the specific heats of certain solids. 271 



from foreign matter on the surface, particularly from vaporizahle 

 matters, which may, by being converted into vapor in passing from 

 the source of heat to the cold water, essentially diminish the tem- 

 perature, or, if in any considerable quantity, may aid in elevating 

 that of the water, and thus give a result too high. I have been 

 sometimes embarrassed by this source of error. In a series of eight 

 experiments, made by heating in a bath of oil on a given mass of 

 wrought iron, at a mean temperature of 236° Fahr., the tempera- 

 ture of the room being. 76° and that of the water at commencement 

 74.86° in a glass vessel of known specific heat, containing at every 

 trial the same weight of water, and measuring the temperatures every 

 time by the same thermometers, I obtained as the mean result 

 .12332, — the lowest being .12131, when the iron was immersed at 

 192°, and the highest .12920, when the metal was at only 190°. 



To ascertain how far this source of error would be obviated by 

 adopting a bath of mercury, I made eight experiments in the same 

 glass vessel, on the same piece of iron, and with all other circum- 

 stances corresponding to the former set, except that the temperature 

 of the metal at immersion was at a mean of 323f°, and of course 

 the specific heat, according to Dulong and Petit, ought to have come 

 out higher than in the other series, instead of which it was at a mean 

 of .12217, the lowest being .12119 at 338°, and the highest .12499 

 at 350°, the higher temperature giving the higher result. 



2. The second precaution relates to the condition of the ivater 

 used in the experiment. — The specific heat of saline solutions and 

 earthy mixtures being different from that of water, care should be 

 taken that only pure water be employed. That which has been re- 

 cently distilled should be preferred as it is less likely to be charged 

 with air than that which has been long exposed in open vessels. If 

 any considerable quantity of air contained in the liquid be suddenly 

 expanded it may rise to the top and escape carrying with it the por- 

 tion of heat which has given it so much enlargement of bulk. This 

 would cause an error in deficiency. 



3. TJie temperature and hygrometric state of the air in which 

 the experiment is conducted, require attention. It is obvious that if 

 we commence the experiment at a temperature below the dew point 

 of the air, the vessel will be accumulating moisture 6e/ore and du- 

 ring the experiment, and if it remain but for a short time at the in- 

 itial temperature before the hot body is immersed, the consequence 

 will be, that the latent heat of the vapor being employed in elevating 



I 



