190 Mr. H. E. Schmitz. On the Determination of [Feb. 2, 



The last six results in Table VI are in good agreement with one 

 another, but are in excess of the result given by the method of 

 mixtures. The author has no doubt that the difference is due to the 

 water imprisoned in the ice coating. 



Table VII exhibits the results of a final series of experiments by 

 the ice method. These experiments were conducted so far as possible 

 in a uniform manner, differing slightly from that previously employed. 

 The time of transference was about 5 seconds, and during the trans- 

 ference the piece of metal was well shaken. The time in the calori- 

 meter was approximately 5 minutes, the time in the drying apparatus 

 J minute; In the case of nickel the ice-coated cylinder was too large 

 for the drying apparatus, and was dried by hand. 



Table VII. 



Specific Heats for the Range 20 to 100. 



In view of the uncertainty prevailing as to the effect on specific heat, 

 not only of impurities, but also of physical state in respect of hardness 

 and density, the author regarded it as of importance to determine the 

 specific heats at ordinary temperatures of the actual specimens used 

 in the low temperature determinations. He used the method of 

 mixtures, the procedure being, so far as the different circumstances 

 permitted, the same as in the low temperature determinations by the 

 method of mixtures. It was necessary in this case to make the time 

 of transference as short as possible. After numerous preliminary 

 trials, the arrangement finally adopted was a fixed steam-heater 



* In comparing the two methods, it should not be forgotten that the ranges of 

 temperature are not the same. The specific heat given by the ice method is for 

 the range 185 to 0, while that given by the method of mixtures is for the range 

 -185 c to +15. 



