INTRODUCTION. 13 



ious common centre; but here even the fossil ma- 

 terial fails or becomes so scanty, unreliable and 

 ambiguous, that all systematic theorizing is un- 

 safe. Nevertheless we have to construct in im- 

 ag'nation the pedigree of the vegetable families, a3 

 branching off from a common centre ; even where 

 practically the undefined character of the more an- 

 cestral types renders it, so to say, very difficult to 

 trace the root of their pedigree. 



At the same time we easily recognize the more 

 modernized types, or as we may call them, the 

 higher evolutions, because their characters are 

 sharply defined and well established. 



The philosophical plan of classification would be 

 to begin with the ancestral types and follow them 

 through all their ramifications to the most modern. 

 It is a charming idea if it could only be done, but 

 it is about as feasible as,to unwind a ball of yarn 

 from the centre. 



The following are a few of the rules by which 

 an evolutionary system of botanical classification 

 may be constructed. 



1. The beginning of all organic life is in the 

 water. 



This axiom holds good equally in the animal 

 and vegetable kingdoms. The lowest or ancestral 

 types of nearly all the evolutionary lines of plants 

 which exist at present, are aquatic. It is, however, 

 difficult to use them in classification, partly on ac- 

 count of their present fragmentary condition, many 

 of the connecting links being extinct ; partly by 



