470 WRIGHT: PETROGRAPHIC MICROSCOPE 



(fig. 1, C). These improvements have greatly extended the use- 

 fulness of this device, which experience has shown to save an 

 appreciable amount of the observer's time in routine work. 



4. A device for use in the accurate measurement of extinction 

 angles. The measurement of an extinction angle involves two 

 operations: (a) the setting on the position of darkness, and (b) 

 the placing of some definite crystallographic direction, as a cleav- 

 age line or a line of crystal growth, parallel with one of the cross 

 hairs in the eyepiece. The precision of the second operation 

 depends largely on the quality of the crystallographic direction 

 which is used as line of reference; in case this is sharply devel- 

 oped there is no difficulty in setting with a precision of 1' of arc. 

 The first operation, on the other hand, depends on a number of 

 factors, one of the most important of which is the sensitiveness 

 of the eye of the observer and his personal equation. The eye 

 is sensitive down to a certain value (threshold value) below which 

 the field appears dark. Under ordinary conditions of illumina- 

 tion in microscope work there is a legion of 1° to 2° within which 

 the average birefracting plate between crossed nicols can be 

 rotated and yet appear to be dark. The common practice in 

 measuring extinction angles is to ascertain the position of maxi- 

 mum darkness on rotation of the crystal plate a number of 

 times to the right and stopping at the position judged to be the 

 darkest; this operation is repeated a certain number of times; 

 and similarly for an equal number of rotations to the left. The 

 average of these readings is then considered to be the most 

 probable position of darkness. In principle this method is ex- 

 cellent, especially if the half shade principle be introduced to 

 increase the precision of each setting; but in practice there is a 

 tendency, which is exceedingly difficult to correct, for the observer 

 unconsciously to attach special importance to the first reading 

 and thus to give it undue weight. 



In all measurements of this type, where settings of relative 

 intensity are made, it is essential for accurate work to eliminate 

 this element of the undue weighting of first readings and to 

 make each setting without knowledge of its agreement or lack 



