486 Mr. A. S. Olliff on Australian 
some of the larve with which we are acquainted. By com- 
paring these with the one similarly affected which we obtained 
here, and also with the delineation on the plate of the living 
larva, together with the appended observations of several 
writers, a fair conclusion can be arrived at respecting the 
genus of moth, which, in its two preparatory stages, is liable 
to the fatal attacks of this fungus. In order, therefore, to 
carry out the necessary comparison, we copy a lignified larva 
found at Barrabool Hills, Victoria, and the well-known New- 
Zealand species. To these we have added sketches of one 
obtained near Sydney by Mr. Shepherd, and another by our- 
selves at the Hunter River.’ Our author then proceeds to 
summarize the observations of Mr. W. H. Hawkes on Cordy- 
ceps Hawkesii found at Launceston, Tasmania; and the 
observations of Mr. John Allen and the Rev. R. Taylor on 
Cordyceps Taylori from Yass, New South Wales. In each 
case he concludes that the host is the caterpillar of a species 
of Pielus, and, after noting that in the case of the New Zealand 
form Cordyceps larvarum, the host has been incorrectly iden- 
tified with the larve of Charagia by Hooker, Dieffenbach, 
Doubleday, and 'l'aylor, he adds:— We think it probable 
that the stems and trunks of the Metrosideros furnish sus- 
tenance for the larve of the Charagia virescens ; but these 
live and undergo their metamorphoses within the wood, effec- 
tually protected against injury from this particular fungus ; 
and it is equally probable that the eternal portions of the 
finer roots of the same or neighbouring plants afford nutri- 
ment to the larve of such genera as pass their lives wholly 
in the earth, a state of existence which would render them 
exposed to the attacks of the Spharia (Cordyceps).” 
In my opinion we have in these remarks the truth of the 
matter, and I am inclined to go even further, and to assert 
that all the larger fungi of the genus Cordyceps live upon and 
at the expense of subterranean larvee and pups. In support 
of this assertion | would point to the fact that all the bulky 
species of which the hosts are definitely known have been 
found on root-feeding insects. As instances I need only cite 
the Dynastide, Melolonthide, Mlateride, and Lucanide 
amongst the beetles, Cicada amongst the Homoptera, and 
Piclus and Trictena amongst the moths. In all these cases 
the hosts are subterranean, and it follows that it is idle to 
speak of any connection between these parasitic fungi and the 
larvee of wood-boring or foliage-eating and free-living insects. 
‘The best-known and the most abundant species are found on 
the early stages—larve and pupa—of Lepidoderma, Lepidiota, 
