Popular Science Monthly/Volume 22/November 1882/Scientific Farming at Rothamsted I

From Wikisource
Jump to navigation Jump to search



IN the literature of every department of agriculture, the references to the Rothamsted experiments are getting to be as familiar as household words, and it is now generally admitted that they have had an important influence on English farm-practice.

In this Country, however, the direct and practical bearing of these experiments on the every-day business of the farm is not fully appreciated, and this is perhaps largely owing to the fact that the American farmer is owner of the soil he tills, and is not therefore compelled to give that strict attention to every detail of the economy of farm management that is essential to the successful practice of farming in Great Britain.

It would seem that the leading object of inquiry at Rothamsted has been the solution of agricultural problems, but the relations of science to agriculture are so broad that what may be considered purely practical lines of investigation can not be limited to considerations that are of interest to the farmer only, as they involve the discussion of questions that are constantly presenting themselves in the progressive development of the sciences of chemistry, botany, vegetable and animal physiology, including dietetics and the laws of assimilation and growth, and thus lead to an examination of topics that are properly included in the domain of social and sanitary sciences.

In fact, when the original object of inquiry is the attainment of some practical end, the dominant work of experimentation, when properly conducted, soon comes to be the investigation of special questions that strictly pertain to some department of the allied sciences.

In the well-planned experiments which have been so ably conducted at Rothamsted for more than forty years, embracing a wide range of topics, there is an abundant fund of information that must be of interest not only to the farmer, who looks for results in pecuniary values, but to the man of science, seeking the truth for the truth's sake, and the intelligent general reader who wishes to trace understandingly some of the leading facts in the world's progress.

Without including numerous newspaper articles, and the annual "Memoranda of the Experiments at Rothamsted," that have been printed for several years for the convenience of visitors, nearly one hundred elaborate papers and discussions of the field, feeding, and laboratory experiments, many of which are in the form of monographs, have been published since 1847, every one of which has had its influence on questions of agricultural practice, as well as on the various theories in science that have been prominent for the past half century.

These papers are to be found in the "Journal of the Royal Agricultural Society," the "Reports of the British Association for the Advancement of Science," the "Journal of the Chemical Society of London," the "Proceedings and Transactions of the Royal Society of London," the "Journal of the Society of Arts," the "Journal of the Horticultural Society of London," the "Reports of the Royal Dublin Society," the "Edinburgh Veterinary Review," the "Philosophical Magazine," and other periodicals, and in official reports on special subjects of investigation.

The titles alone of these papers would require the space of several pages of this magazine.

Rothamsted, with its fine old manor-house, the home of Sir John Bennet Lawes, is in Hertfordshire, England, about twenty-five miles from London, on the Midland Railway, Harpenden Station.

Mr. Lawes was born in 1814, and succeeded to his estate in 1822. He was educated at Eton, and at Brasenose College, Oxford. After leaving the university, in 1835, he spent some time in London, in the study of chemistry, which had been a subject of special interest to him when pursuing his academic course.

Soon after taking possession of his hereditary property at Rothamsted, in 1834, he began a systematic course of experiments with different fertilizers, first with plants in pots, and afterward in the field.

"The researches of Dr. Saussure on vegetation were the chief subjects of his study to this end. Of all the experiments so made, those in which the neutral phosphate of lime in bones, bone-ash, and apatite, was rendered soluble by means of sulphuric acid, and the mixture applied for root-crops, gave the most striking results.

"The results obtained on a small scale in 1837-1839 were such as to lead to more extensive trials in the field in 1840 and 1841, and subsequently."

Dr. J. H. Gilbert has been associated with Mr. Lawes since June, 1843, and has had the direction of the laboratory.

"In 1843 more systematic field experiments were commenced; and a barn, which had previously been partially applied to laboratory purposes, became almost exclusively devoted to agricultural investigations. The foundation of the Rothamsted Experiment Station may be said to date from that time (1843). The Rothamsted Station has, up to the present time, been entirely disconnected from any external organization, and has been maintained entirely by Mr. Lawes. He has further set apart a sum of £100,000 and certain areas of land for the continuance of the investigations after his death."

