236 Sodium and Potassium 
1 H H 
soe oS COT 
8545 85.45 Fy BSA ry BMT 
58.50 74.56 __ 74.56 74.56 
35.45 35.45 
58.50 74.56 
times that in chlorine,? a value which increases as the propor- 
tion of sodium in the mixture diminishes. 
H 
Magnification of Percentage Error in Chlorine (a 35.45 th 
~ 74.56 
ACT SK GL ier, 32 oer ents Sete 9:14 8:2] 7:3 | 6:4] 5:5 | 4:6 | 3:7 | 2:8} 1:9 
Magnification...................| 5.0) 5.3] 6.1] 7.0} 8.1/10.0)13.0)18.4/35.8 
The very small error in the McLean- Van Slyke chloride method! 
can abide such ten- or twenty-folding. 
The steps in the McLean-Van Slyke method are as follows: 
(1) add standard silver nitrate to solution of chlorides; (2) 
make to definite volume, filter, take aliquot part of filtrate; 
(8) titrate excess of silver nitrate in aliquot part with standard 
potassium iodide. 
The error in measuring the silver nitrate with a pipette® (ex- 
perimentally determined) is 0.005 ce. in 20 cc., — 0.025 per cent. 
(We weighed the irregular amounts of silver nitrate used in our 
work; the use of a burette at this point would multiply several 
times the error shown in Table I.) 
If no more than 5 per cent excess of silver nitrate is added, the 
error in “definite volume and aliquot part’ will be only 5 per cent 
of the actual error in the measurements (5 per cent of 0. 10 ne 
cent = 0.005 per cent). 
’The total chlorides (NaCl + KCl) can be weighed accurately to 
0.1 mg.; but if any error is introduced at this point it will be multiplied 
35.45 
74.569 
He 35.45 
7456 times in the final result. 
4 McLean, F. C., and Van Slyke, D. D., J. Am. Chem. Soc., 1915, xxxvii, 
-1128. 
5 We used 0.1 n AgNO; and 0.02 n KI (approximately). 
