8 
The surface temperature of course varied considerably with the weather, 
but the influence of the sun's rays did not appear to penetrate beyond 
about 100 fathoms, and hence while the amount of fall in the thermome- 
ter in the first few hundred feet was subject to considerable variations, 
below that depth the rate of descent in the mercury column, though not 
varying in proportion to the depth, was the same in all the observations 
made within the above named limits. An inspection of the table will 
show that from about 100 to 550 fathoms the temperature decreased 
gradually and with tolerable regularity, while from 550 fathoms to 850 
fathoms the fall was much more rapid, and after that depth the rate of 
decrease was extremely slow, getting slower and slower as the depth in- 
creased. It will presently be shown that there is the strongest reason to 
believe that in the intermediate depths (550 to 850 fathoms) where the 
most rapid fall takes place, the cold current at the greater depths (which 
was traced in the third cruise as entering the North Atlantic basin between 
the Faroes and the Shetlands) mixes with the warmer water at and near 
the surface, which is flowing from the Equatorial towards the Polar 
regions. 
The third cruise was occcupied with a re-investigation of the area ex- 
amined by Dr. Carpenter and Prof. Thomson in the "Lightning" in 1868. 
It started from Stornoway, in the Hebrides ; its northern limit was Thors- 
haven, in the Faroe Islands, its western limit was 9° 20 7 W., and its 
eastern 0° 35 1 E. In no part of this area was a greater depth than 750 
fathoms reached, but careful temperature serial soundings revealed a very 
remarkable state of things. The existence of the warm and cold areas 
referred to in the "Lightning " report was fully confirmed, their bounda- 
ries were determined with precision, and the depth at which the ice-cold 
water flowed, and where it mixed with the warmer surface water, exactly 
ascertained. 
The following tables give the relation of temperature to depth in the 
warm and cold areas respectively, deduced from many observations, both 
by serial and bottom soundings, the results of which agreed closely among 
themselves. 
