VEGA-EXrEDITIONKNS VETENSKAPLIGA AUDETEK. 257 



CHAPTER 2. 



Methods and instruments. 



In the course of the experimental researcli some delicate 

 questions, requiring great accuracy in the determination of 

 temperature, arose. I therefore think it advisable, first to give 

 an exposition of the 



A. Measurements of temperatures and determinations of 

 the coefficient of apparent dilatation of mercury. 



The temperature was observed b.y means of o normal 

 inercnry therraometers, manufactured expressly for these piir- 

 poses bj^ Dr. Geissler in Bonn and his successor, Mr. Franz 

 Muller. In the following these instruments will be denoted 

 hy Gi, Gn & Gi„ 



G^ indicates temperatures from — 1° C to + 26° C 

 G„ » » » — 12° C » + 5° C 



G„i » » » — 22° C » — 7° C 



Gj & G„ were made in 1875, Gm in 1880, consequently the 

 correction for zero at present ought to be constant. It is 

 however liable to httle perturbations. Especially in Gj the 

 final state of equilibrium of the molecules seems not to have 

 been attained yet. 



By a great number of observations it was ascertained, 

 that the indications of the different thermometers above zero 

 corresponded with eaeh other within the limits of 0°.oi or 

 0°.oi5 of a centigrade degree. Below zero the correspondence 

 €ould not be observed so closely, on account of the difficulty 

 of maintaining the temperature steady for a long tinie. It 

 must also be observed, that during the volumetric experiments 

 the dilatometer and the thermometer were immersed in a mer- 

 cur}^ vessel of wrouglit iron. The pressure of the mercury 

 caused a rise of 0^o2 C in the indication of temperature. Tliis 

 is a constant error, to which due reference has been made. 



The correspondence of the different instruments being 

 ascertained to be satisfactory, another question arose of equal 



17 



