64 Lritish Uredinee and Ustilaginee. 
spores.” The spore-formation itself is similar to that of 
Ustilago,* and, commencing above, proceeds downwards. 
Sorosporium,— The spore-formation in Sorosporium 
differs considerably from that which has just been described 
in Ustilago, although it is obviously of the same type. In 
Ustilago the mature spores are separate and distinct, form- 
ing usually a pulverulent mass. In Sorosporium, on the 
other hand, they are in their perfect state aggregated into 
spore-balls, which individually often contain fifty or a 
hundred separate spores. The process of spore-formation 
has been studied by Von Waldheim with S. sapgonaria, 
and is as follows :—The mycelium, which is very abundant 
in the blossom and ovary, rapidly changes into spore-form- 
ing hyphe, from 4 to 74 in diameter, which are gelatinous 
and full of shining protoplasm. The free ends of these 
hyphe have a tendency to curve inwards and roll them- 
selves up (Plate V. Fig. 9). The spore-forming hyphe 
from several contiguous mycelial branches, incline together, 
and twist themselves into a ball, as happens in the forma- 
tion of a lichen thallus. These convoluted and contorted 
spore-forming hyphe, being gelatinous, soon become so 
intertwined and entangled that they cease to be indi- 
vidually recognizable; to all appearances they coalesce 
together in part, if not entirely, and on the exterior of this 
gelatinous ball other hyphz are now seen encircling it 
(Plate V. Figs. 10,11), These latter, also being gelatinous, 
soon lose their individuality, although at times traces of 
their concentric arrangement can be made out. Spore- 
formation takes place only in the central gelatinous ball, 
in the middle of which it commences by the central part 
darkening in colour and becoming differentiated into 
spore-like bodies, which vary in number from four to 
sixteen. Apparently these bodies again subdivide, so 
* De Bary, “‘ Vergleichung,” p. 187, 
