RICHARDS AND JACKSON. — STANDARDIZING THERMOMETERS. 453 



was used, and was read by means of a powerful micrometer telescope with 

 all the precautions used in the most precise thermometry. Hydrochloric 

 acid was the substance chosen for the solution because its concentration 

 may be determined with accuracy and convenience in many ways. 



The third consideration determining the accuracy of the curve is that 

 just referred to, namely, the accuracy of the analysis. While evidently 

 for most purposes volumetric work would give the result quite certainly 

 enough, we nevertheless in the present case preferred to use the gravi- 

 metric method of precipitating the hydrochloric acid with an excess of 

 silver nitrate. The solution was drawn from the neighborhood of the 

 thermometer bulb, after thorough stirring, by means of a chilled pipette, 

 whose capacity had been determined by careful calibration. The solu- 

 tion was measured instantly, while it was still cold, and the precipitation 

 of the silver chloride was conducted in the usual manner. In the calcu- 

 lation of the results, the concentrations were computed in terms both of 

 weight of silver chloride from ten cubic centimeters of solution, and in 

 terms of the normality of the hydrochloric acid solution (i. e. the num- 

 ber of gram equivalents per liter calculated from the true volume cor- 

 rected for the contraction of the glass tube at the temperature of the 

 cold bath). The results were as follows : 



Freezing Points of Hydrochloric Acid Solutions. 



The curve deviates slightly from a straight line, in such a direction that 

 more concentrated solutions produce a slightly greater depression in pro- 



