OCT. — MAE. 1859-60.] Rooms in Tropical Climates. 319 



the expansive action of the air at all, exists in the'Tdifference be- 

 tween the amount of increase in temperature on a certain compres- 

 sion, and decrease of temperature on the same quantity of expan- 

 sion in air ; a difference not practically sensible in low compres- 

 sions, but so very much so in higher ones, as to waste the me- 

 chanical power in producing heat, which can never be made up 

 for by the small amount of subsequent cooling. In any particu- 

 lar compression, a certain per centage only of the acquired heat 

 can be got rid of in the cooler ; unless therefore the decrease of 

 heat on expansion notably exceeds the per centage of heat of com- 

 pression still remaining, no absolute cooling on the initial state of 

 the air will be effected. Thus, let the heat produced by a certain 

 compression be 17°, and let 7 per cent, of that heat be got rid of 

 in the cooler, leaving therefore 5° still outstanding, the decrease 

 of temperature consequent on the same expansion will be 16°, and 

 the amount of cooling consequently 16 — 5, or 11°, and only 1°, 

 or 17 — 16 will have been produced. 



But let such a compression be employed as shall produce an in- 

 crease of heat of 1850°, 7 per cent of this being cooled off, leave 

 outstanding 555° ; and the decrease of temperature consequent on 

 the air expanding from that compression, being only 398°, leaves 

 outstanding still 157° ; or the air sought to be cooled is, after all, 

 heated to that great extent above what it was at first, and the enor- 

 mous amount of 1452° of heat, or 1850 — 398, have been uselessly 

 produced ; and this, however perfect the method or complete the 

 degree of cooling , and the carrying away the heat of the compressed 

 air may have been. 



The effect of this law of the different results of compression and 

 expansion being thus important, it becomes necessary to ascertain 

 its exact amount ;and as this is hardly possible to be done by experi- 

 ment, unless very great expense be incurred, recourse must be had 

 to theory ; and this may be carried out with perfect safety, when one 

 point in the scale has been so satisfactorily fixed by the Kinneil 

 experiment. Both Carnot's and Rankine's theories are. however, 

 rather difficult in their application, and depend on the specific heat 

 of air, a. quantity by no means well denned. A new theoretical view 

 has however been just produced by my friend, Mr. W. Petrie, C. E., 

 and has the advantage of being immediate in its result, and de- 



