ation." What is remarkable, we are also assured that an annual variation of the declination at the same hour of 7 to 8 A. M., “ almost precisely similar in character and amount," obtains at Hobarton, which is nearly diametrically opposite to Toronto, and also at the two intertropical stations of St. Helena and the Cape of Good Hope, and is probably therefore a general phenomenon. It will be seen upon a little examination that it is a simple consequence of the shifting of the position of the radial current proceeding from the point underneath the sun, which may be taken as the representative of all the radial currents in action at the hour in question. To show this let us first suppose the station to be on the equator. At the hour of 7 to 8 o'clock A. M. the point of maximum excitement will be on the 6 o'clock hour circle, as it lags an hour or two behind the sun. (See p. 192.) Now let EQ, Fig. 5, be the equator, as traced on the photosphere, P

=

the zenith of the station of the needle, N the position of the point of highest intensity at the time of the northern solstice, on the 6 o'clock hour circle, and S the same point at the time of the southern solstice. NPE EPS=234°. The meridional component of the current along NP will be directed toward the south, that of the current SP toward the north. The former will displace the north end of the needle toward the east, the latter toward the west by an equal amount. (A transverse current, in the present instance, could have no effect, since it would be equally inclined to the meridian at the two solstices, and solicit the needle with the same force toward the east.)

Next suppose Toronto (N. lat. 4340) to be the station of the needle, and represent it by P' in Fig. 5. P/E will be the prime vertical on the photosphere; and P'EQ=4340, P/E = 90°, and NE SE = 231°. From which we have, NP' = 74° 5', NPE=17° 30', SPE=17° 30', SP=105° 55'. Current NP' impels the north end of the needle toward the East, and SP' toward the West. The meridional component of current NP'= NP sin NP'E; that of SP-SP' sin SP/E; hence the deflecting force at Toronto in the interval of the solstices, is to that at a station on the equator in the ratio of NP' sin NP'E+SP' sin SP/E to 2 NP sin 2330.* Going through with the calculation we find the ratio to be 0.78. NP' sin NP'E=0·312,† SP' sin SP'E= 0-312. The horizontal force at Toronto is to that at the equator nearly as 5 to 9; dividing 0-78 by we obtain 1-40 as the ratio. of deflection at the two stations, or the deflection being 5' at Toronto it should be 3'57 at the equator. I find by calculation, allowing for the west declination of the needle at St. Helena, that the deflection should be very nearly the same at St. Helena as at the equator. At Hobarton it should be about 4'. We may conclude, therefore, that the deflection, so far as it depends on the radial currents, is everywhere in the same direction, from the one solstice to the other, and that it should vary in amount at the four British Colonial Observatories between 5' and 3'6. The diagrams given by Colonel Sabine, (Toronto Observations, Vol. II, p. 20,) represent the deflection as about 0'5 less at St. Helena than at Toronto.

It is obvious that if we compare any two periods of time. equally distant from the equinox, we shall have a deflection of the same character, though less in amount than in the case just considered, since the north and south declinations of the sun, (NE and SE, Fig. 5,) are less than at the solstices. Since the change of the sun's declination is very slow before and after the solstice, the amount of the deflection should be nearly the same for the months that precede and follow the solstices. This fact as the result of observation is exhibited to the eye in Fig. 6, (which is a transcript of Fig. 1, p. 20, Vol. II, of Toronto Observations). From the cause now under consideration the same effect should occur at the hours following 7 to 8 A. M.; but it should be less in amount from hour to hour, because the angle NP'S is less and less as the arc NS is carried farther to the west by the diurnal revolution. This fact is also shown in the diaBy the arcs NP', SP', NP, is meant the force of the currents proceeding from N and S to P and P'.

Current along NP'=

1
sin NP

The effect of the currents proceeding from

any point, as N, varies with the distance by reason of the divergence of the individual currents, which follow arcs of great circles, because of the effect of resistances upon the individual currents, and doubtless also from the very nature of the propagation. The intensity of a galvanic current varies inversely as the square root of the length of the wire traversed.

17h 18h 19h 20% 21h 22h 23h Oh 1h 2h 34

These curves show the deviations of the declination, at the specified hours, in the months named, from the mean declination in all the months and at all the hours, represented by the horizontal line MM. An ordinate lying above MM shows that the declination (West) at that hour is less than the mean, one lying below that it is greater than the mean. 1, May; 2, June; 3, July; 4, August; 5, November; 6, December; 7, January; 8, February. Scale Oin 5 to 1' of arc.

gram, (Fig. 6). It should disappear at noon, and recur in the afternoon, but we shall soon see that antagonistic causes come into operation in the afternoon, which prevent the same result from being realized. We shall see also that the deflection at 7 to 8 A. M., and the forenoon hours generally, in passing from the summer to the winter months, is not wholly due to the shifting of the radial current from the north to the south side of the prime vertical, as we have thus far considered it.

