MARION ISLAND. 169 



evident that these mounds retain and store up a considerable 

 quantity of the sun's heat; and this fact probably yields a 

 partial explanation of their peculiar form, which is that of so 

 many otherwise widely different Antarctic plants, and of some 

 New Zealand Alpine plants [Raoidia, Hastia). No doubt power 

 gained of resistance to wind is one of the chief causes of 

 assumption of this form. 



The island being of such considerable area, and so short a 

 time having been available for the examination of its flora, no 

 conclusions can be drawn from the absence of certain plants, 

 such as Lyallia, which might have been expected to occur there, 

 since they occur in Kerguelen's Land associated with nearly all 

 those found. Although the few plants on such islands as these 

 are, as a rule, widely spread, yet some appear to be local and 

 somewhat scarce, as, for example, the Aspidium, which was 

 only found at the last moment, under the banks of one of the 

 streams. It is thus highly probable that several plants have 

 been overlooked, and amongst them possibly Lyallia. The nine 

 flowering plants collected in the island are all identical with 

 the species growing in Kerguelen's Land ; and the same is 

 the case with the Club-mosses. Of the ferns, two occur in 

 Kerguelen's Land, which has also two others not occurring here. 



Fifteen vascular plants in all were found in the Island of 

 Marion. 



Mr. Darwin suggests that Kerguelen's Land has been mainly 

 stocked by seeds brought with ice and stones on icebergs.* 

 The occurrence of Pringlea on Marion Island, as also on the 

 Crozets and Kerguelen's Land, probably points, however, to an 

 ancient land connection between these islands, which the an- 

 tiquity and extent of denudation of the lavas would seem to bear 

 out. It is difficult to see how such seeds as those of Pringlea 

 could have been transported from one island to another by birds ; 

 and these seeds seem to be remarkably perishable ; besides, the 

 distinctness of the genus points to a former wide extent of land on 

 which its progenitors became developed. The existence of fossil 

 tree-trunks in Kerguelen's Land points to similar conditions. 

 Sir J. D. Hooker, in the " Flora Antarctica," p. 220, expressed 

 * "Origin of Species," 6th Edition, p. 354. 



