« PoprzedniaDalej »
Copper ball 400° C., distant
Copper ball 400° C., distant
140 millims. 280 millims. 140 millims.
off the scale
or consisting of polished metal) of the body on which radiation falls materially influence the movements.
109. The accompanying table gives the results of numerous experiments as to the effect of screens, tried with an exceedingly delicate apparatus, constructed as above desribed, the window, c' (fig. 7), being of quartz. The candle used was the kind employed. in gas photometry, and defined by Act of Parliament as a 66 sperm candle of 6 to the pound, burning at the rate of 120 grs. per hour." The distances were taken from the front surface of the pith when the luminous index stood at zero. They were in the proportion of 1 to 2 (140 to 280 millims.) to enable me to see if the action would follow the law of inverse squares and be four times as great at the half distance. No such proportion can, however, be seen in the results, the radiant source possibly being too close to allow the rays to fall as if from a point. The figures given are the means of a great many fairly concordant observations. Where a dash rule is put I have tried no experiment. The cipher o° shows that experiments were actually tried but with no result.
The sensitiveness of my apparatus to heat-rays appears to be greater than that of any ordinary thermopile and galvanometer. Thus I can detect no current in the thermopile when obscure rays from copper at 100° C. fall on it through glass; and Melloni gives a similar result. (To be continued).
Mean..7'47 The average time of the first half oscillation is therefore 747 seconds, and of the second half 7.3 seconds. This small difference is not unlikely to be due to errors of observation.
After a long series of experiments the zero gradually creeps up, showing that one side of the apparatus is becoming warmed. The conducting power for heat and condition of the surface (whether coated with lampblack
By referring to paragraphs 106 and 107 it will be seen that I have put the time of the first half oscillation as 11'5 seconds. This was with another apparatus, having a glass thread of different torsion.
ESTIMATION OF POTASSIUM AS ACID
By P. CASAMAJOR.
(Concluded from p. 233.)
If we should be in entire ignorance of the quantity of potassium in a compound to analyse, we should take a large quantity of tartaric acid, not more, however, than three times as much as the quantity of material weighed for analysis, as the monosulphide, which is the compound having the greatest percentage of potassium, has 71 per cent, which, multiplied by 4, gives 2.84. The next in order, potassic hydrate, has nearly, but not quite, 70 per
To return to the 10 c.c. of decinormal solution, we may note that they contain 47 centigrms. of potassa, corres
* Read before the American Chemical Society, September 7, 1876.
Dec. 8, 1876.
Estimation of Potassium as Acid Tartrate.
ponding to 39 centigrms. of potassium. We may then weigh approximately 2 grms. of tartaric acid, which is a little less than six times the quantity of potassium to be precipitated. This acid is dissolved, and added to our Io c.c. of decinormal solution. The liquid should now be stirred sufficiently to make a thorough mixture of the solutions of potassa and tartaric acid. The crystals begin to deposit almost immediately, and the deposition increases for about five or six minutes, when it stops, and the liquid clears up. Alcohol should now be added to increase the precipitate. This addition, however, requires a few words of explanation.
As the result of numerous estimations of potassium in compounds of various kinds, I have found it advantageous to add, at first, only a small quantity of alcohol, a volume about one-tenth of the liquid in the beaker-glass. After this addition the liquid should be stirred sufficiently to effect a thorough mixture, and then be allowed to rest five or six minutes, when the liquid above the precipitate becomes clear once more. Finally, the rest of the alcohol should be added, a quantity sufficient to make the whole liquid in the beaker-glass contain at least 60 per cent of alcohol in volume. The liquid in the beaker should be stirred up once more, and after becoming clear it should be thrown on a filter.
