cf. C.A. 49, 12777i.Tablets were compressed during a short period, their weight variation was estimated as coefficient of variance, and the effects were studied of (1) the nature of the tablet base, (2) the size of particles to be compressed (or granulated sieve mesh), (3) the diameter of tablet, and (4) the speed of compression.The tablet bases used were (1) 50% CaCO3, 40% lactose, and 10% potato starch, (2) 100% lactose, and (3) 90% lactomin and 10% potato starch; the sieves tested were of 12, 14, 16, 18, and 20 mesh; diameters of 7, 10, and 13 mm.; and speeds of 76 and 42 tablets/min.Results showed no significant difference among the 3 tablet bases, or between the 2 speeds of compression.Decrease of diameter and of sieve mesh increased the weight variation of the tablets quadratically and linearly, resp.The recommended conditions for compression of tablets were sieves of 19, 16, and 15 mesh (or below) for 7, 10, and 13 diameters, resp.