NEWS

nesian salt was only one-half more than was strictly necessary.

(2). A very large excess of the magnesian salt causes an error of excess.

(3). A very large excess of citrate of ammonia causes an error of deficiency.

Carbonate of magnesia 20

200

Distilled water When the carbonate of magnesia is quite dissolved, 500 cubic centimetres of ammonia at 22° Baumé are added. The liquid heats, and the rest of the citric acid dissolves. It is allowed to cool, and is made up to the volume of a litre with distilled water, filtering if needful. The solution is permanent; it is decidedly acid, but only heats slightly when mixed with a large excess of ammonia. The precipitate thrown down by this liquid is re-dissolved on the filter with dilute nitric acid, re-precipitated with ammonia, collected and washed anew, ignited, and weighed. An equally accurate result may be obtained with greater speed by determining the phosphoric acid contained in the precipitate volumetrically, by means of a standard solution of the nitrate of uranium.

(1). All phosphates in an aqueous solution are thrown down by a solution of nitrate of uranium, and the precipitate is perfectly insoluble in water, even if acidulated with acetic acid. It dissolves, however, in dilute nitric and hydrochloric acids.

(2). If the liquid contains ammoniacal salts, the precipitate formed is ammoniaco-phosphate of uranium, containing 2 equivalents of oxide of uranium to I of phosphoric

acid.

In the absence of ammoniacal salts, but in presence of alkaline acetates, the precipitate has the same composition.

(3). If neither of these two classes of salts be present, the precipitate contains 3 equivalents of oxide of uranium to I of phosphoric acid.

(4). The slightest traces of a uranic salt are shown by the formation of a chocolate-brown precipitate when a drop of the solution is placed in the middle of drop of solution of ferrocyanide of potassium on a slab of white porcelain.

(5). Uranic phosphates held in suspension in water, or very dilute acetic acid, cause no colouration in ferrocyanide solutions.

A. Preparation of the Normal Solution of Phosphoric Acid. -A dilute solution of phosphate of soda is precipitated by a solution of sulphate of magnesia, hydrochlorate of ammonia, and ammonia. The precipitate is washed with distilled water containing 10 per cent of ammonia, first by decantation, and then on a filter, and dried at 100. It is then removed from the filter, and heated to dull redness in a platinum crucible. At the end of the operation the temperature is raised higher. It is then cooled, and the pyrophosphate of magnesia preserved in a wellstoppered bottle. To prepare the standard solution, 3.127 grammes of the pyrophosphate (containing 2 grammes of phosphoric acid) are weighed out, placed in a flat-bottomed flask, in which they are drenched with about 20 cubic centimetres of pure nitric acid. The whole is boiled on

are then poured back into the litre-measure. This solution contains o'i gramme of phosphoric acid in 50 cubic centimetres.

B. Solution of Acetate of Soda.-One hundred grammes of pure acetate of soda are dissolved in distilled water; 50 cubic centimetres of pure glacial acetic acid are added, and the solution is made up to 1 litre with distilled water. C. Solution of Uranium.-The purity of the nitrate of uranium is tested by dissolving a small quantity in ordinary ether, in which it should be perfectly soluble. To prepare the solution 40 grammes of the nitrate should be placed in a test-mixer, and 500 cubic centimetres of distilled water added. Ammonia is then added till a permanent turbidity appears. This is re-dissolved by means of a few drops of acetic acid; distilled water is added nearly up to the litre mark, and the liquid is set aside for 24 hours. The next day water is added, so as to make up the exact volume of a litre, and the liquid is filtered into the bottle in which it is to be preserved.

In determining the respective value of the standard solutions, it is important that the bulk of the liquid in every experiment should be the same. The addition of the acetate of soda somewhat diminishes the sensibility of the reaction; hence one and the same quantity should always be added. The next step is to find the quantity of solution of uranium necessary to produce the characteristic reaction with ferrocyanide in a given volume of liquid.

