fraction in point of bulk, what could be more natural than the supposition that their nitrogen also was derived from the same source? For by this time it was well known that nitrogen gas, in a free or uncombined state, constitutes four-fifths by volume of the air. Observation of the behaviour of growing plants, however, soon made it doubtful if this supposition could be entertained; and in the early years of the present century Theodore de Saussure, a distinguished Swiss savant, as the result of numerous experiments, concluded, not without due caution, that plants do not fix and assimilate the free nitrogen of the air, or get any part of their supply from that source.

The question thus raised continued to be debated for many years; but no certain results were gained until the investigation of the points in dispute were undertaken by the justly-celebrated French savant Boussingault. In a series of careful and masterly researches carried out between the years 1845 and 1865 Boussingault displayed remarkable ingenuity in devising methods of interrogating nature by simple and decisive experiments, and laid the foundation of the experimental methods of to-day, which will be referred to in some detail in the course of this address. His researches showed conclusively that, under the conditions of his experiments, atmospheric nitrogen is not employed in the process of the assimilation of plants. His experimental plants in every case grew vigorously, and produced a normal amount of proteid compounds when he presented to their roots nitrogenous compounds of various kinds, in addition to the other non-nitrogenous substances necessary for their nutrition. They grew very badly, on the other hand, and their proteid substances did not increase in amount, when this supply of nitrogenous compounds was withheld, although the free nitrogen of the air was at their disposal if they could use it. Thanks to the now much further developed art of nourishing plants artificially, and especially by means of water cultures, we are today in a position, by means of simple experiments, to afford ocular demonstration of the important results obtained by Boussingault. Whenever a sufficient quantity of saltpetre is added to the nutritive solution the experimental plants grow vigorously, produce numerous ripe seeds capable of germination, and give on analysis a corresponding increase in the nitrogenous substances they contain. On the other hand, if the nitrate is withheld from the nutritive mixture, the experimental plant grows for a time, it is true, making use of the proteid substances already contained in the seed for the formation of the protoplasm of its cell and organs; and this stunted growth may even continue for some time, since the protoplasm of the first leaves is again dissolved and absorbed

from them to be employed in the formation of new leaves; but eventually analysis of the mature plant shows no increase of proteid or nitrogenous substance, and but little increase in total weight, when compared with the seed from which it has grown.

Soon after Boussingault's results were communicated to the scientific world, Lawes, Gilbert, and Pugh, the great English authorities on agricultural research, undertook at their well-known experimental farm at Rothamstead a fresh investigation of this question among others. Their results entirely confirmed Boussingault's conclusion that plants do not directly fix and assimilate the free nitrogen of the air, but derive the whole of their supplies of nitrogen from nitrates and other nitrogenous compounds in the soil, presented to their roots in a state of solution. As a result of later researches, Lawes, Gilbert, and Pugh drew attention to a very remarkable fact about the leguminous plants which they had investigated. In the case of most crops the nitrogen which they were found on analysis to contain could be entirely accounted for by the combined nitrogen originally contained in the soil, as shown by analysis of samples, or added to it in the form of manure, or washed down from the air by rain. But in particular leguminous crops, such as peas, beans, vetches, lupins, and the like, the crop contained more nitrogen than could be accounted for from the known sources of supply. Here was a thoroughly established fact that could not be explained by any of the recognised canons of the vegetable physiology of the time. The excess of nitrogen fixed by these leguminous plants was for the time being a mystery that no one could solve. The accuracy of the observations and analyses were in due season confirmed by other inquirers, but no clue to the explanation of the unaccountable excess of nitrogen accumulated in these plants could be so much as suggested.

The fact that leguminous plants contain more combined nitrogen than could have been drawn from the known sources of supply in the soil and air was at once seen to be in harmony with the experience of agriculture in ancient as well as in modern times. Even the Romans had believed that leguminous crops grown on impoverished land greatly improved its fertility, and it was a commonplace of agricultural practice that cereals and grass crops grew much better after a crop of pulse than after any other.

