I OO Recent Literature. [April 



ferent from those usuallj assigned it. It embraces not only the typical 

 Thrushes (subfamily Turdince of authors generally), but the Sylviidce 

 (subfamily Sylviince of this work), and the SaxicolidcB, \vhich form part of 

 the present author's Turdince. On the other hand the Mimince, so gener- 

 ally of late regarded as a subfamily of the Turdides, are excluded. In 

 respect to genera, Mr. Seebohm is eminently and, as we believe, laudably 

 conservative, but in matters of nomenclature he, is in some points a law 

 unto himself.' In other respects he shows himself to be refreshingly inde- 

 pendent and untrammeled by traditional methods of working. In his "In- 

 troduction" he strikes the key-note of, for our English friends, a new 

 departure in respect to the treatment of subspecific forms so boldly and 

 with such a true ring that we cannot resist the inclination to let him be 

 heard in part in his own words, even at the expense of extending this no- 

 tice beyond usual limits. • i 

 ^ In respect to genera and generic characters his position is somewhat out 

 of the usual line, as may be seen from the following summary of his 

 remarks on these points, arid he expects to be thought more or less heter- 

 odox in his treatment of genera.* ''If I am accused," he says, "of disturb- 

 ing the existing genera of birds, in some instances by cutting up recog- 

 nized genera, and in others by uniting several together, I can only plead 

 that I haVe not done so capricioiisly, but in order to facilitate the perception 

 of thfe cidm'munity "of origin, which must more and more interest those 

 who accept'the' theory of evolution." After alluding to the fact that it has 

 b'een"acfcepted as an'axiom ampngst ornithologists that genera must be 

 founded upon structural characters," he states that in the first part of his 

 work he ''made usfe of characters that are riot structural in defining some 

 genera,'* and in other cases "considered so-called structural characters as 

 having only value enough to divide genera into subgeneric groups," while 

 in the second part of the work he has convinced himself "that these so- 

 called structural characters have no generic value at all." and further states 

 that he was obliged to fall back upon "colour or pattern of colour as the 

 only character which indicates near relationsliip. In my opinion," he 

 continues, "the pattern of the colour in the family or subfamily of Tur- 

 din£e is a character which is more trustworthy (as showing community of 

 origin), which in fact dates further back than the shape of the wings, tail, 

 and bill." Of the SylviincB., with 104 species, only 7 genera are recognized, 

 which average 15 species each, the largest genus, Phyllosco^us, having 

 25, while the genus Sylvia hia.s 22,. Trie subfamily Turdince, 'v/hlch in- 

 cludes trie Blliebirds, Robins, and Stonechats, as well as trie typical 

 Thrushes, with 246 species, is divided into 11 genera, averaging 22 each, 

 while some of trie larger genera, as Geocichla, Turdus, and Mertila, have 

 respectively 40, 48, and 52 species each. While on the subject of genera 

 we may add that the genus Tttrdus includes all the New World Thrushes 



*That such \viH be the case is already evinced by a paper by Mr. L. Stejneger (Proc. 

 U. S. Nat. Mus., 1882, pp. 449 ei seq.) entitled "Remarks on the Systematic Arrange- 

 ment of the American Turdldce" (received since the preparation of this review), in 

 which Mr. Seebohm's genera and generic characters are considered. 



