converged or superimposed the images, and found them so nearly to coincide, that the common outline was merely elegantly softened by the inequalities. In this experiment it appeared as if the eye, when the figures did not exactly coincide, had some power to complete the work or conceal the imperfections.

I have just succeeded in substituting a blank tablet for one of the pictures, and in tracing upon it with a pencil the illusie image converged from the other tablet. But this is not a very practicable method of copying pictures, requiring unusual command and steadiness of the optical axes for even the most moderate success in the operation.

ART. VII.-Analysis of the Bittern of a Saline on the Kiskiminetas River, near Freeport, Armstrong County, Pennsylvania; by M. H. BOYE, M.D., A.M.

(Read before the American Assoc. for the Advancement of Science, Philad., 1848.)

THE mother liquid, which remains after the brine has been so far concentrated as to yield no more salt by crystallization, is generally termed bittern, and thrown away. A great difference exists in regard to the quantity of bittern which remains at the different salines, some of them yielding a large amount of it, others leaving hardly any. The bittern which is the object of the present examination, was taken in the fall of 1841, from the first saltwork on the Kiskiminetas river, above Freeport, situated on the canal. It was selected on purpose from this saline, which was said to yield it in larger quantity than any other in the neighborhood, with a view to examine its chemical constituents, and more particularly to test it for bromine. Owing, however, to several causes, the examination was not begun till the spring of 1844. In the quantitative estimation of the different ingredients, I have been assisted by Mr. J. Hewston, a pupil of the High School.

The water had during its long standing, deposited at the bottom of the bottle in which it was contained, a crust of common salt, covered with a slight film of oxyd of iron, from which it was separated by careful decantation and subsequent filtration. By a qualitative examination, the water was found to contain lime, magnesia, soda, and a trace of potassa, in combination with chlorohydric acid. By the passage of chlorine through it, and subsequent shaking with ether, the latter assumed a rich orange color, indicating at once the presence of bromine. Chlorid of palladium indicated only a trace of iodine. It contained no sulphuric acid and no lithia.

Quantitative estimation.-1. The specific gravity of the liquid was ascertained, by weighing in a glass bottle, to be 1.339.

content of lime and magnesia in combination with chlorohydric acid, rendered it impossible to determine with accuracy by direct evaporation, its content of solid matter. It will however be seen from the following, that the solid ingredients contained in it amounted to 35-79 per cent.

2. A weighed quantity of the bitter water was super-saturated with ammonia, which produced no precipitate, and mixed with oxalate of ammonia until all the lime was effectually precipitated as oxalate of lime. The latter was collected on a filter, washed and incinerated with the usual precautions, and from the weight of the carbonate the amount of pure lime calculated. The filtered solution was evaporated to dryness, the ammoniacal salts expelled by heat, the residuary mass mixed with a small portion of water and oxyd of mercury and again evaporated to dryness, and ignited. On treating the mass with water, the magnesia remained undissolved. It was collected on a filter, washed, incinerated and weighed. The filtered solution was evaporated to dryness, and the weight of the remaining chlorid of sodium ascertained. By subsequent solution in water, and addition of chlorid of platinum and alcohol, it only yielded a minute portion of chloro-platinate of potassium.

3. Another weighed portion of the bittern was diluted with water and mixed with a sufficient quantity of dilute sulphuric acid; then an equal volume of alcohol of the ordinary strength was added to the whole mass. The precipitated sulphate of lime was collected on a filter, washed with a mixture of 5 vol. of alcohol and 6 vol. water, incinerated and weighed, and from it the amount of lime calculated. The filtered liquid was evaporated to dryness and ignited, and yielded the joint weight of the sulphates of magnesia and soda. These were then dissolved in water and precipitated with acetate of baryta. The filtered solution was evaporated to dryness, charred, and by treatment with water the carbonate of soda dissolved from the insoluble carbonates of magnesia and baryta. The former was separated from the latter by extraction with dilute sulphuric acid, and the amount of sulphate of magnesia estimated. Deducting the latter from the joint weight of the sulphate of magnesia and soda, the amount of sulphate of soda was obtained, from which that of the soda was calculated. As stated above, it only contained a small amount of potassa.*

4. Another smaller portion of water was acidulated with a few drops of nitric acid and then precipitated by nitrate of silver, and the weight of the chlorid of silver with its smaller contents of bromid and iodid of silver ascertained.

* This method of separating lime when in preponderating quantity by sulphuric acid and alcohol, the convenience and accuracy of which I pointed out in an essay on the analysis of magnesian limestones, by K. E. Rogers and M. H. Boyé, Journ. Franklin Inst., March, 1840, I cannot sufficiently recommend.

