102 MR BROUN ON THE RELATION OF THE VARIATIONS OF THE 



the monthly means, as deduced from the observations of horizontal force made at 

 Toronto in 1842, at the following hours : — First, the hours corresponding to the 

 Makerstoun hours ; this could not be done exactly, as the observations were 

 made at Toronto every two liom-s onl}^ and for each of two of the Makerstoun 

 hours, the mean of two of the Toronto hours had to be taken ; thus, one of the 

 observation hours at Makerstoun was 7 a.m. ; the mean of the Toronto observations 

 at 6 and 8 a.m. was taken as equivalent (see the open curve, No. 3, Plate III.) 

 Second, from the whole two-hourly observations (see the open curve, No. 5, Plate 

 IV.) The monthly means from both these methods, gave the same, or nearly the 

 same, annual period as the Makerstoun observations. 



10. I have since then verified this period by the monthly means of the observa- 

 tions made at Makerstoun in the years 1843, 1844, and 1845. I may first refer 

 to the Makerstoun and Toronto curves for 1842, which were exhibited to the 

 Physical Section of the British Association at Cambridge. (See Curves, No. 3, 

 Plate III.) From these I concluded that the horizontal force of the earth's mag- 

 netism has an annual period, consisting of a maximum at each solstice, and a mi- 

 nimum at each equinox ; both curves present a curious inflexion in March, which 

 I then considered due to some in-egularity. The monthly means for the years 

 1 842, 3, 4, and 5, have also been projected together ; (see Curves, No. 4, Plate III.) 

 the whole speak the same language, excepting that the inflexion in March 1842 

 does not occur in the other years, unless it may be said to do so in 1845. All the 

 cui'ves shew a considerable secular change, the horizontal intensity increasing 

 throughout the whole period. Whether this be really an increase of the earth's 

 horizontal intensity, or due to some instrumental cause, cannot be positively 

 stated; it is not at all likely to be due to an increase of the fi"ee magne- 

 tism of the bar, which is suspended at right angles to the magnetic meridian ; the 

 only apparent and possible cause is a stretch of the suspension wires ; it must 

 be a matter for consideration, whether such a cause is likely to operate for such 

 a period, and to nearly an equal amount for two years. Whether a secular 

 change of horizontal intensity, or due to a stretch of the wires, it is evident that it 

 may be considered as a regular increase throughout the year. Upon this hypothesis 

 I have eliminated this increase from the monthly means of the last three years, 

 and projected the mean below the others. This curve shews more strikingly the 

 annual period of solstitial maxima and equinoctial minima. The minima have 

 nearly the same value ; the summer maximum is gi-eater than the winter maxi- 

 mum, but so little, that an error of a thirtieth in the amount of the temperature 

 con-ection, would account for the difference. The annual range from the mean of 

 the three years is 0-000724, or about the mean diurnal range for the three 

 winter months. 



11. M. Hansteen concluded from his observations, that there was a maximum 

 of horizontal intensity in December, and a minimum in June. Colonel Sabine con- 



