24 MH. 3. H. GRINDLEY ON AN EXPERIMENTAL INVKSTTfJATION OF 1 



saturation curve, the number of the experiment being also indicated. When the 

 experiment was made with steam from the large Lancashire boiler, the points repre- 

 senting the^> t relations observed are indicated by small circles. 



Suppose now that through the points thus obtained a series of curves be drawn to 

 represent the law of cooling from any initial condition. It will be found that if 

 a mean curve be drawn through all the points obtained by using steam in a given 

 initial condition, it will also be a mean curve for the points obtained from any single 

 experiment at the same initial pressure and temperature. The relative accuracy 

 which the experiments attain is clearly shown by the diagram, the greatest distance 

 from any point representing the p t relation in any experiment to the mean curve 

 through the points obtained in all experiments under the same initial conditions being 

 little, if any, greater than the expected error of experiment under that particular 

 p t relation. 



The use of these curves will facilitate the deductions from the experiments of the 

 actual law of cooling of the steam and of the variation of the specific heat at constant 

 pressure with both pressure and temperature. 



SECTION XVI. Summai'y of Results. 



The first point of importance brought out by the experiments is that so far as they 

 have been carried the steam never became what is known as a perfect gas. For had 

 such a condition been arrived at, the curve representing the pressure temperature 

 relation in the steam in that condition would have become parallel to the axis of 

 pressures, as the cooling would then have practically vanished. 



Coming now to consider any one of the mean curves drawn on Diagram 5, it 

 may be at once remarked that the portions of the curve which are of the greatest 

 interest were the hardest to obtain. For instance, when the difference in the two 

 pressures causing the flow through the orifice was very small, only a relatively small 

 quantity of steam passed through the orifice, which considerably increased the 

 difficulty of obtaining accurate temperature readings in the wiredrawn steam near the 

 saturated condition. Those results, however, which have been obtained and plotted 

 in Diagram 5 show clearly that the actual fall of temperature with pressure is 

 most accurately represented by a curve which commences at the point on the saturation 

 curve representing the initial saturated condition of the steam before wiredrawing, 

 proceeding for a short distance along the saturation curve, and then branching 

 off from this at an apparently definite angle, proceeding in a regular curve of small 

 curvature. 



The fact that for a short distance the curve approximately coincides with the 

 saturation curve is very important, as it appears to show that even after the steam 

 has been relieved of suspended moisture by a process of drainage, the law of pressure 

 and temperature in the steam follows the law of saturation very closely till saturation 



