On the Variation of the Magnetic Needle. 



225 



Mean temperature at each hour of the day. 



Nov. 

 Dec. 



1835. 

 January, 

 February, 

 March, 

 April, 

 May, 

 June, 

 July, 

 August, 

 Sept. 

 October, 

 Nov. 



44.7 

 36.4 



37.2 

 33.7 



3S.8 

 49.3 

 68.7 

 68.5 

 71.3 

 70.6 



44.0 

 36.9 



34.4 

 33.4 

 39.6 

 51.9 

 62.0 

 70.0 

 71.4 

 70.3 

 60.1 

 61.5 

 i2.8 



45.4 

 39.1 



29.8 

 34.4 

 43.9 

 51.7 

 63.0 

 68.8 

 71.8 

 70.3 

 60.2 

 63.3 

 54.5 



45.9 

 37.1 



.34.1 

 35.7 

 43.3 

 51.7 

 63.7 

 69.6 

 72.7 

 69.9 

 62.5 

 61.2 

 ■2.4 



10 11 12 I 



40.1 



49.6 

 41.4 



.36.6 



38.2 

 44.7 

 .53.4 

 65.5 

 71.1 

 75.5 

 73,4| 

 63,71 

 82.3 

 .55.1 



49.4 

 41.3 



49.3 49.6 48.4 48.9 

 417,42.140.2,39.6 



36.6 

 .38.0 

 46.5 

 54.6 

 65.0 

 72.9i70, 



7 38.2i37.5'35.6 



7 38.9136.8 36.2 



8 45.3144.8 42.5 

 8 .5.3.9155.3 5.3.8 



7.5.9 

 75.2 

 65.5 

 62.0 

 53.2 



4 63.8 

 9 73.3 

 3 78.1 

 2 74.7 



5 68.4 

 664.4 

 9 56.3 



66.4 66.1 

 71.4 70.9 

 76.3 75.2 

 73.7 74.4 

 64.166.0 

 64.2 63.9 

 51.954.6 



46.5 47.4 

 .38.3,38.8 



32.0,35.2 



35.5.35.7 

 39.l|42.4 

 52.9 52.1 

 63.9:61.7 

 71.367.8 

 75.7]74.1 

 72.867.5 

 6.3.764.1 

 64.0;61.7 

 54.1I5I.6 



47.5 

 38.6 



9 1 10 Mean. 



46.946.9 



38.9,38.6 



32.932.0 

 36.7 37.0 

 41.042.6 

 .50.7i51.2 

 64.0;63.6 

 71.2168.5 

 73.1172.5 

 70.5172.9 

 60.0!64.1 

 65.2 62.3 

 52.2'54.2 



47.4 

 39.3 



35.0 

 36.4 

 43.1 

 52.7 

 64.8 

 70 3 

 74.1 

 71.9 

 63.5 

 63.1 

 53.5 



That temperature has an influence on the amount of the diurnal 

 variation can hardly be doubted. Thu,s, in November, 1834, this 

 variation was less than in November, 1835. The thermometer in- 

 dicates it to have been a colder month. The variation, during the 

 winter months, is uniformly less than during the summer months. 

 Yet it does not appear that this variaiion is strictly proportioned to 

 the temperature, for then the variation must have been greatest in 

 July. To attempt satisfactorily to explain the cause of this diurnal 

 variation, with the present limited number of observations, seems 

 almost hopeless. The fact of the daily variation was first discover- 

 ed by Mr. George Graham, in 1722. The discovery, however, at- 

 tracted little attention until 1750, when the subject was taken up by 

 Wargentin, secretary to the Swedish Academy of Sciences, and in 

 1759, Mr. John Canton, an English philosopher, made about four 

 thousand observations on the same subject. 



Since this time, like observations have been made by Van Swin- 

 den, Gilpin, Hansteen and Beaufoy. The following table exhibits 

 the mean diurnal variation for each month of several years, as found 

 by different observers. 



Gilpin, in 1787. 



10' 12'' 



10 24 

 15 



17 24 



18 54 



19 36 

 19 36 

 19 24 

 15 30 

 14 18 



11 6 

 8 18 



Vol. XXX.— No. 2. 



29 



