1 880.] Modifications of the Spectrum of Potassium, Sfc. 



279 



which, at Dr. Gramgee's request, he had made for him: see "Pro- 

 ceedings," vol. xxx. p. 113: — "I have examined spectroscopically 

 for potash a sample of protagon furnished me by Dr. Gramgee, 

 and labelled ' Protagon, twice recrystaliised, Blankenhorn.' I could 

 not detect any potash by the spectroscope in the incinerated mass 

 from O'l grm. of substance. With the carbonised mass obtained from 

 1*0 grm. of substance I obtained the potasssium line {a) very faintly, 

 and from comparative experiments with a dilute solution of a potas- 

 sium salt I estimate the quantity of potash in 1 grm. of the substance 

 not to exceed mgrm. The carbonised residue of L grm. of protagon 

 was carefully oxidised with pure nitric acid, when a small quantity of 

 fused metaphosphoric acid remained after ignition. The residue 

 weighed 0'0278 grm., corresponding to 1*08 per cent, of phosphorus. — 

 (Signed) H. E. Roscoe." 



As regards the bearing of Professor Roscoe's evidence on the 

 question that has been raised before the Royal Society, and as regards 

 some other relations of the same question, I beg leave to submit to the 

 Royal Society the following observations : — 



1. On the Modifications of the Spectrum of Potassium which are 

 effected by Phosphoric Acid. 



It is well known in spectroscopy that the best spectra of metals are 

 obtained in the easiest manner by aid of their most volatile com- 

 pounds, such as chlorides ; and, on the contrary, that salts which 

 are more or less fixed at high temperatures, such as the silicates and 

 phosphates of alkalies and earths, give either no spectra at all or only 

 very feeble spectra. Thus, if a minute particle of potassic bromide 

 be ignited in the aerated gas flame before the slit of the spectro- 

 scope, the spectrum of potassium is at once produced in the most 

 brilliant manner : particularly the line in red is brilliant and sharply 

 defined, while the gas flame to the naked eye shows the violet colour- 

 ing of potassium. But when a similar particle of potassic phosphate is 

 ignited under similar conditions, the flame, instead of showing the 

 violet colour of potassium, shows a greenish- yellow colour produced 

 by volatilised phosphoric acid : ani in the spectroscope, at first, only 

 the continuous spectrum of phosphoric acid (not to be confounded 

 with the continuous spectrum of potassium) is seen, till gradually, and 

 at a white heat, a feeble potassium line in red, with indistinct edges, 

 is perceived. 



Even a large bead of pure potassic phosphate, when ignited before 

 the slit of the spectroscope, never produces, even at white heat, any such 

 intense red potassium line as the smallest bead of potassic chloride, 

 potassic bromide, or potassic nitrate. And, in contrast with the latter 

 salts, which volatilise quickly, the phosphate takes some time before 

 it is completely volatilised. 



u 2 



