FLORA OF GUIANA. 117 
occurred to me that it might be advantageous to modify, in some 
respects, the plan hitherto followed in describing Schomburgk’s 
collections. Instead of confining myself to them exclusively, 
I propose henceforth to give a complete enumeration of all the 
species of each group hitherto published as natives of Guiana, 
commencing with those natural orders not touched upon in my 
former papers. In this enumeration the species which I do not 
possess, or have not examined myself, will be found distinguished 
by an asterisk (*), the stations thus given on the authority of others 
being enclosed in a parenthesis. There are also a few species of 
Sir R. Schomburgk’s first collection which were gathered in 
North Brazil, on the Rio Negro and the Rio Branco; these, 
although not from Guiana, will be enumerated as before, but 
without prefixing any number to them, and they will, moreover, 
be distinguished by a cross (t) before their names. 
The labels of Schomburgk's second collection have generally 
two numbers; of these the first is that of Sir Robert Schomburgk, 
the second, in a parenthesis, is that transmitted to Berlin by Mr. 
Richard Schomburgk, and corresponds, it is believed, with those 
given in Dr. Klotzsch's papers on Equatorial American plants in 
the Linnea. But with regard to all these numbers, useful as they 
are in the determination of distributed collections, and strongly as 
it is to be recommended to monographists and other describers of 
plants, not to neglect them, it must be borne in mind that they 
are liable to many mistakes. The collections are usually hastily 
sorted for distribution, and distinct species, bearing a general re- 
semblance to each other, are often confounded under one number; — 
labels bearing numbers only, when accidentally mis-placed in —— 
herbaria, afford no clue to correct the mistake, and even in publi- — — 
cation, a clerical or typographical error in a figure is more apt to — 
be over-looked than any other. A specimen cannot, therefore, be — — 
considered as absolutely authentic merely because it bears a cor- — 
responding number to one published from the same collector, - 
unless it is found really to agree with the description, or has been 
actually compared. with the individual described ; although in nine. 
VOL. VII. ; L 
