METHODS OF STUDYING VISION JN ANIMALS 89 



the determination trie Lummer-Brodhun screen be fixed mid- 

 way between stimulus area and standard area, and that the 

 coarse adjustment (necessary to bring the intensity within the 

 range of the photometer bar) be made by using a rotating 

 sector (See p. 76) in the path of the Hght from the standard source, 

 and that the fine adjustment be made by moving the standard 

 lamp towards or away from the standard area until a balance 

 is obtained in the Lummer-Brodhun screen. When on the 

 other hand it is desired to set the stimulus lamps so as to give 

 a known illumination value upon the stimulus area the standard 

 lamp is set at the desired distance from the standard area and 

 the stimulus lamp adjusted until the desired balance is obtained 

 in the photometer. Every precaution must be taken to guard 

 the screen from extraneous reflected light. It is recommended 

 that this be done by the introduction of screens as described 

 by C. H. Sharp (Measurement of light intensity, "Lectures on 

 illuminating engineering," vol. i, p. 411). 



Size perception. — The recommendations under light perception 

 apply here also. In addition, it is desirable that the standard 

 plates be chosen so that discrimination shall at first be fairly easy ; 

 that after the ability to choose correctly on the basis of size 

 has appeared check tests be devised which shall exclude possible 

 peculiarities in the plates or in the illumination of the box; 

 that differences in the illumination of the two halves of the 

 experiment box, resulting from differences in the size of the 

 stimulus areas be carefully controlled so that the animal shall 

 not acquire a habit of attempting to discriminate on that basis ; 

 that the photometric values of each size stimulus be deter- 

 mined by balancing it against the light transmitted by a circle 

 of standard area (6 cm. diameter, 28,2743 sq. cm. area) in 

 fixed relation to standard source; that energy measurements 

 be made with radiomicrometer or selenium cell, as in the 

 case of light perception. 



Form perception. — In addition to the above recommenda- 

 tions we would suggest the desirability of making certain that 

 discrimination is dependent upon form alone; that plates of 

 same area and plates whose opening may be inscribed within 

 the standard 6 cm. circle be used in varied relations; that the 

 relative ease of discrimination by form and size be determined 

 in preliminary tests. 



