dipping compasses ; that, M they exist (as we 

 must suppose) they do not exceed eight or ten 

 minutes * ; and that the far more considerable 

 horary changes, mentioned by different au- 

 thors, should be attributed to the imperfect 

 levelling of the instrument. In spite of these 

 well founded doubts, I did not hesitate on the 

 first of November, to place Borda s great dip- 

 ping apparatus in a spot well calculated for 

 delicate experiments of this nature. The dip 

 was invariably 43*65°. This quantity was the 

 mean of many observations made with the great- 

 est care. On the 7th of November, three days 

 after the violent shocks of the earthquake, I re- 

 commenced the same series of observations, and 

 was astonished to find, that the dip became 

 less by 90 centesimal minutes. It was no more 

 than 42*75°. I thought, that perhaps it would 

 again augment, and return progressively to it's 

 first state ; but I was disappointed in my ex- 

 pectation. A twelvemonth after, on my return 

 from the Oroonoko, I still found the dip of the 

 magnetic needle at Cumana to be 42*80° ; 

 though the intensity of the magnetic forces had 

 remained the same before aud after the earth- 



* The annual changes of the magnetic dip appear in bur- 

 climates to be from four to five minutes ; but, according to 

 the analogy of the daily and annual changes of the variation, 

 we are not indispensably obliged to admit, that the daily 

 changes are less considerable than the annual. 



* 2 



