218 MR. W. F. G. SWANN ON THE SPECIFIC HEATS OF AIR AND 



The tap T!* having been adjusted in a preliminary experiment so that when T 2 was 

 completely opened the gas commenced to pass into the bell at about the required rate, 

 air was pumped from .the bellows through T 3 into the bell, T 2 being closed. After the 

 air had passed through the apparatus for some time, and the oil in the oil gauge had 

 settled down, T 3 was turned slightly so as to cause S to hover just a centimetre or so 

 above the end of the lever c. T 3 was then completely closed. S immediately started 

 to fall, touched the end of the lever, and at this instant the pen recorded on the 

 chronograph. T 2 was then opened, and air was allowed to flow out of the reservoir 

 and through the apparatus. T : was continually adjusted so that the brass piece S did 

 not move too far from the contact breaker c. After the pressure in the reservoir had 

 fallen to a value just greater than the minimum required to drive the air through the 

 apparatus at the required rate, T 2 was closed. S then descended, struck the lever, 

 and caused the pen to record on the chronograph. 



Now, obviously, between the two instants recorded by the chronograph, an amount 

 of gas passed through the apparatus equal to that which bad left the reservoir. It 

 will be seen that it was not necessary to turn T 2 at the instant tvhen S touched the 

 lever ; all that was necessary was that the temperature of the gas in the bell should 

 be the same at the start as at the finish,! and it was not necessary that even this 

 should be very accurately true, since the volume of gas in the bell was only a fraction 

 of that of the gas which had passed through the apparatus from the reservoir. By 

 this method everything was rendered entirely automatic, and the times could be 

 obtained as accurately as the chronograph would record them, i.e., to about 

 O'Ol second. 



The pressures recorded on the mercury gauge were measured by means of a 

 cathetometer, about half-an-hour being allowed in each case for the gas to take up 

 the temperature of the reservoir. Each reading of the mercury gauge was accompanied 

 by a reading of the temperature of the reservoir ; this reading was made with a 

 mercury thermometer which was graduated to 0'l G. and read to 0'01 C., its 

 indications being afterwards compared with a standard. Each reading of the gauge 

 was also preceded by a reading of the barometer. The pressure difference H on the 

 oil gauge was measured immediately after the discharge of the gas from the reservoir, 

 tap T 3 being opened immediately after T 2 had been closed, so as to allow the bellows 

 to pump air into the bell before the latter fell right down. The pump was kept going 

 during the whole of the time that T 3 was closed, an arrangement being made whereby 

 under these circumstances the pump could blow air into the atmosphere. Care was 

 taken to insure that the bell was quite stationary when the pressure difference on the 



* I afterwards found that T! was unnecessary. It was quite easy to regulate the gas from the start by 

 means of T 2 alone, and to keep the bell floating in approximately the same position throughout ; in fact, I 

 was very much surprised and pleased at the ease with which the flow of gas into the bell could be 

 regulated by means of this tap. 



t The precaution was taken of measuring this temperature. 



