226 TROPICAL NATURE t 
In the north temperate zone, on the other hand, the winds 
are always cool, and often of very low temperature even in 
the height of summer, due probably to their coming from 
colder northern regions as easterly winds, or from the upper 
parts of the atmosphere as westerly winds; and this constant 
supply of cool air, combined with quick radiation through a 
dryer atmosphere, carries off the solar heat so rapidly that an 
equilibrium is only reached at a comparatively low tempera- 
ture. In the equatorial zone, on the contrary, the heat 
accumulates, on account of the absence of any medium of 
sufficiently low temperature to carry it off rapidly, and it thus 
soon reaches a point high enough to produce those scorching 
effects which are so puzzling when the altitude of the sun or 
the indications of the thermometer are alone considered. 
Whenever, as is sometimes the case, exceptional cold occurs 
near the equator, it can almost always be traced to the in- 
‘fluence of currents of air of unusually low temperature. Thus 
in July near the Aru islands, the writer experienced a strong 
south-east wind which almost neutralised the usual effects of 
tropical heat, although the weather was bright and sunny. 
But the wind, coming direct from the southern ocean during 
its winter without acquiring heat by passing over land, was 
necessarily of a low temperature. Again, Mr. Bates informs 
us that in the Upper Amazon in the month of May there is a 
regularly recurring south wind which produces a remarkable 
lowering of the usual equatorial temperature. But owing to 
the increased velocity of the earth’s surface at the equator a 
south wind there must have been a south-west wind at its 
origin, and this would bring it directly from the high chain 
of the Peruvian Andes during the winter of the southern 
hemisphere. It is therefore probably a cold mountain wind, 
and blowing as it does over a continuous forest, it has been 
unable to acquire the usual tropical warmth. 
The cause of the striking contrast between the climates of 
equatorial and temperate lands at times when both are 
receiving an approximately equal amount of solar heat may 
perhaps be made clearer by an illustration. Let us suppose 
there to be two reservoirs of water, each supplied by a pipe 
which pours into it a thousand gallons a day, but which runs 
only during the daytime, being cut off at night. The reser- 
