206 Prof. K. Olszewski on the Critical and 



dimensions, sometimes of ebonite, sometimes of mica. In the 

 first experiments I used an ebonite frame whose height was 

 20 millim. and diameter 12 millim. ; the distance between 

 the single turns of the wire was 0*5 millim., the resistance of 

 the whole wire was 241*4 ohms at 0°. In a further experi- 

 ment I used an ebonite frame whose height was 11 millim. 

 and diameter 10 millim. ; the resistance of the wire, which 

 was wound also at intervals of 0*5 millim., was 117*5 ohms 

 at 0°. In the final experiments I used a frame made of thin 

 mica sheets : its height was 20 millim., the diameter 11 

 millim. ; the resistance of the wire, wound in intervals of 

 1 millim., was 104*9 ohms at 0°. I devoted the greatest care 

 to the construction of these frames ; for the precise determi- 

 nation of the temperatures depended much upon their careful 

 construction. I endeavoured therefore to make them of bad 

 conductors of heat and electricity, and very light and with 

 very thin wings at the edges where the wire touched them, 

 so that the surfaces of contact were very small. The con- 

 struction of the ebonite frames is easier, but those of mica are 

 considerably less in mass ; so that the larger ebonite frame 

 weighed 0"679 gr. with the wire, the smaller ebonite with 

 the wire 0*218 gr., whilst the mica one weighed with the 

 wire 0*177 gr. ; the wire of the latter weighed 0*011 gr., 

 and was about 595 millim. long. The glass tube /with very 

 thin walls which surrounds the frame b served to insulate it 

 during the expansion from the warmer walls of the steel 

 vessel a. The upper part of this vessel is screwed into it, and 

 a close adhesion is produced by soldering it by means of an 

 easily fusible bismuth alloy. The tube g serves to admit 

 hydrogen, which is contained in a 3-litre iron flask under 

 a pressure of about 170 atm. The lower part of the appa- 

 ratus described is plunged in the liquid oxygen, whose 

 temperature is lowered to about —210° by pumping ; the 

 indiarubber stopper h serves to close the whole apparatus 

 whilst the pumping is going on. The wires k k' connect the 

 apparatus with a Wheatstone bridge, for the purpose of 

 measuring the resistance of the platinum wire wound on the 

 frame b. To measure the resistance I used a very feeble 

 current from one Leclanche element. This current was 

 weakened by passing it through a resistance of 1000 ohms, 

 to avoid heating the platinum thermometer ; for the heating 

 was distinctly to be perceived if this resistance was not 

 brought into play. 



The course of the experiment was as follows : — In order to 

 draw the curve representing the dependence of the resistance 

 of the thermometer on the variation of the temperature 



