INTRODUCTION 5 



embodied in the structure of the pith and bark are, in most cases, 

 least definite, and therefore of minimum value for differential 

 purposes. Furthermore the structural variations which may be 

 assumed to arise in conformity with the general evolution of the 

 species or genus are always most pronounced in the xylem struc- 

 ture, in consequence of which the latter acquires exceptional value 

 for purposes of relationship and phylogeny. Woody plants which 

 are found in the fossil state often show a complete absence of 

 pith structure, due to the operation of extended decay which may 

 have been initiated before the tree or shrub ceased to live. Much 

 more commonly fossil plants are devoid of bark. Instances are 

 on record, as in the case of Juniperus virginiana from the Pleisto- 

 cene clays, in which the plant is so perfectly and hermetically 

 sealed up as to permit of a perfect preservation of the bark as 

 well as of other portions of the stem, but such examples are com- 

 paratively rare. More commonly the extended maceration and 

 decay to which plants are subjected before silicification or calci- 

 fication occurs, involves a loosening and subsequent removal of 

 the bark, especially when the tree is subjected to such mechanical 

 action as is associated with its transport by water. It is therefore 

 obvious that any system of classification which would serve the 

 highest purposes for paleontological research must be wholly 

 independent of both bark and the pith. Another consideration 

 of importance in this connection relates to regional differences 

 of such a nature that different parts of the stem exhibit more 

 or less striking variations of structural detail. As these will be 

 dealt with somewhat in detail in a subsequent chapter, it will be 

 sufficient for the present purpose to indicate that the characters 

 upon which the generic and specific differentiations rest are 

 essentially independent of location, and it therefore matters not 

 whether the sample selected comes from a branch or the main 

 stem, or whether it is derived from the top, bottom, center, or 

 circumference of the latter, though as a matter of preference the 

 wood of a mature stem would be selected as furnishing the best 

 average conditions of structure, and therefore the greatest facility 

 in determination. 



