August 19, 1897] 



NATURE 



365 



THE APPROACHING TOTAL ECLIPSE OF 



THE SUN.^ 



IV. 



nn HE programme of work to be attempted in the Indian 



-^ eclipse of next year, referred to in the last article, 



carries me back very vividly to the eclipse of 1871, also 



observed in India. The shadow path of the eclipse of that 



year also cut the west coast of India, but at a much more 



southerly point than Viziadurg. The coast station was 



then Baikal, and from this point the shadow swept over 



the land in a south-east direction, as shown in the 



accompanying map (Fig. 13). 



The retrospect is very encouraging, for one is reminded 



By 1872 the influence of quantity or density had been 

 made out ; when experiments were made at one tempera- 

 ture the spectrum got simpler as the quantity was 

 reduced, so that the spectrum was finally reduced to its 

 longest line.i 



I am glad to see that Sir William Huggins, who 

 appears to be ignorant of my quarter-of-a-century-old 

 work, has quite recently arrived independently at the 

 same conclusions. 



Next came the influence of temperature. This was a 

 much more difficult problem to tackle, for the reason 

 that enormous changes in the spectrum of each chemical 

 substance were brought about by changing the tempera- 

 ture conditions ; but finally the association of certain 



OF MOON IN TOTAL ECLIPSE \ "^ 



December 12 T-? 1871. 



of the enormous advance in our knowledge of the sun 

 since that time ; and more than this, we have now the 

 supreme advantage that eye observations have been almost 

 entirely superseded by permanent photographic records. 

 The accompanying view (Fig. 14) of my observatory at 

 Baikal in 1871 will show that eye observations of the 

 spectra alone were attempted. 



It must also be remembered that none of the laboratory 

 experiments, referred to in the last article, had then been 

 made. 



Now it was to try to understand such hard solar facts as 

 those referred to in the last article, facts since observed 

 carefully in all their detail, month after month and year 

 after year, that much of my early experimental work was 

 undertaken. 



' Continued from page 321. 



NO. 145 1, VOL. 56] 



lines with certain temperatures was accepted by every- 

 body, though as to the why there were and are contending 

 schools of opinion. 



In connection with certain stellar problems awaiting 

 solution, I have recently been compelled to return to this 

 question, and I have used a more powerful current and 

 larger jar-surface than that I formerly employed ; and, 

 further, the recent work carries the results into the photo- 

 graphic region. The result is important, since the old 

 results have been confirmed and extended. To deal with 

 the case of iron, seven additional lines in the spectrum 

 have been found to have their brightness enhanced at 

 the highest temperature. 



These, as well as the two previously observed, are shown 

 in the following table, which also indicates the behaviour 



1 Phil. Trans., 1873, pp. 253 and 639. 



