68 



THE AMERICAN NATURALIST 



[Vol. XLVII 



generations, respectively, 3, 60, 49, 9, 11 and 21 times. Lata, oblonga, 

 rubrinrrris and schtlillan.-i appeared frequently. 



In the fourth generation, along with 14,000 lamarckiana plants, there 

 appeared 41 gigas, 15 albida, 176 oblonga, 8 rubrinervis, 60 nanella, 63 

 lata and 1 stint iUans, all bred from lamarckiana seed. In the fifth 

 generation, similarly bred from pure lam arch tana seed, among 8,000 

 lamarckiana plants were found 25 aZbtda, 135 oblonga, 20 rubrinervis, 

 49 nanetfa, 142 lata and 6 scintillans. In the fourth generation one 

 plant in 80 was oblonga. In the fifth generation one plant in 60 was 

 oblonga. DeVries himself says: " A species therefore, is not born only 

 a single time, but repeatedly, in a large number of individuals and dur- 

 ing a series of consecutive years. 



Oblonga differs from the parent species lamarckiana not in a single 

 feature, but in an elaborate complex of characters. The other mutants 

 likewise are distinguished from lamarckiana by a complex of characters 

 rather than by a single feature. 



The mutation can hardly be entirely fortuitous if, for several genera- 

 tions, out of every thousand offspring of pure lamarckiana parents, 

 there appear more than ten plants marked by the particular complex 

 group of characters which designate oblonga. Were oblonga demar- 

 cated from lamarckiana by but a single character, it would be remark- 

 able to find it appearing repeatedly and in such numbers. When we 

 remember that it is denned by an extensive series of characters differ- 

 entiating it from lamarckiana and from all other mutants observed, are 

 we not led to the conclusion that mutation in (Enothera lamarckiana 

 is not wholly fortuitous, but is to a degree predetermined; and that there 

 is some tendency to the production of the oblonga and other types in 

 numbers much greater than would be secured by purely fortuitous and 



Mutation in our most carefully observed instance, 

 therefore, is clearly determinate. There is in (Enothera 

 lamarckiana a tendency to mutate in certain definite 

 directions generation after generation. This trend to 

 mutation in certain particular directions is an example 

 of a condition within the organism which might decidedly 

 affect the course of the future evolution of this (Eno- 

 thera and its descendants. 



But we can safely go further. Not only have we evi- 

 dence that there exist tendencies to produce certain mu- 

 tants repeatedly, paleontological records show, I believe, 

 that there have existed trends toward an increasing 

 emphasis upon certain characters and that these trends, 

 in some instances at least, lay along lines that produced 

 no more perfect adaptation of their possessors to their 

 environment. 



