i CIRCLE OF THE TENUIROSTRES. 149 
partaking both of the structure of the hoopoes and of the 
honeysuckers ; and both Cuvier, Lesson, Jardine, Selby, 
and James Wilson concur in believing that our Ptiloris 
paradiseus exhibits such a union: into which of these 
families, therefore, Ptiloris naturally enters, is a se- 
condary question, whichneed not here bediscussed. 'The- 
oretically we might make use of this form, either to 
pass from the Paradise birds to the Meliphagide, or to 
connect the latter with the Promeropide, and thereby 
represent the aberrant circle as closed. Theviews, indeed, 
of the ornithologists above named, are all in favour of 
the latter supposition, seeing that they have not, in every 
instance, separated Ptiloris from Hpimachus ; the latter 
being universally considered as a hoopoe. Such appears, 
to us, an impartial statement of the difficulties which at 
present impede our knowledge of these three groups. 
(168.) Between the honeysuckers and the sunbirds 
the affinity is unquestionabje ; and no one can doubt 
that the latter are followed by the humming-birds. | 
The similarity between these latter and some of the 
small hoopoes is sufficiently strong to point out an 
affinity not very distant between them, so that there 
is every reason to believe the Tenuirostres are naturally 
disposed in one great circle. The Trochilide, never- 
theless, are one of those strongly marked groups in 
nature, which, if the term can be correctly applied, 
are, in one respect, isolated. The species are very 
humerous, and exhibit among themselves a_ great 
variation in the structure of particular members; yet, 
when viewed as a whole, they are stamped with such 
a peculiar character as to be recognised at the first 
glance. In other groups, ornithologists are perpetually 
referring species to families which they closely resemble, 
but to which they have no real affinity: yet no one 
can mistake a humming-bird, a hornbill, a toucan, a 
parrot, or a falcon; nor is there hardly any species in 
these groups partaking so little of their respective 
typical characters, as to excite a doubt to which it 
truly belonged. Now, in applying these remarks to 
L 3 
