1922.] Petrels of the Gemis Ocea.nodvoma. 311 



The culmen, wino-, and tarsus are longer than the 



measurements given by Salvin, GoJman, and Bailey, as is 



shown from the following table : — 



Depth of Measured 

 Tarsus. taU fork. by : 



1-21 = 31-5 — Salvin 



1-2 = 31 — Godman 



1-2=31 1-2 = 31 Bailey 



If the Japanese examples have constantlj^a longer culmen, 

 wing, and tarsus than the true 0. melania in a more extended 

 series, I propose to call them by the new name of 



Oceanodroma melania matsudariae, subsp. n. 



Oceanodroma markhami owstoni (Mathews & Tredale). 



Procellaria melania (nee Bp.) Seebohm, Bds. Jap. Emp. 

 1890, p. 270. 



Oceanodroma fuliginosa (nee Gm.) Stejncger, Froc. U.S. 

 Nat. Mus. xvi. 1893, p. 620. 



Oceanodroma markhaml (nee Salvin) id. op cit. p. 621. 



? Oceanodroma tristrami Salvin, Cat. B. Brit. Mus. xxv. 

 1896, p. 351 (type, Sendai Bay, Hondo, Japan). 



Cymoclwrea oicstoni Mathews & Iredale, Ibis, 1915, p. 58 

 (type, Okinose, in the Sagami Sea). 



The form is a distinctly larger bird than typical 0. mark- 

 hami. In my opinion it is not separable from it as species 

 but is subspecifically distinct. 



A large series was obtained by Viscount Matsudaira off 

 the coast of Sagami Bay, Hondo, Japan, from early to the 

 end of May of the years 1917, 1920, and 1921. These birds 

 agree perfectly with the description of Mathews & Iredale. 



Dr. Hartert believed that 0. tristrami was the young of 

 0. owstoni, and that the name cannot be used for the bird 

 for the following reasons : — 



" Die Beschroibung wurde von Ridgway entworfen, der 

 Name von Stejneger vorgeschlagen ; der Autor kann daher 

 nur eigentlich Ridgway sein. Ich vermute, dass der leider 

 verloren gegawgene Typus eine junge 0. owstoni war, 



