Immobilization antigens of stock 51 of Paramecium tetraurelia were subjected to electrophoresis in NaDodSO4/ polyacrylamide gels. Type A is estimated to have a molecular size of 300,000 daltons; H is estimated to be 288,000, D to be 280,000, E to be 270,000, B to be 253,000, and C to be 250,000. Poly(A)+RNAs have been isolated from cells producing these antigens and subjected to electrophoresis in methylmercury gels. A major band is found to vary in mobility with antigenic type: Its position in preparations derived from paramecia synthesizing antigen A indicates a size of 8400 nucleotide residues; its position from paramecia synthesizing other antigens indicate H, 8200; D, 7900; E, 7500; B, 7600; and C, 7000. Because ofthe sizes and quantities of these RNAs, it is argued that they probably represent the mRNAs for the immobilization antigens. It is concluded that each immobilization antigen probably consists of a single polypeptide and that only one major serotype-determining mRNA is present in each antigenically different paramecium.Differences in gene expression that can be classified as simple environmental modifications are well known. Often, however, differences in gene expression show cellular inheritance; subclones that have alternative phenotypes may arise and reproduce true to type, even though all are under the same environmental conditions. The molecular mechanisms responsible for several instances ofinherited differences in gene expression have recently been discovered. Gene rearrangements have proved to be the basis for all cases. They include mating types in yeast (1), antigenic variation in Salmonella (2) and trypanosomes (3-5), and the production ofantibodies (6). Transposable elements in a variety of organisms (7-9) can also lead to modification of gene action.Since all ofthese cases involve changes in DNA, it is not surprising that they all are relatively stable and are not, as far as we know, directed by environmental stimuli. The production of antibodies may be an exception. Although antigens trigger the formation of specific antibodies, they do so by stimulating growth and synthesis. They select cells that have specific rearrangements, but there is no evidence that they induce rearrangements. Moreover, order in the sequence of changes exhibited by classes ofantibodies and in the succession ofantigenic types in trypanosomes does not necessarily imply specific induction. Nevertheless, a role for the environment in the induction of antibodies may exist, for the rearrangements occur in specific cells at specific times in development, suggesting that unknown inducing conditions may exist in those cells.Differences in the expression of genes determining immobilization proteins or i antigens (i-ags) of Paramecium are in some respects anomalous. They are heritable under certain environmental conditions and some show preferred sequences of change from one state to the other. Brown (10) has suggested that they too will prove to be due to gene rearrangments. However, they are unlike the cases d...