of about 14,500, which is twice as great as the simplest one that can be calculated from its analysis, if there are 2 atoms of sulphur in its molecule. Edestin, therefore, forms salts corresponding to a mono- and bichloride. 7. The crystals of edestin, as well as those of its different salts, are, so far as has been determined, isomorphous, the mass influence of the protein molecule being so great as to prevent the small amount of combined acid from effecting a change in crystalIn this respect, edestin behaves like hemoglobin, the compounds of which with oxygen and with carbonic acid are also isomorphous. 8. The free base edestin, when suspended in pure water, is dissolved by nearly the calculated quantity of hydrochloric acid required for a complete reaction between 1 molecule of edestin and 2 of hydrochloric acid. On adding the acid in successive small quantities, no solution takes place, until one-half the required amount has been added. On adding the second half of the acid, solution takes place at a rate proportional to the amount of acid added. The acidity of the solution obtained with the second half of the acid increases at twice the rate at which the acid is added, in accordance with the conversion of an insoluble monochloride produced by the first half of the acid into a soluble bichloride formed by the second half. 9. Somewhat more than the calculated quantity of hydrochloric acid is required to dissolve a given quantity of edestin, because a more basic hydrolytic derivative of edestin, sparingly soluble in water, is formed by the hydrogen ions set free by hydrolytic dissociation of the chloride. 10. Since solutions of edestin bichloride do not appear to be precipitated by hydrolytic dissociation, it is probable that edestin hydroxide may be formed and remain in solution under the conditions of the experiments tried. 11. Edestin sulphates are less soluble than the corresponding chlorides and, consequently, preparations obtained from solutions containing ammonium sulphate are not soluble in water. Ten times more sulphuric acid is required to dissolve a given quantity of edestin than of hydrochloric acid. Definite compounds with sulphuric acid have not yet been obtained. 12. Hydrochloric acid dissolves more nearly the calculated quantity of edestin than does acetic acid, since the latter, being less ionized than the former, produces, in a given time, only one-third as much of the more basic hydrolytic derivative as does the former. 13. Phosphoric acid reacts with edestin as a monobasic acid, in accordance with its dissociation into the ions H and H,PO,. But little more than the calculated quantity of this acid is required to dissolve a given amount of edestin. 14. Edestin forms a salt with nitric acid, which corresponds to the bichloride. At 30° edestin binitrate is much more soluble in pure water than at 20°, so that a clear solution containing 5 per cent. of this salt yields a considerable precipitate on cooling. 15. Edestin reacts with potassium or sodium hydroxide in a proportion equivalent to that with which it forms the monochloride with hydrochloric acid. A given quantity of edestin is dissolved by an amount of centinormal potassium or sodium hydroxide solution, which corresponds closely to a proportion of 1 molecule of the base to I of protein. Solutions of potassium and sodium edestin, probably in consequence of hydrolysis, become turbid after standing some time and gradually deposit some of the dissolved protein. 16. About thirteen times as much ammonium hydroxide is required to dissolve a given quantity of edestin as of sodium or potassium hydroxide. 17. About three times as much sodium in the form of carbonate is required to dissolve a given quantity of edestin as of sodium in the form of hydroxide. 18. Edestin conforms strictly with the definition of a globulin, being insoluble in pure water, but readily soluble in neutral solutions of sodium chloride of sufficient strength. 19. Edestin monochloride is, likewise, insoluble in water, but readily soluble in saline solutions. Edestin bichloride and potassium or sodium edestin are soluble in pure water, but insoluble in the presence of a small proportion of a neutral salt. In the presence of a larger proportion of the neutral salt they are soluble, and in such solutions they show the properties of globulin. 20. The fact that edestin, as well as its salts with strong acids, is soluble in perfectly neutral solutions of sodium chloride shows that the solubility of a globulin does not depend on the presence of alkali, as Starke has recently asserted. RESEARCH LABORATORY OF THE CONNECTICUT [CONTRIBUTION FROM THE Bureau of Chemistry, U. S. Department of AGRICULTURE, No. 42.] A STUDY OF THE AVAILABLE MINERAL PLANT FOOD IN SOILS. BY C. C. MOORE. Received November 28, 1901. Na study of soils, it is most essential, to have as wide a a variety of types as possible, as at best, the study must be a comparative one. An hypothesis is taken and applied to the maximum number of types, concordant results classified, and the cause of the variations studied. The hypothesis is then amended, after which the work must be repeated according to the revised conditions. If such a study has been systematically carried out, and the variations have been reduced to the limit of error involved in a practical application of the same, then the hypothesis becomes a theory, which is stronger according to the number of types to which it may be successfully subjected. In the conditions of organic work in nature, there are to be considered, what is the result, and how has that result been obtained? And that these problems are independent, one of the other, is of special significance in the presentation of this paper. In the formation of a soil, using the word soil, in an agricultural sense, the practical interest is, to what degree of