57 
inine, but^ where only understood, it might be masculine or neuter — 
we must commend the editor's closing remark: 
'* Perhaps the simplest and most easily recollected rule, would be 
to make the varietal name feminine in all cases, whether the var. or 
varietas^ were expressed or understood. This at least would be in- 
telligiblcj and would conduce to uniformity of practice." 
It would also be logical, and the logic also would require all 
specific names to be feminine; for the word understood, species^ is 
feminine. 
Now we do not suppose that Mr. De CandoUe would tolerate a 
double set of genders for the names of varieties. His doctrine is 
that the " var." should be discarded and the Greek letters only 
employed, not only for numbering the varieties, but for designating 
the fact that the name they are prefixed to is a variety. 
It is not difficult to perceive why it has come to pass that '* Eng- 
lish writers generally use the abbreviation var.^'' and that some con- 
tinental botanical writers follow the practice. One reason is that it 
enables us to cite an author's variety by its name without having to 
concern purselves with its Greek number, whether it is ^ or ;/ or d^ 
which otherwise we should have to attend to. Another is that our 
sense of good form revolts at beginning sentences and paragraphs 
without capitals. In our books, varieties usually stand in independent 
paragraphs. Even in Latin we do not like to begin a paragraph: 
''a indivisum foliis omnibus integerrimis serratisve, non aut vix 
basi auriculatis." 
In English we can still less abide it. So we prefix '' var.," and 
either number our varieties with Greek letters or, preferentially, leave 
them out. . - 
But, we did not suppose that by the employment of the word 
"var." we had interfered with the relation of the name of the variety 
to that of its genus. Var. indivisum^ in this case we should construe 
the phrase: '' Varietas cujus nomen est indivisum. * Var. indivisum ' 
stands on the same ground as -species aniphibiunu^' The latter rank 
we rarely need to express, because we always prefix the generic name 
or its initial. The former may often come in a shape which, renders 
the designating prefix var, necessary, or at least most convenient.^ 
We may, indeed, quite correctly write, var, albiflora, a white- 
flowered variety, var. longfolia, a long-leaved variety; but that is not 
according to the Linn^an pattern nor to the regular practice, nor to 
the strict analogy of the varietal name with the specific. 
Moreover, if the gender of the word which designates the grade 
of the name is to govern the gender of the name, at least when ex- 
pressed, as by var., then all subspecies must be made feminine. Now 
this term subspecies is coming largely into use, and it has to be ex- 
pressed in every case, in this wise: 
Ranunculus aquaiilis, L. 
Subsp. hetcrophyllus. 
Subsp. hederaceus, etc. 
If the proposition which we deprecate is adopted these names would 
have to be written heterophylla and hederaceea by an author who ranked 
them as subspecies, but heterophyllus ^Xid. hederaceus by one who took 
