AbstractWhen rat liver ribosomes were dissociated by removing their magnesium, the two subunits hads 020,w of 52.6 and 28.1 S, and appeared to be quite asymmetrical. In 4Murea they became even more elongated. On restoration of magnesium, both subunits assumed more compact forms, 57.6 and 31.7 S, but only 28% of them formed larger aggregates. After brief treatment with formaldehyde, the magnesium‐free subunits still elongated in urea, but to a lesser degree than the controls. When magnesium was restored they changed to 53.4 and 31.7 S forms, but did not, aggregate. The ribosomal subunits were also separated by reducing the bound magnesium to a low level, then adding 2.2Murea. After the urea had been dialyzed out, most of the large subunits were still in a compact form and when the magnesium concentration was increased, most of the subunits did not reassociate to whole ribosomes. Formaldehyde‐treated subunits, however, did not reassociate. Even in their most compact forms, 60.0 and 35.5 S, both subunits had high frictional ratios. The 60.0 and 31.7 S forms were assumed to be the shapes that are seen in electron micrographs of negatively stained ribosomes. Both subunits seemed to be highly hydrated. Since the large subunit appeared to contain twice as much water as the small one, it might be a structure with either a deep channel or a hollow c