inner cells on the other hand may obtain oxygen from the aerial parts, and thus with less moisture be able to keep pace with the epidermal cells growing with more moisture and less oxygen. In ordinary air the moisture and the oxygen reach the epidermal cells more abundantly than the inner ones, consequently the numerator of the fraction is increased as well as the denominator decreased, and hairs are developed. Upon the upper side of a corn root growing along the surface of water abundant hairs were developed, while the under side remained smooth. The difference in length between the epidermal cells and those of the cortex on the haired side was 20μ, and on the smooth side 6μ. KRAUS's tables (37, p. 254) dealing with the lengths of epidermal and cortical cells in relation to hair production are not very complete, and it seems useless to attempt to harmonize the results with those here reported.

KRABBE (33, p. 491) reports the inner cells of pith to be less turgescent than the outer ones when placed in water at 1-2°C, on account of the resistance to the passage of water offered by the protoplasts. According to VAN RYSSELBERGHE (70, p. 103) the influence of temperature is exhibited not in the total amount of water taken up, but in the rapidity of its passage. In warm water, therefore, the water reaches the inner cells and allows them to elongate sufficiently rapidly to keep pace with the epidermis, which is thus allowed to elongate to its full capacity and shows no hairs.

In the zone of hairs on seedlings in water cultures the available energy and the temporary retardation of growth (evidenced by the short outer and still shorter inner cells, and by the curling of many roots) combine to produce hairs. Also the presence of food may act as a stimulus to cause the cells to divide rapidly and form a thick root, whose inner cells do not get sufficient water, or oxygen, or both, to allow them to elongate as rapidly as the outer ones. Later, in the case of corn, the plumule elongates and probably supplies the inner cells with more oxygen. These are therefore better able to elongate, they are carried further from the food supply, division is less active, the roots grow more slender, the water supply of the inner cells increases, still greater elongation takes place, and the epidermal cells are allowed to stretch to their full capacity. Accommodation to a decrease of oxygen is mentioned by PFEFFER (65,

p. 70), and MER (53, p. 1279) speaks of the roots becoming accustomed to the medium.

The curving of corn roots in water is, according to Miss BENNETT (6) not aerotropic. BEAUVÉRIE (4) considers the turning up of water roots to be due to negative hydrotropism, for by using physiologically dry solutions he was able to get them to grow downward. In an experiment in which a slow stream of tap water was passed into the bottom of a vessel in which the roots of corn seedlings were growing, every one turned down, and grew straight and entirely smooth. The stimulus may have been a rheotropic one, or it may have been the presence of fresh aerated water which caused the omission of the hair zone.

An apparent exception to the explanation offered appeared in one root of sunflower grown in 0.5 N saccharose solution, in which the epidermal cells were shorter than the inner ones and still produced hair. Close to the tip, however, the papillae were found on cells shorter than the cortical cells, which makes it seem probable that the epidermal cells on the upper part of this root were shorter than the cortical cells from the start, as is the case with Elodea. In this plant the epidermal cells at the tip are very much shorter than those of the inner cortex, and the difference does not entirely disappear as the root grows older. Consequently there is not the same relation between the epidermal and cortical cells when hair is produced, as there is in corn. Measurements of the cells of roots of Elodea growing in soil, quartz, and water give the following averages in millimeters:

Upon examination of the table the greatest relative length of the inner cortical cells is seen to be in water, and the least in soil, with the hairs in inverse relation, as was the case with corn.

On the concave side of curved roots of corn the epidermal cells are shorter than the inner ones and at times show more hairs (fig.

8). Here the retarding action of the inner cells upon the epidermis is aided by the compression brought about by the curve. SACHS (71, p. 466) has shown that the average length of cells in a curve is less than in a straight portion of the root. MACDOUGAL (49, pp. 352-3) criticises SACHS' methods and reports the cells on both convex and concave sides longer than those on the same region of a normal straight root. His statement that the hairs are "abundant on the regions apical and basal to the region of greatest curvature, but are also wholly absent from the region exhibiting the shortest radius of curvature," seems to mean that the roots geotropically stimulated elongated at the curve and ceased to produce hair. In curving roots of corn growing in water, the epidermal cells appear to be restricted in their elongation, for curving almost invariably causes hair to develop. SCHWARZ noted this and called it "nutation" (75, p. 159). This term did not seem appropriate, and for want of a better word “kinking" has been used in this paper.

