Or SEA-WATER IN THE DIFFERENT PARTS OF THE OCEAN. 
229 
part of the German Ocean which lies close to the Kattegat 3*15, and in the whole 
German Ocean 2 -87. In a sample from the Black Sea which I analyzed I found it 4 - 22. 
B. On the difference of the contents of Sea-water at the surface and in different, 
depths. 
It would be natural to suppose that the quantity of salts in sea-water would increase 
with the depth, as it seems quite reasonable that the specific gravity of sea-water would 
cause such an arrangement. But this difference in specific gravity relative to the 
increase in the quantity of salts is counteracted by the decreasing temperature from the 
surface to the bottom. We have parts of the sea where the .quantity of solid salts 
increases with the depth ; in other parts it decreases with the increasing depth ; in 
other places hardly any difference can be found between surface and depth ; and, lastly, 
I have found one instance where water of a certain depth contained more salt than both 
that aboveand below. These differences are to a great extent dependent upon currents 
both on the surface and in different depths. The phenomenon of double currents at 
the Straits of Gibraltar has been long known, and in close connexion with these double 
currents the saline contents of the water of the Mediterranean increase in quantity with; 
the depth. There is, however, one exception in the Mediterranean, under interesting 
circumstances, which I shall afterwards discuss more at length. I have made eleven 
complete analyses of the surface-water of the Mediterranean, and calculated another 
quoted in Violette et Archambault, ‘ Dictionnaire des analyses chimiques,’ vol. i. 
p. 358, without a more exact reference to the place where it was taken. Of my own 
analyses, one must be rejected on account of the great quantity of sulphuretted hydro* 
gen that had been formed, and of course caused a loss of sulphuric acid ; but it causes 
also a loss of lime, because the formation of sulphuretted hydrogen is contemporaneous 
with the formation of carbonic acid, which will precipitate the lime when deprived of 
its sulphuric acid. The mean number of the remaining analyses of surface-water is 
20889 per 1000 for the chlorine, and 37‘936 for all salts. The mean number for chlo- 
rine of eight analyses of water taken from a depth of between 300 to 600 feet is 21T38. 
In each case the deep water was richer in chlorine than that from the surfaoe, except in 
one instance, where the chlorine of the surface-water was 21 '718, and all salts, calcu- 
lated from a complete analysis, were 39*257 per 1000, while the chlorine of water taken 
from a depth of 522 feet was 21-521 per 1000. This curious exception occurred 
between Candia and the African coast, where the dry and hot winds from the neigh- 
bouring Libyan desert evidently cause a strong evaporation and a considerable eleva- 
tion of temperature, which counteract each other as to specific gravity. The difference 
between the upper and lower current in the Straits of Gibraltar is, in the surface-water,, 
chlorine 20-160 per 1000, all salts 36*391, and in the depth of 540 feet, chlorine 20-330. 
The cause why the surface-current is Atlantic water flowing into the Mediterranean,, 
and the under-current Mediterranean water flowing into the Atlantic, has long since been 
assigned to depend upon the comparatively small quantity of water that flows from the 
land into the Mediterranean, and the hot and dry African winds that cause more water 
2 i 2 
