April 1, 189G.] 



KNOWLEDGE, 



8*r 



a vessel provided with stopcocks, so that a small quantity 

 could be drawn off from time to time in order to fill the 

 vacuum tubes. Tiie vacuum tubes were subjected to the 

 strictest discipline before the cleveite gas was admitted to 

 them, being completely exhausted, dried at (he air pump, 

 heated as strongly as they would bsar, and a strong induction 

 current passed through them; this treatmont being repeated 

 until the spectrum of hydrogen was no longer shown, or, if 

 shown, only feebly. 



The results obtained were well worth all the trouble and 

 time expended so ungrudgingly on the preparation of the 

 tubes. The wave-lengths of the lines of heUum were 

 determined mainly by the measurement of photographs, 

 the lines of sodium which were g'ven by the vacuum tube 

 itself when the glass got sufficiently heated, and those of 

 iron from a pair of electrodes placed close to the slit, being 

 the chief points of reference. 



The spectrum of the gas from cleveite, as worked out at 

 fullest detail iu this most thorough and painstaking manner, 

 proved at first sight one of great irregularity, and had its 



hydrogen, now known as Hi, B.3, and HJ, and which give 

 rise to the three Fraunhofer lines C, F, and h, were respec- 

 tively the twentieth, twenty-seventh, and thirty-second 

 harmonics of a fundamental vibration whose wave-length 

 in vacuo was 0-131'27714 of a millimetre. But the third 

 line of hydrogen, H-/, — " near G," or G', as it is generally 

 called, — had no place in this arrangement, and attempts to 

 fit similar formula? to other spectra ended only in dis- 

 appointment. 



At that time our knowledge of the hydrogen spectrum — 

 which, as one of the simplest of all spectra, and as given 

 by the lightest of gases, appeared specially marked out for 

 this investigation^did not ascend beyond the violet. The 

 "deep things out of darkness" which the unseen recrions 

 of the ultra-violet were to yield were not then disclosed. 

 But when Dr. Huggins' photographs of stellar spectra* - 

 revealed the long series of lines in the ultra-^•iolet in stars 

 of the first type, it became an almost irresistible inference 

 that these were the completion of the hydrogen spectrum. 

 They made, with the four lines which we knew of old, a 





"1 — r 



?;«' 



JiY-DKO C^J^. 



3.'o' 



/./ TH I U M 



soot U M. 



£C30 



7c oo 6aoo 



&S00 \ SSpo 



fl 



I:,/r«-Xta R«t C.-Hnjt ViCr G-n«.-v Slut V.olsi Vj t ."a. - Vi olc t 



discovery fallen some ten or a dozen years earlier, investi- 

 gators would in all probability have been content to accept 

 it as observed, and to enquire no further. 



But we bad advanced beyond this stage of spectroscopic 

 work. Even in the infancy of spectrum analysis it was felt 

 that there ought to be some close connection between the 

 undulations indicated by the different lines of a glowing 

 gas ; that they must be related to each other by some 

 unfelt rhythm. They seemed like the broken snatches of 

 distant music borne to our ears by fitful gusts of wind ; 

 they were parts of a complete harmony to the knowledge 

 of which we had not yet attained, but we were assured 

 that if it could but reach us in its fulness each of these 

 detached and separate vibrations would find its proper 

 place, and show itself as an integral part of a beautiful and 

 perfect whole. 



It was with some kindred thought to this that in ls71 

 Prof. Johnstone Stoney, in a paper communicated to the 

 Royal Irish Academy,* suggested that the three lines of 



series so obviously rhythmical, so obviously one, that the 

 conclusion could hardly be avoided. 



A further and an important step forward was due to a 

 study undertaken by Prof. Piazzi Smyth. Prof. Smyth 

 was in the habit of drawing his spectra and expressing 

 the positions of his lines on a scale of " wave-numbers to 

 an inch," instead of, as usual, on a scale of wave-lengths ; 

 and in particular had made a fine map of the great green 

 band of the beautiful spectrum of carbonic oxide. This 

 map he submitted to the inspection of Prof. A. S. Herschel 

 in November, 1883, who pointed out that the "wave- 

 numbers " of these lines formed two perfectly similar 

 arithmetical series, so placed with respect to each other 

 that the fifth line of the first series was coincident with 

 the first of the second. Further, the ninth line of the 

 first series showed as a close doublet, and the tenth and 

 following lines as doublets, each slightly wider than the 

 preceding one. I 



• Proc. Boi/. Irish Acad., Jan. 9th, 1S71. 



• JPhil. Traat.. Vol. CLXXI , 18S0, p. titjl). 

 t Trans. Soy. Soc. Jrfiii., XXXIl., 1884. 



