22 DERIVATION OF THE FLORA OF HAWAII 



There are abundant evidences of extensive subsidence in some parts 

 of Polynesia, e. g., the Marshall and Caroline Islands, and Pilsbry shows 

 that there is good reason to believe that this is the case also in the 

 Hawaiian group. It may be that there was a marked subsidence in the 

 latter, early in the Tertiary or possibly somewhat earlier ; but it could 

 hardly have been anterior to the later Cretaceous, as few modern angio- 

 spermous genera were differentiated at that time. In the upper Creta- 

 ceous and early Tertiary, however, many existing genera are known to 

 have existed. 



A very extensive elevation took place in Western America during 

 the early Tertiary, the great mountain systems of both North and South 

 America dating from this time. It is quite conceivable that a correspond- 

 ing subsidence may have occurred in the Pacific regions now occupied by 

 Polynesia. 



This theory would imply that the great volcanic masses of Hawaii 

 are superimposed upon a much more extensive non-volcanic base, which 

 formerly was above sea level. So far as the writer knows, there are no 

 data available as to the thickness of the volcanic formations, or any 

 knowledge of the underlying formations. It is therefore extremely un- 

 likely that any fossils representing the original flora will be found (assum- 

 ing an ancient continental connection), and we are perforce obliged to 

 rely upon a study of the existing plants for evidence as to the original 

 sources of the present flora. 



The great preponderance of Indo-Malayan elements in the Hawaiian 

 flora is sufficiently evident. Not only is the number of extra-American 

 genera common to the two regions very large (about fifty, including Aus- 

 tralasia), but a majority of the endemic genera are evidently derived 

 from Southern Pacific or Asiatic types. 



It may be assumed that these Indo-Malayan and Australasian genera 

 of Hawaii are descendants of an original continental flora to which have 

 been added the endemic genera derived from these, and a much smaller 

 element introduced after the isolation of the archipelago. Among the 

 latter are probably to be reckoned all the genera which are closely allied 

 to Western American types, such as Sisyrynchium and most of the Com- 

 positae. 



There are several genera whose nearest relatives are in South Amer- 

 ica, e. g., Nothocestrum, Kadua. The presence of these may possibly in- 

 dicate a connection of South America with a former antarctic con- 

 tinent and perhaps with New Zealand, and these genera may be the 

 remnants of once widespread types which have disappeared except at the 

 two extremes of their former range. Perhaps the Lobeliaceae of Hawaii 

 may be in the same category, as Guppy believes they are most nearly re- 



