E. K Atkin and A. Bacot 
487 
medium for the organism, or more probably to tbe clearing action of 
tbe larvae. In other cases when infection was suspected owing to the 
progress of larvae in comparison with their fellows which had hatched 
from eggs of the same batch in other tubes, it was only proved when 
cultivations on agar were made with quantities of 0-25 to 0-5 c.c. and 
the tubes were incubated at 60° to 75° F. It is further significant that 
larvae from the same batch of eggs might in one tube make no progress 
at all while in a similar tube perceptible progress would take place, 
only to cease before the third instar was attained, raising a suspicion 
of the presence of an infecting organism which either died out or was 
killed off by the larvae, which subsequently starved. These facts 
suggested that when eggs hatch the conditions are seldom if ever really 
sterile, but that the organism responsible for the stimulus is unable to 
colonize the tube owing to small numbers or want of adjustment 
between itself and the conditions. On the basis of this suggestion the 
difference between hatching and no growth on the part of the larvae and 
slight growth preceding death would be explained by the yeast or 
bacterial infection, probably due to infection from the eggs, dying out 
immediately in the former instance and more gradually in the latter, 
possibly owing to the action of the larvae. 
Nevertheless, in spite of our frequently correct suspicion, that larval 
growth indicated infection, there were a few cases in which adult insects 
were reared under sterile conditions—the final tests being to transfer 
the dead mosquito from the tube in which it had been reared and died 
to a tube of peptone broth in which it was allowed to remain for a week 
or more under a variety of conditions of temperature. In these cases 
only one or two adults resulted from several larvae, and the possibility 
remains that the real source of nutriment was obtained from their dead 
comrades. This last surmise is, however, at issue with the fact that 
the presence of dead larvae did not in most cases enable the living 
survivors to complete their development. 
We desire to take this opportunity of recording our thanks and 
indebtedness to our colleagues on the Staff of the Lister Institute for 
their valuable suggestions and advice, especially to Dr Harriette Chick 
and Professor Harden, without whose assistance many of the experi¬ 
ments could not have been carried out. 
