CHAPTER VII.—PHENOMENA OF VEGETATION,—-PARASITES. 373 
The gonidiophores of these plants on most of the caterpillar-bodies examined, 
which also bore stromata, were observed to be small hyphal branches with whorls 
of branchlets like the germ-plants just described (Fig. 165 £), and these together 
formed a delicate down on the surface. But on other insects they grow into a 
dense mould-like covering some millimetres in height and white with a dust of 
countless gonidia, or else like the Coremium-form of Penicillium they form club- 
shaped Fungus-bodies ı-2 centimetres in height and covered all over or in the upper 
part, which is borne on an orange-yellow stalk, with a felt of branchlets which abjoint 
gonidia. The last-named bodies are known as form-species under the name of 
Isaria farinosa. Both the Isaria-form and mould-covering are found commonly on 
a sclerotioid insect-body by themselves, i.e. without the stromata. I once succeeded 
in obtaining two poorly developed stromata with some large Isarieae from a 
caterpillar of Spinx Euphorbiae infected by ascospores, which had changed to the 
chrysalis state after infection. 
Gonidiophores with the round aerial gonidia are also obtained if cylinder-gonidia 
from the still living insect or portions of the mycelium from the insect converted into 
a sclerotium are cultivated in the air on a suitable substratum; the amount of 
luxuriance with which they are developed varies with the supply of food. They are 
formed too under the same circumstances on small plants, which proceed from germ- 
tubes produced at once in a fluid from the aerial gonidia themselves. These germ- 
tubes do not ultimately penetrate into the skin of the insect, at least not in the 
experiments made with the caterpillar of Sphinx Euphorbiae. When the insects are 
sprinkled with the spores the germ-tubes are seen to enter the tracheae through the 
stigmata, and then to bore through the wall of the tracheae and so reach the cavity 
of the body, where the cylinder-gonidia are then abjointed and disseminated, and 
sprout; at length the insect dies and becomes a sclerotium, exactly in the way 
described above in the case of the direct products of the ascospores. Stromata 
have never been known to be formed on insects killed by infection with aerial 
gonidia, only a fresh crop of aerial gonidia especially of the Isaria-form. 
The Fungus therefore which we are describing can only arrive at its full develop- 
ment, that is, can only form perithecia, as an obligate parasite, and to this mode of 
life it is closely adapted. When it ceases to be a parasite and the dead insect 
becomes a sclerotium, we see a saprophytic stage of the existence of the plant follow 
upon the strictly parasitic stage. When the conditions for a parasitic mode of life are 
withdrawn, facultative saprophytism takes its place (see page 356), and a new form 
of adaptation may make its appearance with the formation of aerial gonidia accom- 
panied also with the normal formation of perithecia. The entrance upon the 
parasitic mode of life is however comparatively easy, because the formation of all kinds 
of spores takes place abundantly in nature, especially in wooded places frequented by 
the insect-hosts. All that is known of the mode of life of other insect-killing Fungi 
allied to Cordyceps militaris agrees entirely with the account here given of that 
species. 
The foregoing account of the life-history of Cordyceps is founded chiefly on the 
facts observed on infecting the caterpillars of Sphinx Euphorbiae with the Fungus, 
and on investigations some of which have not hitherto been published. I make this 
latter remark, because the statements agree so exactly with the account given by 
