A Thesis Submitted for the Degree of Master of Science (1916)
University of Wisconsin

A Hybrid Thornless Rose
Grant Cook

The results of study of experiments with rose hybridization which led to the production of a hybrid rose without thorns and of hardy and vigorous growth are here set forth. The experiments are given with essential details and the results are stated carefully.

Careful descriptions of the thornless hybrid and the parent plants are given and comparisons made as to the likenesses or differences of similar characters of the hybrid and parents, Microscopic studies attempting to fix the more minute comparisons are shown by camera lucida drawings.

The thorn character, left out in the breeding of the hybrid, is thot [sic] to be a separate Mendelian character as it appears to act in accord with the Mendelian theory; but no definite conclusions can be drawn because of the few seeds of roses germinated in the experiments and because no seeds of the hybrid (F. 1) self-fertilized, have been germinated. A descent of the thorn character is shown, however, in a second breeding of the thornless hybrid to another rose.

Very little of the general culture of roses is given, but the appended bibliography will give sources of all the information regarding the rose and its culture which the libraries and catalogues available, can supply.

The microscopic anatomy of roses is partially shown by detailed camera lucida drawings of atoms, leaves, roots and flower parts—from the thornless hybrid rose, principally. No complete anatomical study of the rose has been attempted.

Thornless roses are not new products of experimentation nor in nature. There are at least two thornless hybrid species beside the one considered in this article. One is a German propagation, the name of which I have been unable to learn; the other is a much older French breeding called Rose Zepharine Douhin. Besides these thornless roses, there are several varieties with very few thorns, individual plants of which may be without thorns as Tausendschöne, Boursault and others of more recent breeding. There are also wild varieties with no thorns.

However, a rose without thorns was unknown to me when experiments in hybridization led to that result. My surprise and pleasure in breeding a thornless rose was comparable to the feelings of an inventor or discoverer.

Sometime in June, 1898 or 1899, I found a specimen of wild rose (Rosa blanda species) growing near the shore of Yellow Birch Lake, near my home, Eagle River, Wisconsin. The lower portion of the stalks of this rose were well supplied with thin, needle-like thorns of a delicate texture and not as hard as the usual rose thorns: the upper portions of the stalks, the young stems and petioles were devoid of prickles of any kind. The flowers were of course single and showed no marked superiority over those of other wild roses growing near. I marked the bush and in the fall, transplanted it to a corner of the home garden where it has since spread into a fine mass of shrubbery bearing an abundance of five-petaled roses in June and in winter beautiful because of the dark red stems and cardinal hips.

In 1904, while trimming up the neglected shrubbery in an old garden at a home just purchased, a small, hardy rose bush was found growing in a tangle of grass and shrubbery. It was carefully transplanted in a rose bed as I knew the people who once owned my home, had grown some very choice roses. The rose lived and has bloomed freely every summer since. It is a rose of very slow growing habit and is now not much larger than when found. The flowers are large and very double, produced in profusion, resemble a Boursault or wild rose in color and a cabbage or Centifolia rose in shape. The petals are medium thick and velvety in texture. I have never learned the name of this rose as the mother who cultivated the roses had died and none of the remaining family remembered enough about their roses to name any species definitely.

The thorns on this rose are few in number, short and blunt and are somewhat easily rubbed from the stalks. There are no prickles on the petioles nor midribs of the leaves. The rose is remarkable for its lack of the spreading habit—no suckers nor underground shoots appearing in the ten years it has been a part of my experiments. Usually no stamens appear as the rose is so double, but the hips often develop on flowers not cross pollinated in experiments, showing that fetilization could be effected by pollen carried by bees.

In June, 1905, this rose which I will now designate as the first female parent (1st. ♀ P.) was pollinated with pollen from the wild Rosa blanda first described and which I will call the first male parent (1st. ♂ P. ) and the flowers carefully screened with gauze to prevent the visit of bees.

Three hips out of the five artificially pollinated, grew and developed seeds. Each ripened hip averaged about eight seeds that were et all plump and developed. The seeds were washed from the pulp but were not allowed to dry, being planted at once in a six-inch pot. This was kept in a warm and sunny place over winter, as nearly at a uniform temperature of 70 degrees F. as the conditions in the house would allow. A piece of glass was used to cover the pot to guard against excessive evaporation.

