260 THE AMERICAN NATURALIST.  [Vou. XXXIII: 
Dr. Clark having the advantage of almost exclusively basing his 
observations on fresh birds instead of dried skins, thereby escaping 
many errors unavoidable to his predecessor and being enabled to go 
into such details as are demanded by the more refined requirements 
of modern science. Thus, while Nitzsch only examined the skihs 
of 5 species within the frame of Dr. Clark’s article, the latter could 
work with 65 specimens in the flesh, representing 18 species and all 
the genera of the territory in question. His careful descriptions 
and figures are, therefore, very valuable to the systematic ornitholo- 
gist, and the conclusions he bases upon them entitled to great con- 
sideration. But while thus restricted to fresh material, he is also 
limited to a small number of forms of the groups he treats of, and 
while this limitation in no way lessens the usefulness of the material, 
which he thus places in the hand of the working systematist, it 
naturally interferes with the trustworthiness of the generalizations. 
In discussing his “ conclusions ” (pp. 651—653) it should be borne 
in mind, however, that Dr. Clark has been very careful and guarded 
in expressing his views, and that he has avoided to be dogmatic ; 
nor does he claim that much light has been shed upon the origin or 
relationships of the larger groups. ‘The relationships of the various 
genera within these groups, however, he thinks “is at least suggested 
by these investigations ” to the extent that he ventures to express 
them in diagrams which look suspiciously like ‘“ stammbaums.” 
But I am not certain that they mean more than a diagrammatic 
representation of the pterylographic “relationships” without refer- 
ence to the true phylogeny of the species, unless he really believes 
that the pterylographic characters alone are sufficient to indicate the 
various shades of interrelation or the course followed in the evolu- 
tion of these genera. This uncertainty is indicated by his speaking 
of the diagrams as “pointing out the relationship of the genera,” 
though in the next moment he selects a type at the bottom of the 
series simply for the practical convenience of having “a starting 
point from which to develop the other genera.” Seeing, however, 
that the diagrams have no real value unless meant for “ ephylogenetic 
trees,” whether standing on their roots or their tops, I shall discuss 
them from the latter point of view. 
It is then plain that Dr. Clark, in spite of his guarded reservation, 
really regards the trees as standing on their roots, for he is careful 
to select the genera he starts from, on account of characters which 
Feather-Tracts of North American Grouse and Quail, Proc. U. S. Nat. Mus., 
vol. xxi (1898), No. 1166, pp. 641-653, Pls. XLVII-XLIX. 
