SPECIFIC VOLUME AND DENSITY OF ATMOSPHERIC AIR AND SEA-WATER. 39 



Tables 12 h and 20 h show an increasing resistance of the sea-water against com- 

 pression for increasing salinity, and tables 13 h and 21 h show, for the interval of 

 the temperatures in question, a similar increased resistance against compression with 

 increasing temperature. This dependence of the compressibility upon the temper- 

 ature and the salinity is of great importance for the internal conditions of equilibrium 

 or of motion in the sea. To consider a definite example: At a pressure of 5000 

 decibars, i. e., at a depth of 5000 meters, water of 35 %o salinity and at the 

 temperature of 1 C. will have the same specific weight as water of 35.48 %o 

 salinity at a temperature of -j-i C. But under the diminished pressure of 2000 

 decibars, /. e., at the depth of 2000 meters, the specific volume of the first water 

 will be 0.00015 g reater than that of the second, and at a depth of 9000 meters the 

 reverse will be the case. The extreme importance of these differences of com- 

 pressibility will thus be perfectly clear. 



31. Isosteric and Isopycnic Surfaces. The value of the specific volume being 

 known in a sufficient number of points in the atmosphere or the sea, we can repre- 

 sent the distribution of mass in each of these media by drawing a set of equiscalar 

 surfaces, joining all points where the specific volume has certain constant values. 

 We shall call these surfaces isosteric surfaces. If, on the other hand, the value of 

 the density be known in a sufficient number of points, we may represent the distri- 

 bution of mass by drawing surfaces of constant value of the density, or isopycnic 

 surfaces. 



The two fields representing the distribution of mass are closely related to each 

 other, every isosteric surface being also an isopycnic surface, and vice versa. But 

 one important difference should be emphasized: if the isosteric surfaces be drawn 

 for unit-differences of the specific volume, the corresponding isopycnic surfaces 

 will not have unit-differences of the density, and vice versa. This will be well 

 illustrated if we consider the conditions in the atmosphere. Here the density 

 decreases upwards, converging toward a very small limit, or perhaps toward 

 zero. The specific volume, on the contrary, which is the reciprocal of the density, 

 increases upwards, converging toward a very great limit, or perhaps toward infinity. 



Drawing the isopycnic surfaces for unit-differences of the density (a unit of con- 

 venient magnitude being chosen), the thickness of the unit strata will increase 

 upward, approximately in geometric series. If, on the other hand, the isosteric 

 surfaces are drawn for unit-differences of the specific volume, the thickness of the 

 corresponding unit-sheets will decrease, approximately in geometric series. Even 

 in the sea there is a corresponding difference between the two systems of surfaces, 

 only much less pronounced. 



The equiscalar surfaces of the specific volume or the density being very nearly 

 level, the gradient or the ascendant of these quantities will be directed very nearly 

 along the plumb-line. For reasons which will appear later it will be more con- 

 venient to use the ascendants than the gradients in this case. The ascendant of 

 the specific volume points upward, that of the density downward, forming a very 

 small angle with the plumb-line. 