Early in April, 1906, two seedlings appeared—the great majority of the seeds tailing to sprout altho each had an embryo somewhat developed as I found upon examination &few afterwards. Both seedlings were set out in the garden late in May where they grew rapidly. The seedlings looked very much alike and we did not discover that one was without thorns until along in July. The other was vary thorny and was left out in the rose bed the following winter where it died. The other, the thornless seedling, termed the first filial generation (F. 1.), was carefully cared for in the house over winter and again transplanted to the garden in May, 1907, after being divided into two plants.

The second winter, one of the thornless plants was left in the rose bed to teat its hardiness: the other being again potted and kept indoors. Both plants blossomed in July, 1908, but the bush wintered out of doors gave larger flowers and in greater profusion. Subsequent experiments have proven the hybrid to be a hardy as the parents and more resistant to fungous diseases than any other rose I have ever grown.

The first blossoms clearly showed the relationship of the hybrid to both parents. The flowers are medial in color, size, texture, number of petals, etc. to like characters of parent plants. A detailed comparison of the characteristic growth of parents and hybrid follows in which the likenesses and differences will be pointed out

The flower buds of the Rosa blanda, ♂ parent are rather elongate showing deep rose, almost crimson, petals between the five rather narrow sepals as the bud expands before bursting. The sepals are usually longer than the petals in the bud and et first are folded together about the apex of the folded petals, There are few leafy lobes on the sepals, The calyxes are deeply cup shaped with a waxy bloom that easily rubs off leaving the surface glossy. Pedicellate glands are common but not numerous on calyxes and sepals, being principally found along the margins of the sepals.

The flower buds of the old garden rose, the parent, are sturdy and stout with wider sepals act with several leafy lobes, the petals showing a bright rose between, in the expanding bud. The sepals are also longer than the infolded petals and at first are folded together about the apex of the bud, The calyx is more broadly cup-shaped than in Rosa blanda. The sepals, and calyxes are well supplied with pedicellate glands but not to the extent found in moss roses. The peduncles of the flowers are also somewhat hairy with pedicellate glands and branching, with or without small stipule-like leaves at the axils of the branches. Stipule-like leaves at the branching of the flower stalks are the rule in the Rosa blanda. parent.

The flower buds of the thornless rose are elongated like those of the staminate parent but more sturdy and with a medial number of leafy lobes on the five sepals which are also in contact beyond the apex of the compact bud. Between the spreading sepals, the bright rose to crimson petals show very much like a Rosa blanda bud. The calyxes resemble those of the wild parent but are somewhat sturdier, approaching those of the pistillate parent. Pedicellate glands are freely borne on the edges of the sepals and in lessening ratio on other portions of the sepals, the calyxes and peduncles. The calyxes and flower stems are marked b a waxy bloom as in similar parts of the male parent; a bloom which easily rubs off leaving a glossy surface.

The flowers of the wild Rosa blanda are a bright rose pink in color; in the bud and when first opened, almost crimson. The blossoms average about two to two and one-quarter inches in diameter. There are but five petals which never open to the point of being recurved. The petals are comparatively thin, of a velvety texture and show little veining.

The very double, cabbage-rose-like flowers of the ♀ parent are of a common rose color, a little lighter than the bloom of the ♂ parent. They average from two and three-fourth to three and one-fourth inches in diameter. The petals are comparatively thick, about like the petals of the American Beauty or Killarney roses, very velvety in texture and show the veining of the vascular bundles plainly.

The blossoms of the thornless rose are a brighter rose color than those of the double mother plant—nearly as bright as the first opened flowers of the wild rose parent. They average from two and one-half to two and three-fourths inches in diameter. The number of petals ranges from twenty to forty with an average of about twenty-five. They are medial in thickness to petals of parents and of a velvety texture with some veins showing. In bright sunshine or when the flowers are three or four days old, the outer petals sometimes recurve in more or less marked degree than in the flowers of either parent.

Little difference can be noted in the odor of the hybrid rose and the parents, which are both very sweet scented. If there is a difference, it is in favor of the odor of the hybrid—some claiming that the odor is identical with the wild parent, only stronger. I have not been able to distinguish any difference, all three being equally sweet to me.

