380 JOURNAL OF THE ROYAL HORTICULTURAL SOCIETY. 
for a recognized type. Thus for a period of three and in some cases 
four years every stool of each type has annually had to stand the test 
of the type description. In the main the characters of the dormant 
period have been constant. Whilst the actual shade of wood 
colour may vary slightly from year to year according to seasonal 
conditions, yet the relative tones appear to remain constant, and are 
generally a good guide. Buds show very little variation on " one year " 
wood, except possibly in their angle of insertion on the stem. This 
feature varies mostly with the position of the bud upon the shoot and 
with the vigour of the individual shoot. Buds are less likely to be 
closely adpressed towards the middle and base of the shoot, especially 
on sturdy wood. 
It should also be mentioned that wherever somewhat general terms 
are employed in the identification tables, terms such as " numerous " 
or " few," " large " or " small," " longer than broad " or " broader than 
long," these terms represent actual facts ascertained through a long 
process of taking measurements and counts. These measurements 
and counts have been recorded, but,, except in one instance, namely, 
that of the relative " Leaf Measurements " given in Table V., it 
seemed better not to risk confusion by allowing such minute details 
to preponderate. It is sufficient to say that the terms chosen to 
describe any character are such as to express as nearly as possible 
the relationship of one type with another and the degree of " con- 
stancy " that can be looked for in any particular character. 
Table IV. serves the same purpose in summer as Table III. 
does in winter (see p. 370) . Both these tables have been carefully 
tested as a means to identification by a friend who was not cognizant 
of the nine types, but who was handed them mixed up together in one 
bundle. Within a very short space of time all the shoots had been 
rightly numbered according to type, so that it is hoped, with some 
confidence, that these two tables may prove similarly useful to those 
concerned in the subject. Both Tables may be criticized as being 
needlessly detailed and exhaustive, but it was felt that all previous 
descriptions had erred so much on the other side that they afforded 
no infallible guide to identification.* 
* Some of the best descriptions I have. yet seen are those quoted in the 
Revue Horticole (Dec. 16, 1916) in an article on " Le Pommier Paradis " by 
M. Gerome. In comparing the distinctive characters of the French Paradise 
and the Doucin, M. Gerome gives the descriptions of two French botanists as 
being the most useful. Here are the descriptions : — 
Paradis. " Feuilles vertes en dessous, d'abord pubescentes sur les nervures, 
puis glabres : bourgeons velus non tomenteux. Pedicelles glabres ou pubescents, 
ainsi que le tube du calice ; fruit tres acerbe." 
Doucin. " Feuilles blanches tomenteuses en dessous mgme a l'etat adulte ; 
bourgeons tomenteux. Pedicelles pubescents tomenteux ainsi que le calice ; 
fruit a saveur douce." (G. de Saint-Pierre, " Flore des environs de Paris.") 
The second description, which takes into account characteristics of growth, 
considerably aids the above. 
Paradis. " Arbre peu eleve a rameaux 6pineux ; feuilles adultes glabres sur 
les deux faces ; fruit tres acerbe." 
Doucin : " Arbre assez eleve a rameaux peu ou point epineux ; feuilles adultes 
tomenteuses en dessous; fruit douceatre" (M. l'Abbe Coste, "Flore illustree 
de la France"). 
