16g The Philippine Journal of Science nil 



If Doctor Willis could see Stenochlaena areolaris where it 

 thrives, he would surely be convinced that at least this plant 

 is specifically adapted to its peculiar habitat. It is epiphytic 

 on one species of Pandanus. Its adaptation to the very peculiar 

 conditions presented by its "substratum" is such that it can 

 grow nowhere else. Geographically, it is restricted, therefore, 

 to the few square miles where Pandanus utilissimus occurs. 

 I expect to describe this most extreme case of adaption at 

 greater length. For the point under discussion here, equally 

 valid evidence is presented by thousands of known species of 

 fungi. Almost every species of parasitic fungus has one host 

 species or a single group of host species, which it is able to 

 attack. Is it imaginable (not to ask for a demonstration) that 

 it is anything except specific adaptation of parasite to host — that 

 is, specific adaptation of the fungus plant to its own peculiar 

 environment — that lets the fungus- attack its host, but not the 

 infinitely more numerous other plants growing in the neighbor- 

 hood? Such a question seems to answer itself. 



Finally, the last of Doctor Willis's papers, so far published, 

 deals with the dying out of species, and seeks to show that the 

 extermination that must occur, if natural selection operates in 

 the usually supposed manner, does not occur or seems not to 

 be occurring among the plants of Ceylon. The body of the 

 paper is chiefly a restatement of the facts in the more extensive 

 paper in the Philosophical Transactions, the minor attention, 

 given to the question of dying out, earning one paragraph out 

 of eight in the summary. It may be, as Doctor Willis maintains, 

 that his figures do not furnish any reason to suspect species of 

 being on the downward grade. If there are no species dying 

 out in Ceylon, the number of species in the island must be under- 

 going a constant increase, and, indeed, this is probably happen- 

 ing. Increase in the number of species must result in a de- 

 creased average commonness — that is, abundance in indi- 

 viduals — of all species. Otherwise, the number of individuals in 

 Ceylon is increasing and this is not so probable. If no species 

 is driven to the wall, while the average number of individuals 

 of all species decreases, it is rather strange ; but it must certainly 

 be expected that, as many new species are introduced or evolved, 

 and some of these become very common, the disappearance of old 

 species will be comparatively slow. In a study of the flora of 

 four towns in southern Wisconsin, 10 in which particular atten- 



10 Shriner and Copeland. Deforestation and creek flow about Monroe, 

 Wisconsin, Bot. Gaz. 37 (1904) 139-143. 



