G. cephalantha was described from specimens sent to the Paris Herbarium on May 6, 1887, by Father Delavay who collected them at an altitude of 2800 metres on the mountain called Hu-chan-men near Lankong. The type is in the Kew Herbarium. The species was subsequently collected by Forrest, Schneider, Rock and Fang on the Lichiang Range, Chuntien Plateau, Tongshan in the Yangtze bend and elsewhere in N.W. Yunnan, where it is found in dry situations in pine and mixed forests up to 3,400 m., but it does not appear to have been introduced into cultivation until raised from seed collected on Forrest's 1930-32 expedition. The specimens which formed the subject of the accompany- ing plate were grown from this source by the Marquess of Headfort, of Kells, Co. Meath. Sown in April, 1931, they flowered the following autumn and were drawn on Nov. 4th, 1932. Two shades of flower colour, blue and mauve, were borne by plants raised from the same packet of seed. Not a little confusion of names has arisen in horticultural literature, which the publication of this plate illustrating well-grown specimens of the typical plant should settle. Franchet in describing the species for the first time placed it in Sect. Pneumonanthe. Kusnezew in his monograph (1.c.) transferred it to his Sect. Frigida where it undoubtedly belongs. As however, the seeds of G. rigescens Franch. were unknown to him he left this species in Sect. Pneumonanthe. Thus it came about that these two very closely allied species are widely separated in that work. Herbarium specimens have been examined, which are in all respects intermediate between G. cephalantha and G. rigescens and are probably natural hybrids. G. rigescens differs November 9th, 1936.