70 



Mr. Hall in his investigation used glass frames (made of cylindrical glass rods) 

 of the shape indicated in Fig. 3. He deduced for them equations correspond- 



y^^5 



ing to (6), (7) and (8). He admits, however, that these equations are so compli- 

 cated as to be almost unmanageable, and that the correction is obtained more 

 easily by determining the constants of a frame by using frames of different length 

 and of the same diameter, and again of the same length but of different diame- 

 ters. It is very difficult indeed to make such frames, and to use them after they 

 are made. 



The chief objections to glass frames may be summed up as follows : 



The value of y, and hence the correction that must be applied to the 

 maximum weight in order to obtain the true film weight which measures 

 the tension, depends in a very complicated way upon the diameters of the 

 rods of the frame. 



This correction forms a considerable part of the total maximum weight 

 (see Table I.). Frames can not be made sufficiently rigid and less than 

 0.03 cm. in diameter. Hence the correction is at least ten per cent, of the 

 whole. 



The frames are difficult to make and they require delicate handling at 

 every stage. 



With cylindrical end rods the actual length of the film surface is uncer- 

 tain. 

 It occurred to me that these troublesome corrections and inaccuracies might 

 be partially avoided by using a different kind of frame. After experimenting 

 with frames of various materials, among wliich I may mention thin sheet glass, 

 platinum, aluminum and mica, I found tliat the latter offered decided advantages 

 over glass. The general shape of the mica frame is given in Fig. 4. The frame 

 is supported by a forked glass stem, and the method of using is exactly as with a 

 glass frame. 



/ii4 



