( 562 ) 



between tlie tAvo (letermiiiatioiis oC llip noniial bore, this comparison 

 liiulilv im|»i'()\('> llic acciii-acv of tlic iiiaiioinclcr. 



The lliei'iiiodvnaniical ('<|iiihliriiini in tlic cxpci-inKMilal riibc was 

 ivarhed hv means of an ('l('cIi-oina,uii(Mir slin-in,U' apparatus : to 

 secure llic ('(|iiilil)riiiiii ol' pi-cssiire liclween the experiincnlal liihe 

 ami llie iiianoiiieh'r 1 wailed 5 niiniiles al'tei' each adjuslnienl helni'c 

 rcacHnu'. ()(" each isolhcnnal two series ol' obser\atif)Jis were iisuallv 

 made, one beiiinniji^' witii the lowest and eji(Un^^ with the in<ihest pres- 

 sure, the other in reversed order. Oniv rarelv the results of these two 



1 



diir(Med as much as — . To indue ol this see tabl(> XIX, which relates 

 500 ■ ' 



lo the isolhei'mal for '22".()'^ ('. of a mixlinc (tf ().l(>47 owucn in 



carbon dioxide. Here /■ I'epi'eseiils ilic xoliimc. ƒ), and />/■ the c(»rrc- 



spondinti" pressures in llic two series, willi a coiistantlv risiiiu- and 



a constanllv fallin.ü' pressure respecli\ civ. Lp llie diirerence, z-//,^^. and 



I'U.if the volumes of the li(pnd also in the two series, Z-/',/,^. the 



dilference between them. 



The af2;reenient of the pressures is salisfactor\ , that of the \ (ilumes 

 of the liquid leax'es to be desii-ed. Althoiiuh thi-> is partly explained 

 by the circumstance that the xolume of the li(pnd cannot be read 

 so accurately because the form of the meniscus is not so sharply 

 determined as is the case with mercury, and parallax coidd not 

 so easily be avoided as with the mercury menisci, it yet ap[)eai"s 

 that, if we desire to iii\ estimate this subject more fidly. more care shoidd 

 be taken to procui-e e(juilibriuni by stirring aJid waiting. 



§ 5. I'lu' cünsf((ni-ij (ind tin' nn'tisiircnu'iit of ihc ti'iiiiicintii res. The 

 manometer was surrounded by a Jacivct with llowing watei' kept at 

 constant tem|)ei'ature by a thernatstal. as described by S( iiai.kw i.ik 

 (Comm. N". 70, Proc. May '01) with the modilication in the thermo- 

 regulator described in Conini. X". 78 (Proc. April 1902 |). 7H2). 



In the same way the experimental tul)e was kept at constant 

 temperatui-e by means of a second theiinostat. This dilfered from 

 the former in the following respects: 



The connection l>etween the heating bath and the mixing bath, 

 the mixing bath and the connection between the mixhig bath and 

 the observatioji bath ha\e been insulated more carefully by means 

 of wool, paj)er and felt; this was necessary because greater differences 

 of temperature with the sini-ounding atmos[)here occur here. 



To the glass portion of the thernun-egnlator a side-tube with a 

 cock is sealed ou, which facilitates the adnussiou of a «piaulily of 

 mercio-y from a mercury reserxoir. which by means of an india 

 rubbei' tube is comiected with that side-tid)e, into the tliermorenidalor 



