554 



Mr. H. T. Jessop on Cornu's Method of 



fringes appeared centrally on the beam. The under side of 

 the cover-glass was optically polished, and had engraved on 

 it two scales at right angles by which any distortion produced 

 by the camera lens could be checked. Actually in all the 

 photographs taken, the distortion was found to be negligible, 

 and the graduations then merely served to give the scale of 

 the photograph. 



The source of light employed was a sodium flume from 

 a Meker burner, and this gave a bright, steady light and 

 showed well-defined fringes. 



Originally it was intended to measure the fringes direct by 

 means of a travelling microscope, but the first few sets of 

 readings showed discrepancies the source of which was not 

 obvious, and it w T as decided to use the surer method of 

 photographing the fringes and taking measurements on the 

 negative. 



Fig. 1 shows a sketch of the apparatus, with a microscope 

 in position to measure the fringes. In the experiments, 

 however, a camera clamped to a rigid stand was substituted 

 for the microscope. 



Fiar. L. 



£ 



P 



3& 



fesSJ&ak — Gl ° rs ^ 



u 



u 



^^ 



Flexnre Apparatus 



Method of Elimination of Errors due to Initial Curvature 

 of the Surface of the Beam. 



As a substitute for the cumbersome least-squares method 

 adopted by Straubel, the following method suggested itself 

 as being theoretically sound and much more convenient. 



If the bsnding-moment on the beam be reversed in 

 direction while its magnitude is kept the same, the theoret- 

 ical curvatures of the surface of the beam will be the same 

 in magnitude but opposite in sign, and any irregularity in 



