389 



almost invariable after 664 days. I shall only remark, that to such delicate 

 researches as the variation of the nioiiient of a magnetical needle, which is 

 nearly constant, a single observation is not sufBcient, as the daily regular 

 and irregular variations may make 2" to 3", upon a time of 800". 



The variation of the time T of 300 vibrations, page 17, seems to point out 

 a variation of the horizontal intensity, either periodical (of short period), 

 or undulatory ; for, as the time T has been wthout variation, or even de- 

 creased from 1020, 71, to 1823, 54, in the first period, where, according to 

 the formula M = C + B «""''' the decrease of M should have been greatest, 

 the change of T may have its origin from a variation in the horizontal 

 intensity of the terrestrial magnetism. It had a maximum in 1823, a mini- 

 mum in 1828, or 1829, and another maximum in 1839. I have found the 

 same difference between 1823 and 1827, by different other cylinders, and 

 in different places, as Copenhagen, Altona, and Paris ; also, the same dif- 

 ference between 1832 and 1839 in Gottingen, with the same cylinder. In 

 the letter to Mr Kupffer, I sought to express this variation as a function of 

 the longitude of the moon's ascending node ; and, accordingly, the period 

 should be ISJ years : the time to come will decide. 



I have foi- four yeai-s observed the meteorological instruments here in 

 Christiania, at five fixed hours, 7 — 9 forenoon, and 2 — 4 — 10 afternoon ; in 

 the last year are added O*" and 1^ afternoon. From these observations 1 havo 

 calculated the constants a,, a,, «,, a,, ^ in the formula, 



/3 = f4 + «, sin (a, + /) + «j sin (a, + 2t), 



where /3 is the variable height of the barometer, t the horary angle of the 

 sun, fi the mean value of fi in 24 hours for the whole month. These con- 

 stants are calculated, and curves constructed, for every month in the year. 

 By these constants I have found, that here, in Christiania, lat. 59° 54' 42".5, 

 the barometer has two maxima and two minima in the nine months ; but 

 that in the three months. May, June, and July, when the sun is not six hours 

 below the horizon, the nocturnal minimum vanishes, and the morning and 

 evening maximum coincide in the night. In the winter months, November, 

 December, January, when the day is only six hours, the oscillation in the 

 afternoon is very little, so that it is evident, that, in greater latitudes, near 

 the polar circle, the minimum in the afternoon will vanish, and the two 

 maxima, morning and evening, will coincide in the afternoon. This appa- 

 rent .anomaly is, of course, a necessary consequence of the general rule, 

 modified by the short day. I have in the same manner calculated ten years' 

 observations in Dresden, by Inspector Lohrmann, six times per day, 0'', S*", 

 e"", 9'', 18'', 21''. In Dresden, the two minima are visible through all the 

 twelve months. I have sent the calculations and the curves to Mr 

 Schnmacher, in Altona, and hope he will publish them in his " Astron. 

 Nachrichten." The whole oscillation, between the greatest maximum and 

 lowest minimum in Christiania, is greater than that deduced from the obser- 

 vations of Mr Sommer at Kbnigsberg. The mean temperature at Christiania 

 is less than that stated by L. von Buch, that is, scarce more than4.i°Eeaumur. 

 In 1839 I made a voyage through Denmark and tho northern parts of 

 Germany, and observed the following horizontal intensities, expressed in 

 absolute unities after the roethotl of Gauss (unities of longitude = 1 millim.. 



