152 



NATURE 



\ywie 6, 1878 



not be explained by the hypothesis that the sun acts as a 

 magnet. But, it is said, " May the moon not acquire 

 induced magnetism under the action of the earth, per- 

 petually variable according to the relative position of the 

 two bodies ? If we consider the enormous magnetic 

 power of the earth, that Gauss finds equal to that of 

 464 trillions ' of magnets weighing a pound each, and if 

 we remark besides that the distance of the moon to the 

 earth does not exceed thirty times the length of this 

 gigantic magnet, we may give an affirmative answer to 

 the question proposed. But then the magnetism induced 

 in the moon should in its turn exercise a small action 

 upon the proper magnetism of the earth in the period of 

 a lunar month. The observations alone can decide this 

 provided they are of great precision." 



M. Faye then cites the results obtained from the 

 Toronto observations by Gen. Sir E. Sabine, that for 

 the magnetic declination showing a range of o'"64; and 

 he adds, "All these effects are of double period; they 

 show two maxima and two minima in the course of the 

 lunar month of 29^ days, which proves that they are due 

 to an induced or reflex action, not to a direct action of 

 the moon herself." I shall put my remarks on this sub- 

 ject under three heads. 



1. Is such a result possible for the moon's synodical 

 revolution ? Let us commence with full moon at the 

 winter solstice ; near this epoch the moon is in the plane 

 perpendicular to the ecliptic passing through the earth's 

 magnetic axis and the sun. The north pole of the ter- 

 restrial magnet is then presented to the moon in such a 

 way as to produce the maximum of induction ; when the 

 moon is near her third quarter the two terrestrial mag- 

 netic poles will be equidistant from the moon and the 

 inducing action will be a minimum ; there will be a 

 second maximum near new moon when the south pole is 

 most presented to our satellite and a second minimum 

 near the first quarter. If now we follow the earth in her 

 revolution to the vernal equinox, we shall find all this 

 changed. At full moon our satellite is then equidistant 

 from the two terrestrial poles, and the inducing action is 

 a minimum ; it is a maximum, on the contrary, near the 

 first and third quarters. The consequence will be that if 

 any inducing action existed it would have the same value 

 at all ages of the moon in the mean of observations made 

 during a series of years, such as were employed by Sabine 

 for the variations in question. Such a result, however, 

 as has been imagined by M. Faye might be possible if, 

 instead of the synodical, we employ the tropical revolution 

 of the moon, which occupies nearly 27*3 days. 



2. We may inquire, then, if the moon as a permanent 

 or induced magnet can produce any magnetic variations 

 appreciable by our instruments ? In the first place, Mr. 

 Stoney has shown that if the moon were as magnetic bulk 

 for bulk as our earth, her whole action in deflecting a 

 freely-suspended needle in our latitudes, could not exceed 

 one-tenth of a second of arc (o""i).* In order to consider 

 the question of the variable magnetism induced in the 

 moon by our earth, let us suppose her inductive capacity 

 equal to that of cast-iron. From Barlow's experiments at 

 Woolwich with iron balls I find that the magnetism 

 induced in an iron ball of one foot diameter is about 2'o, 

 in English units, which is nearly twice the magnetic force 

 given by Gauss for the same volume of our earth. Barlow 

 found the induced moments of different balls to vary as 

 their volumes, and assuming that the induced magnetism 

 varies inversely as the cube of the distance of the inducing 

 and induced bodies, we find at the moon' s distance (60 

 terrestrial radii) the induced magnetism at the maximum, 

 under the most favourable condition, could not be more 



than ^, = — :; of that supposed in the first case, 



60^ 108,000 



' M. Faye uses the word trillions, but the trillions are English, not French, 

 the latter being a very different number, 

 a Pkil. Mag. , vol. xxii. p 294. 



that is when as magnetic as the earth. Her whole action 

 on a magnetic needle here, then, due to the earth's 

 induction, could not exceed one millionth of a second of 

 arc. It is advantageous to get rid of hypotheses which 

 are so completely insufficient, and we may put aside for 

 the future any consideration of the moon's action by her 

 own permanent magnetism, or by a variable magnetism 

 induced in her by the earth. 



3. M. Faye has also misunderstood the facts which he 

 wished to explain. The results obtained by Sabine have 

 reference to a variation which occurs in 24I hours, the 

 lunar day, and not the lunar month of 29^ days. The 

 laws of the lunar diurnal variations were obtained first by 

 Kreil for the magnetic declination, and by myself for the 

 magnetic force and inclination. This action of the moon 

 is, however, so very different from what is generally sup- 

 posed, and from what was concluded from the first investi- 

 gation on the subject, that it is of the greatest import- 

 ance, in relation to the whole question of cosmic meteoro- 

 logy, I should state some of the more marked facts which 

 have been deduced from eleven years' hourly observa- 

 tions on the magnetic equator. I shall limit myself at 

 present to the lunar actions on the direction of the 

 horizontal magnetic needle. 



The moon, in a lunar day of 247 hours, produces a 

 variation in the earth's magnetism, such that the mag- 

 netic needle makes two complete and nearly equal oscilla- 

 tions from an easterly to a westerly position in the 

 interval in question. This is the general 7}iean law. We 

 have seen, in considering the law of the solar diurnal 

 variations that, near the magnetic equator, the la\y^ 

 becomes reversed when the sun passes from the one 

 hemisphere to the other, so that when the sun is north, 

 the movement of the needle is like that in high north 

 latitudes, and when south, like that in high south lati- 

 tudes. If, then, the moon acts in the same way as the 

 sun, we should expect a similar phenomenon for the 

 lunar diurnal variation when the moon crosses the 

 equator. This is not the fact. The law differs little for 

 the position of the moon north and south of the equ.itor. 



There is, however, an inversion of the lunar diurnal 

 oscillations ; thus, in the months of December and 

 January the north end of a magnetic needle is farthest 

 east when the moon is on the upper and lower meridians, 

 and farthest west near moon-rise and moon-set ; whereas 

 in the months of June and July the reverse is the case, 

 the north end of the needle being farthest west when the 

 moon is on the meridian (upper and lower) and farthest 

 east when she is on the horizon. It followed from this, 

 as for the solar diurnal law, that the oscillations should 

 be in opposite directions at the same time in the higher 

 latitudes of the two hemispheres, as has been found to be 

 the case. 



It is not then when the moon crosses the equator but 

 near the times when the sun does so, that the moon' s 

 action is reversed. 



The dependence of the lunar action on the position of 

 the sun becomes more evident as the investigation 

 becomes more detailed. When we determine the mean 

 law for each month of the year, we find that the north 

 end of the needle moves equally far east and equally far 

 west at each of the two oscillations in the lunar day ; this 

 is not found to be the case for different positions of the 

 moon relatively to the sun. Thus in the quarter lunations 

 including full moon, in the months of December and 

 January, the greatest west-east-west oscillation of the 

 needle occurs when the moon is on the lower meridian ; 

 not when the moon, but when the sun, is shining on the 

 place of the needle. The oscillation from moon-rise to 

 moon-set, that is to say, while the moon is above the 

 horizon, is little more than one-third of the oscillation for 

 the half day when she is below the horizon ; the two 

 westerly extreme positions when the moon is on the 

 horizon are nearly the same. 



