226 MESSRS. W. R. BOUSFIELI) AND W. ERIC BOUSFIELD 



the standard for the purpose of this series. In the third series the mean error of a 

 single observation is 0'0006, and the probahle error in the result, so far as it 

 depends on casual experimental errors, is only 0'0002. 



9. Point-to- Point Ezperitnents. General Description. For experiments subsequent 

 to the continuous-flow experiments the inflow and outflow tubes shown in tig. 1 were 

 removed, the inflow aperture through the obturator being stopped and the outflow 

 ujxTture used for a second thermometer. The general course of a point-to-point 

 experiment may now be described. A quantity of water was weighed into the 

 calorimeter sufficient to fill it to within about half an inch of the baffle -plate F. The 

 wooden lid D carrying the obturator, heater, thermometer, &c., was then put in place 

 and the requisite connections made. The current was turned on by the switch V a 

 (see fig. 4) through the by- pass Xi, and adjustments of the resistance XiX 3 were 

 made so that on turning the current by the switch V 2 to pass through the heater M a 

 the current at the commencement should be such as to balance the cells S. The 

 reading of the ampere balance A enabled this to be done with sufficient accuracy. 

 The timing was done with the help of the chronometer beating half seconds, the 

 switch being thrown over usually at the exact minute or half minute so as to turn 

 the current through the heater M a or to switch it off. It was then the business of 

 one observer to regulate the temperatures of the bath and obturator, so as to keep 

 pace with the temperature rise in the calorimeter, which averaged about 1 C. rise 

 per minute. The bath was kept at the same temperature as the contents of the 

 calorimeter ; the obturator was kept 10 C. higher, except that whilst the contents of 

 the calorimeter were below 10 C. the obturator was kept at a constant temperature 

 of 20 C. This observer noted down at frequent intervals the excess or defect of 

 temperature of bath and obturator, and also noted down the readings given by the 

 other observer. 



As an illustration the notes of Experiment 125 are given in full in Appendix D. 

 The excess or defect of obturator and bath temperatures is noted in the fourth and 

 fifth columns in tenths of a degree. The other observer was occupied in regulating 

 the current by means of the adjustable resistance X 2 so as to keep the galvanometer 

 at zero. He had also to read the graduations of the thermometer tubes of the 

 resistances M, and M 2 , which are noted in the second and third columns under ra 

 and 7n He read also from time to time the temperature of the watt balance 

 resistance Ro, so long as the watt balance was kept in the circuit (but it had been 

 abandoned before Experiment 125). In the final column are given the uncorrected 

 readings of the temperature of the contents of the calorimeter. At each stop three 

 readings are given which indicate a lag of about 0'005 C. 



''our standard thermometers were in use in this experiment, two being in use 

 simultaneously throughout the experiment. For the range from C. to 27 C. a 

 thermometer was used whose zero depression after heating to 50 C. is only 0'01 C. 

 No correction is made for this, but, assuming that the whole of it persisted at the 



