believes that temperature has a far less influence on the discharge from simple orifices than Mr. Isherwood's results would imply. It is difficult to explain to what the higher results obtained by Mr. Isherwood are due; but the conjecture may be hazarded that the orifice in his experiments was very exceptionally placed. It was at the end of a bell-mouthed tube some 10 inches long, a great part of which was only 2 inch diameter; and there was a plug-cock immediately above the orifice. It seems possible that there was a good deal of friction in this pipe, and that the diminution of friction in this part of the apparatus led to the increase of discharge as the temperature increased *.

Thus far the effect of the temperature on the capacity of the reservoir and the size of the orifice has been neglected. It only remains to examine whether the expansion of these has any material influence on the results.

The effect of temperature on the quantities entering into the

D2 equation of flow is twofold. First, the ratio is altered, because the mouthpiece was of brass and the reservoir of cast iron; and the former expands more than the latter. Secondly, the level marks being attached to the side of the cistern, the distance between these marks and the centre of the orifice increases as the temperature rises. There is, however, an uncertainty in applying a correction for the expansion of the metal, because, its external surface being exposed to the air, its mean temperature would be less than the temperature of the water. The following estimate of the correction is therefore approximate only.

Let e, be the expansion of brass per unit of length and per degree;

e, the expansion of cast iron estimated in the same way; T= the excess of temperature during the experiment reckoned from 60°.

It is impossible to calculate, except roughly, the frictional resistance of the tube to which, in Mr. Isherwood's experiments, the orifices were attached. Taking 4 inches length of pipe 3 inch in diameter, and neglecting the bell-mouthed part, we get, using D'Arcy's coefficient of friction, and putting the data in feet:

So that apparently about 12 per cent. of the head may have been lost in the friction of the pipe leading to the orifice.

Then, in consequence of the expansion of the metal, the ratio

D2

of the areas becomes

and the true difference of the square roots of the heads is

The formula of flow, allowing for the alteration of the dimensions by rise of temperature, is therefore

Hence it is obvious that the effect of the expansion of the reservoir and orifice is very small for the range of temperature in these experiments. Allowing for that expansion, we get for the experiments at 190°,

or slightly smaller values of the coefficient than those given above.

It is rather curious that it is stated in Mr. Isherwood's paper that the results are corrected for the variation of the size of the orifice as the temperature varied, but no mention is made of a correction for the size of the reservoir or the expansion of the vessel to which the index-marks denoting the initial and final heads were attached. If these latter corrections have been omitted, though this is difficult to believe, Mr. Isherwood's results should be divided by

where e, is the coefficient of expansion of the material of the reservoir, whatever that was. This would sensibly diminish the apparent increase of discharge at high temperatures given in his experiments.

XXXIX. Action of Alkaline Solution of Permanganate of Potash on certain Gases. By J. A. WANKLYN and W. J. COOPER*.

IN

N continuation of our work on the oxidizing-power of strongly alkaline solution of permanganate of potash, we have made experiments on the common gases, and have arrived at results of some interest. The solution which we have employed in these experiments contained 16 grms. of permanganate of potash and 5 grms. of caustic potash dissolved in a litre of distilled water.

Binoxide of Nitrogen, NO.

This gas was prepared in the usual manner by the action of diluted nitric acid on metallic copper. On submitting it to the above-described solution of alkaline permanganate there was immediate action, the gas being instantly absorbed at ordinary temperatures, and the solution being instantly decolorized and caused to deposit the brown hydrated binoxide of manganese.

110 cubic centims. of NO and 30 cubic centims. of the potash-and-permanganate solution were shaken up together. Immediately the solution lost its colour and deposited the brown binoxide of manganese, and 85 cubic centims. of the gas was absorbed.

The reaction is

NO+ KMnO1 = MnO2+ KNO3.

Protoxide of Nitrogen, N2O.

This gas, prepared in the usual way from nitrate of ammonia, appears to be quite without action on the alkaline solution of permanganate of potash. Even on prolonged heating of the materials in the water-bath there was no sign whatever of action, the permanganate preserving its brilliancy, and the volume of the enclosed gas undergoing no diminution.

Nitrogen Gas.

Experiments published some years ago by one of us show that this gas is not attacked by the alkaline solution of permanganate, even when the temperature is considerably raised. Carbonic Oxide, CO.

