all the constituents of its ash, gths of which might exist there already in excess.

pro

The first three lines of the second of Liebig's diagrams, as given above, appear to us to illustrate very clearly the facts revealed by Mr. Lawes's experiments, and as before observed, the cases here hypothetically put are undoubtedly more common in nature than those of the first diagram. But we profess that we cannot exactly understand the reasoning of the last line of the second diagram; how if the addition of ammonia to excess of minerals increased the duce proportionately in two degrees, this should cease at the third. It appears to us that in this line the excess of mineral, to represent a true case, should be replaced by mineral the same as in the previous line, and representing the whole amount soluble or available in that season; in which case the ammonia would clearly have no further effect. When Liebig says that ammonia and minerals are both necessary, and then says that, both being present, the beneficial effect will be proportionate to the quantity of minerals, and not to the ammonia, then we think he is straining a point. For it appears to be deducible from the facts before us, that if the two classes of constituents are supplied in their true proportion, either by the soil or atmosphere, or by these and manuring substances in addition, the results will be, up to the point where general physical conditions limit the product, in direct proportion to the amount of the properly mixed supply of food. This is the abstract view. It is illustrated by the high manuring of garden operations as contrasted with field work. Practically speaking, as we have before said, most of the minerals will really generally be in excess. Some however require to be added from time to time, and the more frequently in proportion as the soil has been stimulated by the nitrogenous manures to a greater product and consequently quicker exhaustion.

The result of all this is, that the farmer must look upon his business as a pursuit requiring the active use of the reasoning faculties. Rule of thumb, although more generally in use, is perhaps less beneficial here than anywhere else, since the variations of season, the alterations produced on soil by various crops, &c., should all be taken into account in a rational system of manuring, of restoration or amelio

ration of the soil. If a certain system is found beneficial upon heavy land, the farmer of light land must not expect the same results, unless these depend upon some cause unconnected with this peculiarity, and to be safe the farmer ought to know what these causes are. The present volume has indeed throughout impressed this necessarily upon the mind of the reader, and we think that the attentive study of it will greatly tend to turn the mind from the consideration of particular dogmas or universal formulæ, to the observation of facts under the light of ascertained laws.

EXPERIMENTS OF MR. LAWES AND DR. GILBERT ON THE COMPARATIVE EFFICIENCY OF MANURES.

[Abridged from the Journal of the Royal Agricultural Society of England, vol. xii. p. 1.]

In these experiments a field of fourteen acres was divided into about twenty plots, on which wheat was grown with a great variety of manures, chiefly mineral; ammonia being used in few instances and in small quantity, while rape-cake, as affording a certain amount of nitrogen besides carbon, was used in two or three cases.

The experiments of the first year (1844) gave very decided results for the discovery of the special character of the exhaustion produced by previous heavy crops. While unmanured land (No. 3 plot) yielded only 16 bushels per acre, with farm-yard manure (No. 2) 22 bushels were obtained. The ashes of farm-yard manure (No. 4) gave no increase beyond the unmanured plot. Of nine plots manured with artificial mineral compositions, none gave an increase of two bushels, the average of the nine being 17 bushels. But the addition of 65 lbs. of sulphate of ammonia to some of these mineral manures caused an average produce of 21 bushels. A small quantity of rape-cake added to the mineral manures gave about 183 bushels; and 65 lbs. of sulphate of

ammonia, added with rape-cake to the inefficient mineral manure, gave more produce than the 14 tons of farm-yard manure (22 bushels 3 pecks). The quantities of rape-cake used were small, and the increase attributable to it also small, but nevertheless much what might be expected when compared with that from the ammoniacal salts, if, as appears valid, the effect of rape-cake on grain crops is due to the nitrogen it contains.

The results of the second year's experiment show a produce of more than 23 bushels, instead of 16, without manure of any kind, while farm-yard manure produced 32 bushels instead of 22. The standard produce of the soil and season was estimated at about 23 bushels, and the correctness of the result from constantly unmanured land, is borne out by the produce (221 bushels) of an unmanured plot, to part of which superphosphate of lime had been given the year before.

Plot 9 had received in the former year both superphosphate of lime and a little sulphate of ammonia; this second year it was top-dressed once with 168 lbs. of sulphate and the same quantity of muriate of ammonia; the crop was 33 bushels 1 peck.

Plot 10 had received in the former year superphosphate of lime and silicate of potash. This was dressed in four separate applications with 168 lbs. of sulphate and the same quantity of muriate of ammonia, and yielded nearly 32 bushels. In the case of No. 9, the produce exceeded by 14 bushel that given by farm-yard manure; in that of No. 10 it is about equal. If the total weight of the corn were taken instead of the measure of dressed corn, No. 10, by ammonia only, yielded both more corn and more straw than the farm-yard manure with all its mineral and carbonaceous substances.

