METHODS OF STUDYING VISION IN ANIMALS 73 



likewise can be purchased from Jas. G. Biddle, Philadelphia, for 

 about $15. The Nernst filament is used by some experimenters 

 simultaneously as source and slit. Others mount it imme- 

 diately behind the slit (e.g., S, fig. 6). These two methods of 

 using the Nernst are unsatisfactory for continued use, for several 

 reasons which need not be entered upon here. The best method 

 of mounting is that suggested on p. 47. 



Slits. — The first slit in the optical system, S, fig. 6, is an ordi- 

 nary cheap form: (No. 'i 251 in Max Kohl's catalog.) It is sup- 

 plied through the Arthur H. Thomas Company, Philadelphia, 

 for about $8. The slit as it comes from the factory is greater 

 in height than is necessary for the present work. It is best 

 to limit the height to about 2 mm. by the use of auxiliary metal 

 strips at the top and bottom. These should' be knife-edged 

 in the usual way. The width of the slit is controlled by a micro- 

 meter screw. 



The double slit-wall, fig. 10, is made from our own design 

 by Wm. Gaertner, of Chicago, for about $45. If it is desired 

 the parts supported by casting Y, fig. 9, can be furnished com- 

 plete by the Johns Hopkins University. This includes the 

 casting itself, the whole of the double sHt mechanism, prisms, 

 prism stands, lenses, lens holders, etc. The cost will be 

 about $100, 



b. Lenses. — Suitable lenses of large aperture and short focal 

 length are very hard to obtain for the apparatus at a reasonable 

 price. Many of the large photographic houses in Boston and 

 New York deal in second-hand portrait lenses. In equipping 

 the psychological laboratories of Harvard, Chicago, Illinois, and 

 Johns Hopkins, very satisfactory lenses of this type have been 

 purchased, at prices ranging anywhere between $5 and $35. It is 

 not difficult to get portrait lenses of 3 "-4" aperture with long 

 focal length, but these must be avoided if an intense spectrum 

 is to be obtained. 



To obtain satisfactory condensers is especially difficult. Prob- 

 ably the best and easiest solution of the problem is to get Bausch 

 and Lomb to grind these lenses as single achromats. A single 

 achromat of 3^" aperture and 6" focal length is recommended. 

 The cost of this lens unmounted is S30. The two elements of 

 the condenser should not be cemented, but mounted with an 