perfection has the work been done? That is, what state of fertility has the soil reached? The fertility of a soil is indicated by vegetation, if climatic conditions can be eliminated. To imitate the results obtained by vegetation, and condense the period of work from months to hours, means to forecast the amount of ingredients which go to make up a crop, which means the opportunity for supplying those ingredients, which would otherwise have been found deficient by the crop. And as such an imitation must necessarily be under purely arbitrary conditions, why should we feel called upon to use a single principle found in nature, other than those which suggest themselves on account of their simplicity. The simple fact is presented, that a certain species of vegetation has accomplished a definite result, upon a specified type of soil. That the soil has given to the vegetation that amount of mineral matter, which was in a condition to be attacked and assimilated by the solvent and absorbent properties of the plant. And to estimate the amount of mineral matter which is in such a condition is to forecast the results of vegetation, always barring climatic conditions. The fact that a crop did take up phosphorus pentoxide to the extent of 20 pounds per acre, is the very best of evidence that at least that much was available for the crop. Just how the crop accomplished the result, and what the long list of chemical reactions are, is at most but of secondary interest. It is most apparent that the premise in such a line of reasoning is dependent upon the accuracy with which the vegetation indicates the degree of fertility. The growth of a crop is dependent upon the fertility of the soil and climate. In order to study one, the other must be eliminated. The only way to eliminate climatic conditions, is by culture in pots where the moisture and temperature are controlled. By the use of pots, any number of soils may be subjected to the same artificially perfect climatic conditions, which is to eliminate such conditions. The series of pot experiments, which have been conducted by the department of agriculture for the past five years, has been noticed by most of those in this country, interested in such work. There are 175 pots in use, filled with virgin and cultivated soils. and subsoils, including muck soils from Florida. The cropping has consisted of oats and beans, in duplicate pots, thus subjecting each type of soil to cereal and leguminous cropping. Each crop is always followed by buckwheat as a second crop, the same season, the same condition existing every year. A description of the pot culture, together with a detailed description of the methods that have been used in the planting, watering, harvesting, etc., has been prepared by Dr. Wiley, the head of this bureau, and is shortly to appear as the introduction to a bulletin on soil study. Awaiting this, the writer will not make reference to the cropping in detail, and upon which the accuracy of this paper is based. It is proper to state that the work was originated by Dr. Wiley, who collected the samples with much judgment, and subsequently entrusted the line of research pertaining to the mineral food, and the compilation of such data, to the writer, who, with the exception of the nitrogen determinations, accepts a personal responsibility for the analytical work here presented. To work backward, is many times the simpler method. If we have unquestionable results which were obtained by a crop, and a sample of the soil which was taken just previous to the planting, we have a definite result for which to work. If it is given what amount of mineral matter must be dissolved from a given amount of soil it is possible to vary the arbitrary conditions, until the desired result is obtained. And the more simple and elastic are the conditions made, with greater ease and accuracy can they be varied. Assuming then that it is simplified to the process of obtaining a definite result by the action of a solvent upon a substance, the conditions of solubility naturally suggest themselves as the solvent and its strength, temperature and time of digestion, degree of agitation and proportion of solvent to substance. In a general sense, to vary any one or more of these conditions is to vary the result. In this study, five hours has been adopted as the time of digestion, it being appropriate to weigh out the samples, digest and filter, in a day's work of seven hours. Two hundred grams of soil per liter are adopted as being comparable with custom and accuracy of solvent effect. The temperature of digestion is fixed at 40°, that being the lowest constant temperature obtainable in summer. The question of agitation is of greatest importance, and constant results could be obtained only in the maximum degree, so continuous shaking is adopted. In comparing the effects of shaking by hand, three or four times per hour, with continuous shaking, in the latter case the results were sometimes more than doubled in the potash, other conditions remaining the same. The method employed in this laboratory is that of the slow upsetting device, usually known as the Wagner machine, and making about 40 revolutions per minute. This has the effect of keeping the soil continuously suspended in the liquid. This machine has been modified by Dr. Wiley so as to permit of digestions being made at definite temperatures. The modification comprises a well-fitted double wall sheet iron chamber, in which the revolving shaft together with the attached flasks, are encased. By use of a thermostat, and owing to the circulation of the air as caused by the revolving of the flasks, such a temperature as 40° may easily be maintained for hours with no more variation than 0.5.° For a solvent, the simplest mineral acid, hydrochloric, is adopted. The condition admitting of the greatest suscepti |