1906] OLIVE—DEVELOPMENT OF EMPUSA 197 
that the fat cells, detached in the search for the fungus, must have 
been mistaken by some for stages in the development of Empusa; since 
these bodies may frequently resemble closely short hyphal segments in 
their fatty, granular contents, as well as in their assumption of globular 
or irregular shapes, which give the suggestion of budding cells. 
These detached fat-bodies, which prove in sections to be small aggre- 
gates of insect cells, are particularly abundant in the adult just pre- 
ceding ovipositing. An easy and certain method of distinguishing 
the vegetative cells of Empusa from fat-bodies which oceur with them 
in the body cavity, is to stain with a dilute methyl-green solution, 
acidulated with a few drops of acetic acid, when the fungus cells 
stand out conspicuously, distinguished by their relatively large and 
characteristic nuclei. 
In some preliminary observations on the manner of infection of 
Empusa sciarae, I have not been able to make certain of this point, 
but I wish to record here some notes of interest pertaining to it. The 
small fly, Sciara sp., was sometimes accompanied, in the vessels in 
which the dung cultures were kept, by three other species of small 
flies, which hatched generally in less abundance, Psychoda sp., a so- 
called moth fly, and two other undetermined forms. None, however, 
other than Sciara were infected, although many times I have noted 
conidia stuck on the surface of the bodies of the larvae as well as of the 
adults of the other species. Successive generations of infected Sciara 
larvae as well as adults have appeared with great regularity every 
month or six weeks. The fly lays its eggs on the surface of the dung 
or on the sides of the vessels, and the young larvae, soon after hatch- 
ing, crawl below the surface of the substratum. It would appear 
reasonable to assume that in this case infection would occur with ease 
in this very young condition, when the larval skin is thin and deli- 
cate, and before they had crawled below the surface cf the dung, 
where they would obviously not be reached by shooting conidia. 
My preliminary unsuccessful experiments at infecting healthy adult 
larvae, confined for a week in bottles with diseased ones, at least 
suggest the possibility of infection occurring in young insects only. 
If the cultures in which young larvae were being nourished were 
kept quite moist, perhaps the majority of individuals in this stage 
were killed by the disease. When, however, the conditions were 
