May 30, 1 8 78 J 



NATURE 



\2J. 



consideration ; and that is, that the moon' s heat cannot 

 produce the phenomena in question. M. Faye shows that 

 if the moon' s reflected heat is in the same proportion as 

 the reflected light, such heat cannot produce a change of 

 temperature of a thousandth of a degree Fahrenheit. I 

 would remark that Lord Rosse's carefully-made experi- 

 ments with the most delicate apparatus have shown for 

 the total heat radiated and reflected, nearly ten times the 

 proportion given by the reflected light ; but, as M. Faye 

 observes, if the proportion were increased a hundredfold 

 the effect would still be insensible. " How then," it is 

 added, "can we expect such an action to dissipate the 

 clouds when that of the sun does not always succeed ?" 

 If, however, we can establish that real lunar actions 

 exist which cannot be explained by the moon's heat 

 reflected or radiated, the only philosophical conclusion 

 will be that the moon must act in some other way, which 

 it will be in the interests of science to seek out. 



Influence of the Sun-spots on the Earth's Magnetism. — 

 It has been found that since the first accurate series of 

 magnetic observations, towards the end of last century, 

 to the present time, the maximum and minimum of sun- 

 spot frequency have occurred at the same times as those 

 of the diurnal oscillation of the magnetic needle; but 

 because Dr. Wolf has believed that in the interval 

 between 1787 and 181 8, when the observations of both 

 phenomena were very incomplete, there were only two 

 cycles for both, and because Dr. Lamont and myself 

 believe there were three for both, M. Faye concludes 

 that the mean length of the period for the magnetic 

 needle was different from that for the sun-spots. The 

 true conclusion is— if Dr. Wolf is right the mean length 

 of the period for both phenomena since 1787 is nearly 

 twelve years ; if the other view is right the mean length 

 for both is 10-45 years. I have already considered this 

 question in Nature (vol. xvii. p. 262). The egalit^ 

 rigoureuse of the lengths of the periods sought by M. 

 Faye is thus established whatever view be taken. 



It is next sought to show that any variation in the 

 earth's magnetism in the decennial period cannot be due 

 to variations in the solar heat produced by the spots. 

 Founding on Mr. Langley's observations for the tempe- 

 rature of the photosphere, of the nucleus and the 

 penumbra of spots, M. Faye shows, making use as before 

 of the absolute temperature of the earth, that the variation 

 of temperature due to the spotted surface of the sun 

 cannot exceed ± o°-3 Fahr. 



_ I venture here to remark, in the first place, that I have 

 given strong grounds elsewhere for believing that the 

 sun-spots are not the cause of the decennial period in the 

 magnetic variations ; but that both are due to the action 

 of a common cause.^ I quite accept then M. Faye's 

 conclusion. It may, however, be argued that more heat 

 proceeds from the sun in years of many, than in those of 

 few, spots, but the observations of temperature, which 

 have been made with so much accuracy in many countries 

 for the last twenty years, will prove at once that no marked 

 variation of the mean temperature occurs in the decennial 

 period. 



We must here, however, consider the reasoning which 

 has been employed on this subject. It is shown that the 

 variation of temperature due to the spotted suiface of the 

 sun cannot explain the change in the amplitude of the 

 diurnal oscillations of the needle. It is, however, a mere 

 assumption that this oscillation is due to the solar heat 

 an assumption for which there is really no sound basis' 

 unless some very rude attempts at a hypothesis can be so 

 considered. 



M. Faye says, " This phenomenon is absolutely general ; 



' " On the Decennial Period 

 P- 593). 



and goes on increasing thence towards the poles," This 

 variation of range is afterwards compared with the 

 inverse law for the diurnal variation of the barometer, the 

 range of which diminishes from the tropics towards the 

 poles. 



This view, however, is founded on a misconception of 

 the facts ; we might just as well say that the earth' s mag- 

 netic force diminishes from the equator towards the 

 poles (which is just the reverse of the truth), because its 

 horizontal component does so. The mode in which this 

 movement varies in amount from the tropics towards the 

 poles appears to be imperfectly known, and as it is 

 essential to make this clear before we can compare 

 the facts with any hypothesis, I shall now attempt to 

 do so. 



Diurnal Variation of the Magnetic Declination. — The 

 horizontally suspended magnetic needle performs an oscil- 

 lation in twenty-four hours, during which the north end is 

 most westerlyin the northern hemisphere, and most easterly 

 in the southern hemisphere, between i and 2 P.M. If we 

 consider needles at different stations on the same meri- 

 dian, it might be supposed that as we approached the- 

 equator this opposition of movement would result in the- 

 needle becoming stationary. That was Arago's conclu- 

 sion. This idea, however, neglected the position of the 

 sun. The range of the oscillation is greatest in our 

 hemisphere when the sun is north, and greatest in the 

 southern hemisphere when the sun is south of the 

 equator. It is only when the sun has the intermediate 

 position, near the time of the equinoxes, that the forces^ 

 are nearly balanced at equatorial stations. 



If we suspend an unmagnetic steel needle horizontally 

 on its centre of gravity, so that it can move both in a 

 horizontal and vertical plane, and then magnetise it, the 

 needle dips with one end below, the other above, the 

 horizontal plane passing through its centre ; and the 

 direction in which it lies is that of the earth's magnetic 

 force. If this needle moves up, or down, or sideways,, 

 this is because other forces pull it one Avay or another, or 

 because the direction of the earth's magnetic force has- 

 changed. Let us suppose in the first instance that the 

 latter hypothesis is the true one, that is to say, that the 

 earth's magnetic axis moves with the sun, the north pole 

 having the greatest movement when the sun is in the 

 northern hemisphere. If the dipping needle moves in. 

 the plane perpendicular to the vertical plane, through an. 

 angle u, and we wish to know what would have been the 

 corresponding movements eastwards or westwards, if the 

 needle had been made horizontal, by the addition of a 

 small weight to the end above the horizontal plane, we 

 must divide the angle u by the cosine of the dip below 

 the horizon. Now in England the cosine of the dip is about 

 \, so that the horizontal needle would have moved through 

 an angle of 3 u. We obtain similarly the former from the 

 latter by multiplying by the cosine of the dip. 



We may now see, from observations made at different 

 stations, what is the range of the monthly mean diurnaL 

 variation of the horizontal needle multiplied by cosine 

 dip in a month for which it is a maximum in the northern 

 hemisphere. The following are approximations to the 

 mean ranges thus obtained from several years observa- 

 tions in the month of August for the northern hemi- 

 sphere. 



Station. 



Makerstoun ... 



Toronto 



Simla 



Bombay 



Madras 



Trevandrum ... 



In a similar way we find for the month of February for 

 southern stations — 



