and its application to the observation of the Sky. 21 



of the bands. At an intermediate position, where the angle 

 between these directions is 45°, no effect is visible. Secondly, 

 it is necessary that the direction of the bands should be either 

 parallel or at right angles to the plane of polarization of the 

 incident light ; at an intermediate position, where the angle be- 

 tween these directions is 45°, no effect is visible. This latter 

 property may be expressed by saying that, with this polari- 

 scope, the transition from positive to negative polarization takes 

 place by passing through the value zero, positive and negative 

 being a conventional mode of expressing two directions of po- 

 larization at right angles to one another. 



Now although this property of Savart's polariscope has its 

 advantages, yet in the preliminary examination of regions of 

 faint polarization, where the existence and direction of polari- 

 zation are both uncertain, it is extremely inconvenient. In 

 cases where the polarization is so faint that it can only be seen 

 by glimpses after intense gazing under the most favourable 

 circumstances, it is difficult to map out the distribution of the 

 intensity where it is thus complicated with the question of direc- 

 tion ; for at the moment when one is certain that no bands 

 are seen, it may be that there is a polarization in a plane in- 

 clined 45° to the bands ; and the process of rotating the instru- 

 ment necessary to convince oneself of the real character of the 

 phenomenon adds much to the difficulty of the observation. 

 The polariscope, of which I proceed to give an account, has 

 been devised with the object of removing this difficulty ; it gives 

 bands nearly (if not quite) as distinct as those of the Savart ; 

 and the presence of the bands is independent of the position of 

 the analyzer, and of the plane of polarization of the observed 



light. 



The new polariscope consists essentially of two wedges of 

 quartz, the one right-handed and the other left-handed. The 

 terminal faces are cut at right angles to the axis ; and the in- 

 clined common surface of the wedges makes an angle of about 

 30° with the direction of the axis, which is also the line of 

 vision. 



In the larger instrument which I have constructed, the 

 length of the parallelopiped which is made up of the two 

 wedges is about 1*5 inch ; and the section is about "9 inch 

 square. In the smaller instrument (which was constructed 

 first, by way of trial) the angle of the wedge is about 45°, the 

 length between two terminal faces nearly half an inch, and 

 the dimensions of the area seen through are -8 inch parallel 

 to the bands, by *45 inch. This latter instrument was made 

 out of two small crystals bearing right-handed and left-handed 

 hemihedral faces, which I selected from my collection of such 



