INTRODUCTION 1 1 



represented a great variety of material in common use for struc- 

 tural purposes. They included wood not only from the mature 

 parts of the stem but also from the center of the trunk, em- 

 bracing in some instances the structure of the pith and primary 

 wood zone. For these reasons they possessed special value, 

 inasmuch as they afforded an opportunity to determine the 

 extent and nature of those structural variations which I had 

 some reason to believe existed, as between the earlier and later 

 growths of the stem. In transmitting his material Dr. Fernow 

 stated that he had selected it "with reference to representing 

 typical wood, and it was not taken from butt logs, top logs, 

 nor branches or knots." It therefore represented exactly the 

 problems which would be met with in every-day practice. 



The results of these tests as at first obtained were far from 

 satisfactory. They clearly proved that the genera could be recog- 

 nized with ease, but for the species they made it clear that there 

 was need for a far more detailed diagnosis and differential key 

 than was at first supposed to be necessary. More searching 

 studies were made not only with respect to existing types but also 

 as applied to fossil species from the Devonian to the Interglacial. 

 These studies necessitated frequent recastings of diagnoses and 

 corresponding alterations of the analytical key. They brought to 

 light many important facts and relationships of the greatest value 

 from a phylogenetic point of view as well as from the taxonomic, 

 and they served to emphasize the fact that many of the more 

 detailed structural features of the pines in particular, hitherto 

 supposed to be of little or no value, were in reality of the greatest 

 importance. A final application of the test specimens under the 

 precise conditions which would obtain in ordinary practice showed 

 a verification of 91.5 per cent for all genera and species. In this 

 connection it may be of interest to note that the greatest sources 

 of error were to be found in the second section of the genus 

 Pinus, particularly in P. taeda, P. echinata, and P. glabra in the 

 order given, whence it appears that these species stand out as 

 the most variable of the entire Coniferales and, on the whole, the 

 most difficult to determine. This is in precise accord with the 