The stamens on the wild rose are borne just outside the thickened circular rim of the receptacle or calyx lining. They are many in number, forming a circle of bright yellow about the pistils, the filaments are unequal in length, the longer ones being outermost in the opened flower. Any regular order of the stamens in whorls is difficult to determine as they are all attached so compactly about the receptacle pert of the calyx, but as a regular order can be distinguished in the petals of double roses; which are with the exception of the five outer petals, assumed to be modified stamens, it is fair to assume that the stamens are in regular whorls of five petals each. The anthers dehisce by longitudinal slits, often becoming distorted and twisted about the filaments which are attached to a point near the center of the anthers.

The pollen grains are small, yellow, waxy and smooth and are filled with granular material, The grains quickly dry out in the sun and wind and are not virile if dry as many of my experiments in arose pollination have shown.

The stamens are wanting in the double roses, their places being taken by petals. Sometimes in the later flowers, stamens are found, usually not more than three or four. A modified petal with half stamen is quite common but I have never succeeded in getting the pollen from such a modified anther to fertilize the ovule of the same rose nor those of other roses. Neither have I succeeded in cross pollinating with pollen from the regularly formed anthers but I cannot assume sterility in these, as pollen tubes will form in a culture of water and extract of the pistillate parts of the rose. However, I have never successfully crossed another rose with pollen from the double or ♀ parent.

The thornless rose stamens are very much like those of its staminate parent in size, shape and general appearance in dehiscence. They are of course not so numerous as many are modified to for petals. The pollen grains cannot be distinguished from those of the wild Rosa blanda parent.

I have secured one hybrid by pollen from the thornless rose upon a fourth rose, the details of which will be given later. Modified petals—partly stamens—are quite common in the flowers of the thornless rose, sometimes a gradual gradation from petals to regularly formed anthers may be seen in a single flower.

Dissection of the hips shows the hybrid and parents to be very much alike in fruiting characters. In each, the outer portion of the calyx la fleshy and develops into a coral red to cardinal, fleshy fruit having a thick skin and waxy cuticle. The hips of the parent plants may easily be distinguished from each other by size, shape and color, but the hybrid hips vary so much that some may be found closely resembling those of either parent.

The seeds are borne on the receptacle lining of the calyx cup. Dissection of young buds shows the ovaries to be arranged in quite regular, serried rows around the receptacle. But small number of ovaries are fertilized. This is due to the fact that few of the styles push up thru the narrow orifice of the receptacle and bring the stigmas where pollen may lodge upon them. There is no regular order in this growth of the styles and therefore no regular order of ripened seeds on the receptacle. In the ripened fruit, the seeds are crowded so compactly together that the individual akenes are modified in size and shape and add to the apparent lack of order.

The persistent sepals and the protruding ends of the styles and stigmas become brown and withered. The mouth of the receptacle grows together as closely as possible about the styles as the fruit ripens. A waxy exudation helps to seal the fruit against moisture and fungous spores. The unfertilized ovules with their hairy styles serve to fill the interstices about the seeds and gradually break down into a waxy gelatinous mass as the seeds ripen. The wild rose has by far the greatest number of seeds and the thornless hybrid seeds outnumber those of the double garden rose.

The general structure of Rosa blanda shows a vigorous growth. The sturdy, dark-red stalks grow to the height of three and one-half to four feet with few, if any, branches the first year. Spreading by under ground stems is very common. Thorns are found only on the lower eighteen inches of the new stalks and are fine and. needle-like. They are very flexible when young and more easily broken off than thorns of most roses as the stalks mature. Above the thorny area, the red stems are covered with a waxy bloom that brushes off easily leaving a polished surface of smooth epidermis. The stems branch freely the second year; all branches excepting those that flower, developing for next season's flowering. About three or four years is the life of an individual stalk.

The old garden rose, the ♀ parent, presents an entirely different aspect. The stalks are not vigorous in growth, never exceeding twenty to twenty-four inches in height; are grayish green in color and have thorns scattered without order over the entire surface. The clear spots of young stems are covered with a waxy bloom. The thorns are short and rather stout but are flexible an easily detached. This rose has never spread by under ground stalks and continues to grow solely from the old stem above ground. Cuttings do not root and I have never tried budding nor grafting, so have only the one plant which remains about the same size from year to year.

