562 PROCEEDINGS OF THE NATIONAL MUSEUM. vol. xxvi. 



Necker, May 31. Necker is a dark, forbidding, rather precipitous* 

 rock of volcanic origin, attaining a height of 300 feet. It is about 

 seven-tenths of a mile long and is shaped like a rude lishhook, the 

 "shaft" extending nearh' east and west, the "bar))" being a rugged 

 peninsula extending toward the northeast and inclosing a rocky and 

 turbulent cove. The island is entirely composed of lava, mostly of a 

 sooty gray or black, with streaks of dull dark red through it. The 

 sides of the rock, though steep, are intricately terraced, especialh'^ on 

 the northeast point, where there are a series of shelves and all sorts 

 of knobs and craiuiics, making the island ideally fitted for the occupa- 

 tion of birds. 



The Necker Island tern was found to be fairl}^ common on Necker, 

 but not so abundant as St€7'na fidig hiosa or Gygis alba hittlitzL We 

 first noted them perched on the rocks near what was probabl}^ their 

 nesting site, and recognized them as the curious little terns we had 

 seen ofl' the French Frigate Shoals. A single Qgg is laid in a shallow 

 recess of the rock, but no nest is made; a few sticks and stray feathers 

 only are sometimes gathered. The eggs were not common. The 

 terns usually perch on the rocks somewhere near the "nest," and it 

 was difficult to find a bird on the egg. In fact only one was actually 

 flushed off the ^g^., and that by Mr. J. O. Snyder, on the north side 

 of the island. This made the identification of the few eggs we had 

 collected certain. Incubation was so far advanced that we could save 

 only two specimens. 



At Bird Island this tern is abundant. We were not able to land on 

 the rock, but saw many of the graceful birds from the deck of the 

 U. S. Fish Connnission steamer AVxitross as they flew back and forth. 

 The stomachs of those collected at Necker contained small silver}^ fishes. 



Its near relative Procelsterna cinerea is distributed over "Australian 

 and New Zealand seas, Lord Howe, Norfolk, itnd neighboring islands, 

 and Kermadec Group; also the islet of San Ambrosio, which is nearest 

 to the coast of Chile, but lies outside the cold Antarctic current,"" 

 and Eua, Friendly Islands.^ Procelsterna ceTulca is distributed over 

 "Central Polynesia: Paumotu or Low Archipelago, the Marquesas, 

 the Society to the Ellice Islands, the Phoenix group, and the Fannings 

 (Christmas Island), and a little north of the equator."^ It is of inter- 

 est, therefore, to note that Procelsterna saxatilis is separated from its 

 nearest relative by many thousand miles of ocean, and that a quite 

 different species, cerulea, ranges in between. 



The statement that some of the characters of saxatilis are "interme- 

 diate" does not in any way indicate that this species is a connecting 

 link. The form is quite separate ivom. the two other species, and 

 fortunately can not justify the introduction of trinomials into the 

 genus Procelsterna. 



« Saunders, Cat. Birds Brit. Mus., XXV. p. 135. ''Idem., p. 136. cidem, p. 134. 



