METEOROLOGICAL OBSERVATIONS AT PLYMOUTH. 161 



The results in the three last tables point out, 1st, That the 

 descending semi-oscillations from 10 a.m. to 4 p.m., and from 

 10 P.M. to 4 A.M., approach nearest the mean oscillation of the 

 periods to which they belong. 



Thus in Table XIII. the descending oscillations at these 

 hours differ but little from the mean oscillation of the respect- 

 ive years : — the same is true in Tables XIV. and XV. The 

 descending tides differ but little from the mean of their re- 

 spective seasons; — these differences, however, are not expressed 

 in the tables. 



2nd. It may be perceived that the mean of the two ascending 

 or two descending oscillations is exactly the mean oscillation 

 of the periods to which they belong. 



Hence the differences from the mean are equal, and in 

 opposite directions ; thus in tlie semi-oscillations for summer 

 (Table XIV.), the morning descending oscillation = '014 is as 

 much below the mean oscillation '015, as the evening descend- 

 ing oscillation = '016 is above it. 



3rd. That the ascending tide from 5 to 10 p.m. of summer 

 (Table XV.), approaches very nearly the mean of the year. 



4th. That the descending tides, from 9-10 p.m. to 4 a.m. of 

 summer (Table XIV.), and of summer, autumn and winter 

 (Table XV.), approach also very nearly to the mean oscillation 

 of the year. Hence, if only one set of observations can be 

 made, and those in the day time, it would be desirable to 

 register them during the three months of summer at 4 p.m. 

 and 10 P.M. If two sets of observations at the critical hours 

 can be made, then it would appear desirable to choose either 

 of the periods of the two ascending or the two descending 

 tides, and take the mean of the two, as the mean of the period. 

 If these be continued for a whole year, we thence obtain the 

 mean oscillation of the year ; — if for any season, the mean os- 

 cillation of that season : — whence, knowing the difference from 

 the mean oscillationof the whole year, we may apply the requisite 

 correction. 



So far as these observations extend therefore, we may infer, 

 that four observations in the day in the three months of spring, 

 viz. at 4 and 10 a.m., and 4 and 10 p.m. ; or, otherwise, at 10 

 A.M. and 4 p.m., at 10 p.m. and 4 a.m., would give a mean result, 

 differing from the mean oscillation by a very small quantity. 



There is little doubt, as suggested by Major Sabine, in his 

 Register of the Barometer at Port Royal*, Jamaica, that the 

 amount of the horary oscillation varies with some function of 



* Daniell's Meteorology. 

 1839. M 



