is little difference between the sulphate and chloride. Ordinary superphosphate, 28 per cent. soluble, is better than richer ones. When the potato crop is insufficiently manured, not only is the produce less, but the proportion of small tubers is greater; potash is the most important ingredient for the production of large tubers. A well-balanced manure for raising a potato crop without dung should contain about equal parts potash, ammonia, and phosphoric acid. When applied with dung, half the potash may be used. Manure for potatoes should be within easy reach of the roots, but may be slow acting; dung appears to be the most appropriate manure, producing a crop with a large proportion of large tubers. With dung, nitrate of soda is better than sulphate of ammonia; without it, the sulphate of ammonia is best. Where much dung is used, potash salts may not be required. J. M. H. M.

Production of Nitrates in Arable Soil. By P. P. DEHÉRAIN (Ann. Agronom., 13, 241–261). The present series of experiments was undertaken to determine the conditions under which different substances acting as manures nitrify in the soil. They were made with the nitrogenous black matter of soil, ammoniacal salts, oil-cakes, farmyard manure, and the black matter of farmyard manure, under varying conditions of moisture and for different periods of time; in some instances, the proportion of manure to soil, and the temperature, both varied. The quantities of nitrate formed in the various mixtures from time to time were determined as follows:-100 grams of the mixture were introduced into a litre flask, the quantity of water required to make it up to the mark measured into it, left in contact for several hours, and repeatedly shaken; then 700 to 800 c.c. were withdrawn and evaporated down to about 10 c.c., and the nitric oxide determined according to Schloesing's method by measurement of the gas.

Nitrogenous Black Matter of Soil.-The soil employed contained 0160 per cent. organic nitrogen and 0.015 per cent. nitric acid (NO). Quantities of 100 grams of this soil were placed in glass beakers and moistened with 5, 10, 15, 20, and 25 c.c. of water respectively, and all the beakers were placed together under a bell-glass, and the nitric acid formed determined after several periods up to 90 days. With the 5 per cent. of added water, there was no nitrification at the end of 56 days, but all the other mixtures had commenced to nitrify at the end of the first period. After 90 days, the quantity of nitrate formed in a tonne (1000 kilos.) of soil was 250 grams in that moistened with 5 per cent. of added water; 16 grams with 10 per cent. water; 270 grs with 15 per cent. water; 290 grams with 20 per cent. water; and 370 grams with 25 per cent. water-the nitric acid formed thus attaining a maximum in the most humid mixture. The greatest rate of nitritication observed was 1 gram nitrogen nitrified per day per tonne of soil. Supposing this rate to be maintained during the active life of a cereal crop, say 100 days, the nitrogen nitrified would amount to 300 to 400 kilos. per hectare, a quantity greatly in excess of the requirements of the crop.

Ammoniacal Salts.-Quantities of soil watered with various propor

tions of ammonium sulphate were placed in identical conditions with plain soil freely exposed to the air.

Sulphate of ammonia in quantities equivalent to 0·030 gram and 0010 gram nitrogen per 20 grams of soil far from increasing the quantity of nitrate formed greatly hindered the nitrification in the soil. In a saturated atmosphere, however, with the soil freely supplied with water, doses of sulphate of ammonia up to 0.100 gram nitrogen per 100 grams soil did not prevent nitrification, although the larger the doses employed the less nitrates were formed.

A series of experiments with sulphate of ammonia equivalent to 0010 gram nitrogen per 100 grams soil, moistened with 5, 10, 15, 20, and 25 grams water, showed that with the smaller proportions of water nitrification was less than in the parallel experiments with plain soil; with the larger proportions of water, more nitrogen was nitrified, that of the ammoniacal salt undergoing oxidation as well as that of the organic matter of the soil, and at a much faster rate, so that, for example, the addition of 0010 gram nitrogen as ammoniacal salt to 100 grams soil containing 0-160 gram organic nitrogen doubled the quantity of nitrate formed during the course of the experiments. Frequently as much as one-hundredth of the nitrogen added as ammoniacal salt was nitrified per diem, whilst the nitrogenous matter of the soil scarcely gave rise to one-thousandth of its weight of nitric acid in the same time. The whole of the sulphate of ammonia added as manure is probably nitrified in the course of a few months. Some of the nitrates at first formed disappear afterwards, more especially in the most humid mixtures, and this is attributed by the author to the slow growth of moulds.

Maize Cake. The nitrogen of maize cake is nitrified more quickly than that of the organic matter of soil, more slowly than that of ammonium sulphate, so that if a mixture of the two is used as manure, when the action of the latter is exhausted that of the former still continues. The influence of humidity is the same as in the preceding

cases.

