a Fungus causing a Disease of the Sugar-cane . 5 1 7 
of which a nucleus can in many instances be demonstrated, is 
not understood, but in the case of a segmented spore all the 
germ-tubes usually grow at about the same rate and form 
equally vigorous hyphae ; whereas in unicellular spores several 
germ-tubes are often produced that grow at an almost equal 
rate for some time, one finally continuing to grow, whereas 
the remainder become arrested and die after having apparently 
expended so much energy for no specific purpose. In Oedoce- 
phalum roseum , Cke., the spores are large, unicellular, and 
contain a well-defined nucleus which disappears at the com- 
mencement of germination ; several germ-tubes are produced, 
one of which continues to grow, the remainder soon dying : 
in the persistent hypha a nucleus can be readily demonstrated, 
whereas no such structure can be shown to exist in the hyphae 
or germ-tubes that perish prematurely. Is the presence of 
one or more nuclei necessary for the continuous growth of 
a germ-tube ? 
The conidia generally form two germ-tubes, one from each 
localized portion of protoplasm, which are about 2 ^ thick, 
aseptate, at least for a considerable time, and soon give off 
lateral branches of equal diameter to the parent hypha 
(Fig. 5 )- 
In flask-cultures the conidia behave as already described, 
but development proceeds much further. 
Four conidia were placed in each of three flasks containing 
an equal quantity of sugar-cane solution ; the flasks were 
placed on a shelf fixed over a hot-water coil in the laboratory 
the temperature averaging 7 5 0 F., the extremes being within 
5 0 of that point. At the end of five days the liquid in each of 
the flasks presented an opalescent appearance, which examina- 
tion of the contents proved to be due to very delicate, much 
branched hyphae equally diffused throughout the liquid. The 
nutrient solution had a slightly acid reaction, and the addition 
of potassic hydrate, amounting to one per cent, for the whole 
solution, completely checked further growth of the mycelium. 
Examination of the contents of a second flask after eight 
days’ growth showed numerous filaments of mycelium measur- 
M m 2 
