190 JOURNAL OF FORESTRY 



are silviculturally the most satisfactory to handle. In this connection, 

 a key of anatomical characters may prove a great help. 



I may add that the author's attitude as regards the great variation 

 of anatomical structures seems unnecessarily pessimistic, as being inim- 

 ical to key-making. Is this variation really greater than we find in the 

 use of any botanical key? Do not such keys represent merely the 

 average of conditions from which the single case may vary widely? 

 The main point is to have examined large numbers in order to strike 

 a reliable average and some knowledge of the amplitude of variation. 

 Whatever the immediate use or usefulness of a key based on anatom- 

 ical structure, it is proper to continue the detail of such investigations 

 for the eventual use that all scientific discovery is bound to bring. 



Professor Roth: 



1. Bailey's paper is primarily a statement of facts. They are in- 

 teresting and useful. 



2. Nordlinger in his key of 1881 says in substance: "In answer to 

 the Doubting Thomas who did not believe it possible to distinguish 

 woods by their structure, I published my first table in 1856." etc. 



3. Hartig's key was better, but not carefully worked out. Both 

 were ample for ordinary European practice. 



4. Bailey is right, but in discussing the use of structure study and 

 keys for the practice he is too modest and too "scientific." 



5. The anatomy and histology of wood is interesting and useful in 

 science and in practice ; only the latter interests me here. 



6. In practice we deal with : 



(a) Few species or genera which we wish to distinguish in the yard. 



(&) Locality of growth is usually known. 



(c) We deal with lumber, or the ordinary log stuft' and are not 

 interested in roots and branches, or even stump. 



id) We can avoid irregular,' apparently abnormal stufif, such as 

 wound wood, extremely fine rings, etc., just as we do not use the 

 lobed leaves of a Linden sprout in identification of linden, etc. 



Oak leaves vary more than oak wood in the logs of same tree. 



{e) In most cases we care nothing about species but are satisfied 

 with Genus (spruce, white oaks, hickories, elms, 'even ash, etc.). 



(/) In many cases we care nothing even for Genus ; Thuga, Chamae- 

 cyparis. Cypress. 



