3388 
NATURE 
[ Feb. 2gi 18 82. 
similar experience was made with the first few samples 
boiled out on board the C/a//enger, but it was detected in 
time to prevent any serious loss. Indeed from dust and 
particles of sawdust having got into the tubes, it was 
necessary in every case, before using a gas tube, to 
remove its end tubes, clean the body thoroughly, re-attach 
the end tubes, and draw out the thickened contraction. 
When so drawn out, there is not the same mass of glass 
to be heated, and the contracted part can be heated for 
itself without any danger of softening the wide part. There 
was no instance of a tube cracking after being sealed up 
with these precautions. In recent practice the writer has 
considerably modified and improved the apparatus for ex- 
tracting gases from water on shipboard, but a description 
of the apparatus would here be out of place. 
The figures representing the results of the analyses are 
necessarily affected by errors incident to the collection 
and the transvasing of the water and to the separation 
and analysis of the gas. The combined effect of these 
errors can be appreciated by the study of the following 
table, in which are collected the results of analyses made 
in duplicate. It is not stated whether on each occasion 
two separate and distinct samples were collected and 
treated separately, or, from the same sample of water, 
two portions were separately boiled, and thus two portions 
of gas obtained for analysis. Of the nine waters so 
treated four came from the surface, three from the bottom, 
and one from an intermediate depth of 300 fathoms. 
Table of Duplicate Analyses 
: O+} ce 
— Depth (fathoms). cae ragicor ae Difference. 
| 
a | 2 | —— = 
| | 
700 } 20°5 a 33°0 1 
Be5 (Bottom) | 20°0 mh 336 we 
795 20°6 $ 32°6 ; 
a (Bottom) 19°4 DS al eogaer ae 
1760 19°6 me 34°0 e 
=e (Bottom) = 33'8 <A 
2 1149 21°9 32°2 . 
332 (Bottom) 22°0 ag 31'°8 ns 
24x 300 20°9 . 34°4 
345 (Intermediate) 21°5 o°6 339 ae 
183 Surface ee} — oe toh) 
3 Sartal’ 19'S "2 354 : 
283 Surface 19°5 O73 333 o'r 
_ Surface shal NN 35'8 iPo'4 
77 354 
Bze Surface sey ae ae o7 
| 3 
From this table we see that the mean difference in oxygen 
per cent. as found in the different analyses of the gas 
from the same water was o'5. The differences in the 
total volume of nitrogen and oxygen per litre varied from 
Ol cc, to 1°2 cc. 
Dr. Tornoe has collected at pp. 15 and 16 the results 
of his analyses of ninety-four samples of air extracted 
from water of various depths. Of these thirty-three are 
from surface water, eighteen from intermediate, and forty- 
three from bottom water. The last-mentioned are from a 
great variety of depths, ranging from 25 to 1760 fathoms. 
His results give us a very complete account of the state 
of aération of -the water of the ‘ Norwegian Sea,” or of 
that part of the ocean extending from the Farée Islands 
northwards to Spitzbergen, having for the greatest part 
of its length the shores of Norway for its eastern boundary. 
The investigations were carried on between the middle 
of June and the middle of August, or during the height 
of summer, consequently the temperature of the surface 
water was never either very high or very Jow. In the 
table of results the temperature of the water is given. Of 
the surface waters examined the mean temperature was 
66° C., the highest being 11°8°, and the lowest 0°5° C. 
If the total volume of the oxygen and nitrogen be taken 
to be 100, then the oxygen was found to vary between 
337 and 36°7, mean 35°4. The absolute amounts of the 
gases varied with the temperature. 
The results obtained by Jacobsen in the Pomerania 
showed a very remarkable agreement in the percentages 
of oxygen found in the surface water. The mean results 
of twenty-one observations were as follows :—Tempera- 
ture 16°66° C., nitrogen 11°07 cc., oxygen 5°69 cc., together 
16°76 cc. per litre, and oxygen percentage 33°93. The 
lowest oxygen percentage was 33°64, and the highest 
3414. From this it was concluded that the percentage 
of oxygen in sea water is practically invariable, as it is in 
the atmosphere. For the limited area explored by the 
Pomerania this is undoubtedly proved, but the area was 
comparatively small and the variations in conditions, 
especially temperature, insignificant. In the Challenger, 
waters subject to the most varied conditions of climate were 
treated for the extraction of the gases, and before leaving 
the work of the expedition, now more than four years ago, 
the writer had analysed a considerable number of the 
samples of gas so procured. The results of these analyses 
showed at once that Jacobsen’s conclusion as to the ocean 
as a whole was not justified, while it held good with regard 
to limited areas. If we confine our attention to surface 
water, the highest percentage was 35’o1 in the Antarctic Sea, 
and the lowest 32°35 in the Pacific, between Fiji and Torres 
Straits. This was however a very remarkable water, and 
should be excepted. The next lowest percentage was 32°82, 
so that in round figures the oxygen percentage varied 
between 33 and 35. As the cruise of the Challenger was 
chiefly in tropical regions, the surface water had usually a 
high temperature; but water of all temperatures was 
| experimented on, and if the results are arranged in 
ascending order of temperature, the oxygen percentage 
is seen to decrease very regularly. 
In waters of temperatures above 20° C. the percentages 
ranged between 32°82 and 33°33, the mean of nine such 
observations being 33°09. The mean of five observations 
between 12° and 20° is 34°22, the extremes being 33°52 
and 34°66. We gather from the results of the three ex- 
peditions that the percentage of oxygen is less in warm 
water than in cold. 
In order to judge of the degree of saturation of the 
waters Dr. Tornoe reports some interesting experiments 
on the absorptive power of sea water for the atmospheric 
gases at different temperatures. He experimented on 
four temperatures, namely, o° C., 5°, 10°, and 15° C., and 
from the results so obtained he gives the following formula 
| for the solubility of nitrogen and oxygen in sea water 
exposed to a current of air :— 
N = 1444 —0°'237 
oO 7°79 — 0°24 0°005 7. 
The formula for nitrogen agrees with the facts of his 
observations at the four temperatures: that for oxygen 
begins to fail at the highest temperature, 15° C., and is 
clearly inapplicable at temperatures above 15° C., for it 
gives a minimum of solubility at 20°, and at 40° C. this 
solubility is the same as at o° C., and is increasing. 
The analyses of the gas from surface water do not 
agree very well with the figures given by his nitrogen 
formula. Jacobsen’s results are also higher than would 
be given by the formula, and the Chad/enger results con- 
siderably higher. The last are better represented by the 
formula— 
and 
N = 15'8 — 0'234. 
Dr. Tornoe notices that the oxygen found by him in 
surface water is considerably in excess of what would be 
