96 The Rev. H. Lloyd on the Results of Observations, ^x. 



mean curve is to produce a general increase of the ordinate between these limits 

 of time, as well as the maximum at 10 p.m. 



We learn from the consideration of these facts, that the ordinary mode of 

 grouping the observations, by taking the mean of all the results at the same 

 hour, — although it truly gives the mean diurnal curve for the period embraced 

 by the observations, — does not represent the average actual course of the move- 

 ment during one day. In order to obtain the representative, or type curve (as it 

 may be called), it seems necessary to treat the ordinates and abscissce as inde- 

 pendent variables, and to take, — not the means of the ordinates corresponding 

 to certain definite abscissa, — but the means, both of the abscissse and ordinates, 

 corresponding to the time of the phenomenon. The former of these will give the 

 mean epoch of the disturbance, and the latter its mean amount. 



We find, in this manner, that the mean epoch of the periodical disturbance 

 is a few minutes before 10 p. m. Its mean magnitude is lO'-O ; and its mean 

 duration is an hoiu: and a half 



