C S. Hastings — Double Refraction in Iceland Spar. 63 



and be Q. Then R was ground so that it made equal angles 

 with the cleavage surfaces a b Q and a d g, and an angle of 

 60° with P. As it was desirable to make this last angle tol- 

 erably accurate in order to eliminate 

 all errors of the circle in a determi- 

 nation of the refracting angle, or, in 

 other words, so that a repetition of the 

 angle three times would bring the cir- 

 cle back to the same position within 

 the range of the reading microscopes, 

 the surface P was polished suffi- 

 ciently to yield a good reflection, 

 and then the angle at b was adjusted 

 until it was equal to that of a glass 

 prism known to be accurately 60°. Q 

 was determined in a precisely similar way. The three surfaces 

 were then polished to as close approximations to planes as pos- 

 sible. In this process most interesting differences in the phys- 

 ical properties of the surfaces were found, as might have been 

 expected. R worked almost as readily as glass, except that its 

 departure from flatness tended toward cylindrical surfaces in- 

 stead of spherical. It was not difficult to make P flat, but the 

 slightest carelessness in handling would produce tetrahedral 

 pits in it. The surface Q, being inclined only 15° to the di- 

 rection of cleavage, gave by far the most trouble, because it 

 did not seem possible to get it very smooth by grinding. After 

 carrying this process to its limit of accuracy, determined more, 

 perhaps, by the extraordinary thermal properties of the material, 

 than by purely technical difficulties in working, the faces were 

 cut away until only circular areas were left on the three prism 

 faces. These round faces were then modified, by methods 

 which would only have an interest for the practical optician, 

 until they were optically flat ; that is, until their departures 

 from their average planes was not more than a tenth of a wave- 

 length of light. The test of flatness was the colors produced 

 when white light was reflected nearly normally from the sur- 

 face brought closely in contact with a surface of glass known 

 to be plane. The diameters of the surfaces, in order of letter- 

 ing, were : 



2*8 cm., 2 - 8 cm. and 2.6 cm. 



(3) Spectrometer. 



The instrument with which the measures of the various an- 

 gles were made has some features peculiar to it. The circle is 

 of glass, 8 inches in diameter, and divided to single degrees, 

 except in the case of the first degree, and three others separated 



