274 KBPOET— 1880. 



Christofle and Beilstein : ' C. B,.' Ivi. p. 399 (1863). 

 Mulder : ' Journ. f. Prakt. Chemie,' xci. p. Ill (1864). 

 Lecoq de Boisbaadran : ' Specti-es Lumineax ' (1874). 

 Lockyer : ' Proc Roy. Soc' xxii. p. 374 (1874). 



By passing a spark in hydrogen at atmospheric pressure in which 

 phosphorus was heated, Seguin observed a line- spectrum. Pliicker and 

 Hittorf and Salet observed the same spectrum by treating phosphorus in 

 a vacuum-tube like selenium. The spectrum consists of comparatively 

 few lines, which are chiefly situated in the orange and green. 



A hydrogen flame containing traces of phosphorus takes a green 

 colour, and shows a spectrum which was first drawn by Christofle and 

 Beilstein. Mulder makes the interesting remark that a drop of ether in 

 the hydrogen apparatus altogether prevents the formation of this spec- 

 trnm. He tries to explain the fact by assuming that the ether prevents 

 the oxidation of the phosphorus. 



Salet rendered the green flame more luminous by cooling it. The 

 most successful way of effecting this seems to be to surround the tube, 

 through which the hydrogen escapes, by a wider tube, and to blow cold 

 air through this wider tube. This cooling produces a change in the 

 relative intensity of the bands, the red and yellow bands being strength- 

 ened. Lecoq de Boisbaudran gives a careful drawing of this spectrum ; 

 which very likely is due to some compound of phosphorus. 



XIII. Silicon. 



Troost et Hautefeuille : ' C. R.' Ixxiii. p. 620 (1871). 

 Salet: 'Ann. Chim. Phys.' xxviii. p. 65 (1873). 

 Pliicker: ' Pogg. Ann.' cvii. p. 531 (1859). 



A spectrum of silicon may be obtained by taking the jar discharge 

 iDetween poles of silicon. Kirchhoff has thus mapped two bands, one of 

 which, however, according to Salet, is due to lead. Troost and Haute- 

 feuille mention that they have obtained in this way a great number of 

 lines, but their measurements are all given to an arbitrary scale. By 

 passing a spark through the chloride and fluoride of silicon we can 

 eliminate the lines due to the halogens, and obtain a spectrum of lines 

 which must be due to silicon. This has been done by Salet. Salet has 

 also obtained spectra of the hydrogen flame charged with chloride, bro- 

 mide, and iodide of silicon. Most of the bands observed are common to 

 all three compounds, but whether they are due to silicon or to a compound 

 with oxygen or hydrogen is uncertain. Pliicker observed in a vacuum- 

 tube filled with chloride of silicon a band-spectrum, which very likely is 

 due to that compound. Silicon fluoride also gives a band-spectrum under 

 the same conditions. Bromide of silicon, according to Salet, gives a con- 

 tinuous spectrum in a vacuum-tube, when a weak spark is passed through 

 it. A strong spark decomposes the gas, and the lines of bromine and 

 silicon appear. 



XIV. Boron. 



Troost et Hautefeuille : ' C. R.' ixxiii. p. 620 (1871). 



The spectrum of boron was obtained by Messrs. Troost et Hautefeuille 

 by comparing together the spectra obtained between platinum poles in 

 atmospheres of fluoride of boron and fluoride of silicon. . It consists of 

 several groups of brilliant double lines in the yellow, green, and blue. 



