473 
low on the old Candollean sequence to Frankeniaceae inclusive. 
It is to be regretted that so great a work must be continued on a 
system of classification which has proved itself inadequate. The 
want of system in the matter of nomenclature is also to be regretted, 
for here, as in the former parts, sentiment is the predominating 
guide in place of rule. 
The generic limits as a rule are well taken; some genera, how- 
ever, are too composite according to modern ideas. The interpre- 
tation of species is in many cases not as good as our present 
knowledge would permit. In Clematis, C. Addisonii is taken up in 
its Proper place, but C. viornioides, a very distinct form, if not a 
good species, is merely hinted at. Clematis ovata is reduced to 
C. ochroleuca, a position not warranted by the abundant specimens 
of recent collection. The author, following Coulter, has reduced 
Clematis Scottii to a variety of C. Douglasti; to be consistent why 
fot reduce C. Catesbiana to a variety or state of C. Virginiana? In 
Ranunculus we find R. Allegheniensis Britton inserted as a species, 
while R. micranthus Nutt. stands as a variety of R. adbortivus L. 
In Aguilegia, A. saximontana Rydberg is published for the first 
time and well distinguished from the related A. érevistyla Hook. 
Very scant courtesy is paid to A. Canadensts flavifiora, but almost 
. immediately following we find A. caerulea var. albiflora A. Gray, 
described as a new variety (while there are at least two published 
| available specific names for it),and its characters as given separate 
it from the type in only just the same degree that differentiates 
flaviflora from Canadensis. We can only assume that the editor 
holds a color variety proposed by Gray to be valid, while rejecting 
One maintained by others. On page 76, Nymphaca odorata var. 
minor appears in the old stereotyped form. Why should this 
State be called a variety when it is the original of the species 
odorata? Ifa variety must be maintained it should be done in a 
Consistent way. Little recognition is given an apparently dis- 
tinct species of Castalia from Florida and adjoining territory- 
The editor says ‘this form has been confidently identified with 
N. rentformis Walt. by the collector, . . adisposition in no sense 
Warranted by the brief and wholly dubious characterization of 
Walter’s species.” If there were nothing but Walter’s description 
to Support Mr, Nash’s position, it certainly would be weak, but 
