114 BULLETIN 903, U. S. DEPARTMENT OF AGRICULTURE. 
one of the vines was artificially inoculated. The second vine was 1 
foot distant, and the third 2 feet distant from the first. On October 
10, all three vines were dug up, and it was found that the first had a 
small infestation all over the root system. The second vine had a 
small infestation chiefly on nodosities on roots nearest those of the 
first vine. So far as could be observed, the roots of the two vines did 
not approach nearer than 2 inches at the closest point, but as some of 
the terminal rootlets had died during the summer and autumn it is 
quite possible that earlier in the season rootlets of the two vines were 
contiguous. The third vine was uninfested. Its roots had been sepa- 
rated from those of the first vine by at least 12 inches and from those 
of the second vine by at least 5 inches. 
This experiment did not appear to show that subterranean infesta- 
tion was a common mode of diffusion. The condition of the roots on 
the first vine when it was pulled up showed that its summer infesta- 
tion had been large and that many wanderers had been produced; 
therefore, one would expect that some of these would have found their 
way to both of the other two vines. The earth at the time of plant- 
ing, however, had been jmcked very solidly, and the layer of sand pre- 
vented cracking so that there were very few, if any, subterranean 
passages affording access to the phylloxeras. 
The following experiments also were made : On May 22 two young 
viniferse (Feher Szagos) were planted in a galvanized tin, 8 by 8 by 
10 inches. Two sides of this tin were basally produced in the shape 
of a cone (PL VI, fig. 2, p. 52) , and at each apex was a hole of one- 
half inch diameter. The cones were then tightly fitted into wooden 
tubes, through the centers of which ran a square passageway of one- 
half inch diameter, and the junctions cemented. The cones and 
wooden tubes were buried 8 inches below the soil surface. At the 
farther ends of the two wooden tubes similar galvanized tins were 
connected, and in each of them was planted a single sound vine 
(Feher Szagos). In this experiment the tubes were, respectively, 2 
and 10 feet in length. The conical projections were expected to draw 
the roots toward the hole, therefore toward the tubes. No earth, 
except for about 2 inches at the ends, was placed in the passage in the 
tubes. Black paper was glued on to the top of the outside of the 
wooden tubes so as to prevent entrance of light. Thus the phylloxerae, 
if they passed through the hollow inside of the tube, would not be 
influenced by any light rays. On September 23 the tins and tube were 
pulled up and the vines examined. Both central vines inoculated in 
May were well infested. Their rootlets and those of the two end vines 
had penetrated not more than 3 inches into the hollow of the tube, 
but in all four cases rootlets were abundant inside the conical projec- 
tions. The vine at the end of the 2-foot tube was well infested with 