Inquiry into the source of the excess of nitrogen fixed by leguminous plants was at once taken up, and all sorts of opinions were formed and given to the world. But there was no agreement or approach to agreement among inquirers. More than twenty years had to pass by before any real light

was thrown on the question. Some two hundred years ago Malpighi, an Italian savant, drew attention to certain nodules or tubercles that grew on the roots of leguminous plants; and at a later time Linnæus described these nodules on the roots of Lathyrus (the sweet-pea). From this time the nodules continued to attract the occasional notice of botanists, but nothing of importance was learned about them for many years. In 1856 Lachmann investigated their minute structure, and much discussion ensued, but with no real understanding of the facts. In 1866 Woronin made the important discovery that the root - tubercles contained great numbers of minute organisms, since named “bacteroids”—i.e., bacterium - like bodies. Things were thus getting into train for a true comprehension of the rôle played by the root-tubercles in the nutrition of leguminous plants; but ten more years had to elapse before the connection between the activity of the root-tubercles with the associated bacteroids on the one hand, and the abnormal fixation of nitrogen by leguminous plants on the other, was conceived and investigated.

In November, 1886, two German investigators, Professor Hellriegel and Dr. Wilfarth, who had been engaged for a considerable time on a wide and fruitful investigation of the nitrogenous nutrition of gramineous, leguminous, and other agricultural plants, published their first important results. To these I shall refer at some length, not only on account of their intrinsic importance, but also because they afford excellent examples of scientific method, of skilfully-directed interrogation of nature, and of the cautious and logical interpretation of experimental results.

To ascertain the sources of the nitrogenous food of plants these inquirers sowed and raised a variety of gramineous plants (including cereals and grasses) in a soil deprived of nitrogen and protected from rain, but to which all other mineral substances necessary for the healthy nutrition of plants had been added in proper quantities. The seedlings developed normally until the third leaf appeared, when the reserves of food contained in the seeds were exhausted. At this point growth ceased suddenly. The plants did not die; they lived almost as long as normal plants, but their vegetation was dwarfed. The stunted plants developed stunted and miserable organs, even barren ears, and struggled through the season. Their total dry weight was found to have increased very little beyond that of the seed, and the increase was of non-nitrogenous substances only. When nitrates were added to the soils as soon as the arrest of growth set in, normal growth was soon resumed, and if sufficient nitrates were added it continued without cheek to full maturity. If, on the other hand, the nitrates were insufficient, a gradual passage to the

starved condition supervened. In these experiments it was found that, within the limits of the optimum supply, there was a direct proportion between the amount of nitrates added and the yield of grain.

If, instead of nitrates, ammonia salts or other nitrogenous compounds were added the resumption of growth was delayed; and only after a pause of considerable length did these salts become available as nutriment. In this case the increase in the yield of grain was not proportional to the amount of ammonia added.

These results threw considerable light on a question that had been much debated. Justus von Liebig and others taught that ammonia and salts of ammonia were the most desirable form in which plants could receive their supplies of nitrogen. Hellriegel's results make this view extremely doubtful. The evidence irresistibly suggests that the salts of ammonia had to undergo nitrification—i.e., had to be oxidized into nitrates—before the roots of the grasses could avail themselves of their stores of combined nitrogen. As this has little direct bearing on my immediate subject, I need only say that Hellriegel considers that nitric acid and its salts are the only directly available sources of nitrogen for gramineous plants. The main result of this series of experiments was the proof that the Gramineæ are entirely dependent on the combined nitrogen in the soil for their supplies of nitrogen. They cannot draw upon the stores of free nitrogen in the air, except in so far as rain or dew carry down nitrogenous compounds into the soil.

Hellriegel then proceeds to show that similar experiments with leguminous plants yield totally different results. In a soil devoid of nitrogen and protected from rain, exactly like that in which the gramineous plants had uniformly starved, peas were allowed to germinate and grow; and in nearly every case they flourished and yielded a large increase. Thus peas grown in small culture-vessels little larger than a thimble yielded above ground between two and three times the amount of dry substance that the seeds contained, and this, of course, without any addition of nitrates to the soil in which they grew. The plants grew normally throughout, and even vigorously. Hellriegel notes that such a yield of dry substance from the same kind and quantity of soil could not have been obtained with a gramineous plant such as barley, even with the addition of a sufficiency of nitrates. The result which Lawes, Gilbert, and Pugh had reached by analysis of leguminous plants—namely, that these plants absorb and utilise much more nitrogen than the soil can supply—was placed once for all on a firm basis of trustworthy experiment.

