might enable us to form a fair opinion on the chemical character of the farm. But the difficulties of such analyses, and their cost, are serious objections, if there were no others, to their practical value to the farmer.

2. Of what value are detached analyses of soils a hundred miles apart, compared with a thorough local investigation of the same soil under very different circumstances of culture? With his usual shrewdness, Berzelius led the way in such an investigation by analyzing elaborately a naturally fertile soil, taken from beneath the action of the plough, and the same soil from above the former, where it had been subjected to years of tillage. The differences were tolerably well marked; but, since he did not present us with several analyses of each, we cannot certainly know whether culture alone had the effect indicated by his two single analyses. In a subject so little known, and fraught with difficulty, such elaborate analyses, multiplied a thousand-fold in different localities and under different circumstances, would establish theory on a firmer basis, and then allow of immediate practical benefit; but not till then. The cost of analyses is a serious objection to their practical benefit. If it were only required to determine the amounts of silica, alumina, oxide of iron, organic matter, and perhaps lime included, the analysis might be performed at a moderate cost, and the constitution of a whole farm determined; but the first three of these are precisely those which constitute the groundwork or base of the soil, not the more solvent or diluent of the potential constituents. These last are determined with difficulty, and at considerable cost of time; and since their accurate determination is ncessary, the difficulty and expense of analysis increase in a greater ratio. Having stated that numerous accurate analyses would be necessary to ascertain the chemical constitution of a field or farm, it is evident that expense alone is an impassable bar at present to the wide-spread application of the analysis of soils.

3. The difficulty and uncertainty attendant upon the analysis of soils that has any pretensions to accuracy are such as to render it valueless. Those constituents believed to be of greatest value exist in exceedingly minute quantities in soils, and in an ordinary analysis they are liable to be either left out or grossly exaggerated; in either of which cases, the analysis is useless, because it tells an untruth, and forms an unsound or rather wholly unreliable basis for calculation. Let us examine more narrowly how far what may be called a tolerably correct analysis may be relied on. Alkali, lime, phosphoric, sulphuric and muriatic acids, ammonia, and organic matter, are generally regarded as the fertilizing constituents. Of these, sulphuric acid and lime (including magnesia) may be determined with sufficient accuracy, especially lime, which is often present to the extent of several per cent. The exact determination of chlorine (or muriatic acid) is often impeded from the accompanying organic matter and the volatility of chlorides during evaporation. I would regard inferences drawn from the data of sulphuric and muriatic acids as unsafe in sound farming practice. The precise amount of ammonia is ascertained with difficulty, and the amount given in analyses must be looked upon with some distrust, unless the quantity is unusually large, sufficient to overbalance the errors of analysis. We can ascertain with considerable nicety the amount of organic matter; but of what avail is that knowledge? Are the remaining half decomposed rootlets and organized portions those which give fertility to a soil? or is it that very

unknown humous body, soluble in alkali, and reprecipitable by an acid? This last idea being as yet a mere conjecture, we may summarily dismiss the determination of organic matter, as of no immediate practical utility to the farmer.

Of all the minute constituents in a soil, alkali, or potash, and phosphoric acid, are generally regarded as the greatest cause of fertility; and yet these two are precisely the most obstinate impediments to the accurate analysis of soils. Their precise estimation is attended with difficulty under nearly all circumstances, and peculiarly so where their total amount may fall below one per cent., as in soils. It may be fairly questioned whether the small fraction of one per cent. of phosphoric acid which is usually returned in soil analyses may not often be due to errors of analysis, or be far above or below the true amount. I will assert that no accurate and candid chemist can declare with confidence and truth that he has ever determined, in a soil, the exact amount of this pest of the analyst. Our means of determining it, when in conjunction with alumina, as it is most likely to be or to become in the analysis of soils, are still avowedly imperfect in the hands of the best analytical chemist. There is not much more confidence to be put in the precise estimation of potassa.

It would appear, then, that of all the fertilizing ingredients of a soil, lime can be estimated accurately, but that the precise amounts of the others cannot be given with confidence; while the determination of the most important is the least reliable of all. It is therefore not too strong a conclusion to say that the present practical value of the analysis of soils consists in ascertaining how much lime they contain. Since inferior analyses have been left out of view altogether, and only what may be termed good analyses held under consideration, their uselessness, or rather detriment, to the farmer cannot be too strongly depicted.

