410 



Prof. S. U. Pickering on the 



Although the effect of different pressures on the bulb did 

 not appear to produce irregularities such as would account 

 for all the difficulties experienced in getting two instruments 

 to correspond perfectly throughout their scales, it was thought 

 possible that the expansibility of the bulbs under pressure 

 might be influenced to a considerable extent by the tem- 

 perature of the experiment, and that this might produce dis- 

 cordance between two instruments which had been compared 

 with the same standard but at different temperatures. To 

 investigate this, the behaviour of No. '08 under pressure was 

 examined at 12° C. As it was impossible to keep the tem- 

 perature of the bath absolutely constant throughout the series 

 of experiments, the reading at any pressure P x was compared 

 with that at P by taking the mean of two observations at P, 

 made immediately before and immediately after the observa- 

 tion at P x . In the following Table the value of the constant 



Thermometer '08. Under pressure at 12°. 





R T -E 



p x -p. 



P x -F 



millim. 





-500 



001445 



-400 



0-01445 



-300 



0-01434 



-200 



0-01380 



+200 



0-01470 



J5 



0-01469 



+300 



0-01438 



5) 



001462 



+400 



0-01447 



+500 



0-01455 



+600 



0-01457 



Mean 



0-01453 





obtained from p _ p (Ex and R being the readings at P x 



and P respectively) is given in the second ' column, while 

 Pj — P is given in the first column. P was in all cases the 

 atmospheric pressure. The mean value of the constant (the 

 coefficient of apparent expansion) is 0*0145, a number abso- 

 lutely identical with that found at 0° (table ii., preceding 

 paper), which shows that small alterations of temperature, 

 such as would occur in standardizing delicate calorimetric 

 thermometers, produce no appreciable alteration in the rigidity 

 of the bulbs. The causes producing non-concordance of ther- 



