The distribution of Trichostomum tenuirostre (Hook. & Tayl.) Lindb. var. gemmiparum (Schimp.) Zand. is extended to eastern Asia. While examining a Japanese series of specimens from TNS determined as Tortella japonica (Besch.) Broth., a collection was discovered to have smooth, multicellular, dark orange, thick-walled gemmae. That the specimen had gemmae was indicated on the packet, probably by K. Saito, from whose herbarium many of the specimens examined derived. Since the specimen was not Tortella japonica, having no autoicous buds and lacking other characters of that species, further examination indicated the specimen was in fact Trichostomum tenuirostre (Hook. & Tayl.) Lindb., which has a described variety bearing gemmae, the var. gemmiparum (Schimp.) Zander. This variety has been recently treated by Zander (1994) for Mexico as Oxystegus tenuirostris (Hook. & Tayl.) A. Smith var. gemmiparus (Schimp.) Zand. Specimens from eight states in that country are known, with additional stations in Cuba, Haiti, and Brazil in the New World, and Europe and India in the Old. The present record extends the known distribution of this variety to eastern Asia. The species, however, is so variable and so easily confused with other species, primarily of the genus Tortella, that it is likely the variety is as widespread as the species. Noguchi (1988) did not mention gemmae in his treatment of Oxystegus for Japan. The problematical nature of the placement of Trichostomum tenuirostre is evident in his justification for keeping that species in Oxystegus, because of the absence of a strand in [the] stem. That Trichostomum tenuirostre has no strand was also asserted by Saito (1975). The Japanese specimen of the var. gemmiparum under discussion does, in fact, have a stem strand, which may account for its initial determination as Tortella japonica, which possesses that feature in addition to having a poorly defined region of basal cell differentiation. That the strand is variably present in Trichostomum tenuirostre in Europe has been noted by Hilpert (1933, p. 588) being large and well-defined to absent: gross und gut begrenzt, and fehlend. Flowers (1973) described the species from Utah in the United States as having central strand present, often weak or sometimes lacking. Of four of the specimens from Japan redetermined as Trichostomum tenuirostre, three possessed definite strands in addition to the specimen of the variety being discussed. Since known populations of var. gemmiparus include those with fruiting plants (Zander 1978), one cannot associate the production of gemmae exclusively with depauperate conditions; these plants may develop two distinct strategies for dispersal in the same population. The Japanese specimen was without fruit. For an extensive treatment of this plant and its nomenclature see Zander (1978). Illustrations of the gemmae are provided in Zander (1978; 1994). Specimen examined.-JAPAN Tokyo, Okutama, Nippara, Ogawadani valley, ca 1,000-1,700 m, on decayed wood, K. Saito 10306, Nov. 5, 1971 (TNS).