RESISTAHCE OF CONIFEKS TO CKEOSOTE INJECTION. 13 



are platted in order of the magnitude of the absorptions. The gen- 

 eral tendency of the longitudinal penetrations is to increase as the 

 absorptions increase, but the curve is very irregular. The radial 

 penetrations show even less relation to the absorptions than do the 

 longitudinal penetrations, but inspection of the figure shows that a 

 few of the irregularities would be eliminated by combining the two 

 penetration curves. In other words, a high longitudinal penetration 

 and a low radial penetration may accompany a moderate absorption. 

 Figure 4 shows the relation of absorptions to the combined radial and 

 longitudinal penetrations in a slightly different manner. The ab- 

 sorptions are platted against a quantity obtained by adding twice the 

 radial penetration to the longitudinal penetration. 1 



Although the average curve indicates a constant ratio between 

 penetrations and absorptions, the variation of the individual points 

 from the average is again very marked. Reasons for this behavior 

 were frequently apparent upon examination of the treated sticks. 

 For example, when the creosote followed merely the summerwood 

 band of the annual rings, much less oil was required for a given pene- 

 tration than when the entire ring was saturated. 



GROUPING WITH RESPECT TO TREATMENT. 



While it is desirable to have classifications based upon elementary 

 characteristics rather than upon empirical results, it has been found 

 that the development of the resin canals and cells, which form the 

 most tangible structural differences between the conifers, is insuffi- 

 cient to grade these woods for preservative treatment. The follow- 

 ing method of classification, based upon both empirical results upon 

 air or oven-dry wood and structural characteristics, is therefore 

 proposed : 



A. Species in which the wood other than the resin structures treats with great 



difficulty or is impenetrable: 



1. Containing no resin ducts, Class I. 



2. Containing resin ducts, Class II. 



B. Species in which the wood treats easily: 



1. Containing no radial ducts, or radial ducts not easily penetrable, 



Class III. 



2. Containing easily penetrable radial ducts. 



a. In the heartwood, Class IV. 

 1). In the sapwood, Class V. 



The division is primarily between species which treat with diffi- 

 culty and those which treat easily. As a basis for this distinction 



1 Since the radial and tangential surfaces of the sticks treated in the absorption tests 

 amounted to 12 times the end surface, it might be considered that the penetrations in the 

 two directions should be combined in this ratio, assuming that radial and tangential 

 penetration is equal and that absorption is proportional to surface area. By actual 

 computation, however, this method seemed to give too much weight to radial penetration ; 

 and further trial showed that twice the radial penetration added to the longitudinal 

 penetration gave the most nearly constant relation to the absorption. 



