them to very curious consequences. M. Brandes, in 1823, was enabled to determine the height of a certain number of shooting stars, by observing them simultaneously with persons situated in different places. M. Quetelet, and other philosophers, have also been much occupied in observing these phenomena. M. Arago, in an article inserted in the Annuaire of 1836, has summed up with great care the most certain observations which we have on this subject, and has said, among other things, that the most habitual direction of shooting stars seems to be diametrically opposite to the movement of the translation of the earth in its orbit, but that it would be desirable that the result of M. Brande should be established on the discussion of a great number of observations. This question, then, is far from being cleared up; it is this fact which has induced me to attempt a regular system of observations. I here present a summary of the first year. If the results are not so certain as I could have desired, it is because the observations are not yet sufficiently numerous. I have confined myself as much as possible to observations from between eleven o'clock at night to one o'clock in the morning, whenever the weather was fine, and there was no moon. I always made my observations facing from east to north-east. I will now give my mode of proceeding in my observations. Immediately on observing the appearance of a shooting star, I marked with great care, in short, the hour, the magnitude of the star, the colour, the duration of its appearance, and the arc apparently travelled over; I traced at the same time on a celestial globe placed before me the direction and position in relation to stars and constellations. I took no longer in reporting these particular circumstances than the phenomenon was in presenting itself; the time employed in these annotations was rather more than a quarter of a minute's duration. After each series, I indicated the general appearance of the sky, the temperature, the height of the barometer, the direction of the wind. In transcribing all these observations into a register, I took care to determine the intersections of the ecliptic with the directions of stars directly prolonged, by applying on a celestial globe a metallic semi-circle passing by the extremities of the routes travelled over. I give also the apparent directions relative to the trace of a small circle parallel to the meridian, and passing by the middle of the apparent direction observed; to determine them I have taken account of the hours of observation and of the position of the heavens.

The summary of the observations which I present to-day comprehends a year, from the month of October, 1840, to the month of October, 1841, exclusive. In this interval I have been able to observe 572 shooting stars in 86 days. I may remark that it has not been possible to prove the extraordinary appearances of November, 1840, and August, 1841. The principal results will be found in the following tables.

TABLE III.

Intersection of the Ecliptic by the direction of the shoot- Long.of the ing stars, whether of direction or movement.

Earth, corresponding

MONTH.

Aries Taurus Gemini Cancer
to the
and and
and
and Leo and and Mean of the
Libra. Scorpio. Sagittarius. Capricorn Aquarius. Pisces. Observns.

Virgo

We cannot draw any certain consequences from all the results which I have just presented: there are too few observations to make the actual numbers be the faithful representation of the truth. However, on examining the second table, we see that of the number of stars observed, the 0.31 hundreths moved in the plane which passes by the N. W., and S. E., and 0.22 in that which traverses the S. W. and the N. E.; whilst that the other directions are in the most feeble proportions.

In considering the latter table, it seems that the most general direction of falling stars appeared diametrically opposed to the movement of the translation of the earth, as M. Arago has observed in his instructions for the exploring ship Bonite. Thus, in the month of October, the maximum of the apparent directions, is formed in the planes which pass by the signs Gemini and Sagittarius, of Cancer and Capricorn; in the month of November, this maximum advanced to 30°, a movement equal to that of the earth. There are many months which present anomalies, but the cause proceeds no doubt, from the too small number of stars observed. I do not attach then to these results more importance than my observations allow of; I have merely thought that it was not out of place to make the remark, precisely because there is a certain agreement with that which M. M. Brandes and Arago have said.

On the Heat evolved by Metallic Conductors of Electricity, and in the Cells of a Battery during Electrolysis. By JAMES PRESCOTT JOULE, Esq.*.

1. There are few facts in science more interesting than those which establish a connexion between heat and electricity. Their value, indeed, cannot be estimated rightly, until we obtain a complete knowledge of the grand agents upon which they shed so much light. I have hoped, therefore, that the results of my careful investigation on the heat produced by voltaic action, are of sufficient interest to justify me in laying them before the Royal Society.

CHAP. I.—Heat evolved by Metallic Conductors.

2. It is well known that the facility with which a metallic wire is heated by the voltaic current is in inverse proportion to its conducting power, and it is generally believed that this proportion is exact; nevertheles I wished to ascertain the fact for my own satisfaction, and especially as it was of the utmost importance to know whether resistance to conduction is the sole cause of the heating effects. The detail, therefore, of some experiments confirmatory of the law, in addition to those already recorded in the pages of science, will not, I hope, be deemed superfluous.

3. It was absolutely essential to work with a galvanometer, the indications of which could be depended upon, as marking definite quantities of electricity. I bent a rod of copper into the shape of

a rectangle (A B, fig. 1), twelve inches long, and six inches broad This I secured in a vertical position by means of the block of wood C; N is the magnetic needle, 33 inches long, pointed at its extremities, and suspended upon a fine steel pivot over a graduated card placed a little before the centre of the instrument.

4. On account of the large relative size of the rectangular conductor of my galvanometer, the tangents of the deviations of the needle are very nearly proportional to the quantities of current electricity. The small correction which it is necessary to apply to the tangents, I obtained by means of the rigorous experimental process which I have some time ago described in the "Annals of Electricity."+

• Communicated by the author. + Vol. iv, pp. 131, 132, and 476.

5. I have expressed my quantities of electricity on the basis of Faraday's great discovery of definite electrolysis, and I venture to suggest, that that quantity of current electricity which is able to electrolyze a chemical equivalent expressed in grains, in one hour of time, be called a degree. Now, by a number of experiments, I found that the needle of my galvanometer deviated 33°5 of the graduated card, when a current was passing in sufficient quantity to decompose nine grains of water per hour; that deviation, therefore, indicates one degree of current electricity on the scale that I propose to be adopted. We shall see in the sequel some of the practical advantages which I have had by using this measure.

6. The thermometer which I used had its scale graduated on the glass stem. The divisions were wide, and accurate. In taking temperatures with it, I stir the liquid gently with a feather, and then, suspending the thermometer, by the top of its stem, so as to cause it to assume a vertical position, I bring my eye to a level with the top of the mercury. In this way a little practice has enabled me to estimate temperatures to the tenth part of Fahrenheit's degree with certainty.

Fig. 2.

7. In order to ascertain the heating power of a given metallic wire, it was passed through a thin glass tube, and then closely coiled upon it. The extremities of the coil thus formed were then drawn asunder, so as to leave a small space between each convolution, and if this could not be well done, a piece of cotton thread was interposed. The apparatus thus prepared, when placed in a glass jar containing a given quantity of water, was ready for T experiment. Fig. 2 will explain the dispositions: A is the coil of wire; B the glass jar partly filled with water; T represents the thermometer. When the voltaic electricity is transmitted through the wire, no appreciable quantity passes from it to take the shorter course through the water. No trace of such a current could be detected, either by the evolution of hydrogen, or the oxidation of metal.

B

8. Previous to each of the experiments, the necessary precaution was taken of bringing the water in the glass jar, and the air of the room to the same temperature. When this is accurately done, the result of the experiments bear the same proportions to one another as if no extraneous cooling agents, such as radiation, were present; for their effects in a given time are proportional to the difference of the temperatures of the cooling and cooled bodies; and hence, although towards the conclusion of some experiments this cooling effect is very considerable, the absolute quantities alone of heat are affected, not the proportions that are generated in the same time. [See the table of heats produced during half an hour and one hour, p. 290.]

9. Exp. 1.-I took two copper wires, each two yards long, one of