22 DR. GLADSTONE ON THE RETRACTION-EQUIVALENTS OE THE ELEMENTS. 
Iodine. — Solutions of iodine have hitherto given results which are not comparable 
among themselves. 
From Iodide of Methyl . . . . Iodine =24 ’6 
„ „ „ Ethyl .... „ =24-3 
„ „ „ Propyl .... „ =24-9 
„ » „ Amyl .... „ =24-3 
The average of these is 24'5 ; Haagen gives 24-87 for the line C, as deduced from the 
same series of compounds. 
It would appear that the differences between the three halogens are as follows : 
Br=Cl + 5'4, and 1=01+14-6. This does not exactly agree with the differences be- 
tween the three series of dissolved haloid salts, where Br=Cl+6"36, and 1 = 01+16-46. 
Potassium. — The number of potassium salts in solution whose refraction-equivalents 
have been determined is 26. There are two ways of arriving at the equivalent of the 
metal itself from these data. 1st. If we know the value of any of the radicals conjoined 
with potassium (expressed in Table VI. by Greek letters), it is a simple question of sub- 
traction. 2nd. If we know the value of any other capital letter in Table VII. , we have 
merely to add to it, or subtract from it, the mean number representing the difference be- 
tween it and A, and we arrive at A itself, that is the refraction-equivalent of potassium. 
For calculation by the first method, the numbers already arrived at maybe employed, 
namely, C= 5, 11 = 1*3, 0 = 2-9,8 = 16; and from the ethyl compounds in Table III. 
the following values may also be accepted, NQ 3 = 14-4, Si0 4 = 18-4, C0 3 =12'9, and 
CN = 9T, the two latter numbers corresponding with those of carbonic anhydride and 
cyanogen gas in Table II. We obtain 
From the Formiate 
. Potassium 
= 8-14 
55 
Acetate . 
• • 55 
= 8-08 
55 
Oxalate . 
• * 55 
= 8-05 
55 
Alcoholate . 
• * 55 
= 8-72 
55 
Lactate . 
• • 55 
= 7-92 
55 
Tartrate . 
• 
• 0 55 
= 7-63 
*5 
Nitrate . . . 
. . „ 
= 7-71 
55 
Silicate . . . 
0 0 5 5 
= 8-30 
55 
Carbonate . 
• ‘ 55 
= 7-93 
55 
Cyanide . . 
- . 
CO 
i — i 
cb 
II 
55 
Sulphocyanide . 
• 8 55 
Mean . 
= 8-37 
. 8-09 
Deductions from the chloride, bromide, and iodide are omitted from this list, because, 
as has been shown already, the differences between the refraction-equivalents of these 
halogens in dissolved salts must be somewhat greater than we find them to be in organic 
compounds. It is true this a priori objection does not lie against the chloride itself, 
but the close analogy between its properties and those of the two other halogens renders 
