DE. GLADSTONE ON THE KEFKACTION-EQUIVALENTS OE THE ELEMENTS. 25 
In the oxychloride (Table III.) it seems to have about the same power as in the 
chromium salts, viz. 1 6 *5. 
Mercury. — This metal presents greater difficulties in the estimation of its refraction- 
equivalent than any other in the list, and a comparison of its value, as deduced from all 
its compounds, only increases the difficulty. 
From the Chloride in Water .... Mercury =19*8 
,, ,, Alcohol ... „ = 14*2 
„ Nitrate „ =20-8 
,, Cyanide ,, =16-0 
,, Crystallized Calomel .... „ =22 ’0 
,, Mercuric Methide ,, = 22*7 
These differences are beyond what may be due to errors of experiment. 
There are some other elementary bodies the refraction-equivalents of which may be 
deduced from the observations recorded. 
Tin . — From the Tetrachloride Sn=19'2 
Titanium. — From the Tetrachloride Ti=25*5 
Arsenic. — From the Tetrachloride As=19*9 (line C) 
,, Arsenious Anhydride „ =15 - 7 
,, Sodium Arsenite ,, =15'5 
„ Cacodylic Acid „ = 15*2 
„ Triethylarsine „ = 15*2 
Arsenic acid and some arseniates have been examined, but the results are discordant, 
showing, however, always a lower equivalent than 15. 
Antimony. — From the Terchloride . . . Sb = 3F8 
,, Pentachloride . . . „ =24*5 
Vanadium. — From the Oxychloride . . . V=25*3 
Nitrogen . — From Dulong’s numbers for gaseous nitrogen the refraction-equivalent is 
3*30, but no other means of calculation give so low a figure. The gaseous compounds 
afford the following results : — 
From Cyanogen .... N = 4T8 
„ Nitrous Oxide . . . N=4T6 
„ Nitric Oxide . . . N=3*84 
,, Ammonia .... N = 4T0 
These point clearly to 4*0 or 4T. The hydrogen in ammonia has been taken as 
gaseous hydrogen, viz. 1*5. It is to be remarked that ammonium in the series of salts 
is 11*5; but it is impossible to calculate N from this, as the refraction-equivalent of 
hydrogen is uncertain. Cyanogen in its compounds is 9*1 ; hence the nitrogen is also 
4*1 in this combination. In the nitrates and nitrites, however, it seems to have a greater 
influence on light. 
MDCCCLXX. E 
