NA TURE 



[November 4, 1909 



occupied this region of tlie sun, but its form was in no graph of this series was taken, the extension had almost 



wise remarkable, and afforded no evidence of the phet 

 niena brought to light by the Ha photograph. The struc- 

 ture recorded with the aid of the latter line (Fig. 3) recalls 

 Langley's sun-spot drawings, and suggests the operation of 

 some great force related to the sun-spot group. The same 

 cyclonic structure had been less satisfactorily recorded on 

 the previous day, but a comparison of the two photographs 



(Ha) Flocculi 



failed to indicate such changes as motion along the 

 apparent stream lines might be supposed to produce. 



The close of the rainy season now permitted an active 

 study of the Ha flocculi to be undertaken. Many photo- 

 giaphs wete made daily, and the almost constant associa- 

 tion of apparent cyclonic storms or vortices with 

 sun-spots became evident. During several months ^ 



of the year in California an unbroken succession of 

 •clear days can be counted upon, so that the changes 

 of a given vortex can be followed without interrup- 

 tion. The cyclonic storms were found to be of two 

 principal types, the first associated with groups of 

 spots and represented in such photographs as those 

 of .^pril 30 and September 2, the second associated 

 with single spots, and resembling a simple vortex, 

 as illustrated in the photographs of September q and 

 October 7, igo8 [Fig. 4). The appearance of these 

 simple vortices is such as to indicate rotation in a 

 clockwise direction in the southern hemisphere, and 

 in a counter-clockwise direction in the northern 

 hemisphere (assuming the direction of motion to be 

 inward toward the spot). However, this cannot be 

 taken as a general law, corresponding to the law of 

 terrestrial cyclones. Indeed, many instances have 

 been found of closely adjoining spots, in the same 

 hemisphere and frequently in the same spot-group, 

 having magnetic fields of opposite polarity, produced 

 by vortices rotating in opposite directions. 



In some cases, at least, these vortices seem to 

 exercise a powerful attraction on the surrounding 

 gases, as a series of photographs taken on June 3, 

 1908, illustrates, k long dark hydrogen prominence, 

 first photographed in elevation at the sun's limb on p^^ 

 May 28, had advanced half-way across the solar disc. 

 It lay at the outer boundary of a well-defined vortex, 

 centred on a sun-spot. This spot had been gradually 

 separating into two parts, and on June 3 the separa- 

 tion was complete. The first photograph of a series 

 of nine was made on this day at 4h. s8m. Several 

 successive photographs indicated no appreciable change, 

 but one taken at sh. 07m. showed that the prominence 

 was developing an extension toward the spot. .'Xt 5h. I4ni. 

 this had assumed the appearance illustrated in the next 

 photograph, and eight minutes later, when the last pholo- 



NO. 2088, VOL. 82] 



reached the spot. It will be seen that it divided into two 

 parts, which indicates that each umbra was a centre of 

 attraction. The average %elocity of the motion toward the 

 spot was more than 100 km. per second. Later photo- 

 graphs, made on the following days, show a ring of bright 

 hydrogen surrounding the spots, suggesting that the com- 

 paratively cool hydrogen carried down into the spots was 

 re-heated and returned to the surface, after escaping 

 from the lower end of the vortex. \\'e thus seem to 

 be observing some of the phenomena of an actual 

 vorte.x in the sun ; but it must not be supposed that 

 cases of this kind are common. In many instances 

 the hvdrogen flocculi do not appear to move rapidly 

 toward or away from spots, but undergo changes 

 of intensity, as though the physical condition of the 

 gas were constantly changing ; but before proceeding 

 further with a discussion of these sun-spot vortices, 

 let us turn to another phase of the subject, which 

 will afford much new information indispensable for 

 this purpose. 



We are all familiar with the effect produced by 

 passing an electric current through a wire helix. 

 The lines of force of the resulting magnetic field are 

 parallel to the axis of the helix, and its intensity is 

 determined by the diameter of the helix, the number 

 of turns of wire, and the strength of the current. 

 W'e also know, from Rowland's experiment, that the 

 rapid revolution of an electrically charged body will 

 produce a magnetic field. Thus, if a sufficient 

 number of electrically charged particles were set into 

 rapid revolution by the solar vortices, a magnetic 

 field should result. What warrant have we for 

 assuming the existence of charged particles in the 

 sun, and how could such a field be detected? 



Let me pass rapidly in review a series of pheno- 

 mena with which you are all familiar. Sir William 

 Crookes showed in this lecture-room, so long ago as 

 that the negative pole of a vacuum tube sends out a 

 stream of particles, capable of setting a light windmill in 

 rotation, and deviated from their straight path when under 

 the influence of a magnetic field. He has kindlv consenf'l 

 10 show the same tube again to-night; you now see the 



1S79, 



efl^ect upon the screen. The recent work of Sir Joseph 

 Thomson and others has proved that these are negatively 

 charged particles, called " corpuscles " or " electrons," 

 and that their mass is about 1/1700 of the mass of an 

 atom of hydrogen. Moreover, Thomson has shown that 

 at low pressures these corpuscles are given off from a hot 

 wire or from the carbon filament of an incandescent lamp. 

 He has also demonstrated that this property of emitting 



