678 



In order to avoid in the preceding tables the 

 frequent repetition of the words morning and 

 evening, the hours are counted (according to 

 the ancient method of astronomers,) from the 

 passage of the sun over the meridian, so that 

 the 21st hour corresponds to nine in the morn- 

 ing. The barometric heights are indicated 

 either in millimetres (in the observations of 

 MM. Boussingault and Rivero), or in lines, and 

 hundredths of lines of the French foot (in my 

 observations at Cumana, la Guayra, Callao, 

 Lima* Caraccas, Ibague, Popayan, Mexico, 

 Quito, and Antisana) ; or finally, in inches, and 

 hundredths of the English inch, (in the obser- 

 vations of MM. Kater, Sabine, and Simonoff). 

 The thermometer was suspended by the side of 

 the barometer, when it was not placed in the 

 instrument itself. The barometric heights are 

 not yet corrected by the temperature, that is, 

 they have not been reduced to zero, or to the 

 same degree above the freezing point. It 

 thence results that, as the barometer sinks 

 from 21 hours to 4, while the heat augments, 

 the extent of the diurnal variation is partly 

 masked in the tables, by this increase of temper- 

 ature ; the same thing takes place from 4 till 

 11 ; the movement of the thermometer being 

 still opposed to that of the barometer. On the 

 contrary, the apparent extent of the variations 

 in the atmospheric tides, from ll h to 16 h , and 



