INSULAR FLORAS. 453 



and in South Africa, 30*5 per cent. Assuming that future 

 discoveries will to some extent raise the percentage in 

 Madagascar, and that the genera of South African and 

 Australian plants have even been unduly multiplied, it is 

 highly improbable that Madagascar will ever show so high 

 a percentage of endemic genera as the countries named. 



Comparing the specific endemic element in the Mada- 

 gascar flora with that in each of the countries named, we 

 obtain similar results. Baron points out the high per- 

 centage of endemic species in Madagascar, but he institutes 

 no comparisons. He says : "Of the 4100 indigenous plants 

 at present known in Madagascar, about 3000 are, remark- 

 able to say, endemic ". This includes the vascular 

 ferns. 



In 1888 the percentage of endemic species of phane- 

 rogams known to inhabit Central America and Mexico was 

 70*5, and since that date numerous genera and several 

 hundred new species have been described. Taking a more 

 restricted area in Central America, that is to say, the 

 South Mexican region, as defined in the Botany of the 

 Biologia Caitrali Americana, I estimated the specific en- 

 demic element to exceed 80 per cent. In Australia 84-9 

 per cent, of the vascular plants are endemic, and taking 

 West Australia by itself, 85 per cent, of the species are 

 endemic in that part of Australia. The extratropical South 

 African flora yields nearly as high a percentage. The per- 

 centage of endemic species in British India was found to 

 be considerably lower — barely 69, but within this area there 

 is an overlapping of several distinct floras. 



Assuming the foregoing figures to represent about the 

 same value in each case, and there is no reason to doubt 

 this, it will be perceived, that so far as mere numbers are 

 concerned, at least, Madagascar does not present a more 

 highly specialized flora than certain larger and smaller 

 continental areas, and that it is surpassed in this respect by 

 other areas. I believe, too, that these proportions would 

 hold good for the different regions of Madagascar. Baron 

 divides the country into three regions — an eastern, a 

 western, and a central ; and he gives the results of an 



