488 PROCEEDINGS OF THE AMERICAN ACADEMY 



6.83 



6.28 



7.24 



11.61 



12.56 



13.52 



2.67 



8.61 



4.57 



8.95 



9.90 



10.85 



It will be noted that the exposures Nos. 1, 2, 3, 9, 10, and 11, 

 as shown by the diagrams, last each for two units of time, and 

 may therefore be readily comj)ared with one another. No. 3 lets 

 through the greatest amount of light for any single slot (.58), and 

 No. 11 for a double slot (.79). These are therefore the best forms to 

 use, and if their lengths can be increased in proportion to their breadths 

 so much the better. No. 11 is the better of the two, but presents more 

 mechanical difficulties of construction when high speeds are desired. 

 With No. 8 the exposure is only one lialf that of No. 11, but its co- 

 efficient is somewhat less (.57). This is only a modified form of 

 No. 11, and with No. 7 gives the shortest exposure of any aperture 

 that uncovers the full size of the lens. The ideal practical shutter 

 will then have an aperture of the form No. 3, 8, or 11, as the case 

 may be, and as much lengthened as possible. 



(4.) Motive Power. — Now as to the driving force to be employed. 

 It has been found that, with a very sensitive plate (Allen and Rowell 

 extra-quick, or the Stanley) and a rapid rectilinear lens, an exposure 

 of Tsl^ sec. was sufficient to make a fair printing negative. The ideal 

 shutter should then give a minimum exposure of not more than ttJ^j 

 of a second and a maximum of perhaps ^ a second. Let us suppose 

 that the aperture between the lenses is one inch in diameter. The 

 slot, if single, must then be capable of moving with a maximum ve- 

 locity of two inches in Tji^ of a second. Theoretically this could be 

 obtained by the force of gravity alone only by a fall of sixteen feet. 

 But a shutter of these proportions is evidently out of the question ; 

 therefore, for rapid exposures one must resort to si)rings. These 

 are of three kinds, — india-rubber, metallic coiled, and metallic spiral. 

 The former are convenient and cheap, but cannot be relied upon 

 to give uniform results. Coiled springs, after they are wouud up 



