628 



SCIENCE 



[N. S. Vol. XLVII. No. 1226 



must be younger, are all local, and are no more 

 numerous. I have shown in another paper^ 

 that one can not speak of adaptation in these 

 plants, other than the first adaptation which 

 enabled the families to live in running water. 



Neither of my critics makes any attempt to 

 explain why tmder my hypothesis one can 

 make numerous predictions about the compo- 

 sition and distribution of a given flora — ^pre- 

 dictions which as yet have always proved to be 

 correct on verification of the facts. For ex- 

 ample, in the paper on New Zealand, to which 

 they object, I predicted that the number of 

 endemics found in each zone of 100 miles from 

 north to south in New Zealand would show a 

 maximum at some point (sometimes two) and 

 a regular tapering away from this. This 

 proved to be the case for every one of the 91 

 families and 392 genera — a result hitherto 

 quite unsuspected, and which alone was almost 

 enough to establish age and area in a higher 

 rank than that of a mere hypothesis. In the 

 second of my recent papers mentioned above 

 I predicted that the species which reach the 

 islands outlying round New Zealand (Ker- 

 madecs, Chathams, Aucklands) would on the 

 whole be the oldest in New Zealand, and there- 

 fore very widespread there. This proved to be 

 the case, not only generally, but in detail, the 

 most widespread being those reaching all three 

 groups of islands, the next those reaching two, 

 then those reaching one, and finally those that 

 were confined to New Zealand, however far 

 they might range in the world in general. In 

 this paper I have likewise shown that the area 

 covered in New Zealand by the species that 

 also reach Australia, etc., goes, not with the 

 area covered in the world in general, but with 

 that covered in the archipelago of which New 

 Zealand is the principal part. This result 

 seems to me to exclude any explanation based 

 upon natural selection. 



In the same paper it was predicted that the 

 species endemic to New Zealand and the is- 

 lands roimd it would also be very widespread 

 in New Zealand. This proved to be the case; 

 they are even more widespread than the aver- 

 age of the species which range beyond New 



5 Proc. Boy. Soc, 1914. 



Zealand to Australia and the rest of the world. 

 This also is a fact that is quite inexplicable by 

 natural selection. 



It seems to me that a hjTpothesis that shows 

 itself thus capable of being used as a basis for 

 successful prediction deserves at least a very 

 careful investigation before being rejected. 

 One of the best proofs of its general appli- 

 cability is the fact that other workers are 

 beginning to apply it to the solution of various 

 problems. Breakwell has already worked out 

 the distribution of the grasses of Australia, 

 and shown that it agrees with age and area. 

 Small, in work now being published on the 

 Composite," has found it to confirm the evi- 

 dence of other lines of work in phylogeny, and 

 a third worker is applying it to the Legvimi- 

 nosse. 



In the last of the three papers mentioned 

 above I have shown that the orchids of 

 Jamaica are least widespread in that island 

 when endemic to it, more widespread when 

 also found in Cuba, and most widespread of 

 all when ranging yet further than Cuba. It 

 is also shown that the flora of the Hawaiian 

 Islands fits the age and area hypothesis, as 

 does the distribution of Callitris (Conifers) 

 in Australia. The ferns of New Zealand and 

 Hawaii are then considered, and shown to 

 obey the same law, while confirming previous 

 work on the greater youth of endemic species. 



Finally, in two further papers, as yet im.- 

 published, I have applied age and area in more 

 detail. In my first (a reply to the criticisms 

 of Professor Sinnott, who objects that age and 

 area will not explain the flora of New Zealand) 

 I have shown that New Zealand was peopled 

 with plants in all probability by two chief in- 

 vasions, one northern from Indomalaya, one 

 southern. In the second paper I have dealt 

 with the flora of Stewart Island (the southern- 

 most of New Zealand) and have successfully 

 made no fewer than 15 predictions about its 

 composition and geographical relations, bring- 

 ing out a number of points hitherto unsus- 

 pected or unnoticed. John C. Willis 



Claeendon Eoad, 

 CAMBKmGE, England 



8 See his review in Science Progress, January, 

 1918. 



