191 8.] Recently published Ornithological Works. 507 



Alcedininae and Daceloninae alone occur in Australia, the 

 status of other possible subfamilies hardly affects the present 

 part; Mr. Mathews accepts both, while demurring to the 

 present method of defining the latter subfamily. 



The genus Alcyone is used for the species azurea, and five 

 subspecies are allowed (of which diemenensis used to stand 

 as a species) besides some seven that are extralimital, two 

 being new. They lead to a good deal of consideration, as 

 do those of the next species, Micralcyone pusilla, where three 

 are allotted to Australia, yorki being new. Several former 

 names are shown to be synonyms. 



Several new genera are next propounded for non-Australian 

 forms, viz. Cyanonyx for lepida and its nearest allies, Ceyc- 

 alcyon for cyanopectus, Aryyronyx for argentata, Ispidella 

 for leucogaster, and Ceycoides for madagascariensis. Therosa 

 is, moreover, accepted for sotitaria and meeki, while their 

 relationships are fully discussed. 



The part ends with Syma, where flavirostris is taken as a 

 subspecies of torotoro ; but further details are to be given in 

 the following part of this work. 



Murphy on Oceanites. 



[A study of the Antarctic Oceanites, By Robert Cusliman Murphy. 

 Bull. Amer. Mus. N. H. New York, xxxviii. 1918, pp. 117-146.] 



The main object of this paper appears to be to controvert 

 Mr. Mathews's suggestion that there are two forms of 

 Wilson's Petrel — one in the south Atlantic and one in the 

 north. This is, of course, a new suggestion, as it has been 

 generally acknowledged that Wilson's Petrel breeds in the 

 southern hemisphere and makes a trans-equatorial migra- 

 tion during the (northern) summer months to the north 

 Atlantic. Mr. Mathews, however, relying mainly on his 

 belief in the comparative non-migratory habits of Petrels 

 and other " Tubinares," is confident that the Wilson Petrels 

 of the north Atlantic breed in some hitherto overlooked 

 colonies in the West Indian or North African Islands. 

 This appears to us to be carrying a pi'iori reasoning rather 



