216 
THE VOYAGE OF H.M.S. CHALLENGER. 
.Y< — No. 2G5. — "Sample shaken with air, so as to saturate it at 22° - 8 C.” (J. Y. B.). This, 
according to my laboratory experiments, as above referred to, would remove carbonic acid even from 
:t water of the eomposition given in the table, and as the water even after having been thus manipu- 
late. 1 apparently contained fully saturated bicarbonate, it should originally have contained a very 
appreciable proportion of free carbonic acid. 
No. 3SJ. — “ Gases boiled out of water and 225 c.c. of remaining water distilled with BaCl 2 as 
usual. Hence CO., in gas-tube must be added” (J. Y. B.). I have done so; without this correction 
the loose carbonic acid per litre would be 82‘9 mgrms. ; and the deficit, consequently, —29 0. 
No. 556. — “C0 3 and gases boiled out” was on the label of the bottle which came to me. Hence 
the alkalinity may to a slight extent be owing to dissolved glass, and the recorded value 57'4 be a 
little too high. 
Mr. Buchanan’s numbers for the “carbonic acid per litre,” even when taken in a 
purely empirical sense, i.e., as referring to the carbonic acid which actually was eliminated 
in the distillation, are, in my opinion, uncertain by about ±2 mgrms. Admitting this, 
the results are compatible with the following conclusions. 
Seven out of the 27 waters contained their surplus base substantially as unmixed 
bicarbonate. Five of these 7 waters came from the bottom or great depths, only 2 
being from depths of 50 fathoms or less. 
Two of the 27 waters contained free carbonic acid in addition to bicarbonate. I 
n-fer to Nos. 378 and 383, both bottom waters. 
In No. 383 the actually free carbonic acid would appear to have amounted to 42 
mgrms. per litre ! 
In 18 of the 27 waters the carbonic acid deficit has tangible positive values; they 
all contain their surplus base as a mixture of normal carbonate and bicarbonate. In 
regard to their depths they may be arranged as follows : — 
Surface. 50 Fathoms. 400 Fathoms. Bottom. 
3 1 2 12. 
To pa-- now to the remaining 168 determinations of loose carbonic acid by Bucharian, 
f-ir which we have not the corresponding alkalini tics : — To be able to interpret them at 
all, W< must assume my average value of 54*7 for the alkalinity per litre to hold for 
them all, and we may well permit ourselves to use the approximate value 55. 
If w. do so it takes in each case 55 mgrms. of loose carbonic acid to produce bicarbonate, 
of tin- 168 waters under consideration only 15 yielded more than this limit proportion of 
‘ o . ror arid, and two of the determinations (Nos. 114 and 497) are probably erroneous.* 
* v HI. > Mr. !;>i hnrwin w;i» lightly turlii.l with carbonate of lime. Hence the high carbonic acid (04 
Mgiw*. jwr litas) which he (bund count* for nothing. It i- «urc to have been produced partly by the action of the 
■jdrwhlon •rid liberated from tin- chloride of magnesium during tin- process of distillation. No. 497 gave 59*4 mgrms. 
' : I |. r litr> , t ut tl,. ic.»ult, in Mr. Buchanan * opinion is “probably false.” (Compare remarks on 
U>-lr, jwgv | id •(,,] of memoir.) 