The thornless rose is of very vigorous growth. Stalks of bright vintage-red make an average growth of nearly six feet in a season. The bark is covered with a waxy bloom that resembles that on a Delaware grape. There has never been a thorn on this rose. The manner of growth is very much like that of the ♂ parent but the height equals the sum of the annual growths of both parents, a character Mendel found in the hybrids of the first generation in the case of certain garden pea hybrids. The hybrid rose resembles the staminate parent in the great number of blossoms and the pistillate parent in the weighting down of the bushes with the burden of bloom.

The differences in leaves are in the number of leaflets: the wild rose bearing 7, 9 and 11 (very often 11) leaflets; the garden rose 5 or 7 or rarely 9 leaflets and the hybrid 7 or 9 or rarely 11 leaflets. The hybrid (F. 1) rose is like the ♀ parent in glossiness of the upper surfaces of the leaves and in their red color when young. The leaves of all three roses are about equally supplied with hairs upon their under surfaces.

The roots of all three roses are apparently alike and no minute comparisons were attempted. The F. 1 rose roots were used for microscopic studies and will be described later. No essential differences could be found with the microscope, between the roots of Rosa blanda end. the hybrid thornless. The roots of the female parent were not examined with a microscope.

One curious feature of the experiments was a pollination of a fourth garden rose with pollen from the thornless rose. The mother plant bears a pale pink but quite double blossom, the leaves have five leaflets and the strong thorns are borne in pairs directly beneath the leaf nodes and on no other parts of the stems. The bush is medium in size, never growing over three feet in height. The stems are green and the leaves somewhat thick and glossy. The flowers are about two and one-half to three inches in diameter and although quite double, have a plentiful supply of anthers which bear presumably virile pollen.

The anthers were very carefully cut off before the roses were quite ready to unfold their petals and later, pollen from the thornless hybrid was carefully applied with a camel's hair brush. The flowers were then securely covered with gauze which had been over the blossoms since the buds were opened and the anthers removed.

Only one plant was the result of about thirty seeds planted, all of which looked well developed. This second hybrid has the character of the ♀ parent in the color of the flower only. It resembles the ♂ parent in size of the plant and in general appearance. The stalks are slightly tinged with red—a blending of stem color. The pale but double flowers are the size of those of the wild, Rosa blanda grandparent. The stems are thorny for about eighteen inches above the ground like the Rosa blanda, but the thorns are a blend in size and shape of those of the immediate parents. This hybrid is of no value but is kept in cultivation to show the curious inheritance of its freak features.

Neither of the hybrids sends out underground stems and they therefore do not spread as the Rosa blanda and also the pale-flowered garden rose spread. Cuttings of the hybrids and of Rosa blanda root quite easily. This ability to grow from cuttings is marked characteristic in roses. Some varieties will grow from cuttings and others will not root, so that budding and grafting are resorted to as a means of propagation. Greenhouse men know these facts well and bud or graft their roses that will not grow from cuttings. An attempt is then made to get rid of the graft stock by planting the grafted portion so low that it will root, after which the stock may be severed. With a rose on graft stock, there is always the chance that the original stock will send up shoots that will take all the strength and spoil the grafted portion.

Most of the looks on rose hybridization state that a rose will be fertilized by its own pollen more quickly than by pollen from another rose. I do not believe this to be true from experiments tried. If the anthers are left on the rose, the dehiscence scatters the pollen at exactly the most virile part of its existence right where it will naturally fail upon, or near, the pistils. Self pollination will therefore have many times more chances to occur than when pollen is introduced artificially.

The anthers of the rose to be crossed must therefore be removed before they dehisce. In some roses, this occurs before the flowers fully open so that the petals must be unfolded or plucked off and the anthers removed. Some experimenters always pluck off all the petals before clipping off the anthers but I have found it better to unfold the outer petals carefully, then clip the anthers. The petals left on the flower help to maintain a more natural condition for fertilization even tho the rose be covered with gauze.

Some English experimenters claim that cross-pollination of roses should be done on a very bright, sunny day. This may be a good weather condition for England; but in northern Wisconsin, where my experiments have been made, I have found a very sunny day, especially if there was a drying wind, to be unfavorable, I believe that my best results were from the cross-pollinations made before nine in the morning and I prefer very warm but humid weather. I have concluded that the pollen is very much affected by the weather conditions, especially dry winds, not sun or continued rain.

