Finally Professor Norton insists that, though the force of gravity is not sensibly intercepted "this does not prove that a tendency to interception does not exist : " and to prove the existence of this tendency he argues from "the supposition of a wave-transmission of the force." To this I reply that nothing proves the existence of the slightest tendency to the interception of the action of gravity or any other action. Such an interception can moreover be demonstrated to be absurd, as there is no power in nature which is calculated to intercept action. As for the "supposition " (the author seems here as elsewhere to accept suppositions as proofs) I reply that it is only a supposition: which moreover has nothing to do with the exertion of active power; since the propagation of waves is a phenomenon of motion, not of its causality with which alone we are here concerned. But my critic, who employs the indefinite word " force " instead of "action," could not help confounding the exertion of power with the consequent motion and its propagation.

Out of fairness to Professor Norton, I must quote a short passage that I find in one of his articles (Phil. Mag. vol. xxviii. pp. 277, 278) where he gives his reason for admitting the interception of action. He

He says:

If two molecules are in equilibrium under their mutual actions, the attractive and repulsive impulses exerted by each upon the central atom of the other must be equal, and therefore no effective action, either attractive or repulsive, can be transmitted to other most distant particles on the same line. Under these circumstances, one molecule in receiving the action of another, intercepts the action that would otherwise take effect upon other most distant molecules. This being admitted" &c.

I recommend this curious reason to the consideration of physicists, of those especially who deal most with dynamics. As for myself, I have italicized the word therefore, but shall not stop now to give a special answer; as the passage is not taken from the paper to which I intended to reply. On the other hand, the author must himself solve his argument in the case of gravity, which, as he allows, is not sensibly intercepted. To conclude. What I have said suffices, in my opinion, to show that Professor Norton has been unsuccessful both in his answers to my objections and in his arguments against some of my conclusions. The cause of his ill-success lies almost wholly in his too great facility to take a stand upon gratuitous hypotheses and mere conceptions, which he also confounds too frequently with "established truths." This fact is now so evident that I might be allowed to consider his last paper not as an answer to, but as a striking confirmation of, my previous objections.

I cannot end without sincerely thanking the learned author for the honour which he has bestowed on me by condescending to criticise my 'Elements of Molecular Mechanics.' The more so, as his criticism, whilst impugning a few points of secondary importance, contains what to my great satisfaction I consider to be a virtual approbation of my leading principles. Nothing can be more gratifying to the author of a scientific work than to find his fundamental views shared by men of such accomplishments in science as Professor Norton is known to be, whatever unavoidable differences may remain in some matters of detail.

LXI. On a Metallic Connector to replace the Vulcanite Tube used with Bianchi's Air-pump. By WILLIAM SWAN, LL.D., F.R.S.E, &c., Professor of Natural Philosophy in the University of St. Andrews*.

IN

N the air-pumps constructed by M. Barthélemi Bianchi of Paris, which are driven with a continuous motion of rotation, the cylinder C (fig. 1), like that of an oscillating steam-engine, vibrates on trunnions, T, U, through a considerable arc. The trunnion T is pierced to form an air-passage, to which a tube (A)

Communicated by the Author, having been read before the Literary

and Philosophical Society of St. Andrews, April 17, 1869.

is screwed on air-tight. In order that the cylinder shall be free to oscillate, it is evident that the tube A does not admit of being rigidly attached to any fixed piece of apparatus; and accordingly, to connect A with the tube B communicating with the pressure-gauge and air-pump plate, M. Bianchi employs a flexible tube of vulcanite, which is prevented, by a helix of wire inside, from collapsing with the pressure of the atmosphere when the pump is working.

Such a vulcanite connector, when of the quality furnished by M. Bianchi and when new, answers its end most satisfactorily; but, like all other vulcanite tubes, in course of time it becomes brittle, and, especially where it is attached to the metal tubes A and B, it begins to leak. Thus it happens, when an experiment is to be made, that the pump, in every other respect in perfect order, may be found to be useless until the vulcanite has had its ends cut off and is again attached, or until it is replaced by a new tube. I believe all who have been in the habit of using this form of air-pump must have experienced the inconvenience thus arising, and will agree with me in thinking that, although the use of a vulcanite tube may sometimes be convenient, the necessity of employing it always is the only objectionable feature of Bianchi's otherwise most excellent instrument.

