THE BOTANY OF GOUGH ISLAND. 35 



of the branches are thickly encrusted with lichens. Tree-ferns grow in the rich 

 ground beside the stream, and reach a height of 4 feet or more. The beach is thickly 

 strewn with water-worn sterns of these ferns, which have probably been brought down 

 by the stream when in spate, carried into the sea, and washed up on to the beach. 

 Several species of ferns grow in nooks and crannies of the moist rocks, and apparently 

 obtain an easy footing in the relatively soft volcanic ash. Mosses are plentiful every- 

 where, and in the bed of the stream I "got several specimens of a liver- wort. 



The only plants in flower were Sonchus oleraceus and Apium australe and two 

 species of Rume.c, and the majority even of these were in seed. Gnaphalimn pyra- 

 midale bore withered flowers, and Phylica nitida and Empetrum nigrum, var. 

 rub-rum, were in fruit in a few places. 



I found no trace of any plants introduced for cultivation by the settlers whose ruined 

 huts we found. Beyond the huts was half an acre of ground beset with tree-stumps, 

 the remains, no doubt, of the native tree which had been cut down for firewood. 



The phanerogams and ferns of Gough Island, as might be expected, have proved 

 to be very similar to those of Tristan da Cunha. The present collection contains 17 

 species of phanerogams and 10 of ferns. Four of the 17 species of phanerogams are 

 almost without doubt introduced (Hijpocliceris glabra, Sonchus oleraceus, Rumex 

 obtusifolius, and Plantago major}. Of the remaining 23 species of Gough Island 

 plants, 20 are recorded from Tristan da Cunha one (Hydrocotyle leucocephala) is a 

 South American plant and two are endemic (Cotula, sp. nov., and Asplenium, sp. nov.). 

 Of the 18 species also recorded from Tristan da Cunha, four certainly, and probably six, 

 are endemic to the group. The mosses collected by me at Gough Island comprise 21 

 species, of which 11 are new. A discussion of the affinities of the Gough Island flora 

 as exhibited by the mosses is contained in Dr Cardot's paper in the present volume 

 (pp. 57-66). Excluding the 11 endemic species and one which is only geuerically 

 determined, but which is probably new, 9 species remain. Of these two are more or less 

 cosmopolitan, and four others are of wide distribution in the southern hemisphere, so 

 that their presence in Gough Island proves little from a geographical point of view. 

 Two species occur in both Gough Island and Tristan da Cuuha, and one in Gough Island 

 and Ascension. One would expect the relationships to Tristan da Cunha to be more 

 marked, and I agree with Dr Cardot that further exploration will probably prove this 

 to be the case. Otherwise the Fuegian affinities are most marked in the moss flora of 

 Gough Island, but I do not feel that our knowledge of that flora is anything like 

 adequate enough to justify our drawing from it any deductions of a geographical nature 

 regarding former land connections. For though, as Dr Cardot points out, 6 of the 

 9 extra-Gough Island species are found in Fuegian lands, their cosmopolitan nature 

 or wide distribution in high southern latitudes militates against their being used as 

 evidence in this respect. Nor must it be forgotten that a species of wide distribution 

 in high southern latitudes would most likely be found in Fuegian lands, owing to the 

 greater land area available there than elsewhere. 



