OF THE MAGNETIC DECLINATION. 



23 



and for the whole year 7'.89; all corresponding to an epoch removed about one 

 year and a half from the epoch of a minimum of the solar period. 



The numbers expressing the diurnal range exhibit three remarkable features, 

 viz., the maximum value in the month of August, the sudden falling off in the 

 months of September and October (see the graphical representation), and the 



DICBNAI RANGE OP THE DECLINATION. 



minimum value in November or December. Otherwise the progression is regular; 

 the curve is single-crested, a feature equally true for the eastern as well as for the 

 western deflection when viewed separately. This latter circumstance is of special 

 importance, since it is probable that it is mostly by the interference of these two 

 separate curves that we observe at other stations the curve of the diurnal range 

 at some stations apparently, to be a double-crested one. The curves for Milan, Munich, 

 Gottingen, Brussels, Greenwich, Dublin, etc., for instance, exhibit two maxima, 

 one after the vernal equinox, and a second, generally the smaller one, about the 

 summer solstice, with more or less regularity. The system to which Philadelphia 

 belongs is exemplified by the annual curve of the diurnal range at Prague and at 

 some Russian stations, especially at Nertschinsk, but principally at Toronto, for 

 which last station the curve is shown in the diagram. Neither station appears to 

 have a tendency to a secondary maximum about the month of April, leaving the 

 maximum about a month and a half after the summer solstice, a well-marked North 

 American feature. 



Annual Variation of the Declination. In connection with the preceding discussion 

 the annual inequality in the magnetic declination next claims attention. 



This subject presents greater difficulty, inherent in the observations, than the 

 diurnal inequality; not so much on account of the length of the period as on 

 account of the difficulty of keeping the instrument in precisely the same condition 

 of adjustment throughout the year. In the first part of this discussion I have 

 already had occasion to refer to this circumstance while investigating the annual 

 effect of the secular change, and it was there shown that the Philadelphia observa- 

 tions share in this respect the difficulties of those of other stations, 1 in consequence 

 of which the results must be received with caution. 



1 It may be proper to give here, in full, Dr. Lloyd's instructive note on this subject, in his discussion 

 of the Dublin observations : " The determination of the annual variation is much more difficult than that 



