¢ 
variations of the Magnetic Forces. - 339 
given above, it will also be seen that an irregularity occurs in the 
month of October. In the curve of the declination there is a 
ats. corresponding irregularity; the declination was less in October 
4 oe! September, and much greater. in ‘November than in 
to 
_ shows that the same general law obtained during the eutire pe- 
4 riod of the observations comprised between June, 1840, and the 
beginning of 1844. The following ‘table shows ‘that the varia- 
tions of the difference of temperature between sige and 
Lancaster followed the same law during t this peri 
; Mean on Differences of Temperature between Philadelphia 
d Lancaster, Pa. 
t ‘given in the Report of the Girard College observations, ex- 
- Cept in the case of the year 1841, for which the temperatures 
published. ‘in the Journal of the Franklin Institute i used. 
“The e tempeiatn res for-Lancaster were obtained ri: J 
of the Franklin os and from the origina manuseri 
- W.-M. Atlee’s Reports to the Franklin ii. hen be bed 
the ‘indy ess to Shae in my hands. .The’ obse rvations for Get- 
tysbur ilisle, Harrisburg, and W. Chester were also abtained 
from the ‘published repotts to the, Franklin Institute, under the 
general. system of: meteordlogical | ‘observation established by this 
mise » Phe. ar. observations were made at 7 aA. M., 2 P.M., 
| The temperatures at Philadelphia used in the calculations are 
Tt is*ton obi that the comparatively large plus differen- 
“dot be tiperail that occur in the year 1841, as given in the 
above table, are to be attributed to the fact that ‘the temperatures 
at Philadelphia as reported to the Franklin Institute are higher 
The curve of annual variations of declination above referred to, 
- | Feb. arch April. {| May, | June.|-July..| Aug. | Sept.) Oct. | Nov. | — |: s 
° Oe 
“yh ae _ ; > 41-19 123], [08 |-07 ts 
3°38] 150] 0:06} 023] 038) O91K.... | 0°93) 150) 416 4°71 
e2 —0) —2°3 4 08 12) 17 -09 |-0O7 | 1 49 (518 
-15 |-O4 1-13 Pl -0'6 | 04: | OO -02 
as given in the Girard College observations. Making use of oe 
the latter we ‘obtain for the 
Diff erences of Temperature between Philadelphia and Lancaster 
for 1841. 
Nov. Dec. 
Feb, 
March,| April. | May. | June. | July. | Aug, { Sept. ; Oct, 
10 
° ° ° ° ° 
-13|~051-18 | -01 06 
3 ° 
-18)} ~ —~30 | 
_ The Sata which we have Re leave little room to doubt 
ncipal annual movements of the needle are connected 
iain corresponding movements of the isothermal line ; 
