70 



BEPOBT 1886. 



apparent too frequently to leave any donbt that it is the law ; it could only 

 be demonstrated by lithographing examples. I enclose a tracing of mean 

 curves from the tracings of undisturbed days, together with an abstract 

 of the numerical values. The Greenwich turning points agree very closely 

 with those at Dublin. That they are so much earlier than they are at To- 

 ronto seems to me likely to be traceable to some influence of the Atlantic 

 on the mean direction of the currents. I should add to my explanation 

 that, to obtain three months for the winter solstice, January from the 

 beginning of the year was grouped with November and December at the 

 end, as I had not the next following month of January. 



Senhor Capello has sent me his curves showing the general corre- 

 spondence of character between the diurnal changes of magnetic inclina- 

 tion and those of the tension of vapour in the atmosphere. Lloyd in 

 1849 ' showed that the area of the diurnal curves of declination gave 

 an annual progression closely resembling the corresponding progression 

 given by the area of the daily curves of temperature, with which the 

 tension of vapour is so intimately connected ; so that it would seem that 

 there is a closer connection between meteorologic and magnetic jjheno- 

 mena than has been supposed. 



Abstract of mean Solar-diurnal Curves of Declination by measurement from photo- 

 graphic records at Greenwich and Toronto, 1850. Undisturbed days only, grouped 

 in astronomical seasons. 



Note. — The mean date of each group only corresponds approximately to the equinox 

 or solstice, the days being irregularl)' distributed. 



— = E of mean. + = W of mean of the 24h. 



The Greenwich observations were '20m. after the hours named ; the 

 Toronto observations 2^m. before the hours named. 



' Trans. li. Irish Arddnii//, vol. xxii. Pi. I. 



