MOUNT WILSON SOLAR OBSERVATORY. 211 



rent lines of this hypothetical whirl would be logarithmic spirals 

 tending more and more to become circles, according as the structure 

 of the hydrogen whirl tended to be more nearly radial. Combining 

 the results of the Director's investigations on the magnetic phe- 

 nomena of sun-spots with the recent work of Mr. St. John on radial 

 motion of the spot vapors, the conclusion is reached that the magnetic 

 field is produced by the motion in the vortex of negatively charged 

 particles. The current lines in this low-lying vortex are orthogonal 

 to the spiral hydrogen flocculi, so that spectroheliograph plates fur- 

 nish the means of determining the form of the current lines. 



Proceeding in this way, Professor Stormer has calculated the 

 magnetic field over a spot vortex. As observation shows that the 

 intensity of the magnetic field decreases rapidly in passing upward, 

 it becomes necessary to choose such dimensions of the vortex as 

 make the results of the calculation harmonize with the observed facts. 

 A discussion of the formulae shows that if the thickness of the vortex 

 is small compared with the diameter of the whirl, and if the inner 

 diameter of the whirl is of the same magnitude as the thickness, then 

 the magnetic force will decrease very rapidly along the axis, as found 

 by the Zeeman effect. For example, if the thickness of the vortex is 

 assumed to be 500 km., the inner diameter of the whirl 500 km., and 

 the outer diameter 50,000 km., other probable assumptions as to 

 the form of the current lines and the charge lead to the result that 

 at the level just above the whirl the force is of the order observed 

 by the Zeeman effect and decreases upwards so rapidly that it can 

 hardly be detected at the level of the hydrogen flocculi. 



A very striking resemblance appears if we compare the field of 

 force projected in a meridian plane with spectroheliograph pictures 

 of calcium prominences around a sun-spot at the sun's limb, such as 

 those reproduced by Slocum in his paper on the attraction of sun-spots 

 for prominences. 



Professor Stormer is continuing the investigation and will soon 

 present the detailed results for publication by the Carnegie Institu- 

 tion of Washington. 



Ultra-violet LiiMit of the Solak Spectrum. 



Mr. van Maanen began in July a series of photographs to deter- 

 mine whether the ultra-violet limit of the solar spectrum is subject 

 to important variations in position. The solar spectrum is photo- 

 graphed once in about 10 days with a quartz spectrograph, used in 

 connection with the Snow telescope. An iron-arc comparison spec- 

 trum is employed to give the wave-length of the limiting region, 

 and a record is kept of tne condition of the sky and of the mirrors. 

 For the sun on the meridian the observed limits have been XX 2973, 

 2970, and 2962, while with a solar altitude of 32° the limit came out 

 X 3006. By making the photographs with the sun at a given altitude 



