262 A. D. Bache on the Horizontal Component of Magnetic Force. 



ism of the bar and 1-8 divisions for the effect of one degree of 

 temperature of Fahrenheit's scale. 



A new term involving the difference between the square of 

 mean epoch and any single epoch with a coefficient, was found 

 to improve this result giving for the monthly change 17-6 scale 

 divisions, for the effect of one degree of Fahrenheit, 1-62 divi- 

 sions, and for the coefficient of the term involving the square 

 0-31. This gives for the annual progressive change 211-2 scale 

 divisions, and for the effect of one degree of the thermometer, q- 

 1-62 X 0-0000365 = 0-0000591 in terras of the magnetic momeat 

 of the bar. This agrees with the best direct determinations, be- 

 ing those in which the Observatory was alternately heated and 



To test these results a combination of the six warmest months 

 ■with the six coldest by alternate means was made, beginning 

 with May, 1841, and ending with April, 1845. The result ior 

 correction for temperature was 1-3 scale divisions for 1° Fahren- 

 heit, confirming the foregoing. In reducing the observations To 

 scale divisions has been used as the correction for each degree 

 of Fahrenheit's scale, q = 0-0000548 in terms of the horizontal 

 force and the observations are reduced to 63°-0 Fahrenheit. 



This discrepancy between the results of the direct experiments 

 for the correction of temperature, and those derived from ne 

 general discussion of all the observations occurred also m tbe 

 Toronto observations, in those at Makerstoun, and in those at bt. 

 Helena, and has been the subject of comment by General Sabine, 

 Mr. Broun, and Dr. Lamont. , 



_ Table No. IT, of the full memoir, contains the results in scale 

 divisions after correcting each value of the former table for tem- 

 perature. 



Table No. Ill gives the resulting monthly means for each year 

 of the whole series from 1840-'41 to 1844- '45 both incl^sive-^^ 



The corrections for regular progressive changes of read'^o 

 in the horizontal force, and for certain irregular changes are "e^ 

 carefully stadied and appended in Table No. IV which shows 

 the monthly means of the bi-hourly and hourly readings oi i 

 bifilar magnetometer referred to 63°0 Fahrenheit, and correc 

 for irregularities in the progressive change. The monthly rr^ec 

 for each year and for the five years are given in abstract iQ ^^ 



The differences in the successive annual means indicate 

 the progressive change may be assumed to have been unn 

 from year to year, and, applying the usual method, an au- 

 progressive change of 220 scale divisions is the result. 



From a study chiefly of other independent observation:- 

 total annual change of horizontal magnetic force is deterr::. 

 to be 16-5 scale divisions, hence of the 220 scale divisi'^- 



