166 
TODUS. 
diminutive, not more than three inches and a quar- 
ter long, of which the head and bill actually occu- 
pies one inch and a tenth, or nearly one-third the 
length of the entire bird. It differs also from 
all the other species in having the hill more gra- 
dually attenuated in its breadth as it approaches 
the end ; just, in short, as that of Conopophaga , of 
which it becomes, by its very short tail, large head, 
and bristled front, the complete prototype. Never- 
theless, the sides of the mouth are strongly bristled, 
and the tarsi are considerably shorter than those of 
its congeners. Now, all these variations from the 
typical species are absolutely necessary to the har- 
mony of nature. They indicate, as plainly as pos- 
sible, that this bird is an aberrant species, and that 
it is of the fissirostral or aquatic type. Nor can it 
possibly he removed from this position among the 
todies, since there are now before me three other 
species which establishes the gradation between 
tnelanocephalus and megacephalus in the most deci- 
sive manner. And yet, if the bills of these two 
kindred species are viewed separately, one might 
be tempted to place them in different genera. The 
situation, also, of Todus megacephalus in the se- 
ries of species, is further confirmed by the structure 
of T. platycircus, which, from being clearly a re- 
presentation of the Rasores, will stand at the opposite 
side of the circle to that of Megacephalus; the gral- 
latorial, or intervening type, as before observed, 
being wanting. We thus find, that even in a sub~ 
genus the variations are sufficiently marked to give 
