and which, with equal accuracy and fewer joints, would be applicable to all purposes to which the Fresenius and Will apparatus is suited.

The apparatus described in this paper, has proved all that is desirable. It is represented in the drawing of the natural size. A, having been washed. out and dried, by warming, and drawing air through by means of a tube

and cork inserted in the neck, is weighed, after having stood sufficiently long to become cool and to have attained the hygrometric state of its balance case. The carbonate is then added, and the weight a second time determined. After the addition of not too much water to A, the pipette, c, is adjusted, filled with nitric or hydrochloric acid in the case of insoluble sulphates, and empty in the contrary case, where it may be replaced by a simple tube. The pipette is filled by means of a tube and perforated cork, and closed by a wax stopper. After closing, it is warmed by the hand, so that a few drops of acid may escape, and that the acid may not, when cool, reach the extremity of the tube. This tube, if the bore is one millimetre in diameter, does not require to be drawn out at the extremity. is now adapted. The addition of water to the carbonate (in the case of dry alkaline carbonates) heats the flask, so that it must become cool before weighing, prior to the evolution of the gas. This is indicated by the rise of the acid in B, which acts as an air thermometer; by which means also the tightness of the joints is ascertained. When acid is used in c, care must be taken that some of it be not expelled by the alteration of temperature of A. When sulphuric acid is used to decompose the carbonate, the air is exhausted by suction at D, by means of a bent tube and perforated cork. The acid may thus be brought into A, and managed with the same control as in the apparatus of Fresenius and Will. The tube, E, is drawn out at both its extremities. It is filled with oil of vitriol by dipping in the acid, and exhausting at D by the bent tube and perforated cork. The last drop adhering to the extremity of E, must be

B,

carefully removed before connecting the apparatus. This acid is saturated with carbonic acid, as will be shown subsequently. During the decomposition of the carbonate, the gas passes through the drop of acid at the end of E, left by forcing the oil of vitriol over, and passing through the tube, the sides of which are also moistened with oil of vitriol, becomes partially dried before reaching the flask B. At the end of the operation, air is drawn through the apparatus. The wax stopper is then fitted to the pipette, and the flask, A, held over a small flame until the contents are in a state of ebullition; the wax stopper is slightly removed, a little more air drawn through, and the apparatus, when perfectly cool, weighed. In analyses where it is necessary to effect the decomposition of the carbonate by acid in the pipette, c, the operation is carried on in an analogous

manner.

This apparatus unites lightness with great strength, which is at once felt on handling it. This arises from the nature of the joint of E with B, and from the construction of the decomposition flask, which allows E to be made of a piece of stout tube. The one I first constructed and used for the following analyses, weighed when empty 36 grammes; and when charged for analysis, and with much sulphuric acid, together with the hook, between 50 and 60 grammes. It fits the hook which accompanies Oertling's balances for holding the chloride of calcium tube, &c., in organic analysis. An apparatus made of German glass, by Mr. Storm, professional glass blower, weighed when empty, with its corks, 20 grammes. The Professors Rogers, by violent shaking for 20 minutes, in a peculiar apparatus, having caused sulphuric acid of common density to absorb 94 per cent. its volume of carbonic acid, have, rather hastily I think, rejected the apparatus of Fresenius and Will as liable to error. In one respect every mode of analysis is liable to error; the question should rather be, what accuracy is possible by careful manipulation? The Professors Rogers appear to have rejected this mode of carbonic acid determination, a priori, without having given it a fair trial. It has been in almost universal use for some time, and by careful analysts, without objection. Mr. Noad, who also employed it frequently, testifies (Chemical Gazette, 1848, p. 176,) to its accuracy, and communicates the results of some of his analyses. Carbonate of Soda yielded, 41-41 and 41.48 per cent. CO2; theory requiring (Na=23·178) 41.37. Carbonate of Potassa 31.81 and 31.84; theory 31.88. Carbonate of Strontia 29-75; theory 29.8.

