io8 



NA TURE 



[May 30, 1895 



first line of the triplet should be. But this one is out of 

 the question on account of its enormous cncrj;y, which | 

 would be quite out of comparison with the other lines. | 

 So we must suppose that the first line of the triplet is con- ; 

 cealed by the strong line. Indeed, on the plates Prof. ] 

 Kayser and 1 have examined, it would be impossible to , 

 detect a line close to 2S52. .Ati-tin' as in the spectrum 

 of thallium, these triplets form two series i.scc Fig. s"), 

 and again we find that the . wave-numbers of the first, 

 second, and -third lines in each series are vcr\- accurately 

 represented by'a fomiula. 



.A - 15/7 



C« -', 



n standing for the row of entire numbers. For each series 

 there are three values of .A, but only one vahie of B, and 



sponding rays have not been identified with certainty. 

 There are many lines: beside those forming the triplets. 

 For magnesium, the triplets contain 33 out of 56 lines, 

 for calcium 33 out of 106, for strontium 29 out of 97. We 

 have found that, as a rule, the higher the melting point of 

 an element, the greater is the percentage of lines in the 

 arc spectrum that do not belong to the series. From 

 magnesium, to calcium, and from calcium to strontium, 

 the triplets widen and shift to the less refrangible side of 

 the spectrum. The same thing happens in the spectra of 

 other groups. of chemically-related elements, the difference 

 of wave-numbers of the doublets or triplets being some- 

 what proportional to the square of the atomic weight. 



There is one more feature which seems interesting in 

 regard to the connection of the spectra of different 



W IBI 201 gal g*l 261 281 301 3g| Ml 36| 



481 50| ' 



one value of C. The three values of .A are very nearly 

 the same in both series, indicating that the ends of both 

 series coincide. The lowest number for which the formula 

 gives a positive value is /; = 3. To this value corresponds 

 the strong green triplet. But in the other series the 

 corresponding triplet ought to be found near 13000 10-* 

 cm. where photographic methods fail. It may be that it 

 is identical with the lines that Bccquerel has found near 

 iiooo and 12120, the first of which, he says, is possibly 

 double. The deviation between these and the calculated 

 values is not so very great, considering the wide extra- 

 polation of the fonnula. .\ small change in the value of 

 C would alter the formula much more for // = 3 than for 

 the higher values of //. Besides, we believe the formula 

 only to be an approximation to the true function which 

 may be developed into a series of descending powers of 



elements. In all the formuho of scries that have liccn 

 observed, the coefficient of «"-' does not varj' more than 

 about 10 per cent, from its mean v.alue, if we except one 

 of the two series of doublets in the spectrum of aluminium 

 where the variation is somewhat larger. 1 think, when 

 in some time a satisfactory theoretical explanation of the 

 symmetry in the spectra of the elements will be given, 

 this co-efficient will pro\e to be an important physical 

 constant. C. RuNc.K. 



KARL VOOT. 



THE life of Prof Karl V'ogt, who died on May 6, was 

 no tranquil scientific career, for he was a fight- 

 ing philosopher. He first comes into notice in 1839, 

 working with Agassiz, then Professor at Neuchatel, on 



■ A l«l 151 If! 171 rel I9| iOl 2l| 22l 231 g4| 2Si 281 27! 28| gpj 30l 3l| 321 331 34! 351 36l 37| 3b] 



Mg. 



Ca. 



Sr. 



I 



I 



/;-. Il iliis IS Ml. the neglected terms would affect the 

 values of the formula much more for the low values of ;/, 

 than for the higher ones. The separation of the triplets 

 into two scries is not only suggested by the .symmetry of 

 the distribution, but also by the aspect of the lines. 



In the spectra of calcium and strontium, we also find 

 triplets with the same differences of wave-numbers, and 

 their appearance teaches us in each spectrum to separate 

 them into two scries. We then sec that the distribution 

 of triplets shows a remarkable similarity to that in the 

 spectrum of magnesium. 



The dotted Imcs in the figure mean that the corrc- 



the "Freshwater FishesofCcnlr.il Europe." This great 

 work, never completed, determined the direction of Vogt's 

 best research during the rest of his long life. It was only 

 in 1888-94 that t'lc " Traite dWnatoniie Compar('e," by 

 \'ogt and Jung, was published in Paris, taking high rank 

 as a standard authority, and likely to retain it. He 

 returned from Paris to his native town of ("licssen, where 

 hehad been a))poinled Professor. But the revolution of 

 184S soon burst forth, and we hear of him as an advanced 

 Democratic Deputy contending for liberty and progress 

 with the trenchant orator)' he could use alike in politics and 

 science Political forces were too strong against liini. ;ind 



