174 H. B. GUPPY, M.B., F.R.S.E., ON PLANT-DISTRIBUTION 



Such a scale of cliances is directly indicated in the foregoing 

 table. Here we perceive that in a sample number of the 

 families that are distributed in both the eastern and the western 

 hemispheres, about two-thirds of the tribes, 12 or 13 per 

 cent, of the genera, and one per cent, of the species, are common 

 to the Old and i^ew World. I have added this table, since it 

 gives the data on which this important inference is based. The 

 mere outlining of the numerous principles involved in its 

 columns would afford material for a paper of some length, so 

 that I will make no further reference to it here. 



It is strange that the old doctrine of multiple centres was 

 supported by a great Darwinian evolutionist. It was held by 

 Sir William Dawson, a leader of the opposing school, who in 

 one of his last works (Some Salient Points in the Science of the 

 Earth, 1894), observed that the upholders of the theory of 

 I^atural Selection would get rid of many difficulties of time and 

 space," if they would admit the possibility of more than one 

 centre. Like Professor Huxley, Sir William Dawson believed 

 in the differentiation of " generalised or synthetic primitive 

 types " ; and since they Ijoth held the doctrine of multiple 

 centres, they were fighting for the same cause, though oddly 

 leading contending factions. 



In this connection it is important to notice that in some 

 families where the monographer has worked on genetic lines, 

 similar to those adopted by Professor Huxley in the case of the 

 Gentians, the same possibility of the independent development 

 of plant-forms over the area of the primitive type presents itself. 

 For instance, with the Eriocaulacere, the type-genus (Eriocaulon) 

 from which Euhland traces the descent of all the other genera 

 of the family, is the only genus that is universally distributed.* 

 So also with the Juncacese, the sub-genus which is regarded by 

 Buchenau as nearest to the parent-type is widely spread over 

 the world. f Now, it cannot be pretended for a moment that 

 these forms, which come nearest to the original type of the 

 family, are indebted for their wide distribution over the area of 

 the said family to their exceptional capacities for dispersal. 

 Kather ought we to assume that they have been developed in situ 

 over the area originally held by the primitive type of the 

 family, that they represent the earliest stage of the differen- 

 tiating process, and that they have in their turn given rise to 

 various centres of differentiation from which the other more 



* "Eriocaiilaceje," by W. Euhland. Das Pfiaiuenreich, iv, 30 ; 1903. 

 t "Juucaceae,"' by Fr. Buchenau, Das Pflanzenreicli^ iv, 36; 1906. 



