316 POLARISATION, 



a perpendicular and a horizontal axis, and also for reading the 

 angles. 



Hotation round a perpendicular axis is most simply effected by 

 means of Welcker's rotating-plate arranged for centering and 

 graduated for determining the angles. Revolving stages of the 

 ordinary (Continental) construction are less convenient, because 

 the simultaneous turning of the body -tube changes also the posi- 

 tion of the analyser, unless the latter is prevented from moving by 

 the hand or by a special contrivance. Moreover, the rotation of 

 the objective lenses may give rise to deceptions where slight differ- 

 ences of colour are the criteria, for if the object is moved a more 

 or less evident change of colour frequently takes place, owing to 

 the anisotropy of the glass. It is therefore advisable, under all 

 circumstances, to rotate the object alone. 



The determination of the azimuth, in which an object has been 

 inserted with regard to the polarisation planes of the Nicols, is, 

 moreover, quite impossible, even with the aid of a graduated scale, 

 with the same degree of accuracy which is otherwise attained in. 

 angular measurements. A given diameter of the rotating-plate 

 may certainly be brought tolerably accurately into the diagonal 

 plane of the polariser, since the position of the latter is determined 

 by the edge of the prism, and similarly no difficulty is connected 

 with the adjustment of the analyser. The right-angled crossing of 

 the Nicols is accurately determined to within one degree by observ- 

 ing the greatest possible darkness in the field of view. The result 

 is quite different if the azimuth of a microscopic object, such as the 

 edge of a crystal, is to be measured by the aid of the graduated 

 scale upon the rotating-plate. It is here a question of producing 

 parallelism of the lines, of which only one can be seen with the 

 right eye, the other with the left, at one and the same time ; and 

 in this circumstance lies a very considerable source of error for the 

 novice. If, for instance, we look into the Microscope with the 

 right eye, and place a pencil or a rule upon the rotating-plate in 

 such a manner that looking with both eyes it lies parallel to, or 

 exactly coincides with, the object situated in the field of view, and 

 then look with the left eye into the Microscope, the supposed 

 parallel directions make an angle of from 6 to 10 if viewed again 

 with both eyes. It is clear that under such circumstances we can 

 no longer rely upon our own vision. The cause of this striking 

 phenomenon, which obtains also in stereoscopic vision, is of a 



