EXPERIMENTS ON SOLID AND GASEOUS EXPLOSIVES. 361 



With this instrument the work is very tedious, and no information is obtained as 

 to the rate of combustion of the explosive. The experience gained during the course 

 of the above preliminary investigation was, however, of the greatest use in the design 

 of the final apparatus. 



Recording Manometer. 



The requirements for a reliable recording gauge are somewhat complex. In the 

 case of gases, the explosive pressures to be dealt with range from 100 to 800 

 atmospheres ; in the case of solid explosives it was desirable to extend the research 

 to pressures of 2000 atmospheres, or above. The combustion of several gaseous 

 mixtures is much more rapid than that of the fastest explosives used in ballistics, and 

 the time period of a recorder designed for this work must, therefore, be exceptionally 

 small. 



Before passing on to a description of the instrument it may be well to recall in a 

 few words the law which governs the time period of vibrating bodies. 



If A represent the force required to produce unit deflection of the vibrating system, 

 W the weight of the moving parts, the time period will be 



We have, therefore, two variables at our disposal, namely, the weight of the 

 moving parts and the controlling force. The former must be made a minimum, the 

 latter a maximum. 



In most instruments where a short period is desirable, the strains to which the 

 parts are subjected are very small, and the desired result is obtained by decreasing 

 the size of all moving parts, and using, wherever possible, materials of low density. 

 This method is employed in the case of all oscillographs, telegraph recorders, 

 phonograph receivers, galvanometers, &c. 



In the present case, the instrument having to withstand pressures of 20,000 or 

 30,000 pounds per square inch, applied with extreme suddenness, strength becomes a 

 condition of vital importance, and steel is the only material which will withstand the 

 strain. We cannot, therefore, use materials of small density, neither can we reduce 

 the dimensions of the moving parts below a certain limit. 



It is thus evident that we must have recourse to the second variable factor to 

 secure the short time period which is necessary. As we have seen above, the 

 controlling force brought into play per unit length of motion must be as great as 

 possible. In other words, we must use the stiffest spring we can obtain. 



The stiffness of a spring will vary with the material of which it is made and with 

 its shape, increasing as the shape approaches more nearly to that of a solid bar 

 subjected to longitudinal strain. This bar can be made as short as may be desired 



VOL. CCV. A. 3 A 



