ae 
148 ON THE CLASSIFICATION OF ola 
the other, we immediately perceive that the straig 
of the bill is a typical perfection of the h 
In the genus Lampornis the bill is eee auc 
depressed at the base, and the tail is broad and even ; 
while the type of the genus Campylopterus seems to be 
the recurved-billed humming-birds. Last of all come 
the genus Phethornis : hitherto the form of the tail has 
either been square, forked, or rounded; but in this 
group the tail is considerably and regularly graduated, 
the side feathers very short, and the middle pair far 
exceeding all the others; the bill is not merely bent, 
but so much curved in the typical species as nearly to 
assume the form of a sickle. The gay and beautiful 
green which ornaments the upper plumage of all the 
groups here gives place to a brown colour, and even the 
throats of the male birds are destitute of ornament. 
The genus Phethornis, in fact, obviously represents the 
rasorial type, and is a miniature likeness of the mini 
or to that family with which we began our survey 
the Tenuirostres. 
(167.) Although the circle of this tribe is by no 
means so complete as some others, it must still be 
_ remembered that nothing yet known tends to invalidate 
the succession of the forms here traced. On looking to 
the hoopoes, and more particularly to the genus Epi- 
machus of M. Cuvier, we see an evident tendency in 
nature towards some extraordinary developement of 
plumage, quite different from the usual arrangement ; 
nor in no one family is this more apparent than among the 
Paradisiade. On looking again to the genus Ptiloris, we 
have such a close resemblance both to the Epimachus and 
to the Paradise birds, that some writers have actually 
placed it as species of Epimachus ; and such it would 
unquestionably be, but for its possessing the feet of the ; 
Meliphagide. It is quite immaterial to this view of the 
subject, whether a whole or a part of the Paradise birds 
enter between the Promeropide and the Meliphagide: it 4 
sufficient to show thestrong probability of the interval being — 
filled up bysome birds of a similar construction,—that is 
