| Dykes, The Genus Iris 213. 1913;Description. Rootstock, an ovate bulb covered with thin membranous tunics not splitting into fibres at the apex and producing bulblets ini pairs on opposite sides at the base (see Plate XLIII). Leaves , 12-24 in. long, glaucous, channelled. Stem , 12-18 in. high, usually with only one head of flowers but occasionally a very strong bulb will produce a side branch, very similar to those of I. spuria. Spathe valves , up to 4 or 5 in. long, narrow, green, 1-2-flowered. Pedicel , varying in length, often equal to, or a little longer than, the spathes, that of the second flower being the shorter. Ovary , 1-1¼ in. long, oblong, narrow, usually exserted from the spathes. Tube , practically none. Falls . The suborbicular blade is separated by a gradual and slight constriction from the oblong-oval haft. The colour is very variable but there is always a yellow or orange streak or patch on the blade. The oblique veining on the sides of the haft is usually conspicuous. Standards , oblanceolate-unguiculate, usually of a slightly different shade of colour to that of the falls. Styles , broader than the haft of the falls. Crests , subquadrate, large. Stigma , bifid, with two rounded teeth. Filaments , varying in colour. Anthers , varying in colour. Pollen , yellow or orange. Capsule , long and narrow (2-3 in.), with a hollow running down each face (cf. Fig. 27). Seeds , small, yellow-brown, thick D-shaped, compressed.Observations.There is little doubt that Linnaeus included under the one name of I. xiphium both that species and also I. xiphioides. This is clear from the fact that he quotes C. Bauhin, Pinax, p.38 I. bulbosa latifolia, caule donata, which is almost certainly I. xiphioides, and also the Hortus Cliffortianus which refers us to Bauhin's I. bulbosa lutea inodora, which in its turn is based on Clusius, Hisp. p. 276 I. bulbosa angustifolia flore luteo, which is undoubtedly a form of I. xiphium. (Cf. the specimen, named I. xiphium, in the Smithian Herbarium at the Linnaean Society which is opviously I. xiphioides.)This confusion is the more surprising in view of the fact that in this case, as in several others, Clusius was well aware of the difference between the two species and carefully described them both. There are doubtless many forms of this Iris even in the wild state but we should scarcely be justified in separating as species forms which differ merely in size or in the time at which they flower. The largest, as well as the earliest, form is probably that which is known to dealers as "I. filifolia." It has nothing to do with the true I. filifolia, as will be seen by comparing Plates XLIII and XLIV. The pseudo-filifolia has the characteristic short tube of I. xiphium, while that of I. filifolia is always of some length. Another obvious difference is to be found in the shape of the standards. This pseudofilifolia flowers in April and May and grows apparently in the neighbourhood of Gibraltar, while the true plant grows actually on the Rock itself.On the other hand, at an altitude of over 5000 feet, further north, specimens of I. xiphium are found which do not flower until August or September. The Roquehaute form also flowers late.Many garden forms of this Iris have long been known. Gerard in his Herball ( I 597) mentions among "certaine bulbose or Onion rooted Flower-de-luces" an Iris bulbosa jlore vario and the name probably means that he knew more than one form of the plant. The Hortus Eystettensis (1613) has about a dozen forms and in Simula's Flora exotica ( I 720, BM) eight varieties are depicted.In recent years many fine large-flowered forms have been introduced into cultivation under the name of Dutch Irises. These were raised by the well-known firm of C. G. Van Tubergen, Junr., of Haarlem, who tell me that I. Boissieri, I. tingitana, and other species were used as well as I. xiphium. The pollen of these species seems however to have had little effect except in increasing the vigour of the plants and the size of the flowers, for I can see no trace in any of the specimens that I have grown or seen illustrated of the long perianth tube that is found in all the species except in I. xiphium. It may be that the absence of tube is dominant over its presence and that specimens with obvious linear perianth tubes will occur in the next generation of these hybrids, but on the other hand there is no doubt that hybrids with flowers at least <).S early and as fine as those of these Dutch Irises can be obtained by crossing the early form of I. xiphium already described, and for which I. xiphium var. praecox would be a not inappropriate name, with pollen of garden varieties of I. xiphium. This, at least, has certainly been the result of some crosses which I made several years ago and from which I have obtained a number of fine seedlings, which come into flower in the last week of May. The absence of the linear perianth tube in any of these so-called Dutch Irises is the more remarkable because there is in existence a hybrid of I. tingitana crossed with pollen of I. xiphium, raised by Foster, in which the linear tube is half an inch or more long. This hybrid is of great garden value for the growth and the flowers are practically identical with those of I. tingitana and yet the plants flower freely and the flowers survive when those of I. tingitana succumb in the bud stage to late spring frosts. In 1912 this hybrid was actually in flower in the open on April 15th in Surrey. For the cultivation of I. xiphium, see the introduction to the section, p. 210.