591.] 



Young's Modulus for Crystals, tyc. 



383 



The cement is applied to E and E' before the mirrors are placed in 

 gauge ; when they are in position, the beam is laid with one end 

 on E and the other on E', and warmed. When cool, the experimental 

 beam with the mirrors attached is removed from the gauge and laid 

 in the proper position on K 1} K 2 . 



On looking through the telescope T, two images of the scales S 

 will be seen side by side. One of these images being formed by 

 successive reflection from the four mirrors M 2 , MI, M 3 , M 1} in the 



ler named, and the other by reflection from M 3 and M 4 only. 



The course followed by the two sets of rays is indicated by the 

 >ws on the dotted lines in fig. 2. 



The appearance of the scales in the field of the telescope is shown in 



. 4. If MI and M 2 were absolutely parallel in a horizontal direc- 



FIG. 4. 



on, the scale images would at times exactly overlap one another, 

 irhich would make readings difficult. 



As long as the condition of approximate parallelism of the inter- 



3ctions of the pairs of mirrors is fulfilled, any shifting of one image 



the scale past the other is due to an alteration of the angle between 



pair or other of the mirrors, and to that alone, and since M 3 , M 4 



in a rigid mounting, any relative motion of the images is due to 



alteration in the angle between Mj and M 2 . 



The experiments were made by noting the relative motion of the 

 lages when force was applied to the beam by the central knife-edge 

 The course usually followed was to start with no load on K 3 , and 

 aving noted the relative position of the scale images, to move W 

 long D until the scales had moved relatively through one division. 