In 1854 a subscription was made by agriculturists for a testimonial to be presented to Mr. Lawes as an expression of their appreciation of the great value of the services he had rendered to British agriculture. The committee in charge of this fund, instead of expending it in plate as had been intended, devoted it, at the suggestion of Mr. Lawes, to the erection of a new laboratory, so that the facilities for experimenting were largely increased.

The eminent services of Drs. Lawes and Gilbert, in the improvement of agriculture and the advancement of science, have been repeatedly recognized. In 1854 Dr. Lawes was elected a Fellow of the Royal Society, and in 1867 the royal medal was awarded to him conjointly with Dr. Gilbert, by the council of the society. The gold medal of the Imperial Agricultural Society of Russia was awarded to Dr. Lawes, and last year the Emperor of Germany, by imperial decree, awarded the gold medal of merit for agriculture to Dr. Lawes and Dr. Gilbert jointly, "in recognition of their services for the development of scientific and practical agriculture."

As a national recognition of the great value of the investigations to which he has devoted his life, Dr. Lawes has this year been created a baronet.

The number of assistants engaged in the work of experimenting has gradually increased. At first only one laboratory-man was employed, but soon a chemical assistant was needed, and then a computer and record-keeper.

"During the past twenty-five years the staff has consisted of one or two and sometimes three chemists, and two or three general assistants, one of whom is generally employed in routine chemical work, but sometimes in more general work."

The general assistants superintend the experiments with animals and the field experiments—the making of manures and their application—the harvesting and weighing of the crops—the selection of samples which are prepared for preservation or analysis, and they also make determinations of dry matter, ash, etc.

"There are now more than thirty thousand bottles of samples of experimentally grown vegetable produce, of animal products, of ashes, or of soils, stored in the laboratory. A botanical assistant is also occasionally employed, with from three to six boys under him, and with him is generally associated one of the permanent general assistants, who at other times undertakes the botanical work." Several computers and record-keepers have for some time past been occupied in calculating and tabulating the field, feeding, and laboratory results. Additional chemical assistance is frequently engaged in London and elsewhere. Professor Way, Dr. Frankland, and Dr. Voelcker, have done more or less work on material obtained at Rothamsted, and their published reports are of great interest. Mr. R. Richter, of Berlin, has for some years past been almost constantly occupied with analytical work sent from Rothamsted. A considerable, but of course varying, force of agricultural laborers find employment in the field-work.

"The general scope and plan of the field experiments has been to grow some of the most important crops of rotation, each separately, year after year, for many years in succession on the same land, without manure, with farm-yard manure, and with a great variety of chemical manures; the same description of manure being, as a rule, applied year after year on the same plot. Experiments on an actual course of rotation, without manure and with different manures, have also been made. In this way experiments have been conducted as follows:

"With wheat, thirty-nine years in succession: thirteen acres, thirty-seven plots, many of which are duplicates of others. On barley, thirty-one years in succession: four and a half acres, twenty-nine plots. On oats, ten years (including one year fallow): three quarters of an acre, six plots. On wheat, alternated with fallow, thirty-one years: one acre, two plots. On different descriptions of wheat, fifteen years: four to eight acres (each year in a different field), now more than twenty plots. On beans, thirty-two years (including one year wheat, and five years fallow): one and a quarter acre, ten plots; also twenty-seven years: one acre, five plots. On beans, alternated with wheat, twenty-eight years: one acre, ten plots. On clover, with fallow or a grain-crop intervening, twenty-six years: three acres, eighteen plots. The land is now devoted to experiments with various leguminous plants, commenced in 1878. On turnips, twenty-eight years (including three years barley): about eight acres, forty plots. On sugar beets, five years: about eight acres, forty-one plots. On mangold-wurzel, seven years: about eight acres, forty-one plots. On potatoes, seven years: two acres, ten plots. On rotation, thirty-five years: about two and a half acres, twelve plots. On permanent grass-land, twenty-seven years: about seven acres, twenty-two plots.

"Comparative experiments, with different manures, have also been made on other descriptions of soil, in other localities. Samples of all the experimental crops are taken, and brought to the laboratory. Weighed portions of each are partially dried, and preserved for future reference or analysis. Duplicate weighed portions of each are dried at 100° C, the dry matter determined, then burned to ash on platinum sheets in cast-iron muffles. The quantities of ash are determined and recorded, and the ashes themselves are preserved for reference or analysis. In a large proportion of the samples the nitrogen is determined, and in some the amount existing as albuminoids, amides, and nitric acid.