To this end drops of the ferrocyanide solution are placed upon a white porcelain plate with the end of a glass rod. These drops should not be more than 5 millimetres in diameter. On the other hand, 75 cubic centimetres of distilled water are measured into a beaker, and a Mohr's burette, graduated into tenths of a cubic centimetre, is filled with the uranium solution. The best way to fill the burette is to plunge its lower opening into the liquid, and suck it up by means of a caoutchouc tube fixed to the top. When the burette is full the liquid is let fall, drop by drop, into the measured quantity of distilled water, stirring after every drop. The end of a glass rod is then lightly moistened with the liquid, and applied to one of the drops of ferrocyanide solution on the porcelain plate. If a slight reddish colouration does not appear in the centre of the drop, more nitrate of uranium is dropped into the water. About ten drops of the solution have to be added before the colouration is produced. A second test is then made, adding to the 70 cubic centimetres of distilled water 5 cubic centimetres of the solution of acetate of soda, and proceeding as above; 20 drops, or 1 cubic centimetre, will now be required before the colouration is produced. This amount is marked on the label of the bottle, and is deducted from the result found in titrating any sample. In determining the value of the standard solutions, 50 cubic centimetres of the phosphoric liquid are put into a beaker of Bohemian glass, and 5 cubic centimetres of the acetate of soda are added. The beaker is then set on a sand-bath, and brought to the boiling-point. Into the boiling liquid 18 centimetres of the uranium solution are poured without testing the liquid. After that it is necessary to test at each halfcentimetre till the colouration is obtained. The number of degrees of the burette consumed is then read off. It will often be found that after no colour has been produced, the next half-cubic centimetre will give a deep red. In such a case it is necessary to repeat the trial, testing after every drop when near the point found above, until the very faintest colouration appears. Suppose, e.g., that 21 cubic

centimetres of the uranic solution have been consumed; as I cubic centimetre is required to obtain the colouration with the same volume of liquid without phosphate, the amount of uranic liquid which precipitates o 100 gramme of phosphoric acid is 21 120 cubic centimetres, and each cubic centimetre corresponds to 5 milligrammes of phosphoric acid. For actual analysis it is well to mark the bulk of 75 cubic centimetres, on a sufficient number of beakers.

1st Case.-Phosphates soluble in water and neutral or alkaline. Phosphates of potassa, soda, and ammonia. Five grammes are weighed and introduced into a flask marked at 100 cubic centimetres. Distilled water is added, and it is dissolved by means of heating and stirring. When the solution is complete, the flask is cooled down to the temperature of the atmosphere, and it is filled exactly up to the mark with distilled water, closed with an indiarubber stopper, and shaken. If the solution is not clear, it is quickly filtered. Should this operation prove tedious, the flask and the filtering-funnel are both covered with a bell-glass, to prevent evaporation. When the liquid is clear 5 or 10 cubic centimetres are taken, by means of a graduated pipette, previously washed with a little of the same solution. The liquid is put into one of the marked flasks, 5 cubic centimetres of acetate of soda solution added, and the volume made up to 75 cubic centimetres with distilled water, and the whole heated to a boil. The nitrate of uranium solution is then dropped in as above, until, on testing, the colour is obtained.

2nd Case. The phosphate is insoluble in water, but readily soluble in dilute nitric and hydrochloric acids, tribasic and bibasic phosphates of lime, and magnesia and phosphates of iron and alumina.

