METHODS OF ANALYSIS USED. 13 



country is on record. For this reason, the results obtained on 

 samples collected from all parts of the country are recorded for the 

 information of the public, to show the extent of the practice of 

 adulteration, and to furnish more complete data for the grading of 

 American turpentine and the preparation of specifications therefor 

 which shall be fair to the producer and also insure that the con- 

 sumer secures the article for which he pays. 



The samples have been classified into those secured within and 

 without the turpentine-producing States, and each of these classes 

 has been subdivided into three groups. First, samples taken from 

 the producer, or in the yards of primary buyers, but which the 

 primary buyers stated, and the evidence indicated, were in the 

 original containers in exactly the condition in which they were 

 received from the producer. Second, samples taken from the tanks 

 or other containers of primary buyers after they had accepted the 

 goods from the producers. Third, samples taken from the stock 

 of wholesale and retail druggists, paint and hardware merchants, 

 and other dealers. These classifications are made to show as clearly 

 as possible at what point in the turpentine industry adulteration is 

 most practiced. 



METHODS OF ANALYSIS EMPLOYED AND THEIR DISCUSSION. 

 GENERAL EXAMINATION. 



Owing to the nature of turpentine, there are quite wide variations 

 in its chemical and physical constants, and as these natural varia- 

 tions do not appear within limits to affect its suitability for technical 

 and medicinal purposes, rigid control of analytical procedure usually 

 is not insisted upon in the examination of the article, and, indeed, 

 great refinement of procedure is scarcely warranted. Thus specific 

 gravity and refractive index need not be expressed beyond the third 

 decimal, flash point and distilling temperature should be stated in 

 whole degrees, while the residues from polymerization and the 

 results of distillation should be given in whole numbers as percent- 

 age. For these reasons, and in order that the results might be 

 comparable with other published data, the usual procedure in the 

 examination of turpentine has been followed in the main in analyzing 

 the samples reported in this bulletin. It became apparent, how- 

 ever, as the work progressed that certain precautions and modifi- 

 cations in making such examinations were desirable in order that 

 concordant results might be obtained by different analysts, and con- 

 ditions affecting the results on adulterated and certain oxydized 

 samples be better controlled. The details of the modified methods 

 arc givon, beginning on page 26. 



