A Note on the Geology of Gough Island. 259 



Madeira, with its bold jagged outlines and steep slopes 

 deeply trenched by ravines, but its high, sheer cliffs are 

 more like those of St Helena, only differing in being much 

 more clothed with vegetation. The steep slope of the land 

 appears to be even greater under water, for at a distance of 

 3 miles from the coast (to the S.-E.) a depth of 1322 fathoms 

 was obtained. 



These cliffs bear witness to the rapidity with which the 

 coast-line is being eaten back by the waves, at a rate seem- 

 ingly much greater than that of the sub-ferial denudation, 

 for most of the stream valleys have their courses sharply 

 truncated by the sea-cliffs, the streams ending in cascades 

 which pour out from the hanging valleys and tumble down the 

 precipices into the sea, forming very prominent features of the 

 landscape. Only Very few of the streams — notably that one at 

 the mouth of which a landing was eff'ected — have a deeper, 

 mature course, by which the water reaches the sea without 

 any terminal waterfall. These have excavated valleys which 

 apparently penetrate to the very heart of the island, and 

 should afford fairly easy access to the interior. 



The highest part of the island appears to be of the nature 

 of a ridge, cut into two slightly separated summits. JSTothing 

 indicative of a crater was observed. On the contrary, 

 through a telescope conspicuous horizontal terracing was 

 visible very near the summit. If these terraces are due to 

 successive sheets of lava, as seems probable, one must be 

 prepared to admit a former considerably greater extension 

 of the island. It may possibly even have been in continuity 

 with Tristan da Cunha : the fact that the latter has almost 

 certainly been continuous at one time with Nightingale 

 Island (one of its outliers), although now separated by water 

 over 1000 fathoms deep, makes this hypothesis at least 

 worthy of mention. 



Some of the shore cliffs also show, though less distinctly, 

 a succession of lava flows, and in some of the beds a trans- 

 verse columnar structure is visible, but neither of these 

 features was seen in the rocks near the landing-place. 



The stream at the mouth of which we landed might be 

 well compared with a typical small Highland burn, — there 



VOL. XVI. 2 B 



