2*26 
THE VOYAGE OF H.M.S. CHALLENGER. 
collected, a good number were analysed by himself, but the majority were analysed by 
me, and my report contains both his results and mine. 
The method which I used in the absorptiometric determinations referred to, consisted 
essentially in this, that I saturated a quantity of sea-water with air at a known 
temperature, and then from a measured volume extracted the gases by a method similar 
to Jacobsen’s, but in its final form differing from it in this, that the vacuum in the gas- 
collccting tube was maintained by a mercurial air-pump, which sucked out and removed 
the gas as quickly as it was liberated. I was led to adopt this improvement, because I 
had found it impossible to obtain sufficiently constant results by the Jacobsen method in 
its original form, and ascribed the fluctuations to the obvious fact that the vacuum 
originally existing in the gas-collecting tube, is soon destroyed by the gas going into it, 
so that necessarily a small but variable portion of the gas, corresponding to the coefficient 
of absorption at the temperature at which the water boils at the end and to the final 
pressure in the tube, must remain in the water. The results of our analyses, as inter- 
preted on the basis of my absorptiometric work, agreed on the whole with the inferences 
which have just been deduced from known physical laws. In the surface waters the 
volumes of nitrogen and oxygen present in a litre of water were found to be functions of 
the temperature whose general course was similar to the theoretical functions determined 
by my laboratory experiments. In the deep-sea waters the volumes of the nitrogen varied 
within the same limits as those in the surface waters; but the volumes of oxygen were, 
in general, less than those calculated from the nitrogen-volumes on the hypothesis of 
surface absorption of air at the temperature corresponding to the nitrogen found. 
In waters from great depths the actual volume of oxygen was often very small, as 
illustrated by the following two examples : — 
C.c. per litre. 
No. of Water. 
N a 
o 2 
0 2 calculated. 
8 
1001 
1508 
0-6 
8-21 
2875 
1645 
13-38 
2 04 
6-95 
1500 
The occurrence of such very small values for the dissolved oxygen proves that at many 
places of the ocean bottom the progressive motion of the water and the rate at which it 
exchanges gases, or mixes, with the upper strata, must be very slow indeed. 
It is worth noting, however, that very small quantities of oxygen present themselves 
o< ca-ionally even at moderate depths, as shown by the following example : — 
