234 BOTANICAL INFORMATION. 
old species, nearly complete sets of Blume's Javanese an 
Galeotti and. Linden's Mexican plants received by the Mu- 
seum from Holland and Belgium.* He had also access to 
the most complete botanical libraries in existence; and when 
we consider in addition to these advantages, the high autho- 
rity of Robert Brown, whose principles the author generally 
follows, his own well known scientific views, ability and accu- 
racy, and his admirable talent for botanical delineation, we 
should think it vain to attempt a criticism of his work with- 
out having followed him through it step by step. And if on 
taking up an Asclepiadeous plant to determine, we are alarmed 
on observing in this very natural order, one hundred and 
thirty-three genera, of which many appear to us exactly alike 
in habit, inflorescence and general aspect of their minute 
flowers, and are tempted to exclaim against them as artificial 
and fanciful, we may at least rely upon their definitions being 
accurate, and provided we do not fail in our own powers of 
observation, we shall be safely led to the genus we are 1 
search of. T 
The Loganiacee (partly by the elder De Candolle, partly 
by his son) with which the ninth volume commences, present, 
at first sight, an anomalous assemblage of genera, expelled 
from other orders; but, if we strike out three or four genera, 
which may possibly find their place elsewhere, and reduce the 
Loganiacee to those limits which were probably originally con- 
templated by Brown, we shall have an intelligible group, well 
designated as “ Rubiacee with a free ovary? Brown orig 
nally said they might be either considered as an independent 
order or be united with Rubiacee ; Torrey and Gray adopted 
the latter course, De Candolle has preferred the former; 
this appears to us the most convenient, there being nothing 
to interfere with drawing a positive line of demarcation, even 
between the semi-free ovary of Hedyotis (Houstonia) cer’ 
lea, and the supposed partial adherence of the ovary in Mi- 
treola, which we confess escapes our observation. 
* Of Gardner's Brasilian collections, those from Goyaz and Minas Ge- 
raes appear to have reached him too late for insertion, 
