572 DISEASE IN WILD MAMMALS AND BIRDS 
time, but gaming entrance to the animal's body via 
wounds made by sharp sticks or the beards of grain. 
While circumstantial evidence offers some support to 
this general idea, it cannot be accepted as proven. I 
have not made studies of the feed for the purpose of 
isolation of the streptothrix but cultures from the lips and 
gingival margin of healthy and infected animals were 
made for its cultivation. These attempts were fruitless, 
and without wonder since the germ when isolated from 
a fully developed case is quite finical in its manner of 
groA\i;h; there are so many kinds of bacterial life that 
they may easily overgrow the one in quest. So, too, cul- 
tures made directly from subcutaneous necrotic areas may 
not always give a positive growth although smears from 
the same material may reveal numerous threads under 
the microscope. 
The idea that sharp grasses are responsible for the 
origin of Kangaroo disease finds a protagonist in Dr. A. 
S. LeSouef, Director of the Zoological Garden at Sidney, 
Australia, a gentleman whose judgment carries weight. 
He \vrites : "We have found that it is entirely due to get- 
ting spear or barley grass in their food; o%\ang to the 
formation of the mouth, tliis grass gets wedged in be- 
tween the teeth and the cheeks, penetrates the flesh and 
allows the bacteria to get a footing, this in time heaps up 
on the inside and forms an abscess that bursts exteriorly. 
Formerly all the Australian Zoos lost animals through 
this cause, but now, through being very careful not to 
give any rough spined grasses, we never have the 
trouble." Since recei\'ing this letter we have re- 
moved straw bedding, and feed only soft alfalfa which is 
carefully inspected for foreign substances. During this 
time we have had two cases but the period of observation 
is too short for final judgment as to the value of feeding 
grasses mthout sharp beards and spines. The appear- 
ances of cases sporadically without reference to the arri- 
val of new specimens, the low communicability of the 
