37. Affinity of potassium for oxygen.-Twenty grains of potassium were combined with about ten ounces of mercury. The amalgam was poured into a wooden cup, into the bottom of which a copper wire connected with the galvanometer (5) had been let. At about half an inch above the surface of the amalgam I secured a piece of platinum, also in connexion with the galvanometer. On pouring dilute sulphuric acid into the cup the needle was deflected 74° (= 0o·05 Q) during three successive minutes, but the local action of the amalgam was so vigourous that at the end of this interval of time most of the potassium was dissolved, and the needle declined very fast. On treating 20 grains of zinc in precisely the same manner, I had a deviation of 49° (= 0a·0152 Q). Hence

and whence

But from (35), z = potassium, 4:06.

1.93, and h = 1; therefore k, the affinity of

38. It is necessary, however, to pay attention to the circumstances under which the experiments were made, in order to obtain correct ideas concerning the above intensities of affinity. The increase of the intensity of the voltaic apparatus by heat is by no means great; and as all the experiments were conducted at common temperatures, no regard need be paid to it. But then the intensities of affinity were obtained by comparing currents which had been produced under peculiar circumstances with regard to the condition of the elements of the galvanic arrangements; in one case the hydrogen was evolved in a gaseous state; whilst in the other, the hydrogen, by combining with free and condensed oygen, did not escape. Now we shall see from the following experiments that electric intensity is expended in the act of converting a body into the gaseous state.

39. I took ten glass jars (see Fig. 3), made them perfectly clean and dry*, and placed them in series on a non-conducting substance. Into these I poured a quantity of dilute sulphuric acid, taking care not to wet the glass within an inch of the top of each. Pairs of platinized silver and amalgamated zinc were placed in the jars, and connexions, furnished with the mercury cups 1, 2, 3, &c., were established between them seriatim. A decomposing cell, d, furnished with platinum wires, was connected on one hand with the battery, and on the other with the galvanometer (5). Lastly, I provided a copper wire, w, by means of which connexion could be conveniently made between the galvanometer g, and any of the mercury cups, 1, 2, 3, &c.

40. Into d I poured a small quantity of dilute sulphuric acid. Then, by placing the wire w in each of the mercury cups, beginning

• It is necessary to be very careful in insulating the apparatus, in order to obtain the maximum intensity of a battery. The divided porcelain trough has frequently great conducting powers (particularly when the glaze has been partially destroyed), which render it unfitfor accurate experiments.

at 10 and ending at 10, I observed the deviations of the galvanometer contained in the following table.

41. Now if we divide the straight line, AB, Fig. 4, into ten equal parts, representing pairs on Mr. Smee's plan, and if at each division we erect straight lines, perpendicular to A B and proportional to the comparative quantities of electricity just given, the principles of electric action demand that the line drawn through the extremities of those perpendiculars should be straight. It is in fact so nearly a straight line, that its slight discrepancies there from may be properly referred to unavoidable errors of experiment. Produce the straight line C D so as to meet A B in X, and the straight line A X, equal to 2.8, will indicate the number of pairs necessary to decompose

water.

42. Fig. 5 represents an experiment of the same kind, with a solution of sulphate of oxide of zinc in the decomposing cell. Oxide of zinc was decomposed, the oxygen being evolved at the positive, and the zinc being reduced at the negative electrode. The intensity necessary to decompose oxide of zinc is equal to that of 3.7 of Mr. Smee's pairs.

43. With sulphate of protoxide of iron I did not at first succeed, on account of the formation of peroxide at the positive electrode. However, by placing the negative electrode among some crystals of the salt, pouring water thereon, and suspending the positive electrode in the water, I obviated that difficulty, and obtained the results which are projected in Fig. 6, and which indicates 3.3 pairs as the intensity necessary to decompose protoxide of iron.

