THE DYNAMICS OF DISTRIBUTION 475 



species) will, on the average and taking groups of allied genera, have a wider 

 distribution than a small genus, whether the distribution be measured by the 

 range of the genus as such or by the range of the most widely spread species 

 in it. But to return to Chapters VI and VII. In examining the flora of 

 Ceylon, Dr. Willis found that the range of endemics was usually less than the 

 range in Ceylon of species which extended to the peninsula of India. Since 

 some of the endemics were presumably species that had arisen after the 

 separation of Ceylon from the mainland, and since the average age of endemics 

 was, therefore, less than that of species shared by Ceylon and India, Dr. 

 Willis thought that these facts showed that age and area varied together. 

 Furthermore, he found that the average range in Ceylon of the species that 

 extended into peninsular India (but not farther) was less than the average 

 range in Ceylon of the species which extended into peninsular India and into 

 the outside world as well. These were the facts that formed the basis of 

 the conception. Casting round for another region on which to test these 

 views, the author then thought that New Zealand and its neighbouring 

 islands would furnish such a test. The Kermadecs, the Chathams, and the 

 Aucklands, were once part of a greater New Zealand. Therefore the New 

 Zealand plants found in these islands should be ancient species. Therefore 

 they should, on the hypothesis, be more widely spread in New Zealand, on 

 the average, than those New Zealand species which do not reach the islands. 

 Moreover, those found in only one group of islands should have a less wide 

 range than those found in two groups, and those found in all three groups 

 should have the widest range of all on the mainland. And it was found that 

 the figures were entirely in accord with aU these expectations : the range in 

 New Zealand of the species which do extend to the islands is markedly greater 

 than the range of those that do not extend to the islands, and the batches of 

 species arrange themselves in the required graduated series, according as 

 they occur in one, two, or all three groups of islands. And the author says : 

 " There is no conceivable reason why ranging to one or more of these little 

 groups of islands, and to any one of them . . . should make a species more 

 widespread in New Zealand than the average, unless it be the mere fact that 

 to have been able to reach the islands at all it must have been above the 

 average age in New Zealand, and thus have had more time in which to spread." 

 Here we have the essence of the matter. And the discussions in later chapters 

 largely depend upon the validity of this fundamental " Age and Area " 

 proposition. It should be said that Dr. Willis considers that his hypothesis 

 supports the theory of evolution by mutations as against the Darwinian theory 

 of the Natural Selection of continuous variations ; and it is for this reason, 

 of course, that we have the chapter by De Vries, though it is stated that 

 neither he nor the other writers above mentioned are necessarily committed 

 to all Dr. Willis's doctrines. 



The subjects discussed by Dr. Willis extend into many branches of 

 biological science, and in a brief space it is only possible to consider a few of 

 his most salient points. The most obvious comment is, of course, that species 

 and genera do die out, and that therefore there are diminishing ranges as 

 well as expanding ranges ; of course Dr. Willis does not ignore this point, 

 but he thinks that very few of the small-range species are, in fact, dying 

 species. If this be so, the total number of species in the world must of course 

 be increasing. It may be true that the number is increasing ; and it is also 

 certainly true that, since range is measured " by the most outlying stations," 

 and since the decay of a species is likely first to manifest itself by discontinuities 

 in distribution rather than by steady contraction of range, the dying species 

 will not show so much in small area groups as will the young species, even 

 supposing that the total number of species in the world is remaining constant. 

 But apart from this general and very obvious consideration, there are some 