Farmyard Manure.-2 per cent. of farmyard manure was added to the soil, corresponding to a dressing of 72 tonnes per hectare. Even with 5 per cent. of added water some degree of nitrification occurred, whereas with plain soil of the same humidity there was none. With the largest proportions of water, nitrification was very irregular, no more nitrate being formed after 90 days than after 26, and in some cases the nitrate formed was less than with plain soil. These irregularities are doubtless due to the growth of moulds, &c., of which the spores and mycelia abound in farmyard manure. In point of rapidity of nitrification, farmyard manure comes next to ammonium sulphate, and before maize cake and plain soil.

Black Matter of Farmyard Manure. This was obtained by extracting decomposed farmyard manure with water, filtering, and evaporating the filtrate to dryness by exposure to air alone. It contained 3.5 per cent. of nitrogen, and was added to the soil in the proportion of 03 per cent., thus introducing the same quantity of nitrogen as that contained in the 0·050 gram ammonium sulphate used in a former series of experiments. With these mixtures, nitrification occurred

exactly as with plain soil, that is to say, the nitrogen of the black matter appeared to undergo no change. It appears to lose the capacity of being nitrified by the act of drying, since in trials with the undried extract of farmyard manure the nitrogen of the extract was nitrified in sensible quantities. J. M. H. M.

Free Phosphoric Acid and Superphosphates. By M. WEILANDT (Landw. Versuchs-Stat., 34, 207-215).-Ritthausen has previously found that a solution of superphosphate and calcium carbonate produces dicalcium phosphate, and generally in the crystallised condition. The author has now studied the results of the reactions of a 10 per cent. solution of orthophosphoric acid on several alkaline carbonates. Pure barium and strontium carbonates produced 65.2 and 45.7 per cent. respectively of the insoluble phosphate, but when carbonic anhydride was passed in rather less of the bibasic and more of the monobasic phosphate and free acid were produced, as also many crystals. When, however, the concentration of the solution was raised, the results were reversed, but when the time of action was lengthened there was rather less of the bibasic salt produced, but in the filtrate was found a large conversion of the free acid into the monobasic salt. It would appear, then, that the crystallised phosphoric acid acts somewhat differently from the superphosphate.

The author, therefore, prepared a superphosphate from Mejillones guano and phosphoric acid, and with the solutions thus obtained he experimented on marble and other forms of carbonate, of which marl appeared to react most energetically. From the results obtained, it is evident that if superphosphate is applied to a well marled soil none will be lost by passing through unabsorbed in the monocalcium form, but all (98 per cent.) of the acid will be retained at the surface. As was to be expected, the more crystalline compounds are least readily acted on, whilst marl, which is readily decomposed, and in which the calcium carbonate is finely dispersed, becomes coated with the bibasic compound and prevents any further reaction.

E. W. P.

Basic Cinder and other Finely-ground Insoluble Phosphates as Manures. By AITKEN (Trans. Highland and Agr. Soc., 1887 [4], 19, 245-253).-The phosphates tried were Curaçoa phosphate with 87 per cent., Canadian phosphate with 59 per cent., Carolina phosphate with 57 per cent., Belgian phosphate with 40 per cent., and basic cinder with 40 per cent. of phosphate of lime. These were tried against each other and superphosphate of 28 per cent. at four different stations on turnips. It was found that only 40 to 50 per cent. of the ground Curaçoa, Carolina, Canadian, and Belgian phosphates would pass through a sieve of 120 wires per linear inch, whereas all the basic cinder passed through this sieve. A large sieve of this gauge was therefore made, and sufficient quantities of the mineral phosphates passed through it to try the sifted against the unsifted portions. Sulphate of ammonia and sulphate of potash were added in all cases to make a general manure, and quantities of the phosphates containing equal weights of phosphoric acid were employed. In almost all cases, the sifted phos

phates produced a larger crop than the unsifted phosphates, amounting on an average to a gain of 1 to 1 tons of roots per acre, and it seems reasonable to suppose that this increase would pay for the extra fineness of grinding. The superphosphate proved little if any superior to the sifted phosphates or the basic cinder, and this last phosphate appears to be at present the most finely ground as well as the cheapest in the market. The nature of the mineral phosphate used appears to be of much less importance than the fineness to which it is ground. J. M. H. M.

