THE MOSAIC DISEASE OF CUCURBITS. 
25 
filtration the juices were at once passed through a sterile normal 
Berkefeld bougie, the bougie, receiving flask, and all connections 
having been previously sterilized with steam for 45 minutes at 15 
pounds' pressure. 
Owing to the large amount of finely divided material which re- 
mains in suspension in the expressed juices, the filtration process is 
very slow and with the Berkefeld filter 12 to 14 hours are necessary 
to obtain 200 c. c. of filtrate from a bougie 1 by 6 inches. The 
filtrate was removed to sterile test tubes or small flasks by means of 
sterile pipettes and inoculation made from each lot to beef bouillon, 
Tubes which proved free from bacteria were used for inoculation, 
inoculations being made with the unfiltered juice at the same time. 
These were made by pricking the filtrate into the young leaves at 
several points and into a wound at the base of the stem. The 
results given in Table VII show that the filtration of the infectious 
juices did not destroy their power of infection, most of the filtrate- 
inoculated plants developing mosaic symptoms as rapidly as those 
inoculated with the unfiltered juices. 
Table VII. — Effect of filtration through a Berkefeld (normal) filter on the infectivity of 
the expressed juice of mosaic cucumber -plants. 
Date inocu- 
lated. 
Aug. 31,1915 
Do 
Do 
Sept. 6,1916 
Do 
Do 
May . 12, 1917 
Do 
Do 
May 18,1917 
Do 
Aug. 30,1917 
Do 
Do 
Treatment. 
Filtered mosaic juice 
Unfiltered mosaic juice 
Unfiltered healthy juice (control) . 
Filtered mosaic juice , 
Unfiltered mosaic juice 
Unfiltered healthy juice (control) . 
Filtered mosaic juice 
Unfiltered mosaic juice 
Unfiltered healthy juice (control) . 
Filtered mosaic juice 
Unfiltered mosaic juice 
Filtered mosaic juice 
Unfiltered mosaic juice 
Unfiltered healthy juice (control) . 
Number 
of plants 
inocu- 
lated. 
Results. 
Number 
ofmosaic 
plants. 
Date last 
observed. 
Sept. 15, 1915 
Do. 
Do. 
Sept. 14,1916 
Do. 
Do. 
May 21, 1917 
Do". 
May 31, 1917 
Sept. 12,1917 
Do. 
Do. 
Chamberland filter. —Lii contrast to the results with the Berkefeld 
filter, the Chamberland type of porcelain bougie has given filtrates 
that were noninfectious. A considerable number of inoculations 
have been made, using the filtrates from different types of Chamber- 
land filters, but no positive results have been obtained. 
The earlier work was done with filters of the "F" and "B" types 
about 1 by 8 inches in size, the "B" type being supposedly the finer. 
Later a set of smaller bougies was obtained which gave a gradation 
in porosity. These were five-eighths by 6 inches and were graded as 
L2, L3, L5, L7, L9, and Lll. The L2 and L3 grades were supposed 
to be permeable to the more minute forms of bacteria, and the other 
