Day, DONALD J.~MATTHEWS. 
to dissolve off the paper completely, so the following method was adopted. 
Five cubic centimetres of strong hydrochloric acid were poured into the 
beaker in which precipitation had taken place and distributed over the 
walls to dissolve any iron ; then 10 ccm. of water were added, the beaker 
was warmed for a few minutes, and the hot dilute acid was allowed to 
drop on to the filter paper by means of a small pipette provided with a 
rubber-teat head. The beaker and paper were well washed and the 
latter incinerated in a platinum crucible; the ash was dissolved in a 
little strong hydrochloric acid and added to the rest of the solution, which 
was then evaporated to dryness on the water-bath ; the chlorides were 
converted to nitrates by evaporation to dryness with 10 ccm. of the 
dilute nitric acid, taken up in more nitric acid, using 7 ccm. of 25% by 
volume if the final bulk was to be 50 cem. The analysis was then carried 
out as described in the previous paper. In an extreme case the dirty 
residue on the paper was found to contain 0-0024 mg. of P,O;. The 
origin of the greasy matter is unknown ; it may be due to the drainage 
from the port, or, on the other hand, it may be the volatile oily substance 
which so often renders the distillate turbid when sea-water is boiled with 
an alkah for the determination of ammonia. 
It was mentioned in the earlier paper that the amount of phosphates 
found was increased if the water was previously oxidised with a per- 
manganate. Attempts to find a suitable method for determining this 
excess phosphorus have been only partly successful. It is necessary to 
use strong oxidising substances which do not interfere with subsequent 
operations and are easily purified. In the end the following process was 
adopted. From 200 ccm. to 400 cem. of the water were evaporated in 
a porcelain basin holding nearly 200 cem. until the bulk of the salts had 
separated; then 10 cem. or 20 cen. of strong nitric acid were added, accord- 
ing to the amount of sample taken, the dish was covered, and heating 
continued until the evolution of brown fumes had ceased. The cover was 
removed and the evaporation was continued to dryness; the dish was 
then heated over an argand burner until the salts were in gentle fusion 
and the nitrates of the earths were decomposed with evolution of brown 
fumes. 
The dish was allowed to cool and the salts were dissolved by warm- 
ing for an hour with 150 cem. of water and 1 ccm. of strong hydrochloric 
acid. tron was added and 5 cem. of 2 N sodium carbonate solution and 
the analysis completed as before. There is danger of the porcelain being 
attacked unless it is very carefully heated, and unless it is certain that 
it gives up no phosphates under this treatment it would probably be 
better to use fused silica basins. 
The blank on the reagents used in the analyses reported in the previous 
