198 
ON THE classification OF BIRDS. 
to arrogance and self-sufficiency. It may be so : and 
yet it has only been these fits of enthusiasm which have 
borne me onward, amidst those doubts, and discourage- 
ments, and difficulties which always attend an original 
line of investigation. When I reflect, that what is now 
laid before the public, is the fruit of thirteen year-s’ study 
and meditation — almost incessant — I feel at times even 
painfully sensible that these results are so inadequate to 
what might have been expected ; and that much more 
might have been accomplislied, if the same labour had 
been bestowed upon the subject by others. 'Few, indeed, 
can be more dissatisfied with the imperfect conclusions 
containeil in these volumes, than myself; and I almost 
regret that I have brought them before the public in 
such an unfinisheil state. Had health, and fortune, and 
freedom from professional duties permitted, 1 should 
have elaborated this part of my researches much more, 
by visiting the public Museums of Europe, adding more 
to my own by extensive purchases at home, and by em- 
ploying collectors abroad : but few private fortunes can 
command these resources ; and no assistance or patronage, 
in cases like this, can be expected from our government. 
The scientific world must therefore rest satisfied, or dis- 
satisfied, with what I have been able to do, small as it 
appears. Misapplications of my own theory will, doubt- 
less, be found in the subsequent pages ; and those who 
will point them out by making a better arrangement of 
the groups, upon principles equally universal and com- 
prehensive, will deserve well of me, and of science at 
large. But these errors, however numerous they may 
be, touch not the theory itself. Whatever may befall 
other circles, those of the Psittacidm and the Pician/e 
I have found the most perfect; and they are conse- 
quently those which, in the most minute details of 
variation, are the best calculated to test the truth of 
the theory itself. In the latter, more especially, I 
have even found three out of the five generic circles 
to be actually perfect, even so far as my own limited 
knowledge extends. There is also a peculiar advantage 