The source of the nitrogen gained by these pea-plants had now to be considered. The soil contained none, for it consisted

of pure quartz sand repeatedly washed, and was enriched by a nutritive mixture that contained no nitrogenous compound. The water used was specially prepared, and free from ammonia and nitric acid. Moreover, the starvation of the grasses grown under the same conditions proved the absence of nitrogen compounds in the soil. It was thus quite clear that the soil was not the source of supply of the accumulated nitrogen of the leguminous plants.

The only other possible source of supply is the nitrogen in the atmosphere. Now, nitrogen exists in the air in two distinct forms—(1) As free nitrogen, which forms by far the largest part of the air; and (2) as combined nitrogen in the forms of nitric acid and salts of ammonia (carbonate and nitrate). Hence we must assume either that leguminous plants have an extraordinary capacity for collecting and absorbing by their leaves the very scanty nitrogen compounds in the air, or that they somehow make use of the abundant stores of free nitrogen which it contains.

The following considerations led Hellriegel to favour the latter of these hypotheses: When peas are cultivated in a soil free from nitrogen, and under the conditions described above, two remarkably sharp periods of growth are to be noticed. So long as the reserves of food stored in the seeds last, the seedlings grow naturally, luxuriantly, and with the normal colour. As soon as the reserves are exhausted, a somewhat sudden change occurs—growth stops, the leaves turn pale, and the plant evidently begins to starve. Sooner or later, however, the pale or yellow leaves again turn green, and a second period of active growth sets in, after which the plants grow normally to the end. When the arrest of growth sets in each plant has about six leaves; and, if these can fix and assimilate the nitrogen compounds in the air, we are quite unable to understand why the starvation phase should appear at all. If the leaves cannot perform this function when in vigorous growth and of normal green colour, how can we suppose that they begin to do so when in a sickly and discoloured condition? Hellriegel was not, however, satisfied with these general considerations, but proceeded to carry out a series of new experiments that would enable him to decide with certainty between the rival hypotheses.

Four vessels were filled with soil devoid of nitrogen, and peas were put in and allowed to germinate. The vessels were then placed under four bell-jars completely closed, except where they were joined by connecting tubes. The four vessels were arranged in series, so that a constant stream of air was drawn through from No. 1 to No. 4. Absorption vessels were placed between each pair of bell-jars, and matters so adjusted that the ordinary air passed into the No. 1 jar unaltered; but

before entering Nos. 2, 3, and 4 the stream of air was deprived of any ammonia or nitric acid it contained. The plants all grew normally, and passed successfully through the usual starvation stage. When the experiment was stopped they had grown to a height of 120cm., and had entered on the flowering and fruiting stage. Subsequent analysis showed that the plant grown in the ordinary air did not yield as much dry substance in the straw and roots together as did the plants grown in the purified air. Several repetitions of this experiment made it quite clear that the small traces of combined nitrogen present in the air did not supply the experimental plants with the nitrogen which they somehow obtained. The only hypothesis that could now be entertained was that somewhere and somehow the leguminous plants do make the free nitrogen of the air available for their nutrition.

In the course of Hellriegel's experiments he noticed that, while some plants grew in a soil devoid of nitrogen and under the conditions already described, others grew badly, and some never recovered from the starvation stage. Repeated and careful examination of the experimental plants brought to light a remarkable fact. The plants that remained in the starvation phase had either no tubercles on their roots or very few and insignificant ones, whereas the plants that throve and grew luxuriantly had many well-developed tubercles on their roots. The more plants Hellriegel investigated the more was he convinced that the development of the root-tubercles stood in the closest and most direct relation to the growth and assimilation of the whole plant. The origin of the root-tubercles had now to be investigated. It was likely that they were caused by the invasion of the root-tissues at certain spots by micro-organisms existing in the soil. To test this supposition Hellriegel carried out numerous experiments. A brief account of two of these will show his methods and his results.

