encountered infrequently. For the Pinon shot the images will be approxi- mately hemispherical. The fireball edge will be only moderately sharp, fsy--e because of scattering in the 20-mile photographic path; scattering may ene also result in poor exposures of low contrast. Furthermore, natural clouds will probably obscure portions of the fireball, since it will grow to a considerably greater height than the normal low-lying cloud cover. These factors introduce the necessity for exercising some degree of judgment in measuring the fireball diameters. The orientation of the film on the comparator stage is such that the images will appear to be on their sides, with the horizon vertical. The image of the selected picture is centered, making use of the micrometers and the rotational adjustment, so that the base of the fireball coincides with the vertical diameter of the circular grid pattern. It will undoubtedly be necessary to adjust the illumination for both stages, and in some cases to switch the upper stage light on and off to aid in locating the fireball edge in reference to the grid. The observercan usually bracket the average diameter quickly by noting the greatest and the least circles of contact. These limits can be narrowed by running the eye around the intermediate circles, finally interpolating to arrive at a best estimate of the average or effective diameter. Experience ‘and practice are required to make this determination quickly and consistently, and without undue fatigue. It is always best, in our experience, to have each picture measured by several persons; in the event of disagreement between them a short discussion usually suffices to reach a joint reading agreeable to everyone. Before moving the film to read another frame, it will be necessary to measure with the micrometer the distance from the film-speed mark to the trailing sprocket hole. These measurements only need to be precise to about 0.03 inches, or one-tenth of a frame, anda visual estimate may suffice instead of a measurement. They will enable the calculation of the time for the particular picture. 25. Calculation of Total Yield Measurements and calculations are best worked out in tabular, systematic form. We usually record all measurements and calculations on vellum, which may be reproduced in any desired quantities easily and ree - 36 - a \ we