476 



THE BELL SYSTEM TECHNICAL JOURNAL, MARCH 1953 



quantitative loss such as those shown in Fig. 31. Now let it be assumed 

 that these hypotheses can be applied to treated wood, with full realiza- 

 tion that the application may err in the direction of oversimplification. 

 Bateman and Cislak accept the general principles of the 355 °C division 

 point between volatile and nonvolatile creosote constituents in debating 

 the theoretical aspects of creosote losses. ^^^ Rhodes^^ indicates some 

 actual loss in the fraction above 355°C in his test blocks, possibly in 

 part the result of oil displacement in the water phase of his tests, rather 

 than any increase that might occur as a result of oxidation. ^^ For the 

 purpose at hand in this paper it is convenient to use the relations shown 

 in the percent loss — residue increase curves in Fig. 31. 



Table XXXII represents a theoretical approach to what would happen 

 to the gross characteristics of the oils if, in some time period X, the losses 

 from creosoted wood treated to 8-, 10- and 12-pound retentions became 



Table XXXIII — Creosote Losses* from Southern 

 Pine Sap wood Blocks t 

 Calculations of residual fractions below 355°C; weathering wheel tests. | 



* Losses based on ether extraction. 

 t Blocks 3^^ X J'^ X 3 inches, 

 i See Bibliography, References 50 and 89. 

 § Assuming that all loss occurs in the fraction below 355°C. 

 Ij Retention by soaking in undiluted creosote. 

 The 12-week figures were calculated from extrapolations of the loss curves. 



