26 THE VOYAGE OF H.M.S. CHALLENGES. 



Both Burchell and Melliss record the casting ashore of the seeds of various Leguniinosa? 

 (see Part II., p. 80) in St Helena, some of which occasionally germinate; yet, in con- 

 sequence of the nature of the shore being unfavourable, they fail to establish themselves. 

 Furthermore, not a single species of Leguminosse is native in any of the islands of the 

 southern Atlantic and Indian Oceans, from the Tristan da Cunha group to Amsterdam, the 

 Macdonald group, and the outlying islands of New Zealand, if we except the problematical 

 existence of Sophora tetraptera in the Chatham Islands (Mueller, Vegetation of the 

 Chatham Islands, p. 13). In New Zealand itself native Leguminosse are rare, and form a 

 very insignificant part of the vegetation : only five or six genera exist, and about fifteen 

 species, mostly belonging to the anomalous genus Carmichaelia. Passing on to Juan 

 Fernandez, we there find only the widely dispersed and variable Sophora tetraptera ; yet 

 Leguminosse abound in Tasmania and Chili (though absent from the Falklands and south 

 of Magellan Strait), and introduced species flourish wherever strewn on the islands. 

 This want of, or poverty in Leguminosse, is less apparent in the Polynesian groups, the 

 Galapagos, and the equatorial groups in the Indian Ocean. Nevertheless, endemic genera 

 are unknown, and endemic species are few or altogether wanting. There are no endemic 

 species either in Rodriguez or the Seychelles, and only one, Acacia heterophylla, is restricted 

 to the Mauritius and Bourbon ; and this is remarkable as being one of the very few phyllo- 

 dineous Acacias outside of Australia, and also on account of being so closely allied to the 

 endemic Sandwich Island Acacia Jcoa, that Bentham (Trans. Linn. Soc. Lond., xxx. p. 

 482) doubts whether the two forms should be admitted to specific rank. The more remote 

 Pacific Islands possess a few peculiar species associated with others of almost world-wide 

 distribution ; and eleven out of twenty-nine species found in the Sandwich Islands are 

 regarded as endemic by Brigham ; but this number is almost certainly too high. Thirty 

 species of Leguniinosse are recorded from the Galapagos, being fully one-tenth of the 

 number of flowering plants ; yet only seven of them are peculiar to the islands. 



The Gyninospermese (Conifers, Cycadacese, and Gnetacese) seem absolutely unrepresented 

 in the remoter Oceanic islands, as well as in many relatively near continents. Exceptions 

 are: Bermudas, one species of Juniperus; Azores, one endemic species of Juniperus; 

 Madeira and the Canaries, three species of Ephedra, two or three species of Juniperus, 

 and one endemic Pinus ; and in the southern hemisphere, Norfolk Island and New 

 Caledonia possess Gymnosperms ; while in New Zealand there are five genera and about a 

 dozen species, forming a large proportion of the forests. The outlying insular groups, 

 Chatham, Auckland, Campbell, &c, on the other hand, are destitute of Gymnosperms. In 

 the Mauritius there is none, while in Madagascar there is one endemic species each of 

 Cycas and Podocarpus. 



Another group of plants very poorly represented in, or entirely absent from many 

 oceanic islands, is the Petalifcrous Monocotyledons. They are quite absent from 

 Ascension, St Helena, South Trinidad, the Tristan da Cunha group, and all the islands 



