EEPORT ON THE PRESSURE ERRORS OE THE THERMOMETERS. 19 



or less unsatisfactory. The very fact that I was dealing with thermometers whose bulbs 

 were protected from pressure, at once suggested an unprotected thermometer as some- 

 thing perfectly well suited to the purpose so long as the glass might be trusted to follow 

 Hooke's law. [I have since found that the invention of such an instrument, to be 

 used as an clatcrometre, is due to Parrot. 1 His investigation of the effects of pressure 

 is wholly incorrect, as it takes no account of distortion ; but the device, and the recogni- 

 tion of the fact that its indications are proportional to the pressure, are wholly his.] 



These instruments, which, like the thermometers, are fitted with a needle-index with 

 hairs attached, have only one defect, which is that they act Uke thermometers as well as 

 pressure gauges. That defect I managed to remove almost completely by the simple 

 device of enclosing in the bulb a closed glass tube which all but fills it. The Hquid then 

 occupies only a small space between the interior tube of glass and the exterior tube form- 

 ing the bulb, and is as ready as ever to give indications of pressure, while it is not in 

 sufficient volume to be more than slightly disturbed even by a serious change of tem- 

 perature. 



As will be seen in Appendix A to this paper, it is epiite easy, by comparing two 

 instruments of this kind in which the ratios of the internal to the external radius of the 

 cylindrical bulb are different, to find by trial through what range its indications are 

 strictly proportional to the pressure. Thus all the requisites of a perfect gauge, so far as 

 the experiments required, were met by this simple apparatus. That I have obtained a 

 sufficient accuracy in the graduation of these instruments is proved by the close agree- 

 ment between my results for the volumes of air at different pressures as measured by 

 means of them, with the volumes corresponding to these pressures in Amagat's table. 

 If Boyle's Law had been even approximately true for these high pressures, this mode of 

 verification would have been fallacious. It would, however, be easy to make an inde- 

 pendent verification, by sinking some of these instruments, each thoroughly imbedded in 

 a mass of lard (as a protection from shocks), to a measured depth in the sea. This idea 

 is worthy of consideration, especially if the gauge be made to register by means of a 

 silvered tube. The only probable cause of error in such a case would be the breaking of 

 the mercury column by a jerk, and to this all other forms are at least equally liable. 



1 "Experiences de forte compression sur divers corps, par M. Parrot " (Me'moires de V Academic Imperiale des Sciences 

 de St Petersbourg, 6me Serie, tome ii., 1833). The pages are headed " Parrot et Lenz," and it was by mere accident 

 (seeking in the Royal Society's Catalogue of Scientific Memoirs for a reference to Lenz's thermo-electric writings) that I 

 lit on the paper. I was much surprised at some of the statements it contains, till 1 found at the very end a footnote by 

 Lenz, in which he disclaims all responsibility for the writing of the paper, and for the conclusions drawn in it. 



