34 THE VOYAGE OF H.M.S. CHALLENGER 



The figures in this table are, of course, in many instances, merely rough approxima- 

 tions, and they are not all based on one generic standard ; still, they are sufficiently near 

 the truth to give some idea of the proportions. Several of them are taken from Engler's 

 Versuch einer Entwieklunesgeschichte der Pflanzenwelt, and the others which are not our 

 own are taken from the various works enumerated in the Bibliography, p. 71. Unfortunately, 

 there are few published statistics relating to large continental areas. It is clear, however, 

 that the endemic element is equally as high in certain continental areas as it is in any of 

 the oceanic islands. Even if we reduce the indigenous element of the flora of St 

 Helena to the endemic species, and that would be going too far, there would be only 

 five endemic genera out of twenty-seven. In Juan Fernandez the proportions, after 

 deducting several genera, which are probably indigenous, are ten out of forty-six ; in the 

 Sandwich Islands thirty-nine out of 253, and in the Galapagos seven out of 164. In the 

 larger islands of New Zealand these proportions are unequalled, while in Australia the 

 endemic genera amount to considerably more than a quarter »of the whole, the exact 

 proportion being 30*5 per cent. In Madagascar. Mr J. G. Baker informs us, out of about 

 730 genera of flowering plants known to be represented in the island, about 100, or 13'7 

 per cent., are endemic. In Extratropical South Africa, the proportion of endemic genera 

 is certainly not quite so high as in Australia, for many of the genera which add largely to 

 the total of endemic species, such as Pelargonium, Oxalis, Phylica, Mes&mbryanthemum, 

 Erica, and Gladiolus, have a wider distribution. On the other hand, the proportion is 

 much higher than in Madagascar, so far as the flora of the latter has been investigated. 

 Engler (loc. cit., p. 190) gives the number of endemic dicotyledonous genera in Brazil as 

 215, with 124 more restricted to North Brazil, Guiana, and Venezuela. This is a large 

 proportion ; but the figures are not available for comparison, because we do not know the 

 total number of genera represented iu Brazil. Taking the whole of Tropical America, 

 there are, according to the same authority (p. 173), 1448 genera of dicotyledons; and the 



6 Taken in conjunction with Marion, the Crozets, and Kerguelen, this little flora comprises twenty-one 

 species of flowering plants belonging to eighteen genera, two of which are endemic ; and six of the species are 

 endemic. 



7 Five of the reputed endemic genera since reduced in Bentham and Hooker's Genera Plantarum. 



s The latest enumeration of Iceland plants is by Groenlund (Botanisk Tidsskrift, 2, iv. p. 36), who admits 

 317 flowering plants, and places a note of interrogation before all the names of plants of which he had no 

 authentic specimens. 



'•' A few additional genera and species have been described since the publication of Hooker's Handbook, 

 from which these figures were taken. Engler's numbers for New Zealand, including vascular cryptogams, 

 are: — Genera, 343, of which 20 are endemic [but a species of Carmichaelia has since been found in Lord 

 Howe's Island ; therefore, 19], and 1094 species, of which 671 are endemic. 



10 The total numbers of genera and species are probably too high : see remarks, Part II., p. 55. 



11 The number of dicotyledonous genera, according to Grisebach's computation, was 1030, comprising 1789 

 endemic species belonging to 540 genera, and 1866 not endemic species belonging to 763 genera. Of the total 

 number of endemic species given in the table, 849 are peculiar to Cuba, 275 to Jamaica, Trinidad 83, 

 Dominica 29, and St Vincent 12, the other smaller islands being almost destitute of endemic species, i 