Other experimenters write that the hips should not be allowed to become too ripe but should be picked and the seeds be planted at once or the hips kept packed in sand until the next spring, t have never obtained a needling from seeds that were not from fully ripened hips. I believe that the hips should be left to ripen until the fruit softens and the skin is a dark coral-red color. I have never tried to keep the hips packed in moist sand. I did try keeping the hips and washed-out seed dry one winter and from about a thousand seed not one germinated. Planting at the time of gathering the seed is advisable as the seed coats soon dry and harden.

Roses are valuable for their lasting qualities as cut flowers. In this respect, the thornless hybrid excels both its parents. The thornless rose is distinctly a garden rose and does not do well indoors nor in the greenhouse. Its value as a winter bloomer is therefore not to be considered. But as a shrub on the lawn or in the garden, the great profusion of flowers which make the whole plant a huge bouquet for three weeks or more, fixes it in the class of roses well worth cultivating, The thornless habit makes it particularly desirable where a rose with thorns would not be suitable.

Plate 1 shows a five-year-old bush of the thornless rose in full bloom—a bouquet nearly six feet high. There were more than two hundred roses and buds on this bush at one time by actual count.

 

Plate 2 shows a bunch of roses cut from the thornless rose. The largest rose in the bunch measured three inches in diameter.

MICROSCOPIC STUDIES OF THE ROSE

Cross sections or the year-old stems of the thornless hybrid and parent plants are chiefly remarkable in comparison because or their great similarities and little differences.

The waxy cuticle shows about the most consistent variation in sections from the three stems. The male parent has a cuticle that ranges from 12 to 16 mm in thickness, the female parent cuticle ranges from 9 to 12 mm in thickness, the hybrid is medial, ranging from 10 to 14 mm. This characteristic cuticle difference is shown in plates 5, 6, and 7. Too much is not to be made of this difference as an inheritance character in the hybrid as not enough experimentation has been made to determine whether or not the waxy cuticle may not vary in all the plants with different conditions of culture.

Plate 5 also shows a great difference between the wild Rosa blanda and the tame garden rose in the thickness or the annular woody growth in a year old stem and that the thornless hybrid is medial in this respect. The difference holds true in regard to the parent plants as may be expected from a gross comparison of the size and vigor of annual shoots from each The comparison does not hold true in ease of the thornless plant as the large shoots show an annual growth that is much thicker than that of the gild parent. Here again the conditions or growth may be considerably changed by culture so that no definite conclusions may be drawn.

Examination of a number of stems of the garden rose indicated that the woody annular growth was not uniform in thickness, but thicker or thinner in different parts of the same year's growth. As there was no evidence of a diseased condition, the conclusion that this may be a characteristic of the plant, is offered.

The epidermis of the year-old stem is composed of a single layer of cells and no cork is formed, The cutinized outer walls or the epidermis are constantly renewed by the living protoplasts as long as the stalk lives. Aside from the difference in thickness in the cuticle, the epidermis of parents and hybrid are essentially alike in structure, excepting the thorns of the parents. No experiments were made to determine what relation the cuticle thickness bears to hardiness; but Stevens (Plant Anatomy, pp. 64-65) says that there is such relation—that Alpine plants have a thicker cutinous surface than the same species growing in warmer climates.

The waxy excretions upon the surface of the stems are identical in form in parents and hybrid. They are globular to rod-like in form and give the "bloom" to the rose stems by their great number altho each wax particle is very minute—a small fraction of a mm in diameter. Bloom is more noticeable on the thornless rose stems than on those of the parent plants.

The thorns on the parents are wholly outgrowths or modified cells of the epidermis and outer collenchyma tissues. Where thorns are broken off, the collenchyma beneath forms a compact, corky-cutinous layer of very small cells that effectually seals the scar from cell evaporation and prevents the entrance of disease producing organisms. The thorns are also covered with waxy cuticle, continuous with that on the stems. As soon as the thorns cease growing, their cells apparently die, the whole thorn becoming hard and woody.

Collenchyma cells contain chlorophyll bodies in more or less abundance, some scattering starch grains and more or less red coloring matter. This red coloring matter in the cell sap is most noticeable in the stems of the male parent and is nearly lacking in the female parent's stems, The hybrid thornless stems are medial in this color character.

Stomata furnish a connection between the air spaces of the collenchyma, and parenchyma tissues and the outside air. (Plate 11) The guard cells are two in number and are covered externally with the continuous cutinization of the stem well into the opening. The cells directly beneath the stomata are crowded with chlorophyll bodies.

