NO. 2 MKTTIOD OF REACHING EXTRliME At/inUDKS 23 



Straight with the spring. The simple harmonic motion was, there- 

 fore, much the more preferable, but could not be used when the 

 powder charges were large. 



MEANS OF ELIMINATING GASEOUS REBOUND 



It should be remembered that the real object of the vacuum experi- 

 ments is to ascertain what the reaction experienced by the cliamber 

 would be, if a given charge of powder were fired in the chamber 

 many miles above the earth's surface. A container is therefore neces- 

 sary, which, for the purpose at hand, approaches most nearly a con- 

 tainer of unlimited capacity. A length of 3-inch pipe, closed at the 

 ends, is evidently unsuitable, because the gas, fired from one end, is 

 sure to rebound from the other end with considerable velocity, and 

 hence to produce a much larger displacement than ought really to be 

 observed. Moreover, any tank of finite size must necessarily produce 

 a finite amount of rebound, from the fact that the whole action is 

 equivalent to liberating suddenly, in the tank, one or two liters of gas 

 at atmospheric pressure. 



There are two possible methods for reducing the velocity of the 

 gas sufficiently to produce a negligible rebound : a disintegration 

 method, whereby the stream is broken up into many small streams, 

 sent in all directions (i. e., virtually reconverted into heat) ; and 

 secondly, a friction method, whereby the individual stream remains 

 moving in one direction, but is gradually slowed clown by friction 

 against a solid surface. 



As will be shown below, accurate results were obtained by the 

 first method, in what may be called the " cylindrical " tank ; and these 

 results weve checked satisfactorily by the second method, in what 

 will be called the " circular " tank. 



The cylindrical tank was 10 feet 5 inches high and weighed about 

 500 lbs. It consisted of a 6-foot length T, figure 4, and plate 8, 

 figure I, of 12-inch steel pipe, with threaded caps on the ends. 

 Entering the upper cap at a slight angle was the 3-inch pipe P, 

 4^ feet long which supported the cap C of plate 6, figure 2, and 

 plate 7, figure i. The 12-inch pipe was sawn across at the dotted 

 line To, so that any device could be placed in the interior of this tank, 

 or removed from it, as desired. The upper section of the tank was 

 lifted off as occasion demanded by a block and tackle. The two ends 

 to be joined were first painted with the wax previously described ; and 

 after the tank had be6n assembled, the joint was painted on the o'ut- 

 side with the same wax, W, and the entire tank thereafter painted 

 with asphalt varnish. 



