TRINOMIALS 335 
partial American practice as stated in Britton and Brown’s 
Finite Flora (i. p. xi) is that the former retains and the 
latter omits the word ‘“var.”: that in the latter case “ varieties 
are printed as trinomials, e.g. ‘Ryne hospora glomerata paniculata, 
the pnea AIG var. or the Greek letters a or 8B commonly inserted 
between the specific and varietal name being dispensed with, “sf 
sacha ws Moss writes “ Populus tremula (a) var. sericea,’ In 
the former case, however, Rynchospora glomerata has a definite 
Bstatiog apar rt from the forms which the Americans term sub- 
species—“ the name avacietyt is felairatid to horticultural species ” 
whereas” in the latter Populus tremula is distributed into a 
varieties, we the binomial has no separate being. 
It will make my meaning clear if I quote Dr. Moss’s explana- 
tion of his “ “ position with regard to the subdivision of species 
into varieties.” ‘There are,” he says, “two plans, each of which 
finds favour in certain circles. One is to regard a certain form of 
a species as typical of that species, and to regard any deviations 
=e ig type as varieties. The second plan is to subdivide the 
e species wholly into Mets Ba as a genus is wholly 
subdivided into species. Populus tremula may be taken as an 
illustration. ati varieties of this at fouognitned as British. 
e is a form whose young leaves are silky, and the other form 
howe leaves (excepting the leaves of the suckers) are always 
glabrous or almost glabrous. If the firs + of the above plans be 
adopted, it becomes necessary to decide which of the two varieties 
shall be regarded as the type. Supposing the silky variety to be 
regarded as the type, the British forms would be written thus :— 
opulus tremvu 
(6) var. glabra. 
If the glabrous a be regarded as the type, then the British 
forms would be written thus: 
SPipaiee tremala 
(a) va 
However we ourselves have doula phat to adopt this first plan 
but the second; and accordingly we write the British forms thus :— 
a 
We have decided on this plan for two reasons. First, it is (so 
far as we are able to Judge) quite att to decide which of the 
forms of a species is the type; and it is unusual to find agree- 
is 
type | and which the deviation from the ooh We frequently find 
the is merely the form 
is more abundant in the district which he usually investigates ; ; 
and we find that this view of the of the species sometimes 
prevents him from taking a broad view of the relationships of the 
different forms of the species. Secondly, it is impossible, if the 
* Bull. Torr. Bot. Club, 1904, 251. 
