216 



Prof. W. A. Norton on Terrestrial Magnetism. 



•500. Now by taking L = latitude of New York, and I = latitude 

 of the place for which the horizontal intensity is to be computed, 

 and making use of the observed temperatures, the respective val- 

 ues of C determined as above, are those which have been empir- 

 ically substituted in equation (14), and used in calculating the 

 following table. 



Table VII. 



Place. 



Montreal, 

 Washington, 

 Chapel Hill, Lat. 36°, 

 St. Augustine, Lat. 30°, 

 Key West. Lat. 2U°, 



IIU1. MI 



itMJMiy. 



Diff. 



Values of C. 



• 



1 Computed. 



Observed. 



•44o 



•409 



+ u3t 





•5?3 



•5 7 6 



- *oo3 



•38o 



•635 



•645 



- -oio 



•33o 



•759 



*75o 



+•009 



•323 



♦8l6 



•826 



- -oio 



•211 1 



By taking the data furnished by the third column of the small 

 table of temperatures just given, I find that on the western coast 

 of Europe and Africa, the difference between the temperature at 

 the equator and the temperature at a given latitude, varies as 



1 -cos wlat., n having various values for different latitudes, from 



2 to 15. This makes the temperature proportional to 1+ 

 cos wlat. 4-C, C being some constant number, either positive or 

 negative. For, supposing this to be the law of variation of the 

 temperature, and denoting the temperatures at the equator and a 

 given latitude (/) by T and T', we have T : T*c: 2+C: 1+ 

 cos«/+C. Whence T : T -T' : : 2 + C : 1 -cosnf, and T-T' 



2^q( 1 - cos nl). For any other latitude (L) we have in like 



manner T - T" =-- 



T 



2 + q( 1 ~ cos wL )- These equations give T - T' 

 . 1-cosm/:1-coswL. We have therefore for the 



calculation of the horizontal intensity on the coast of Europe, the 

 following formulas. 



:T-T 



) 



Hor. intensity =iH 



Or, assuming C •= 0, 



Hor. 



H 



1+COS7 1/+C 



1+coswL + C' 



1 4- cos nl 

 l+cosnh 



Table VIII. contains the results of numerous calculations made 

 with this formula. The values of n, as given in the table, were 

 obtained from the consideration that for each place the excess of 

 the observed temperature at the equator over that of the place, 

 was proportional to 1 - cos wlat. H was taken = -814= hor. in- 

 tensity at latitude 25° on west coast of Africa. The temperature 



^k ^ qUat ° r ° n the west coast of Africa was taken at 27 '8* 

 i he difference between this and the temperature at latitude 2o , 

 was obtained by means of the table given on page 2 15, by sup* 



