V ,°g 93 X ] Recent Literature. 6j 



divided into three 'sections,' according to the presence or absence, or 

 partial absence, of serrations near the tip of the bill, viz. (i) 'Trochili 

 serrirostres,' (2) 'Trochili intermedii,' and (3) 'Trochili laevirostres.' The 

 difficulty of finding trenchant characters for the subdivision of the 

 Trochili into natural supergeneric groups is well known, but the one 

 here chosen seems the most unfortunate and artificial that could have 

 been selected. Besides being one of degree merely of a character rarely 

 strongly marked, the major divisions based on it separate widely genera 

 which in general characters are often closely related. This results in 

 extensive transpositions of genera from the order in which they have 

 heretofore usually been placed. As already stated, the number of species 

 recognized as valid is 482 ; as, however, a number of species are mentioned 

 in foot-notes as unknown to the author, and thus not included in the 

 above number, the total of species and subspecies may be estimated in 

 round numbers as not far from 500. This is a considerable increase over 

 the number (426) recognized by Mr. Elliot in 1878, since which time, 

 however, many new species have been described. The method of treat- 

 ment is of course similar to that of the preceding volumes of the 'Cata- 

 logue.' We note one new genus, Neolesbia (p. 145, type Cyanolesbia 

 nehrkorni Berl.), but the eight apparently new species, as thev stand in 

 the body of the work, were really described in 1891, in the 'Annals and 

 Magazine of Natural History' (Vol. VII. pp. 375-379), as duly noted in 

 the 'Addenda' to the present volume. 



The Trochili occupy pp. 27-433 of Vol. XVI, and are followed by the 

 first four families of the 'Coracise,' namely, Cypselidre. Caprimulgidae, 

 Sleatornithidae, and Podargidae, by Mr. Hartert. The Cypselidae number 

 78 species, and are represented in the British Museum by 1500 specimens; 

 the Caprimulgidae number 86 species, represented by 1S00 specimens; the 

 Podargidae number 24 species, represented by 271 specimens; and the 

 Steatornithidae by 1 species and 14 specimens. Respecting the relation- 

 ships of some of these groups, Mr. Hartert observes : "The Cypselidae 

 have been placed, along with the Trochilidae and Caprimulgidae, in the 

 order 'Macrochires' or 'Strisores.' This arrangement is based on good 

 grounds, and has much to commend it. Some recent anatomists deny 

 the near relationship of these families. If raised to the rank of orders, 

 Cvpseli, Trochili, and Caprimulgi should be placed near together" (p. 



435)- 



Mr. Hartert prefers to retain the name Cypselidae for the Swifts, in 

 place of Micropodidae, on the ground that the law of priority should apply 

 to families as well as to genera and species. With this we agree, but we 

 differ from Mr. Hartert in his interpretation of its application to families. 

 When Cypselus is shown to be a synonym of Micropus, and Micropus 

 becomes the name of the group formerly currently known as Cypselus, 

 from which the name of the family was derived, we believe the name of 

 the family should be changed to conform to the correct name of the genus 

 from which the name of the family is taken. In cases like the present it 

 might seem better to retain well-known family names, although based on 