Note by Abstractor.-The author says of basic cinder, "I do not doubt that its efficacy as a phosphatic manure is in the main to be attributed to the extraordinary fineness to which it is ground." No doubt fineness of grinding is extremely desirable, but the basic cinder appears to possess a distinct advantage over the mineral phosphates in the great relative solubility of the phosphate of lime contained in it. On some soils this tells very much; in the Downton experiments, conducted in 1885 by the Abstractor and another (this vol., p. 176), the coprolites were much more finely ground than the basic cinder employed, yet the latter produced far better crops than the former. J. M. H. M.

Ground Felspar as a Potash Manure. By AITKEN (Trans. Highland and Agr. Soc., 1887 [4], 19, 223–255). The result of two small experiments, made at Pumphuston on turnips and at Boon on peas, is that felspar when ground so as to pass through a sieve of 120 meshes per linear inch undoubtedly acts as a potash manure. On the turnips, the felspar did better than an equivalent quantity of sulphate of potash, on the peas not so well.

J. M. H. M.

Butter from various Districts. By E. DUCLAUX (Compt. rend., 104, 1727-1730).-The author has determined the proportion of volatile acids in various butters by the methods previously described. The butters examined were obtained from Isigny, Gournay, Brittany, and Cantal.

The ratio of butyric acid to caproic acid is practically constant in butter from the same district, but varies considerably from one district to another. In butter from Gournay, it is 16, Brittany 19, Isigny 21, Cantal 24. The total quantity of the two acids varies in different samples even from the same district. The production of butter of fine flavour and good quality is not simply a question of the method of preparation. C. H. B.

Analytical Chemistry.

Iodine Detection in Laminaria. By F. A. FLÜCKIGER (Arch. Pharm. [3], 25, 519-522).-One decigram of the seaweed is ample for the detection of iodine. The weed in as small pieces as possible is mixed with twice its weight of not too fine pumice powder, boiled with water, and dried on the water-bath whilst being well stirred.

The

mass is then roasted over a very gentle flame and extracted with 5 c.c. water in which the iodine can be detected by the addition of ferric chloride solution to acid reaction and shaking up with carbon bisulphide. Or a very dilute solution of chlorine-water may be used with one portion of the filtrate and ferric chloride with another.

J. T.

Iodimetric Studies: Application of the Method to the Analysis of Commercial Substances. By G. TOPF (Zeit. anal. Chem., 26, 277-302. See this vol., p. 688).-Commercial sodium thiosulphate is nearly pure, and usually contains such minute traces of sodium carbonate that its titration by iodine is not affected thereby; nevertheless if present in larger quantity this salt would cause an error in the result equal to double its own amount, unless acidified iodine solution is used.

The iodimetric method serves well for the assay of commercial sodium sulphide. It is necessary to add the sulphide to an excess of acidified iodine solution, otherwise the alkali present induces oxidation of the sulphur, and causes the results to be too high.

Bunsen's explanation of the necessity for diluting sulphurous acid to 0.05 per cent. before titrating with iodine, namely, that above that strength a reverse action occurs, does not seem to be correct. Finkener has shown that with an excess of iodine the action is complete in a solution of any strength. Even sulphuric acid of 20 per cent. strength liberates no iodine from potassium iodide. Stronger acids are reduced, especially when much iodide is added, but up to 71 per cent. the product is hydrogen sulphide. Above that strength, sulphurous acid is also produced, but is not the sole product till 81 per cent. is reached. If Mohr's plan of neutralising the sulphurous acid before titrating it is adopted, it is still necessary to use an excess of iodine and to acidify before titrating back with thiosulphate.

In titrating stannous salts by iodine, the best results were obtained by adding tartaric acid and sodium hydrogen carbonate, then an excess of iodine, and then acid before titrating back. Zinc powder can be assayed by shaking it with excess of neutral iodine solution (in a bottle containing glass balls), and acidifying with acetic acid before titrating the excess.

In examining commercial iodine containing chlorine, it is necessary to dissolve in soda and add a small quantity of a sulphite to reduce the iodate formed by the action of the chlorine. A ferric salt and hydrochloric acid are then added and the iodine is distilled over into potassium iodide. The author figures a special distillation apparatus in which caoutchouc joints are avoided, and which gives excellent results. The same apparatus serves for the analysis of chlorates and hypochlorites by distillation with hydrochloric acid.

Diehl's method of titrating Weldon mud by digesting with potassium iodide and acetic acid and determining the liberated iodine is found to be a very practical one. Lead dioxide and red lead can be estimated in the same manner if sufficient sodium acetate is added to keep the lead iodide in solution. The results are more trustworthy than those obtained by distillation with hydrochloric acid. Nitrates,