WE wished to isolate the cell membranes of the smooth muscle of the uterus for investigation of the sialic acid components (see following communication) and have therefore applied the method of Rosenthal, Edelman and Schwartz1, which is a modification of McCollester's method for emptying broken skeletal muscle cells2. The uterine muscle cells are small, tapering and enmeshed in collagen so that the tissue is too tough to be disintegrated in a blender. We therefore protected the uterine horns against the effects of freezing by impregnating with glycerol3, and then froze them between slabs of solid carbon dioxide and scraped off the endometrium. Frozen muscle segments are made into a block with drops of 50 mM calcium chloride, and sectioned on a freezing microtome. The mush on the knife was periodically transferred to 50 mM calcium chloride at 0° C and this homogenate was freed of cell contents by modifying Rosenthal's method as in Fig. 1. The membranes sedi-mented readily during centrifugation because the collagen fibres were retained; if they did not settle after the ATP treatment, but remained as a fluffy dispersion, then phase-contrast microscopy showed that they were still surrounded by extruded protein containing nuclei and mitochondria, and a second ATP treatment was necessary.