698 



It results from this table, that, even in the 

 places where, near the extremity of the equi- 

 noxial zone, the horary variations become less 

 regular than at Cumana, the epocha of the 

 maximum does not vary as much as the dura- 

 tion of the stationary state. We find for 

 Calcutta : 



Mean of the app.max . 9h 17' 



of the real max. 10 5 



of duration. 1 36 



Now, the variations of the apparent maximum 

 are separated from the mean, more than eigh- 

 teen minutes of time, fourteen times on twenty- 

 six ; while the same separation of 18' is found 

 in the duration of the stationary state, nineteen 

 times. The epochas at Calcutta of the apparent 

 maxima and minima are, employing the ob- 

 servations of a whole month, 9 h 15' in the 

 morning, 2 h 36' afternoon, 9 b 32' of the evening, 

 and 3 h 12' of the morning. The real maxima, 

 that is the real epochas of the maxima, succeed 

 nearly an hour to the apparent maxima. 



Are the epochas of the extreme limits, which 

 we have called, with M. Ramond, the tropical 

 hours *, the same over all the earth ? That 



* JVendestunden, hours in which the movement returns on 

 itself, and which must not be confounded with the hours of 



the tropical or equinoxial year. 



