68 G. S. Hastings — Double Refraction in Iceland Spar. 



Of these the first and fourth, giving them equal weights, 

 yield 



Pb = U° 38'-26 



which, with the second and third, give 



44° 37'-29 



as the angle between the crystalline axis and the normal to a 

 cleavage plane. The last of the measured angles implies 



44° 36'-70 



for the angle between the axis and the normal to a cleavage 

 face. This value, however, rests upon two observations only 

 and cannot therefore be regarded as of great weight. We 

 may, perhaps, attribute to it a weight £ that of the value de- 

 rived from the other measures, whence the accepted value 

 becomes 



44° 3V-19. 



This value gives, for the direction of the axis drawn from P 

 inward, an inclination 



£=1'4"-1 



from the normal to P towards Q, and 



7] = 0' 12", 



i. e., 12" below the refracting plane of the prism QR ; they 

 can hardly be in error as much as \b". 



It is perhaps worth noting that the accepted value 44° 3Y /- 19 

 gives 105° 5' - 07 for the dihedral obtuse angle of the rhombo- 

 hedron at 20° C, which is practically the value accepted by 

 mineralogists. 



(6) Angles of deviation. 



Minimum angles of deviation were determined in each case ; 

 there are thus two angles for each prism-angle. The line 

 pointed upon was the more refrangible component of the D 

 line of the solar spectrum, except in the case of the extraor- 

 dinary image by the faces Q R, of which the dispersion was 

 too small to admit of easy separation, and, by mistake, in four 

 pointings on the double deviations for the ordinary image by 

 the same refracting angle when D 1 was observed on one side. 

 Care was taken to adjust the collimator, telescope and prism, 

 so that the axial ray passed through the center of the prism in 

 both positions for minimum deviation, i. e., right and left. 

 The lines of collimation were made at right angles to the 

 axis of the circle and to the refracting faces by means of the 

 plane glass plate and the collimating eyepiece. For observing 

 the spectrum a magnifying power of 31 was employed. Table 



