We now come to the consideration of the most delicate part of the battery-the conducting plate. The function of this plate is solely the elimination of the hydrogen of the electrolyte. The plate, indeed, does conduct the positive electricity from the electrolyte. But by the chemical theory of the voltaic generation, the elimination of the hydrogen from the electrolyte is the conduction of the positive electricity; and the conduction of the positive electricity is the elimination of the hydrogen-the one is inseparable from the other, conduction and decomposition being identical in an electrolyte.

By this it will appear that the liberated hydrogen should not be suffered to adhere to the conducting plate, for the gas being a very bad conductor, it will resist conduction, and consequently chemical action.

The office of this part of the battery has generally been considered a mystery, being subject to very great and sudden changes, which being unaccountable by mere inspection, were attributed to occult power called the "electro-motive force." But certainly rational investigation can refer all the battery changes to known forces. By late investigations we know that the gases adhere to, and condense on, all solid bodies. If the attraction between a sheet of metal and a molecule of gas is regulated by those same laws which govern the planetary masses, then we can easily conceive that the more definite the plane of attraction, the more strongly will the molecule be drawn, and consequently the denser and higher will be the adhering layer of gas. And, by the same laws, we can see that an indefinite plane will act with opposing forces, as we see in the deviation of the plummet in the vicinity of mountains; as we see in the perturbation of the planets; and as we suppose when a body descends toward the centre of the earth from the surface. The former of these conditions is fulfilled when the surface is polished; the latter when it is rough or unpolished.

We can conceive the mounts of the roughness to rise so high above the plane of maximum force, that the adhesion to the prominences will be almost destroyed. And we can conceive the mounts to react so much against the pits that the plane of attraction shall be nearly or quite destroyed. We can conceive of a surface thrown into such fine points and recesses that a molecule of gas might float in equilibrium in the cavities, or adhere with the least determination to the prominences. Hence the great advantage of the deposit of finely divided metal.

A surface atomically rough will hold only an atmosphere of the least possible height and density. But this is not attainable. We get the nearest approximation to an atomic roughness when a surface has been covered with amorphous metal by electro deposition. Then it may be said to be rough or unpolished to the greatest degree for that metal.

Could we view such a surface, or rather I should say want of surface, we doubtless should find it many thousand fold more rugged, uneven, and porous, than the common sponge.

The various metals let go the hydrogen in the voltaic circuit with very different degrees of readiness. From my observations, I conclude that the attraction of the various metals for the gas is directly as their specific gravities. All the less dense metals decompose water, (evolve hydrogen). Sodium and potassium evolve the gas in torrents. The base metals proper have less action on water, and a stronger attraction for the gas. The noble metals hold the gas very firmly, and are without action on

water.

The order in respect to evolution is the reverse in respect to gravity; and the order in respect to gravity is consequently the reverse of the order of fitness for a conducting plate in respect to the evolution of hydrogen. Platinum, gold, and lead, hold the gas very hard. When polished plates of these metals are used, the hydrogen adheres in large bubbles, which very slowly creep up the plate. Mercury I do not compare, because its mechauical form is the best possible for adhesion; but could we but polish the solid noble metals as perfectly as the atomic polish of the mercury, I have no doubt but that the mercury, according to its density, would follow after gold. Silver answers better than the other noble metals. Experiment has not enabled me to decide that copper is better than silver, but I am much inclined to consider the copper as best. Iron is decidedly better than any metal above it in density, and requires no special preparation to make it evolve freely. Zinc is so prone to evolution that it is with difficulty that the hydrogen can be made to adhere. The metals of the alkalies cannot be invested with hydrogen like the denser. A mere particle of zinc will coat a surface of copper or iron with hydrogen, and protect it from oxydation forever; but as soon as potassium or sodium is deposited, it is instantly recombined with oxygen, because it cannot be coated with hydrogen. Here I may remark that the newly reported aluminum which is said to have the nobility of silver, with the density of only 2.5, ought, by the above views, to make a most admirable conducting plate.

By the above view, the adhesion of hydrogen is very nearly the reverse of the affinity for oxygen. Here we find silver with a medium adhesion and a low affinity. This at once indicates that it is the metal which will be generally used for making batteries. Iron, which is the most oxydable metal that can be employed for conducting plates, has a very low adhesion, and fortunately a mechanical advantage from its ever-fibrous or granular form, which greatly increases its fitness for evolution. Could it remain as iron in the battery, it would probably be all we

should ever desire. Yet though it acts vigorously when newly cleaned, its affinity for oxygen soon makes it worthless. This objection holds not only for iron, but for some kinds of batteries. holds even against silver, and we are sent at last to the more noble metals.

The difference between gold and platinum in respect to the adhesion, and also in respect to the liability to chemical change, is so small as to make the employment of one or the other merely a question of economy. But there is another property-one which quickly determines the preference; this is the capability of being put in the best mechanical form for non-adhesion, or making the closest approximation to atomic roughness.

Of all the metals, platinum has the greatest tendency to the amorphous state, (excepting its relatives, rhodium, iridium, &c.) I do not remember having seen that its crystal has ever yet been determined. Not so with gold, its crystalline tendency is so strong, that it aggregates so much in precipitation, even from extremely dilute solutions, that the deposit has a decidedly yellowish tinge, and the slightest pressure makes the deposit conglomerate. I here need scarcely remind you of Wollaston's tedious process for metallizing spongy platinum.

