The difficulties involved in the direct measurement, by eyepiece-micrometer or from photomicrographs, of small microscopic objects such as living bacteria are discussed. The accuracy with which this can be done is limited by the numerical aperture of the optical system and the wavelength of light used. With visible light it is scarcely possible to determine the dimensions of an object more accurately than to the nearest 0.4µ. It also seems probable, from the nature of the diffraction pattern at the edges of images of objects of circular cross-section such as bacteria, that direct measurement of the width of the image will tend to give an underestimate of the true width of the object. An interference microscope enables thickness measurements to be made that are not subject to these particular limitations, because with it, the phase-change in the light passing through the middle of a bacterium can be measured very accurately. This phase-change is proportional to the product of the refractive index of the bacterium minus that of the mounting medium, and its true thickness. Two methods were used to determine the mean thickness of the living bacilli in a number of different cultures of Lactobacillus bulgaricus. With the first, the mean refractive index of the bacilli was measured directly by the method of immersion refractometry first used by Barer and Ross (1952), and phase-change measurements were made on the bacilli mounted in dilute saline. Their mean thickness was calculated from these measurements. With the second method, phase-change measurements were made on the bacilli mounted in saline and also mounted in protein solutions with refractive indices ranging from 1.365 to 1.376; and, from these, both their mean thickness and their mean refractive index were calculated. The phase-change measurements were made with a Smith interference microscope and half-shade eyepiece (manufactured by Messrs. Charles Baker). The values for the mean thickness of the living L. bulgaricus from 14 different cultures obtained by the first method ranged from 1·13 µ to 1·23 µ; and those from 9 different cultures obtained by the second method ranged from 1·02 µ to 1·14 µ. The mean refractive indices of the latter calculated by the second method agreed very closely with that obtained by immersion refractometry, and differed by a maximum of 0.009 in all the cultures measured. It therefore seems unlikely that the mean thickness measurements obtained by either method are wrong by more than about ±0.1 µ.