"In selected cases—illustrating the influence of season, manures, exhaustion, etc. complete ash-analyses have been made, numbering in all more than seven hundred. Also in selected cases, illustrating the influence of season and manuring, quantities of the experimentally grown wheat-grain have been sent to the mill, and the proportion and composition of the different mill-products determined.

"In the sugar-beet, mangold-wurzel, and potatoes, the sugar in the juice has in most cases been determined by the polariscope, and frequently by copper also.

"In the case of the experiments on the mixed herbage of permanent grass-land, besides the samples taken for the determination of the chemical composition (dry matter, ash, nitrogen, woody fiber, fatty matter, and composition of ash), carefully averaged samples have frequently been taken for the determination of the botanical composition. In this way, on four occasions, at intervals of five years—viz., in 1862, 1867, 1872, and 1877—a sample of the produce of each plot was taken and submitted to careful botanical separation, and the percentage, by weight, of each species in the mixed herbage determined. Partial separations, in the case of samples from selected plots (frequently of both first and second crops), have also been made in other years."

This condensed statement of the plan of the field experiments, and brief outline of some of the work performed in connection with them, from the "Memoranda" for June, 1882, will give some general idea of the extent of the Rothamsted Station, and of the thorough manner in which all operations are conducted; but, in our enumeration of the other lines of inquiry now in progress, we can only mention the subjects of investigation without referring to particulars in the methods practiced, as we wish to save space for a discussion of some of the leading results that have been thus far obtained.

More than one thousand samples of soil have been taken from the experiment-plots, at different depths, for the purpose of analysis, to ascertain the rate of soil-exhaustion under different conditions, and to trace the relations of the soil to the crops grown and to the manures applied.

For nearly thirty years the rain-fall has been measured in a gauge having an area of one thousandth of an acre, and frequent analyses have been made to determine the available supply of combined nitrogen in the form of ammonia and nitric acid that can be obtained by plants from this source. In some cases the chlorine has also been determined. The absorptive capacity of soils and subsoils for water and ammonia has likewise been investigated.

The quantity and composition of drain age-waters under various conditions have been the subject of elaborate and extended experiments for many years, and the results obtained are of the greatest importance.

In 1870 three "drain-gauges" were made, each having an area of one thousandth of an acre, and inclosing the soil and subsoil in a natural state of consolidation to the depth of twenty, forty, and sixty inches, respectively. As the surface-soil in these gauges is kept free from vegetation, and no fertilizers are applied, their drainage represents, in effect, that of a bare, unmanured fallow. In the separate drains of the permanent wheat-plots facilities were provided for collecting samples of drain age-water from soils growing crops without manure, with barn-yard manure, and with a great variety of chemical manures.

Determinations of the nitrogen in rain-water were made at Rothamsted as early as 1846. The ammonia in the rain-fall for fifteen months, in 1853-'54, was determined in the laboratory at Rothamsted, and again in 1855-'56 by Professor Way. Dr. Frankland made analyses of the rain-fall, and also of dew and hoar-frost in 1869-'70, since which time a series of systematic investigations have been conducted in the Rothamsted laboratory.

A large number of samples of the drainage-waters from the experimental wheat-field were analyzed by Dr. Voelcker, the able chemist of the Royal Agricultural Society, and by Dr. Frankland, previous to 1875, while over thirteen hundred samples have been analyzed since that time at Rothamsted. The drainage of the "drain-gauges," from 1870 to 1874, was analyzed by Dr. Frankland, and since that date it has been systematically investigated at Rothamsted.

A full report of these drainage experiments is given in an elaborate paper "On the Amount and Composition of the Rain and Drainage Waters collected at Rothamsted," published in the last three numbers of the "Journal of the Royal Agricultural Society" (1881-'82), which, from its direct applications to questions of farm-practice, and the light it throws upon the obscure subject of soil-exhaustion and on the economy of manures, is undoubtedly the most valuable contribution to agricultural science that has appeared for many years.

Experiments were made for several years with plants representing the gramineous, the leguminous and other families, and also with evergreen and deciduous trees, to ascertain the amount of water given off during their growth.

