptical sporules, which generally form chains consisting 
of two or three cells by acropetal growth. The sporules soon fall 
the parent mycelium, and continue to reproduce 
themselves by ecd soon rendering the nutrient solution 
turbid by their immense n 8 
The product of REDE just described corresponds to what 
has been described as Dematium pullulans, well known as 4 
uat in the life-cycle of Cladosporium. 
he Dematium sporules or conidia readily ge the disease 
e placed on a wounded surface of the hos 
Fragments of the sporophores of CISD, when placed in 
water, also give origin to the Dematium form of reproduction. 
