Frazer.] t)4o [April 6, 



ascent at Unit period passed below Washington. It certainly passed above 

 Norfolk. * 



The maximum declination will probably be = 4^.42. 



In Mr. Schott's discussion of the secular cliange in C. S. Report for 1859 

 he mentions tiiat no entire cycle has yet been completed on either the East 

 or West coast. The linear form of the formula first applied to all obsejva- 

 , tions in 1855 (which does not involve any great length of period) is retained 

 for the observations on the Western coast while a circular function is 

 chosen for the others, of which the length of period and other numerical 

 co-efficients have been obtained by the method of least squares. But he 

 adds, "as long as the cause producing the secular change remains altogether 

 unkntiirn it is not safe to trust too far to the continuation of the law thus 

 empirically derived." 



He tinds that "if the stations be arranged geographically, the minimun 

 (West) declination occurred earlier at the Eastern than at the Western and 

 Southern stations." * * * "and if xoe proceed to the Western coast we find 

 that the Eastern declination has not yet reached, its maximum {equivalent to a 

 Western minimum.)" 



This Report concludes with a record of all declination observations em- 

 ployed in the forgoing papers. 



From the preceding imperfect memorandum of the analytical methods of 

 some of the ablest mathematicians and magneticians, who have taken this 

 subject in hand, and especially of those of Mr. Schott, it will be seen : 



1st That to the best informed the cause of variation in declination re- 

 mains still unknown. 



2d The mathematical analysis proves the existence of a period including 

 secondary periods, the latter resulting from perturbations due to some 

 cause not identical with the main cause. 



3d Over small areas and during short intervals of time the magnetic va- 

 riation can be predicted and a formula established which shall express it. 



4th Tiie linear formula having shown the change of the kind of variation 

 in abscissa after a certain epoch and in i\wkind of variation in the ordinate 

 after another epoch establishes the cyclic curve as one of unequal axes. 



5th. Tlie easternmost North and South line which is tangent to this 

 curve passes nearly through Washington and some distance west of Phila- 

 delpiiia. The westernmost North and South tangent is not j^et determin- 

 able with exactness. 



* Hy reroronce to the Isogenic chart previously mentioned, kindly corrected 

 and sent me by Mr. Schott, March 17, 1870, the hitflicst position attained by the 

 Zerocurvc is marked in red ink as above Annapolis and Haltimore in 17!i'i, wliile 

 by computing tlie mean annual declination from the tables from Report of 1S74 

 (Philadcipliia) it will be found that the maximum was from 1740 to 17(50 when It 

 amounted to lO'.S, and the minimum from IHOOto IHIO when it amounted to 0' .0 



The column for Washington only goes back to ISOO but Its minimum seems to 

 have l)een between 1790 and ISOO. From this to the present time this annual 

 maan for Wasliington presents several irregularities. 



