300 THE VOYAGE OF H.M.S. CHALLENGER. 



time before a plant was found ; subsequently a good many were met with, but grow- 

 ing in groups of only four or five. Some were found on the very verge of the shore, 

 within reach of the spray, and the rest on the banks of a small rivulet. The Cabbage 

 was mostly in full flower and bud, with sepals and anthers complete ; no plants were 

 found with seed at all ripe, and the last year's seeds were decayed. This plant at least 

 would appear to have a regular summer flowering-season, since Sir Joseph Hooker found 

 only the fruit at Kerguelen Island in the winter. 



Hymcnojihyllum tunbridgcnse var. wifooni, the well known British Fern, and 

 Polypodium (Grammitis) australe grow abundantly on the sheltered sides of the pro- 

 jecting rock-masses already mentioned, but are dwarfed and almost hidden amongst the 

 mosses ; they grow in greatest luxuriance on the damp banks of the stream. 



The mosses are in most striking abundance, 1 and, in some very wet places, form 

 continuous sheets over the ground many square yards in extent. Lichens are not in very 

 great quantity, except the incrusting forms, which are tolerably abundant on the rocks. 



An attempt was made to reach the actual upper limit of vegetation, but failed from 

 being commenced too late in the day. The ascent was up the bed of the small stream 

 already mentioned, which lay at the verge of one of the modern lava-flows, where it 

 abutted on a low cliff exposing a more ancient flow in section. The more recent flow 7 had 

 a very gradual inclination of not more than 8°. When the swampy moss-covered 

 ground, the uniformly dull green colour of which was relieved here and there by the 

 snowy plumage of the nesting albatrosses, had been left behind, the stream was found 

 to flow over an apparently very recent stream of black cellular lava, the ripples and 

 eddies in wdiich were still perfectly fresh, except in the very centre, where they had 

 suffered some slight abrasion ; there was no trace of any hollowing action on the part of 

 the water, the windings and little waterfalls being still determined by the original inequali- 

 ties of the solidifying rock. The lava was basaltic, containing much olivine. Close 

 by the bed of the stream rose several of the above mentioned red conical hills. One of 

 these, the highest within reach, consisted of a heap of loose scoriae disposed in layers, 

 dipping away on all sides at a regular and very steep angle. Few of these pieces of 

 scoriae were more than six inches in diameter ; and bad it not been for the occasional 

 clumps of moss which alone afforded a sure footing, the ascent would have been a matter 

 of considerable time. At the top was a perfectly conical pit, and slightly below the 

 summit, on the noith side, were three smaller and similar pits. The scoriae of which the 

 hill is made up consisted of a highly cellular red ground mass, with indications of augitc, 

 without, however, any perfect crystals being discernible. Besides the red scoriae, there were 

 some of a chocolate-brown colour, with frothy exterior and compact kernel. The form of 

 some of them resembled the almond-shaped bombs found in many volcanic districts ; but 

 none were noticed with the dense outside and highly cellular core so characteristic of the 



1 Tl:irty-one species were collected, five of which are described by Mr. Mitten as new in Bot. Chall. Exp., part ii., 1884. 



