the Island of Tristan da Cunha, 4c. 487 
from time to time against the sides of the dome, and gave us à 
transient glimpse of the peak at a height and distance that were 
by no means encouraging. After resting, however, for a few mi- 
nutes, we proceeded across the base of the dome, trusting that the 
cloud would be dissipated by the meridian sun; nor were we in 
this respect altogether disappointed. In the mean time, we found 
the ground as we advanced a perfect swamp, studded with tufts 
of small rushy plants, that gave way under the slightest pressure. 
Here also we had to pass through extensive patches of fern (Lo- 
maria robusta), the stems of which, like junks of old cable, trail 
along the ground, and cross and recross each other in such an in- 
tricate manner, that it required all our circumspection to avoid 
stumbling over them. Further on, the ground becomes more 
firm, but is perforated in all directions by the various species of 
Petrel, which resort in myriads to the island during the season of 
incubation, and burrow in the earth. The weaker tribes of these 
birds are devoured in vast numbers by the Skua gulls, which 
pounce upon them as they come out of their holes in the evening, 
and leave nothing but the bones and feathers to attest the havock 
made among them. 
The surface of the dome is furrowed on every side with ravines, 
which take their rise among the scoria of the peak, deepen as 
they descend, and open in tremendous chasms on the edge of the 
precipice. The various portions of the surface thus cut off in a 
great measure from all mutual communication, grow narrower 
and narrower as you approach the base of the peak, and dwindle 
at last into bare ridges of scoria, so sharp and so steep, that the 
wild goats of the mountain dare hardly venture to thread them. 
‘That ridge in particular over which we must either have passed 
or returned without accomplishing our object, is for at least fifty 
yards not more than twelve inches in diameter. The wind blow- 
ing 
