1900] on Solar Vortices and Magnetic Fields. 623 



and suspect it to be due to a magnetic field, we must apply the test 

 for circular polarisation. 



The simplest means of testing for circularly polarised light is to 

 transform it into plane polarised light by passing it through a quarter- 

 wave plate or a Fresnel rhomb. In the case of a Zeeman doublet, we 

 would then have issuing from the rhomb the light of the two compo- 

 nents, polarised in planes at right angles to one another. A Nicol prism, 

 standing at a certain angle, will transmit one of these plane polarised 

 beams and cut off the other. Turning the Nicol through {)(f will 

 cause the component previously cut off to be transmitted, and the 

 other to be stopped. 



Consider a sun-spot at the centre of the solar disk, and suppose 

 it to be produced by a vortex, the axis of which lies on the line 

 passing from the eye of the observer through the spot to the centre 

 of the sun. Under these circumstances, if a strong magnetic field is 

 produced by the vortex, the spectral lines due to vapours lying 

 within this field should be widened or transformed into dou])lets. 

 Moreover, the light of the components of these doublets should be 

 circularly polarised in opposite directions. This Avould be true if 

 the spot vapours were emitting bright lines, identical in character 

 with those emitted by a radiating vapour between the poles of a 

 magnet. The experiments of Zeeman, Cotton, Konig, and others, 

 show, however, that dark lines, produced by the absorption of the 

 spot vapours, should behave precisely in the same way as bright lines. 



The spectrum of a sun-spot was observed for the first time by 

 Lockyer in 1866. He found that many of the lines of the solar 

 spectrum were widened where they crossed the spot, and the obser- 

 vation of these widened lines has been carried on systematically by 

 many observers ever since. Conspicuous among these observers was 

 Young, whose last observations were made with a powerful grating 

 spectroscope attached to the 28-inch Princeton refractor. This instru- 

 ment showed that some of the spot lines are close doublets. Dr. 

 AValter M. Mitchell, who at first worked in conjunction with Professor 

 Young and later by himself, gave special attention to these double 

 lines, which he found to be particularly numerous at the red end of 

 the spectrum. He called them "reversals," and the existing evidence 

 favoured the view that they were produced by the radiation of a 

 hotter layer of vapours overlying the spot, which would give rise to a 

 narrow bright line at the centre of the widened dark line. True 

 reversals of this kind actually seem to occur in the case of H and K 

 and other lines in the spot spectrum, and it was therefore natural 

 that Mitchell should attribute the similar phenomena of the spot 

 doublets to a similar cause. It was generally supposed that the 

 widening of the dark lines was due to the increased density of the 

 spot vapours. The diverse character of the lines in the sun-spot 

 spectrum is w^ell illustrated by this drawing, which is due to Mitchell. 

 In addition to the ordinary widened and " reversed " lines, we find 