This gas, prepared by the action of excess of sulphuric acid on ferrocyanide of potassium, is attacked by the alkaline solu

* Communicated by the Authors.

tion of permanganate. The action is not instantaneous, as in the case of the binoxide of nitrogen, but is comparatively slow.

118 cubic centims. of CO and 30 cubic centims. of the above solution of potash and permanganate of potash were sealed up and heated in the water-bath, being frequently taken out of the bath, cooled, and shaken. Altogether the heating occupied some three or four hours. On opening the tube under water it was found that great absorption of gas had taken place.

Of the 118 cubic centims. of CO taken for experiment, 92 cubic centims. were absorbed,

26

118

The action appears to be

residue.

CO+0=CO2,

regard being had to the amount of KMnO, which had been reduced during the operation. At ordinary temperatures the action takes place, but very slowly.

Hydrogen

is also absorbed by the alkaline solution of permanganate. In an experiment in which 64 cubic centims. of hydrogen were sealed up with 16 cubic centims. of alkaline permanganate and heated for some hours in the water-bath, an absorption of 34 cubic centims. of hydrogen was noted.

We are continuing the investigation.

XL. Researches on Unipolar Induction, Atmospheric_Electricity, and the Aurora Borealis. By E. EDLUND, Professor of Physics at the Swedish Royal Academy of Sciences*.

[Plate VIII.]

"Res ardua rebus vetustis novitatem dare, novis auctoritatem, obscuris lucem, dubiis fidem, omnibus vero naturam et naturæ suæ omnia."PLINY, Hist. Nat. t. i., præfatio.

§ 1. Unipolar Induction.

WE represent to ourselves a steel magnet in a vertical

position, readily set in rotation about its geometric axis; and we picture it besides surrounded by a metallic muff in the form of a cylinder, which can likewise turn about the same axis. If the current of a pile be caused to pass into this • Translated from a copy, communicated by the Author, of the Kongl. Svenska Vetenkaps-Akademiens Handlingar, vol. xvi. No. 1.

Phil. Mag. S. 5. Vol. 6. No. 37. Oct. 1878.

U

cylinder in such manner that one of the electrodes is in contact with it in the vicinity of the poles of the magnet, and the second electrode at a point situated between the two poles, experience shows that the cylinder begins to rotate about the magnet. The direction of rotation depends on that of the current in the cylinder, and also on the situation of the poles. As to the magnet itself, it remains perfectly motionless; consequently the galvanic current does not exert upon it any rotatory action. It is therefore possible to turn the magnet mechanically round its axis without the slightest obstacle being offered by the reciprocal action of the magnet and current; the sole resistance to be surmounted in the mechanical rotation of the magnet is occasioned by the friction in the sockets of the axis, &c., a resistance which has nothing to do with the current. In a previous memoir* I have demonstrated that, according to the mechanical theory of heat, every phenomenon of the sort mentioned will be accompanied by a phenomenon of unipolar induction. In fact, if the pile be removed and replaced by a galvanometer inserted between the two abovementioned electrodes, and in contact with the cylinder, the galvanometer indicates the rise of a current as soon as the cylinder is mechanically put in rotation. The electromotive force here consists of the mechanical work necessary to overcome the reaction of the magnet upon the current in that part of the circuit which is set rotating. This species of induction has received the name of "unipolar induction." The current produced is proportional to the velocity of the cylinder. Of course the simultaneous mechanical rotation of the magnet produces no augmentation in the current generated by the rotation of the cylinder, since this augmentation would be made without the consumption of a corresponding amount of mechanical work-which would be perfectly absurd. Plücker has also proved by experiment that in this case the rotation of the magnet is incapable of producing a current. Here is another reason:-It has been demonstrated by experiment that, in the case in question, for the magnet a solenoid can be substituted, producing the same effectt. I have myself shown, on a previous occasion, that the rotation of the solenoid about its axis cannot produce a unipolar induction current, whether proceeding from a single fluid or from two fluids in translatory motion. It is therefore, in this case, the rotation of the cylinder about the magnet that gives rise to the observed uni

Öfversigt af kongl. Vetensk.-Akad. Vörh. 1877; Wiedemann's Annalen, vol. ii. p. 347.

↑ Öfversigt, April 1877; Pogg. Ann. vol. clx. p. 604.

Öfversigt, Jan. 1877; Pogg. Ann. vol. clx. p. 617.