the quarter of an hour which has elapsed since the taking of No. 4 sample, a large proportion of its carbon, equal to 20 per cent. of its weight, whilst the silicium, on the contrary, has remained nearly stationary.

6th Sample, taken out at 1h 40m P.M.

The reason why this sample was taken only five minutes after the last sample, was, that the mass in the furnace was rapidly transforming itself into two distinct products, viz. the scoria on the one hand, and small globules of malleable iron on the other. We attached some importance to this sample, as the workman was on the point of beginning the balling or agglomerating the globules of iron, so as to form large balls of about 80 lbs. weight, to be hammered and rolled out into bars. Whilst the mass taken out for analysis was cooling, small blue flames of oxide of carbon issued from it. These were similar to those observed in Nos. 4 and 5, but were not so abundant. The appearance of this sample was very similar to the last one, with the exception that the scoria was not so intimately mixed with the globules of iron, and that these were larger, and slightly welded together when hammered. The proportions of carbon and silicium were as follows:

When these figures are compared with those of the previous analysis, it is interesting to observe, that whilst the silicium remains nearly stationary, the carbon rapidly diminishes; for in the five minutes which elapsed between the taking out of the two samples, there was 28 per cent. of the carbon burnt out. This rapid decrease of carbon in the iron is maintained during the remaining ten minutes of puddling. In fact, in one quarter of an hour, viz. from 1h 35m to 1h 50m, the iron lost 50 per cent. of the carbon which it contained at 1h 35m.

7th Sample, taken out at 1h 45m

This sample was obtained when the puddler had began to ball. The appearance of the sample, although similar to the last, differs from it by the granules being rather larger, and nearly separated from the scoria, which forms a layer at the top and bottom of the mass. These granules are also much more malleable, for they are easily flattened under the hammer. This

last fact is easily accounted for by the small amount of carbon which it contains, as stated above and shown by these results:

8th Sample, taken out at 1h 50m.

This last sample was taken a few minutes before the balls were ready to be removed from the furnace, to be placed under the hammer, and was a part of one of the balls which were separated and placed to cool. It was observed that no blue flame issued from the mass as it cooled. The appearance of the sample showed that the mass constituting the ball was still spongy, and granulated similar to the previous ones. The only difference was, that the granules adhered together sufficiently to require a certain amount of force to separate one from the other, and also that they were much more malleable under the hammer. They were found to contain the following quantities of carbon and silicium per cent.:

We should observe here, that the black coating which covers the granules of iron, even of No. 8 sample, preserves the iron from all oxidation; for none of the samples became oxidized during the nine months they were in the laboratory, exposed to the atmosphere, and to the various acid fumes floating about. This black coating is probably composed of a saline oxide of iron.

9th Sample.-Puddled Bar.

The balls taken out of the furnace were hammered, and then rolled into bars, and in these we found the following:

The puddled bars were cut into billets of about 4 feet in length, and heated in a furnace to a white heat, and then rolled

into wire iron. The proportions of carbon, silicium, sulphur and phosphorus, were as follows:

To complete the series of products in the conversion of pig iron into wrought iron, we analysed the scoria or slag which remained in the furnace after the balls had been taken out, and found its composition to be as follows:

Therefore in the scoria are found the silicium, phosphorus, sulphur and manganese which existed in the pig iron; and probably the phosphorus and silicium are removed from the iron by their forming fusible compounds with its oxide.

We shall conclude this paper by giving our results in a tabulated form, so that the removal of the carbon and silicium may be better appreciated by those who may consult it with the view of obtaining such information as may lead them to those improvements to which we think our investigations tend.

Finally, we wish to express to Mr. Siméon Stoikowitsch our best thanks for the ability and perseverance which he has shown in helping us in these long and tedious analyses.

XX. Researches in Statical Electricity.
By Sir W. SNOW HARRIS, F.R.S.

[Continued from p. 100.]

Charging of the proof-plane and other insulated conductors.
Coulomb's experiment with a hollow globe and circular plate of twice

the diameter.

Expression representing electrical charge.

