i6 



NATURAL SCIENCE. 



July, 



has been shown by Kriimmel and Schott, where the winds are 

 strongest and steadiest, and we accordingly find the densest 

 surface waters in the regions of the great permanent winds ; the 

 typical examples being those of the trade winds in the North and 

 South Atlantic. Comparing this map with that of mean annual 

 surface temperature, it appears at once that the areas of highest 

 temperature do not in general coincide with those of greatest salinity, 

 but that over the latter there is usually a considerable gradient of 

 temperature from one side to the other. Hence, we derive the 

 important result that the heaviest water occurs very much on the 

 polar sides of the areas of maximum salinity, and that so far as gravity 

 alone is concerned we must expect to find there the chief tendenc}' 

 for the water to sink below the surface, a fact which may come to be 

 of vital interest in connection with the distribution of animals. 

 Space does not permit' us even to mention the many points of great 

 interest suggested by Dr. Buchan's map of surface salinity ; we may 

 merely observe that in the case of the North Sea and the East Indian 

 and Chinese waters it proposes a distribution markedly different from 

 that hitherto accepted. With reference to the latter, great stress is 

 laid on the influence of rainfall in lowering the salinity of the surface 

 waters, and we must confess to some difficulty in following Dr. 

 Buchan's explanation, especially in view of Schott's observations and 

 experiments in this very region. 



In constructing the temperature maps for the depths, Dr. Buchan 

 states that after full investigation for intermediate depths down to 

 200 fathoms, the plane of 100 fathoms was recognised as being beyond 

 the limit of seasonal variation ; single observations are accordingly 

 accepted as mean temperatures. We hope Dr. Buchan may be 

 induced to publish the maps to which he refers ; for a complete 

 knowledge of seasonal variations, to the greatest depth at which they 

 occur in the open ocean, is of first importance in discussing the 

 circulation of waters in closed areas. We know, for example, that 

 temperature in the Faeroe Channel is not constant even at a depth 

 of 400 fathoms, and it is probable that variations in the strength of 

 the current flowing over the W^yville-Thomson ridge are in part 

 responsible for great changes of temperature and salinity in the North 

 Sea and the Baltic. 



The mean temperatures, for all the oceans, at the different levels,, 

 are given as follows, and show an average vertical curve of extreme 

 interest : — 