In order to remedy this defect, I have recently constructed a metallic connector for the Bianchi air-pump belonging to the Natural-Philosophy Museum of the United College, St. Andrews. This consists of a brass tube, H (fig. 1), about 6 feet long, of an inch bore, and of an inch outside diameter, coiled into a helix of three and a half turns about 6 inches in diameter. To the air-tubes, A of the pump and B of the pressure-gauge and pump-plate, are attached, by soldering, tubes with conical ends. That which is soldered into the air-tube A is carefully centered. It is represented at D (fig. 2). The D are fitted air-tight, by grinding, into hollow conical pieces, E, F, attached by soldering to the ends of the connecting helix H. To prepare the pump for working, the cones D having been slightly oiled, the piece E of the connector is slipped on to the cone D of the pump at A; and the pump-plate and gauge are slid on the table on which they stand towards the pump until the conical tube D at B (not seen in the figure) entering the hollow piece F, takes its seat and slightly compresses the helix H lengthwise. The helix of brass tube, although quite rigid enough from point to point to preserve its shape, yet possesses ample elasticity as well as flexibility to keep the cones D in their places, notwithstanding any jolting motion which may occur, even when the pump, for rapid exhaustion, is driven with the utmost speed. It need scarcely be further explained that the cone D at A turns

freely within the hollow piece E, and thus the cylinder C has perfect liberty to oscillate without necessarily twisting the connecting tube H; although, owing to the unavoidable friction between the rubbing surfaces, at each stroke of the pump a slight torsion of the connector actually does take place. In order to ensure the connexions remaining perfectly air-tight, oil-cups, O, are placed on the pieces E, F, and a groove, G (fig. 2), is cut round each of the cones D, so that when a little oil is dropped into the cups it runs down into the grooves and effectually prevents the passage of air. In the apparatus as constructed, wings, W, w, were added at the ends of the connector, and also placed on the tubes A and B with the intention of stretching vulcanite bands over the wings to maintain the connexions until such time as the atmospheric pressure, owing to the progress of the exhaustion, should of itself be sufficient for that end. The wings have proved convenient in handling the connector; but it has not been found necessary in practice to make use of vulcanite bands. A pair of stops might be attached to the pumpstand immediately below and in contact with the wings, W, so as effectually to hinder their oscillation. Thus, were it deemed desirable, the slight torsion of the tube H, already noticed, might be prevented. It is indeed quite possible that, failing this precaution, in course of time the tube may become brittle by continued twisting. Not only the wings, but, I believe, the oil-cups also, or at least that at B, might be dispensed with; but I am not disposed to advise that either of the cups be omitted.

The apparatus, as described, which has been exceedingly well constructed by Messrs. Kemp of Edinburgh, was designed by me for use with a moveable air-pump which is placed on the floor, while the pump-plate stands on a table. If the air-pump be permanently screwed to a lecture-table, which, where room can be spared, is a good plan, the apparatus might advantageously be simplified. The helix might probably be made of a shorter tube, and the fastening at B might be permanent. A straight tube might even perhaps replace the helix, or a tube bent at right angles, where the table does not admit of the pumpplate standing in front of the pump. In either case sufficient play might be given to keep the cone D in its place while the pump is working, if the pump-plate were mounted on a board with small castors or rollers; and, if necessary, vulcanized bands might be stretched over the wings on the tubes A, E, to draw them together. With a permanently fixed pump, I would also suggest that, by making the helical connector long enough, the connexions at A and B might be made absolutely rigid and permanent, the oscillatory motion of the cylinder being rendered possible simply by the torsion of the helix. Per

haps in this construction it would be necessary to employ a tube of tempered steel, but probably one of brass or copper might suffice.

I will only further add that the moveable connector which I have described has been found to work most satisfactorily. It can be mounted or dismounted in an instant; and if at any time it be desirable to revert to the use of a vulcanite connector, all that is needed is to pull the vulcanite over the cones D on to the tubes A and B.

In conclusion I cannot omit the opportunity now afforded me of warmly recommending, with or without the new connector, Bianchi's most elegant and useful air-pump. Owing to the very great rapidity with which, by its use, a vacuum can be obtained, it is particularly well adapted for lecture-experiments. The pump-cylinder and nearly all the apparatus being made of cast iron, is not liable to suffer corrosion from oil, and will therefore, I have no doubt, last much longer than a brass air-pump. I have found a glass case very effectual in preventing external rusting; and the instrument is so beautifully finished as well to deserve such protection. M. Bianchi has adopted in his air-pumps Babinet's ingenious arrangement for double exhaustion. By turning a stopcock, one end of the barrel of the pump is made to exhaust the other; and thus a very perfect vacuum, as is well known, can be obtained.

LXII. Upon the new Conception of Electrodynamic Phenomena suggested by Gauss. By R. CLAUSIUS*.

IN

Na letter written by Gauss in 1845 to W. Weber, the brief remark occurs that he considered the corner stone of electrodynamics to be the deduction of the accessory forces (which are superadded to the reciprocal action of electrical particles at rest) not from an instantaneous action, but (as is the case with light) from one requiring time for its propagation. Based upon this remark, three very interesting papers, by Riemann, C. Neumann, and Betti, have lately appeared. All three authors arrive by different ways at the result that the forces which two currents exert upon each other are explicable on the assumption that a certain time is necessary for the propagation of electrical actions.

I have read these researches with the greatest interest, but must confess that I am not satisfied with them; and I think that both the well-deserved reputation which these authors possess in the scientific world, and the importance of the subject, amply justify me in stating my objections. The fact that three distin

* Translated from a separate impression, communicated by the Author, from Poggendorff's Annalen for December 1868.