162 PROCEEDINGS OP THE AMERICAN ACADEMY 



from the total weight the small amount of base assumed to exist as 

 bicarbonate, corresponding to the excess of CO^ shown by analysis, 

 and making also the very probable assumption that the calcium found 

 replaces the magnesium in the triple salt, we should obtain very 

 nearly the same results as before; but such a calculation would rest 

 on uncertain data, and add nothing to the strength of our general 

 conclusion. 



It is evident that the crystalline scale is a double carbonate of 

 magnesium and sodium united in molecular proportions with common 

 salt, and mixed with a small amount of impurity ; but the amount 

 of impurity is very small, considering the conditions under which 

 the scale is formed. That the chloride is combined, and not sim- 

 ply mixed, with the carbonate, is shown, not only by the definite 

 proportions, but also by the fact that the material is so slowly acted 

 on by water, which evidently acts as a decomposing agent, and not 

 solely as a solvent. The scale is then a definite crystalline product, 

 having a very interesting constitution not unlike that of several well 

 defined mineral species. 



III. — ON THE CONSTANCY IN THE HEAT PRODUCED 

 BY THE REACTION OF ARGENTIC NITRATE ON 

 SOLUTIONS OF METALLIC CHLORIDES. 



By Theodore W. Richards, A. B. 1886. 



• 



Having observed in some experiments made solely for practice that 

 the heat produced by the precipitation of silver chloride from aqueous 

 solutions of several metallic chlorides was directly proportional to the 

 amount of silver nitrate used, the following investigation was made in 

 order to determine whether this relation was really exact. 



A standard solution of argentic nitrate was prepared by dissolving 

 100 grams of AgNOg in one litre of water, and diluting to 1250 cc. 

 50 cc. of this solution diluted to just 250 cc. were used in each de- 

 termination (except once when twice that amount was used), and the 

 solutions of the various chlorides were made up so that 250 cc. would 

 contain a gram or so more salt than was necessary to precipitate the 

 quantity of silver nitrate used. 



