Cyclopedia of American Horticulture: E-M

throwing up slight ridges, with a plow or other implement, and the water turned into these basins in succession and allowed to accumulate to a sufficient extent. This method is particularly applicable to fruit trees, although it is occasionally used in other crops. In very sandy soils the water is occasionally carried through the field in wooden troughs, which admit of sufficient seepage to water the land. This prevents the undue seepage which might occur in such soils if the water was flowed over the surface. Another method is to distribute the water through the field in iron pipes, with opeuings at frequent intervals, in which nozzles can be attached to deliver a fine spray over a small area. With four or five such nozzles an attendant can water a considerable area of ground in the course of a day. Such an irrigating outfit in Florida was supplied with a power equivalent to about one horse-power per acre. The mains and laterals were of 1-inch or lX-inch iron pipes laid near the surface of the grouud, the laterals about 100 feet apart, with hydrants every 50 feet. Tuuks were originally used, but it was found desirable to pump directly into the mains to insure a sufficient pressure.

Care should be exercised in applying water to the land. Where water is plentiful there is a common practice of using such an excess as to injure the flavor of fruit, increase the liability of disease, and eventually injure the land by the accumulation of seepage waters and of alkali. As a rule, there has been very much more damage from over-irrigation than from the use of too little water. The first two or three years a soil usually requires a considerable amount of water, but after becoming well moistened to a considerable depth it should require comparatively little water thereafter to maintain its fertility. As it is not easy to apply just the proper amount, the excess should be provided for. If there is any reason to fear lack of drainage, the land should be thoroughly underdrained before Irrigation is started, or at any subsequent time when the need of it becomes apparent.

Irrigation always should be supplemented by the most thorough cultivation. After going to the expense of watering the soil in this way, it is poor economy to allow the water to escape by evaporation or otherwise; therefore every precaution should be used in thorough, subsequent cultivation and in the exclusion of weeds, to conserve the moisture so applied. The intelligent horticulturist will find that in the use of this expensive method of maintaining a proper water supply in the soil, it is incumbent upon him, even more than if the method were not used, to give careful attention to all the ordinary methods of preparation and cultivation in order to maintain the advantages he has established by the Irrigation plant. Milton Whitney.

Si'b-irrioation In The Greenhouse. — The term subirrigation is used to describe a method of supplying water to the roots of plants by means of some form of conduit placed below7 the surface of the soil. In greenhouse operations, the essential features of the plan are a level, water-tight bench-bottom, and tile or pipes to serve as conduits for the water. The tile, or pipes, are laid directly on the bench bottom, and over these the soil is spread, usually to the depth of alK>ut 6 inches. When water is introduced in sufficient quantities through the tile or pipes, it passes out at the joints or perforations into the soil.

When applied to greenhouse operations, the term subwatering has been proposed by Goff and (Vanefleld for the reason that Irrigation is used to denote watering on a large scale out-of-doors. It may be said, however, that the words watering and Irrigation do not indicate the scale of operations with any degree of accuracy, hence it seems as well to use an old word as to coin one, especially when the familiar word expresses the meaning intended.

Experiments in watering plants bv this method were begun in the winter of 1890 and 1891. at the Ohio Experiment Station. The suggestion came from the result obtained in an effort to check the lettuce rot. Water was introduced to the soil in boxes by means of a pipe, in a similar manner to the method often employed in watering hills of melons and cucumbers. When the plants were watered in this manner, the lettuce showed

so much more vigor than that watered in the ordinary way that operations were begun at once on a larger scale: first in a beil on the ground having a clay bottom, ci.en on a water-tight bench, made of lumber, and finally, on tile benches, covered with cement.

In all of the earlier experiments the water was introduced through pipes, or drain-tile, laid about 2 feet apart on the bottom of the benches. Uoff and Cranefleld have used brick instead of tile, placing them near enough together to touch. They were set on edge in a galvanizediron pan, made for the purpose. J. C. Arthur clipped off the corners of the bricks, so as to facilitate the flow of water. The Ohio Station has modified this plan by using common drain-tile, laid so as to touch, thus covering the entire bench bottom, instead of a line of tile every

2 feet, as at first.

Benches made of lumber have proved unsatisfactory because of the swelling and warping of the boards. Solid beds on the ground have not been successful, except where an impervious clay bottom existed. Galvanized-iron adds greatly to the cost of construction, and lasts only a short time. The only suitable bench for greenhouse sub-irrigation is one made of materials which are not acted upon by water.

