PROCEEDIXGS OF UNITED STATES NATIONAL MUSEUM. 451 



but which possess the above characters at least as well determined as 

 his "aberrant" Geocichhe. 



Thus, besides the cases mentioned by Mr. Seebohm himself, '■'Turdus^' 

 pallasii has the light color ou the inner web "very abruptly defined;" 

 his Turdus albiventer likewise has "the pale portions of the inner webs 

 of the quills greatly developed and very abruptly defined," and so 

 further on. But he gives no characters by which these aberrant 

 species (aberrant of both genera) may be distinguished, and he also 

 gives no reason why he places these aberrant forms in different genera. 

 It would be interesting to know why Turdus albiventer is not as good a 

 Oeocichla as G. sinensis, when the pattern of coloration is the only 

 character which has generic value; or, in other words, why he does 

 not place G. sinensis in another genus. May there not, perhaps, be 

 other characters of more importa4ice and generic value, and which indi- 

 cate a nearer relationship than the coloration? But one ought not to 

 suspect that, as Mr. Seebohm, in another place, retains a species in the 

 genus Cafharus, for the reason that its " general style of coloration" so 

 closely resembles the other species of that genus, although it "is a 

 typical JSrithacus so ftir as what are called structural characters are 

 concerned." Here, again, "the general style of coloration" is the only 

 generic character of v^alue! In the one genus it thus apparently has 

 no value at all, while in the other it is the only important one! It is 

 curious to see that Mr. Seebohm, when neither pattern nor general 

 style of coloration is sufficient to sepaitite two genera, hastily takes 

 refuge in a structural character; for example, p. 362, and p. 334, and 

 especially the "Key," p. 146, d'", e"',J"', and //'". Curiously enough, 

 he separates two genera, in either of which several species are simply 

 inseparable as to general style and pattern of coloration from certain 

 ones of the other genus, and j^et such similarly colored species, he says, 

 are typical members of the other genus, so far as structural characters 

 are concerned! How, then, will ]\Tr. Seebohm tell Catharns graciliros- 

 tris or occidentalis from " Erithacus " luscinia and pJiilomela ? In colora- 

 tion C occidentalis and E. luscinia agree so closely that it would be 

 very difficult to sei)arate them even specifically, if we had no structural 

 characters, and he expressly makes the statement that (7. gracilirostris 

 is, as to structure, a typical Erithacus. We will attempt, by his " Key," 

 to unravel this intricate question. In this the distinctive marks of each 

 genus are given as follows: 

 a'". General color ofuiuler parts slate gray, shading only into brown 



or white. Legs never black 5. Catharus. 



h'". Throat generally brilliant in color and frequently in violent con- 

 trast to the cheeks ; if not, legs pale 6. Erithacus. 



Unfortunately, the "key" is of no use; both the species of Catharus 

 and Erithacus mentioned above have not a brilliant colored throat; and 

 further, the legs are in both pale, and never black. If no structural 

 differences are to be found, the separation of the two genera would, in 



