JOURNAL OF THEARNOLD ARBORETUMVo.. XXXII JULY 1951 NUMBER 3 STUDIES IN THE BORAGINACEAE, XXI SINO-INDIAN SPECIES OF ONOSMA IVAN M. JOHNSTON THE GENUS ONOSMA has its center of distribution, greatest diversity,and maximum concentration of species in Iran and westward into Syriaand Turkey. Over seventy of the 128 species of Onosma recognized inthe enumeration by Stroh, Beih. Bot. Centralbl. 15": 430-454 (1939),occur in that area. Eastward of its center the genus is represented bytwo groups of species. One occurs from Turkestan northeastward to theAltai. Its species are conventional members of the genus and most haveevident affinities with species in the Near and Middle East. The othergroup, a more southerly one, occurs across northern India and adjacentTibet into Burma and China. Among its members are the most southerlyand most easterly ranging species in the entire genus. Its western mem-bers, those found in Baluchistan. Afghanistan, and adjoining India, mayhave evident Iranian affinities, but those ranging east of Kashmir havenone. In its migration eastward from its center, across northern India toChina, the genus has been confined to a narrow band of territory. Isolated,subjected to high altitudes, and influenced by monsoonal climate, its repre-sentatives there have been modified and become very distinct species,some with peculiarities of habit and structure unknown elsewhere. Thepresent paper is primarily concerned with this distinctive group of species.In order that the treatment be given wider usefulness, however, it hasbeen expanded into a regional treatment, and all species of Onosma fromwestern India and 'eastern Afghanistan and Baluchistan included also."India" has been used as a geographical rather than a political term andhas been applied to all of the former, continental British India exclusiveof Baluchistan and Burma. In the past the species of Onosma have been distinguished almost exclu-sively by external features of the plant, with general habit, indument and size and shape of the flowers most emphasized. In general, though the differences that have been used are real, they are of a sort difficult to express in precise terms. Published accounts of the genus, as a result, have been far from satisfactory as a means of identifying species. It has