To ensure in an easy manner a quantity of alcohol equal to 60 per cent of the total volume of liquid, I mark on the side of a beaker-glass a line corresponding to 50 c.c., which is easily done with a file, previously moistened with petroleum or spirits of turpentine to prevent the abrasion from cracking the glass. 50 c.c. are quite sufficient when we take I grm. of material for analysis. The volume of the beaker-glass should be at least 200 c.c. This volume of 50 c.c. is for the solution in water before adding alcohol. In the case we have on hand, we have dropped 10 c.c. of decinormal solution of potassa in a glass, and added a solution of tartaric acid. We may now add water up to the mark indicating 50 c.c.
On the other hand, we have strong alcohol-say, of 93 per cent-which is the strongest common alcohol found in the market, and which is sold under the name of 95 per cent alcohol. I have no intention of giving rules for mixtures of alcohol and water, which are familiar to most chemises. In this case I will call your attention to this-that if you add 100 c.c. of 93 per cent alcohol to the 50 c.c. of liquid in the beaker-glass, the result will be 150 c.c., and if we divide 93 by 150 the result, 62:3, will be the strength of alcohol required.
the 100 c.c.
After adding about 10 c.c. of strong alcohol to the 50 c.c. of solution in our beaker-glass, we finally add the rest of After the deposition of crystals has stopped, the contents of the beaker are thrown on a filter.* The liquid that filters through gives a distinct red colour to litmus paper. The precipitate on the filter should now be washed with alcohol of 60 per cent until the filtered liquid ceases to show a red colour with litmus paper. The precipitate after this is ready to be washed down into a beaker-glass to be tested with potassa, after the liquid in the glass has been sufficiently heated and coloured with litmus. The glass containing cream of tartar in water is placed under a burette, and, if the operation has been carefully conducted, it will take exactly 10 c.c. of the decinormal potassa solution to turn the liquid in the beakerglass from red to blue.
The condition of a solution containing only potassa and water is one that very rarely, if ever, presents itself in chemical analysis, and we have in the next place to ascertain the influence of bodies which are usually found in combination, or in a state of mixture, with potassium.
If we drop 10 c.c. of a decinormal solution of potassa in a glass, and add a few drops of solution of litmus, we will be able to find the quantity of sulphuric acid, added drop by drop, which will neutralise the 10 c.c of potassa. *When soda is present in the solution it is expedient not to delay too much in throwing the precipitate on a filter to avoid errors in the result. I propose in a future communication to examine this question.
After doing this, if we add as before 2 grms. of tartaric acid dissolved in water, a very slight precipitate will be obtained, even after standing for hours, and however much the liquid may be stirred, or whatever quantity of alcohol we may add, the precipitate does not increase perceptibly. If, instead of stopping at neutrality, a sufficient excess of sulphuric acid is added, tartaric acid will not show the least turbidity after continued agitation and addition of large quantities of alcohol. Hydrochloric acid in the same circumstances behaves exactly in the same manner, as is also the case with nitric and phosphoric acids. From the behaviour of potassic bromide and iodide, when in presence of an excess of tartaric acid, we must conclude that hydrobromic and hydroiodic acids belong to the same category.
With all these acids, a quantity sufficient for neutralisation of the potassa gives a slight precipitate, while an excess prevents precipitation. In the first case, the precipitate produced can only take place by liberating a quantity of the acid in combination, and after a sufficient quantity of free acid has been formed further deposition is prevented.
The acids experimented on were powerful mineral acids, whose affinities for potassium are so great that, although the acid tartrate is more insoluble than any of their potassic compounds, they only yield a small portion of potassium to tartaric acid. If, therefore, a weaker acid than the tartaric was chosen to combine with potassium, it would not prevent the production of an abundant deposit of acid tartrate. Acetic acid naturally suggested itself, and, on being tried, was found incapable of preventing this precipitation. Here, then, was our way out of the difficulty.
Before describing the manner in which this property of acetic acid was utilised, we must, for the better understanding of the subject, state that salts of sodium in a solution containing 60 per cent of alcohol do not prevent the precipitation of cream of tartar. The sulphate, the nitrate, the chloride, iodide, and bromide, the tartrate, and acetate seem equally powerless to prevent the formation of the precipitate. This is an important point, as by means of soda or its carbonate we may separate the bases that accompany potassium and ammonia, whose acid tartrate is very insoluble, and may in presence of soda be driven off by heat.
