262 THE AMERICAN NATURALIST. [VoL. XXXIII. 
We shall now apply these principles in an attempt at reconstruct- 
ing the genealogic tree of the North American partridges or quails, 
chiefly from the pterylographic characters. 
In viewing these forms it seems thus evident that we have two 
groups, the members of which bear closer relationship to each other 
than to those of the other group. On the one hand, we have Colinus 
and Cyrtonyx; on the other, Callipepla, Oreortyx, and Lophortyx. The 
relationship of the former is obvious. The speckled-backed Cyrtonyx 
cannot well have developed through stages like those of the plain- 
backed Zophortyx and Callipepla, as suggested by Dr. Clark; and this 
character alone goes a long way to show the more generalized status 
of Colinus and Cyrtonyx. Cyrtonyx, with its higher developed crest, 
differentiated ventral feather-tract, reduced tail, and wanting claw to 
the thumb, is manifestly the more highly specialized form, thus leav- 
ing Colinus on the bottom round of the ladder. This is the same 
position given it by Dr. Clark it will be observed. The only differ- 
ence between us is that he places it there because, among other 
generalized characters, it possesses 12 rectrices, while I give it a 
similar place 77 site of this fact. By placing it lower than Callipepla, 
with 14 rectrices, I suggest that while Co/inus has already lost a 
pair, Callipepla is, on the whole, a more specialized form, in spite of 
the fact that it has retained the greater number of rectrices. If, 
therefore, the two groups of quail have a common origin, this ances- 
tor is obviously below the line of the beginning of our tree, having 
in all probability a speckled plumage like Colinus, 14 rectrices and 
16 secondaries. 
Of the 3 members of the second group it is only necessary to say 
that Callipepla, with its 14 rectrices and loose crest, is the more 
generalized form, probably descended from an ancestor with 16 
secondaries, by reduction to 14. From this ancestor Oreortyx as- 
cended, on the one hand, by further specializing the crest, though 
retaining the 16 secondaries; while, on the other hand, Zophortyx 
lost two of the latter, besides assuming a still more complicated 
top-knot. 
Assuming, then, the common origin of these genera, I substitute 
the following diagram as probably representing more nearly the 
true course of the development diagram of these forms. 
The question of the genealogy of the North American Zetraonine 
is one greatly more complicated, for the reason that two of the genera 
apparently hold closer relationships to forms not occurring in this 
hemisphere than to the genera which are restricted to our continent. 