On a certain day (25th May) were taken forty vessels filled with soil devoid of nitrogen, and two pea-seeds were planted in each. Ten of the vessels were then watered with washings from the fertile soil of the culture field in which micro - organisms existed in abundance. Thereafter these vessels, and from the first all the others, were watered with pure distilled water. In about three weeks the aspect of the plants was changing, and all turned pale as the seed reserves were exhausted. So far there was no difference to be seen between the forty cultures. But a difference soon set in, and in another week it was most decided. In the ten vessels supplied with bacteroids all the plants had regained their fresh-green colour, and commenced to grow vigorously. Of

the other thirty vessels in which the introduction of micro organisms was left to chance, only two at this stage presented a similar appearance, all the rest were still starving and in part yellow. A fortnight later the plants in the ten vessels were developing their tenth leaf and growing luxuriantly. Of the sixty plants in the other culture-vessels which had not been supplied with bacteroids ten were now nearly as flourishing as the above, while five were nearly dead. Among the remaining forty-five plants all stages between these extremes were to be found. An examination of the roots of the best grown and the worst grown plants fully confirmed the relation between the growth of the sub-aërial parts and the development of root-tubercles. This experiment did not, however, furnish a decisive answer to the question of the source of infection of the roots with the organisms that produce the tubercles. Abundant tubercles had appeared on plants grown in culture-vessels to which no soil-washings had been deliberately added, the micro-organisms having been introduced in some unknown and accidental way.

The question was attacked anew, and this time with more decisive results. Two cultivations were made in soil free from nitrogen, to which the usual non-nitrogenous nutritive mixture, and, in addition, a small amount of the above-mentioned soil-washings, were added. The culture vessels thus charged were then sterilised by heating, after which the seeds were sown and the surface of the soil covered over with sterilised wadding. All went well until the development of the sixth leaf and the setting in of the usual starvation stage. After this the plants made no further progress, and before long all of them died. No tubercles were formed on the roots under these circumstances, and numerous repetitions of the experiment always gave the same results. These experiments show conclusively that the formation of the root - tubercles is dependent on the presence of living micro-organisms in the soil, which infect the roots, and that in the absence of root - tubercles leguminous plants are exactly in the same case as other plants—they pass into a condition of permanent starvation as soon as the food reserves of the seed are exhausted, just as gramineous plants do when grown in a medium devoid of available nitrogenous compounds.

Hellriegel sums up the outcome of his researches in these terms: “Leguminous plants, in contrast to the gramineous ones, are not dependent on the soil for their nitrogenous nutrition.” As to the source of their supplies of nitrogen, he adds, “Not one of my experiments supports the idea that it is to be found in the minute quantities of combined nitrogen

that exist in the atmosphere. Thus, probably, the only remaining assumption we can make is that leguminous plants have the power of making use of the free nitrogen of the air. To their nutrition, and especially to their assimilation of nitrogen, the so-called tubercles and the organisms that dwell in them stand in the closest active connection.”

Many other observers have taken part in the elucidation of this subject, and Hellriegel's results have been abundantly verified both in England and on the Continent. The most rigorous confirmation has been supplied only two or three years ago by Laurent and Schlœsing. These inquirers grew leguminous plants from seed in a sterilised glass apparatus of great beauty and ingenuity, which provided for the supply of pure water and carbon-dioxide to the foliage exposed to the light, as well as for the circulation of air known to consist of oxygen and nitrogen in the free state only, all nitrogenous compounds being taken out. Samples of the air could at any time be taken out and analysed, and at the close of the experiment the whole of the air affected by the growing plants could be collected and examined. In this way Laurent and Schlœsing proved not only that the nitrogen increased in the growing plants, but that a corresponding amount of nitrogen had been removed from the air that surrounded them. This result was got only when the experimental plants were infected with bacteroids and well provided with root-nodules. If the bacteroid organisms were not added no tubercles appeared, and the plants starved.

