at Leith Fort, every Hour of the Day, in 1824 and 1825. 29 



I have given these examples principally with the view of 

 showing the application of the results of the hourly register, 

 and not with the design of contrasting the hours employed by 

 different observers ; for it yet remains to be determined how 

 far the form and dimensions of the daily curve, as determined 

 for Leith, are applicable to places in different latitudes, and 

 situated at different heights above the sea. At Paris, for ex- 

 ample, the mean temperature of the day occurs before 9 

 o'clock in the morning ; at Tweedsmuir in Scotland, J 300 

 feet above the sea, it happens before 7| h A. m., and at Salem 

 Massachussets, before 8 o'clock a. m. ; but it must be remark- 

 ed, that the observations at 9 o'clock, and at 7| h and 8 h , are 

 compared with a calculated mean temperature, and not with 

 the mean temperature of the whole 24 hours. 



It is curious to remark, that, with the exception of the 

 hours of 10 a. m., and 10 p. m., no similar pair of hours has 

 been used by meteorologists. The following table will show 

 how nearly at Leith the mean of every similar pair of hours 

 approaches to the mean temperature of the day. 



TABLE, showing the Difference between the Mean Temperature of 

 every similar pair of hours, and thut of the Day. 



In some instances, meteorological registers have been kept, 

 in which the thermometer has been observed only once a-day. 

 These registers may now be rendered useful, by means of the 

 following table, which shows the relation between the mean 

 temperature of each hour, and that of the whole day. 



