Lerts anpD BLhake—Estimation of Oxygen in Solution. 461 
The following experiments show the extent of accuracy of the 
process, with different samples of natural waters, sewage effluents, 
and sewage itself :— 
Distilled Water saturated with air at 15°C.—(Titrations by per- 
manganate 1 c.c. = 1 c.c. oxygen at N. T. P.) 
A. Blank experiment required, ; ‘ «BASS CaGs 
Oxygen determination required, . : les) se, 
Dissolved oxygen = 2°35 ,, 
5. Blank experiment required, 3 : . 3°89 Ge. 
Oxygen determination required, . : Se oi. 
Dissolved oxygen = 2°35 ,, 
Dissolved oxygen per litre, . 7:06 ¢.c. found. 
Roscoe & Lunt’s result for 
OSCE distilled water at 15° C. 
1) ) 9) 
Difference, 0°10 ,, 
Distilled Water saturated with Air at 13-°7°C.—(Titrations by 
permanganate | c.c. = 0:1 c.c. oxygen at N. T. P.) 
A. Blank experiment required, : : 5 Seril Ge, 
Oxygen determination required, . : BeBe o) oa 
Difference, . BENDA ICO e 
Dissolved oxygen = 2°420 c.c. 
B. Blank experiment required, : : . 38°05 c.c. 
Oxygen determination required, . , 5 1 4 
Difference, 3 5) DOI) a5 
Dissolved oxygen = 27415 c.c. 
Whereas assuming that the 7 c.c. of reagents contained xo oxygen, the figures be- 
come— 
2350 
325°5 
The difference is 0-152 c.c. per litre, which, for most practical purposes, is unim- 
portant, and the actual error would certainly be much less. The error, so far as the 
ferrous sulphate solution is concerned, could be eliminated by saturating it with air at 
the given temperature previous to a determination, while that from the small volume 
of ammonia added is too insignificant to need correction. 
= 7:°219 c.c. dissolved oxygen, per litre, at N.T.P. 