In answer to Mr. Noad's communication (Chemical Gazette, 1848, p. 477,) the Professors Rogers give the results of two experiments, which they say will indicate the amount of error which may arise in the process of Fresenius and Will. They have calculated their loss in carbonate of soda equal to a per centage of 0.65 and 1:05. When their experiments are calculated for a loss of carbonic acid, 70 grs. gave a per centage loss of 0.27, and 60 grs. a loss of 0.43 carbonic acid, which is not so large that one might despair of lessening it by perhaps more careful manipulation, which the Professors Rogers admit might be done by drawing more air through the apparatus. But these I think are not perfectly fair experiments, since they do not take into consideration small constant errors on the other side, which would tend to more general accurate results; and also they should have been performed on perfectly pure and dry carbonate of soda, and the

results compared with each other and with the theoretical per centage of carbonic acid. Lastly, they did not consider that the sulphuric acid in Fresenius and Will's apparatus is under different circumstances from the acid in their experiments, and also that carbonic acid is readily expelled from liquids in which it is dissolved by a current of air.

Nevertheless, as in a small apparatus like the present, where the small amount of substance taken for analysis renders precautions necessary, which may be neglected in a large one; and also in order to lessen the quantity of air drawn through, I prefer saturating the acid in в with carbonic acid; which is readily effected by adding to dilute sulphuric acid in a test tube (which the cork of the apparatus fits) a couple of crystals of bicarbonate of potassa, and afterward drawing air through.

The following experiments performed with the apparatus, will show its accuracy. The sulphuric acid used was pure and of density 1-76.

About one gramme of perfectly pure carbonate of soda, (which was tested,) yielded in two experiments, a per centage of carbonic acid 41.42 and 41.35, which differ 0·07 per cent. from each other, and of which the mean 41.38 differs 0.01 from the theoretical per centage of carbonic acid 41.37, Na. taken at 23.178, as in Weber's tables.

In order to obviate the use of the tube c, and thereby simplify the apparatus by omitting the tubulus of A, I tried three experiments respectively upon 0.83, 0.859, and 1.0193 grammes of the same salt, which yielded a per centage of carbonic acid 40.93, 40-34, and 40-49. The liquid in the decomposition flask was boiled about as long as in the former series of experiments; but after detaching the apparatus, a more prolonged boiling evolved carbonic acid, which was shown by its precipitating lime water, the precipitate being soluble in hydrochloric acid. The presence of the tube, c, is therefore not to be avoided. The tubulus on A, (though most convenient) may be avoided by passing c through the large cork.

Carbonate of lime, (precipitated, commercial,) which contained sulphate and iron, gave the following results, the carbonic acid being evolved by nitric.

0-63025 and 0.61625 grammes gave a per centage of CO2=41·65 and 41.70; difference, 0.05. The theory for pure carbonate of lime requires 43.88.

In order to ascertain the amount of error arising from evolving the CO2 by sulphuric acid, two more experiments were made, which gave for result 38-79, and 39.17 carbonic acid.

Since for a second analysis with this apparatus, it is only necessary to detach the bulb B, (to which a little more oil of vitriol is added if requisite,) and wash and dry A, which is readily effected owing to its size, a series of analyses may be more quickly performed than by Fresenius and Will's larger apparatus, and its accuracy enables it to be employed on small quantities of substance. Its lightness suits it to all balances in general use. The joints are but few and the connexions readily made, while the cork surface is reduced to its lowest possible value. It may be applied to the analysis of manganese, and to all analyses for which Fresenius and Will's apparatus is suitable.

As it is possible that an objection may be raised to the flask B, I would state that it is of quite simple construction. The first one that I attempted,

with which the foregoing experiments were performed, succeeded, as did two others made subsequently.

The construction of the apparatus is effected in the following manner: A tube of sufficient size to form the bulb, B, is drawn out, true, and thin at both ends; one end is cut off close to the large part of the tube, which is then slightly flanched. The tube E is then taken of glass of sufficient substance, (of 4 or 5 millimetres exterior diameter and 1 mm. thickness of glass,) and by rotating it in the flame, heating a narrow zone, and pushing together, a thickened ridge, called technically a marize, is raised. The extremity which is to dip into the oil of vitriol, is now drawn out and cut off, and the marize is joined to the flanched opening, and well incorporated with it by careful melting. The tube D, the small end of which is closed, is now joined to the large one in the usual manner, inclining it slightly upwards, and the bulb is then blown. If the heating has been properly conducted, the apparatus will result pear-shaped, which is the most suitable form, and D will be inclined at the proper angle. Should D turn out in too horizontal a position, it may be raised by carefully heating its lower junction with the bulb, and bending it upwards, at the same time blowing. The bottom part of the bulb is now cut off, and flattened as in the figure. Lastly, the other extremity of E is drawn out, and E is bent at two right angles. The end of the tube D being cut off, the flask B is complete. It is necessary in drawing out the small end of D, to melt the glass well together, so that the extremity be of the required strength.

