( 534 ) 



calonlations tlie data about tlic manometer will be pven. The gia- 

 (hialed liibc lias bc(Mi calibi-alcd Iwice. in U)()l (.1) and in liM)2 (/)'). 

 I sliall {^ivc lioi'c the data and llie calcnlalions concerning' cali- 

 bration /). ('alibralion .1 lias been made iji eniirclv the sanu^ wav : 

 the resnlts ot'il will be ,uiven at the end of lliis |)aper ajid compar(Ml with 

 those of B. 'i'altle 1 shows the r(>,snlts of the calibrations with tli(> 

 iTierciiry cfdnnin ; (obimn .)/ contains the means, L the lenulli of 

 the mei'curv thr(>ad, Z_ th(^ dili'er(Mice IVoni the mean ienutli. The 

 temperature coidd be considered as iia\iii,Li' i'emaiiie(l constant. 



TABLE II. 



ImkIs iif tlic int!rcurv column: 

 1st [.(.sition: 3.79'' iO.OS-" t.Mn|i. ^2(17 



2(1 )' ;iO.I{5 in the fiTuJuatrd stem,... '.I.'ir) iiiiii. alidvc 



tilt" division of tln' 

 tliin (■a|iiil;irv C;, ; » !20.t)5 

 ;{(! » 2.80 49.00 » 20.0 



Wci^rJit of tlic incrcury (in vucuo) : H.3777 pr. 



(Calculation : 



Let .s' be the boi'e of the tube at au ai-i)itrai'v place, indicated bv 

 (he coordinate ,/■ (from O to 50). We may j>ut . 



where .s^, is a particular, normal, bore. 



The volume between the 2 divisions j> and q will be : 



'/ ,'' -J' 



I =z Cs d,r = I (,s„ + ./) d.c == .s„ (v - />) 4- ƒ '/ ''-r^ ■ ■ (1) 



The length of the mercnrv column : 



q — /' = '" H~ ^ 



if ;// = the mean length (comp. table I). If f", represents the xolume 

 of the mercury column then ; 



7 



Vt =z ,„ m -f .s, Z- + ƒ </ . dx. 



/' 

 We may choose .v„ so tliat : 



then : 



L — \ <l . 'I.r = — I '/' . (/.'• 



I 



-iJ" •"■'■=-ƒ' 



p p 



d • /f ■ I 1 ■ 



if </' = . If we knew the foi-m of the fnnction ^/ , \ve nnulit den\(^ 