Observations on the character and range of the roots of different plants, the relative development of leaf and stem, and their composition at various stages of growth, have been made in connection with experiments to determine the differences in the amount and constituents assimilated by plants of different botanical families, under similar conditions, and of the same plant under varying conditions. From these investigations, so far as they have been published, it appears that the chemical relations of the plant and soil are, to a great extent, determined by botanical and physiological conditions.

In the experiments with the mixed herbage of "permanent meadow," for example, it was noticed, even in the first years of the experiments, that "those manures which are most effective with wheat, barley, or oats grown on arable land—that is, with the gramineous species grown separately—were also the most effective in bringing forward the grasses proper in the mixed herbage; and again, those manures which were the most beneficial to beans or clover, most developed the leguminous species of the mixed herbage, and vice versa."

In the produce grown continuously without manure the average number of species was forty-nine. Of these seventeen are grasses, four leguminous species, and twenty-three of other orders. By weight the grasses averaged sixty-eight per cent, leguminous species nine per cent, and species of other orders twenty-three per cent.

In the produce of the plot most heavily manured and yielding the heaviest crops, the average number of species was nineteen; of which twelve to thirteen were grasses, one only (or none) leguminous, and five or six only of other species. By weight the grasses averaged about ninety-five per cent, the leguminous species less than 0·01 per cent, and other orders less than five per cent.

On the plot receiving annually manures that are of little avail for gramineous crops grown separately in rotation, but which favor beans or clover so grown, the average number of species was forty-three, of which seventeen were grasses, four leguminous, and twenty-two belonging to other orders. But by weight the grasses averaged but from sixty-five to seventy per cent, the leguminous species nearly twenty per cent, and all other species less than fifteen per cent.

The "struggle for existence" and the "survival of the fittest," therefore, determine the character of the species contained in the produce under the conditions, and the chemical composition of the crop varies accordingly. With an increase of the leguminous produce the nitrogenous constituents are increased, and with a decrease in the leguminous produce the nitrogenous constituents are diminished.

Experiments with leguminous, gramineous, and other families of plants were made for several years in succession, at Rothamsted, to determine whether plants assimilate free or uncombined nitrogen.

The relations of nitrogen to the growing plant and to the soil and the sources of the nitrogen of vegetation have been prominent subjects of investigation in all the Rothamsted field-experiments.

It is not my purpose, in this connection, to discuss the various theories of vegetable growth, or to give an account of the many controversies that have arisen in the developmental progress of science, but simply to call attention to some of the leading lines of investigation at Rothamsted which have had an influence in correcting our theories of vegetable nutrition and soil-exhaustion and in improving our methods of agricultural practice.

The legitimate aim of all systematic, exact experiments is to lay a foundation of well-ascertained and closely related facts on which may be developed a superstructure of science to supersede the theoretical speculations which form an important part of the prelude of scientific discovery. In this work of reconstruction, Drs. Lawes and Gilbert have for many years occupied a prominent position, and a full account of their labors would involve in the record a history of agricultural science for the past half-century.

From an agricultural stand-point one of the first steps in the study of the laws of plant growth and nutrition is to ascertain the relative influence of the soil and the air in the supply of plant-food, as they are the only sources from which plants obtain the elements which enter into their composition.

The atmosphere is a mixture of gases, of which more than three fourths is nitrogen, and less than one fourth oxygen, with something less than one part in ten thousand of carbonic acid. In addition to these there are traces of ammonia and a variable quantity of vapor of water.

As the carbon, which forms about one half of the dry substance of plants, is all derived from the minute proportions of carbonic acid found in the atmosphere, it has been assumed that the comparatively small amount of nitrogen required by plants could be readily obtained from the abundant stores of this element in the atmosphere, and that wide-leaved plants, like clover and beans, could more readily assimilate it than those with narrow leaves, like the grasses.

Experiments by Boussingault and the elaborate researches at Rothamsted, however, show that free nitrogen, the most abundant constituent of the air, is not assimilated by plants. The atmospheric sources of nitrogen for plant-growth must, therefore, be limited to the minute quantities of combined nitrogen in the form of ammonia and nitric acid.

Important data as to the amount of nitrogen in various field-crops, grown under a variety of conditions, and the sources from which it is obtained, are furnished in the Rothamsted field-experiments.