A well-made tile- and cement-bench seems to be the only form of construction that will meet the requirements. Such a bench does not cost so much as to preclude its use, and will last as long as any other part of the greenhouse. In describing such a bench, it will not be necessary to enter into details, except such as relate to the method of watering under discussion. The bench must be water-tight, or nearly so, and this condition is secured by spreading a layer of cement, an inch or more in thickness, over the tile bottom. It is not a matter oi any moment whether flat tile or common drain-tile are used, except in the quantity of cement required. The cement must be spread with care, so as to secure a perfectly flat, level bottom, otherwise the water will not flow uniformly in all directions. The sides of the benches are made of cement also, but need be only 2 or

3 inches high, or of sufficient height to merely retain the water. Hoards or slate are placed outside the cement wall to retain the soil. The tile-bottom may rest on iron or wood cross-pieces. Wood has been in use for this purpose at the Ohio Station for seven years and shows no signs of decay, because it is out of reach of the water.

Nine years' experience shows that a perfectly constructed bench bottom, with the tile laid 2 feet apart, will serve satisfactorily in distributing the water to all parts of the bed, provided the tile are straight, so as not to impede the flow of water. The tile are laid in the

have been used over the entire bench bottom with good results, and it seems probable that this will be found to be the best form of construction, as it appears more certainly to insure an even distribution of water. The method of construction is the same as above described, for the two plans differ only in the number of tiles employed to distribute the water. When the bench bottom is covered with tile, placed near enough together so that the soil will not fall between, it will be readily seen that water introduced at any point will flow to all parts of the bed in and around the tile. It needs simply to be brought up to such a level that it will reach the soil, when capillary attraction will complete the distribution. Fig. 1183 shows a bench in a tomato house constructed after this plan. AA are the inlets; B the irrigating tile, from which the soil has been removed; C is the tile bench bottom, covered with cement. The same size of tile, viz., 2%- or 3-inch, is used both above and below. D is thecement side, which has been broken away to show the method of construction. The outer board has been removed also.

The cost of construction need not be discussed here, except to state that the only items extra, more than are required in any well-constructed greenhouse, are the cement bottom and the tile in which the water is distributed.

A plan has been devised for applying water to small plants in fiats which may properly be mentioned under this head. The fiats are shallow boxes with slatted bottoms. When the plants require water, the flats are placed in a shallow vat of water and allowed to remain until the surface of the soil appears to be damp, or even wet.

A watering in this manner is far more efficient than by the ordinnry method. Taken in connection with subirrigation in the benches, a crop of lettuce can be brought to marketable size nearly two weeks earlier than when surface watering is practiced. Anything like a full discussion of results of experiments in watering plants in the greenhouse by sub-irrigation would be too voluminous for an article in this connection. A brief review of the results obtained at some of the stations, together with a short discussion of some general principles, will serve the purpose intended. The Increase in weight of lettuce from sub-irrigated plats over those watered in the ordinary manner has been reported by Rane, of West Virginia, as 25 per cent and by Ooff and Cranefleld as 2fi per cent. At the Ohio Station the range has been from 25 to 100 per cent. In the latter case the result was obtained by commencing with the plants as soon as taken from the seed-bed, and carrying

the two lots through to the termination of the experiment, one by watering altogether on the surface of the soil, the other by sub-irrigation. Each of the experimenters speaks of a gain in earliness of several days by sub-irrigation. Rane secured similar results with long-rooted radishes by this method of watering, but not with the turnip-rooted sorts, while Munson, of Maine, doubled the crop by watering below. Better results have usually been secured at the Ohio Station with the turnip-rooted than with the long varieties, but in all cases there has been a gain in favor of sub-irrigation, varying from 50 to 100 per cent. Rane found that sub-irrigation increased the yield of tomatoes, but the gain was not large. Essentially the same results have been secured in Ohio. The tomato crop has not been greatly influenced by the manner in which the water was applied, and the same is true of beets, while subirrigated cucumbers and parsley have shown a decided gain over surface-watered. Carnations, roses, chrysanthemums, sweet peas, violets and smilax have been under experiment by the two methods of watering, and while no such marked results have been secured as with lettuce and radishes, the sub-irrigated plots have shown superiority over those watered in the ordinary manner, In nearly all cases. With carnations the improvement has been mainly in length and stiffness of stem.