A is formed by commencing precisely as with B; but instead of cutting off and flanching one end, the tubulus is at once formed by joining a tube like D, and then blowing the bulb. It will be requisite to incline this tube which forms the tubulus, so that the pipette c will reach the surface of the water in the decomposition flask. The upper extremities of the tubes which form the neck, and the tubulus are now cut off and flanched. When cool, a straight open tube is adapted to the neck by the cork, and the bottom of the bulb cut off and flattened. The joint of the tube E, with the flask B, is very secure, if the glass has been well melted together, owing to the marize. I have heated this flask repeatedly, in drying it after washing it out, without cracking.

I am not aware that a tubulus is ever joined to a bulb before blowing it, as here described; and yet such a junction is very readily effected, by one but little skilled in glass blowing, and when well made gives a joint which is absolutely not to be distinguished from the substance of the flask. I have made small retorts with tubulus in this manner; and if it be not already known, I would strongly recommend it to (at least amateur) glass blowers, and especially to those who use lead glass. Messrs. Bullock & Crenshaw, at the corner of Arch and Sixth streets, have procured a supply of this apparatus, made according to my directions.

New Art of Ornamenting Metallic Surfaces.*

Numerous as are the inventions or methods which have been applied for the ornamentation of metals within the last few years, we are not aware of any which, for simplicity and beauty, at all equal that recently * From the London Mining Journal, No. 906.

VOL. XXV.-THIRD SERIES.--No. 5.-MAY, 1853.

29

invented and patented by Mr. R. F. Sturges, of Broad street, Birming ham. It affords the means of decorating plain surfaces of objects forined of metal at a reduction of cost which throws all other processes, devised or invented, into the shade, while at the same time it materially improves their appearance. The invention depends upon the compression of a material between two or more plates of metal in the operation of rolling. It may astonish our readers to learn that the most delicate thread lace, such as is used in ladies' attire, perforated paper, or wire webbing, when passed through a pair of rolls, leaves an impression upon the sheet of metal, corresponding in depth to the compressibility of the material used as the pattern, and the density of the metal upon which the pattern is required to be impressed or indented. Various articles in electro-plate, Britannia metal, &c., such as tea services, salvers, waiters, &c., demonstrate the value of the invention, and show that for all ordinary purposes it is equal to the much more expensive process of decoration by engraving. In the brass foundry trade, the proprietor of the patent, R. W. Winfield, Esq., of the Cambridge street works, intends to introduce the same for the ornamentation of his patent metallic bedstead pillars, curtain bands, &c. Several of the objects we have examined prove satisfactorily the applicability of this class of ornamentation for the purpose. Mr. Winfield, we understand, will supply the metal from his extensive rolling mills in the sheet, which manufacturers and others may convert into such objects as may be required to be made therefrom, and ornamented by the new process. The opportunity now afforded for the production of articles of elegant form and delicately ornamented surface, will no doubt be appreciated.—Wolverhampton Herald.

Notice of the Explosions of Steam Boilers.*

A frightful accident occurred on the premises of the London and North Western Railway Company, at Longsight, near Manchester, on Tuesday morning, March 8th, whereby four men lost their lives, and at least a dozen received serious injuries. It seems that the engine No. 1, an old fourwheeled engine, made by Sharp, Roberts and Co., some fourteen years ago, and one of the first put upon the line, if not the very first, having undergone "a thorough repair," was in the Polygon shed, at Longsight, ready to start to act as pilot engine at the Standedge tunnel, on the Huddersfield and Manchester line. In a few moments more the driver would have taken the engine forward upon the turn-table in the centre of the building, in order that its head might be reversed, when suddenly the men in the Polygon were alarmed by a tremendous report and concus sion of air. The place was immediately filled with steam and smoke, and a portion of the building came rattling down, overwhelming a very large number of them. As soon as the smoke and steam had cleared away, the persons uninjured found the tender standing where it had been before the occurrence, with the back wheels of the engine and the plate over them, on which the driver stands, still adhering to the tender; the engine itself was some thirty paces off, beneath the centre of the building, and it was from that the firebox had exploded. On looking for the persons injured, they discovered four men dead in the ash pit next to that upon which From the London Artizan, April, 1853.