Chemistry and Physics. 223 



phide with lead fluoride, preferably in a tube of lead. Bismuth 

 fluoride may also be used but a higher temperature is required 

 for the reaction. The gas has also been prepared by heating a mix- 

 ture of sulphur, phosphorus and lead fluoride, the last substance 

 being added in excess to moderate the reaction. It may also be 

 produced by heating a mixture of arsenic trifluoride with thio- 

 phosphoryl chloride in a sealed tube at 150°. The first method, 

 however, is the best since it is the most convenient and gives a 

 pure product. Thiophosphoryl fluoride is a colorless gas, liquefi- 

 able in the Cailletet apparatus. It spontaneously ignites in con- 

 tact with the air, burning, as it issues from a jet, with a pale 

 yellow-green flame tipped with blue. If a considerable quantity 

 of the gas be exposed to the air, it produces a beautiful blue 

 flash, f ollowed by the yellow-green flame. It is sparingly soluble 

 in water, has no action on mercury, is easily decomposed by heat 

 or by the electric spark, and, heated for some time in a glass tube 

 over mercury, the volume alters, phosphorus and sulphur are 

 deposited on the tube and the resulting gas is silicon tetrafluoride. 

 It is slightly soluble in ether, is insoluble in alcohol and benzene, 

 is completely absorbed by lead peroxide and forms a white solid 

 with ammonia. Heated sodium takes fire in it, burning with a 

 red flame, yielding a residual mass evolving spontaneously inflam- 

 mable hydrogen phosphide on treatment with water. — J. Chem. 

 Soc, liii, 766, Aug. 1888, g. f. b. 



4. On the Relation between the Absorption-spectra of Organic 

 Compounds and their Chemical Composition. — Kettss has ex- 

 amined the spectra of various organic compounds with a view of 

 obtaining a relation between the absorption-spectra of these com- 

 pounds and their chemical composition. Solutions of indigo and 

 anthracene derivatives in chloroform or concentrated sulphirric 

 acid were used, and also solutions of fluorescein derivatives in al- 

 cohol or water. The paper gives the wave-length of the spectrum 

 lines of maximum absorption for each of the sixty-four com- 

 pounds examined. In general, the author finds that the substitu- 

 tion of a methyl, ethyl, methoxyl or carboxyl group, or even of 

 bromine, for hydrogen, moves the absorption lines farther toward 

 the red; the similar substitution of a nitro or amido group pro- 

 ducing a displacement toward the violet. In four cases only were 

 exceptions found to this law : dibromamido-indigo and bromali- 

 zarin being the exceptions to the former, and tetranitro- and di- 

 bromodinitro-fluorescein, both in alcoholic solution, to the latter ; 

 the displacements produced by these substances being exactly 

 opposite to that normally produced. In the case of the last two 

 compounds, indeed, the aqueous solutions are normal in their dis- 

 placements. To some extent also the amount of change which 

 the introduction of any of the above groups into a compound pro- 

 duces in its spectrum, appears to depend upon the nature of the 

 compound itself. If absorption lines be considered as produced 

 by the absorption, by the molecules of a compound, of those ether 

 waves whose period is the same as their own, then the vibration- 



