( 564 ) 



been gmdiiated in mm. ovci- a length of 4 cm., wliicli facilitates a 

 finer adjustment liv moxing n|) and down the supply tube. 



For temperatures iiiuher than '^rt ihe water sireamed succes- 

 sivelv tln-ouuh two liealing baths, each pro\ ided ^^ith a tliermo- 

 reg-nlator ; the lirst brought the water to 35° ('., the second to the 

 desired temijerature. To avoid exchange of heat with the surrounding 

 atmosj)here the lasi hath ^^ as coated with asi^estos, except where 

 it came into contact ^^ ith the flames. 



One thermore.uulatoj- had ijeen constructed for higher temperatures, 

 filled Willi xylene and g(ies as far as -j- 00" C, the other for lower 

 temperatures was filled Avith benzole, of which the coetlicient of 

 expansion was determined al O.OOTl and goes from — 4()to-}-40°C 



For tenii)eratures below tliat <»f the wafer supply, tlie Abater 

 streamed between the o\ei-flo\N' and the heating l)atli through a zinc 

 vessel insidated with felt, where small piece? of ice were continually 

 brought in. 



Hy leadiuu llic water back to the oxerflow affer it has llown 

 fhrouüli I he observation jacket by means of a membrane pump ^) 

 fhrouuli lubes enclosetl in wool, we ccudd make measurements to 

 about 1 .5) {'. Viw ihc expci-iiueuls wiih the mixttii-e 0.3. mentioned 

 in § •>■ ^^ ^' used tlic ciin-ciil of >olutiou of calciumchloride at constant 

 temperalure, docribcd in ( omm. N". 87. Zittingsversl. Deel XII 

 Juni 1903 (see PI. 11 of ihai Comm.. Avhich shows the whole arran- 

 gement wilh llie lliermo>l;tt and the obser\;ilion hath). 



For lem[»eratures abo\ e 3(1 , and al>o foi' teniperatuj-es as much 

 again below the iem])ei'alui'(.' of the loom, it appeared necessary 

 to replace the observatioii Ijalii. which first consisted of a simple 

 glass tube, by a \acuum jacket ' . After thi> the difference in tem- 

 perature at 34°('. over a height (tf 33 cm. without stirring amounted 

 to O^.O'i ('. Moreoxei" a\ (^ ha\e always stirred before each observa- 

 tion. The water w a-> insulated from the co])JK'1- j>icce with Avhicli 

 the jacket \\ a> faslcucd to die >feel tlanued tube of die experimental 

 tube by a layci' of sulj)liur. Tiii^ |)revented the cenient from softening. 



Ill oi'der lo prexeiil \ai'ialioii> of temjterafure of the gas compressed 

 in the exjiei'imenfal tube dwiuLi' lo conductioii of heat along the 

 mercury coluuin (o dw lucrcury in ihe coni|tres>ion tubes, we look 

 care thai in the ob^er\ati(Ui>, where the mercury meniscus was lowest 

 the mercury coliiiuji >lill >UM)d o\ei' a length of 40 cms. in the jacket. 



1) Oomp. DK Haas, Tliesis for the ducloi'ate fig. f. 



2) Comp. Comm, N". 85, Vers). Deel XI! .luni f903, p. 21 f . Tfie case acting as 

 a spring was placed as low as possible to leave room for the coil wiiich moves 

 the eleclio-inasuetic stii-rer. 