The property that acetic acid posseses, of allowing the complete deposition of cream of tartar to take place, suggusted at first the following process :-Given a compound containing potassium, phosphoric acid, if present, would be separated as ammonio-magnesic, as tricalcic, or in any other convenient phosphate. The volatile acids could be driven off by excess of sulphuric acid and heat until fumes of sulphuric acid began to appear. Sulphuric acid could afterwards be precipitated with acetate of barium, thus leaving acetic acid as the only acid in the solution, in combination with all the bases.
This process is simple in theory, but long, and altogether detestable in practice. An analysis was already begun on this plan, when another, much more simple and convenient, suggested itself, which gave on trial the most satisfactory results. This process consists in adding to the compound to be analysed, if it contains a strong mineral acid, a certain quantity of acetate of sodium and, afterwards, tartaric acid. The effect of adding acetate of sodium is that if a strong mineral acid is in excess it forms a sodium salt by acting on the acetate, and liberates a corresponding quantity of acetic acid. When tartaric acid is afterwards added, and a quantity of acid tartrate is precipitated, the strong mineral acid set free reacts on the acetate, and acetic acid is again liberated. This action goes on until all the potassium has been precipitated as acid tartrate, and all the strong mineral acids originally combined with potassium have been combined with sodium, and a corresponding quantity of acetic acid has been set free.
The quantity of acetate of sodium that I usually add is equal to the quantity of tartaric acid. The theoretical
Estimation of Potassium as Acid Tartrate.
CHEMICAL NEWS, Dec. 8, 1876.
quantity necessary is 55 per cent of the quantity of tartaric, but another part, although converted into cream of tartar, acid, but I always obtain excellent results by using equal quantities; the excess does no harm.
This action of acetate of sodium, in promoting the precipitation of cream of tartar, is one of importance in testing for potassium qualitatively. In our chemical books we find directions for precipitating potassium from its compounds by means of tartaric acid, as if it was a difficult and delicate operation. An addition of acetate of sodium in conjunction with tartaric acid, and a discreet use of alcohol, will give indications of potassium in a few minutes, even when present in small quantities.
To give an example of analysis of potassium, let us take a sample of chloride, and weigh I grm. This is dissolved, and 2 grms. of acetate of sodium are added and dissolved. We now dissolve 2 grms. of tartaric acid, and add them to the niixture of potassic chloride and acetate of sodium. We note the total volume of liquid, and, after the deposition of acid tartrate has stopped, we add about one-tenth as much of alcohol of 931 per cent, when a further deposition takes place. Afterwards, the remaining quantity of strong alcohol is added, which must be such that the total volume of alcohol is double that of the ori
ginal aqueous solution. The precipitated cream of tartar obtained is tested, exactly as before, by a decinormal solution of potassa.
Instead of estimating the precipitate of acid tartrate volumetrically, it may be dried at 100° C., and weighed. The volumetric analysis is, however, preferable, apart from its rapidity and convenience, because by the action of alcohol some compounds, such as sulphates, may in certain cases be precipitated, but as they have not an acid reaction their presence would not interfere with the estimation by a titrated alkaline solution.
To enable me to verify the accuracy of the results obtained by this process of analysis, I have in all cases taken a stated volume of titrated solution of pure potassa, and I have added sulphuric, nitric, and hydrochloric acids, and varying quantities of sodium salts. The potassa dissolved to form a.titrated solution was Tromsdorff's potassa by alcohol, containing no soda. The acids leave no residue by evaporation, and can therefore contain no potassium.
