160 Mr J. D. FORBES on the Horary Oscillations 



The general results of the monthly means fully justified my 

 expectations. The extent of the deviations on both sides, from 

 the mean oscillation given by the monthly results, was greatly 

 diminished, and likewise the number of times which the oscilla- 

 tion came out negative, or in the wrong direction. 



6. By consulting Tables I, II, III, and IV, at the end of this 

 Memoir, the partial means of each month during which observa- 

 tions were made will be found for the years 1827, 1828, 1829, 

 and 1830. As there cannot be a doubt that the seasons pro- 

 duce a very sensible effect on the amount of the oscillation, I 

 prefer deducing the general result from the mean of the different 

 seasons, which are contained in the four following Tables, the 

 absolute numbers being restored from the data of the previous 

 ones. The following are the general means, the spring period 

 being understood to include March, April, May ; summer, June, 

 July, August ; autumn, September, October, November ; and 

 winter, December, January, February. 



It thence appears that the principal oscillation from the first 

 morning observation to the afternoon one, amounts to .0114 

 inch ; that from the afternoon to the second evening one, to -f- 

 .0097 : the ratio is 1 : 1.18, which is considerably less than the best 

 observations give for the south of Europe. In the north we have 

 no data of comparison, since, as far as I know, no register capable 

 of affording satisfactory results for the evening maximum has 



