﻿BOTAMCAL GAZETTE 



more numerous and the cells themselves enormously larger. The 

 minor differences, as shape of berries and leaves, in which the 

 distinctions are of the vanishing order, are not at variance with the 

 mutation hypothesis, for they are also found when known mutations 

 are compared with their parent forms; for example, in Oenothera 

 rubrinervis the foliage characters are not sharply differentiated from 

 those of O. Lamarckiana, but are quantitatively separated. 



Finally, Spiraea tomentosa and 5. alba constitute a spurious pair 

 of species. In reality S. tomentosa is paired with S. alba in one 

 part of its distribution and with S. latifolia in another part; but 

 S. tomentosa has itself been derived from a tomentose ancestor 

 represented by a fossil form from Alaska. Hence its relation to the 

 other two species is more remote, and it only forms a pair with either 

 of them through the accident of their present distribution. 



It seems clear that the mutation conception can be applied with 

 advantage to the consideration of all such species relationships, 

 but, of course, crossing experiments and cytological investigations 

 provide the only final answer to the specific questions involved, and 

 it is to be hoped that such investigations will be undertaken, at least 

 in some of the genera discussed in this paper. 



The photographs which illustrate this paper were kindly taken 

 by Mr. C. H. Thompson. They are all from specimens in the 

 herbarium of the Missouri Botanical Garden, and all specimens 

 cited in this paper are from the same source. I am indebted to the 

 Director, Dr. George T. Moore, for the facilities provided for 

 making these observations, and to Dr. J. M. Greenman for much 

 kindly help in connection with the work in the herbarium. 



