224 METEOROLOGY. 



other engagements multiplied so much that I fonnd it utterly impos- 

 sible to do 80. The clock I used is a first rate time-piece, and was 

 compared, from time to time, with the meridian line laid down by 

 observation of the north star. 



No. 7 contains the mean barometrical height* of all the half-hourly 

 observations made in 1857, recapitulated chiefly from table No 6, to 

 "which are added the monthly means of the barometer at 7 a. m., 

 2 p. m., and 9 p. m. These latter means are not the means of the 

 tvhole months, but of those days only on which half-hourly observa- 

 tions have been made. 



No. 8 exhibits the curves of mean height of the barometer in its 

 course from 7 a. m. to 9 p. m., laid down according to the numbers 

 in table No. 7. We find these curves to be much more regular, and 

 more gradually and smoothly rounded off in proportion to the number 

 of days, of which they are the mean result. So is, for instance, the 

 curve c, resulting from the mean of eight months, more smoothly 

 rounded than the curve of the month of August or that of May ; and 

 these again more smoothly rounded than the curves resulting from 

 single days, which I have sent in my previous communications to you, 

 and whicii are more or less angular. 



The curve c in diagram No. 8 may therefore serve to illustrate the 

 true and normal course of the periodical rise and fall of the barome- 

 ter from 7 a. m. to 9 p. m. Its rise and fall from 9| to 11 a. m., 

 and from 4 to 5 p. m., are very inconsiderable ; butfrom 1 to 2| p. m., 

 and from 7 to 9 a. m., far more rapid. There exists also a dilference 

 in the shape of the curve of the month of August compared with that 

 of the month of May. 



No. 9 contains the mean amplitudes of the barometer, calculated 

 from table No. 6, for periods fi'om six to six successive days; also, the 

 mean am])litudes for periods from 12 to 12 days, and likewise those 

 for the different months. All these mean amplitudes are laid down 

 in diagram No. 10, in their proper position, according to their numeri- 

 cal value. 



No, 10. The monthly means in the first curve exhibit a pretty 

 regularly rounded curve. The second curve, that is, the curve of the 

 twelve daily means, is somewhat more irregular. The third, or six 

 daily' curve, is still more irregular, exhibiting many projecting 

 corners. The fourth and lowermost, which can no longer be called 

 a curve, and of which, to save time, I have given only a small portion, 

 exhibits in a striking manner the great fluctuation of the daily am- 

 plitudes. 



At first view there seems to be not the^ least tendency in them to 

 follow a certain law with regard to their mean value. This law, 

 however, becouies apparent, when we look at the next curve above, 

 and still more so in the two uppermost curves. The third curve 

 shows, also, that the nature of the curve of mean amplitudes is not 

 the same in every year ; in 1857, for instance, it is much higher from 

 October to December than it was in LS56. All this indicates plainly 

 enough that the daily amplitudes of the barometer are subject to 

 great disturbances by -some cause or other. 



Great as the irregularities caused by such disturbances are in tlie 



