SCHÖNBEIN'S TEST FOR NITRATES.

Storer has explained at length the reaction proposed by Schönbein for the detection of minute traces of nitrates. Two modes of applying the test were originally given. In the one, dilute sulphuric acid and iodo-starch paste were added directly to the nitrate solution, and the mixture was stirred with a zinc rod. In the other and better mode, the nitrate was first reduced by means of zinc or cadmium, and then the solution was acidulated with sulphuric acid and the iodo-starch paste added. He finds, contrary to the opinion of Carius, that no objection seems to lie against this test on the score of delicacy; but that it has a fatal defect in the forms in which it is commonly used, in that mere water, entirely free from any nitrates or nitrites, will, on being treated with zinc or cadmium as in the process of testing, react upon iodo-starch as if these compounds were present. This coloration, thus produced, the author shows is due to the hydrogen peroxide which is formed by the action of the metal itself upon the water-a reaction observed by Schönbein himself. Hence, whenever the degree of coloration of the iodo-starch obtained in testing for a nitrate by this method is less intense than the tint obtainable from 0.000187 gramme of potassium nitrate in 50 cubic centimeters of water, it is diffi cult to decide whether the coloration may not be wholly due to hydrogen peroxide. Having thus shown the defect, the author set himself at work to remedy it, and to devise a mod ification of the process which, while preserving the delicacy of the test, should yet be easily applied. The very simple plan was adopted of acidulating the solution very slightly, before boiling the nitrate in it, with metallic cadmium. Under these circumstances no hydrogen peroxide is produced, while the reduction of the nitrates goes on quite as well. The only precaution necessary is to prevent the loss of any of the nitrous acid, which is easily accomplished by attaching to the flask a small inverted Liebig condenser during the boiling. In his experiments, 0.0001 gramme of nitrogen pentoxide, placed in the solution as potassium nitrate, in 50 c. c. of water containing two drops of dilute sulphuric acid, gave, after boiling for five minutes, a reaction in less than fif teen minutes. Even 0.00005 gramme gave the reaction in

the course of half an hour. Zinc, amalgamated zinc, aluminum, iron, lead, and magnesium were also used as reducing agents, but none of them acted as efficiently as cadmium. No experiments were made with the alkali metals.—4 D, III., xii., September, 1876, 176.

HYDROCELLULOSE,

In some of the processes in the arts in which woody fibre is used, as in paper-making, for example, the fibre appears to undergo during the process of manufacture a peculiar transformation, by which it is rendered friable. Aimé Girard has investigated the matter, and finds that this change is owing to the assumption of a molecule of water by each molecule of the cellulose, thus producing a new body having the composition C12H22O11, to which he gives the name of hydrocellulose. To prepare it some form of purified cellulose, such as carded cotton, is placed in sulphuric acid of 45° Baumé in the cold for twelve hours. It is then well washed, pressed, and dried. After it is dry, its fibrous character is destroyed by pressure; rubbing between the fingers converts it into a white powder. Girard supposes that this substance may be formed in the process of bleaching paper-pulp by chloride of lime; and accounts in this way for the peculiar brittleness of certain papers found in commerce.-6 B, LXXXI., December, 1875, 1105.

A NEW OXIDE OF Sulphur.

Weber has investigated the cause of the intense blue color which is developed whenever sulphur is allowed to act upon sulphuric oxide or disulphuric acid, and has shown that it is due to an oxide of sulphur hitherto unknown, which he has succeeded in isolating and examining. To prepare it, carefully dried flowers of sulphur are thrown in small portions at a time into sulphuric oxide containing sulphuric acid. At the instant of contact, the sulphur is converted into dark blue liquid drops which sink to the bottom of the liquid and there solidify. Care should be taken to keep the temperature at 15° Centigrade, since below this point the whole liquid solidifies, and above it the blue body decomposes. After the operation the excess of liquid is poured off, the blue crystalline crusts are drained, and the excess of sulphuric oxide driven

off at a temperature not exceeding a blood heat. Bluishgreen crusts are thus obtained which are very friable, and which have a structure similar to malachite. They decompose without fusion slowly at ordinary temperatures, more rapidly on heating, evolving sulphurous oxide, and leaving sulphur behind. In a cool place the decomposition is so slow that the substance may readily be weighed for analysis. Moist air decomposes it rapidly, and it hisses when thrown into water. Alcohol and ether also decompose it and set free sulphur. A mean of five closely accordant analyses showed that it contained 57.12 per cent. of sulphur, thus giv ing it the formula S,O3. The author names it sulphur sesquioxide, or dithionic oxide. No compounds of it have yet been made. Selenium gives an analogous compound, having the formula SeSO3. It is dirty green in mass, but is yellow when in powder.-Poggendorff's Annalen, CLVI., December, 1875, 531.

CORROSION OF PLATINUM STILLS BY SULPHURIC ACID.

