368 A. D. Bache on a Magnetic Survey of Pennsylvania 
Table continued. 
Observed Compnted Diff. obgerved 
Dip. Hip. — computed, 
Group lV. 
Berlin Tavern,* 72°88 73°09 -o-21 
ercer, 72°95 73°02 -0'07 
arren,* 3-00 73:03 -0'03 
Ashtabula,* aa: 73 60 —o'2I 
rie, 53:98 73°85 -0'07 
Group V. 
Duncan’s Island,* 72°58 92°45 +013 
ewistown, 72°50 9757 -0'07 
Huntingdon, 72°30 72°48 -o'18 
rmagh,* 72°31 72°40 -o 
Bellefonte,* 72°70 9686 -o'16 
Curwinsville, 72°83 72°86 -0 03 
Group V 
Belvidere,* 74:16 74:03 +o13 
Ellicottsville, 7430 74°05 +0°25 
Dunkirk,* 74:2 74°21 +0:08 
Niagara Falls,* 74 83 74:76 +0'09 
T 0, 75°19 75°28 -o'0g 
up VIL. 
Bushkill,* 93-53 73°19 +0°33 
Williamsport 72°91 73°19 -0'28 
ilkesarre,* qo 19 73°24 —0'07 
Silver Lake, pe 73°88 -0 19 
ego,* 74.2 74:05 +orl 
Group V II. 
74°46 74.19 +0'27 
Rochester,* 74°72 74°87 -0 | 
7455 74 -ol 
Syracuse, 74 85 i 04 
Oswego,* * 95°19 75°21 -0'09 
West Point,* 73:20 73:4 0°29 
1 73-79 73 38 +041" 
tica, 54°84 74:96 -o'12 
Schenectady, 74-91 74°77 +014 
Troy,* 74 7472 | 4008 | 
° ; : “ 019= 
The probable error of any single observation is +0 12 ae 
72; the probable error of any observation with the regular id 
needles and the Lloyd needles combined is 0°18; rh ig 
latter needles alone, +0°'11. is shows that the Oe mt 
in the observed dip are due to local attractions rather than 
imperfections in the needles employed. It is proper thel : 
to assign equal weights to results by the direct and im 4 
method stpieivin : tions - 
If we apply Peirce’s criterion for the rejection of observa’ ae : 
differing too much from the regular value indicated by all ° 98; a 
observations, we find the limit of rejection to be +0°46 oe as 
7 Maximum difference = 25’. 
