685 



In 1756, a naturalist, whose sagacity and 

 rare merit were not sufficiently appreciated by 

 his contemporaries, M.Thibault de Chanvalon*, 

 first reduced the horary observations he had 

 made in the West Indies, into tables. " The ba- 

 rometer/' he observes, in a work which was not 

 published before 1761, "is entirely useless at 

 Martinique to indicate the variations of the 

 weather; but it affords a singularity which 

 merits to be studied in all its details, and which 

 had been already perceived by an observer at 

 Surinam ; but either from the small confidence 

 which travellers generally inspire, doubt was 

 preferred to investigation, or because it re- 

 quires some celebrity to give credit to extraor- 

 dinary facts, the truth was never clearly pre- 

 sented to the public. The regularity of the 

 horary variations may be said to have been un- 

 known till the journey of M. Godin to Quito. 

 Soon after my arrival at Martinique, I per- 

 ceived that the barometer mounted insensibly 

 the whole morning, and after having remained 

 some time without movement, began to lower 

 at sunset. The most considerable revolutions 

 of the atmosphere do not alter this periodical 

 movement of the barometer, which coincides 



an analogous observation, the table I have given for the ho- 

 rary observations applied to the calculations of the height of 

 places, in my collection of Astron. 0bs. 3 Vol. i, p. 289. 

 * Voyage to Martinique, p. 135 ('20, 21,25). 



