Geology and Natural History. 155 
species, and then distinguish v var. a. villosa, and var. 6. glabrata; 
named form, and append the variety #. glabrata. 
has its advantages and is likely to be employed in certain 
cases. The former classifies the erat under the species, 
ly, exhi 
acter of what we call a varia oe spealdes ony andolle con- 
— me it will prevail in proportion as ‘the ss-of a species 
co o be well known; the latter holds sisi to the bibli- 
are etek: as Sopa a from this. ven wiinn the chek 
character is drawn so as generally to cover the varieties (as 
from a pelts 
As DeCandolle peste out, ae: is much res tew and loose- 
ness in the use of this word type, which it would be well to avoid. 
Properly the type of a species és the species or genus, or the full 
sarge of it, which no one individual or species may embody, which 
the case of a a group no single sare Maa or member can 
fully exemplify. To apply the term to a form which well exem- 
i 
which very often proves not to be the best representative of the 
group, sometimes not even a fair one. Finally a particular speci- 
men which the rel author described, or ‘an authentic speci- 
men, is said to b type, or a typical speci imen; and this De- 
Candolle objects to. eB after all, such terms can hardly be held 
to a single sense in technical any more than in ordinary language. 
Something must be left for the context to determine. 
In drawing up the characters of groups, such especi ially as 
orders and genera, are exceptions or what we call exceptions to 
be indicated in the character, or shall this express only what is 
generally true? DeCandolle discusses = ‘perineal _ leaves it, 
as must needs be, for practical judgment to dete On the 
e hand the point or the usefulness of a paiceeaiombey is » handio’ or 
diecipsted by the intercalation of alternatives and exceptions, yet 
aracters must be somehow made to correspond with the facts. 
The method of Bentham and Hooker, of a separate specification 
of the principal known exceptions, is commended. 