Scheurer-Kestner having communicated to Hofmann, in 1862, certain results which he had obtained in the process of concentrating sulphuric acid in platinum stills (which results were published by the latter in his Report on Chemical Industry), and these results having been since that time called in question, the author has examined the facts in the case still more fully, and now publishes a new set of obser vations. From 1851 to 1861, 4309 tons of sulphuric acid were concentrated to 66° Baumé, in an alembic the body of which weighed 40 kilogrammes. The entire loss of this part of the still during this time was 12,295 grammes, being 2.859 grammes for each ton of acid worked. Perceiving that the cause of this large loss in platinum was the presence of nitrous products, ammonium sulphate was added to the acid in amount just sufficient to destroy them. In 1862, 1843 tons of acid were concentrated in the still, with a loss of 2490 grammes, being only 1.22 grammes of platinum for each ton of acid, a marked decrease. From 1864 to 1875, 17,516 tons (of 1000 kilogrammes each) were concentrated to 66° in a still the body of which weighed 50 kilogrammes. The acid contained sulphurous acid, but no nitrous compounds. The loss of the still was 16,178 grammes, or 0.925 gramme to the

ton of acid. It may be assumed, therefore, that to produce an acid of 66° Baumé (containing 94 per cent. of real acid), there is a loss to the still per ton of acid of one gramme when nitrous compounds are absent, and of 24 to 3 grammes when they are present. If, however, the concentration be carried above 66°, these numbers are much increased. In a still weighing 30 kilogrammes, 180 tons of acid, containing 97 to 98 per cent. of real acid, were produced. The still lost 1092 grammes of platinum, or 6.07 grammes per ton of acid. In producing 47 tons of acid of 99 per cent., there was a loss of 8.8 grammes of platinum per ton of acid, and an analysis of the acid showed that it contained 8.38 grammes of platinum to the ton, present in it in solution, thus proving the loss to be a chemical one. To the figures here given for the loss of the body of the retort, about 13 per cent. should be added for the loss of the other parts. The use of a platinum-iridium alloy for the stills prevents to a large extent this action, but the brittleness and consequent fragility of the alloy is a serious objection to it.-Bull. Soc. Ch., II., xxiv., Dec., 1875, 501.

NON-OXIDATION OF CARBONIC OXIDE BY OZONE.

The question of valency among the elements is a fundamental one in chemistry; and while all chemists are agreed that a given atom may form a series of compounds with the same substance, they are divided on the question of the interpretation of this phenomenon. On the one hand, it is claimed that the valence of an atom is fixed and invariable, and consequently all but one of its compounds, and that the highest, must be unsaturated. On the other, it is asserted that the valence is variable, and that by twos, and consequently that in the lower of two compounds formed by an element it is as fully saturated as in the higher. Carbonic oxide, for instance, whose molecule contains one carbon and one oxygen atom, has, according to the former view, two free bonds. But according to the latter it is saturated, though at a lower stage than in carbon dioxide. The weak point in the theory of variable valence is the law of, or the cause for, the variation. The constitution of carbonic oxide being thus in doubt, Remsen and Southworth have sought to throw some light upon it by acting upon it with ozone, a body

which is supposed to give up its extra oxygen atom with great readiness. For this purpose they passed the pure monoxide through potassium hydrate and lime-water into a flask. Into the same flask oxygen was passed after treatment with potassium hydrate and lime-water, and also after being ozonized. The interior of the flask was moistened. From the flask a third tube led the products first through lime-water and then over potassium hydrate. But, though the currents of gas were slow and the action was long continued, not a trace of dioxide was formed, even when the experiment was repeated in direct sunlight. This result is the more suprising from the readiness with which carbonic oxide is oxidized by chromic acid solution. The authors propose to study the action of hydrogen peroxide on this substance.-4 D, III., xi., February, 1876, 136.

ON THE HYDROCARBONS OF COAL GAS.

Berthelot has made a somewhat exhaustive investigation of the hydrocarbons existing in coal gas, the results of which go to confirm his theory of the action of acetylene and hydrogen at high temperatures. In the gas which is supplied to the city of Paris he finds that benzene may be detected very readily in so small a quantity as two or three cubic centimeters by exposing in it a drop of fuming nitric acid. On diluting the drop with water, the peculiar almond odor of nitrobenzene is at once recognized. If fifty liters of the gas be passed through 8 or 10 c. c. of the acid, enough nitroben zene separates on dilution to weigh; and from the amount thus obtained it appears that this gas contained two or three volumes of benzene vapor in the hundred. More accurately, the amount present is from 3 to 3.5 volumes. Next to marsh gas, therefore, benzene is the principal hydrocarbon in coal gas, and is the substance to which the gas mainly owes its illuminating power. Besides benzene, the gas contains ethylene, propylene, and butylene of the olefine series, and acety lene, allylene, and crotonylene of the acetylene series. These bodies were detected by absorbing them with concentrated sulphuric acid, diluting, and then fractioning. In one million volumes of the Paris gas, the author concludes that there are by this analysis: Benzene in vapor, 30,000 to 35,000 volumes; acetylene, 1000 (about); ethylene, 1000 to 2000; propylene,