and particularly of the Genus Milium. 
25J 
most unaccountably failed. It is difficult to understand his defi- 
nitions of the two generic characters, and next to impossible to 
divine by what rule he distributed the species under each. 
From the Hedwigian school, which had thrown so much light 
on the generic characters of mosses in general, and which had 
done the most essential service in removing from Mnium and 
Bryum those vast and discordant tribes in which the peristomium 
is simple, every thing was to be expected upon the point in ques- 
tion. For my own part, after having contemplated with raptu- 
rous admiration the physiological discoveries of the illustrious 
Hedwig, and yielded that implicit assent to his assertions and 
deductions, which his clear and candid manner commands, I 
turned with eagerness to the methodical part of his works. My 
primary object was to learn to distinguish with certainty the ge* 
nera of Hypnum, Bryum and Mnium, about which botanists had 
ever been in dispute. But here I was disappointed. In his dis- 
tribution of the mosses with a single peristominm all is lucid order, 
so far at least as his principles are admissible. In the arrange- 
ment of those which have an inner per istomium, he appears to me 
to run into refinements which neither lead to the knowledge of 
natural genera, nor can easily be followed up by common ob- 
servers. I found with some concern that we must rely on the 
old mode of distinguishing llypmim, by its lateral fruit-stalk, 
from Bryum, the difficulties attending which are however hap- 
pily removed by the separation of the single-fringed mosses from 
the latter : I found moreover that Mnium remained at least as 
unintelligible as before. Indeed Hedwig has rather confused it 
by reversing the original characters. His Bryum has a round or 
capitate male flower; his Mnium a flat or discoid one. Mis most 
able ■ followers, Schreber, Swartz and Roth, well aware of the in- 
sufficiency of such distinctions, united the two genera into one. 
:»j vol. vil. 2 l while 