23. QUESTION has not unfrequently arisen in regard to the charging of the proof-plane and simple electrified conductors, which admits of an easy solution upon the elementary principles and subsequent experimental inquiries we have been considering. It has been doubted whether these bodies take up electricity upon all their surfaces, or upon one only. Now it is evident, (3) and Exp. 4, that neither the proof-plane nor any other conductor can take up electricity upon either face, except we displace from the face of contact a portion of its own electricity (3). The charge virtually consists of its own displaced electricity (9), consequently the charge it receives will be entirely dependent on this induced change, as we have already seen (17). In the case of an insulated conducting surface exposed to the operation of surrounding matter, a stratum of accumulated electricity must always be found upon all its surfaces upon the principles already exposed (10). Let, for example, a, b, c, fig. 18, be three concentric hollow spheres placed one within the other, and so sustained as not anywhere to touch; let the interior middle sphere be perfectly insulated, whilst the spheres a and c communicate with the ground; then in communicating electricity to the insulated sphere b, we find it charge on both its surfaces, and so produce an electrical stratum, bc, on each side of it. If we cut off the influence of the sphere a or c, then it charges on one surface only; at least the charge on the opposite surface is so small as to admit of being neglected in such an experiment. In the case of the double induction, twice the quantity of electricity may be accumulated under the same degree of the electrometer. The following experiments are instructive, and very conclusive.

Exp. 14. Let a light circular disc of gilded or silvered wood, c, fig. 19, about a foot in diameter and something less than 25 of an inch thick, be suspended from a varnished glass arm by an insulating thread of varnished silk gut (2); connect it with the hydrostatic electrometer E, or with the fixed ball of the balance, as in fig. 17 (22). Let two other perfectly similar plates, a, b, be suspended in like manner at an equal distance upon each side of the plate c.

Connect one of the external plates, a, with the ground, and proceed to charge the centre plate c with a given number of

measures of electricity, the opposite plate b being turned aside during this process. Observe now the electrometer intensity in degrees at a given distance of the attracting discs p, n. Suppose the number of measures 5; distance of the plates a, c 5 of an inch, distance of the attracting discs p, n 6 of an inch, and the index indication 10 degrees.

Replace plate b, and put it also in connexion with the ground. Discharge the air between e and a by a communication between the two coatings or plates a, c in the usual way, and repeat the former process; the number of measures now requisite to produce 10 degrees of force will be 10 measures, or just double the former. The plate c, therefore, or rather the air, has now charged in proportion to the two surfaces of the plate, which in this case have been each exposed to a free inductive action, and the electricity proper to the plate c has been enabled to retire, as it were, outward toward the stratum of air on each side of it.

24. This is precisely what happens in the case of what we term the charging simple insulated conductors, except that from the distance of other conducting matter taken as the opposite or uninsulated coating to the charge, the accumulation is less complete and more feeble. If we look at this question somewhat critically, we may probably be led to conclude that the essence of the charge is virtually the displaced electricity proper to the plate itself (9), (17). Suppose, for example, that we are about to communicate a given measure of electricity to the rectangular plate adb, fig. 20, through the intervention of a small carrier-ball c, charged to saturation, suppose positively, and applied to any point c of the plate. The first action will be that of direct and reflected induction, already explained (3), fig. 2; that is to say, the electricity proper to the plate recedes in every direction, ca, cd, cb, from the point c, and a powerful negative space, c, is for a moment produced all around it; at the same instant, the charge, before distributed uniformly over the carrier-ball, concentrates itself by the reflected induction immediately upon the point c in contact with the plate; a neutralization of the opposite forces ensues, and the plate and surrounding air remain charged with the displaced electricity. If the magnitude of the plate be considerable in respect of the carrier-ball, all its charge, or nearly all, will have disappeared on a first contact; and the charge communicated as in the Leyden jar will be so greatly masked by the inductive action of surrounding matter, that comparatively little effect is produced on an electrometer, n, placed in connexion with the plate. On a repetition of this operation, the same actions are apparent another quantity of electricity becomes displaced, and again the carrier-ball becomes robbed of its electricity. The Phil. Mag. S. 4. Vol. 14. No. 92. Sept. 1857.

N