PATTERN-DISCRIMINATION IN VERTEBRATES 335 



cular openings of St 2 about the axis of Y. All movement of Y 

 except in the direction longitudinal to the axis of Scr is minimized. 

 The spring Sp 1 by its constant action against Y, assures that all 

 resistance to the micrometer screw is made against one face of 

 the threads. The micrometer readings are therefore highly 

 accurate. The holder H, carrying the parts just described, is 

 fitted in a ring B, flanged as shown in the construction drawings, 

 so as to permit of free rotation about an axis normal to its plane 

 surfaces, within limits determined by the setting of two stops, 

 BS 1 and BS 2 , attached to the rim of B. A 360° scale is divided 

 on B, on which, from the indicator I, the direction of the test- 

 bands may be read. This feature is useful for making quick 

 and measurable changes in the direction of the test-bands 

 between two given settings. When it is not needed, H can be 

 secured so that the test-bands lie in any desired direction, by the 

 action of the set-screw SS, which bears against the convex 

 cylindrical surface of B. 



All the metal parts are of acid-blackened brass, except the 

 two springs, the pin P, the millimeter scale and the shaft of the 

 micrometer screw. These are of steel. Three weeks were re- 

 quired to make the mountings for the two instruments. The 

 gratings used are ruled 240 lines to the inch. The ruled lines, 

 which are filled with an opaque substance, are of the same 

 width as the clear interspaces. The gratings are obtainable 

 from Max Levy of Philadelphia at $18.00 the pair. 



As appears in Behn's formula, the grating Width being given 

 the separation of the test-bands varies inversely as twice the 

 sine of half the angle of rotation. The scale-reading on the present 

 instrument divided by 59 (half the distance in millimeters from 

 tip to tip of Y) gives the tangent of half this angle. In the present 

 work it is not necessary to correct for the cosine factor, since the 

 angle used is so small that the sine and tangent are practically 

 equivalent. 



In the present work the writer is using a cheap substitute 

 for the double photometer-box recommended by Yerkes and 

 Watson. Two instruments are used, so that the Yerkes- Watson 

 stimulus-adapter is unnecessary. In each compartment of the 

 double photometer-box is placed a wooden track on which rests 

 a carriage with a 60-watt tungsten lamp, protected by suitable 

 diaphragms. A screen of two superposed panes of opal flashed 



