In the observation of lunar eclipses, penumbral eclipses in particular, significance is attached to the determination of the threshold at which the lunar crepuscular effects of the penumbra become visible or invisible to the unaided eye as the moon enters or emerges, respectively, from the shadow. During such eclipses the eye perceives the moon under conditions which are quite different from those in which objects are ordinarily seen. The following factors, among others, contribute to the complexity of the situation : ( 1 ) the moon subtends a small visual angle (about 30 minutes) ; (2) the luminance of the object is high; (3) the object's background is very dark; (4) the reflectance of the object is not uniform over its surface; (5) the shadow's radiant-energy gradient is sufficiently low to produce the effect of a diffuse boundary between the light and dark areas ; (6) the absorption of the earth's atmosphere is highly variable ; and (7) the duration of eclipses is often long enough to cause general fatigue in the observer. The psychophysical relationships involving factors 1 and 5 are of prime importance in the evaluation of the threshold and may be of considerable significance in the viewing of mosaics whose smallest units subtend similar visual angles. Data regarding the radiant-energy gradient in the penumbra have been obtained for few eclipses. Early results such as those of F. W. Very1 consist of measurements by visual photometry of the total integrated luminance of the moon during the successive phases of the eclipse. M. A. Dan j on,2 with a specially designed polarizing visual photometer which permitted direct com^ parison of the opposite limbs of the moon as it passed through