376 JOURNAL OF THE ROYAL HORTICULTURAL SOCIETY. 
any of the other Paradise types, but though the growers, from whom 
it was received as a true sample, remember having used it for at 
least fifty years, they are unable to trace its origin. Moreover, 
they do not know its real name, for they sent the stock as a special 
strain of Broad-leaved English Paradise, from which stock it is so 
far removed as to bear no close resemblance to it. This stock appears 
to be in common use at least in Holland and Germany. It con- 
stituted the bulk of the Dutch collection and of several German 
collections. It may thus be possible to establish its identity.* 
Type VII. was similarly received true from only one English source. 
The growers have possessed it for a very long time and believe that it 
originally came from France. Although they still have the stock 
true, they have given up using it commercially, as the main demand 
is for trees worked on the ' Broad-leaved English.' Type VII. also 
appeared in a mixed collection from one other English source — where 
it can be remembered for a good many years — but its origin is again 
forgotten. In the foreign collections this type is entirely absent. 
Table II. shows that Type V., the ' Doucin Am chore/ is very 
widely distributed, especially as a minority in mixed collections of 
types, though it has only once appeared under its true name. 
Type VI., Rivers' Nonsuch, appears more widely distributed 
amongst English nurserymen than the Broad-leaf (Type I.), with 
which it is unfortunately not infrequently mixed. 
Type VIII., the French Paradise of to-day, seems not to be exten- 
sively used in this country, though it is one of the principal stocks on 
the Continent. The Table (II.) shows that it is universally recognized 
by its true name. 
Type IX., the Paradis Jaune de Metz, seems to show a similar 
history, though so far it has appeared from no English source direct. 
XI. The Comparative Tables III. and IV. 
Table III. shows a comparison of the nine types during the winter 
season. In all cases the descriptions (see p. 369) refer to characters 
of the growth made during the preceding summer. The buds, which 
are the main help in winter identification, are much more variable 
in size, shape, and angle of insertion on wood two and three years of 
age, hence it is advisable as far as possible to identify from " one year " 
wood. Even on one-year bedded stocks there is nearly always a 
certain amount of young growth available for identification purposes. 
The method by which these descriptions were compiled was such 
that any very noticeable variations should have become obvious. 
Each winter, from the time the collections were planted (191 3), the 
annual growth of every stool has been examined and described in 
detail if any variation seemed to appear from the characters laid down 
* Mr. E. A. Bunyard showed me in his own garden a bush he had received 
from Germany as Malus pumila. This I discovered to be identical with our 
Type IV. Paradise. I am indebted to Mr. Bunyard for several literary references 
to Paradise Stocks. 
