Prof. W. A. Norton on Terrestrial Magnetism. 



219 



"Ver. intensity 



C 



cos 2(1- 5°) -cos 2(2+5°) 



— ^ m m ^^^^ — ^— ,, - - - i r - ^^^^^^ ~ 



cosrf 



Or, designating by V the observed vertical intensity at some par- 

 ticular place, by L the latitude of this place, and by D the decli- 

 nation of the needle there, we have 



Ver. intensity 



V 



cos 2(1 -5° ) - cos 2(1 + 5°) cosD 

 cos2(L-5°7-cos2(L + 5°) 'cosrf 



It was remarked on page 10, of this volume, that it might be 

 found necessary to increase the results of formula (16), except in 

 high latitudes, by reason of the deviation of the change of tem- 

 perature from the supposed law of uniformity. If any term or 

 factor is to be added to formula (17) on this account, it must be 

 some function of the differential of cos 2 /, that is, of sin 21 



In the calculation of the following table, V was taken equal to 

 17236 = vertical intensity at New York ; except for St. Louis, 

 Dubuque, Louisville, and Hudson, (the first result,) whose verti- 

 cal intensities were computed from that of Cincinnati (1*643.) 



Table IX. 



Place. 



Washington, 



Hadaon,Ohio, 

 do - do. 



1-766 



*'747 

 i'743 



*'74i 

 1*742 



i- 7 34 

 i-73i 



1-719 



1-714 



Ver. Intensity, 

 computed, observed. 



---■ 11 !■ m n-i— rrm 



1*7236 

 1.758 

 1*720 

 1*755 



i-744 



1-741 



1-757 



1 '706 

 1-725 

 1-716 



1-690 



1-727 

 1-727 



Diff. 



-f-oo8 

 4-027 



- -012 



- -oo3 



4-'OOI 



-•023 



4"*025 



- -006 



- '002 



4-006 

 -•070 



- *oo6 



Ver. Intensity 

 computed, observed. 



Cincinnati, 



Prairie du Cbien* 



Dubuque, 



Louisville, 



St. Louis, 



Lat. 35°, E. coast 

 of U. States, 



Lat. 30°, do. do. 



Lat. 25°, do. do. 



Lat. 20°, W. In- 

 dia Isles, 



Lat. ]0°,E. coast 

 of S. America. 



I'7o5 



x-643 



1-690 



1-707 



1-685 



1-690 



1-633 



i-646 



1-649 



i-636 



1-638 



r648 



i-5io 

 1-335 



i" 1 20 



600 



1-479 



i-3i4 



i- 160 - .o4o 



•771 I --171 



ao. 1-721 1-727 -•00b ot a. nmern-a.i -uw //* i - »/* i 



South of latitude 20°, the difference becomes large. For this 

 •««e reasons may be assigned. 1. Errors in the "observed" el- 

 ements obtained by estimation. 2. The deviation of the change 

 of temperature from uniformity. 3. The inaccuracy, in the vi- 

 Cu »ty of the equator, of the supposed law, that the variation of 

 temperature is proportional to the variation of the cosine of twice 

 toe latitude ; for this law supposes that on each meridian the tem- 

 pature has its maximum value at the equator, whereas Captain 

 U'iperrey has found that the warmest point of each meridian is 

 ^e point of intersection with the magnetic equator. This is 

 Pmbably the principal source of the errors in low latitudes, 

 toagnetic equator traverses South America at a distance of some 

 lu south of the geographical equator ; and thus in low latitudes 

 ° n the western continent, the computed vertical intensities ought 

 f come out considerably too small. In the calculations which 

 l01 W, f or the eastem continent ( Ta ble X,) this source of error 



The 



