Horizontal Component or Magnetic Foece. 



355 



of June and July as the most perfect representative of the diurnal curve at Makerstoun for the sun north of 

 the equator, and compare this curve with that similarly obtained for the months of December and January 

 when the sun is farthest south, we obtain the following variations corresponding to the differences of the hourly 

 means for the two curves : — 



121> It i.M. Jl" 3l> lb Sh et Tl" Sh 9I1 101> nil 0>' 1'' P.M. ■i'' 3^ *>> Si" B'' 7*' 81" 9li 10'' 11'' 



3'-43 4'-12 4'-22 5'-0O 5'-69 7'-12 7'-89 7'-77 6'-85 5'-6e 4'-47 2'-29 l''32 0'-40 O'-OO O'-IS l'-44 2'-09 2'-39 2'-21 2'-09 2'-39 2'-64 3'-09 



It follows, therefore, that by the addition of the above quantities, representing the motion of the magnetic 

 needle for the southern hemisphere, which has its maximum of westerly declination about 6*^ a.m., and its mini- 

 mum about 2^' P.M., to the quantities representing the diurnal motion for midsummer at Makerstoun, we obtain 

 variations which represent the midwinter motion for the same place. The same result is obtained in the com- 

 parison of the midsummer curve with that for November and February, and even for October and March, 

 though with less accuracy as regards the epoch of minimum for the differences. 



It will be interesting to determine, from other observations, the latitude for which the midwinter curve 

 ceases to have the characteristic of a northern curve. 



Horizontal Component op Magnetic Force. 



TABLE XXII. — Mean Values of the Variations of the Horizontal Component of Magnetic Force, the 

 whole Horizontal Component being Unity, for each Civil Week-Day, Week, and Month of 1844. 



I 



