272 JOURNAL OF THE ROYAL HORTICULTURAL SOCIETY. 
is twisted so as to bring an eye forward at the proper level, wherewith 
to develop the second pair of branches, and so on for further develop- 
ment. It may be asked, Why not allow the mid-stem to grow on 
naturally ? the reply to which is that the central stem must be checked 
to allow thorough development of the lower tiers before it is allowed to 
proceed, otherwise basal denudation is sure to occur ; the procedure 
also has the advantage of getting the more productive (as is claimed) 
stipulary wood for the continuation. I may add here that Lorette 
advises the use of the weaker or second stipulary shoot for obtaining 
the main stem of the simple cordon and also for the main branches of 
the dwarf distaffs or " fuseaux." It will be noted that in many of 
these forms the eventual support branches are " doubly stipularized," if 
one may coin the phrase, as for instance is the case with the " doubled 
U." It is important that the maiden tree should be thoroughly well 
rooted before commencing operations, and if, as I usually find, growth 
is not good in the spring following plantation, it is better to await the 
next April, or more time may be lost. At planting the maiden, Lorette 
advises a mere removal of the top few eyes : I am not sure whether it 
is not best to leave all alone. Where straight continuations are 
needed the pruning cuts should be made obliquely ; if only a U-form is 
being made the cut should be transverse for preference. In some 
cases M. Lorette was developing his stipularies from a lower eye, 
as if a continuation was wanted, but the upper part was to be cut 
away, and was merely retained as a sap drawer (" appel seve ") ; in 
the same way some shooting may be allowed below the chosen eye 
to draw sap for awhile ; these variations may be useful where growth 
is not strong enough, as I find often to be the case where maidens have 
not been planted for a good year. At any rate, in one case I have 
obtained direct growth of stipularies in a tree in its third year ; two 
sets broke without the central shoot. There is, of course, the risk that 
the eyes may have received damage without the protecting shoot base. 
In these procedures only pairs of branches can be obtained, but 
sometimes one wants more branches at or close about one level, as in 
a " staged pyramid," for the natural branches are always so separated 
that the upper ones will always cause trouble and delay in securing 
an equilibrium, which indeed is never final. Curiously enough, 
Lorette does not write of this in his book, and I cannot be sure in what 
author I found the procedure. The leading shoot is cut back to an 
eye at a level slightly below the point at which the branching is desired, 
the pruning cut being oblique to allow prolongation in line (fig. 41). 
When the shoot from the terminal eye is ready, and has a good basal 
cluster of leaves, it is cut back to one or two apparent eyes ; when 
these have shot well, a transverse cut is made to the basal cluster. 
The first cutting yields the basal cluster, the second throws nourish- 
ing sap into the latent eyes of the basal cluster leaves, and the third 
one causes them to shoot ; in this way one can obtain some five to 
seven shoots originating almost at one level, anyhow more so than 
can be done in any other way except by the insertion of grafts or 
