62 C. 8. Hastings — Double Refraction in Iceland Spar. 



by Stokes, and so difficult to explain by any plausible hypothe- 

 sis as to their cause, suggested a not too remote probability 

 that they indicated a physical reality. 



With these ends in view, all methods except those based upon 

 prismatic refraction were practically excluded. Again, since it 

 is impossible to get cleavage faces which admit of very accurate 

 determinations of their angles of inclination, e. g., to within a 

 second of arc, it seemed necessary to arrange the experiment so 

 as to be independent of such accurate determinations. The 

 method chosen, then, was to measure the various angles in- 

 volved in an equilateral prism of Iceland spar in which one 

 face was normal to the crystalline axis, the other two as nearly 

 equally inclined to the axis as possible, and all three refracting 

 edges as perfectly at right angles to the axis as practicable. 

 Such a prism restricts the range of wave velocities which can 

 be observed, but on the other hand, it enables us to find the 

 direction of the crystalline axis from the observations them- 

 selves by mere considerations of symmetry, wholly independ- 

 ently of all assumptions of the law of double refractiou. 



(2) Description of Prism. 



Since the accuracy of a determination of a refractive index 

 depends largely on the character of the prism used, and espe- 

 cially in this case of extraordinary refraction, it may be worth 

 while to describe the method employed to secure satisfactory 

 results. 



After selecting a good block of spar, a wooden model of the 

 largest prism of desired orientation which could be obtained 

 from the block was made. As this model represented the 

 cleavage faces as well as the prism faces, it served as a guide 

 as to how far any process of grinding should be carried. One 

 of the obtuse trihedral angles was ground down, so that when 

 the block rested upon this ground surface under a fixed tel- 

 escope nearly perpendicular to it, the images of a distant object 

 reflected by the three opposite cleavage faces could be brought 

 to the crosswires of the telescoj3e by merely rotating the block 

 on the ground surface. This admitted of securing a face, P 

 in the accompanying figure, very nearly perpendicular to the 

 crystalline axis. The limit of accuracy was restricted only by 

 the character of the reflection from the cleavage faces. The 

 size of the face was determined by reference to the model. 

 The next step was the formation of the surface, dbefg of the 

 figure, to serve as a base for the prism and a rough guide for 

 the other two faces of the prism. It was ground perpendicular 

 to P, and, by a process similar to that used in fixing the 

 direction of P, equally inclined to the cleavage planes a b Q 