5. The accurate quantitative determination of bromine in connection with larger quantities of chlorine, is as is known connected with difficulties, and different experiments performed after the different methods recommended, yielded different results varying from 0:3 to 1 per cent. The result adopted in the following, was obtained by distilling a certain quantity of the water with peroxyd of manganese and a small but sufficient quantity of chlorohydric acid, receiving the bromine in a small globular receiver containing ammonia, and in the tubulure of which was fixed a modification of Liebig's potassa bulb, as employed by Will and Varrentrap in the analysis of nitrogenized organic compounds, but in this case containing, like the receiver, ammonia. The chlorid and bromid of ammonium thus obtained were decomposed by evaporation to dryness with caustic baryta, the mass treated with water, the excess of baryta precipitated by carbonic acid, and then the bromid and chlorid of barium precipitated by nitrate of silver, and the weight of the bromid and chlorid of silver ascertained after perfect desiccation on a counterpoised filter. It was then treated with nitro-muriatic acid by heat and its weight again ascertained, and from the difference in weight the amount of bromine calculated.

6. The amount of iodine was determined by precipitation of chlorid of palladium, and ascertaining the weight of the iodid of palladium.

Taking the mean of §§ 2 and 3, the results of the analysis were as follows:

from which the following composition of the water may be

derived.

ART. VIII.-On the Comparative Value of different Kinds of Coal for the purpose of Illumination; and on Methods not hitherto practised for ascertaining the Value of the Gases they afford; by ANDREW FYFE, M.D., F.R.S.E., F.R.S.S.A., Professor of Chemistry King's College University, Aberdeen, &c. Read before the Royal Scottish Society of Arts, April 24, 1848.*.

In a paper published in the Transactions of the Society for 1842, I gave an account of numerous trials made with the view of ascertaining the value of different kinds of coal for the purposes of illumination. Having been again engaged in conducting experiments for the same purpose, I have been induced to make some of the results public, because I conceive that they are interesting, and lead to valuable practical applications.

The objects which I have had in view were to ascertain not only the comparative amount of light afforded by the gases which the coals yield, but also the durability of these gases, so as to enable me to fix, as far as possible, their comparative value, and, consequently, their comparative expense, for the purpose of illu mination. Besides these, however, my attention was drawn to other circumstances connected with the consumption of gas, which, strictly bearing on the other part of the inquiry, are of importance.

1. Quality of the Gases.

In estimating the quality of coal-gases, and, consequently, fixing their comparative value for the purpose of illumination, we must take into account both the light afforded, and the time required for the consumption of equal volumes.

In trying the former, I have, in the following experiments, invariably had recourse to the method mentioned in my former papers, viz., the condensation by chlorine, and in which, now that the trials have been greatly extended, I place the utmost confidence.

For ascertaining the latter, I have followed the usual method,an accurately-adjusted experimental metre; by which the times. required for the consumption of equal volumes, burning under similar circumstances, and also the quantity consumed in equal times, were easily determined. The jet burner was the same in all the trials.

I consider both of these circumstances absolutely necessary, for, though some have insisted only on the one, and others on the other only, yet, unless both be taken into account, we do not

• From Jameson's Jour., Edinburg, vol. xlv, pp. 37-49.

arrive at the true value of the gases, and, consequently, cannot compare one with another for the purpose of illumination. Thus, if two gases afford, by their combustion, from the same or similar burners, with the same height of flame, the same light; but if a foot of the one lasts an hour, and a foot of the other an hour and a-half, then the latter is one-half more in value than the former for yielding light, because it is giving the same light for one-half more time; or, which is the same thing, one-half more of the former must be used so as to complete the time which the latter will burn. This, I regret to say, has been too often overlooked by many in estimating the value of coal-gas.

As the chief object I had in view was, not the comparison of the light afforded by coal-gas, or its expense, as compared with other sources of light; but merely the value of the gases as compared with one another, when obtained from different coals, I shall commence with that from English caking coal, and take it as the unit for comparison.

English Caking-Coal Gas.-The gas from this kind of coal, on which my experiments were made, was that from Newcastle; others were also procured by means of an experimental apparatus, fitted up expressly for the purpose. The condensation by chlorine in the former, was, on an average of several trials, 4.33 per cent.

The specific gravity, at Th. 60, B. 30, was 420.

The durability, with a four-inch flame, from a platinum jet, of an inch in diameter, was 1 cubic foot in 50 minutes 30 seconds. The pressure by water-gauge at the burner was of an inch.

From 1 ton of coal, about 8000 cubic feet of gas are obtained. The gases obtained with my experimental apparatus, from a variety of samples of the same kind of coal, both lately and several years ago, were very nearly of the same composition. Different heats were used in driving off the gas, with the view of finding the best heat. The condensation by chlorine varied from 3.5 to 55; the average of the trials, amounting to eight, was nearly 5. The durability varied from 47' 20" to 53' 30"; the average being 50′ 25′′.

The average specific gravity of eight different gases was 464, the highest being 512, the lowest 414.

As above mentioned, I take the gas from this kind of coal for illuminating power and durability, and, consequently, for value, as my standard of comparison.

English Cannel-Coal Gas.-The gas obtained from this kind of coal, such as that from Wigan in Lancashire, with which Liverpool, Salford, and other places are supplied, and that from coals found in different parts of Yorkshire, which are occasionally used at Manchester, are very nearly of the same quality.