386 THTSICAL GEOGRAPHY OF THE SEA, AND ITS METEOROLOGY. 



than those of the sea, are therefore, though cokler, yet lighter 

 (§ 42()) than the warmer waters of the ocean. And thus we have 

 repeated here, though on a smaller scale, the phenomena as to 

 the flow of cold waters from the north, which force the surface 

 isotherm of 60° from latitude 56'^ to the parallel of 40° during 

 three or four months. Changes in the colour or depth of the 

 water, and the shape of the bottom, etc., are also calculated to 

 cause changes in the temperature of certain parts of the ocean, 

 by increasing or diminishing the capacities of such parts to 

 absorb or radiate heat ; and this, to some extent, assists to bend 

 or produce irregular curves in the isothermal lines. After a 

 careful study of this plate, and the Thermal Charts of the 

 Atlantic Ocean, from which the materials for it are derived, I 

 am led to infer that from January to August the mean tempera- 

 ture of the atmosphere between the parallels of 56° and 40° north, 

 for instance, and over that part of the ocean in which we have 

 been considering the fluctuations of the isothermal line of 60°, is 

 at least 60° of Fahrenheit, and upward, and that the heat which 

 the waters of the ocean derive from this source — atmospherical 

 contact and radiation — is one of the causes which move the 

 isotherm of 60° from its Januaiy to its September parallel. It 

 is well to consider another of the causes which are at work upon 

 the currents in this part of the ocean, and which tend to give the 

 rapid southwardly motion to the isotherm of 60°. We know the 

 BGiean dew-point must always be below the mean temperature of 

 any given place, and that, consequently, as a general rule, at sea 

 the mean dew-point due the isotherm of 60° is higher than 

 the mean dew-point along the isotherm of 50°, and this, again, 

 higher than that of 40°, this than 30°, and so on. Xow suppose, 

 merely for the sake of illustration, that the mean dew-point for 

 each isotherm be 5° lower than the mean temperature, we should 

 then have the atmosphere which crosses the isotherm of 60°, 

 with a mean dew-point of 55°, gradually precipitating its vapours 

 until it reaches the isotherm of 50°, with a mean dew-point of 

 45-° ; by which difierence of dew-point the total amount of pre- 

 cipitation over the entire zone between the isotherms of 60° and 

 50° has exceeded the total amount of evaporation from the same 

 surface. The prevailing direction of the winds to the noiih of 

 the fortieth parallel of north latitude is from the south wai'd and 

 westward (Plate YIII.) ; in other words, it is from the higher to 

 the lower isotherms. Passing, therefore, from a higher to a 



