in the Proportion of Oxygen in the Atmosphere. 433 
water. If facts do not agree with these deductions, the suppo- 
sition that a large part of the processes of oxidation on the 
surface of the globe takes place in sea-water within the tropics 
1s contrary to the facts. . 
In the fourth place, it is doubtful whether rivers carry an 
such amount of organic matter as is required by the theory. 
Determinations of the amount of oxidizable matter contained 
in the water of rivers have been chiefly limited to the water 
supply of towns. But some observations have been made on 
the water of the Nile. Tidy found by the permanganate pro- 
cess that 0-283 grain of oxygen was given up to a gallon of the 
water of this river. If we take this result to represent the 
amount of oxygen absorbed by river water after the water 
reaches the tropics, we shall concede much for argument. Such 
water could remove 0-001 oxygen from about ten times its own 
volume of air. Of course it is difficult to suppose that the 
consumption of oxygen can be localized in a small volume of 
air. Now, if such water be diluted with sea-water, and if it 
absorbs oxygen from a hundred times its volume of air, 
through several degrees of latitude, and if the deficiency of 
oxygen to be explained is several times 0-001, it is hard to 
believe that the cause is sufficient. 
5. It is very doubtful whether the whole consumption of 
oxygen on the globe would account for the observed deficien- 
cies of oxygen, even if we suppose this total consumption for a 
certain short period to be taken from one and the same small 
volume of air. 
Dumas and Boussingault made an approximate estimate of 
the amount of oxygen used in a century by all process of oxi- 
dation. If we take this estimate we shall find that all the oxy- 
gen absorbed from the air in a week, if taken from the same 
volume of air covering but half a square degree of the earth’s 
surface, and containing only the lower third part of the atmos- 
phere, would produce in this limited volume a deficiency of 
oxygen of but one-eighth of one per cent. But we have to 
account for deficiencies several times as large, and we cannot 
suppose the consumption so limited to a small volume. Then 
Am. Jour. Sct.—THmpD foe Vou, XXII, No, 132.—DEcEMBER, 1881. 