=

44. Now from (41) and (42) we have (using the same letters as before) 28 pairs h, and 3-7 pairs = z; whence 2.8 z = = 3.7 h, and z = 1.32 h; or, in other words, the intensity required to separate oxide of zinc into metal and gas is to the intensity required to separate water into its gaseous elements as 1.32: 1. But from (35), the

intensity produced by the union of non-gaseous oxygen with zinc is to the intensity necessary to separate water into non-gaseous oxygen and gaseous hydrogen, as 1.93:1; and 1·32: 1 :: 1·93 + 1·9: 11.9. Wherefore, the intensity necessary to give oxygen the gaseous form is to the intensity necessary to separate water into nongaseous oxygen and gaseous hydrogen as 1·91: 1.

45. Thus we see that a very great intensity of current is employed in changing the condition of bodies, as well as in separating them from their combinations. The field of investigation here opened is very extensive, but I may not at present enter further upon it. I will only remark, that if the intensity necessary to convert a body into a different state, compared with the heat or cold due to the mechanical or other production of that different state, be such as accords with the relations of intensity and heat which we observe in the voltaic apparatus, we have proof that some of the effects which are usually referred to "latent heat," are in fact nothing more than the recondite operations of resistance to electric current.*

46. In our investigation into the cause of the heat of combustion, it will be necessary to deduce our calculations from the electric intensity which is required in order to reduce the product of combustion to the state in which its elements were prior to combustion. The following is a list of these intensities, reckoning the decomposition of water into its gaseous elements as unity.

47. Intensity necessary to decompose oxide of zinc into gaseous oxygen and metal, from (42) and (44), is 3·7 pairs of Smee's battery, or 1.32 h.

48. Intensity necessary to decompose protoxide of iron into gaseous oxygen and metal.—From (43), 3.3 of Smee's pairs = i: and from (41), 2.8 pairs =h; whence 28 i 33 h, or i = 1.18.

49. Intensity necessary to decompose potassa into potassium and gaseous oxygen.-From (144) and (37) we have 1.93 + 1·9: 4:06 + 1·9:: 1·32 h: 2.05 h the intensity required; which may be otherwise expressed by 5.74 of Smee's pairs. Heat evolved by Combustion, when it terminates in the formation of an Electrolyte.

50. Finding that our information on the quantity of heat evolved by the combustion of metals was not very satisfactory, I have without wishing to depreciate the labours of Dulong, Despretz and others, thought it right to bring forward such of my own experiments as are necessary in order to make my investigation complete.

51. I provided two glass jars. The smaller had an internal capacity of 90 cubic inches; and when placed within the other jar, as represented by Fig. 7, the space left between the two was sufficient to contain three pounds of water. By means of a scale, s, suspended by wire from a thick fold of moistened paper, I was able to introduce a combustible within an atmosphere of oxygen, and by means

Some experiments, which I have not time to refer to at present, render this hypothesis more than probable.

of a heavy weight I could keep the paper, sufficiently close to the top of the jar to prevent the escape of any considerable quantity of heated air, while at the same time it was not so tight as to prevent the admission of air as the oxygen was consumed. The increase of the temperature of the water was measured by a thermometer of great sensibility.

52. The heat evolved by the combustion of zinc was ascertained in the following manner. The smaller jar was filled with oxygen, placed in the other jar, and surrounded by three pounds of water, the heat of which was contrived to be as much below the temperature of the surrounding air as it was expected to exceed it at the close of the experiment. A piece of phosphorus, weighing 0.4 grain, was then put into the scale, aud over it I placed a heap of fine zinc turnings, weighing 50 grains. I now ignited the phosphorus, and plunged the scale into the inner jar. After the combustion had terminated, and the heat thereby evolved had been evenly distributed throughout the water by stirring, the increase of temperature was noted. The contents of the scale were then thrown into dilute sulphuric acid, and the volume of hydrogen thereby evolved indicated the quantity (generally about 15 grains) which had not been burnt. Two-tenths of a degree of heat were deducted from the observed heat, on account of the phosphorus, and an allowance having been made, on account of the capacity of glass for heat, the results were reduced to the standard of one pound of water.