Aside from the increase of crop secured by sub-irrigation, there are other considerations which may be urged in its favor, and these are embodied in the following general propositions:

(1) Wittering by nub-irrigation in the greenhouse, saves labor. The amount of labor saved depends mostly on tin completeness of the arrangements for watering, but there is a saving in the number of applications as well. It is possible to reduce the time eraployed in watering a house, or series of houses, to oneflfth the time usually required.

(2) Watering by sub-irrigation assures an abundant and uniform supply of water to all parts of the bed. Perfect construction of the benches is assumed in this case, but with such construction watering becomes almost automatic, the only care necessary being to look after such portions of the beds as may, by position, be subject to unusual conditions of air or sunlight.

(3) Where sub-irrigation is practiced in the greenhouse, the surface of the soil does not become compacted, but retains its original loose, friable condition. It is true that where frequent syringing is practiced the surface of the soil becomes more or less hardened, but not to the extent that occurs in surface-watering, and the condition is easily remedied, whereas in the other case it is not. It follows that a heavier soil may be used for sub-irrigation than with surface-watering.

Still other considerations might be urged in favor of this method of watering, but many of them would apply to special cases only. Regarding the effect of the method upon insects and diseases, but little can be said. Lettuce rot is less prevalent upon sub-irrigated plots than upon those which are surface-watered, but in extreme cases plants succumb to the disease, whichever method of watering is practiced. Munson found that radishes suffered more from the attacks of millipedes upon sub-irrigated plots than upon plots watered in the usual manner. Nematodes work upon the roots of roses, whichever way the plants are watered. The manner of watering has no apparent effect upon the red spider. Even in houses watered wholly by sub-irrigation this pest is no worse than in houses in which the water is applied to the surface of the soil. It may be said, however, that nearly all classes of plants are more easily kept in a healthy growing condition, and are thus better able to resist enemies of all sorts, when sub-irrigated than when supplied with water in the ordinary WBy.

This method of applying water to plants in greenhouse benches has now been sufficiently tested to determine its value. All that now remains is to devise ways and means to utilize what is known concerning it. The adaptation to suit particular cases must be made by individuals, but this will be far easier in the future than in the past, because better methods of construction prevail than formerly. The success of sub-irrigation in the greenhouse is now simply a question of mechanics.

W. J. Gkeen.

18ATIS (meaning obscure). Crucifertr. This includes the Dyer's Woad, /. tinetoria, formerly cult, for a blue dye, but no longer advertised. Cffisar relates that the

1184. Isoloma Tyda:a(XK).

ancient Britons used the Woad for staining their bodies, and the word Britain itself comes from an old Celtic word meaning painted. Before indigo became common in Europe, the Dyer's Woad produced the chief blue coloring matter for woolen cloth. The introduction of indigo in the seventeenth century destroyed this important Industry, not without opposition. Dioseorides and Pliny mention both the Dyer's Woad and indigo.

/. tinetdria, Linn., is rather tall, glabrous and glaucous: stem-lvs. lanceolate, entire, sessile, somewhat arrow-shaped: lis. small, yellow, borne in early summer, on panicled racemes. Instead of apod, opening lengthwise by valves, it has a closed fruit like on the samara of an ash, 1-cclled, 1-seeded. indehiscent, wing-like. It is a biennial, and common in Europe.

ISCHARUM. See Biarum.

ISMENE. Now referred to Hymenoeallis.

ISNARDIA. Includes a few species of Ludu-itjia.

ISOCHtLUS (Greek, equal lip). Orchidileeie. A genus of no commercial value. Plants epiphytic, with tall, slender, leafy stems, without pseudobulbs, bearing a few small fis. at the summit. Sepals erect, free, keeled; petals similar but plane; labellum like the petals and united

with them to the base of the column, somewhat sigmoid below the middle: column erect, long, without wings: pollinia 4. About 5 species in Braz., Mex., and W. Ind.

linearis, R. Br. Slender, 1-lJi ft. high, leafy: lvs. distichous, linear, striate, obtuse, einarginate, IX in. long: Hs. purple, borne in a short, terminal spike. March. Growing on rocks and trees in thick woods, Jamaica, Trinidad, Brazil, etc. B.R. 9:745. L.B.C. 14:1341.

