130 



White. 



if we are to believe what chemistry and physics teach us. A character 

 is always the result, of a chemo-physical reaction and not a continuous 

 entity existing from generation to generation. When a character is 

 handed on from generation to generation, it is formed anew each time. 

 On this viewijoint, historical knowledge is the only criterion by 

 which the newness or the oldness of a character may be established. 

 Dominance has often been suggested (de Vries, Castle, etc.) as a 



Fig. 29. Fasciated plants of P. sativum mnbellatum arranged like Fig. 28 in 

 order to show their close morphological similarity to those described by Blodgett. 



means of distinguishing between old and new, progressive and retro- 

 gressive characters. But since dominance itself is an expression 

 phenomenon, due to both environment and heredity, it must be 

 dispensed with as a criterion. So that in the end, the primitiveness 

 as well as the progressiveness of characters must be determined by 

 palaeontological evidence and logic. 



This brings us to a still more important point, i. e., to the question 

 of the validity of the work of many morphologists who have drawn 

 deductions as to which characters in certain gTOups are primitive and 

 which are progressive. These deductions are not infrequently made 



