32 



SCIENCE. 



[N. S. Vol. XIX. No. 470. 



for discontinuous surfaces of different tem- 

 peratures sliding past each other with different 

 velocities, and rolling up vortex tubes between 

 them. The layers are warmer around the 

 axis of rotation of the sun, and have slower 

 angular velocities than those more distant 

 from it. The vortex tubes have the shape 

 indicated in Fig. 1, right-handed in the 

 northern hemisphere and left-handed in the 

 southern. If the constituents of rotating 

 matter carry electric cljarges in their atoms 

 and molecules, this vortex entrain m en t will 

 produce polarization and a true magnetic field 

 extending outside the sun. The rotation 

 period of the magnetic field near the poles, 

 since it is primarily seated in the nucleus, is 

 the same as that of the surface at the equator, 

 namely, 26.68 days. The earth's normal mag- 

 netic field has a component system impressed 

 upon it which is directed from north to south 

 perpendicular to the ecliptic, and these vec- 

 tors are probably portions of the lines here 

 described as springing from the solar nucleus. 

 Furthermore, all large cooling masses, con- 

 tracting by their own gravity and rotating on 

 ah axis must, in conformity with the equa- 

 tions of motion, set up such a polarized in- 

 ternal structure, and, therefore, all stars are 

 probably magnetic. The earth still possesses 

 a residual magnetism originally produced in 

 this manner, which is gradually fading away 

 as the earth cools, and will become very feeble 

 as the loss of convective heat progresses, some- 

 what like that of the moon at the present 

 time. The belt systems on the planets Jupiter 

 and Saturn afford examples of rotations with 

 discontinuous surfaces, and minor vortices be- 

 tween them, under this law. The granulated 

 surface of the sun is probably due to this 

 vortex motion, where each granule represents 

 the discharge of a single vortex tube. 



The Solar-terrestrial Synchronism. (Fig. 

 2.) — The eleven-year cyclic period of the sun- 

 spot variation gives a curve with one principal 

 maximum and one principal minimum, but 

 this register of the solar action is not so sen- 

 sitive as that recorded in some of the other 

 elements. The eruptions of the sunspots and 

 the faculse are confined to the gaseous en- 

 velope, and do not directly represent the 



working of the viscous nucleus. The prom- 

 inences of higher latitudes, 40° to 80°, pro- 

 duce the same fundamental curve, but there 

 are minor crests superposed upon it, some- 

 times one on the ascending branch. A, and 

 usually two on the descending branch, C, D. 

 In some of the eleven-year cycles A does not 

 appear, and one might count the length of 

 the short period from the three crests, B, G, D, 

 and make it, 11.1 -^ 3 = 3.7 years, as Lockyer 

 has done. I have taken the four crests, A, 

 B, C, D, and make the average period, 11.1 -^- 

 4 = 2| years, as in ' Weather Bulletin ' No. 

 21, page 125. This more sensitive curve 

 registers primarily the action of the solar 

 nucleus, and the minor crests are the re- 

 crudescences of a contracting and congesting 

 medium seeking to free itself of supercharged 

 energy. The curve is found to be repeated 

 in a remarkable manner throughout the cos- 

 mical system. Thus, we have found, (1) that 

 the periods of rotation in the higher zones of 

 the sun, 50° to 70°, reproduce the curve in 

 a secular variation, and refer its cause, with- 

 out doubt, to the effects of internal circula- 

 tion;* (2) that the magnetic field at the earth 

 synchronizes with it;t (3) that the terrestrial 

 temperatures, in the tropical zones give the 

 same curve directly, but in the temperate 

 zones they synchronize in an inverted form; 

 while the terrestrial pressures synchronize di- 

 rectly with it in the regions around the Indian 

 Ocean, Australia, South Asia and Africa, but 

 in an inverted form throughout North and 

 Soiith America. This inversion implies that 

 there is a surging of the earth's atmosphere in 

 the process of its general circulation, whereby a 

 portion rises in pressure and temperature while 

 another portion falls. This opens up a new field 

 of meteorological research. A laboratory ex- 

 periment, by means of cathode rays within a 

 magnetic field, matches the. observed distri- 

 bution of the solar corona, and this is also 

 in harmony with analysis of the sun's physical 

 condition here outlined. The computed sys- 

 tem of ordinary magnetic deflecting vectors 

 and of the large magnetic storms which dis- 

 turb the earth's normal field and fluctuate in 



* See Monthly Weather Review for October, 

 1903. 



t See Monthly Weather Review for July, 1902. 



