3.20 Mr. 13. Loewy on the Behaviour [Mar. 4, 



gauge indicated a pressure of half an inch the mercury stood at 31°'2o ; 

 on readmitting the air it rose again to 32°. The experiment was repeated, 

 with precisely similar results ; and a correction was ultimately adopted, 

 corresponding to the varying pressures in the receiver, in order to reduce the 

 pendulum-experiments to the true temperature at which they were made. 



2. It was generally admitted that this apparent fall of the mercury arose 

 from a change in the capacity of the interior of the thermometer ; and the 

 physicists, especially the pendulum-experimenters who followed in General 

 Sabine's steps, never neglected this correction when their object was to 

 discuss the results of experiments made in a vacuum, and in the reduction 

 of which the temperature entered as an element. 



In the pendulum-experiments which were made at the Kew Observatory 

 in connexion with the Great Trigonometrical Survey of India (vide Pro- 

 ceedings of the Royal Society, No. 78, 1865), the thermometers used were, 

 before the discussion of the observations, subjected to independent experi- 

 ments, to determine their " vacuum-correction," which was found nearly 

 the same for each of the two thermometers employed, viz. o, 43. In these 

 experiments the two thermometers were suspended, together with another 

 (the latter enclosed in a sealed glass tube, and hence surrounded by air), in 

 the receiver, and their readings taken some time after the exhaustion, 

 sufficient to equalize its effect upon all three thermometers, bearing in 

 mind the fact that the thermometer in the glass case would take a some- 

 what longer time for showing changes of temperature than those without 

 such an enclosure. The arrangement of the experiments was precisely 

 the same as that originally adopted by General Sabine ; and the precaution 

 taken as regards the time of reading the different thermometers left no 

 doubt on my mind that the observed difference of o, 43, by which amount 

 the thermometers exposed to the effect of exhaustion were in every experi- 

 ment found to read less than that enclosed in a glass tube, gave the 

 required vacuum-correction in this particular case. It is also clear that in 

 this method of carrying on the experiment the refrigeration due to the 

 work done by the expanding air during the process of exhaustion will 

 affect all thermometers alike, and that consequently the residual difference 

 must be due to other causes. 



3. One point, however, was overlooked in these experiments, viz. to 

 wait a number of hours and then to take another series of readings, in 

 order to determine whether the effect of the removal of the atmosphere 

 upon the capacity of a thermometer was only transient or permanent. 

 Professor Oscar Meyer, in Breslau, was the first to call attention* to this 

 question. While making some experiments on the internal friction of 

 gases, he found that the primary effect of the exhaustion upon a ther- 

 mometer was quite in accordance with the observations of General Sabine, 

 but that after some time (for the thermometer employed by him, after 

 about half an hour) this effect entirely disappeared. Captain Basevi, who 



* Vide PoggendorfP's 1 Amialen,' vol. csxv. p. 411. 