There is no distinct division between collenchyma and parenchyma tissue. The latter forms the larger portion of the bark. Parenchyma cells have thicker wails which are more pitted near the bast. (Plates 7 and 11) Some chlorophyll and some red coloring matter characterize the cells of the outer layers or the parenchyma. The cells of the inner layers contain little or no chlorophyll and no coloring, but are filled with proteids and starch.

No regular starch sheath is found in rose stems. All cells excepting those in the thorns, the large cells of the pith and the petals, anthers and pistils in the flower contain some starch. The collenchyma and cortical parenchyma and the small cells of the pith lead in starch storage. (See plates 10, 11, 12 and 20.)

The starch found in both the parents and the hybrid was of the same construction—mostly simple, but compound grains were common in the same cells with simple grains. The starch from all three plants, tested with iodine, showed a concentric striation like that of bean starch

Scattered thru the lower cells of the collenchyma and in the parenchyma were cells containing simple or compound crystals of calcium oxalate. (Plate 6, 11 and 13.) These crystals were variable in size and in shape ranging from an irregular hexagonal form to a pyramid. The largest crystal was 8 x 10 x 18 mm.

No typical stone cells were found in any of the roses studied, altho some of the older parenchyma cells had thick and stone-like walls. All the cells with thickened wails had pits none of which were found to be branched.

The regularity or the ring or bast tissue just outside of the cambium and phloem parenchyma is irregularly intercepted by the medullary rays. The bast is thereby left in radial groups of tightly packed cells with thick walls and very little protoplast spaces. (Plates 5 and 8.) The bast bundles form a strong and tough protective tissue for the phloem parenchyma just beneath and allow the stem to bend easily without breaking. Bast cells are small but very long—the length being many times the width. They become lignified and their protoplasts die but their function seems not to be impaired. Many pits were found in the older bast fibers that had become lignified.

The cambium differs in no essential from the cambium in other plants. The cells are small an in the growing season, are in process of constant division. They are arranged in a crowded mass in not very regular order between the cells of the woody tissue and the bark. From the cambium cells, the phloem and xylem arise. No differentiation for phloem or xylem is marked in the cambium and a single cambium cell may divide into daughter cells one of which may be phloem and the other xylem. This is an inherent character of cambium but it is not understood.

The phloem parenchyma consists of rather small, thin-walled cells between the bast and the cambium. A few sieve tubes are found in the phloem, elongated cells or which may function as sieve tubes or become sieve tubes in the older stems, The phloem parenchyma cells contain very little proteid or starch.

The woody cells of the rose stem are arranged in comparatively regular segments which are separated by the medullary rays. The woody cells are long, their thick lignified walls always pitted. They have small protoplasts which are utilized for the storage of food, mostly starch. Wood fibers are about the same diameter as the bast cells but not so long and not nearly so tough and pliable. (Plates 5, 9, 12, 18 and 19.)

Broad bands of medullary tissue separate the wood into segments and smaller radiating bands divide the segments, The main rays are three or four cells thick and form a connection between the pith and parenchyma cells for the transferrence of food materials. The medullary rays end at the pith, an irregular line of demarkation being plain; but the connection with the parenchyma shows no such abrupt ending but a gradual change in form to parenchyma cells. The narrow bands within the wood fiber bundles are usually only one cell thick. They form radial connections between the pith, wood fiber cells and tracheal tubes among the wood fibers. In longitudinal sections or rose store, the medullary rays were seen to be interrupted bands and to be in close proximity to the tracheal vessels.

The tracheal vessels are scattered among the woody cells and are more numerous and of larger diameter next to the pith in the annular woody growths.. They are of varying sizes and have pitted and pocketed walls. The cells composing the tracheal vessels become lignified and their protoplasts die but the function of the tubes is not impaired. The pitted tracheal vessels serve to connect all the larger parts of the plant with each other. The spiral forms are found in the young wood of stem and roots and in leaves and flower parts. (Plates 5, 9, 12, 13, and 16-19.)

The cells of the pith region or the stems are of two distinct sizes and functions. Some pith cells are large, larger than any other cells of the plant, and contain very little organic material but are filled with water or air. The small cells are about the size of parenchyma cells and contain a store of organic materials—starch and proteids. All pith cells have comparatively thin walls which are much pitted. The smaller calls are usually connected and fill in the interstices between the larger cells. In longitudinal sections, the small cells are seen to form irregular columns of closely connected cells for the transferrence of sap and food materials. (Plates 5, 10 and 12.)

