Ole ne eee 
Chemistry and Physics. 487 
and may be regarded as sensibly constant. Moreover, with differ- 
ent compounds this ratio only varies between limits but little 
ena: from each Sti Thus we have for the mean value of 
this rat 
Substance. Ratio < 
Obloride-of urawiumy 2. <i. 00.4 St os 0-000081 
Chlor. of uranium and potassium, _.....- 83 
Fluor. of uranium nr potassium, joi 81 
Sulphate: of wrewium. oi... 5520. 232. 2 85 
Sulphate of uranium si potassium, . . --- 84 
Oxplate of: ureniane: 045 pies 8 86 
Phosphate tn uranium and ‘sem Bt ge 82 
Nitrate of ura DOM, 5 ss Gk cas we 88 
_ Arsenate of uranium, os 2 ss en Sok 83 
(6.) There does not appear to be es simple relation between the 
wave-lengths corresponding to the homologous lines or bands of 
the same luminous group in different compounds and the chemical 
ht gd of these de leesy es. 
7.) When we illuminate the solid compounds of uranium with 
transmitted violet or nite vie et light, we observe in the most 
differ for each compound, and which appear to correspond in this 
part of the spectrum to less refr. igible’ groups of bo Peres 
7 bands, and to continue their succession.— Comptes Rendus, 
XXV, 296 
n the e spectrum of the Aurora Borealis.—V ocr. has ‘made 
an tiem to identify the spectrum of the aurora with that of air, 
and has arrived at results which, if absolutely tnt scl 
render the identification at least proba e author employed 
a direct vision spectroscope with 5 prisms, “soliimator and eer rving 
telescope, which last by means of a micrometer ond could be 
moved so as to oe different parts of the spectrum into the center 
Wave-length. Probable error. 
0°0006297 00000014 = Very nang? band. 
00005569 0-0000002 —_ Brightest line. 
00005390 Very faint line. 
0°0005233 070000004 ciate ot tiss se Bowes 
right where the 
stcscdaet er Sh vamos peste otherwise faint. 
00005004 =. 00000003 ~~ Very bri 
Broad ae somewhat less bright 
00000003 inthe center; very faint w 
0°0004694 
the red line appears. 
0°0004663 + 
0°0004629 
