166 



The Journal of Heredity 



that of other oceanic islands and of 

 other more or less isolated regions, 

 finally covering the flora of the whole 

 world. For this reason it seems de- 

 sirable to consider the island of Ceylon 

 here also, in the first place. Such a 

 process will at once put the meaning of 

 the general law in a clear light. 



The flora of Ceylon embraces 2809 

 species of Angiosperms, of which 809 

 are endemic to the island. Of these 

 latter about 200 are confined to very 

 small areas and about half of these 

 occur upon the tops of single moun- 

 tains or on small groups of mountains. 

 There are 1027 genera, of which 

 twenty-three are confined to Ceylon, 

 and among the 146 families this is the 

 case with six. Most of the endemic 

 genera are represented by one species 

 only, four by two or three and only 

 two (Doona and Stemonoporus) by a 

 larger number. The endemic types 

 are well marked Linnean species and as 

 such sharply distinguished from their 

 nearest allies, but a comparison of their 

 characters does not reveal any qualities 

 which could be regarded as useful un- 

 der the conditions of their localities. 

 On the contrary, it is evident that they 

 do not surpass their more widely dis- 

 tributed allies in this respect. More- 

 over, many of the endemics occur as 

 a very few individuals, say only a doz- 

 en or more, and the places where they 

 can thrive are often so small that it is 

 obvious that they can never have been 

 much more numerous. 



These facts lead to the conclusion 

 that the endemics are not, as a rule, 

 the remnants of forms which previously 

 had a wider distribution, but which are 

 now gradually dying out. In the tem- 

 perate regions of the northern hemi- 

 sphere there are a great many endemics 

 of this kind, and they are usually 

 designated by the name of relics. They 

 are the survivors of species which were 

 widely spread during the tertiary pe- 

 riod, as has been proved in many in- 

 stances by the study of their fossil 

 remains. But there is no good reason 

 to apply these results to the Ceylon 



endemics, nor to those of tropical and 

 subtropical regions in general. More- 

 over, the northern relics are rare as 

 compared with these. North America 

 has only about 400 such forms, where- 

 as the small island of Ceylon has 800 

 endemics, as has already been stated. 

 Brazil has 12,000 such species. New 

 Zealand and many other oceanic islands 

 are even richer in them, when areas 

 of the same size are compared. The 

 real relics may be estimated at one or 

 two per cent of all the endemics and 

 therefore they cannot be considered as 

 indicating the general condition of the 

 forms of this group. 



Willis proposes the conception that 

 most of the endemics of Ceylon and of 

 other similar regions are, quite on the 

 contrary, the youngest members of 

 their flora. They must have originated 

 on the same spots where they are 

 found now. And, since most of them 

 have the same means of dispersal as 

 their more common allies, it seems evi- 

 dent that if time were allowed, they 

 would spread like these. The smallness 

 of their area can therefore only be 

 explained in a satisfactory way by as- 

 suming that they have not had the 

 time to gain a larger territory. As 

 soon as this main point is conceded, 

 it is obvious that the size of the terri- 

 tory occupied by an endemic species 

 is to be considered as the result of the 

 lapse of time during which it has been 

 able to spread. Or in general terms 

 that the area covered is a measure of 

 the relative age of the species under 

 consideration. Age goes parallel to 

 area, and this, condensed into its small- 

 est form, is the meaning of the phrase. 

 Age and Area. 



Of course, this rule cannot be re- 

 stricted, either to the plants of Ceylon 

 or, on the other hand, to endemics only. 

 If true, it must embrace all species of 

 organisms and all geographical regions. 

 It must be the main law which gov- 

 erns the geological and geographical 

 evolution of the whole living world. 

 Distribution must go on everywhere 

 on the same general lines, it must be 