We must now direct attention to the existence of another form of electric current in the photosphere of the earth, already briefly alluded to, from which several interesting effects ensue. This is a current, or rather system of currents, induced by the sun's magnetic action on the photosphere, and running in a direction parallel or nearly so to the ecliptic and from east to west. I conceive them to be developed by the inductive action (the simple propagation of an impulse probably) of currents traversing the sun's surface in a direction parallel, or nearly parallel to his equator. The physical theory of their excitation does not fall within our present inquiry. The following discussion will, I think, serve completely to establish the fact of their existence. To have a clear conception of their diurnal and annual change of position, we may regard them (or at least those which originate in the lower latitudes) as represented by a single current following the direction of the ecliptic traced on the earth, and observe that this is carried around with the sun during the day; and at a given hour of the day goes through in the course of a year the same changes of position that it does during a day. It is also to be observed that the force of this current will be the greatest at or near the point directly underneath the sun. At either equinox, and at the hour of noon, this current will be inclined 234° to the meridian of the station; at the vernal equinox passing from the north to the south side of the equator, and at the autumnal equinox from the south to the north side. The meridional component of this current will then be directed from north to south at the vernal and from south to north at the autumnal equinox. The north end of the needle ought therefore, to stand farther to the west at the autumnal than at the vernal equinox, at the hour of noon and thereabouts, at all stations. In fact there is an excess at Toronto, at noon, of 3/23.* (See Table I.) In what precedes I have only considered the action of the ecliptic currents near the equator. In point of fact, the sun acts upon the high as well as the low latitudes, and develops at each point of its action a current which sets out in a direction parallel to the ecliptic and follows the course of a great circle. The more northerly currents may in general be approximately represented by a single current passing through the zenith of the station. Throughout the year,

*This inequality of declination, which has its positive maximum at the autumnal and negative maximum at the vernal equinox, and is zero at the solstices, does not appear hitherto to have been noticed.

+ More correctly by a single current originating in about the latitude 45°. The station is supposed in the text to be in a high northern or southern latitude (as Toronto). So long as we are only attempting to explain the laws of the phenomena it is not important that we should know the precise starting point of this representative current. It will at any given hour be shifting its position in the same direction, from whatever point in the higher latitudes it be supposed to issue. It is to be observed that, other things being the same, the effect of a current will be the greatest when it passes through the zenith of the station.

the current developed within the Torrid Zone (or rather within a few degrees of the ecliptic) may be represented by a single current following the course of the ecliptic traced on the earth's photosphere. At the equinoxes the northerly currents will cross the meridian at noon, under larger angles than this single ecliptic current, but their tendency will be the same in the inequality just considered. Let us endeavor to obtain a general conception of the entire system of currents now under consideration. At the solstices the currents will everywhere, at the outset be parallel to the equator, and the circle traced through the various points from which they proceed at any one instant, will be a meridian and correspond to the solstitial colure. The entire system of currents will pass through the pole of this circle, lying on the equator 90° to the west of the circle. As the earth rotates the circle of excitement with its pole will be carried toward the west, and the inclination of each of these currents to the meridian of any particular station will vary continually. The currents in question will also pass through the other pole, 90° to the east of the circle on which they originate, but they will flow from this pole toward the circle, and from the circle toward the other pole. To the west of the circle they are leading currents, to the east of it following currents. The two poles will in all cases be diametrically opposite to each other, and on the ecliptic. The starting point of the current that passes through the zenith of the station also varies continually. At 6 A. M. and 6 P. M. this current issues at the pole, at noon its starting point is in the zenith of the station. At the equinoxes the circle of excitement will coincide with the circle of latitude (that is circle through the pole of the ecliptic) which passes through the equinoctial points. The currents will set out perpendicularly to this circle and all meet at its pole. This pole will, at the autumnal equinox, lie in the northern Tropic, and move along it toward the west. At the vernal equinox the pole will follow the course of the southern Tropic. During the year it will move along the ecliptic from west to east, keeping always 90° behind the sun. It will therefore pass gradually from one tropic to the other, as the sun does, but be on one of the tropics when the sun is on the equator, and vice-versa. The two poles, or points of concentration of the currents, will always be on opposite sides of the equator, and in the same latitude. At 6 A. M. and 6 P. M., on the day of the autumnal equinox, the current which traverses the zenith of the station originates at a point on its meridian 234° beyond the pole; at noon at a point a few degrees to the south and east of the station, and crosses the meridian under an angle of 73°. At the same hours on the day of the vernal equinox, the current in question originates at the point on the meridian 2330 on this side of the pole ; at noon a few degrees to the south and west of the station, and deviates 73° from the meridian.