106 ON THE CLASSIFICATION OP BIRDS. 
both, as a whole, have the upper mandible more curved 
and deeper notched than any other. It is only among 
the Glaucopinat and the ScaphiduriruB that we find that 
unusual developement of tail so characteristic of rasorial 
types. The Crypsirina temnura in one group, and the 
whole of the typical boat-tails in the other, verify these 
to be rasorial types, and consequently analogous. It is 
at these points that the circles of the crows and the 
starlings touch and pass into other ; and we thus find 
that it is here the two groups evince the closest affinity : 
a singular circumstance corroborating tliis is afforded by 
the fact of the commissure of the bill, in both groups, 
being sinuated in the typical genera. From this point, 
as the affinities recede of the two families, and become 
more’ and more distant, so, as a natural consequence, 
do their analogies also become more remote. Hence, 
those between the two most aberrant groups, Icturina; 
and CoraciruB, appear to rest only on the fact of both 
being strictly composed of arboreal birds, having pecu- 
liarly short feet, and living principally upon soft fruits. 
Fin^y, the most remote analogy is that between the 
Agelaitta and the Frugilinm, because they stand at the 
coniines of their respective circles ; and yet, as these 
two possess the common property of having the longest 
wings in their respective circles, they must be repre- 
sentations of each other : this structure, we know, 
is eminently characteristic of fissirostral types ; so that 
the circle of each having now been compared, we find 
they produce a perfectly uniform result, adding one 
more to those innumerable instances we have brought 
forward to demonstrate the law of represent- 
ation. 
(122.) We shall now throw these two groups into 
their proper circles, because the analogies will not only 
be at once placed before the eye, but because these 
circles will also illustrate that apparent transportation of 
groups we have just before alluded to : 
