374 JOURNAL OF THE ROYAL HORTICULTURAL SOCIETY. 
X. Isolation of the Nine Types. — The Comparative Tables. 
When once the principal distinguishing features of each type 
had been ascertained, it became fairly easy to sort out the mixtures 
by a series of systematic examinations. For the initial purposes of 
classification the various trade names were dropped, and as each 
type was segregated, it was simply numbered. Thus the types have 
been known as Types I: to IX. Since at least three types — one 
of which is very frequently found — are as yet unidentified and did 
not come under any recognized trade name, it is essential to retain 
this method of numbering in this Report. At the same time, when- 
ever the identity of a type seems certain and the nomenclature satis- 
factory, the actual name of the type will be added to the numeral. 
Table II. gives an " historic " summary of the nine types as found 
in the collections. It also gives the names under which each type 
was received and shows the different mixtures of one type with 
another. It is worth noticing how certain mixtures of two types 
constantly recur. 
Perhaps the most outstanding feature in this table is the 
frequency with which the name Broad-leaved English Paradise 
occurs, despite the fact that it was only received at all in six 
collections, and then only in two cases was the collection true. It 
is frequently mixed with the Nonsuch Paradise. Though the 
two stocks (Types I. and VI.) originated from the same source, 
there is no reason whatever for any confusion between them, as 
their characteristics are most distinct and their general habit very 
different. 
The next point to notice in the table is the prevalence of the 
Doucin stock. Though it has been sent true in the great majority 
of collections, yet trade names for it are very erratic and misleading. 
This is a matter of considerable importance, as the true Doucin seems 
to possess marked individual characters and is perhaps farthest 
removed in its rooting system from any of the other ' Paradise ' 
types, as will be seen later. Hence it is important that it should 
be known under its true name. When mixed it is generally to be 
found in company with Type III. 
Type III. is a remarkable case which shows clearly the necessity 
for such standardization work as the present. It will be observed 
that though this type appears to be, and undoubtedly is, one of 
the most widely used and distributed amongst Paradise stocks, yet 
it was never sent as a true sample under a separate name. It fre- 
quently appears mixed with Types II. and VIII., from both of which 
it differs very considerably. It is this same type which was recently 
discovered so plentifully mixed with the 10,000 ' Doucin ' stocks 
from France, that have already been referred to. 
Types IV. and VIL show a somewhat puzzling record. Type IV. 
was received true from only one English source, whilst it appeared 
in another mixed English collection. It is remarkably distinct from 
