Notes on Some South African EntomopJithoraceae. 
81 
means of its legs, on August 8. The above determination is based entirely 
on the shape and size of the conidia, which agree very closely with Thaxter's 
description and figure. Until the resting spores are observed, however, the 
determination must be regarded as somewhat doubtful, especially as Thaxter 
states that his specimens (Tipulid larvae and imagines) were found floating 
on water. 
Umpusa grylli, Fresenius. (Plate III, figs. 7-12.) 
Conidia ovoid to pear-shaped, 25-45 /x x 20-35 /^, average about 
35 X 28 fx, containing one or more large oil-globules. Conidiophores simple, 
club-shaped. Secondary conidia like the primary and produced by direct 
budding. Kesting spores spherical, colourless, very regular in shape and 
size, 30 jjL in diameter, with thick, double, hyaline walls. Host attached to 
tips of grass, etc., by the contraction of its legs. 
Hosts. — Orthoptera (according to Thaxter, also Lejndoj^tera and Diptera). 
Habitat. — South Africa, U.S.A., Europe. 
This disease was first noticed on grasshoppers on January 10 at Cedara. 
It was exceedingly common from this date until the end of March, attacking 
impartially several different species of Acridiids. 
According to Sacharov (6) and Perez (7) this fungus only attacks spent 
adults, but at Cedara all stages from the second instar to the adult stage 
were found dead of the disease, and several dead females were found whose 
ovaries contained eggs. 
As was noticed in the case of E. muscae, all the infected individuals that 
were kept under observation died in the afternoon, the great majority dying 
between 5 and 7 p.m. Sluggish individuals could be found in the field 
clinging to grass-stems at 1 p.m. Earlier in the day no dying individuals 
could be found. This remarkable characteristic of the fungus is probably 
explained by the climatic conditions at Cedara. During the summer the 
mornings are usually warm and dry, but in the afternoon mists and thunder- 
storms came up, making the atmosphere very moist. 
During the period that this fungus was kept under observation about 
1 per cent, of the individuals that were found infected failed to throw off 
conidia, even when kept under favourable conditions. On dissection these 
individuals were found to contain numerous resting spores (fig. 12). The 
mode of formation of these spores was not observed. Attempts to germinate 
them in water all failed, although hyphal bodies and conidia germinate quite 
freely in drops of water. 
In view of Hesse's claims regarding E. muscae, it is interesting to recall 
the work of Edington and Black on the locust fungus, carried out at 
Grrahamstown over twenty years ago. These two authors cultivated a 
fungus on a large scale, and this was distributed to farmers as the locust 
fungus. The reports of the efficiency of this fungus as a means of control 