For a period of thirty-two years, wheat, on plots without manure, yielded an annual average of 20·7 pounds of nitrogen per acre. The yield, however, declined from an average of more than twenty-five pounds during the first eight years to an average of but sixteen pounds during the last eight years of the experiment.

Barley, for a period of twenty-four years, on plots without manure, yielded annually an average of 18·3 pounds of nitrogen per acre, with a decline from an average of twenty-two pounds over the first twelve years to an average of 14·6 pounds over the next twelve years. Mineral manures, containing no nitrogen, applied to barley and wheat did not materially increase the yield of nitrogen in the crop.

A succession of root-crops (with three years of barley intervening after the first eight years), dressed with a complex mineral manure, yielded an average of 26·8 pounds of nitrogen per acre, per annum, over a period of thirty-one years; with a decline from an average of forty-two pounds over the first eight years to 13·1 (in sugar-beets) over the last five years. Afterward, with the change of crop to mangolds, the yield of nitrogen was somewhat increased.

Beans, for a period of twenty-four years, yielded an average of 31·3 pounds of nitrogen without manure; and, with a complex mineral manure, an average of 45·5 pounds of nitrogen per acre. The decline in yield of nitrogen was, however, from an average of 48·1 pounds over the first twelve years to only 14·6 pounds over the last twelve years, when unmanured, and from an average of 61·5 pounds over the first twelve years to but 29·5 pounds over the last twelve years, when mineral manures were applied.

An annual average yield of nearly two hundred pounds of nitrogen per acre was obtained in the clover grown for twenty-seven years in succession on a plot of old garden-soil that was exceptionally rich in nitrogen at the beginning of the experiment. As in the case of other crops, there was a marked decline in the average yield of nitrogen in the last half of the period, and there was also a great reduction in the stores of nitrogen contained in the soil.

The leguminous crops, beans and clover, it will be seen, contain a larger amount of nitrogen per acre than the gramineous crops, wheat and barley. In the Rothamsted experiments it was, however, found that manures containing nitrogen benefited the gramineous crops, while they had but little, if any, influence upon the growth of leguminous crops. The chemical composition of the crop was not, therefore, an index of the manurial constituents required to promote its growth.

When turnips, barley, clover, or beans, and wheat were grown in rotation for twenty-eight years, the average annual yield of nitrogen per acre was 36·8 pounds, and in the mixed herbage of permanent grass-land, when unmanured, the annual yield of nitrogen averaged thirty-three pounds per acre.

The larger average yield of nitrogen per acre in the crops in rotation and in the mixed herbage of the permanent grass-land, as compared with the yield of nitrogen in gramineous crops when grown separately, is not entirely due to the larger amount of nitrogen in the leguminous species themselves, but also to their influence upon the gramineous species which are able to take up and assimilate more nitrogen when the highly nitrogenous leguminous crops have been priating their larger supplies of the same element, as will be seen from the following experiments:

"In alternating wheat and beans, the remarkable result had been obtained that nearly as much wheat and nearly as much nitrogen were yielded in eight crops of wheat in alternation with the highly nitrogenous beans as in sixteen crops of wheat grown consecutively without manure in another field, and also nearly as much as were obtained in a third field in eight crops, alternated with a bare fallow."

And again: "After the growth of six grain-crops by artificial manures alone, the field so treated was divided, and in 1873 on one half barley, and on the other half clover, was grown. The barley yielded 37·3 pounds of nitrogen per acre, but the three cuttings of clover yielded 151·3 pounds. In the next year, 1874, barley succeeded on both the barley and the clover portions of the field. Where barley had previously been grown, and had yielded 37·3 pounds of nitrogen per acre, it now yielded 39·1 pounds; but where the clover had previously been grown, and had yielded 151·3 pounds of nitrogen, the barley succeeding it gave 69·4 pounds, or 30·3 pounds more after the removal of 151·3 pounds in clover than after the removal of only 37·3 pounds in barley."

We will now examine some of the evidence furnished by the Rothamsted experiments, in regard to the sources from which the nitrogen of field-crops is obtained.

As free or uncombined nitrogen cannot, as we have seen, be assimilated by plants, we will next consider the supply of combined nitrogen in the form of ammonia and nitric acid, existing in the atmosphere.