The substance is powdered, and 5 grammes are weighed out and put in a marked flask as above. A little distilled water is added, and 20 cubic centimetres of dilute nitric acid, applying heat if needful: when the solution is complete it is made up to 100 cubic centimetres with distilled water. It is then shaken up, and filtered exactly as in case 1; 5 cubic centimetres of the solution are taken, put in a beaker, and mixed with 10 cubic centimetres of the citro-magnesian solution, and a large excess of ammonia. The ammonia should give no immediate precipitate, but on stirring a crystalline precipitate of ammoniacomagnesian phosphate is deposited. It is allowed to settle for 12 hours, and then carefully filtered through paper free from lime and phosphates., The glass which held the precipitate is repeatedly washed with water containing 10 per cent of ammonia, and the liquid each time thrown upon the filter. The precipitate is washed on the filter with ammonia-water. A beaker marked at 75 cubic centimetres is placed under the funnel. The glass in which the double phosphate was precipitated is repeatedly washed with water containing 10 per cent of nitric acid, and the liquid each time poured upon the filter, washing the dissolved precipitate into the flask. The filter is then washed with distilled water. The volume of the liquid in the flask will not exceed 20 to 30 c.c. The liquid is then saturated with ammonia, adding it drop by drop, till the precipitate formed no longer dissolves on stirring. When the liquid remains very faintly turbid, one or two drops of nitric acid are added, to restore the transparency of the liquid. The 5 c.c. of acetate of soda are then added, the liquid made up to 75 c.c. with distilled water, and the titration performed as above. It has been found by direct experiment that the presence of sulphate of lime, chloride of calcium, salts of iron and alumina, whether severally or jointly, has no influence on the result.

We arrive at the following conclusion:

(1). The precipitation of phosphoric acid with ammonia and magnesia in presence of an excess of citrate of ammonia is an excellent method for separating phosphoric acid from the bases with which it is commonly combined. But if we are satisfied with weighing the precipitate immediately after ignition, we obtain, in the majority of cases, too high a result.

(2). The determination of phosphoric acid with a standard solution of nitrate of uranium, with the precautions pointed out above, gives results perfectly exact when the phosphoric acid is combined with magnesia or alkalies. But if lime, iron, or alumina be present, the results are inaccurate, and generally too low.

(3). By employing the citro-magnesian liquid to separate the phosphoric acid, and the uranic solution to determine it in the precipitate obtained, accurate results are obtained even in presence of an excess of iron and alumina. In a succeeding chapter the author examines the principal applications of the method to the substances most frequently requiring analysis.

THE PHYSOMETER, A NEW INSTRUMENT FOR THE DETERMINATION OF VARYING VOLUMES OF AIR AND OTHER SUBSTANCES.* (Concluded from p. 216).

THE apparatus described was primarily intended to be used in physiological researches.

In order to estimate the influence of the bladdermembrane on the changes in volume of contained air, the bladder of a large gilthead was filled with air, and the volume of this air, after correction for barometric pressure and temperature, ascertained directly (by weighing in distilled water, &c.), to be 76'403 c.c.

This bladder was put in the physometer-cage, and the cage three several times raised and sunk about 45 c.c. in the cylinder. The positions of the water-column in the measuring-tube are indicated in the following table :

The mean difference, with necessary corrections, is 191*54. Every millimetre of the scale corresponds with 8.022 centims. Hence, the entire difference of air-volume in the bladder in the two positions=1536 centims., i.e., about th of the entire air-volume contained. Now, deducing from this difference, simply in accordance with Boyle's law, the original volume of the contained air of the bladder under atmospheric pressure, this is found 69.81 c.c., a result which is 7.586 c.c., or 10 per cent, lower than the actual air-volume found directly, a difference too great to be attributed to error in method or observation. There is evidently another cause in action than the increasing pressure of the water. The calculation, according to Boyle's law, supposes a free expansion and contraction, whereas the resistance of the membrane must affect the result. This effect becomes most apparent in using a very dry bladder. At first the raising or sinking of the cage produces hardly any change at all in the measuring-tube; as the membrane becomes saturated, the water-column shows more motion; but, even when it is fully saturated, some allowance must be made. From a considerable number of observations, a coefficient of inertness (Trägheit) may be determined, at least for bladders of the same fish species.

It was also found that the bladder in the free state expanded and contracted somewhat more, for the same differences of pressure, than when in the body of the fish. The difference in the case of a tench was about 4 per cent.

