120 Dr. C. Chree on the Law 
_ When neither P nor Q is negligible, and they have to be 
eliminated by observations at 22°5, 30, and 40 cms., the value 
of H requires to be increased by 2°94 x 2°79 or 8:2. 
_§ 6. Other likely distances are dealt with in Table IL., 
which is to be interpreted on similar lines to Table I. All 
distances are in centimetres. | 
TABLE II. 
Distances one, two, or three of the following :— 
"por {2 ; | . it 
given Single distance. Two distances. _ Three distances. 
een 25°25 | 35 | 45 2625 & 85] 26-25 & 45 35 & 45° 26-25, 35, & 45 
2625. 1 |...|..| tae 40059 |. ° | eee 
2 alll wok aed eee ae 1:68 || —3:50 
AG | Se) De +152 | 42:53 +3-84 
| 95 |95|45| 25¢a5 | 25645 | 35 & 45 | 95, 85, 46 
1 t |... |u|) —t0k | 04 |. | ol 
te 1 jas glue —153 | —3:13 
Sid Ly | | 4145 | +4253 | 4366 
Sie a Van 85 | 40 30 & 35 30640 |35&40!  30,35,40 
A Mie As “BIT | =120 |. 
35. Laid ahead -32 |  =19a0e 
40 1 | 4299 (4497 | +4975 | 
The results in Tables I. and IL. depend in reality only on 
the ratios of the distances, not on their absolute values. 
The last combination dealt with in Table II. exemplifies 
what happens when the distances 7;~7, and 7.~73 are com- 
paratively small. It clearly cannot be recommended when 
observations are to be taken at three distances. Unless Q 
were exceptionally large, it is clear that a trifling error in 
the 35 cms. distance might more than neutralize the advan- 
tages to be derived from not neglecting Q. 
§ 7. When observations are made at three distances, and 
more than one of them is in error, the following Table will 
be found convenient. In it 6H, 674, 579, ér; are all corrections, 
or else all errors. 
