50 THE CROWN-GALL AND HAIRY-ROOT OF THE APPLE TREE. 
These experiments prove that soft crown-gall of the apple is com- 
municable to some extent in wounds by inoculating with .pieces of 
living galls, but that pieces of hard crown-galls and the simple 
hairy-root may not readily communicate the disease under similar 
conditions. 
Increase of the disease in root-grafted trees due to the inoculation of 
grafts.—In order to test the communicability of the soft form of apple 
crown-gall, a set of nursery experiments was conducted. The grafts 
were prepared from healthy scions from the varieties Red June, 
Rambo, York Imperial, and Minkler, grafted on the roots of healthy 
selected seedlings from trees grown from the same seed in the same 
field. The grafts:were prepared in three lots: The first as ordinary 
tongue or whip grafts; the second lot had a piece of clean cross section 
of healthy apple seedling root inserted beneath the lower end or 
tongue of the scion before wrapping each graft; the third had a piece 
of living apple crown-gall of the soft form taken from the outside of 
the gall and inserted under the lower end of the scion in each graft 
before wrapping it. Each of the three lots of grafts was separated 
into eight equal sets, which were planted and grown in three experi- 
ments (1, 2, and 3) in the same localities and which were dug in the 
‘same manner as the grafts in the experiments whose results are given 
on page 62. The results of the experiment are given in Table XXIV, 
in the appendix. 
Briefly summarized, the results from the experiments are as follows: 
Where a piece of living healthy root chip had been inserted in the 
eraft union it caused an increase of crown-gall amounting to from 1 
to 14.2 per cent and where a piece of tiving apple crown-gall had been 
inserted there was an increase of crown-gall amounting to from 2 to 
33.8 per cent as compared with the amount of crown-gall present in 
trees grown under similar conditions from ordinary root grafts. In 
the former case the average increase was 4.7 per cent and in the latter 
12.6 per cent. The resulting disturbance did not vary the amount 
of hairy-root present in either case. 
On the whole, this set of experiments shows that crown-gall is com- 
municable in root grafts if they have an opportunity to become 
infected at the time the healing processes are taking place in the 
union; or, in other words, at the time callus is formed abundantly in 
the wounds. 
The data in experiment 1 indicate not only that there is no further 
spread of crown-gal! in the nursery rows the second and third years, 
but that there is an actual decrease of the disease due to the recovery 
of many trees. This must have taken place, since in experiment 1 
every third tree was dug with spades the first year, the second year 
one-half of those remaining also with spades, and the balance of the 
186 