The results of the third year's experiments were as follows the constantly unmanured land (No. 3) gave not quite 18 bushels of dressed corn as the normal produce of the season. Part of plot 10, so highly manured with ammonia in the preceding year, now unmanured, gave rather more than 17 bushels. The near approach, again, to identity of results from the two unmanured plots, at once gives confidence in the accuracy of the experiments, and shows how effectually the preceding crop had, in a practical point of

view, reduced the plots-previously so differently circumstanced both as to manure and produce to something like a uniform standard as regards their grain-producing qualities.

Plot 2 received, as before, 14 tons of farm-yard manure, and produced 27 bushels, or between 9 and 10 bushels more than the unmanured land.

A part of plot 10, which had in the preceding year given with ammoniacal salts a produce equal to that of farm-yard manure, now gave with 2 cwts. of sulphate of ammonia, more than 274 bushels, the weight of corn exceeding that of the produce of plot 2 by 24 lbs. per acre, while the weight of the straw was rather less. A plot (No. 5), unmanured in 1845, was subdivided into two parts, to one of which was applied the ash of wheat-straw (in the proportion of the ash of 3 loads per acre), by which there was yielded rather more than 19 bushels (an increase of about 1 bushel) of dressed corn per acre; to the other half was given the same amount of ash with 2 cwts. of sulphate of ammonia as a top-dressing, the result being 27 bushels,-that is, only 77 bushels more than the half-plot of No. 5 without the sulphate of ammonia; while on plot 10, the difference with and without the ammonia was 93 bushels. This may be accounted for by the plot 5 having been top-dressed with the ammonia, while in plot 10

this was drilled in with the seed.

Plot 6 had received in 1844 superphosphate of lime and phosphate of magnesia, in 1845 superphospate of lime, rapecake, and ammoniacal salts. It was divided in 1846 into two portions; a received 4 cwts. per acre of Liebig's patent wheat manure, and yielded 20 bushels, or rather more than two bushels more than the unmanured plot; but the manure gave off a very perceptible odour of ammonia, so that some of the increase at least might be due to nitrogenous substances contained in it in addition to the minerals peculiar to it. The other division b, received, besides the 4 cwts. of patent manure, 1 cwt. each of sulphate and muriate of ammonia, and the produce was 29 bushels; that is to say, the addition of the ammonia increased the produce by nearly 9 bushels per acre, while the increase over the unmanured plot by 14 tons of farm-yard manure, was only 9 bushels. In 1847 the produce of the constantly unmanured plot 3 was

about 17 bushels per acre, and that of plot 2, annually supplied with 14 tons of farm-yard manure, nearly 30 bushels.

Plot 9, of which section a had received in the preceding year 4 cwts. of rape-cake, and b 4 cwts. of rape-cake with 2 cwts. of sulphate of ammonia, with no direct mineral manure since 1844, was this year chosen for the trial of some substance rich in carbon (but not in nitrogenous or mineral matter), side by side with pure nitrogenous supply. One half of 9a (9a1) received ground rice, in the proportion of one ton per acre; the other half (9a2) 150 lbs. of sulphate and the same quantity of muriate of ammonia; 9b received the same. The produce with one ton of rice was 223 bushels of dressed corn, while the mineral salts of 9a2 and 9b were respectively 26 and 26 bushels. So that while there was only a difference of half a bushel in the two cases, the ammoniacal salts caused an average increase of 9 bushels over the unmanured plot.

The authors next show by tables the results of experiments on six plots compared with the produce on the constantly unmanured plot, from the year 1844 to 1850 inclusive.

Table V. shows the produce of the unmanured plot compared with that of the plot (10a) which in 1844 received superphosphate of lime and silicate of potash (giving less than one bushel of increase); but subsequently always ammoniacal salts alone.

It is remarkable that plot 3 (the unmanured one) of the previously corn-exhausted soil, yielded seven successive crops of grain and straw without any manure whatever, and that under this treatment it showed no signs of diminished fertility; for the average of the seven seasons was about 17 bushels of dressed corn and 16 cwts. of straw. Whence it appears, that upon a soil of given quality the produce will only vary with the climate and the variations of the seasons, materially affecting the supply of ammonia available from natural sources, upon which depends the assimilation of the other constituents. The results of plot 10a are sufficient to show, that in spite of any deprivation by previous cropping, the soil still contained, relatively to the ammonia available from natural sources, an excess of the necessary mineral constituents.