ElectricUy — Magnetism . 433 



and December 1819, and in March, April, and May, 1620, 

 seven or eight times every day, the time of 300 oscillations, 

 (performed by a magnetic steel cylinder, suspended by a very fine 

 silk thread,) by which I have found : 



First, That the niagnetic intensity of the earth is subject to a 

 diurnal variation, so that it decreases from the first hours of the 

 morning till about ten or eleven, when it arrives at its minimum ; 

 from that time it goes on increasing till four in the afternoon, and 

 in the latter months till six or fcven in the evening. This force 

 afterwards decreases anew during the night, and about three 

 in the morning reaches its maximum, when it again returns, |by 

 little and little, to its minimum about ten or eleven in the 

 morning, and so on continually. 



Second, That whenever the moon passes the equator, the 

 niagnetic intensity is considerably weaker in the two or three fol- 

 lowing days. 



Third, That the magnetic intensity is still more reduced during 

 the appearance of an aurora borealis, and is so much the weaker 

 as this meteor is extensive and powerful. The common intensity 

 returns only by degrees, and 24 hours afterwards. 



Fourth, That the magnetic intensity appears to have a very 

 considerable annual variation, being stronger in the winter months 

 than in the summer months. 



«' My second magnetic discovery is the following:— I have 

 fouitd that every vertical body S. N. whatever it be, and of what- 

 ever kind of matter has a north pole at bottom, and a south pole 

 at top, as in all vertical bars of iron, for 

 otherwise it would be impossible to explain 



N. the phenomena which 1 have observed, and 



c which I have determined by a great number 

 of incontestible experiments, namely, that 

 the magnetic cylinder oscillates more quickly 



N. towards the north in a, and more slowly 



, a towards the south in b. And, on the con- 

 N. trary, it oscillates more slowly towards the 



north in c, and faster towards the south 

 in d. I have found this law constantly confirmed by my 

 experiments near the walls and partitions of houses, whe- 

 ther of wood or stone, and even near large trees in the gar- 

 den. This action must necessarily exert its influence on the 

 direction of the compass-needle on ship-board. The whole mass 

 of wood in a ship has in this way a magnctical axis, and the 

 observed variation of the compass ought rather to be attributed 

 to this influence than to that of iron, guns, and ballast carried by 

 the vessel. Hence it results that all observations on the mag- 



