114 



of. W. A. Norton on Terrestrial Ma 



The second result for the equator, at the end of the table, was 

 calculated from -814, the observed horizontal intensity at latitude 

 25°, W. coast of Africa. 



The following observed elements were obtained by estimation 

 from the tables and charts of Sabine, Gauss, and Loomis : Chapel 

 Hill, intensity = 1-77, dip = G8° 37'; St. Augustine, intensity = 

 1-66, dip =:63°; Key West, intensity =1-55, dip ^58°; equa- 

 tor, E. coast of S. America, intensity =1-06, dip =28°; latitude 

 25°, W. coast of Africa, intensity =1-3, dip =51° 15'; equator 

 W. coast of Africa, intensity =-920, dip =0°. 



It will be observedjhat the differences for Europe are greater 

 than for America. 

 form ula 



i 



The following table was calculated from the 



* 



Hor. intensity 



H 



_ t±2Q° 

 T+20° 



Table V. 



(12.) 



Place. 



Moscow. 



Chrisliania, 



Konigsberg, 



Berlin, 



Edinburgh, 

 Gottingen, 



Hor. Intensity. 



computed. 



4o8~~ 



438 

 48 1 

 486 

 5oo 



observed. 



5o4~ 

 •43 7 



•4?9 

 •509 



•446 

 *5io 



Diff*. 

 •094 



oi3 



o4i 

 028 



Place. 



i Hoc. inten sity. ... - 



London, 

 Milan, 



Marseilles 



Naples, 



+-o4o] Lat. 25°, W. coast of 



Africa, 

 Equator, do. 



•010 



computed, 



56 1 



5 7 3 



6i4 



716 

 800 



ibserved 



•571 

 •543 



.65 7 



•8i4 

 •920 





+ 023 



010 



Lf-o3o 

 -•o43 



—•098 



-• 1 20! 



The differences here are generally less than in the previous 

 table. The same formula would however give results wider 

 from the truth than the first one, for the western continent. 



More accurate results might be obtained by making the calcu- 

 lations from the observed horizontal intensities of a greater num- 

 ber of places. 



Theoretical investigations have conducted to the law that the 

 change of temperature of the earth is proportional to that of the 

 square of the cosine of the latitude, and observations are said to 

 lead to the same conclusion. Assuming this law to be true, the 

 temperature must every where be proportional to the square of 

 the cosine of the latitude, plus or minus some number. That is, 



±C. This gives the formula 



Hor. intensity =H 



cos-1 



(13.) 



nor. intensity =H — . TT - rw* ™^^ ^6.) 



cos=L±C- M l+ C os2L±C ' ' * 

 / and L denoting two different latitudes, and H the horizontal 

 intensity at L. The results given in the following table are cal- 

 culated from this formula, taking C=--0, and H= horizontal & 



I and L denoting two different latitudes, and H 



tensity at New York 



5291. 



