Geologic St rue tut c — Habits of Petrels. 87 



to rather astonish an incautious intruder. Nevertheless, I subse- 

 quently ascertained, by dissection of specimens taken from the 

 nests, that both male and female birds take part in the duty of 

 hatching. The rock in this locality was almost completely sterile; 

 only three or four plants (stunted undershrubs) were found, which 

 eked out a miserable existence among fragments of crumbling 

 rock. 



The island is of volcanic formation. The cliffs which we 

 examined displayed a section, fully 1,000 feet deep, of various 

 layers of tuff, laterite, and scoricne, which, for the most part, stretched 

 out horizontally, and were intersected in every conceivable direc- 

 tion by dykes of basalt. In some places ridges or spurs of rock 

 projected like buttresses from the vertical cliff; and where we 

 landed the spur was composed of a vertical djke of basalt flanked 

 by a crumbling scoriaccous rock, which latter was being worn away 

 by the action of waves and weather much more rapidly than its core 

 of basalt. The columnar blocks of which the basalt was composed 

 were bedded horizontally ; i.e., at right angles to the plane of the 

 dyke, so that the appearance of the whole was strikingly suggestive 

 of an immense stone staircase. After a stay of an hour and a 

 half we were signalled to return on board, as Captain Maclear was 

 obliged to get under way, and accordingly at half-past twelve 

 we were sailing away to the southward, leaving this comparatively 

 unknown island as a prize for future explorers. 



In the course of this cruise we were followed by great numbers 

 of petrels, among which were the giant petrel {Ossifraga giganted), 

 the Cape pigeon {Daption capensis), and two species of T/ialassi- 

 droma (I think T. leucogaster and T. Wilsoni'). I noticed on this, 

 as on several subsequent occasions, that the little storm petrel is 

 in the habit of kicking the water with one leg when it is skimming 

 the surface in searching for its food. This mo\emcnt is usually 

 seen most clearly when the sea presents a slightly undulating 

 surface ; and when the bird strikes the water in performing a 

 slight cur\e in its flight, one can see that it is invariably the outer 