53. The mean of several experiments conducted in the above manner, showed that the heat evolved by the combustion of 32.3 grains of zinc is able to increase the temperature of a pound of water by 10°.8.

54. The heat evolved by the combustion of iron was ascertained in a similar way. The iron was in the state of fine wire, and that portion of it that was not burnt was carefully collected, weighed, and deducted from the original quantity. The mean of several trials indicated that 28 grains could increase the temperature of a pound of water by 9°·48.

55. Heat evolved by the combustion of potassium.-This metal, in pretty large lumps, was introduced into an atmosphere composed of equal bulks of oxygen and air. 1 then introduced a stout iron wire, sharpened at the end, into the jar, and with it I cut the potassium into small pieces. Under this treatment it soon became so soft, that every time the rod was lifted it would draw out a string of metal. In this state it often ignited, and the experiment was spoiled on account of the partial formation of peroxide. However, by careful management, I succeeded in making some good experiments, in which nearly all the potassium was converted into potassa; and the exact quantity of unoxidized metal was ascertained by observing the volume of hydrogen evolved when the contents of the scale were exposed to the action of water. The mean of these showed that the heat evolved by the conversion of 40 grains of potassium into potassa is able to increase the temperature of a pound of water by 17°·6.

56. Heat evolved by the combustion of hydrogen.--The gas was burned in an atmosphere of oxygen, diluted with common air, by means of a jet furnished with a very narrow bore. A grain of hydrogen evolved as much heat as is able to increase the temperature of a pound of water by 8°36.

57. We shall now proceed to examine how far the theory of resistance to electric conduction agrees with the above experimental results.

58. We have seen (47), (48), and (49), that the intensities of the affinities which unite gaseous oxygen with zinc, iron, potassium and gaseous hydrogen, are as 1:32, 118, 2·05 and 1; and the proportional quantities of heat which were generated by the combustion of the equivalents of these bodies, are 10°-8, 9°-48, 17°-6, and 8°-36, or 1-29, 113, 2-105 and 1, a ratio which is very nearly the same as that of the intensities just given. Hence we see that the quantities of heat which are evolved by the combustion of the equivalents of bodies are proportional to the intensities of their affinities for oxygen. Now I proved in my former paper* that a similar law obtains in the voltaic apparatus, in consequence of its heat being produced by resistance to conduction. And hence we have an argument that the heat of combustion has the same origin.

59. But our proof of the real character of the heat of combustion is rendered more complete by regarding quantities as well as ratios of heat. From the quantity of heat generated by the motion of a given current along a wire of known resistance, we can deduce the quantities of heat which, according to the theory of resistance to electric conduction, ought to be produced by the combustion of bodies; and then these theoretical deductions may be compared with the results of experiment.

60. The mean of three careful experiments detailed in my former papert, shows that if a wire, the resistance of which is an unit, be traversed by an electric current of 1°.88 Q‡ for one hour, the heat evolved by that wire will be able to increase the temperature of a pound of water by 15°-12. Now I have ascertained experimentally, that a pair consisting of amalgamated zinc and platinized silver, excited by dilute sulphuric acid, is able to propel a current of 0°.168 Q against the whole resistance of the circuit, when that resistance is 5.2; consequently, a similar pair can propel a current of 0°·168 Q × 5·2 = 0°·874 Q against the resistance which I have called an unit. But from (42) the intensity necessary to separate oxide of zinc into zinc and gaseous oxygen, is to the intensity of one of Smee's pairs as 37:1; consequently, the electricity produced by the union of zinc and gaseous oxygen must be sufficiently intense

⚫ Philosophical Magazine, October 1841, S. 3, vol. xix, p. 275. (70.) + Ibid. p. 266.

I beg to remind the reader that my degree, expressed thus (1°Q), indicates that quantity of current electricity which, after passing constantly during one hour, is found to have electrolized a chemical equivalent expressed in grains; as, 9 grains of water, 36 grains of protoxide of iron, &c.