440 i Mr. J. Gill on the Dynamical Theory of Heat. 



ties, I could not rest satisfied until I had planned a repetition of 

 his principal experiments on this point in a new shape. The 

 only circumstance in his mode of proceeding about which I 

 could have any doubt was the fall of pressure from the varying 

 density of the air in the large reservoir to the constant lower 

 pressure maintained in the calorimetric part of the apparatus ; and 

 I imagined it might be more satisfactory, at least as a variation of 

 the experiment, if the heating and cooling worms were traversed 

 by a current of the compressed air from a mass contained in an 

 adjoining reservoir, and caused mechanically to circulate through 

 the apparatus without appreciable change of density or tension. 

 " To effect this I copied as closely as possible the arrangement of 

 Regnault's heating and cooling worms (PL VI. fig. 2) a, b, and 

 added the receiver, c, of sheet copper, 7 inches in diameter and 

 30 inches long, divided into two compartments by the air-tight dia- 

 phragm d. To this diaphragm is fixed the circulating-pump, e — a 

 brass model of a steam-engine cylinder accurately finished, with 

 but slight "lap" and "lead" on the valve, the piston being worked 

 by a handle on the fly-wheel fixed on the outer end of the crank- 

 shaft,/^ which passes through a stuffing-box on the side of the 

 receiver. The air is taken into the pump through the pipe g 

 (the steam inlet-pipe), and discharged into the other compart- 

 ment of the receiver through the pipe h (the exhaust) ; and it 

 thus circulates in the direction of the arrows, the pump being 

 worked steadily by hand at the given rate. By means of a con- 

 densing syringe attached to the tube k, the receiver is filled with 

 air of any required pressure above that of the atmosphere, indi- 

 cated by a pressure-gauge not shown in the figure. The gauge 

 i shows the pressure resulting from resistance to the free passage 

 of the air through the worms. At sixty revolutions of the pump 

 per minute the quantity of air at atmospheric pressure passed 

 through the worms corresponds approximately with the mean of 

 Regnault's experiments ; then at four atmospheres' pressure fif- 

 teen revolutions per minute would send through the worms the 

 same mass of air in equal time; and so on for other densities. 

 Suffice it to say, in a few words, that the results of careful expe- 

 riments with this apparatus completely confirm Regnault's results, 

 and I have experienced sincere satisfaction from the dissipation 

 of my long pending doubts on this very essential point. 



It was well remarked by Hirn that, as the capacity of gases 

 for heat is constant, or nearly so, the facts of the heating and 

 cooling of a gas by compression and expansion are inexplicable 

 and really without a cause, unless we allow a direct relation be- 

 tween the work expended or produced and the heating and cool- 

 ing of the gas. The same direct relation of cause and effect 

 ought to exist in the case of steam ; and as saturated steam can- 



