Roots are either annual, biennial, or perennial, but in all roots, and under any mode of management, the fibrous parts (radiculae) are strictly annual; they decay as winter approaches, and are produced with the returning vigour of their parent in the spring. Hence the reason that plants are transplanted with most success during the season of their decay: for, as the root almost exclusively imbibes nourishment by the mouths of these fibres, in proportion as they are injured by the removal, so is the plant deprived of the means of support; that sap which is employed in the formation of new fibres, would have served to increase the size of other parts.

The quantity of root I have always observed to increase with the poverty of the soil in which it is growing. A root always proceeds to that direction where food is most abundant; and from a knowledge of this fact, we should be circumspect in our mode of applying manures, according to the crop and ob- ject we have in view. The soil in my own garden being shallow, never produced a carrot or a parsnip of any size; but almost every root consisted of numerous forks thickly coated with fibres; digging two spades deep produced no material advantage, the gardener applying as usual manure to the surface; but by trenching as before, and turning in a small quantity of manure at the bottom, the roots always spindled well, grew clean, and had few lateral fibres. For late crops of peas, which mildew, chiefly from a deficiency of moisture to the root, it is an object to keep their radiculae near the surface, for the sake of the light depositions of moisture incident to their season of growth; hence it will always be found of benefit to cover the earth over the rows, with a little well-rotted dung, and to point it in lightly.

If it be desirable to prevent the roots of any plant travelling in a certain direction, the soil on that side should be excavated, and the cavity refilled with sand, or some other unfertile earth, whilst the soil on those sides of the plant whither the roots are desired to tend, should be made as fertile as is permissible with its habits.

It may be accepted as a universal maxim, that whatever causes an excessive development of root, prevents the production of seed; and vice versa, the productiod of seed, especially in tuberous-rooted plants, reduces the amount of root developed. Thus, frequent transplanting the young plants of the lettuce, brocoli, and cauliflower, causes the production of numerous fibrous roots, and is found effective in preventing the mature plants advancing early to seed.

The early varieties of the potato do not naturally produce seed; but if their tubers are removed as soon as they are formed, these early varieties blossom and bear seed as freely as the latter kinds, a fact suggesting many experiments in the cultivation of shy-blooming tuberous-rooted flowers. Again, if the blossoms of these later varieties are plucked off as they appear, the weight of tubers produced will be very materially increased.

According to the usual acceptation of the term, the roots of plants do not emit excrements, yet it is quite certain that, in common with all the other parts of a plant, they perspire matters differing in their amount and composition in every species. The earth in contact with the tubers of a potato fully ripe contains mucilage, and has the peculiar odour of the root; that in contact with the roots of peas is also mucilaginous, and smells very strongly of that vegetable; and the freshly upturned soil where cabbages have been growing, always smells offensively.

MM. Sennebier and Caradori found that if roots of the carrot, scorzonera, and radish, are placed in water, some with only their extremities immersed, and others with their entire surfaces plunged in except the extremities, the former imbibe the water rapidly, and the plants continue vegetating, but the others imbibe no perceptible quantity, and speedily wither. It suggests also the reason why the gardener in applying water or manure to trees or shrubs, does so at a distance from their stems.

A good rule, for ascertaining the proper distance for such applications, seems to be to make them beneath the circumference of the head of the tree; for, as M. De Candolle observed, there is usually a relation between that and the length of the roots, so that the rain falling upon the foliage is poured off most abundantly at the distance most desirable for reaching the extremities of the roots.

This explains why the fibrous points of roots are usually annually renewed, and the caudex (or main limb of the root) extended in length; by these means they each year shoot forth into a fresh soil, always changing their direction to where most food is to be obtained. If the extremity of a root is cut off, it ceases to increase in length, but enlarges its circle of extension by lateral shoots.

The roots of plants, unless frozen, are constantly imbibing nourishment, and even developing parts; for if the roots of trees planted during the winter be examined after an interval of a few weeks, they will be found to have emitted fresh radicles.

The food they imbibe is slowly elaborated in the vessels of the stem and branches, and there deposited. In general, roots have no buds, and are, therefore, incapable of multiplying the plant to which they belong. But it constantly happens in some species, that they have the power of forming what are called adventitious buds; and in such cases, they may be employed for purposes of propagation.

There is no rule by which the power of a plant to generate such buds by its roots can be judged of: experiment is therefore necessary, in all cases, to determine the point. When there is a difficulty in procuring a suitable stock, pieces of the roots of the plant to be multiplied are often taken as a substitute, and they answer the purpose perfectly well ; for the circumstance which hinders the growth of pieces of a root into young branches, is merely their want of buds. If a scion is grafted upon a root, that deficiency is supplied, and the difference between the internal organization of a root and a branch is so trifling as to oppose no obstacle to the solid union of the two.