THE MOSAIC DISEASE OF CUCURBITS. 29 
of such a dilution when pricked into the leaf of a healthy plant will 
produce the disease as rapidly and with as intense symptoms as 
when the undiluted juices are used. We have evidence here that the 
virus possesses the power of rapid increase, since all parts of the 
leaves and stem of the inoculated plant contain the virus within three 
to four days after inoculation. If we adopt the theory of an ultrami- 
croscopic organism as the causal factor, this increase is readily ex- 
plained, but examples of similar increase or multiplication are not 
commonly associated with enzyms. 
The behavior of the virus in filtration may be used to support the 
theory, although the fact that the finer filters hold back the virus 
indicates that the particles composing the virus are probably colloidal 
in nature and of relatively large size. These qualities, however, 
neither exclude it from the class of enzyms nor from that of ultra- 
microscopic organisms. 
In attempting to prove either of the above hypotheses it must be 
recognized that a great part of the available data is equally applicable 
to either theory and that no definite conclusion is as yet possible. 
It is the writer's opinion, however, that the causal agent possesses 
characteristics which tend to place it as an ultramicroscopic organ- 
ism rather than as an enzym, but no property has yet been discovered 
which may not be characteristic of either agency. 
On the other hand, the examples of diseases attributed to a so- 
called "filterable virus" or "ultramicroscopic organism" are numer- 
ous in animal pathology and are generally accepted, while as yet 
there are no demonstrated examples of diseases due to enzymic 
causes alone. In the case of tobacco mosaic the recent work of Allard 
(4) has nearly eliminated the oxidases from consideration, so that we 
at present can hardly attribute the disease to a definite type of 
enzym. 
INFECTIOUS NATURE OF THE DISEASE. 
The epiphytotic character of cucumber mosaic and its rapid spread 
to plants adjoining those first infected at once indicated that it was 
of an infectious nature, 'as proved by Jagger (17) and the writer (11). 
The first inoculation experiments by the writer were begun at Ham- 
ilton, Mich., in 1914, but as the plants used had no protection from 
insects and mosaic was already present in the field the results, although 
positive, did not permit definite conclusions. In 1915 the work was 
continued at Big Rapids, Mich. This district was practically free 
from the disease and the experimental fields had never before grown 
cucumbers. The possibility of outside infection, particularly from 
insects, was practically eliminated by using cages covered with cheese- 
cloth to protect all plants in the inoculation experiments. 
As the cages in most cases had no openings, they were lifted when 
inoculating the plants, but by exercising proper care and using the 
