Cheeseman. — On the Flora of the Kermadec Islands. 161 



It is not quite clear whether Norfolk Island and Lord Howe's 

 Island are to be considered as remnants of this former extension 

 of New Zealand, or as evidences of volcanic activity since its 

 subsidence in whole or in part. Their flora certainly lends 

 some support to the first view, for, in addition to possessing con- 

 siderable affinity with that of New Zealand, it is undoubtedly 

 much more closely allied to that of North Australia and New 

 Caledonia than it is to that part of temperate Australia nearest 

 to them and situated in the same latitude. The large propor- 

 tion of endemic species also goes to prove that the islands are of 

 considerable antiquity. With respect to the fauna, the presence 

 of a species of Ocydromus in Lord Howe's Island, and of Nestor 

 in Norfolk Island, cannot but be regarded as highly suggestive. 



No doubt a considerable amount of probability attaches to 

 the supposed former extension of New Zealand to the north- 

 west. But an extension to the north-east has not nearly such 

 cogent evidence in support of it. In the first place, the sub- 

 marine ridge connecting New Zealand with the Kermadec 

 Group, and through it with Tonga and Fiji, is much narrower, 

 and, so far as can be judged from the few soundings taken, is 

 covered with much deeper water. This would put back the 

 subsidence of the land to a more remote period. But the flora 

 of the Kermadec Islands is composed almost wholly of plants 

 living in New Zealand or Polynesia, the proportion of endemic 

 species being only ^, against \ in Norfolk Island and Lord 

 Howe's Island. This undoubtedly demands that the connection 

 should be of recent date. Again, had there been land uniting 

 New Zealand with Tonga and Fiji, there should, as Mr. Wallace 

 states, be more community between the natural history produc- 

 tions of the two localities. Every botanist knows that the 

 Polynesian element in the New Zealand flora is small and 

 unimportant ; and, although some branches of the fauna are 

 perhaps more closely connected, even there the affinity is by no 

 means close. The scanty and fragmentary nature of the 

 Kermadec flora, and the still more scanty fauna, are not what 

 we should expect to find had the islands been connected with 

 better peopled countries ; while they are quite in harmony with 

 the view that they have received their inhabitants by trans- 

 oceanic migration. I must admit feeling much scepticism as to 

 the likelihood of any recent extension of New Zealand towards 

 the north-east. In all probability, if a land connection with 

 Polynesia in that direction ever existed, it had disappeared long 

 before the origin of the Kermadec Islands. According to Mr. 

 Percy Smith's observations, the two chief islands are mainly 

 composed of distinctly stratified pumiceous tuffs, evidently 

 deposited under water. In the absence of fossils, their age 

 cannot be precisely fixed ; but they must be assigned to a 

 comparatively recent period. 



11 