By saturating 10 c.c. of potassa solution with sulphuric, hydrochloric, and nitric acids, the following results were obtained. The numbers represent the c.c. of the same potassa solution, which saturated the acid tartrate precipitated in each case :
will remain in solution. If the precipitate is thrown on a filter without the addition of alcohol, and if we free it from excess of tartaric acid by washing with water, a further loss will take place. If, instead of using pure water to dissolve the tartaric acid added to our 10 c.c. of potassa solution, and to wash the precipitate of acid tartrate left on the filter, we use a liquid incapable of dissolving cream of tartar the loss would be very much reduced. This we can easily do by using a saturated solution of acid tartrate. When the impurities on the filter have been removed by washing with this saturated solution, the last portion of this solution may itself be removed by washing with a small quantity of alcohol of 60 per cent. There will then remain one cause of error, due to the water introduced with the 10 c.c. of titrated potassa solution, which may, with sufficient approximation, be considered as 10 c.c. of water. To estimate this quantity, we must use the table of Chancel already given, and if we suppose that the temperature of the liquid is 25° C., we shall find that at that temperature 100 c.c. of water will dissolve 67 centigrms. of acid tartrate, amd therefore Io c.c. will dissolve 67 milligrms., corresponding to 16 milligrms. of potassa, and consequently to 0.35 of I c.c. of decinormal solution. If we have operated with care, we will find that we can account for the original 10 c.c. of potassa solution within one-tenth of a c.c.
The results, however, are not nearly so accurate when salts of sodium are mixed with the potassa solution, and there is always a deficiency in the acid tartrate precipitated. I found, however, that a small quantity of alcohol will in great part overcome this difficulty, but by using too much alcohol, as much as 10 per cent, the results are too high. After a great many experiments, I have been led to adopt 3 per cent as the strength of alcohol that gives the best results. The quantities of acid tartrate dissolved by 3 per cent alcohol at various temperatures are as follows:
100 c.c. of Water having 3 per cent of Alcohol will Dissolve Cream of Tartar. Grammes.
In making these experiments it was of the utmost importance that my solution of potassa should be pure, but, in commercial tests, it matters little whether the standard alkaline solution be of potassa or soda, as both neutralise cream of tartar.
The process for the estimation of potassium which I have described is not always advisable, as the strength of alcohol required may in some cases interfere with the results. When a great many tests are to be made, as happens in a factory, and at the same time extreme accuracy is not required, the expense attending the use of so much alcohol may be worth consideration.
For I grm. of substance weighed for analysis 100 c.c. of strong alcohol are required, besides about 50 c.c. more for washing the precipitate. Altogether, we may calculate that the alcohol for every test costs about ro cents. In testing the low products of sugar houses for potassium strong alcohol cannot be used, because they contain substances which become adhesive and unmanageable in presence of strong alcohol.
To explain the process that I have used in such cases, let me take once more 10 c.c. of titrated potassa solution in a beaker-glass, and add enough tartaric acid to precipitate the potassium. A certain portion will be deposited,
This alcohol of 3 per cent, saturated with acid tartrate, is used to dissolve the potassium compound weighed for analysis, to dissolve tartaric acid and acetate of soda, to dilute our solutions, and wash our precipitates. As the temperature of a laboratory need not vary during an operation, no error need result from the quantities of cream of tartar which alcohol of 3 per cent will take up at different temperatures.
Let us take a low sugar-house product to test for potassa, in the shape of a syrup of 42° B. Suppose we take 100 grms. of this, the quantity of water in our weighed sample will be 20'6 grms. To this we add a quantity of alcohol equal to 3 per cent, or 3.3 per cent of alcohol of 931 per cent, which is seven-tenths c.c. We may now dilute our syrup of 42° B. with alcohol of 3 per cent until it is quite thin, and add about 5 grms. of tartaric acid if the low product is from cane-sugar, or about 20 grms. if a beet product. We should also add about the same quantity of acetate of sodium, and after allowing deposition to take place for about fifteen minutes, the precipitate may be treated as we have seen in other cases.
After the acid tartrate has been saturated by the titrated potassa solution, we should add to the result obtained by the burette, the quantity of potassa corresponding to the acid tartrate dissolved by the 2016 c.c. of water which
Development of the Chemical Arts.
