563 



deep that the correction which waa made for surfaces amounted to only 

 about 1%,* The averaged result for Kj^, A»A1, is in good agreement with 

 the value of Table in, 4.39. 



Further results of work with pressed tetryl charges are shown in 

 Tables V emd VI. These charges weighed slightly less than an ounce, 

 and may be considered as being in free water (surface correction much 

 less than 1%) . Note that both series of shots give a value of the 

 period constant which averages about 1% lower than that for the 3/2 lb 

 charges . 



The results of a slightly less reliable series of measurements in 

 free water are shown in Table VII. The depth in this case was measured 

 from the known length of cable let out, multiplied by the cosine of the 

 angle the cable made with the vertical at the siorface. As long as the 

 angle is small, this gives a good measure of the depth; however, since 

 the amgle in this series ranged up to 32°, the uncertainty in the depth 

 may be as much as 5%> The period constant is, as a result, somewhat 

 less precisely known in this series, the standard deviation of a single 

 obsei^ation being about 2%. The average value, 4.31, is also in this 

 case found to be lower than that for the l/2 lb charges. 



TABLE V 



FIRST PERIOD CONSTAOTS FOR PRESSED TETRYL IN FREE WATER 



Charge Weight: 0.0558 lb 

 Total Depth of Wrter: 85-100 ft 

 H - 55 ft 



Shot Number Period K]_ 



Tl 

 (msec) 



EDG 1 39.52 4.32 



2 39.94 4.36 



3 39.89 4.36 



4 39.90 4.36 



5 40.07 4.37 

 6 39.81 4.35 



Average 4.354l0.0O7 



* The correction for surfaces was made by using the equation in Ref. (5), 

 page 5. For convenience, this was transformed toi 



\ 



1.179 KwV3 j(y) 

 D Z,V3 I 



by using the relationship 5.78 K^ = rQ where K is the value of the period 

 constant. A value of K is put into the correction term as a first approxi- 

 mation, and if the value of K found by solving the equation differs greatly 

 from this, a second aporoxiraation should be made. The values of K reported" 

 in Tables IV, VIII, and IX were found by this method. The niomerical con- 

 stant 5.78 is not a universal constant, but may be used in all cases con- 

 sidered here without incurring much error. 



-7- 