H. Hasselbring. IS6LEPI8. See Scirpus.

ISOLOMA (equal border). GesnerAceie. Includes Tydasa. Sixty or more tropical American plants, very closely allied to Gesneria and Aciiimenes. From Gesneria distinguished by absence of well-formed tubers and characters of capsule and anthers, and the 5 lobes of the disk equal; from Aciiimenes in the more tubular flowers and lobed disk. The culture is the same as for Achimenes and Gesneria. Seeds of thenewerhybrids genninatequickly, and plants bloom the same year. It is probable that the pure species are not in the trade. Like Achimenes, Gesneria and Gloxinia, they have been much hybridized and varied. It is probable that they are hybridized with Achimenes and (Tesneria. Tyda?a is a garden genus. It is not known how the current forms have originated. Some of the recent ones have fringed fls. (Gn. 55:1223).

Tydeea (Achimenesplcta, Benth. Ty<?<>« picta. Dene.). Fig. 1184. One to 2 ft., hairy: lvs. cordate-ovate, coarsely serrate, spotted and reticulated with pale green or silvery green, with a broad light zone down the center: fls. single, on long, axillary stems, nodding, the orifice oblique and lobes obtuse, the upper longitudinal half of the fl. red, the lower half yellow and redspotted. Colombia. B.M. 4120 (adapted in Fig. 1184). B.R. 31:42. F.S. 1:1718. —On this species Decaisne founded the genus Tyda?a in 1848. This species has been called Isoloma pirtum, but this name was taken by Planchon in 1850 to '51 for the Gesneria picta of Hook., which is a very different plant. See F.S. 0:580. B.M. 4431. This latter plant, the first Isoloma piclum, is apparently not in commerce, i, Mottet (Tydma amdbilis. Planch. & Lind.). hairy: lvs. ovate, more or less tapering to the

Erect,

Cecilias, Nichols. {Tydvea Cecllitpt Andre). Much like J. amabile, but lvs. marked with violet and silvery zones or blotches: fls. 2 or 3 from each axil, the fls. pale rose outside and striped in the throat, and the limb purple-spotted. Colombia. I.H. 23:200.

ocelli turn, Benth. & Hook. {Achimenes ocellita, Hook.). Short-hairy on the stem: lvs. ovate-acuminate, serrate, green: fls. small, on peduncles shorter than the lvs., the tube and short, rounded lobes red, the segments marked with whitish and black spots. Panama. B.M. 4359.

Jaliscanum, Wats. Fig. 1185. Herbaceous or some what woody at the base, 1 ft., pubescent: lvs. opposite, oblong-lanceolate to ovate-lanceolate to ovate, shortacuminate, short-stalked, serrate: fls. 2-4 on an axillary peduncle, the corolla an inch long, tubular and shortlobed, [pubescent, scarlet. Hex.—A worthy plant, not yet in the trade, but has been cult. L Jj_ 3.

ISONANDEA (Greek, equal anthers). SapotAcea>. Isonandra Gulta is a large-leaved E. Indian tree, which furnishes the best commercial gutta-percha. The name has appeared in one southern catalogue, but the plants were found to be not true to name. This plant should be called Dichopsis Gulta. In Dichopsis the floral parts are in G's, stamens 12, and the seeds have no albumen, while in Isonandra the floral parts are in 4's, the stamens 8, and the seeds albuminous. See Subber Plants.

Giitta, Hook. Properly Dichdpsis Gtitta, Benth. & Hook. Gutta-percha Tree. Lvs. leathery, elliptic, abruptly pointed Malaya. R.H. 1898, p. 441.

IS6TOMA (Greek, equally cut; referring to the corolla, and true only by contrast with Lobelia). Companulaceat. This includes a plant treated as a half-hardy annual, which grows about a foot high, has curiously cut foliage, and odd fls. with a slender bent tube 1 in. or more long, and 5 slender spreading lobes, each %in. long. Among allied generaof garden value, Centropogon and Siphocampylus have the stamens fastened at the base of the tube, while in Isotoma they are at the top or above the middle. (Centropogon has an indehiscent berry: Siphocampylus a capsule 2-valved at the top like Isotoma.) Downingia has a tube of stamens free from the corolla.

axillaris, Lindl. Perennial, flowering the first year so as to appear annual, but forming at length a hard rootstock, erect, with few spreading branches: lvs. linear, irregularly pinnatifid, 2-3 in. long, lobes linear: pedicels axillary, 2-6 in. long: fls. large, bluish purple, pale outside. Australia. B.M. 2702 (as Lobelia senecioides) and 5073 (as Isotoma senecioides, var. subpinnatifida). — Not in cult.

petreea, F. Muell. Identical with the above, except that the lvs. are ovate-oblong or elliptical. Australia. The plant in the trade is said to have cream-colored fls., and is sold as a"Lemon Verbena," a name which properly belongs to Lantana.

