919 
TAB LE, II 
W Wi | f Ww d Corres- 
Ji | W—W, WW | ponding 
observed | computed | computed | AT 
| 2 2 | 042 2 | 1042 0-001 
14.70 1.453 1.45406 | — 10.6 1.45385 | — 8.5 + 16 
16.04 1.531 1.53201 | — 10.1 1.53184 | — 8.4 | + 13 
17.01 1.601 | 1.59983 | + 11.7 1.59969 | + 13.1 | 17 
18.04 | 1.685 | 1.68326 | Ses) fg Ast. 68317 | + 18.3| — 21 
19.45 1.819 1.81786 | + 11.4 1.81784 | + 11.6) — 11 
20.43 | 1.924 1.92624 | — 22.4 1.92626 | — 22.6| + 19 
56.93 | 15.119 15.1255 | — 65 15.1260 |— 70 + 13 
60.57 | 17.097 | 17.1026 |— 56 17.1027 |— 57 + 10 
68.38 | 21.491 | 21.4859 | + 51 21.4847 |+ 63 = 12 
86.36 | 31.904 | 31.9010 | + 30 31.8966 + 74 — 13 
90.27 | 34.180 34.1808 | — 8 34.1759 | + 41 — 7 
142.66 | 64.189 | 64.2008 lS 64.1969 — 79 + 14 
159.53 | 73.629 | 73.6366 |— 76 1546361 eetl js 
170.39 79.674 | 79.6655 | + 8 79.6676 |+ 64 | — 12 
193.52 92.422 | 92.4111 | + 109 | 92.4196 | + 24 | — 4 
211.59 102.280 | 102.2987 | — 187 102305 EI 315: | reas 
230.00 112.278 | 112.3206 | — 426 | 112.3407 | — 627 + 115 | 
273.09 135.450 | 135.6372 | — 1872 | 135.6754 | — 2254 + 418 | 
77.91 26.988 | 26.9797 |+ 83 | 26.9771 | + 109 — 19 | 
| | 
5. Computation with least squares. In order to get my final 
solution 1 yet applied corrections determined by the method of least 
squares to the values found for w, a, 6, c, and d, though it was 
hardly to be expected that the experimental values would thus be re- 
presented much closer. Thence the rather lengthy computation was 
principally made for another reason, viz. for the determination of 
the weights of the unknowns. Indeed the preceding calculations 
showed clearly that the constants 4 and e could not be found very 
accurately from the data, a small change of the constant terms of 
the equations causing a great deviation in the unknowns. 
