184 



JOURNAL OF FORESTRY 



These trial and error methods of constructing keys have certain 

 practical advantages. They are so simple that anyone who has an 

 elementary knowledge of plant anatomy can readily handle them. 

 Furthermore, the keys that are constructed by their use, although not 

 sufficiently accurate and reliable for certain scientific and economic 

 uses, are, nevertheless, of some value in the absence of more de- 

 pendable systems of identification and classification. 



A more direct and scientific method of attacking this problem is 

 to study the limits of variability of anatomical characters in different 

 plants, to endeavor to isolate and analyze the factors which control 

 or regulate this variability, and to attempt to formulate laws for fore- 

 casting the behavior of selected characters in any given species. How- 



Fig. 24. 

 Curves showing the variation in size of tracheary elements at different ages of 

 a tree. A , Variation in the length of the tracheids in a specimen of longleaf 

 pine. B, Variation in the length of the fibers in a specimen of shagbark 

 hickory. C, Variation in the length of vessel-segments in this specimen. 

 D, Variation in the diameter of the vessels. A, after Shepard and Bailey; 

 B, C, and D, after Prichard and Bailey. 



ever, the progress that can be made by this method must necessarily 

 be slow, particularly in the beginning, but the results secured should 

 be reliable and accurate. Furthermore, such fundamental and in- 

 tensive investigations should incidentally throw considerable light 

 upon certain problems connected with the seasoning, pulping, pre- 

 servative treatment, and chemical utilization of wood. 



