470 SORGHUM. 



It may be well to state, for the information of those interested in 

 this matter, that the National Academy of Sciences, incorporated by 

 Congress in 1863, is, by the fundamental law of its organization, con- 

 stituted the adviser of the government in all matters of science referred 

 to it for investigation by any department of the public service, and 

 has often acted in this capacity. 



The following is taken from the unanimous report of this committee 

 upon the results of the above investigation : 



The Analytical Methods Employed. 



The committee, after a careful examination of the analytical 

 methods employed by the chemical division of the Department of 

 Agriculture, find that they are entirely sufficient for the work to be 

 done. The details of the processes for the volumetric determination of 

 sucrose and grape sugar are fully exhibited. These methods have 

 been skillfully adapted to the character of the proximate constituents 

 of the complex juices to be analyzed, and are among the best known 

 t>) science.''^ 



These methods have been employed with precautions adapted to the 

 exigencies of the special problems for the solution of which the investi- 

 gation has been instituted. By a judicious system of checks and con- 

 trol, and by the reduction to the lowest limits of the personal error of 

 the observer, the accuracy and constancy of the results have been as- 

 sured as far as, in the present state of our knowledge, such end can 

 well be attained. 



The care with which the methods for the determination of cane 

 sugar have been tested, and the probable error determined, enlists our 

 confidence. The reserve with wliich the chemist has refrained from 

 accepting the results as conclusive, until, by repetition and variation 

 in the methods, he had exhausted the means at his command to prove 

 them to be erroneous, is in the true spirit of scientific research. 



The analytical work prior to 1882 comprises the enormous number 

 of nearly 4,500 analvses of forty varieties of sorghum and twelve varie- 

 ties of maize, covering all the later stages of development of the grow- 

 ing plant. Such an amount of analytical work as is implied in the 

 careful conduct of nearly five thousand quantitative analyses by the 

 most rigid system and subdivision of labor in the work — a system in 



* The limits of error, as shown to the committee from a considerable 

 number of unpublished determinations, sustain the conclusion that the 

 method employed for the estimation of cane and grape sugars was excep- 

 tionally accurate, and more subject to a minus error of 0.2 per cent on a 

 10 per cent solution of pure sugar, than to a plus error. 



