Variations of the Barometer. 291 



when the thermometer was highest, and generally higher in 

 the night than the day, by from two to four lines ; and that 

 this instrument suffered a greater change from the morning 

 till the evening, than from the evening till the morning. 



It was in 1722 that the phenomenon of the hourly varia- 

 tions was observed, for the first time, and with great exact- 

 ness, by a Dutch philosopher, whose name has not descended 

 to our times. In the Journal Litteraire de la Haye, it is said. 

 in a letter from Surinam, " The mercury rises in this part of 

 Dutch Guiana every day regularly, from 9 h in the morning 

 till half-past ll h , after which it descends till 2 or 3 o'clock 

 in the afternoon, and afterwards it returns to its first height. 

 It suffers nearly the same variations at the same hours of the 

 night. The variation is only one-half or three-fourths of a 

 line, or, at most, a whole line. It would be desirable that the 

 philosophers of Europe would make their conjectures on this 

 subject.''''* The observations which I have made seventy-seven 

 years later on the same coast of Surinam, on the banks of the 

 Orinoco, have confirmed, with the exception of the hour of 

 maximum in the morning, the accuracy of the first determina- 

 tion of the periods. 



From 1740 to 1750, Father Boudier observed the barome- 

 ter at Chandernagore in India. He found that the greatest 

 elevation of the mercury took place every day about nine or 

 ten o'clock in the morning, and the least elevation towards three 

 or four o'clock in the evening ; and he remarks, that, during 

 the great number of years that the barometer was marked at 

 Chandernagore, there were only eight or ten days in which 

 this uniform motion of the mercury was not observed. 



The academicians sent to Quito in 1735, observed also the 

 horary variations of the barometer, and they found that it 

 reached its maximum at nine in the morning, and its minimum 

 at three in the afternoon, the mean difference at Quito being 

 lj of a line. 



In 1751, M. Thibault de Chauvelon reduced into tables the 



• We trust that M. Arago will now have the candour to divest his coun- 

 tryman, M. Godin, of the honour which he bad erroneously conferred upon 

 him, of having been the first to observe the horary variations of the baro- 

 meter. — Ed. 