This research proves in the most conclusive way that the nitrogen fixed by infected leguminous plants is taken from the air, and must be taken as free nitrogen.

So far as leguminous plants are concerned, the modern results are of course in direct conflict with those which Boussingault was thought to have established about the middle of the century. The discrepancy is due to the fact that bacteroid infection was excluded in Boussingault's experiments, one main object of which was to show that humus and its immediate derivatives were not necessary, and, indeed, not directly available, for the nitrogenous nutrition of plants.

The fact of the fixation of the free nitrogen of the air by infected leguminous plants being thus finally and firmly established, the exact mode or process of the fixation remained to be explained.

At first various opinions on this subject were held by competent inquirers. Hellriegel's and Laurent and Schlœsing's investigations have made some of these untenable. One view, which is still maintained by Frank of Berlin, was that the Leguminosæ took up the free nitrogen of the air by their leaves. The experiments of Hellriegel, already described, and others

conducted with even greater rigour by Laurent and Schlœsing, show that this view is untenable. Frank is its only prominent advocate at the present day, and the experiments on which he relies are generally held to be inconclusive. Another view was that the Leguminosæ, with their abundant and deeply penetrating root-system, could take up nitrogenous compounds from the lower layers of the soil which other plants could not reach. On this view we cannot understand why the presence of numerous well-developed tubercles in the roots should be a necessary condition of the abundant assimilation of nitrogen. It is, moreover, directly refuted by the abundant fixation of nitrogen by plants grown in shallow culture-vessels and in soil devoid of nitrogenous compounds of any kind.

Two other views have been entertained, and both are more or less consistent with the results of experimental inquiry. But before dealing with these let me describe somewhat more fully the nature of the nodules and their contained organisms, and of the relations of the latter to the living plant which they affect.

In 1887 Professor Marshall Ward showed that a perfectly definite organism from the soil penetrates the root-hairs, and invades the cortical tissues of the roots. The organism lives in a sort of helpful association (or in symbiosis as the specialists have it) within the tissues of the root-tubercles, and not merely as a hurtful parasite. Instead of injuring the leguminous plant, the nodules and their living invaders in the long run add to its health and vigour. What follows the infection is briefly as follows: The invading organism causes the cells deep down in the cortical tissues (the tissues that correspond to the bark in stems) to divide and form a delicate tissue of well-nourished cells. As these enlarged cells multiply the organisms keep pace with them, and send brandies into each new cell. Eventually myriads of minute bacterium-like bodies fill the cells. The cells are not, however, killed by the bacteroids; on the contrary, they show all the signs of intense physiological activity, accompanied by the destruction of copious supplies of carbohydrates brought to them from the leguminous plant. After a time this metabolic activity abates, and the myriads of bacteroids begin to get disorganized. Then the now victorious leguminous plant absorbs the disintegrated contents of the interior of the nodules, taking up everything that is capable of solution and absorption, and leaving the nodule a nearly empty, limp, collapsed shell, in which such bacteroids as have passed into a sort of resting-stage alone remain alive, to be scattered in the soil as the débris of the exhausted nodule rots away.

It looks as if the invading organism acted at first as a parasite, but though it lives at the expense of the cells which

it has enslaved and forced to intense metabolic activity it does not destroy them. The next stage is one in which the struggle for mastery between the pseudo-parasite and its host is most intense. For a time the bacteroids take everything the cell can give them; they multiply and fill the cells, and then gradually pass into a passive condition. The leguminous plant now gains the mastery, and the cells of the nodules absorb and pass on into the plant almost the entire store of nourishing substances which the bacteroids have accumulated.

