182 PROCEEDINGS OF THE AMERICAN ACADEMY. 



of iodine. With these three specimens of acid six determinations of 

 the specific heat were made. They were all essentially of the same 

 strength and the differences between the specific heats of the different 

 samples did not exceed the error of experiment, the determination of 

 sample A giving a value of 0.9210, three of sample B giving 0.9215 and 

 two of sample C yielding 0.9211 as the respective values of the specific 

 heat. 



Nitric Acid. 



This solution was prepared from the best obtainable nitric acid 

 through fractional distillation, using a quartz condenser. To avoid 

 the action of sunlight in forming oxides of nitrogen the operation was 

 performed in a carefully darkened hood. The middle third of the 

 middle third was reserved. This was diluted to the proper strength 

 with freshly distilled water. The concentrated acid was not abso- 

 lutely colorless, but the exceedingly pale tinge of yellow could not have 

 indicated an amount of lower oxides large enough to affect the determi- 

 nations of the specific heat of its dilute solution. The solution was 

 analyzed by titration and properly diluted. 



Perchloric Acid. 



A solution of this substance was made from a sample of acid of pure 

 material prepared for the determination of the atomic weight of silver 

 and lithium by Dr. H. H. Willard. The authors take pleasure in 

 expressing their thanks for this courtesy. The preparation is described 

 elsewhere.^ The concentrated acid was diluted to the proper strength 

 with freshly distilled water. Two analyses each showed the presence 

 of 5.788 percent of perchloric acid which corresponds to a concentration 

 of HCIO4.99.99 H2O. 



Sodium Hydroxide. 



The preparation and maintenance of the alkaline solutions in a 

 condition free from carbonate offers well known difficulties. The 

 purest sodium hydroxide of commerce contains appreciable amounts 

 of this impurity; and the contamination is not diminished by the 

 usual methods of handling. The frequency wuth which this fact is 

 overlooked will be discussed in another place. The available data 



5 Richards and Willard, Journ. Am. Chem. Soc. 32, 4 (1910). 



