Chase.] 432 [Oct. 6, 



Aggregate, S., .00000815; S, .735; H 718,725,600,000; G 49,821,227,000,- 

 000 ; C 614,813,000,000,000,000,000.000,000,000,000. 



391, AnotJier Comparative Basis. 



In the above comparisons it seemed best to exclude tbe elements that 

 were exactly phyllotactic multiples of the assumed divisors (H in the 3d 

 and 5th columns ; N and C in the 5th). If those elements were consid- 

 ered as uncertain to the amount of .001 H, the results would be modified, 

 b}^ introducing all the elements, as follows : residuals ; H, monatomic, 

 .05859, perissad, .07144, mean, .11154 ; Chase, monatomic, .05859, tri- and 

 pentatomic, .03891, di- and tetratomic, .03538, perissad, .04916, artiad, 

 ,05393, mean, .05168. The comparative mean probabilities would be as 

 follows: S2.5O5O, Si .6035, II 1.0000, Gerber .9936, Chase 3.1581. The 

 mean probability of the hydrogen unit, as deduced from the mean acci- 

 dental residual, would be, 1.9386 ; of the phyllotactic divisors, 3.5589. 

 That the test of the mean accidental residual is sufficiently severe is evi- 

 dent from the probabilities which it indicates for the surd divisors, Sj 

 and S^. 



292. Objection Answered. 



The uncertainity, even of (Clarke's recomputation of atomic weights, has 

 been urged as an objection to the acceptance of any apparent probabilities 

 which may be inferred from their examination. If our conclusions were 

 absolute, the objection would be valid, and it must be admitted that even the 

 comparative probabilities will doubtless be greatly modified by the more ac- 

 curate determination of doubtful atomicities. The modifications, however, 

 would be quite as likely to increase the evidences of phyllotactic influence 

 as to diminish them, if there were no reason to look for such influence. The a 

 priori grounds for expecting proof of harmonic action in some shape or 



