THE MOSAIC DISEASE OF CUCUKBITS. 39 
it does when inoculation is made in the stem tissues. The fact 
that the pores of a Chamberland filter can hold back the particles 
of infectious material also leads to a belief that such particles may 
be colloidal in nature and not likely to pass readily through the 
^ cell membranes. 
Allard (5) has shown in the inoculation of older leaves of tobacco 
plants that severing the lateral veins from the midrib or cutting the 
base of the midrib itself does not appreciably increase the time 
usually required for the virus to reach the stem and pass to the young 
leaves. He believes that in such cases " multiplication and diffusion 
of the virus from cell to cell, aided perhaps by the fine anastomosing 
lateral veins, would sooner or later allow the virus to reach the 
petiole and pass to the rest of the plant." The writer is inclined to 
believe from the results of work with cucumber aphids that conduc- 
tion through the fine veins is a more important means of transference 
I than diffusion. These insects apparently attack only the small veins 
of the leaf, as shown both by observation and by stained microtome 
sections of aphids killed and embedded while still attached to the leaf. 
In all such sections the sucking apparatus of the insect extended 
into the small leaf veins. Further support of this fact is given by 
Woods (31) in his work on stigmonose of carnation, in which he 
shows drawings and photomicrographs which indicate that the aphis 
attacks the vascular elements. Since aphids are the most consist- 
ently successful agency of inoculation for cucurbit mosaic yet dis- 
covered, it seems likely that their high percentage of infection is due 
to the introduction of the virus at a point where it is most rapidly 
carried to all parts of the plant. Stem inoculations made by cutting 
off a petiole and pricking the juices of mosaic cucumber plants into 
* the region about the bundles of the leaf trace give about double the 
percentage of infection that occurs when the virus is pricked into 
the stem at random, indicating again that the vascular system is 
concerned. 
The portion of the vascular system especially concerned is still 
rather indefinite, but the work with aphids again furnishes a clue. 
Woods's work before mentioned (31) indicated that the soft bast 
parenchyma cells of the phloem were the ones particularly attacked. 
Other observers have also recorded the same facts, and the writer's 
work along this line, although not extensive, indicates that the 
cucumber aphis punctures the cells of the phloem. If this is true 
* it may be possible that the phloem is at least as important as the 
xylem in the distribution of the virus, since the inoculation work with 
aphids is so universally successful. Beijerinck (6) is of the opinion 
that the phloem is the means of distribution of tobacco mosaic, 
since inoculation of the older leaves first produces symptoms in the 
young leaves, and he therefore concludes that the virus is carried in 
the descending sap flow. 
