The intermittency effect has been investigated with monochromatic light at several wave lengths between 2537A and 4358A. Light from a quartz mercury arc was reduced in intensity by known amounts using a set of wire mesh screens, and the beam passed through a rotating sector and was focused on the slit of a quartz spectrograph. The Eastman Speedway plates showed only a very small intermittency failure, independent of wave length, while the Wratten Panchromatic and Cramer Contrast Process gave a very considerable intermittency effect, increasing with an increase in wave length. The increase of the latent image as set forth by Brush was found not to exist, nor were the sector scale and intensity scale curves found to coincide as indicated by Weber. In attempting to check Weber’s results it was observed that the intermittency failure is in the same direction as the reciprocity failure, but of less magnitude. Further evidence is cited indicating that there is an intimate relation between the intermittency and reciprocity failures. For moderately high frequencies of flash the intermittency effect was found to increase exponentially with the ratio of rest to flash, being almost independent of the frequency of flash, but at very low frequencies of flash the effect was dependent on frequency and not on ratio. If the plate were previously exposed, the intermittency failure was found to be reduced, and if this previous exposure brought the latent image up to the foot of the straight line portion of the characteristic curve, the intermittency effect was found to be almost eliminated. The association of the intermittency effect with the underexposure region was further indicated by the fact, deduced from the relative shapes of continuous and intermittent characteristic curves, that the difference between them increases up to the foot of the straight line portion and then remains constant.
© 1927 Optical Society of AmericaFull Article | PDF Article