Now, there can be no reasonable doubt that the fixation of the free nitrogen occurs in the underground part of the plant. The known anatomical structure of the nodules, especially the complex and very regular system of vascular bundles that connect them with the mother-roots, and the abundant supply of carbo-hydrates passed into the nodules, make it extremely probable that the seat of the action lies in the nodules themselves. We may therefore suppose either that the inert nitrogen molecules are forced into the organic synthesis in the cells of the bacteroids, or that the metabolic activity of the cells composing the nodules, and it may be adjacent parts of the roots also, is so exalted by the stimulating action of the bacteroids as to be able to build up compound proteids by the synthesis of gaseous nitrogen with the already elaborated carbo-hydrates. It is difficult to arrange experiments that will settle decisively which of these views is the correct one. Attempts to demonstrate that the bacteroid cultivated outside the plant can assimilate free nitrogen have hitherto failed. But we must not lay too much stress on such negative results, because the cultivation of an organism so highly adapted to its symbiotic life as this evidently is may well fail outside the cells of the host. On the other hand, it is improbable that the cells of the bacteroids would assimilate, and by their disintegration and absorption pass on to the infected plant, such quantities of nitrogen as are known to be fixed from the air by the Leguminosæ. On the whole, we may conclude that the view adopted by Hellriegel and many other inquirers, that the cells of the nodules are the actual seat of the fixation of the free nitrogen of the air, is the true one.

How the fixation is brought about we cannot in the present state of knowledge explain, though theoretical considerations carry us some way towards an explanation of the phenomenon. In the ordinary course of plant metabolism nitrogenous bodies —e.g., amides—preceded, it may be, by simpler compounds, are built up by the living cell protoplasm from carbo-hydrates and nitrates, the energy required to effect this synthesis being chiefly if not wholly derived from the indirect oxidation of

part of the carbo-hydrates supplied to the cells. It is this indirect oxidation that produces the carbon-dioxide that is continuously given off by living plants. This constructive metabolic work of the protoplasm is an act that we cannot explain in detail; when we can, we shall indeed have solved the mystery of life. We can only dimly perceive that the synthesis depends on some peculiar power possessed by the protoplasm of presenting the atoms and molecules to each other. We may therefore suppose that the cell machinery of the roots or the nodules is so exalted in activity by the presence or the products of the bacteroids as to be able to force the notoriously inert nitrogen into combination with other molecules, so as to produce nitrates, or amides, or other similar bodies. The energy required for this transformation is no doubt supplied by the oxidation of part of the abundant stream of starch and other carbo-hydrates, as well as salts, that is conveyed, to the nodules from the roots, the surplus of these substances being alone available for entering into combination with the nitrogen directly assimilated from the air. It has been suggested that, as the sap expressed from the active tissues of the nodules is alkaline, the cell protoplasm in the presence of an alkali and free nitrogen may be able to build up ammonium-nitrite or some such body. On this view, all that seems required for the forcing of nitrogen into the organic synthesis is a sufficient supply of carbo-hydrates. But this does not explain the whole difficulty, else why cannot the well-nourished cells of any plant do the trick?

Though we are unable, and are likely to remain unable, to explain in detail the process of nitrogen fixation, the results of the researches I have detailed do not on that account lose anything of their importance. They constitute a clear and a most valuable addition to our assured knowledge of the nutrition of a large class of agricultural plants. To the farmer, the forester, and the gardener, the new knowledge is of the very highest significance. The farmer now knows that he has at his disposal a simple, certain, and most economical means of enriching his land in nitrogenous compounds by the spontaneous bounty of nature. To avail himself of this he needs only to grow a judicious mixture of leguminous and gramineous plants in the same crop, or to use a suitable rotation of leguminous and other crops in successive years. Clover, lucerne, peas, beans, and sainfoin are good examples of useful leguminous crops, and they all possess a very extensive and well developed system of roots. When the crops are harvested or eaten off by stock, the abundant root-system decays in the soil, and leaves it much richer in nitrogenous compounds than it was before. The soil is thus, by the mere act of nature, endowed with large supplies of a prime necessary of

plant life; one, too, which it costs the farmer much more to supply in the form of manure than any other important ingredient that cultivated soils are apt to be deficient in.

