AbstractA significant fraction of theTetrahymenaclones isolated from natural habitats self (mating occurs within a clone). Early attempts to study such clones failed because stable subclones were rarely, if ever, observed, and isolated pairs all died. Isozyme analysis revealed that these wild selfers were a diverse group; some were very similar toT. australis, a species with synclonal mating type determination and toT. elliotti, shown recently to have a karyonidal mating type system. One originally stable clone ofT. australisincluded some selfing clones after a few years in our laboratory. Other clones manifested unique zymograms.Subclones isolated from 18 selfer strains were heterogeneous. All subclones of several selfers mated massively at each transfer through 100 fissions. Selfing among subclones of other selfers was highly variable or not observed. Although 77% of the pairs isolated died, and 9% of the pair cultures selfed, 15 selfers yielded some viable nonselfing “immature” progeny. Additionalimmatureprogeny were obtained by isolating pairs from macronuclear retentionsynclones.Although some “immature” progeny eventually selfed, most remained stable. Giemsa staining revealed macronuclear anlagen in nearly all mating pairs and some anomalies. Crosses among the F1 progeny clones of theT. elliottiselfers yield viability data comparable to those from crosses amongnormalstrains. Perhaps perpetual selfing is a mechanism of getting rid of deleterious combinations of genes and uncovering better combinations in homozygous state by playing genetic roulette. © 1992 Wiley