ITALIAN MAY. Spiraa kypericifolia.

ITEA (Greek name of the willow; because it has willow-like lvs., and grows near the water). Saxifragaeeat. A genus of trees and shrubs, numbering about 5 species, inhabiting eastern N. Amer. and eastern Asia, whose one representative in cultivation is /. Virginica, a low, upright, somewhat coarse shrub, best known by its long, upright racemes of small white fls. appearing about July l,in Massachusetts, and its brilliant autumn coloring. In nature it inhabits low, wet places. In cultivation it seems to adapt itself to almost any soil. It is not perfectly hardy, but grows rapidly and seems enduring of both sun and shade. In ornamental use it is planted in masses or mixed with other shrubs of similar character in the shrubby border or at the edge of woods. Its somewhat coarse character does not favor its approach to more refined objects. In autumn it becomes a brilliant red. It is prop, from seed, by cuttings and by division of roots, which spread slowly and form flumps of stems. It may be collected from the wild.

Virginica, Linn. Virginian Willow. Fig. 1186. A shrub, lk-6 ft. high, usually not more than 2-3 ft. high, of upright, somewhat slender habit: lvs.deciduous, alternate, oblong, pointed, minutely serrate, smooth green above, pale and slightly pubescent below, petioled, without stipules, 1-3 in. long: fls. regular, 3 lines long, fragrant, white, in solitary, erect, hairy, simple, dense, terminal racemes 2-6 in. long, given a greenish white effect by the stamens and pistils, not particularly showy, appearing late June and July; calyx 5-cleft, persistent, nearly free from the base of the ovary; corolla of 5 lan

1186. Itea Virginica (X%).

ceolate, nearly erect petals and longer than the 5 stamens: capsule slender, longitudinally 2-furrowed, 2celled, many-seeded, splitting through the simple style and partition. Pa. and N. J. to Fla. and La. B.M. 2409.

A. Phelps Wyman. IVA (named after Ajitga Tva, from its similar smell). Comp6sita>. This includes I. frutescens, Linn., the Marsh Elder or High-water shrub, a native hardy perennial of no garden value, which is, nevertheless, on record as having been cult. It grows 3-12 ft. high in salt marshi's and on muddy seashores, hns serrate lvs. and fls. as inconspicuous as those of a ragweed. See B.B. 3:292 and Gray's Manual.

IVfiSIA. All referred to Potentilla.

IVY. The common or English Ivy is Hedera. Bolton J,=Ampelop8t8 tricuspidata. German I. = Climbing Senecio and JTerniaria glabra. Ground \. = Xepeta Olechoma. KenilworthI.=//inarta Cymbalaria. Poison l.=Bhu8, B. Toxicodendron. Some authors think that two species of Rhus are confused, R. radicant being the common Poison Ivy of the North, and B. Toxicodendron being a shrub of the South.

tXIA (Greek, bird lime; said to refer to the juice). Iriddceir. Ixias are delightful tender bulbs originally from the Cape of Good Hope, with attractive grass-like foliage and spikes of flowers borne in early spriug, exhibiting an exceptionally wide range of colors. They grow about 114 ft. high on the average, with an nnbranched stem, a spike 3-8 in, long, containing G-12 Hs. each 1J-5—2 in. or more across. The fls. have a very slender tube usually about in. long, and 6 segments. The following colors are all well marked in Ixia: white, yellow in nt least 3 shades, orange, lilac, rose, pink, crimson, light and dark purple, ruby red, pale blue, and even green. Perhaps the only important colors lacking are sky-blue anil red in the bright shades of scarlet and vermilion. The flowers may be concolorous (all of one color) or these same shades may be combined wi'li an eye. Most of our cultivated forms seem to have an eye of brown, purple or almost black, but there have been kinds with a white, blue or green eye. Occasionally there is a ring of brown color above the purple. Add to this that the backs of the segments may be more or less suffused with various colors) usually, however, that of the eye) and the interesting possibilities of Ixias in color combinations can be imagined. Sooner or later all good gardeners yield to the fascination of bulbous plants, and whoever lias not yet succeeded in growing Ixias has something to live for.

