14 



MR. J. H. GRINPLEY ON AN EXPERIMENTAL INVESTIGATION OF 



SECTION IX. The Correction of the Tliermometers. 



Before proceeding to describe the method of correcting the thermometers used, it is 

 necessary to repeat, as stated in the first part of this paper, that the research is based 

 on REGNAULT'S determinations of the relatipns between the pressure, temperature, 

 and total heat of evaporation of saturated steam, and hence the definition of 

 temperature assumed for the purposes of this paper is that saturated steam under 

 a certain pressure has a fixed temperature given by REGNAULT'S tabulated results. 



The method of correction to be described was adopted since it removed the 

 necessity of correcting the thermometers for the length of stems in the oil, and also 

 any error which may arise from any of the junctions not finally attaining the same 

 temperature of the steam or oil in which they are immersed. This method was to 

 correct the thermometers in position in the apparatus, using the thermo-j unctions, and 

 without making any alterations except to substitute for the orifice plate another plate 

 containing a large hole, which would not in any degree wiredraw the steam, so that 

 saturated steam at a known pressure would occupy the whole of the channel and the 

 steam chest ; the outflow of steam, or the velocity of steam through the apparatus, 

 could then be regulated by the valve on the low pressure side of the orifice. The 

 pressure gauge relied upon to denote the pressure in the steam was the one used to 

 give the pressure in the reservoir, its readings having been corrected as already 

 described. 



The operation of correcting the thermometers was then proceeded with as in an 

 ordinary experiment, the only difference being that now the steam is always saturated 

 in the apparatus. 



Experiments were conducted on six days with this object, and from the results of 

 these experiments the necessary corrections for both thermometers were obtained 

 throughout the range of temperature required. 



Hence it will be seen that the thermometers were used merely as instruments to 

 effect a comparison of two temperatures, one of which was the temperature of 

 saturated steam under a known pressure, and the other the temperature in the 

 wiredrawn steam, so that the basis of the whole method has been reduced to a 

 comparison between the temperatures of the wiredrawn steam and of saturated steam 

 under known pressures when flowing with approximately the same velocity through 

 the same portions of the apparatus. 



SECTION X. Results of Experiments. 



The experiments on this subject were commenced in January, 1897, the earlier ones 

 being chiefly devoted to determining the precautions necessary, and the best form of 

 apparatus to use (see Section II.). These experiments showed that the chief source 