CHEMICAL NEWS Dec. 8, 1876. accompany the 100 grms. of 42° syrup and the 07 c.c. of alcohol added, which are equal to 213 c.c. of alcohol of 3 per cent. If the temperature of the liquid at the time of filtration is 30° C., we will find in the table I have given that 100 c.c. of alcohol of 3 per cent will, at that temperature, dissolve 60 centigrms, of cream of tartar, and consequently 213 c.c. will dissolve 13 centigrms., which represent 32 milligrms. of potassa, which should be added to the result.
The process based on the use of weak alcohol, saturated with cream of tartar, is of older date than the process I first described. The results obtained are not uniformly satisfactory, for, although they are generally good, sometimes there will be errors of 2 or 3 per cent, which cannot be attributed to any cause that I could discover. These discrepancies induced me to try the other process, in which the solutions are made to contain 60 per cent of alcohol, and this has always given satisfactory results.
DEVELOPMENT OF THE CHEMICAL ARTS
The Sulphur Industry of Sicily.
By Dr. ANGELO BARBAGLIA.
THE true home of sulphur is Sicily, where the deposits extend over a great portion of the island bounded on the south by the mountains Delle Madoni, and comprising almost the whole of the provinces of Caltanisetta and Girgenti as well as a part of Catania as far as Caltagirone, Rammacca, and Centuripe. Besides this there are isolated deposits at Lercara, in the province of Palermo, and at Gibellino, in the province of Trapani. The number of sulphur mines scattered in the above-mentioned provinces is very considerable. According to a statistical conspectus for the year 1872 the number exceeded 250, with a total yearly production of 1,861,700 metric quintals, requiring an outlay of 2,472,935 lire.
According to the recent and highly interesting investigations of the mining engineer, Mottura, published in 1871, Sicilian sulphur is a product of the tertiary formation, and is found in the upper miocene between foliaceous crystalline gypsum and massive limestone (calcinari); its associates are bituminous marl (tufi) and gypsum. The sulphuriferous deposits (veins, courses, beds) vary exceedingly in inclination, thickness, extent, and in richness. In these deposits and on their outer boundary, there is invariably found a granular, friable, whitish rock consisting chiefly of gypsum. The miners of the island name this rock briscale, and suppose that from the purity and thickness which it displays on the surface they can infer the richness and extent of the sulphur deposits. The ores are divided into three groups Real Percentage.
"Berichte über die Entwickelung der Chemischen Industrie Während des Letzten Jahrzehends."
Prospecting for Sulphur.-The existence of sulphur underground may be almost always concluded from characteristic indications on the surface. As such the briscale is especially regarded, and where it crops out to daylight it is as a rule certain to lead to deposits of sulphur. The occurrence of siliceous limestone and of sulphur springs are regarded as favourable indications. The first operation consists in driving strongly sloping adits, known by the native miners as buchi or scaloni. The latter name refers to the circumstance that they are laid out in stairlike flights, which are distinguished as sani and rotti according as they run on in a right line, or turn off at an angle. (To be continued.)
FROM the remotest time burning sulphur has been employed to fumigate and purify infected air, and to destroy fermentative and putrefactive action. There is no agent mere powerful in its effects than this. Unlike chlorine, it not only acts as a disinfectant or destroyer of diseasegerms and of the results of putrefaction, but it is also a powerful preservative agent, and, like carbolic acid, is a preventive of chemical changes in dead organic matter of every kind.
understood, its use is necessarily limited by the difficulty Although the value of sulphurous acid is thoroughly which exists in the way of producing it in a form in which rating it by burning sulphur is cumbrous and very uncerit can be readily applied. The ordinary method of genethere are also many situations in which the process cantain, owing to the difficulty of keeping up the combustion; it is inconvenient and but little under control. The not be carried on at all, and under the best circumstances evolution of the gas from its solution in water is scarcely it may be said that there is no ready, convenient, and more convenient, while it is much less effective; indeed, easily controllable way of producing this valuable agent in use at present; and this is the more remarkable when it is considered what a ready and simple means we really have at hand for this purpose.
