ANDREWS: PROTOPLASMIC STREAMING IN MUCOR 463 
on the tubes and make it water-tight. In case there should be 
vertical joints the same scheme as shown by Fic. 3 could again 
be used, if the lower end is closed by a split cork as before and the 
opening D closed by a rectangular piece of rubber or cork. Any 
leakage in a joint so surrounded with water would, of course, be 
easily detected, while the apparatus has the advantage of being 
quickly and easily changed from one connection to another. 
n some cases metal gas chambers of the Engelmann type 
were used for the investigations, but in some others, where hydro- 
gen especially was not used, the ring form of gas chamber was 
a. ¢ 2 cA 
id 
-—“«e 
: 
i 
— 
“a: 
SN gq 
Fic. 3. Method of testing for leaks in glass tube connections under water. 
employed. These were fastened to the slide with sealing wax 
or paraffin. A still better way is to use the glass ring form of gas 
chamber having a continuous glass base of its own. 
In those cases where it was not necessary to pass gas of any 
kind over the specimens under investigation, simply glass rings 
Were used and these were cemented to the slide by sealing wax 
fr paraffin as above described. The cover glass to which the 
medium containing the specimens hung was made gas-tight by 
cementing it to the gas chamber, of whatever form used, by means 
of vaselin or a mixture composed of vaselin and wax as indi- 
fated by Pfeffer.* In some cases, as mentioned by Pfeffer, 
shellac of the proper thickness may be used to advantage especially 
™ experiments requiring some considerable time. The gas chamber 
and apparatus were always thoroughly freed from any gas foreign 
0 the investigation before beginning the experiments. 
* 
Pfeffer, W. Plant Physiology 3: 239. 1905. 
