The Heniibasidii. 
359 
sometimes exclusively in the latter. Several species produce the 
chlamydospores in the tissues of the leaves or of the axis of the 
host-plant. 
A peculiar phenomenon is observed in the dioecious flowers of 
Lychnis dioica and L. alba when they are attacked by Ustilago 
Melandryi [ = U. violacea (Pers.) Tul.]. The anthers in the 
staminate flowers are filled with the blackish-violet chlamydo¬ 
spores. The pistillate flowers normally possess no stamens, but 
on plants infected by the parasite stamens are developed in these 
flowers and the anthers become filled with the spores. Thus, 
under the influence of the fungus the female flowers again became 
hermaphrodite. 
Chapter II. deals with the perpetuation of the Hemibasidii 
both externally and within the host-plant. The chlamydospores 
are typical resting-spores, and by means of their thick, cuticularized 
cell-walls are well adapted to withstand unfavourable external 
circumstances. In fact the spores of many species will germinate 
only at the end of a long resting-period. The spores of several 
species were found to retain their viability for at least ten years, 
while the spores of other species show immediate germination but 
soon lose their vitality. 
The perennation of the fungus frequently takes place within 
the tissues of the host-plant, principally in the subterranean axes, 
from which it passes each year into the growing points of the 
subaerial shoots. The mycelium may be best observed in the 
parenchymatous tissues of the nodes, and in the older nodes it is 
doomed to remain sterile. But if the top of the plant containing 
the fungus in the tissues of the growing-point be decapitated, the 
mycelium in the nodes then penetrates into the growing-points of 
the new shoots produced in the axils of the lower leaves, which 
take the place of the main axis. 
A long-extended series of observations shows that, while in 
slow-growing plants, e.g., Primula officinalis, the duration of the 
fungus in the annual shoots is assured year after year, the hyphae 
are unable to keep pace with the extension in length of the shoots 
of plants of rapid growth, so that the fungus disappears after a certain 
number of years, and the plant finally becomes free from the parasite. 
In Chapter III. Brefeld brings forward good reasons for 
removing from the Hemibasidii certain forms placed in this group 
on account of a superficial resemblance in the production of their 
chlamydospores, because the latter show striking differences in 
their germination from anything hitherto observed in the Hemi¬ 
basidii. These doubtful forms are comprised in the genus Genii- 
mil a (== Schrceterid) and among the root-parasites known as 
Entorrhiza, which are transferred to the Fungi Imperfecti. Also 
Ustilago virens (= Tilletia oryzce Patouillard) when cultivated 
produced sclerotia, which on germination gave rise to ascocarps, 
very similar to those of Claviceps. On this account it has been 
assigned to the genus Ustilaginoidea belonging to the Hypocreaceae. 
Chapter V. is devoted to a consideration of the relationship of 
the Hemibasidii to the Basidiomycetes proper. An interesting new 
species, Heptasporium gracile, is described, the hyphae of which 
show clamp-connections at all the septa, and bear numerous 
isolated, scattered basidia each producing five to eight, but most 
commonly seven, basidiospores. No other form of spore was 
