the defending Node of Saturn. 4r 
0/3 , the error in the heliocentric longitude of the node, 
b + & 
ad 
b + l 3 
If the fault in the fouthern latitude = — d, in the 
northern latitude = -\-d, the fame formula is flili true; but 
then ENzEw, and the place of the erroneous node will be 
between E and N. In both cafes the errors in the place of 
the node are directly as the errors in the latitudes. 
Let us now fuppofe, that only the 
one latitude is erroneous ±d. Then 
N« = ±E«^EN = rtxr-*- 7 ^— == 
V ~b+8±:d 
8 \ abd 
0 = 
b+ej~ {b+eyetd(b+8)' Iri the caie 
when the error in both latitudes is politive ~ + d, and /3^b, or 
fi^b, the refulting error in the place of the node — 
— ad f h —f — j n the cafe when the error in both latitudes is 
(b+8) % +2d(b-\-&) 
negative - d , and (2~zb, or (2^b 9 then the error in the node = 
ad {~^ b± ^ . In thole two cafes the error is lefs than in 
(b + 8) — ld(b-\-8 ) 
any of the former, and quite nothing when b=/ 3 . If the 
radius of the inftrument, with which the meridian altitudes are 
obferved, is given, the quantity of d is alio given. In a 
mural quadrant of 6 or 8 feet d*=$ or 3 feconds. Take a — 
3^ 3" * 56'', jQ 135 2 l" 9 d ~ 5" > an ^ t ^ le error i 11 the fouthern 
latitude + d, in the northern = - d ; then N n — — ? -- 5 . — 
83 
2' 1 2". Take now the error only in the fouthern latitude 
— + d; then N« = - T j 8^ ; in the cafe of -d; N n- 
57 JO *° — j' 28". From hence it appears, that in comparing 
Vol. LXXVII. G two 
