264 LOTS^'s THEORV OF EVOLUTIOxX. 



dilTerences can only be explained b\' the assumption of slight 

 differences in the C(jniposition of their protoplasm, and I think 

 it is not far-fetched to assume that such physiological races may 

 form the material on which natural selection can act in the pro- 

 duction of new species. 



The almost incredible persistency of species through vast 

 geological ages and the enormous present distribution are proofs 

 that we cannot look upon species as mere creations of syste- 

 matists. I have referred to a few striking examples in 

 my paper just quoted, and I w'ill refer here only to 

 one case which I have recently studied with the aid 

 of an almost overpowering mass of material. Crassiila 

 miiscosa (L.) Roth (= Tillo:a luuscosa L.) is a well-known 

 annual first described from Central Europe. A large num- 

 ber of allied species were described from the countries round 

 the Mediterranean, Persia, India, Tropical Africa. South Africa, 

 Australia, Xew Zealand, Tasmania, South America, and a small 

 part of North America. A most careful study of thousands of 

 specimens has convinced me that they all belong to one specie.s. 

 Some specimens from Europe agree exactly with some from 

 South Africa, except that they dift'er in the number of floral 

 parts (a common variation in many Crassulas), yet the typical 

 form is absent from Tropical Africa, India, and Australia. 

 There are a number of ferns which grade into one another, and 

 there are, lastly, some from India, Tropical and South Africa 

 which stand by themselves, being perennials (one wdth a 

 tuberous root), which I shall for convenience' sake treat of 

 separately. Now here are thousands of specimens from different 

 parts of the world sticking persistently to a number of easily 

 recognisable characters. Is it not reasonable to suppose that 

 they all had a common origin ? Is it not reasonable to give them 

 a common name, and note simply the characters in which they 

 vary? There is no fact known to me that its wide variations 

 might be due to crossing. To my mind the " seed-pan 

 botanists," however valuable their work has been, are apt to 

 rather over-estimate their work and obscure the issue. They 

 seem quite to forget that the term " species " denotes an entity 

 in a logical sense only, and that it is just as easy to defend the 

 assertion that there is no such thing as an individual as to say 

 that there are no species except so-called " homozygotes." 



But to come back to Lotsy's theory. As stated above, I 

 miss in his paper any reference to the countless number of fungi 

 in which crossing is impossible. How^ have such species arisen? 

 Lotsy's theory cannot supply the answer, and, if it is thought 

 to be generally applicable, it must fall, unless such cases are 

 satisfactorily explained. 



It seems to me also that, if it were generally applicable, it 

 should show^ itself in two facts. 



Fi)^stly, the genera with the most numerous species should 



