[Vor. 8 
350 ANNALS OF THE MISSOURI BOTANICAL GARDEN 
precaution was taken with reference to contact with the clothing 
or hands. Between different inoculations the hands were washed 
with soap and water, then washed or rinsed with 1-500 formalde- 
hyde, which has been found an effective antiseptic for the purpose, 
although when added to the juice in this concentration it is 
relatively ineffective. 
Results of inoculation experiments.—There are given in table 
II the results of a series of inoculation experiments, with the 
filtered juices already described, conducted during November, 
1921. In accordance with the indications previously given the 
inoculations were made on plants about 3 months old, which had 
been grown under greenhouse conditions and at this stage were 
in 5-inch pots. Good growing conditions were maintained through- 
out the experiment, since it has been repeatedly shown in our 
work that such conditions are favorable for most rapid pro- 
duction of unmistakable symptoms of the mosaic disease. 
TABLE II 
INOCULATION EXPERIMENTS MADE ON HEALTHY ibn ot 
WITH FILTERED JUICES OBTAINED FROM PLANTS v dae 
THE CHARACTERISTIC MOSAIC DISEA 
Yos. Number ee A No. of plants 
à of à . with mosaic 
No plants inoculation after 18 days 
1 20 Filtrate, Mandler filter 19 
2 20 Filtrate, spherical atmometer cup 18 
3 20 Viltrate, cylindrical atmometer tube A 1 
4 20 Filtrate, cylindrical atmometer tube B 0 
5 20 Filtrate, Atm. C. infilt. with Al(OH); 0 
6 20 F'iltrate, Atm. D. infilt. with Al(OH); 1f 
7 20 Filtrate from celloidin membrane E it 
8 — Filtrate from celloidin membrane F* — 
9 20 Control, juice from diseased plant 19 
10 20 Control, distilled water 0 
* This membrane leaked and no inoculations were made. 
f These two plants exhibited pronounced symptoms of mosaic in so short a time 
after inoculation that they are thought to have been accidental contaminations. 
From the results obtained it was clear that particular interest 
would attach to the spherical atmometer cup and to the filter 
