The depth of earth, as well as the time and manner of applying, vary considerably; it should never be put on until four or five days after the bed is formed : before it is applied, the edges of the bed should be raised full eight inches higher than the middle, as from the additional weight of the frame they are sure to sink more and quicker, thereby often causing the earth to crack and injure the roots of the plants.

The roots of plants being liable to injury from an excessive heat in the bed, several plans have been devised to prevent this effect. If the plants in pots are plunged in the earth of the bed, they may be raised an inch or two from the bottom of the holes they are inserted in by means of a stone. But a still more effectual mode is to place them within other pots, rather larger than themselves; a space tilled with air being thus interposed between the roots and the source of heat, an effectual security is obtained. To prevent the same injury occurring when the plants are in the earth of the bed, a moderate layer of neats'-dung laid between the earth and the fermenting mass, is an efficient precaution, and is much preferable to a similarly placed layer of turf, which interrupts too much the full benefit of the heat. A plan recommended by Bradley is well worthy of notice. A woven hurdle somewhat larger than the frame being placed upon the dung, on this its woodwork can rest, and the earth is laid within it, thus the whole can be moved together without disturbance.

This would especially he of advantage when bark is employed, which requires occasional stirring to renew its heat in case of emergency, when time cannot be allowed for the bed becoming regular in its heat before the plants are inserted. Besides these precautions, vacancies should be left in the mould, and holes bored with a thick pole into the bed, which must be filled up with hay or dung when the danger is passed.

For ascertaining the internal temperature of the bed, the thermometer is the only certain guide, as it also is for judging of the temperature of the air within the frame; the mode of introducing it into the body of the bed, is to have the thermometer inclosed in a wooden case of the size and form of an ordinary dibble, which is to be lined with baize and fitted with a cap of thinned iron to exclude the exterior temperature. The end which enters the earth is shod with perforated copper. In conjunction with the thermometer, trying sticks may be employed for occasional observation; these are smooth laths of wood, about two feet in length, thrust into different parts of the bed, which, being drawn out and grasped quickly, afford a rough estimate of the heat of the bed.

The small extent of the frame, and the rapid deterioration of the air within it by the plants, render its frequent renewal necessary. To effect this, the common practice is to raise the glasses in proportionate heights according to the state of the air; and to prevent any injury arising when necessarily admitted during inclement weather, mats are hung over the opening; but notwithstanding these precautions, the supply of air can seldom be regular; hence, and from sudden chills, the plants are often checked, and sometimes essentially injured. It may be remarked here, that raw foggy days, if anything, are more unfavourable than those that are frosty for the admission of air. A complete remedy for all these difficulties is afforded by a plan, which succeeds on the principle that warm air ascends, and simply consists of a pipe passed through the body of the bed, and one end communicating with the exterior air, the other opening into the frame, at one of the top corners of which an aperture must be made; the heated air of the frame will constantly be issuing from this aperture, and its place supplied by that which rises through the pipe.

A pipe of lead may be used, about two or three inches in diameter, bent nearly at a right angle, and each limb being three feet long, one of these to be placed horizontally, as the bed is forming, with its mouth extending in the open air, that of the other opening into the frame; a cap should be fitted to the first, and by a slit on its under side, the quantity of air admitted can be regulated.

Although stable manure is generally employed for the constructing of hotbeds, yet there are several other vegetable matters that are also in use for the same purpose. Tanner's bark, from its long continuance and regularity of heat, is much to be preferred, especially for very tender exotics. In many situations it can be obtained at a cheaper rate than stable dung; it should be employed when fresh drawn from the vats, or at most when a fortnight or three weeks old; it must lay in a heap for six or eight days to allow the escape of the superfluous moisture: in summer this is not of such material consequence, as an excess of wet is, at that season, not so liable to prevent fermentation.

If the ground is dry, a pit three feet deep may be dug, and is better lined with slates, boards, or brickwork, but whatever may be the nature of the soil, it is best to form this case or bin of a similar height upon the surface. Without some support the tan will not form a solid bed, and if mould becomes mixed with it, the fermentation is retarded or entirely prevented. The breadth must not be less than five or six feet, or of a length shorter than ten or twelve, otherwise the heat will not be lasting. When the bark is laid, it must be gently settled with the fork, but never trodden upon; for if violently compressed, it loses the power of fermenting; if the bark is fresh and not ground very small, it attains a sufficient warmth in a fortnight for the insertion of the plants, and will continue in heat for two or three months; the larger the fragments of the bark are, the longer time it requires to ferment, but in an equal proportion it attains a higher temperature and preserves it much longer; a middle sized bark is, therefore, in general to be preferred; and added to the above consideration, it is | to be remarked that, when made of large fragments, violent and sudden excesses often arise, even after the bed has been constructed two or three months: on the contrary, if very small, the fermentation soon passes off.