158 REPORT — 1900. 



The Absorption Spectra of Ammonia, Methylam,ine, Hydroxylamine, 

 Aldoxime, and Acetoxime} 



It was shown by L. Soret that commercial ammonia, even after many 

 recrystallisations as sulphate, still shows an absorption band. Hartley 

 and Huntington (' Phil. Trans.,' 1879, Part I., 267) contirmed this observa- 

 tion, and, believing the absorption to be due to traces of some constituent 

 of gas-liquor, examined specimens of what was sold as ' volcanic ' ammonia 

 of special purity for analytical purposes. Three separate samples were 

 examined, each measuring half a gallon, with the result that all the rays 

 beyond '/'^ 2638-2 (A 2747'7) were absorbed by the strong solution in acell 

 15 mm. in thickness. A very distinct absorption band was visible on 

 diluting the liquid with eight volumes of water, and was still seen until 

 sixteen volumes had been added. 



This result appeared remarkable in view of the fact that gaseous 

 ammonia, at atmospheric pressure, in a tube 1 metre in length showed 

 no selective absorption, and that ethylamine, even when solutions 

 containing as much as 33 per cent, of the base were examined in cells 

 25 mm. in thickness, transmitted continuous spectra with very little 

 absorption. 



Carbamide also showed no absorption band, but transmitted a con- 

 tinuous spectrum.^ A 10 per cent, solution of carbamide in a cell 15 mm. 

 in thickness transmits all rays to A 2140, rays more refrangible than 

 X 2750 being slightly weakened. 



When we remember that practically all the ammonia of commerce 

 is obtained from coal tar, and is liable to contain minute traces of the 

 volatile bases of the pyridine and other series, which can only be com- 

 pletely separated with great difficulty, it is obvious that great care must 

 be taken to obtain chemically pure ammonia for examination before any 

 trustworthy conclusion can be arrived at as to the character of its absorp- 

 tion spectrum. 



The following investigation was undertaken with the view of definitely 

 ascertaining whether or not chemically pure ammonia shows selective 

 absorption. An examination of ordinary aqueous ammonia was first made 

 in order to determine the exact position of its absorption band. A tube 

 150 mm. long was used. With this thickness of layer a solution containing 

 5 grams of ammonia in 100 c.c. water showed complete absorption beyond 

 ^/A 3638 (\ 2749). A layer of the same thickness containing 2'5 grams 

 of ammonia in 100 c.c. gave a continuous spectrum to '/'^ 3694 {\ 2707), 

 a broad absorption band occupying that portion of the spectrum 

 which lies between '/X 3694 (\ 2707) and ^ jX 4306 (X 2322), the spectrum 

 again showing beyond this point. This band is persistent, being still 

 traceable in a solution containing only 0'625 gram of ammonia in 100 c.c. 

 All the samples of commercial ammonia examined showed selective 

 absorption, but by converting the base into ammonium chloride the 

 absorption band was found to become less marked in the spectrum after 

 successive crystallisations of the salt. 



In order to^try the eiiect of crystallisation of one of the less soluble 

 salts, ammonia was converted into oxalate and the salt repeatedly 



' Hartley and Dobbie, Trmis. Chem. Soc, vol. Ixxvii. 



2 From unpublished experiments on the determination of aromatic substances in 

 urine. See note, Dublin Journal of Medical Science, June 1882. — W. N. Haetlet. 