Transference from a solution of low osmotic pressure to one of high osmotic pressure appears to withdraw so much water from the epidermal cells that they do not grow into hairs. When the reverse order is followed there is a better chance for the epidermal cells to absorb water and to grow before the inner ones, and in this case some hairs appeared. The problem of the effect of osmotic solutions upon roots is quite different from that relating to filamentous algae and fungi. In the last two cases each cell is bathed in the solution to be tested, while with roots the action of the neighboring cells influences the epidermis, and on account of the thickness of the root the inner ones are not affected just as the outer ones. If a solution could be made which by its osmotic strength or chemical composition would retard the growth of the inner cells and allow the epidermal cells to grow, hairs might be expected. In one or two instances the epidermal cells of roots of sunflower and corn growing in o.-0.2 N solutions seemed to become accustomed to the solution before the inner cells, and thus were able to grow out as hairs while the growth. of the deeper cells was still retarded.

The retarding effect of diminished food supply on the production of hair on the internodes of the stem of potatoes is reported in a short note by KRAUS (38). In experiments with half seeds, one or two

cases occurred in which the central cylinder was torn apart at regular intervals by the stretching cortex, the epidermis bearing no hairs. The food supply seemed not enough to give the cells of the central cylinder sufficient strength to retard the stretching of the outer cells.

No change in turgor is needed to explain the appearance of root hairs, for according to PFEFFER (642, p. 29) there is no change when growth is accelerated by a rise of temperature or by absence of light, or when growth is retarded by lack of oxygen or (66, p. 296) by pressure. In the first three cases hairs disappeared or were diminished, while in the last they appeared.

An interesting relation was noticed between the epidermal and the hypodermal cells of some corn roots. In roots growing in the air and producing hair, the nuclei of the hypodermal cells were usually larger than those of the epidermal or cortical cells (fig. 9). This may indicate that the hypodermal cells were passing food to the outer cells, the starting of the lateral growth thus initiating a movement of material in that direction. SAUVAGEAU (73, p. 171) reports small hypodermal cells under the piliferous cells in Zostera. This demand for food by the outer layer would decrease the supply in the central cylinder and may account for the inverse relation between root hairs and lateral roots, noted by LESAGE (42, p. 110), COSTANTIN (9, p. 149), MER (52, p. 666; 53, p. 1278), SACHS (71, p. 589), et al. In Eichhornia the lateral roots extend nearly to the tip, but there are no root hairs. This activity of the central cylinder, contrasted with that of the epidermis, is in harmony with the results of the experiments here reported.

In spite of the structural and functional similarity which often exists between root hairs and rhizoids, it does not seem appropriate to consider them together. In the first place, they are not morphologically similar, rhizoids being of gametophytic origin and root hairs developing from the sporophyte. The fact that rhizoids arise usually from a rather small gametophyte, all the cells of which retain in large measure their primitive condition, may account for the irritability they display toward geotropic, phototropic, and thigmotropic stimuli. Root hairs, on the other hand, are developed on a highly differentiated organ of a highly differentiated sporophyte, and are not thus sensitive, a difference pointed out by HABERLANDT (22, pp. 194-5).

It would be well to limit the term "root hair" to hairs borne by morphological roots only.

SUMMARY.

1. Light and darkness appeared to have only an indirect effect, through their influence on growth.

2. High temperature with sufficient moisture tended to decrease hair production by increasing the elongation of the internal cells. 3. The slower the growth in air the better the hair development. 4. Retardation of growth by glass tubes, by wounding, or by resistance of the substratum favored hair production.

5. Roots of seedling corn in water first curled and produced hair, possibly because of the retardation of growth by the diminution of oxygen or its presence in the dissolved state. Later the roots grew straight and smooth, either on account of accommodation to the oxygen supply or because the gas was supplied by the aerial parts. 6. Saturated air with high temperature tended to suppress hair development (cf. 2).

7. Saturated soil tended to suppress hair in corn and wheat, but other factors must be considered when Elodea develops hair in the substratum.

8. Osmotic solutions gave very irregular results on account of some undiscovered disturbing factor.

9. Less hair was developed in distilled water than in tap water. 10. Air deprived of oxygen stopped hair production and retarded growth.

II. Curves and swellings had a favorable effect upon hair development, probably because they represent the retardation of the growth of the root.

12. In all these examples of retardation favoring hair development, not the mere rate of growth, but the differential elongation of the inner and outer cells was of prime importance. Hair production depends on the ratio between the capacity of the epidermal cells to elongate and their ability to do so.

13. The activity of the epidermis may be in inverse proportion to the activity of the central cylinder, lateral roots often appearing when hairs are suppressed, and vice versa.