212 



KNOWLEDGE 



[September 1, 1890. 



shown. This necessitates the photographing of two 

 ■differently coloured regions of the spectrum on the same 

 plate, and greatly adds to the difficulty of the problem of 

 selecting suitable dyes. 



Practically all that is ordinarily required is the relative 

 wave-length ; from it the absolute wave-lengths may be 

 determine<l when we know the absolute wave-length of any 

 line. One of the most thorough investigations of an absolute 

 wave-length has been made by Mr. Louis Bell, of .Johns 

 Hopkins University. He gives the wave-length of the 

 Dj line as 5890-188 tenth-metres, which exceeds the mea- 

 sure as obtained from Thalen's correction of Angstrom's 

 value by about one-fifth of a tenth-metre. Prof. Rowland 

 has followed this by a table of the relative wave-lengths of 

 about 4.50 standard lines based upon Bell's investigation. 

 He gives the wave-length of — 



C = Go63-042 Dj=5896-1.5G D„= 58yu-188 



K. 1474 = 531G -877 Ei = 5270-497 E.^ = 5260-720 



Aj = 5183-798 A^ = 5172-867 /<= 5169-159 A = 5167-580 

 G = 4293-245 



The spectrum as photographed by Mr. Higgs is on about 

 one-fourth the scale of the enlargements shown in the 

 illustration. Four portions of the spectrum are given on 

 ■each plate. The reader may possibly be confused by two 

 different regions having been placed adjacent to one 

 another, but the line of junction separating the two 

 regions will easily be seen ; they were placed in contact so 

 as to lose no room and get as much of the spectrum on to 

 the page of Knowledge as possible. 



Section 1 represents the Great B group. Most of the 

 lines in this group are certainly telluric, as well as in the 

 curiously similar group Great A, which has been photo- 

 graphed by Mr. Higgs, but is not given in our plates. 

 These groups are so curiously similar that one is tempted 

 to suspect some harmonic relation between them ; but 

 there are no similar groups in other parts of the spectrum, 

 although a region corresponding to much more than two 

 octaves of wave-lengths has now been photographed. The 

 first line in what has been termed the head of the l! group 

 is triple, as seen in Mr. Higgs's negatives, and it can be 

 just recognised as triple in our plates. The tail is a 

 fluting, consisting of several pairs thinning off and mixing 

 •with other solar lines. At 6930 there is a group of five 

 lines, the first and fourth of which belong to the fiutiug, 

 and increase markedly in blackness with a low sun, 

 becoming much more conspicuous than the true solar 

 lines. 



Section 2, the D line region. This portion of the spec- 

 trum has been photographed by Mr. Higgs with dry plates 

 stained with erythrosine and cyanine, which greatly in- 

 creases their sensitiveness to the green and orange region 

 of the spectrum, and has enabled him to photograph all 

 the lines which can be seen with the eye, a result which 

 had not been before obtained. Cyanine adds sensitiveness 

 to the ordinary bromo-iodine dry plates down to the infra 

 red regions. Many interesting experiments have been 

 made with respect to such colouring substances by Mr. 

 Higgs. He finds, for instance, that in the neighbourhood 

 of the A region the sensitiveness may be increased a 

 thousandfold by staining with some of the dyes of the 

 anthracene series. 



This group embraces the 1) lines, and the two so-called 

 rain-bauds (wave-length ,5911 to 5926 and oOlO to 5960) 

 close to it. Most of the lines in the rain-band groups, and 

 certainly ten of the lines between Dj and D„, are due to 

 aqueous vapour in our own atmosphere. With a high sun 

 these atmospheric lines are much reduced in intensity, 

 and some appear to vanish altogether. The spectrum here 



reproduced was taken with the sun from 10" to 15° high. 

 5884 is a double line on the original negative. One of 

 its components is atmospheric, the other solar ; and coin- 

 cident with an iron line ; the same remark applies to 

 5914-3. One of the components of each is consecjuently 

 variable, while the other is constant. 



Midway between the two D lines is a nickel line, but 

 the other lines between the Dj and D, have not yet been 

 matched with laboratory spectra. Mr. Higgs has taken 

 photographs showing the extreme faintness of the telluric 

 lines during a keen frost ; even with a low sun this seems 

 to be strongly confirmatory of their being due to aqueous 

 vapour. 



Photography will enable the position of the lines visible 

 in the spectrum of the setting sun to be registered as 

 never would have been possible by the laborious method 

 pursued by M. Thollon ; their number seems to increase 

 almost indefinitely as the last limb of the sun sinks over 

 a sea horizon. I have on more than one occasion watched 

 the spectrum of the setting sun at sea with a hand spec- 

 troscope, and been reminded of the spectrum of a red 

 star, so entirely does the character of the solar spectrum 

 change owing to the added lines and absorption bands. 



Section 3, the E group and adjacent lines. Peculiar 

 interest attached to this group, as on the original negatives 

 not only is the E line divided but also the line at wave- 

 length 5264-4, which is twice as close. Both Angstrom 

 and Kirchoff give this as being the jiosition of a line com- 

 mon to both iron and calcium ; but the duplicity of the 

 line in the solar spectrum suggests that possibly the two 

 elements have not even a vibration in common. It has 

 been customary with some speculative physicists to speak 

 of two elements having a common line in their spectra, as 

 if they were thus shown to be not elementary, and were 

 proved to be built up of still more elementary substances, 

 one of whicli was thus proved to be common to both of 

 them ; but the fact that some parts of two molecules have a 

 common period of swing, no more proves them to be iden- 

 tical than a violin is proved to be identical with a trombone 

 because they are both capable of sounding the same note. 

 It should rather be taken as proof that under particular 

 circumstances there is some common ratio in the architec- 

 ture of the two molecules, or some similar ratio between 

 stresses and distances, which may be altogether different 

 in the two molecules — but there is no such thing in spec- 

 troscopic analysis as proving two lines to be identical ; a 

 higher dispersion may at any time show that lines which 

 were believed to be identical are really distinct. In this 

 case there is only a dift'erence of wave-length between the 

 two lines which amounts to jryTiTTTTths of the whole wave- 

 length. 



The duplicity of K. 1474 is another case in point. It 

 is shown distinctly double in the plate, Section 3, at wave- 

 length 5316-9. It was at first spoken of as an iron line, 

 and theories were formed as to the existence of this one 

 iron line in the corona. Prof. Young first saw that the 

 dark line was double in the solar spectrum ; and after- 

 wards, during the eclipse of 1870, identified the coronal 

 line with the chromosphere line, which he had already 

 identified with the component of 1474 that does not 

 belong to the iron spectrum, h^ is another case in point ; 

 on the maps of Angstrom and Kirchoff it is marked as a 

 line common to magnesium and iron, but it is now seen 

 to be a double line. 



Section 4, the G region. This is a most difficult region 

 of the spectrum to draw, owing to the vast number of 

 lines it contains ; one can hardly put a pin's point down 

 anywhere than on a line dark or faint, broad or narrow, 

 but the photographic plate registers them all just as readily 



