390 JOURNAL OF THE ROYAL HORTICULTURAL SOCIETY. 
season Type V. stood out as possessing a very promising root system 
in its first year. It rooted very readily and the individual roots were 
of considerable length and well developed. The two following summers 
(1915 and 1916) were considerably more rainy, and Type V. showed 
a marked falling off in its rooting quality, both in freedom and develop- 
ment. Exactly the reverse appeared to be the case of Type VIII. 
(French), the rooting habit of which undoubtedly showed a slight 
improvement during 1915 and 1916. The latter case may,. of course, 
partly be accounted for by the fact that the stools had become more 
established. We also learned with certain types that it was easy to 
break the layers away from the parent stool, a method which gave 
more root system to all those types which tended to root more freely 
at the base and on a heel of two-year wood. Previously we had cut 
each layer from the parent as close to the stool as was conveniently 
possible. In view of the future condition of the stool and the obtain- 
ing of fresh annual shoots, the latter method seems most to recommend 
itself generally. Type IX. has so far only been tested during the two 
moister seasons, so that it is impossible to speak very definitely of its 
rooting characteristics, but it appeared to root more freely during last 
summer (1916) when we experienced the heaviest rainfall. 
The question as to whether differing soil conditions affect this early 
adventitious rooting is now being tested, but in view of the fact that 
the French choose the Doucin, the French Paradise, and the Jaune de 
Metz as free rooting types, this assumption appears true, since these 
types root least well at East Mailing.* On the other hand, com- 
plaints have been received from a nursery in another part of England 
that the bedded Doucin stocks root so poorly that the raising of them 
was likely to be discontinued. I have also tried to root Doucin layers 
on my own soil — which is considerably heavier than that at East 
Mailing — but the results were almost identical. The fact that so 
many of the Doucin stocks in common use all over the country are 
bought in from France is sufficient to account for its predominance in 
our nurseries, despite its shy rooting habit here. 
B. The Four -Year Old Roots. — How far early-rooting powers in- 
dicate future promise has to be considered. Sufficient data have not 
yet been collected to enable us to follow closely the life-history of the 
roots at this early stage and so differentiate their future functions and 
character. However, the examination of numerous roots of each type 
at later stages of development seems to show that in the main it is 
possible to draw safely certain inferences from the early habit. 
Types I. (Broad-leaf) and VI. (Nonsuch), for instance, which in 
their earliest stages exhibit sturdiness together with an ample root 
system, continue to develop along these lines. The early rooting of 
the Nonsuch is distinctly coarser than that of the Broad-leaf, and this 
characteristic is maintained. Type II. (Doucin) , as would be expected, 
* I have also heard it asserted that the True Broad -leaved English — one of 
our most freely rooting types (Type I.) — will not root successfully in the French 
nurseries, but I cannot state this authoritatively. 
