AVES PROCELLARIID^. 121 



a true Pyocellaria at all ; and this view is confirmed by other characters, 

 such as the shape of its nostrils, the elongated tarsi, which are much 

 longer than the mid toe and covered anteriorly with transversely arranged 

 scutellas, the very minute hallux, and the lamellar, concave form of the 

 claws. It belongs, in fact, to the group of Oceanites, Fregetta and Pelag- 

 odrofiia, but is not exactly congeneric with any of them. I propose 

 therefore to make it the type of a new genus, to be called Garrodia, in 

 memory of my lamented friend A. H. Garrod, not only as a token of my 

 personal esteem for and indebtedness to him, but also as some slight 

 recognition of the thanks ornithologists generally owe him for the addi- 

 tions he made to our knowledge of the anatomy of birds. 



"The genus Garrodia may be shortly defined as follows: 



''Garrodia. Gemts ex ordine Tubiiiariitni Oceanitas luaxime affine, tar sis 

 pro digitis longioribus et antice scutellatis, necnon margine sterni posteriore 

 integro distiugneuduni. 



" Type Procellaria nereis, Gould. 



" Garrodia is perhaps most closely allied to Oceanites, as already stated, 

 but differs from that genus in having the tarso-metatarsi covered anteri- 

 orly with a series of transverse scutellae instead of being 'entire,' in their 

 slightly greater proportional length as compared with the third M;oe, in the 

 even more minute hallux, and in the more flattened and lamellar form of 

 the claws. The sternum too is posteriorly entire, whereas in Oceanites 

 oceanicits it is slightly notched. The coloration of the two genera is also 

 quite different. From Fregetta, Garrodia may be easily distinguished by 

 the very different proportions and forms of the nails and feet in that 

 genus, and from Pelagodroma by its much shorter feet and entire tail. 



"These four genera — Oceanites, Garrodia, Pelagodroma and Fregetta — 

 form a very well-marked family of the Tubinares, which may be called 

 Oceanitidae, as distinguished from the remainder of the group, or Ful- 

 maridae of Prof. Garrod. Anatomically, these four genera agree together, 

 and differ from the Fulmaridae (on nearly all the genera of which, includ- 

 ing Diomedea and Puffiiinria, I have notes), in the two important charac- 

 ters already mentioned — the absence of caeca and the presence of the 

 accessory semitendinosus muscle. Externally they may be at once recog- 

 nized by their peculiar elongated tarsi, lamellar nails, and by never having 

 more than lo secondaries, Procellaria and Pnffinuria having 13, and the 

 remaining Fulmaridae more (in Diomedea, according to Nitzsch, as many 