4. There is a confirmatory argument against the practical value of soil analyses, which has been so clearly set forth by Major J. F. Lee, of Washington, that I take the liberty of quoting his letter to me on the subject: "We know that on all poor land of proper texture, the application of 200 pounds of guano to the acre will produce fair crops of grain and roots; and this is the difference between a barren and a tolerably fertile soil. Now this guano applies only 6 pounds potash, 24 pounds phosphoric acid, and 34 pounds ammonia. But the acre contains 3,920,000 pounds of soil, (to the depth of a foot.) Can analysis now, or will it in any progress we may reasonably expect it to make, ascertain 1 part of potash in 600,000 parts of foreign matter; or 1 part of phosphoric acid in 150,000 parts of foreign matter; or 1 part of ammonia in 100,000?” It may be answered, without the slightest fear of contradiction, that such determinations are greatly beyond the present power of chemical analysis. Whether they will continue so, I will presently inquire; but the argument is strong against the present value of analysis as applied to soils.

5. Another and I fear a greater objection to the immediate value of soil analyses, is the difficulty of ascertaining how much or what part of a soil should be analyzed. Soil consists of mineral and organic matter in a more or less comminuted state. Suppose that an ultimate analysis were made upon a fair average of soil, ground to the finest powder, would it express the fertile value of the soil during the time we look for remu

nerative crops? If, besides finely comminuted matter, it contained gravel or coarse sand, consisting of quartz, feldspar, greenstone, &c., how long a time will be required for the disintegration of the cohering mineral masses, so far as to allow plants to extract the alkali, &c., which they ask for? Even if we make a previous mechanical separation of the very fine from the coarse matter, and subject only the former to analysis, who would be so presumptuous as to predict how much of this fine matter would disintegrate and yield its rich stores to the husbandman in the course of one or more years, or even of a century? The farmer would, doubtless, prefer to know how much benefit he is to reap in his own life. time, than to leave it to posterity in a future of uncertain length.

Guided by these considerations in the analysis of soils, I employed water, slightly acidulated with acid, to extract the fertilizing ingredients, supposing that my analysis would thereby express the now potential qualities of the soil. I am now, however, more thoroughly convinced that, in our present ignorance of the rate of decomposition of mineral aggregates from atmospheric influences and from culture, such assump. tions, and, of course, their deduced inferences, are merely conjectural. The farmer has enough to contend with, in varying seasons, the depredations of insects, &c., without basing his practice upon conjecture.

6. Assuming that we could obtain a fair average of a soil from a field, that we could analyze it with accuracy and at little cost, and that we know the rate at which mineral aggregates would yield up their sources of fertility, would such knowledge assist us in determining how much of the several active ingredients is wanting to render that soil fertile? Can any one presume to assert, in the present state of our knowledge, how much each kind of plant demands to insure its luxuriance or productiveness? From the observed effects of guano, bones, ashes, lime, and green, sand, as well as from the analyses of ashes of plants, it is fair to infer that ammonia, phosphoric acid, potassa, and lime, possess fertilizing qualities; but the numerical measure of their value is hypothetical, if not conjectural. Much of what we term our knowledge on this subject is an idea floating in the region of hypothesis, and until it alights upon the ground, and can be handled with some degree of certainty by weight and measure, the practical farmer would do well to keep to his well-trodden paths of practice, and rather be content with the accumulating experience of practical trials than depend upon the results of analysis. When lime is applied to land, why is it that one kind is found to produce much more than another? It is certainly not merely because magnesia is present in larger proportion in one than in another; for by far the greater portion of the lime applied to the soil in the United States contains notable quantities of magnesia. Is it because the land has already been saturated with lime? This has not yet been proved by facts. May it not be that one kind of limestone contains more alkali or phosphoric acid than another, although in exceedingly minute quantities? Their presence in limestone has been only recently demonstrated, and the question cannot, therefore, be answered positively. These questions are offered merely to show that we use lime from observation of its value, and not from an absolute knowledge of the cause of its fertilizing effects.

In wood ash, is it alkali, phosphoric or sulphuric acid, or lime, that constitutes its more active principle? In guano, does ammonia or phos

phoric acid give its chief fertilizing character? Does phosphoric acid act without reference to the base with which it is united, whether potassa, lime, magnesia, or ammonia? Does sulphuric acid act with greater potency in combination with alkali or with lime? In general, is it of inferior moment in what combination a so-called fertilizing body is employed, or does it always act as a particular compound?