Another point to be considered was the possible presence of air in the intestinal canal, which might affect the column in the measuring-tube like the air in the bladder. The fishes were opened after experiment; in

* Abstract of a paper in Poggendorff's Annalen, by P. Harting.

As an example of the observations made, a tench, weighing o 125 kilo., was put in the apparatus, which was supplied with fresh spring water at a temperature of 10°. At frequent intervals during three days (at the end of which time the fish died) the cage was raised and depressed, and observation made of the measuring-tube. Afterwards the fish was opened, the bladder (still filled with air) was taken out, and a series of physometer observations made with it.

It appears from the tabulated results that, although between individual observations there are considerable inequalities, the average difference in the water-column (from raising and sinking the cage) was in the one case 727, and in the other 719, showing a close agreement. The volume of air in the bladder (in free air), estimated by direct method, was compared with the volume calculated from the indications of the instrument by formula, and the coefficient of inertness thus obtained was 121. On comparing this with the coefficients of other bladders, the coefficient is found, as might be expected, to increase with the size of the bladder.

Lastly, the air contained in the bladder was analysed. Oxygen had quite disappeared. The constituents wereN 90.6 per cent

9'4

CO2 Soon after the fish had been put in the apparatus (at II a.m.) the volume of air in the bladder was 14°105 c.c. The animal was very sluggish in its movements, and its breathing was slight and irregular; the impure state of the water from which it had been taken had probably been unfavourable to its health. It continued vertical in the cage. Gradually the respiratory movements became fewer, and by about two hours had almost quite ceased, the volume of air being then 12.152 C.C. In the course of the day, however, the breathing was renewed, and the fish revived; and next day, twenty-four hours after it had been put in the apparatus, the quantity of air in the bladder had risen to 14.821 c.c.; five hours later it was 16.709 c.c.; next morning it was 21.158 c.c. The specific gravity of the fish was thus considerably diminished, and, from the position of the bladder in the body, the centre of gravity was so displaced that the animal could not retain the vertical, but was pressed with its right side against the lid of the cage. If the fish were in the normal state, the superfluous air found exit by the ductus pneumaticus (as was observed with several of the Cyprinida), and the vertical position was recovered. But in the present case there was no such liberation of air, and, by the evening of the third day, the quantity was 21.722 c.c. The breathing movements were again very much enfeebled, and it was evident these would soon cease, on account of the failure of oxygen in the water. It was, however, highly probable that the air in the bladder consisted in great part of secreted oxygen, which, when the gill-breathing ceased, would return to the blood. In the following night the fish died, and next morning it still lay somewhat on one side, but not pressed against the inner surface of the lid. The volume of air had diminished to 14'95 c.c., and, as stated above, there was not a trace of oxygen, but about o'r of CO2.

The above case is instanced as showing the use to which the physometer may be put. The number of experiments made is as yet too small to permit of much generalisation. As to the question whether fishes possess the power of changing their specific gravity by compressing at will the air in their swimming-bladders, experiment did not furnish a sufficient answer. Only with one of the fish (a whiting)

In the case of a perch, the influence of the gill-breathing on the water in the tube was very apparent, each movement of the gills (when the fish was at rest) corresponding to a rise and fall of the column, so that one might count the respirations without seeing the fish. Sometimes the gills would open a little wider than usual, and this, too, had its effect in greater rise and fall of the column. The sensitiveness of the instrument is thus shown.

Not only fishes, but other animals which contain air, might be experimented on with the physometer. If it were practicable to introduce a nautilus into it, some light might be thrown on the disputed question how this animal rises and sinks in the water.

Another use to which the physometer might be put is that of determining the quantity of air in the lungs of the new born, who have breathed only a short time and then died. The results thus obtained would be much more certain than with the simple lung-probe, as the smallest quantity of air might not only be indicated but measured.

Again, the physometer might be employed to make visible the diminution which takes place in muscles during contraction. The wires of an induction apparatus might be connected with the knobs of the brass rods.