If the above views of the nature of the adhesion are correct, then it follows that the surface of the conducting plate should be amorphous platinum, and nothing but amorphous platinum; and consequently, if we wish our battery to retain its capacity to remove the hydrogen from the electrolyte, which, let it be borne in mind, is the capacity to conduct electricity, then there should not be the remotest liability of the amorphous platinum to have a deposit of any other metal, or any oxyd formed on it.

On this consideration I have carefully avoided using any reducible base metal about the battery, in such a way that its salt might get into the cell. I have before shown what would be the consequences of this on the zinc plate, and equally injurious would be its action on the conducting plate, whether it were deposited on it as metal or as oxyd. In either case the hydrogen would adhere, the conduction resistance would be increased, the tension would rise proportionally and react against the affinity; the chemical action, the soul of the battery, would proportionately decline. That the mutations of the battery from adhering coats of hydrogen, metals, or oxyds, on the conducting plate, are to be attributed to conduction resistance, I shall expect will be regarded by the advocates of electro-polar forces as wholly untenable, and the resistance to be considered as incompetent to produce the effect. But that the gas resists is indisputable, and that it adheres to the conducting plate is equally indisputable, for we know that the very dust of the fields attracts and condenses the gases; and is it not, therefore, but as fair an inference that it ad

heres somewhat to dense metallic plates? The thickness of the adhering film may be extremely small, but its resistance may be quite considerable, for the resistance of airs is almost incomparable to that of metals. We know that a battery has penetrated over 3,000 miles of iron wire, and when a battery of 2,000 pairs had the poles parted only the least distance that could be manipulated, then the galvanic action could not be exhibited.

It remains now only to notice the electrolytic changes, with reference to continued action. The generated sulphate of zinc alters the conditions of action, not only by saturating the acid and water, but the dissolved sulphate itself is an electrolyte, and therefore may coat the conducting plate with zinc, and deteriorate it just as was shown would result from the salts of the other base metals. Fortunately, there is not so much danger of the plate becoming wholly coated with zinc as with the other base. metals, for the deposited zinc is rapidly removed by its great tendency to become salt, in which it is assisted by the close proximity of the uncoated portions of conducting plate, forming good local circles with it. Should there be no portions of the plate bare to reduce the counter-tension generated by the resolution of the deposited zinc, then we should have the tension acting against the battery current. This probably can never happen, yet the plate is often made nearly inefficient by the reduced zinc, when the acid is mostly saturated.

The acidulated water or sulphate of hydrogen is electrolyzed by a far less tension than decomposes sulphate of zinc; it is only, therefore, when the quantity of sulphate of hydrogen becomes proportionately small, and causes the tension to rise by its increased resistance, that the sulphate of zinc is decomposed.

But it is unquestionable that that force which is the result of the combination of the elements of sulphate of zinc, cannot of itself undo that combination; yet while the battery is working, zinc is constantly being deposited and re-dissolved. In considering this action of the galvanic current, which is apparently so anomalous to the exhibition of every other known force, I have concluded that we should look for some additional force acting conjointly with the current, rather than for a moment admit the absurdity of an "electro-motive force," with its supposed capacity of acting infinitely without expending itself. Such an additional force I conceive can be found in the attraction of the matter of the conducting plate for the heavy element of the electrolyte. If the conditions under which the deposition of the zinc takes place be considered, it must appear that it is the attraction which makes the determination. In the first place the deposition is nothing when the proportion of sulphate of zinc about the plate is small in comparison to the sulphate of hydrogen; but as the proportion of sulphate of zinc increases, the decomposition of it

begins to show itself, until it becomes very copious in a nearly saturated solution. The supposition I have made is, that the deposition is effected by the conjunction of the attraction with the current or electrical tension; consequently the deposition can only take place when the tension is so high that the addition of the attraction enables it to overcome the affinity. This exactly conforms to the conditions; the good conducting sulphate of hydrogen being removed, the bad conducting solution of zinc will cause the tension to rise. I cannot now go into the discussion of the specific weights of the elements of the two electrolytes, to show that the attraction will act in the same direction with the electric tension. It is at once evident that if we admit that the matter of the plate attracts the elements or atoms-and what physicist at this day would think of denying it ?-then it follows that altering the aggregation of the surface of the plate will diminish that attraction, just as it diminishes the adhesion of the hydrogen; yet, as the molecule of zinc is so much heavier than the hydrogen atom, the disturbed aggregation should extend much deeper into the plate for destroying the attraction for the zinc than is merely required for preventing the adhesion of the gas. On these principles I have made the conducting plate, with the disintegrated state of the surface extended to the greatest depth admitting of the requisite mechanical durability, for which the plate is electro-plated to the beginning of roughness before putting on the coating of platinum.

I have sought to describe the peculiarities of this battery, by exhibiting the actions of the various parts, and the principles which guide me in their construction. These principles, I acknowledge, are new in their application to the galvanic phenom

I have only to say for them, that they are the acknowledged principles of matter and motion, and consequently the principles of universal nature. But it may be that my solutions are wrong, and that further research will not sustain these views; yet, I ask for them a trial as to their conformity with the admitted solutions of the great multitude of natural phenomena. Thus we know that oxygen is condensed with a force of nearly a thousand atmospheres on spongy platinum; and does not geometry show us that if the disintegrated mass attract thus strongly, the solid surface will attract enormously? and if oxygen is so strongly attracted by the solid surface, then why may it not attract hydrogen, which is only sixteen times lighter, sufficiently to condense a layer which the battery liquid cannot displace because it is denser than the liquid? I must here ask that I may not be misunderstood by supposing that I refer to the bubbles of gas which adhere to smooth surfaces by the superincumbent pressure: geometry, indeed, shows us that these bubbles are dispersed by a rough surface, but it also shows that these bubbles