( 588 ) 



a \ / r — O 



h-h 



-11 



1 ^-K 



2 h—h,. 



'J his cx|)i"('ssi()ii (or ., l)ec()iii('s tlierelorc, i( we j)iil =(■?,: 





p 



/>fiy 



i+.M^^-iwi 



1 !>,,--/>, 



2 In—h. 



This vichls \\ilh Ihc xnliic's cak'iihikMl nltoxc (sec § 8): 



0,9772 



}'=4 X 



0.9430 



= 4 X 1,<-»^^ LI + 0,0117 X 5,747 X 0,4578] 



, 1080 \f 1 270 



1-f 0.0117 11 . 



249 A 2 249 



or 



)' = 4.140 X 1,0308 = 4.2(37 



T ,/j, 



Fiiialh' \\(' iii\('sliual(\ w hclhcr ihis xahic of ( ' I iiia\' he 



hroii^i^iil into a^rcciiicjil w itii ihc I'cw c'xpei'iinciital (hila of Dkwah. 

 Dewak found iiaiiiclv (I.e.): 



T= 20°k2r j /> = 1 Htm. 



7'/, = 30°a32° ( pi,= 1-5 atui. 

 The l\\() (hila \i('hl lt\ i>u\-iiis of iho iiilcural f(M-ninla 



for /' the Aahie 



IK'p Uuj — = / ',,- — 1 

 P V^ 



nci) l(ui 15 / 5 7 \ 



/•— „^___j — 2,708 X - - '> — I 



(I — 

 20 21 



aeeordiiiu' as \v(> take 20^ ajid '^'l or '21 and 'Mf . The lowest 

 vahie is 4,51, so still hiiiher than the ealeidatcd \ahie 4,27. We 

 must farther Jiole that 20' diHers eomparativelv verv nnieh fi-oni !)■ 

 (being "^/^ 7'/) and tiial therefore at ^O"* the factor /' \\ill cerlaiidv 

 he fonnd to he U'rcalcr than near 7/, hence 4,51 is prohalilv too 

 great. 



From the aho\e an e mav in anv case eojielnde, that the hirge 

 extrapolation, l>v 'ini^ms of \\ hieh \\ e ha\e calcnlated tiie \alne of 

 A„ at —242' from the \ahies of A at (»^ 100' and 200^ really 

 yields the critical (hita with a snflicient degree of accuracy — at 

 least in so far as we may judge from the few data, that are available. 

 ( )nly )" is |)r()l)al)ly too h»w . 



We have reason to expect a prioi'i that the new e(pialion, deri\ed 

 bv VAN Di'.u Waai,s for the \arial>iHtN of h \\\\\\ lh(^ xolmne, does 



