| Dykes in The Genus Iris, 1913;Description.Rootstock , a stout rhizome. Leaves , glaucous, striated, 15-20 in. long, of a yellowish-green. Stem , about 2 ft long, with a terminal head of three flowers and one or two side branches with 1-2 flowers each. Spathe valves , pointed, slightly ventricose, persistently green, the outer valve keeled, only very slightly scarious at the tip, 3-4 in. long. Pedicel , not more than ¼-½ in. long. Ovary , 1 in. long with three larger and three smaller grooves. Tube , about I in. long, light green, rounded trigonal. Falls , obovate cuneate ; the haft marked with greenish-yellow veins on white reaching to the end of the beard, which is white in front, becoming tipped with yellow towards the base; the blade of a somewhat creamy white, not so pure as in albicans, sometimes slightly tinged with blue and extending almost horizontally. Standards , white, faintly veined with yellow on the short haft and at the base of the blade. The canaliculate haft bears three or four whitish hairs, sometimes very short and almost indistinguishable. Styles , cream coloured, keeled, gradually becoming slightly broader. Crests , large, triangular, overlapping, especially along the keel of the styles. Stigma , entire. Filaments , cream, equal in length to anthers. Anthers , cream, narrow. Pollen , cream. Capsule , Seeds , Fragrance , very marked and quite distinct, somewhat like that of the lemon-scented verbena.Observations.This Iris is very distinct and yet, owing to some weakness in its constitution, which does not allow it to thrive in England, it is imperfectly known. My experience of it has been that newly imported rhizomes grow well and flower once and then the plant either dies altogether or becomes very weak and refuses for years to make any vigorous growth. I have consequently been unable to obtain seeds of this Iris, which would seem to be a possible means of raising plants that would succeed here. A hybrid of this Iris, known in commerce as the Shelford variety, was raised by Foster but there is no record of the pollen parent. It is perhaps the best white garden Iris, but like the wild plant it is sometimes liable to collapse unaccountably and probably needs warmer conditions than it usually obtains here. It is distinguished from the wild species by the narrower, more rigid and less markedly ribbed leaves, by the less widely branching stem and by the colour of the flowers, which are of a less milky white.Of the other white Irises in cultivation, I. kashmiriana is most likely to be confused with I. florentina and I. albicans. From the latter it is at once distinguished by its more widely branching stems and by the fact that its standards always bear a few hairs on the inner side near the base. These never occur in I. albicans, though they are always present in I. florentina. From this latter, however, I. kashmiriana may be distinguished by persistently green spathes, by its broad, yellow green, ribbed leaves, by the stiffly spreading and not drooping falls, and by the milk-white flowers.I am inclined to believe that this Iris is the albino form of a pale mauve-purple Iris, which has more than once flowered imperfectly with me, but the whole question needs further investigation with fresh material from Kashmir'. In this connection, a warning may be given that the common purple Iris that apparently grows freely in Srinagar is not this possible purple I. kashmiriana but the form of I. germanica, which Foster named Kharput, after the town in Asia Minor from which it was first sent to him (see p. 163).There seems no good reason for separating from I. kashmiriana Foster's I. Bartoni: The yellow-green ribbed or striated leaves, the green spathes, the bearded standards and the thick, somewhat flattened stem are all characteristic of I. kashmiriana, which moreover is not mentioned by Foster in his description and notes of I. Bartoni. The latter he described from plants obtained from Kandahar in 1880 by Colonel Barton, who was informed that the rhizomes came from the ancient ditch surrounding the ruins of the old town, which are situated at a distance of four miles from the present site. Obviously the plant was growing in Kandahar in a semi-cultivated state, and the fact that it flourished in the damp soil of a ditch also agrees with what we know of I. kashmiriana.The typical I. Bartoni, which I have grown and flowered and compared with Foster's MS. notes, has a very long beard ( 1 in.) on the haft of the standards, but I find that Foster records (MS.) that, in some plants that flowered in 1885, the hairs on the standards were "much less developed," "hardly visible often."It should be noticed that the Botanical Magazine figure of~. Bartoni does not show any purple markings on the blade of the falls. These occur on the" typical plant but, as Foster records in his notes, they are variable. I have noticed similar variations in plants of I. kashmiriana which I have received direct from Kashmir.