For the observation of simple physical phenomena, the physometer might prove serviceable. Two experiments, similar to those with the bladders, were made with the small caoutchouc balloons which, filled with hydrogen, are often used as toys. In the first, the balloon was but slightly distended with air, and the mean difference obtained in the water column from raising and sinking the cage through 115 centims. was 116.65 m.m., which corresponds to o'9451 c.c. The air volume, directly observed by weighing, was 36'16 c.c., and the volume, as calculated from the physometer experiment by formula, was 31°55 c.c., showing the proportion : 1:24.

In the second experiment, the balloon was distended so as to have an air volume of 178.83 c.c., or more than four and a half times as much as in the first experiment. The volume, as calculated from a series of five raisings and sinkings, was 131'09 c.c., showing a proportion to the other of 1: 137. A comparison of these results shows that, with stronger tension of the balloon, its resistance and influence on the expansion and contraction of the enclosed air increase.

But caoutchouc itself, as Wertheim showed many years ago, increases in volume when drawn out. His method of measurement (with compasses) could not give very accurate results. M. Villari last year made similar experiments by a much better method, finding the specific gravity of the same band when extended and when unextended. It might, perhaps, be objected that possibly in the extended caoutchouc microscopically small fissures were produced, into which the air entered, and that this was the cause of the change of specific gravity. With the physometer, this change of volume might be easily rendered visible and measured. The following experiment shows this:

On the bottom of the glass cylinder was placed a weight of 3 kilogrammes. By an iron hook, two rings of vulcanised caoutchouc were attached to it; a second hook, catching the opposite part of the rings, being attached to the end of one of the brass rods. The diameter and thickness of the rings were very carefully measured, and the volume of both was found to be 5506.88 c.m.m. The apparatus having been filled with water, and all bubbles removed, a measuring tube was inserted, on which each millimetre corresponded to a volume of 3.3003 c.m.m. The second brass rod was now pressed down, and thus

the one with the caoutchouc rings raised, till the weight | hung free, the rings being extended. The water in the measuring-tube rose about 5 m.m. Allowing the weight and rings to return to their first position, the water column descended to its former height. This was repeated several times, and always with the same result. The increase of volume in the caoutchouc thus indicated was 16:5015 c.m.m., that is, 0'003, or of the original volume. Doubtless with a heavier weight and thinner measuring tube a still greater effect would be obtained.

ON A NEW SENSITIVE FLAME.

By GEORGE J. WARNER, F.C.S.

PERMIT me to call the attention of your readers to the peculiarly sensitive character of a new gas-burner lately introduced into the market.

As Wallace's burner may not yet be universally known, I append a brief note of its construction.

It consists of a hemispherical chamber, into which the gas is introduced through a cone fixed horizontally at a tangent, the position of the jet with regard to the cone being so adjusted that the quantity of air injected by the velocity of the gas at all ordinary pressures is always the proportion required for its perfect combustion.

The upper part of the interior of the chamber is lined with wire gauze, and from it issue one or more tubes, at which the gas is burned. The burner which I have used had only one tube.

At ordinary pressures the flame is of the colour of a Bunsen burner, but with a central cone, clearly defined, of pure green, whether it be turned high or low. But if the gas be reduced below the ordinary pressure on the main, the flame becomes white-tipped, and there is no longer perfect combustion, as in a defective Bunsen. We then find that the flame is sensitive to sound, to all sound in fact, but to high notes particularly.

I consider this a curious fact, as I believe it has been generally supposed that a high pressure is necessary to produce such a flame.

The first effect of the sound is to elevate the flame several inches, after which, if the sound be prolonged, it shrinks down, producing the same perfect combustion as at ordinary pressures, and this continues as long as the

sound.

It would appear, therefore, that the gas at a very low velocity does not carry with it sufficient air, and that the effect of the sonorous vibration is to increase the velocity of the gas so long as the vibration continues; so that by sound, by the rattling of a bunch of keys at a distance, we can bring the flame from a state of "imperfect" to one of "perfect" combustion.