Leaves of the rose show a typical structure. The epidermis consists of a one-celled layer of irregular interlocked cells with thick cutinized outer walls. Stomata are very plentiful in the lower epidermis but none are found in the upper epidermis which has an unbroken, waxy covering more or less glossy, Hairs are common on the under surfaces of the leaves. They are but extensions of the epidermal cells and are unicellular. There are pedicellate glands on the leaves of some roses but none were found on the leaves of the roses studied, the stipules are of the same general character as the leaves but contain less chlorophyll unless they are large and leaf-like. (Plates 3 and 13.)

The rose sepals are five in number and vary very much in different rose varieties The thornless rose sepals have a very hairy Inner surface. Pedicellate glands are common, especially on the edges. These glands stained with osmic acid showed the presence of steroptine in agglomerated drops. All cells of the epidermis of sepals except those differentiated to form hairs, contain steroptine, a basis for cheaper rose perfumes. The epidermis of the sepals has a very thick and waxy outer surface and many wax particles forming t bloom easily rubbed off. Stomata are found in the inner epidermis which is very much more hairy than the under surfaces of the leaves. The inner structure of the sepals is essentially like that of the leaves and functions in the same manner. (Plates 15 and 17.)

Rose petals are the source of attar of roses. This is an oil that is secreted in the epidermal cells. Tested with osmic acid, the emulsion of oil globules were seen nearly to fill the cells. The structure of rose petals is simple. There are three or more layers of closely packed cells between the epidermal layers, spiral tracheal tubes are common. The papillate surface of the upper (inner) side of the petal gives the velvety appearance. The coloring matter is contained in all cells of the petal but especially in the epidermal cells. (Plates 15 and 16.)

The root structure differs somewhat from the stem structure. Very young roots have a well defined epidermis with numerous root hairs. In older roots the epidermis is generally much broken, almost corky. There is little bast tissue and the bark is not a thick as on a stem. The cambium layer is very distinct and is not so much intercepted by the medullary rays as in the stem. Tracheal vessels are numerous—about as in the stem and the wood cells are appearantly the same as in the stem. There is little or no pith in the root. Starch was found packed in the cells of the parenchyma, medullary and woody cells. (Plates 18-20.)

Roots from cuttings grow out from cells of the cambium region. In one case, a complete circular, scar-like growth covered the surface of the severed end of the stem before roots grew out. In another, the wild Rosa blanda, a single root grew out of the epidermis at the verge of the cut surface. (Plate 18.) The root branches arise from the woody area and contain branches of it.

As I have not succeeded in getting a generation by self-pollination from the thornless hybrid rose, I cannot tell what such a generation (F. 2) would give. From Mendel's experiments, one plant with characters of the male grandparent, one like the female grandparent and two like the thornless may be expected from every four plants. However, there may be a different linking up of any separate or unit character and any possible recombination may result in breeding. That the thornless rose carries the thorn character as a recessive is shown by the breeding to the fourth rose where a thorny stem like that of the male grandparent resulted. (Plate 4.)

No microscopic examination of the nuclei of the hybrid and parents to determine their chromosome number, which is presumed to be constant for all roses, was made. According to the latest theories, not the number of chromosomes but the characters carried in the chromosomes much as a granular wax rod may carry bits of materials adhering to its surface or as an impression made in the component arrangement of such a rod, would account for the leaving out of the thorns in breeding. The loss or thorns is then supposed to take place during the reduction division in the formation or the gametes which unite to form the new sporophyte, a plant with new combination of characters.

No one has yet seen the hypothetical character determiners on the chromosomes and their actual existence is not proven. The hypothesis, however, answers the conditions and may later be proven to be the explanation of inheritance. At present, examination of the nuclei of the plants bred would throw no light on their breeding.

A better prospect to get a class of thornless roses of fine flower form, so that their propagation shall be a financial success, lies in breeding the present hybrid to the hardy roses of the hybrid perpetual and hybrid tea classes that have few thorns. The experimenter in breeding new plants always hopes for the best possible combination of characters in the new varieties. Thus a class of roses without thorns but with all the beauty of blossom that the popular roses of to-day possess, may be the result.