444 SCIENCE PROGRESS 



not arrive sufficiently rapidly to give proportionate increase 

 in the number of species occupying only area i or areas i and 2 

 Given that the rate of spreading of a species is rapid com- 

 pared with the rate of origin of new species, the endemics as 

 well as the non-endemics will show a larger proportion of widely 

 distributed species in ABCD than in A'B'C'D'. By comparing 

 the average area occupied by the species in Ceylon and in 

 New Zealand and the numbers of widely distributed non- 

 endemic and endemic species in these two islands, Willis has 

 confirmed these predictions. 



Another prediction made with the aid of the age and area 

 law and verified in the New Zealand flora is that if the species 

 enter ABCD at / instead of at AD, then if the area of dis- 

 tribution of any one species extends to the line cd it will pro- 

 bably extend to gh also, if to ab then to kl also, etc. Simi- 

 larly with endemic species, those produced early in the history 

 of the flora would extend over the greatest area ; being pro- 

 duced early they would have arisen near /, so that if the area 

 occupied by an endemic species includes area 1 or area 6, the 

 whole area of the species is likely to be large. This also has 

 been verified. 



Therefore, the age and area theory, having been subjected 

 to the test of prediction and verification, may be taken as 

 proved, and its universal application subject to modifying 

 causes being proved by published and unpublished work it 

 remains no longer a theory, and must be regarded as a law. 



It is necessary to point out that even with the age and 

 area law there may be endemic species which are older (in the 

 country) than some recent arrivals. For instance, an endemic 

 may be old enough to have spread all over ABCD, while an 

 immigrant may be so recently arrived that it has spread only 

 over AabD. This, however, does not interfere with the general 

 statement that the number of endemic species increases with 

 their rarity, while the number of non-endemic species decreases 

 with their rarity, nor with the conclusion that, therefore, the 

 endemic element in a flora is, on the whole, a younger element 

 than the immigrant element. Students of endemism are apt 

 to concentrate their attention on the peculiar relic endemics 

 (which undoubtedly exist) thus producing the impression 

 that the whole endemic flora is composed of such species. 



This part of Willis's work has recently been criticised very 