Ixias number their cultivated varieties by the hundreds. Next to Crocuses and Freesias they have no rivals in point of popularity among spring-blooming bulbs of the important Iris family, which rejoices in the possession of such splendid summerblooming bulbs as Iris, Gladiolus and Montbretia. Culturally they belong to the same class with Babiana and Sparaxis, which are also desirable and distinct in general appearance and coloring, but are outstripped by Ixias in popularity and in number of varieties. Botanically, these three genera belong to the Ixia tribe, in which the lis. are spicate. not fugitive and never more than one to a spatbe. The stamens of Ixia are equilateral: those of Babiana and Sparaxis unilateral. Ixias have about C erect grass like lvs. arranged in 2 ranks; Babiana has plaited, hairy lvs.

Bulb catalogues give no bint whatever as to the parentage of the numerous named varieties. Not one of them mentions /. mnculala nor /. columellaris, which were the two all-important parent stocks. Of the 23 species recognized by Baker in Flora Capensis, vol. 6, 1896, only /. viridiflora appears as a trade name, but /. speciosa and panieulata are advertised under their synonyms craterioides and longiflora. I. crocata is Tritonia crocata, and /. hybrida of the trade is not tne hybrida of the botanists, but means nothing more than mixed varieties.

Before speaking of the dominant types, it is convenient to mention some very distinct species which are still cultivated in a condition not essentially different, botanically, from the wild types. /, panieulata is instantly distinguished from all other Ixias in cult, by its very long tube, which is often 3 in. long. It is also the last to bloom. /. viridiflora is unique in the genus for its green flowers, and it is one of the few green-flowered plants that are attractive. Whether this species has hybridized with the other dark-eyed species is conjectural. At any rate, the prototype is a popular plaut to-day.

Of 86 named varieties received from 3 leading dealers in America, England and Holland, and supposed to be a representative collection, all but a bare dozen seem to be the offspring of /. maculata and /. columellaris. Both of these species have a purple or purple-black eye, sometimes brown, and thewhite and yellow colors of the segments are derived from maculata, while the lilac and purple shades of the segments are derived from columellaris. Baker makes no distinction between these two prototypes except that of color. The common opinion is that Ixias hybridize freely, both at the Cape and in cultivation, and it is usually said that they are now so thoroughly mixed by hybridization and selection that it is impossible to refer any of the named horticultural varieties to their proper species. Nevertheh ss, from a study of the specimens mentioned above and the colored plates cited below, the writer ventures the opinion that the vast majority of cultivated Ixias are eyed forms, which, with the exception of viridiflora, can be readily referred either to maculata or to columellaris, and that all such forms could be reproduced without hybridization if the original types were reintroduced from the Cape and subjected to an equal period of selection.

The real mystery in Ixia is why the self-colored forms are so little cultivated. There are at least 7 species with self-colored fls. which should be obtained directly from the Cape, if necessary, for they would all make decided additions to the Ixias that are in common cultivation. These are: polystachya, pure white; tlexuosa, white, veined rose, with fine red and purple varieties; aristata, a superior pink; lutea, orange; patens, bright red; speciosa, ruby-red, and odorata, yellow. Of the eyed forms ovata, a bright red flower, should be added, as this color seems to be lacking among the varieties that are commonly cultivated; also monadelpha in its variety with pale blue segments.

One of the most desirable of all these little-known types is /. speciosa, which is shown in the Botanical Magazine, with a delightful ruby-red color, untouched with any suggestion of purple, lilac or allied shades. This form would seem to promise to the hybridizer the possibility of several distinct shades of red that now seem to be practically unknown in cultivated Ixias. A synonym of J. speciosa is /. craterioides, which is a common trade name, but it is doubtful if the ruby-red form is in general cultivation. At any rate, it has not been sufficiently exploited. The dearth of good colored plates of modern cultivated Ixias is out of all proportion to their commercial and artistic value. The writer has no record of any good one since that published in 1884 in "The Garden."

Ixia flowers are charming in every stage of development. At first the flowers are erect and cup-shaped. They close at night and remain closed on dark days. As they grow older they open wider and become more star-shaped. The reader may judge by Fig. 1187 of the beauty of the flowers in theirdrooping stage. The plants remain in flower for three weeks, though the faded flowers Bt the bottom of the spike should be taken off toward the end of the period. As cut-flowers, they are presentable for a week or two. XV. M.

Culture Of Ixias Out-ok-doors.—The writer has always liked Ixias, but has considered that it is too much trouble to grow them under glass. They are vastly more satisfactory when grown outside. The planting of the bulbs should be delayed until the last moment, because Ixias are more inclined than most things to make an autumnal growth. They should be planted 3 inches deep, as late as November 30. In planting bulbs it is always