285 
of Edinburgh, Session 1876 - 77 . 
equator from east to west the specific gravity appears to increase 
pretty steadily. In the Central Pacific the maximum saltness is 
observed in latitude 18° S., where the specific gravity is 1*02725; 
in the north the maximum is 1*0263, in latitude 22° N. The 
minimum of 1*02488 was found in latitude 7°26' N. In the 
Western Pacific the series of observations was not so continuous, 
there being a considerable gap between latitude 15° S. and 2 S. ; 
the northern portion, however, is complete. Here, again, the 
absolute maximum is in the south, reaching 1 *0268, in 22°30° S. ; 
in the north the maximum is 1*0263 in 22° N., but the dense water 
of the Southern Pacific penetrates into the northern part of the 
ocean, giving another and higher maximum of 1*0266 in lat. 2' N. 
The actual maximum observed was 1*0242 close to Humboldt Bay, 
in New Guinea. Other very low specific gravities were observed 
occasionally even at considerable distances from the coast; but the 
fact that there was frequently drift-wood floating about, and that 
the light water was confined to a stratum of 10 to 20 fathoms thick- 
ness, showed that we had here to do with shore water. If the 
general run of the curve be followed, the oceanic minimum will be 
found to occur about 10° N., marking 1*0258. 
The saltness, which increases on both sides of the equator until 
a maximum is reached in the trade wind districts, decreases again 
towards the poles and the whole of the ocean to the southward of 
latitude 40° S., has a specific gravity below 10260, and for the 
greater part below 1*0255. In the neighbourhood of ice in sum- 
mer great and sudden variations are frequently observed, the specific 
gravity amongst loose park ice frequently falling as low as 1*0240, 
Confining our consideration, then, in the meantime, to the 
surface, we find that the maxima in the Southern Pacific and 
Atlantic are situated nearer the equator than those in the northern 
parts of these oceans; that the maxima in the Atlantic are higher 
than the corresponding ones in the Pacific ; and that the maximum 
maximorum occurs in the South Atlantic about latitude 15° S. 
The equatorial minima in both oceans are to the northward of the 
equator ; and from all these facts we see how intimately the con- 
centration of the sea-water is connected with the distribution of 
the trade winds. 
If we turn now to the consideration of the vertical distribution 