Most of the readers of The Lancet are no doubt familiar, at least theoretically, with the substance called bisulphide of carbon. This is a compound of one atom of carbon with two atoms of sulphur (CS2); it is a dense, mobile liquid, heavier than water, and intensely inflammable, burning in the air like spirit of wine. During combustion the constituents of the bisulphide combine with the oxygen of the air, producing sulphurous and carbonic acid gases; but as 100 parts contain, by weight, as much as 84 parts of sulphur, which will give, in burning, 168 of this gas from a given quantity of bisulphide greatly exparts of sulphurous acid, it will be seen that the volume ceeds that of the carbonic acid, and is comparatively very large. Suppose the above quantities to be in grains: as 168 grs. will measure upwards of 245 cubic inches, or 100 cubic inches of sulphurous acid weigh 68.5 grs., the sulphurous acid obtainable from 100 grs. of bisulphide. about one-seventh of a cubic foot, which is the volume of
The bisulphide of carbon can be burned in a common acid and carbonic acid only, in relative proportion to the spirit lamp, and in that case the products are sulphurous atomic composition of the bisulphide, as I have stated; but by a modification of the method of burning, the amount of sulphurous acid produced in a given time can be regulated to any desired extent.
It is a property of the bisulphide of carbon to dissolve in fat oils and hydrocarbon liquids, such as petroleum; so by mixing it with any one of these liquids and burning
Protection of Buildings from Lightning.
the mixture in a properly constructed oil or petroleum lamp, sulphurous acid will be generated with the other usual products of the combustion of such materials, and in proportion to the quantity of bisulphide present in the mixture of combustible liquids; any proportionate quantity of sulphurous acid can in this way be thrown into an atmosphere, and the action may be continued for any length of time.
As the sulphurous gas is generated pari passu during the combustion of the bisulphide, it diffuses itself in the air, which in a short time will become completely impregnated with it. In a room containing about 1300 cubic feet of air it was found that by burning 280 grs. of the bisulphide the atmosphere was so far charged with sul phurous acid that it was impossible to remain in the room for more than a few seconds. In five minutes after the lamp was lighted litmus paper began to be reddened at some distance from it; in ten minutes the air had become very oppressive, and the litmus paper was reddened in the extreme corners of the room; in fifteen minutes the air was so charged with the gas that it could scarcely be breathed, and in twenty minutes it was unbearable. In that time, as I have said, 280 grs. of bisulphide were consumed in a simple single-wick lamp.
Sulphurous acid generated in this manner can be applied with facility to the disinfection of any place or object. In the case of rooms in which infectious or conagious disease has prevailed, it is only necessary to light the lamp and allow it to burn until the atmosphere has become impregnated with the gas to any desired extent, and then to remove or extinguish it just like a common spirit lamp. In the simple form of apparatus which I suggest for this purpose, the lamp is enclosed in a metal case, about 3 inches in diameter and 8 or 9 inches high, furnished with holes near the bottom for the admission of air, and others in the top for the emission of the sulphurThis can be conveniently moved about, and placed, while the lamp is burning, in almost any locality. Receptacles for infected clothing, or the clothes or linen used in connection with disease, or carriages which have conveyed fever or other patients, can we thoroughly purified without difficulty and with very little trouble. For the disinfection of ships, too, the lamp is particularly suitable, as it can be carried into the remotest part of a ship and burned without the least danger, and with the certainty of effecting its object completely.
It must be observed that the bisulphide of carbon is extremely volatile, having its boiling-point as low as 110° F.; t is therefore necessary that the lamp in which it is burned should be furnished with a well-fitting screw-cap, to prevent the liquid from evaporating, and at the same time to keep its peculiar odour from escaping. This odour is often very nauseous, but the bisulphide is now manufactured by Messrs. C. Price and Co., of Thames Street, so pure, that it possesses very little smell, and can be used without the least inconvenience.-The Lancet. Printing-House Square.
