549 



be constant to within 1^ for such a displacement of the charge from 

 position. Thus for our set-up of k gaviges at 90° intervals around the 

 charge, a gauge displacement woxild have a more serious effect on the 

 average result than a charge displacement, but vas much less likely to 

 occur. Table X gives the data. 



Table X. Variation of damage with distance 

 (charge 2000 gm loose tetryl) 



(ii) Due to variation of charge level and tilt . Since in normal 

 practice the chsirge was tied into the center of a square frame by fo^lr 

 tie-lines of marlin (twine), it was of interest to know how much error 

 might be introduced by a change in level, or tilt of the charge. By 

 placing the charge above and below the level of the gauges it was shown 

 that a displacement of 1 ft. in a line perpendlcvilar to the plane of the 

 gauges changed the damage about k'f). 



Results on tilt of the charge varied widely but indicated that an 

 inclination of 30° of the charge axis did not result in more than a 5^ 

 variation in damage. 



(iii) Due to variation in gauge orientation . Gauges were twisted 

 so that the angle between the diaphragms and the charge-to-gauge axis 

 was 120°, 135°, and l80°. In the custcanary set-up the diaphragm is 

 perpendicvilar to this axis. When the diaphragm was parallel to the 

 cheirge-to-gavige axis (l80° angle), the damage was decreased 5^j in the 

 other cases, about 3^- 



(iv) Due to miscellaneotis variations . Some of the steel diaphragms 

 were slightly warped by the shearing process. These could be straightened 

 by a pressvire of a few pounds and the amount of warping was never greater 

 than about l/61t- in. (measured perpendicular to the diaphragm) . Warping 

 to this extent introduced no measxirable error. 



Variation in cap screw tightness (see Section II) was found to 

 introduce errors as large as 10^ with copper diaphragms and somewhat 

 smaller errors with steel diaphragms. Having one man tighten all cap 

 screws with a 10 in. wrench an equal amount seemed to be satisfactory. 



