164 The Philippine Journal of Science 1915 
feeces for inoculation, pressing it out of the abdomen and diluting 
with broth as the directions recommended. For convenience in 
reference we have designated this series as 30x (Table I). 
We early found that our inoculated insects did not show the 
one characteristic described—the liquid excrement. This may 
have been present in a few cases, but was decidedly rare. Since 
we often failed to find feeces with the abundant actively motile 
bacilli as described in the directions, we did not always dilute so 
highly with broth as the author of the directions recommends. 
It was found that if we gave too light a dose, even after a con- 
siderable number of insect transfers, the death of the insects was 
long delayed. The volume of the dose as given in the directions, 
“two or three drops,” is rather indefinite, and we may have 
averaged a somewhat smaller volume of material. But since 
the effective dose depends on the number of viable bacteria, this 
ought not to make a material difference, especially in view of 
the fact that increase of virulence of the bacteria should go 
on as well with small doses as with large, provided only that 
enough is given to cause a fatal infection. 
On account of the inconstancy of the number and character 
of the bacteria in the feces, we undertook a second exaltation 
series carried out in another way. Here any material from the 
gut was carefully avoided, and the bacilli were taken wholly from 
the body cavity. This was accomplished in most inoculations 
by injecting broth into the body cavity of the dead or moribund 
insect, withdrawing it by means of the pipette, and using this 
liquid for inoculation—always after microscopical examination 
to assure us of the presence of motile organisms. This series, 
carried through 15 insect transfers, we have designated as 15x 
(Table IT). 
Later, when field experiments with 30x and 15x had given no 
practical results, a third exaltation series was carried out. In 
this series the inoculation material consisted of fluid pressed with 
aseptic precautions from the leg of a dead or moribund insect. 
When full of actively motile coccoid forms, as was usually the 
case, this liquid was diluted with sterile broth and used for the 
next set. Wingless locusts in the later stage were used for the 
most part, and the series was carried to the twelfth insect 
transfer. This series we have designated 12x. 
In the first two series especially, we often divided the lot of 
insects to be inoculated at any one time into several sets, one 
set receiving material from the gut, another from the body cavity 
of the dead insect used as a source of material, or one lot was 
