On the Hwary Oscillations of the Barometer at Rome. 225 



From a view of the two sets of results at the bottom of it, 

 it is easily evident that those derived from the second supple- 

 ment merely destroy the others, and they have only been in- 

 serted to show the complete execution of the ti/pe, for the num- 

 bers are too few to correct each other, and the difference of 

 — 005 between 10 a. and 10 m. (of the same day) is utterly 

 insufficient for correcting them. The morning observations are 

 therefore quite destroyed, while the evening ones are less af- 

 fected. Let us then notice the Jlrst results. The morning maxi- 

 mum occurs between 10 and 11, probably near the latter hour, 

 as it diiTers from 10 positively ^ merely loloo^h of an inch. 

 Whether the mercury is actually falling at 7 and 8 o**clock it 

 is not so easy to decide. The observations at the former time, 

 from their small number, are not much to be depended on, but 

 they are abundant at the latter hour, and appear to indicate 

 the conclusion of a fall, but it is probably inconsiderable, some 

 fluctuation between the morning minimum and maximum. 

 The mercury sinks regularly from 11 till 3. At that hour a 

 temporary elevation is shown ; but I am inclined to think it 

 erroneous, and arising from an overplus of observations with 

 the sign + , and an accidental deficiency of the others, which 

 an inspection of the column will confirm. At 4 it decidedly 

 attains the afternoon minimum ( — 0170), and fluctuates till 

 about 7, when the height is nearly the same, and then rises 

 constantly for four hours. By 11 it has rapidly risen, and at 

 that hour we have the evening maximum : at 1 2 it is lower. 

 During the night I can give no account. Part second of this 

 table shows all I know concerning it from three sets of obser- 

 vations, but two are so affected by irrelevant causes that I can- 

 not pretend to discover the truth. They tend to indicate a maxi- 

 mum at 3 M., when we ought rather to have a minimum. I 

 Tiere give the mean results in inches of mercury found by ap- 

 plying the means in the columns to the height of 10 m. It only 

 includes part first of the third table. 



4 a. 

 5 

 6 

 7 



29.9730 

 9782 

 9804 

 9729 



D. 



— 188 



-f 52 

 + 22 



— 75 