From the earlier investigations of the rain-fall at Rothamsted and likewise on the Continent, it was estimated that from eight to ten pounds of combined nitrogen per acre was precipitated annually in the rains of Western Europe. Later observations at Rothamsted show that this estimate is probably too high, and Drs. Lawes and Gilbert, after a full discussion of their records for twenty-seven years, fix the probable amount at four to five pounds per acre.

As this is only one fourth of the average annual yield of nitrogen per acre of the unmanured wheat over a period of thirty-two years, and but little more than one fourth of the average annual yield obtained with barley over a period of twenty-four years, to say nothing of the much larger yield of nitrogen in leguminous crops, it must be admitted that it is an entirely inadequate source of supply of nitrogen for vegetation. The nitrogen condensed by the soil from dew and atmospheric vapor has not been definitely determined, and is not, therefore, included in this estimate; but it is probable that it is less than that brought to the soil by the rain. On the other hand, it has been shown by numerous experiments, including those at Rothamsted, that free nitrogen is evolved in the decomposition of organic matter, and that the soil, under certain conditions, may suffer a loss of nitrogen in this form.

Of the rain falling upon the drain-gauges, the unmanured soil of which, it will he remembered, is in a natural state of consolidation and kept free from vegetation, for a period of ten years, about forty-four per cent has appeared as drainage-water, and about fifty-six per cent has been evaporated. Approximately, two thirds of the evaporation takes place during the summer months, and one third during the winter months.

The annual loss of nitrogen in the drainage-water has been, upon the average, at the rate of 43·7 pounds per acre. This represents, approximately, the loss involved under the conditions of a bare summer fallow.

When the roots of growing plants are distributed through the soil, they take up the nitrogen as it becomes soluble in the process of nitrification, and the loss by drainage is to that extent diminished.

The drainage from the unmanured wheat-plots contained nitrogen at the rate of only from 12·56 to 18·62 pounds per acre each year, and during two seasons of excessive drainage, when every running from the drains was analyzed, "it was estimated that from fifteen to seventeen pounds of nitrogen were lost per acre, per annum, by drainage from plots which had received no nitrogenous manure for many years," and the average for thirty years was from ten to twelve pounds per acre. During its period of active growth, the crop appropriated the nitrogen of the soil, so that there was little or none lost by drainage, and nearly the entire loss took place after harvest, and during the winter and spring months.

The nitrogen lost by drainage on land receiving no nitrogenous manure is therefore considerably more than can reasonably be estimated in the supply from atmospheric sources.

When nitrogenous manures were applied, the loss of nitrogen by drainage was materially increased, and on the average for more than thirty years, and under the most favorable conditions of growth, less than one third of the nitrogen applied as manure was recovered in the increase of the crops, and much less than this when there was a deficiency of potash or phosphoric acid in the soil.

In connection with these facts, relating to the amounts of nitrogen removed in the crops and lost by drainage, and the inadequate supplies of available atmospheric nitrogen for the purposes of plant-growth, it becomes a matter of particular interest to trace the influence of this system of continuous cropping, without nitrogenous manures, upon the nitrogen contained in the soil itself.

The nitrogen in the soil of the unmanured wheat-plots has gradually diminished, and Dr. Gilbert says, "So far as we are able to form a judgment on the point, the diminution is approximately equal to the nitrogen taken out in the crops, and the amount estimated to be received in the annual rain-fall is approximately balanced by the amount lost by the land, as nitrates in the drainage-water."

Where the great decrease in the yield of nitrogen was observed, in the case of the root-crops which were grown for thirty-one years in succession, the soil at the end of twenty-seven years was found to contain a smaller percentage of nitrogen than any other arable land of the farm.

In the experiments on the mixed herbage of permanent grass-land, "the soil of the plot which, under the influence of a mixed mineral manure, including potash, had yielded such a large amount of leguminous herbage, and such a large amount of nitrogen, showed, after twenty years, a considerably lower percentage of nitrogen than that of any other plot in the series."

The soil of the garden-plot, which gave so large a yield of clover over a period of twenty-seven years, was analyzed at the end of twenty-six years, and Dr. Gilbert remarks, in regard to the loss of the nitrogen of the soil, that "the diminution, to the depth of nine inches only, represents, approximately, three fourths as much as the amount estimated to be taken in the clover in the intervening period; and the indication is, that there has been a considerable reduction in the lower depths also."

