106 Temperature and Sallness of the River Hugli. [March, 



Hence it would seem that the mean temperature of the surface water 

 exceeds 81° Fahr. every where between Calcutta and the sea. The ab- 

 sence of observations during stormy weather, may have given some trifling 

 excess to the apparent result ; but as the mean of each month is taken 

 from the extremes, and I was upon the river immediately after several 

 violent storms and during many north-westers, and registered the 

 temperature at sun-rise, as well as during the day, I do not think the 

 true mean temperature is lower than 81. Indeed I should rather 

 apprehend an error in the opposite direction ; for in tropical countries I 

 have remarked that the high temperatures, whether of the day, the month, 

 or the season, are of longer duration than the low. A violent storm from the 

 north has been known to depress the temperature to 50° Fahr. in January, 

 at Vera Cruz (Lat. 19° 11) ; but in a few hours it rises again above 60: 

 whereas for months together the temperature there exceeds 80° ; but, 

 except on very rare occasions, never passes 90°. The mean of the 

 extremes of the year would thus give a temperature of only 70° for Vera 

 Cruz, while the true mean, according to different observers, is between 79 

 and 81. In the climate of Calcutta, if we take the extremes at 50 and 

 100, we obtain a mean of 75 instead of the true mean 78 ; and in the 

 preceding table, the extremes of the year, (91 and 68) give a mean 

 79.5, two degrees lower than that of the months duly cast out. I 

 have observed also, that, although a thin surface of water may occa- 

 sionally be brought to the freezing point by radiation in the still clear 

 nights of winter, the hottest sun will not raise it to 110 ; which gives 

 a mean of only 71. 



It may seem extraordinary, that the temperature of a river should 

 exceed that of the climate of the country through which it flows. I 

 believe the reverse is generally supposed, on account of the cooling 

 influence of evaporation, and I should take it to be so with the rivers 

 of England. In tropical rivers, however, especially the Ganges, there is 

 this peculiarity, that the whole, or nearly the whole supply is obtained 

 at that season when the power of the sun is the greatest, and when the 

 evaporation is very much reduced by the saturated state of the atmos- 

 phere. Hence the mass of waters has not time to cool very considerably, 

 before the short winter is over ; and when the sun begins again to act 

 with power, their depth is at its minimum, and the sun's influence rapid 

 in proportion. This difference of 2 or 2j degrees is, however, not much 

 more than Humboldt remarked within the tropics upon the ocean, which, 

 except where influenced by currents, he always found to be warmer 

 than the air immediately above it. We have an example of the same 

 fact in the Bay of Bengal, which, at a short distance to the southward, 



