J^8 ON COLOURS OF THIN PELLICLES. 



only a given quantity of nitrogen as a liquid, and tliat, if it 

 be taken thus saturated, it constantly skives the precise pro- 

 portion of 'Ql in close vessels, no objection can be brought 



But objection- ajrainst its use ; while a very stroni^ oue may be urged 



able, as boiled " . . , . , ' 



■water absorbs against the use of the electric spark, since either boiled wa- 

 ■omeoftheox- ter is employed, and then it willabsorb a small portion of 

 *d *nb 1 d ^^^ oxigen very g'eedily, or the water will be saturated with 

 jiTcs out air. air, and then the pressure occasioned by the detonation will 

 always lorce out nome bu!)bles. One or other of these in- 

 conve ieiices appears to me unavoidable ; and the latter in 

 particular oiten teamed me, in the numerous experiments I 

 had occjisiou to inake on the analysis of gasses by the elec- 

 tric spark, cither with Tiienaid, or alone, on the Alps* It 

 is true tVe diherences hence arising, when we operate with 

 care, amouiit only to some thousandth parts : but it is of 

 thou^alidth parts we are speaking; and if there be another 

 process, wiiic^h gives at least equal if not supTerior accuracy, 

 with less trouLde, it appears to me to deserve the prefer- 

 ence. 

 Air merowded Finally, to return to the memoir of Mr. de Marty, I shall 



cKnrchesand add, that he has equally tried the air contained in theatres 



theaties not 



deficient in ox- aod m churches, when a great concourse of persons were 



igen. assembled in them, and that he constantly found the same 



quantity of oxigen ; an experiment which von Humboldt and 



Gay-Lussac made likewise at Paris. 



xiy. 



Summanj Considerntmis on the Prismatic Colours of Bodies 

 reduced to thin Pellicles; with an Explanation of the Co- 

 lours of Amiealed Steel, and those of the Peacock's Fea,- 

 thers. A Fragment of a Work on Colours: by C, A. 

 Prieur*. 



Tliin substan- ^ ERTAIN extremely thin substances, the thickness of 



CCS exhibit ^j^jch varies proj^ressively from one part of them to another, 

 prismatic CO- i ?? j • n ^ p i- cf 



lour^. exhibit, as is well known, a series oi colours ot ditierent 



• Annales de Cliimie, Vol. LXI, p. 154, Feb. 1807. 



tintSa 



