[Vor. 8 
348 ANNALS OF THE MISSOURI BOTANICAL GARDEN 
for a short while, and allowed to dry over night at laboratory 
temperature. Since a very permeable membrane was desired, 
the film was put into 96 per cent ethyl alcohol for 24 hours at 
20° C. and then thoroughly washed in water for a day. The 
films were cut into sizes large enough to fit over the broad end 
of a thistle tube. Tests of these membranes for leakage by the 
air bubble method were concurrent with the filtration experi- 
ments. 
Preparation of the juice from diseased leaves.—A simple standard 
method, long in use in this laboratory for preparing the infected 
juice to be employed in experimental work, was adopted. This 
consists in pulping a known weight of the diseased leaves in a 
large mortar with a heavy pestle, then adding an equal weight of 
water and continuing the pulping until the leaf tissue is thoroughly 
crushed. The material is then filtered through cotton on a 
Buchner or ridged funnel. This diluted juice is used directly 
in the inoculation experiments. 
Filtration of the juices.—In these experiments it was necessary 
to use every precaution possible to prevent accidental contamina- 
tion of surfaces or vessels that might come in contact with the 
filtered juice. It was soon found that this could best be done by 
lowering the wet filter into the vessel containing the diseased 
juice to a suitable depth and then drawing the filtrate into the 
tube, rather than to draw the current from within outward. 
By the method indicated, as soon as sufficient filtrate had been 
drawn into the filter cup or tube, the filtration was stopped, 
and with sterile pipettes a quantity of the clear filtered juice 
was taken from within the cup and placed in clean vessels for 
use in inoculation. 
With the various porcelain filters the water pump reduced the 
air pressure to 1/15-1/30 of an atmosphere. The filtrate was 
rapidly obtained in the case of the Mandler filter and also very 
nearly so rapidly in the case of the spherical atmometer cup. 
In fact, the time required to obtain a sufficient amount of the 
filtered juice was about 15 minutes with the spherical atmometer 
cup, and 30-45 minutes with the cylindrical ones. According 
to all the evidence at present available, such differences in 
pressure as were used do not materially influence the size of the 
