243 THE JOURNAL OF BOTANY 



In our view, both niutation and ci-ossing are true causes of new- 

 species. On the latter point one or two quotations may sei've to 

 indicate the trend of expert opinion. Mr. Kolfe {Orchid Review, 

 xxiii. p. 229), in an article on "The Mechanism of Heredity," says : 

 "A complete blending- of character " — i. e. in the case of a perfectly 

 successful ci-ossing — " would result in a batch of uniform secondary 

 hybrids, and Mendel himself a])preciated the fact when he pointed 

 out that hybrids in which the diverse elements were permanently 

 accommodated together reproduced themselves true from seed, and had 

 all the attriUutes of sj)ecies." Here it may be noted that, if a cross 

 affecting several characters is fulh' fertile, not only are there types in 

 which perfect and stable blending may occur (in the second hybrid 

 generation), but types in which it must and will occur. From that 

 point, granted continued fertility and equally favourable environment, 

 those types can never be swamped and lost. 



Mr. Bateson evidently regards mutation as the chief cause of new 

 species, crossing coming in as a further contributory cause. In his 

 Pi-esidential Address alread\' quoted he sums up the case for variation 

 as fundamentally due to an "accidental" change, i.e. niutation, of 

 germinal tissue, and then says, " Distinct types once arisen, no doubt 

 a profusion of the forms called sjiecies have been derived from them 

 by simple crossing and subsequent recombination. New species may 

 be now in course of creation by this means, but the limits of the 

 process are obviously narrow." 



This seems to us the true relation of the two processes. The^rs^ 

 deviation from the norm arises, from causes as yet hardly even 

 guessed, in some individual, as in the case of the Shirley poppy or 

 the Victoria plum. If now self-fertilization is possible, the strain 

 may survive or be artifici'illy preserved ; but if it is cross-fertilized, 

 it is only in the second hyl)rid generation that it may reappear in its 

 pure form, and continue as a species. Such a history may underlie 

 our acquisition of a "thrashable wheat," of which Mr. Bateson says, 

 " the original may have occurred once only in a single germ-cell." 



As to whether a similar or even identical mutation might arise in 

 several individuals simultaneously, in the same or in different places, 

 we are not aware of any dehnite evidence to guide us. It is the sort 

 of thing one might expect, seeing that no form can vary just any- 

 how. It has already, by being what it is, been cut off from a vast 

 number of modes of change. The group of organisms to which 

 it belongs has acquired a certain constitution, which may determine 

 that variations must be in some dehnite general directions. The 

 same consideration will make it reasonable to suggest, without at all 

 denj'Ing the initial " general " variability of all organisms, that 

 mutations in any given species are likely to follow in a serial order 

 which may appear pre-determlned, but is mainly self-determined. 

 From an oi'Iginal parent stock several such series might arise, follow- 

 ing a roughly parallel course. 



Upon the contention that mutations affect only single characters, 

 and that therefore a " mutation " can never become a " species," it 

 should be observed that since the change is first of all in the germinal 