The lesson which the farmer should learn from the new knowledge is obvious enough. Every few years he must break up his land and grow a crop of one or other of the common useful leguminous plants, such as peas, beans, lucerne, or clover. His land will thus be sufficiently supplied with nitrogenous compounds, and he will have to add to it only such other less expensive and scarce elements of plant food as the soil may require. Clovers, too, will be grown along with grass-crops, but this method is not so likely to yield satisfactory results, as clover-plants die out in a few years, and are apt to induce the condition known as clover sickness. This undesirable condition of the soil can be partly avoided by providing it with sufficient stores of potash and lime, and more effectually, and also more economically, by following a sub-rotation of clover-plants in the general rotation of crops—red-clover, white-clover, and alsike being sown in alternation, when the leguminous crop of the general rotation becomes due. It is, however, clear that lands laid down in permanent pasture cannot be greatly enriched by gains of nitrogenous compounds accumulated by the roots of leguminous plants, since hardly any useful plants of this kind can be grown in the same soil for a long series of seasons. It is therefore only in lands that are cultivated with some regularity that the farmer can draw to the best advantage on the bounty of nature for the enrichment of his land.

I need not dwell upon the application of the new knowledge to the economic management of forests, orchards, and gardens. It must suffice to point out that it has obvious and important bearings on each of these pursuits.

There are many other points connected with the relation of the nitrogen of the air to the soil to which it would be interesting to refer, but the time at my disposal will not allow me to touch on more than one of them.

Observers, even thirty years ago, had no doubt that the soil in which plants grow was a comparatively simple medium, consisting of a mixture of sand, clay, lime, humus, and traces of other mineral substances. The modern botanist has learned that the soil is a vastly more complex substance than that. He has learned to recognise that every fertile soil abounds in microscopic organisms, often to a degree that is truly startling. The earlier estimates may be passed over as less trustworthy than more recent ones. The latter, however, are sufficiently startling, for they show that in a cubic centimetre of rich sandy soil from ten to forty millions of germs may be found. In fact, we may look for from one to ten

millions of germs in a thimbleful of such soil. These numbers are, it must be admitted, sufficiently bewildering. And what are these microscopic organisms doing? A few of them are undoubtedly injurious parasites; others are the fungi and bacteria that live as saprophytes on decaying vegetable and animal remains, and are the active agents in the fermentations and putrefactions which resolve organic substances into the forms of carbon-dioxide, water, and ammonia. Others have the power of oxidizing ammonia first to nitrous and then to nitric acid, and are doubtless the chief agents in nitrification, while still others undo the work of the last by degrading the highly oxidized salts of nitrogen, deoxidizing nitrates to nitrites, and the latter to ammonia and even free nitrogen.

Besides these, evidence is accumulating that points to the existence of forms which, provided there are plenty of carbo-hydrates available, can actually fix free gaseous nitrogen in their living machinery, and compel it to enter into the organic synthesis. Winogradsky, who has isolated and studied organisms of this kind at the Pasteur Institute Laboratory, suggests that in its life processes nascent hydrogen is set free, which combines with the free nitrogen of the air to build up ammonia.

These results are still so recent that I have not heard of their verification by other observers, but should they be confirmed they show us that the very process of nitrogen fixation, which occurs normally in the root-nodules of leguminous plants, takes place in microscopic organisms that lie detached in the soil, and indicate possibilities of nitrogen storage in the soil that have hitherto not even been suspected. Winogradsky's observations, too, lend additional probability to the hitherto doubtful view that it is the bacteroid cells that constitute the actual seat of nitrogen fixation in the root-nodules of the Leguminosæ. However this may be, it is obvious that the whole question of the supply of nitrogen to vegetation is taking new turns. We feel that we are steering on the brink of new and fruitful discoveries that may revolutionise the whole practice of agriculture, and even restore the prosperity of the agricultural interest in spite of the world-wide decline in the values of nearly all cereal and other products which the soil yields. Let us, in conclusion, note that the advances in knowledge which I have endeavoured to explain have been won by eminent investigators, hailing from every one of the great civilised nations of Europe. Original memoirs relating to the investigations I have been discussing have appeared in nearly every European language, and only a polyglot could fully follow the more important researches in this one subject. In these days when the jealous nations are snarling at each other's heels, it is no small consolation to