374 Results of Makerstoun Observations, 1844. 



The balance-needle was removed on the 27th January 1844, for the purpose of determining its 

 temperature coefficient; it was readjusted on the evening of the same day; all the observations before that 

 date have been rendered comparable with those after it by the application of a correction. — See Introduction. 



In order to render the means in Table XV., p. 238, 1843, comparable with the means in the preceding 

 Table, the following formula must be employed : — 



V= '" ^ % 0-00223. 



Where V is the reduced value of the mean for 1843, v is the value in Table XV., 1843; c is the 

 correction for the mean of 9 observations to the mean of 24 observations, as given, line 6 from foot of page 

 240, 1843 ; 223 micrometer divisions being the difference of the zeros for 1843 and 1844. 



The mean value of the variations of the vertical component, from Table XXXVIII. = 0"006781. 



Secular Change. — When the monthly means for 1843 are rendered comparable with the monthly means 

 for 1844, we obtain the following comparisons for the secular change of the vertical component : — 



1843, . 0-00 



1844, . 000 

 Secular j q.^^ 

 chanse, 



")"■' 



Jan. Fell. March. April. May. .lunt. -luly. Aug. .Sept. (Jet. Nov. Dec. 



9922 9717 9312 8989 8794 8879 8733 85.54 8163 8120 8093 7998 



7838 7458 7341 7384 7061 6847 6552 6341 6267 6129 6155 6003 



2084 2259 1971 1605 1733 2032 2181 2213 1896 1991 1938 1995 



The mean secular change corresponding to the year, from July 1, 1843, till July 1, 1844= — 0-001994. 



A comparison of the mean for 1844 with that for 1845, gives the secular change for the year from 

 July ], 1844, till July 1, 1845= - 0001837, whence 



The mean secular change corresponding to the year, from January 1, 1844, till January 1, 1845, 

 = -0-001915. 



The mean secular changes for three years are as follow : — 



July 1, 1842, till Julv 1, 1843, =-0002298 



1843, "... 1844, =-0-001994 



1844, 184.5, =-0-001837 



Annual Period. — The mean vertical component diminishes with the greatest rapidity from .January to 

 February, and in the months from May till August ; it increases slightly from March to April, and from 

 October to November. If the secular change be eliminated from the means at the foot of Table XXXVIII. 

 by the application of the corrections + »i. 0-00016, where n is the number of the month after January, we 

 obtain the following quantities : — 



0-00 



These quantities give a result considerably different from that obtained from the observations in 1843; 

 whether tliis difference be real, or mei-ely instrument.al, cannot be determined at present. From the above 

 mean values of the variations, the vertical component is a minimum in August, and a maximum in December 

 or .January ; it is also a secondary minimum about the end of February, and a maximum again in April. 



Differences of the Daily Means from the Monthly Means. — The following are the average differences of the 

 daily means from their corresponding monthly means : — 



In order to eliminate the accidental irregularities, the mean of each three months has been taken as a 

 mean for the middle month ; these are given in the second line above, from whence, the differences of the daily 

 means from their corresponding monthly means are a maximum in April, and a minimum in August ; they are 

 a secondary maximum in October or November, and a minimum in December. These epochs are, on the whole, 