Of course the burner is small, burning only about 1 or 2 cubic feet per hour; but I have a larger one in course of manufacture, which I expect will be infinitely more sensitive.

Ardwick Bridge Chemical Works, Manchester, May 5, 1873.

PROCEEDINGS OF SOCIETIES.

CHEMICAL SOCIETY.
Thursday, May 1, 1873.

Dr. ODLING, F.R.S., President, in the Chair.

THE minutes of the previous meeting having been read and confirmed, Messrs. William A. Prout and Alfred Payne were formally admitted Fellows of the Society. The list of donations was then announced, and the names of Messrs. Isidore Bernadotte Lyon and John Henry

These gentlemen were

Baldock read for the third time. balloted for and duly elected. Dr. H. SPRENGEL then read his paper "On a New Class of Explosives," in which, after stating that Mr. Nobel's important discovery in 1864 of effecting the explosion of nitroglycerine and analogous substances by means of a detonating charge, had initiated a new era in the history of explosives, he observed that an explosion may be regarded as the sudden release of that force which previously held together the molecules of gaseous matter and explosives as those solids and liquids which can suddenly assume the gaseous state. Heat is usually the cause of this change, so that explosions, as a rule, are simply rapid combustions of compounds which yield gaseous products: the higher the temperature produced, of course the more these gases are expanded, and the more powerful the explosion. On examining a large number of mixtures of oxidising and combustible substances, it was found that mixtures of nitric acid, density 15, and nitro-compounds of the hydrocarbons, fired by a detonating cap, were the most effective. A mixture of nitrobenzol, or picric acid with nitric acid, exploded with the greatest violence, comparable to nitroglycerine. In conclusion, the author drew attention to the harmlessness of the materials as long as they are kept separate, and the comparative ease with which they may be mixed. Something is gained even when one only of the components is liquid, and with this object he had proposed the use of porous cakes of potassium chlorate, saturated with a combustible liquid, such as bisulphide of carbon or nitrobenzol. had been found to be five times as effective as an equal weight of gunpowder in open granite quarries.

These

The PRESIDENT said they were much indebted to Dr. Sprengel for his account of these explosives, and although extremely interesting from a theoretical point of view, the question still remained as to whether they could enter into competition with those with which we are already acquainted.

Professor ABEL said he had witnessed some experiments with Dr. Sprengel's explosives which had been very successful. They no doubt opened out a new field for experiment, although there might be some difficulty in the practical application of corrosive liquids like nitric acid, or volatile ones like bisulphide of carbon. The mechanical state of the explosive had certainly a great influence on the result; for instance, slightly compressed gun-cotton did not detonate readily even if warm, since to obtain a satisfactory result there must be a certain resistance to mechanical motion; but well compressed gun-cotton could be detonated even when wet by employing a small portion of the dry cotton as an initiative; moreover, a low temperature did not in all cases make a difference, frozen gun-cotton detonating readily. When the cotton was SO circumstanced that it could resist mechanical motion it was very sensitive to the detonating charge, but when it could yield to mechanical motion it was not sensitive. These conditions had doubtless some bearing in the case of liquid explosives, as might be inferred from the fact that a shell charged with finelydivided gun-cotton suspended in water could readily be exploded by a small detonating charge.

A paper "On Zirconia," by J. B. HANNAY, was then read by the Secretary. The author has carefully examined the effect of precipitating zirconia at various temperatures, and finds that when it is formed at a high temperature, a considerable proportion of the precipitate is very difficultly soluble in tartaric acid. This explains the results obtained by Mr. Forbes when examining zircons for jargonium; the supposed new earth, insoluble in tartaric acid, being really zirconia altered by the temperature at which it had been precipitated. Zircons of low sp. gr. were found to contain uranium, iron, cerium, and didy. mium, besides silica and zirconia. The author also gives an account of the absorption-spectra obtained on gradually heating a borax bead previously saturated with zircon, and to which a little boracic acid had been added. For this