What duty does organic matter perform, and what is its most suitable condition? Does it enter by the rootlets or by the leaves, to fulfil its functions in the organized structure? Is it crenic or apocrenic, or humic acid, or perchance some condition as yet unknown to the chemist, that chiefly exerts its beneficial influence upon vegetation?

When these and numerous other like questions shall have been answered by a fair union and agreement of sound theory and long practice, then may we hope for numerical data for determining how much of each ingredient is required upon a soil. And when analysis shall become so far perfected as to determine with tolerable precision the quantities of the minute ingredients contained in a soil, then can we apply the required substances by weight and measure, and predict with measurable confidence the results of the application.

Lastly, it will be observed that in the preceding part of these remarks I have confined myself exclusively to the consideration of the practical uselessness of the analysis of soils at the present time. Can we look forward to a period when such analyses can be performed with such accuracy, expedition, and moderate cost as to be available to the art of agriculture? I am well satisfied that such an expectation is well founded. Our assay balances can now show the millionth part of the weight placed in them, and may be still further improved. Reasoning from the past, the methods of analysis admit of almost indefinite improvement; and it is highly probable that new analytic processes will be devised of much greater power, rapidity, and accuracy than those at present known, because every journal of chemical science conveys to us monthly, and even weekly, notices of the progress of chemical analyses. But although soil analyses may not be useful at present to the operative farmer, they may be made available for the advance of scientific agriculture; and for this purpose the enlightened agriculturist should lend his aid, by having analyses of soils most accurately performed; not one or two, but numerous analyses of the same soil under varying conditions. Such investigations keeping pace with the advance of vegetable physiology, will the sooner tend to deliver husbandry from the thraldom of empiricism, and place it under the dominion of a rational system. Besides the analysis of soils thus performed, the analysis of ashes of plants and of manures, by throwing light on vegetable physiology, will contribute to the progress of rational agriculture. Above all other things, frequent and carefully-conducted experiments on manures of known composition, and close and continued observation on their effects on various crops, will accumulate a treasure of experience from which sound theory will draw her data, which will then react most beneficially upon the culture of plants. Then may we look for a literal fulfilment of the expression, that "the desert shall blossom as the rose."

Remarks on the foregoing communication, by Daniel Lee.

Baron Liebig raised the expectations of farmers far above what the infancy of chemical science in its application to agriculture, and especially to the analysis of soils, would justify.

Now there is a strong tendency to run into the opposite extreme, and as greatly under-estimate the value of chemical researches as the distinguished Geisen professor prompted the public to over-estimate their importance. The truth lies between these extremes; and we will endeavor to come as near to it as possible in the suggestions that follow.

A few things in agricultural science may be regarded as settled; and taking these as our place of departure, we are to advance (if we can) from things known to such as are unknown, and thereby increase our present stock of scientific knowledge. It is known that solids in a soil do not, and probably cannot, pass through the walls of cells in plants to nourish them as water and gases do; and therefore all solids must be dissolved in water, or reduced to a gaseous condition, before they can enter the roots and circulate through the stems and leaves of vegetables. Without water to permeate every living cell in a growing germ until the plant reaches its full development and maturity, its organization is impossible. Now what is there to prevent a good chemist, familiar with soil analyses, from taking not one or two hundred grains of earth, but a million grains, and ascertaining what organic and inorganic substances rain-water, as it falls from the clouds upon the farmer's fields, will dissolve in three or four summer months, when the growth of plants is most rapid? Let us suppose that there is found only a tenth of a grain of gypsum. This may be separated after Boussingault's plan without difficulty; and the delicate balance used by Professor Booth in the United States mint, or a balance less delicate, will weigh the gypsum. In the case supposed, the quantity of available sulphate of lime in the soil is in the proportion of one part in ten millions, as determined in the most satisfactory manner.

Common rain water contains many foreign bodies besides pure air, ammonia, and carbonic acid, and the presence of these may vitiate the ultimate analysis. The impurity of rain-water is governed in a great degree by the amount of volatile substances and fine dust diffused through the atmosphere in the neighborhood where the rain is precipitated. M. Barruel, a distinguished French chemist, studied the foreign substances contained in rain-water as it fell in the last six months of the year 1851, in Paris, and found in a cubic metre the following bodies:

Allowing the rain to be only 24 inches in twelve months, there would fall 227 pounds of the substances named on an acre in a year. Of this matter 45 would be nitrogen, 103 nitric asid, 194 ammonia, 12 chlorine, 35 lime, and 11 magnesia. Different localities will doubtless furnish unlike results; but the investigations of M. Barruel present a new feature