ANDREWS: PROTOPLASMIC STREAMING IN Mucor 479 
present. In those cases where a good many spores were grown 
under the same cover glass some of the filaments or their branches 
just entered or passed entirely through some of the air bubbles. 
In these, however, the air was saturated and no streaming oc- 
curred. Even when they had just grown into the bubble no 
streaming was seen. When the position of a saturated bubble 
under the cover glass was shifted so that it was placed over a fila- 
ment or a part of a filament that had previously been entirely sub- 
merged, no streaming of the protoplasm began. This shifting can 
be easily done by very slight pressure on the cover glass when it is 
nearly swimming in the solution which submerges the filaments. 
If this pressure is properly applied, a bubble which is a little 
distance from a filament can be caused to flow over it. At the 
Same time, since the cover glass does not rest directly on the 
filaments, the pressure is not sufficient in this case to cause any 
injury, as the experiments showed. If, however, a dry air bubble 
is gently brought under the cover glass directly in contact with 
the filaments by means of a glass capillary tube, streaming will 
be induced immediately and will continue for a time. A con- 
venient way of growing the filaments so as to prevent them from 
from being affected by pressure in moving a bubble around under 
the cover glass is to make a cell of a layer of filter paper and fill 
it with the solution and a few air bubbles. For the use of the 
capillary tube here mentioned the filter paper cell should be cut 
away on two sides to freely admit the tube. At first I forced 
air bubbles under the cover glass through a very fine capillary 
tube by means of an automobile pump. By the use of the auto- 
mobile pump directly connected it was difficult to regulate the 
°w and constancy of the air bubbles. To overcome this diffi- 
culty the air was forced directly into an ordinary autoclave, A, 
aS an air receiver, by the pump B and from this was conducted 
through the capillary glass tube to the specimen. A mercury 
Manometer, C, was connected with the apparatus to show the pres- 
sé more accurately than the autoclave gauge would do. From 
to } of an atmosphere was necessary to force the air quickly 
through the capillary glass tube D, according to the size used 
(Fic, 2a). By care, with this method a number of bubbles or a 
Stream of them may be forced in, and their number and size con- 
