Cockayne. — Observations on Coastal Vegetation. 321 



slopes ill the east are covered with brown tussocks of Poa ccBSfi- 

 tosa, from amongst which rise np dark-coloured semi-pyramidal 

 masses of the liaiie Muehlenbeckia coniplexa, which has assumed 

 a shrubby habit, and stiff, erect, dull-green bushes, quite leaf- 

 less, of a species of Carmichcelia. 



Many other physiognomic peculiarities could be cited, but 

 it would lead to a general account of formations, here obviouslr 

 out of place. Other characters which affect the landscape 

 depend upon the blooming of various plants, and this is dealt 

 with to some extent in another section. 



6. The Small Coastal Islands ; Endhmism. 



Not the least interesting portious of the coast, if they may 

 be so included, are the islands situated at no great distance 

 from the mainland with which they have been connected at 

 some earlier period. Of such islands some have been inhabited 

 for a considerable time, while a few are still in their \argiii 

 state (14). In the former case the effect of man and herbivorous 

 mammals on a primeval vegetation \?- fairly easy to estimate,, 

 since their area is small enough for close observation. Dog 

 Island and Centre Island, which are to be treated of, are of 

 this class, while the Open Bay Islands and the yet unexplored 

 Solanders belong to the other category. The limited size of 

 these small coastal islands, too, permits detailed observations 

 to be taken as to the distribution of formations, the effect of 

 introduced plants, the change of edaphic conditions on the 

 vegetation, and the like. Finally, it is in such spots that relics 

 either of former vpgetation not existing elsewhere or of incipient 

 species might be expected.* As for the latter, a variety of 

 Brachi/come thomsoni named minima is reported by Kirk for 

 Dog Island (34 ; p. 260), and I mentioned a form of Veronica 

 elliptica as peculiar to the Open Bay Islands (14a; p. 371). On- 

 East Capo Island, in February, 1905, I collected a very robust 

 and fleshy plantain, which is either a new speciesf or Colenso's 

 P. picta. Other examples are afforded by the more distant 

 islands— e.g'., Veronica gigantea of Chatham Island, closely re- 

 lated to the common V. salici folia, and Acrena sanguisorhce var. 

 arifarctica of the Soutbern Islands. 



But it is as a haven of refuge that islands figure more conspi- 



* See also, regarding this question, Laing and Blackwell (35; p. 302 

 ct seq.). 



t This plant I have raised from seed in a soil containing no excess of 

 salt, and in a greenhouse. The early leaves of the seedlings are, however, 

 extremely tieshy. Of course, such fleshiness may not be shown in future 

 generations, but that it is present at all is suggestive and interesting. 



11 Trans. 