PROCEEDINGS OF SOCIETIES.
PHYSICAL SOCIETY. November 2nd, 1876.
Professor G. C. FOSTER, F.R.S., President, in the Chair.
THE following candidate was elected a member of the Society:-G. Waldemar von Tunzelmann.
M. JANSSEN made a brief communication, in French, with reference to a method which he has proposed to the Académie des Sciences for ascertaining whether planets really exist between Mercury and the Sun. After mentioning the importance of photography from an astronomical point of view, he explained his reasons for hoping
CHEMICAL NEWS, Dec. 8, 1876.
that a series of solar photographs-taken regularly at in tervals of about two hours, at a number of places on the earth's surface—would enable us to determine this question which is now agitating the scientific world, since any spots which crossed the sun's disk would be at once registered. As it is necessary that such observations be made at several places and in several countries, M. Janssen hopes that other countries besides France will ere long arrange to have such a series of observations taken, and he considers that in a few years the circumsolar regions would thus be explored with a certainty which could not possibly be attained by any other method. He exhibited some of the original photographs taken in Japan of the transit of Venus, and explained the advantage of placing a grating in the focus of the camera in order to eliminate distortion.
Mr. CROOKES Showed the spectrum of a small specimen of chloride of gallium which he had received from its discoverer, M. Lecoq de Boisbaudran. The discovery of this metal is of peculiar interest, as M. Mendeleef had previously, from theoretical considerations, asserted it to exist, and had also correctly given some of its chemical and physical properties. The most prominent line in the spectrum was a bright line in the blue, somewhat more refrangible than that of indium.
Mr. LODGE briefly described a model he has designed to illustrate flow of electricity, &c., which is fully explained in a paper in the Philosophical Magazine for November, and he showed how similar considerations can be applied in the case of thermo-electric currents. model in its simplest form consists of an endless cord passing over four pulleys, and on one side of the square thus formed it passes through a series of buttons held in their positions by rigid rods or elastic strings, according as they represent layers of a conducting or non-conducting substance. When considered in connection with thermoelectricity the buttons are assumed to oscillate on the cord, and if they move in one direction with greater velocity than in the other, the cord will tend to move in the former direction. Now, at a junction of copper and iron, since the metals have different atomic weights and their kinetic energies are equal, the velocities must differ on each side of the junction, and an unsymmetrical oscillation of the molecules must ensue, analogous to that assumed by Mr. Stoney to take place in Crookes's radiometer, and the cord, or electric current, will advance when two junctions are at different temperatures. Mr. Lodge showed experimentally that for a given difference of temperature the maximum thermo-electric current is obtained when one of the junctions is at 280° C., and beyond this point the amount of deflection decreases. This fact led Sir W. Thomson to discover the convection of heat by electricity; that is, if we have a circuit composed of copper and iron, and one of the junctions is at the above temperature, the current, in passing from hot to cold in the iron or from cold to hot in the copper, absorbs heat. This fact was experimentally illustrated by Mr. Lodge. A strip of tin plate is symmetrically bent so as to nearly touch the two faces of a thermopile, and is heated at the bend by steam passing through a brass tube on one side (not end) of the thermopile, and kept cold by a current of water on the other side. As the arrangement is symmetrical no current is found to pass through the thermopile, but when a powerful voltaic current passes through the strip of metal a reversible deflection of the needle is observed, in accordance with the above law. MANCHESTER LITERARY AND PHILOSOPHICAL
Ordinary Meeting, October 17, 1876.
E. W. BINNEY, F.R.S., F.G.S., President, in the Chair. MR. BAXENDELL drew attention to a paper, "On the Protection of Buildings from Lightning," read by Prof. J. Clerk Maxwell at the late meeting of the British Associa