When nitrogenous manures were applied in the form of ammonia salts or nitrate of soda there was little or no decrease in the nitrogen of the soil, and in some cases there was an actual gain; but the loss from drainage was much greater, and it increased with each increment of the manures applied under the same conditions. On plots receiving 43, 86, and 129 pounds, respectively, of nitrogen in the form of ammonia salts, mostly applied in the autumn, the estimated loss of nitrogen by drainage was 19, 31, and 42·4 pounds; "and with 86 pounds of nitrogen applied without, or with different mineral manures, the estimated loss ranged from 31 pounds with the most liberal manure to 43-2 pounds with the ammonium salts continuously used alone."

The nitrogen of barn-yard manure, which from its comparative insolubility is more slowly available for purposes of plant-growth, was not lost by drainage to the same extent as the chemical manures, and there were decided indications of considerable accumulations of it in the soil. The nitrogen applied in manures is not all accounted for in the amounts removed in the crop, lost by drainage, and stored up in the soil; and it therefore seems probable, in the absence of any other known disposition of it, that the estimated losses by drainage, based on the materials discharged by the tile-drains, are altogether too low. As the drainage-waters, with the substances they hold in solution, pass from the upper to the lower strata of the soil, we can not avoid the conclusion that a large proportion must pass below the level of the drains without entering them. The drains of the experimental wheat fields are nearly twenty-five feet apart, and the underlying chalk at the depth of from ten to fourteen feet furnishes good drainage for the spaces between the drains.

The nitrogen applied in the manures that is not taken up by the crop or stored up in the soil, or lost in the waters discharged by the tile-drains, may therefore be fully accounted for in the amount that must be carried, under the conditions of the experiments, by the drainage-waters to the lower strata of the subsoil, without entering the drains. The estimated losses of nitrogen by drainage, based on the amounts detected in the waters discharged by the drains, may therefore with good reason be increased by the amount not accounted for in the crop and in the accumulations of the soil.

Practically, then, in the light of the Rothamsted experiments, we may look upon the soil as the great source of the nitrogen of plants, as the atmosphere can furnish but a small proportion of the needed supply, and this is more than counterbalanced by the losses from drainage.

In connection with this imperfect outline of some of the leading lines of investigation that have been so successfully prosecuted at Rothamsted, it would be interesting to examine the data that indicate the relations of nitrogen to other elements of plant-growth, as supplied in manures and assimilated by crops when cultivated in succession or in rotation with other species; but these, with other cognate topics, must be omitted, as we can not at this time undertake anything like an exhaustive discussion of the results of these valuable experiments.

The great importance of physiological researches and the comparatively subordinate influence of purely chemical methods in solving the great problems of agricultural science, have been so fully illustrated in the experiments at Rothamsted that we must accept them as the basis of a new departure in the development of a consistent science of rural economy. In the light of these experiments the generally accepted theories of soil-exhaustion must be reconstructed, and the action and relative value of manures must be investigated from a new stand-point.

The exhaustion of a soil can no longer be estimated by the constituents removed in the crop, Wheat and oats, with other cereals, are generally considered as exhausting crops, and a summer fallow is looked upon as a means of increasing or restoring the fertility of the soil; but the grain-crops when grown by themselves, and the summer fallow itself, are alike the occasion of a loss of fertilizing materials, and in precisely the same way. In both cases there is a long interval in which there are no living roots of plants in the soil to take up the nutritive materials as they are transformed into the soluble form, and they are lost by percolation to the lower strata of the subsoil out of the reach of vegetation.

Many of what are called restorative plants feed in the deeper layers of the soil, and they may, by their scattered foliage and thick roots, add to the stores of plant-food in the surface-soil, which may be used by plants of a different habit that are not deep feeders.

The physiological peculiarities of the different botanical groups of plants must, however, be better understood before we can fully explain the influence of one crop upon another that succeeds it. That the special formula manures, so widely advertised in this country, and which are claimed to provide, in due proportion, the constituents required by a particular crop, are based on false assumptions, is abundantly shown in the Rothamsted field-experiments; but we can not now discuss the fallacy in detail.

The experiments with animals at Rothamsted must form the subject of a separate article.