CANADIAN FISHERIES EXPEDITION, WlJf-lo 229 



The above simple example will serve to give an idea as to the complicated move- 

 ment caused by the wind in water layers of stable relative position. The system of 

 strata and movement of the water shown in figs. 5 and 6 are of very frequent occurrence 

 in the sea; in some places, however, the density increases more continuously with the 

 depth, and isosteric surfaces are then found in all parts of the water. Such waters 

 may be regarded as consisting of an infinite number of infinitely thin strata. The 

 movement of the water here is highly restricted, which gives rise to very peculiar dyna- 

 mic phenomena. The water assumes a remarkable power of resistance against the 

 action of external forces, and when these cease to operate, it returns to its original 

 position. 



Let us now endeavour to ascertain the cause of this. Taking any one of the 

 isosteric sections in either of plates VIII and IX, we imagine a water sample from 

 one of the isosteric surfaces transferred to a greater depth. It will here be lighter 

 than its surroundings, and will therefore, according to the Archimedean principle, 

 j'ise until it once more reaches the isosteric surface from which it was taken, and 

 there it will remain. In the same way, if a sample of water be shifted to a point 

 above that whence it was taken, it will be heavier than its surroundings, and will 

 sink until it reaches the isosteric surface corresponding to its own specific gravity. 

 It is otherwise, however, when a sample of water is moved along the isosteric surface. 

 Here its specific gravity remains equal to that of its surroundings, and no force 

 arises which would occasion its return to the original position. Thus we see that 

 the water can only move along the isosteric surfaces, and not transversely through 

 them. In other words, the scope of movement of the water, instead of being tri- 

 dimensional, is restricted to the two dimensions. 



The constant validity of this principle throughout the whole of our present area 

 of investigation is most clearly shown by the existence, and extraordinary permanence, 

 of the cold water layer. It is at once evident that no vertical convection can take 

 place through this layer, the movement of the water being strictly confined to the 

 horizontal. 



It is therefore of particular importance to ascertain at what depths the water par- 

 ticles composing one and the same isosteric surface are to be found. For the sake of 

 convenience we may here content ourselves with examining the isosteres i;i=400, 500, 

 600, etc. These depths may easily be found, by interpolation, from the Vi column in 

 table ], and are here included in table 2. In this table, B indicates that the isosteric 

 surface in question touches the bottom, and * that it cuts the surface of the sea before 

 reaching the hydrographical station concerned. Plates X and XI show some of these 

 figures for depth at their proper position within the area of investigation. Here 

 also, lines are drawn to indicate the intersection of the isosteric surfaces with the 

 surface of the sea. This figure shows at a glance why it is that the surface water, 

 daring the cold season, keeps to the coastal zone, but is able during the warmer 

 months to move farther out. Another point immediately evident is' the very high 

 degree in which the heating of the surface water by the sun's rays contributes to its 

 freedom of movement. 



4. FOECES DEEIYED FEOM THE DISTEIBUTION OF DEN"SITY. 



The Archimedean principle itself teaches us a great deal as to the forces in the 

 sea which are derived from the distribution of density. When the water at a certain 

 point :s specifically lighter than its surroundings, it tends to rise, while if heavier, 

 it will tend to sink. Let A, in the section fig. 7, represent light, and B, heavy water, the 

 two strata being separated by an oblique surface. The deeper portion of the light 

 layer A is then surrounded by heavier water, and is specifically lighter than its sur- 

 roundings, according to the Archimedean principle, therefore, it wiU be driven 



