128 Mr Rankine's Letter to Prof. P. Smyth. 



quiescent cold air should then be turned on through the large 

 orifice, with just so little pressure as to drive it through the 

 tube at the same rate as the air from the jets previously 

 passed. If after each part has thus acquired a constant 

 temperature, the air is observed to become less cold as it 

 slowly rises in the tube, the amount of such effect should be 

 noted as an error to be deducted in the case of the air from 

 the jets. The longitudinal conduction and the radiation 

 obliquely through the tube, will cause errors in an opposite 

 direction to the preceding. To diminish these, the tube must 

 be either of very thin metal or of pine wood, and gilt inside. 



William Petrie. 



Letter on the Re-heating of Jets of Air, and the Relation be- 

 tween Temperature and Compression of the same. By W. J. 

 Macquorn Rankine, C.E. 



30 Great George Street, Westminster, 

 16th February 1851. 



My Dear Sir, — I herewith return to you, with my best thanks, 

 the two papers by Mr Petrie, of which you have been so kind as to 

 give me the perusal. 



With respect to the re-heating of air by friction, there can be no 

 doubt of the existence of the phenomenon, which has been established 

 experimentally by Mr Joule. Professor William Thomson, of Glas- 

 gow, is of opinion that this re-heating is the cause of the dryness of 

 a jet of high-pressure steam escaping from a boiler, which otherwise 

 would (as I have shewn) leave a portion liquefied to supply the 

 heat necessary for the expansion of the remainder. I think the ap- 

 paratus proposed by Mr Petrie well calculated to investigate the 

 laws of the phenomenon. As for his theoretical idea of the gra- 

 dual subdivision of currents, until they become converted into 

 what I have called molecular vortices, it is undoubtedly ingenious ; 

 but of its soundness I can as yet form no opinion. 



The part of Mr Petrie''s other paper, which is independent of 

 hypothetical ideas, is founded on the principle, that the pressure of a 

 gas, when its volume is varied, without its receiving heat from, or 

 communicating it to, external bodies, varies proportionally to a cer- 

 tain power of the density whose index is a number 7, greater than 

 unity, the temperature, at the same time, varying as the y—\ power 

 of the density, when measured from a certain absolute zero. This 

 principle is that adopted by Laplace to account for the fact, and 



