137 



which I offered to the Royal Irish Academy. I should be 

 glad to learn that it has proved acceptable, however trifling. 



" As I have always remarked that scientific discoveries 

 are long known either in Ireland or in France, before they 

 travel from one country to the other, I think it may be 

 gratifying to you to become acquainted with a few parti- 

 culars of my voyage. 



" We made three observations every hour, day and night, 

 from the 11th of February to the end of March. The in- 

 struments examined were, 1st, the horizontal magnetic 

 needle ; 2nd, the thermometer ; 3rd, the barometer, a 

 niveau constant; 4th, Saussure's hair hygrometer. The 

 direction and force of the wind, state of sky, &c. were also 

 observed. 



" The variations of the needle were far greater than in 

 Paris ; the hours of maxima and minima agreeing very well 

 together, except near the time of the sun's passage through 

 the zenith of Olinda, (lat. 8°. 0'. 58". long. 2 h 19 m W.) I 

 then remarked two important phenomena; 1st, the extreme 

 digressions, a. m. in one sense, became p. m. in the same 

 direction, when the sun began to culminate in the other 

 hemisphere, after passing through the zenith. 2ndly, this 

 remarkable alteration was preceded by sudden and perma- 

 nent changes in the variation of the needle, amounting to 

 more than one degree. The first of these variations took 

 place twelve hours after the sun's centre had reached a 

 declination equal to the latitude of the place. All these 

 sudden changes were accompanied with feeble storms con- 

 fined to one small part of the horizon. Referring to the 

 problem as laid down in the Annuaire du Bureau des Longi- 

 tudes for last year, it would seem that the transition from 

 the daily variations belonging to the northern hemisphere to 

 those which characterize the southern part of our globe, is 

 not on the magnetic equator, but depends on the sun's path 

 in the heavens ; and the sun acts here not as a source of 



o2 



