AIR TEMPERATURE n 



their approach, even before the barometer begins to 

 fall, it seems not improbable that the study of the con- 

 ditions of the upper atmosphere may throw some light 

 upon the problem. Storms in general appear to be 

 eddies in the steady flow of the great air currents 

 which carry on the normal circulation of the atmo- 

 sphere, and they are apparently fed from the per- 

 manent high-pressure areas. 



The storms of high latitudes (such as the extra- 

 tropical parts of the North Atlantic, the North Pacific, 

 and the Southern Ocean) are most frequently cyclones 

 of far less intensity than those of the tropics, but in- 

 volving a very much larger area of the ocean surface 

 and travelling with much greater velocity ; but on 

 account of the gentler pressure-gradients the winds are 

 rarely of such intensity, though frequently of longer 

 duration. There are some peculiarities, still im- 

 perfectly understood, in the storms of the Southern 

 Ocean, which require elucidation. Speaking generally, 

 the force of the wind depends only on the pressure- 

 gradient — i.e., a great fall of the barometer in a long 

 distance may produce a much lighter wind than a 

 small fall of the barometer in a very short distance. 

 The use of wireless telegraphy now permits ships at 

 sea to compare their barometers and so to measure 

 the gradient directly ; but where this is not possible 

 gradient can be judged by the rate of fall of the 

 barometer. Thus, a fall of a tenth of an inch in half an 

 hour shows a far steeper gradient, and forecasts far 

 stronger wind, than the fall of a whole inch in ten or 

 twelve hours. 



Observations show that the air temperature over the 

 open ocean is subject to a much smaller diurnal and 

 annual range than over the land. It is only in such 

 enclosed waters as the Red Sea and the Persian Gulf 



