December 29, 1911] 



SCIENCE 



897 



known as the Doppler effect and had been 

 well established for sound waves (a loco- 

 motive whistle appears of higher pitch 

 when approaching and lower when reced- 

 ing) but it was only confirmed for light by 

 Huggins and by Vogel in 1871, by the ob- 

 servation of displacements of the solar and 

 stellar spectral lines on observing in suc- 

 cession the advancing and the receding 

 limb of the sun. 



It may be worth while to indicate the 

 accuracy necessary in such measurements. 

 The velocity of rotation of the sun's equa- 

 tor is approximately two kilometers per 

 second, while the velocity of light is 300,- 

 000 kilometers per second. According to 

 Doppler 's principle the corresponding 

 change in wave-length should be 1/150,000 

 — a quantity too small to be "resolved" by 

 any prism or grating then in existence. 

 But by a sufficient number of careful 

 micrometer measurements of the position 

 of the middle of a given spectral line, the 

 mean values of two such sets of measure- 

 ments would show the required shift. It is 

 clear, however, that if such radial veloci- 

 ties are to be determined with any consid- 

 erable degree of accuracy, nothing short of 

 the highest resolving power of the most 

 powerful gratings should be employed. 



Another extremely important application 

 of spectroscopy to solar physics is that 

 which in the hands of Hale and Deslandres 

 has given us such an enormous extension 

 of our knowledge of the tremendous activi- 

 ties of our central luminary. 



The speetroheliograph, devised by Hale 

 in 1889, consists of a grating spectroscope 

 provided with two movable slits, the first 

 in its usual position in the focus of the col- 

 limator, and the second just inside the 

 focus of the photographic lens. A uniform 

 motion is given to the two slifs so that the 

 former passes across the image of the solar 

 disc, while the other exposes continually 

 fresh portions of the photographic plate. 



If the spectroscope is so adjusted that 

 light of the wave-length of a particular 

 bright line in a solar prominence (say one 

 of the hydrogen or the calcium lines) passes 

 through the spectroscope then a photograph 

 of the prominences, or sun spots or faculae, 

 etc., appear on the plate. But the character 

 of this photograph depends on the portion 

 of the bright spectral ' ' line ' ' which is effec- 

 tive, and as the entire range of light in 

 such a line may be only a thirtieth part of 

 the distance between the sodium lines, it 

 would require a resolving power of at least 

 100,000 to sift out the efficient radiations 

 so that they do not overlap. 



As another illustration of the importance 

 of high resolving power in attacking new 

 problems, let us consider the beautiful re- 

 sults of the investigations of Zeemann on 

 radiation in a magnetic field. The effect 

 we know is a separation of an originally 

 simple radiation into three or more, with 

 components polarized at right angles to 

 each other. This is one of the very few 

 cases where it is possible to actually alter 

 the vibrations of an atom (electron) and 

 the fact that the effect is directly calcul- 

 able, as was first shown by Lorentz, has 

 given us a very important clue to the struc- 

 ture and motions of the atoms themselves. 



The experiment is made by placing the 

 source of radiation (any incandescent gas 

 or vapor) between the poles of a powerful 

 electromagnet and examining the light 

 spectroseopically. Now this experiment 

 had been tried long before by Faraday but 

 the spectroscopic appliances at his disposal 

 were entirely inadequate for the purpose. 



Even in the original discovery of Zee- 

 mann only a broadening of the spectral 

 line was observed, but no actual separa- 

 tion. In fact, the distance between com- 

 ponents which had to be observed was of 

 the order of a hundredth of the distance 

 between the sodium lines, and in order to 

 effect a clear separation and still more to 



