May, 1915. 



KNOWLEDGE. 



141 



In potassium the common convergence-frequency 

 of the two subordinate series is 21967-5 for the 

 less refrangible lines, and 22025-37 for the more 

 refrangible lines of the pairs. Subtracting these 

 numbers from 35005-56, the convergence-frequency 

 of the principal series, we get 1 3038-06, and 1 2980-1 9 

 for the two P 2 red lines ; and these numbers 

 correspond to the wave-lengths 7667-77 and 7701 -96 

 respectively. The observed values of these lines 

 are : 



7668-54) T . 



7701 •9'> Lehmann, in the arc. 



7665-61 ,, , „ 



7699-3/ Jva y ser an " Runge, in the arc. 



7699 ! ^er an d Valenta, in the flame. 



For rubidium the mean convergence-frequencies 

 of the subordinate series are 20872-7 and 21 1 10-68 

 and Poo = 33688-20. These numbers give 



P2=jnnip.5 )> the observed values being: 

 1 7947-7 ( ^er an< ^ Valenta," in the arc. 

 7nf7 a" Saunders, in the arc. 



I 7947-6J 



For caesium the corresponding numbers are 

 19672-9 and 20226-86 and Poo =31404-6. These 



numbers give P 2 =i omT.-v;, > the observed values 



\ 8943-96 J 



Lehmann, in the arc. 



, . | 8527-50 | 

 «*ing l8949 .G7l 



Here the correspondence is not quite so good : 

 a result to be expected perhaps from the difficulty 

 of determining accurately the wave-lengths of 

 lines lying far in the infra-red. 



The observation that the pairs of lines in the 

 subordinate series have the same difference in 

 oscillation-frequency is important. In the case 

 of sodium the interval 17-2 occurs twenty-three 

 times; in potassium the interval 57-9 occurs 

 thirteen times, and the interval 238- occurs sixteen 

 times in rubidium. These are instances of what is 

 known as the " law of constant differences," 

 discovered by Hartley, who observed (in 1883) 

 that three triplets in the spectrum of zinc corre- 

 sponded with three triplets in the spectrum of 



cadmium, and that in each of these spectra the 

 triplets have their lines equally spaced if mapped 

 on the scales of oscillation-frequency. The photo- 

 graphic reproduction of Lecoq de Boisbaudran's 

 drawings of the spectra of zinc and cadmium shows 

 these triplets plainly (see Figure 123). Such triplets 

 characterise the spectra of magnesium, calcium, 

 strontium, zinc, cadmium, and mercury, the " scale " 

 on which the triplets are spaced being proportional, 

 more or less nearly, to the squares of the atomic 

 weights. 



In the reproduction of Lecoq dc Boisbaudran's 

 spectra of zinc and cadmium the first triplets of 

 the sharp series are marked «, y, 8 and a, /3, 6 

 respectively. The wave-lengths and oscillation- 

 frequencies [in vacuo) of these lines are : 



If these differences are proportional to the squares 

 of the atomic weights, we can calculate that of 

 zinc from that of cadmium, namely. 112-40; 

 making the calculation we obtain for zinc the number 

 65-358, whereas the atomic weight of zinc deter- 

 mined by chemical methods is given in the Inter- 

 national Table for 1915 as 65-37. 



The diagram which accompanies this article 

 (see Figure 124) is a reproduction of the diagram 

 given by Kayser and Rungc in their paper of 1890, 

 and shows the spectra of the alkalies as known 

 at that time. The spectra are mapped upon the 

 scale of oscillation-frequencies, and the scale of 

 wave-lengths is also shown under each element. 

 The complete spectrum i- shown in the top line. 

 the next line shows the principal series, the third 

 line shows the diffuse series, and the fourth line 

 the -harp series. The scale of the drawing is too 

 small to allow of the pairs of lines being properly 

 represented : they are shown too far apart. 



[To be continued.) 



KHAKI DYH. 



The present shortage in this country oi synthetic yellow 

 dyes has put considerable difficulties in the way of manu- 

 facturers of khaki cloth. A temporary way out of the 

 trouble was found, however, by the increased u 

 fustic, a natural yellow dyestufl, consisting of the wood o) 

 a tree {Chlorophora inutoria) which grows freely in Jamaica 

 and also in British Honduras. At the outbreak of the 

 war there was only a small supply available in Europe, 

 though fortunately one of the prizes captured from ( lermanv 

 had a considerable consignment on board. A certain 

 amount, too, has been obtained from France, which country 

 and the United States have hitherto taken the bulk of the 



fustic exported from Jamaica, When the shortage of 

 yellow dyestuffs in tins country lirst became apparent the 

 imperial institute took steps to place British dye tinns in 

 touch with exporters oi fustic in Jamaii a Only a moderate 



amount ol CUl fustic wood was, as it happened, then avail- 

 able in the island, but as a result "t the Institute's action 

 the Government >>t Jamaica has offered to purchase from 

 the growers further supplies, and carry these, at Govern- 

 ment cost, to Kingston, the pert .>t shipment. The Govern- 

 ment of British Honduras is also taking action in this 

 matter, and a further supply of the wood may possibly be 

 forthcoming from that colony. 



