340 Scientific Proceedings, Royal Dublin Society. 
sea-water ; but the first explanation appears the more probable 
one.? 
A very instructive result of our determinations is the com- 
parison of the “oxygen absorbed ” by the permanganate test with 
the oxygen actually absorbed as shown by the analyses of the 
dissolved gases, which clearly show how fallacious the indications 
of the first-named test are. 
This is shown in two ways, and to illustrate the point it will 
be sufficient to take the figures obtained in Series 1. First, if we 
take the results obtained from the “oxygen absorbed ” test in the 
original mixture, and contrast it with the oxygen actually absorbed 
in 48 hours, the figures are, respectively, 0:144 parts by weight 
per 100,000, and 3°21 c.c. per litre. The latter is equivalent to 
4-6 milligrammes per litre, or 0°46 parts by weight per 100,000. 
In other words, the oxygen actually absorbed was about three 
times as much as the amount shown by the permanganate test. 
Next, if we contrast the differences in different periods of the 
two tests, and translate them into the same units, we obtain the 
following figures— 
Dissolved Oxygen Differences between 
actually disappearing ‘‘ Oxygen absorbed ”’ 
after :— originally, and after :— 
6 hours, 6 é 0:0048 0:018 
24 85 : 6 0°2996 0:036 
A Sens EB c 0°4602 0:063 
The figures represent parts per 100,000, and should, of course, 
correspond if the oxygen of the permanganate acted in the same 
manner as the dissolved oxygen. This, of course, was not to be 
expected, but it is surely a remarkable fact that a more energetic 
oxidation is induced by micro-organisms and free oxygen than by 
the nascent oxygen of the permanganate solution. 
Turning next to the determinations of nitrogen both “ unoxi- 
dized’” and “ oxidized,” we have, in the Table, expressed their 
amounts in the orthodox manner of parts per 100,000, but owing 
to the smallness of the quantities involved, and the magnitude of 
1 It has been suggested to us that possibly both the absorption of oxygen and evolu- 
tion of carbonic anhydride may be due to an act on the part of the micro-organisms 
analogous to ordinary respiration. 
