DR. W. M. HICKS : A CKITICAL STUDY OF SPECTRAL SERIES. 353 



In the case of Si it appears as if the e = 7315 is much superior. But, as it happens, the value of S 3 (3) 

 as calculated from S 2 (3) by v 2 is 42151-20, and so is very close to Si (1). It will be shown that for this 

 the 7314 link gives very close values, and the linkage probably belongs to S 3 (3). The -e-u is 

 doubtful. 



8(2). 



25936 27714 28530 



-e(3) 18622 -00, --19 -t>(2) 23282 -51 v (2) 32958 -00, -17 



-e + M(10) 22755 -08, -16 -^(-281) -00 

 -2e [15441 -84] 



In Si the intensity of 22755 would suggest that its linkage is a coincidence. L.D. give a line 15447 

 which may possibly be ( - 2e) Si, for their measures are only to the nearest A.U. 



8(3). 



39559 41336 [42151] 



SeeS 3 (l) -e(<l) 34022 -'03, --10 -e(<l) 348361 '02, --05 



- + (<!) 38155 -'03, --06 -2e (4) 27523 -00, -17 



-e-w(l) 29892 -'20, --31 -e-w(3) 30700 -22, '14 



-e + v(<l) 38453 --16, --22 -t'-u(-2S l ) -01 



30843 -'14, --16 



t Si(l) and S s (3) are very close. This sounder probably is correct here and does not hold for Si(l), 

 for which it differs by about 3. 



8(4). 



-e-v(<l) 34518 -'08 



The 38940, 39763 are too far out to be dependent on e links. They could be dependent respectively on 

 e ( - 2S t ) and e ( + 2Si). Or more probably the lines 38940, 39754 may be S 2 (4) ( - 28).e, S 3 (4) ( - 28).e. 



8(5). 



[46799-33] [48577-23] [49392-28] 



-e(ln) 39491 -28 -e(l) 41257 -'23 



-e-w(l) 35350 -08 * - e-v(<l) 36832 -11 



-2e(l) 33948 -01 * 



39491 is too far out for a link. It would give a reading for Si = 46805-45. It is curious, however, 

 to notice that we have sounders for a set with this value, viz. : 



46805-45 1779-80 48585-01 815 '52 49400-53 



-e(ln) 39491 -e(3) 41271-15 -e-w(l) 37593-25 



8(6). 



[48072-77] [49850-67] [56665-97] 



-e (2) 40758 -02 * -e [42536] -e [43352] 



-e-v(ln) 38108 -00 * -e-w(l) 39218 . '00 * 



3 C 2 



