22 SMITHSONIAN M ISCI-'.I.I.A N1':()US COI-F-EC TIONS VOL* /I 



then painted on the outside with melted wax consisting of equal parts 

 bee's wax and resin with a httle Venice turpentine. 



When a determination was to be made, the cap was screwed into 

 position, a wide rubber band was slipped over the junction between 

 cap and lock nut, and the outside of this rubber band was heated 

 with an alcohol blast torch. The result was a joint, for all practical 

 purposes, absolutely air tij^ht, which could, nevertheless, be dis- 

 mounted at once after iMilUn^- oil (he rubber l)and. 



■rill':oK\' OK Till'] ICXri-lKIMENTS IN VACUO 



The expressions for the velocity of the expelled gases are easily 

 obtained for the two types of motion of the suspended system that 

 were employed, namely, simple harmo'nic motion produced by a 

 spring, and direct lift. 



Siinple harmonic inotion. — Results obtained with simple harmonic 

 motion (slightly damped, of course) were naturally more accurate 

 than with direct lift, as it was impossible in the latter case to eliminate 

 friction. The theory, for simple harmonic motion, in which account 

 is taken of friction is described in Appendix B, page 60. The spring 

 was one niadc to specifications, particularly as regards the magnitude 

 of the force per-cm.-increase-in-length by the Morgan Spring Com- 

 pany of Worcester, Mass. Care was taken to make certain that 

 in no experiment was the extension of the spring reduced to such a 

 low value as not to lie upon the rectilinear line part of the calibration 

 curve. 



Direct lift. — The theory of the motion, in this case, has alreatly 

 been given under Appendix A. page 60. In this case it might be 

 assumed that a correction could be made for friction by multiplying- 



the displaccnionl s, bv some particular dccronicnt, ^ /, obtained in 



\ do 



the experimonls with simple harmonic motion, that might reasonably 



apply. This, as will bo shown ])oU)w, was found to give results in 



good agreement for the two typos of motii)ii, if the direct lift was 



about 2 cm. ; but not if it was much larger. It was found that very 



littlo frictitinal rosistaiioo was experienced when the mass M was 



raisetl In' hand, provided the axis of the gun were kept strictly 



vertical, but a very considerable resistance was experienced if the 



axis was inclined to one side so that the holders H, H' rubbed 



against the rods R, R. This sidewise pressure did not take place 



when the spring was used. It was also found that the trace upon 



the smoked glass was always slightly sinuous, with direct lift, and 



