THE SEMI-CENTENNIAL ANNIVERSARY 53 



While it should not be rejected, the merits of other theories must 

 not be overlooked. 



Chief among these is the theory that rests on the very probable 

 assumption that every molecule is a magnet. If the magnetism 

 is accounted for as the effect of the rapid revolution of electrons 

 within the molecule, a gyrostatic action might be anticipated. 

 That is, each molecule would tend to set itself with its axis 

 parallel to the axis of the earth, just as the gyrostatic compass, 

 now coming into use at sea, tends to point to the geographical 

 pole. The host of molecular magnets, all acting together, might 

 account for the earth's magnetic field. 



This theory, in its turn, is not free from some points of 

 weakness, though they may disappear as the result of more ex- 

 tended investigation. Its chief advantage lies in the possibility 

 that it may perhaps explain the secular variation of the earth's 

 magnetism by a precessional motion of the magnetic molecules. 



On either hypothesis, it is assumed, in the absence of knowl- 

 edge to the contrary, that every molecule contributes to the pro- 

 duction of the magnetic field. Thus the density of the rotating 

 body may prove to be a factor. Perhaps the change of density 

 from the surface to the center of the sun must also be taken into 

 account. But the observational results already obtained suggest 

 that the phenomena of ionization in the solar atmosphere may 

 turn out to be the predominant influence. 



The lines which show the Zeeman effect originate at a com- 

 paratively low level in the solar atmosphere. Preliminary 

 measures indicate that certain lines of titanium, which are widely 

 separated by a magnetic field in the laboratory, are not appre- 

 ciably affected in the sun. As these lines represent a somewhat 

 higher level, it is probable that the strength of the sun's field 

 decreases very rapidly in passing upward from the surface of the 

 photosphere — a conclusion in harmony with results obtained 

 from the study of the corona and prominences. Thus it may be 

 found that the distribution of the electrons is such as to give rise 

 to the observed field or to produce a field opposing that caused by 

 the rotation of the body of the sun. It is evident that speculation 



6 



