VEGA-EXFEDITIONENS VETENSKAPLIGA ARBETEN. 



265 



to assume its original temperature, ana the contraction of vol- 

 nme caused by the melting process is determined by weighing 

 the remaining quantity of mercury in the cup. Afterwards b 

 is shnt, c is opened again, and the expansion of the liquid 

 for every centigrade degree measured by raeans of the scale- 

 tube as before. But in this case the dilatometer needs much 

 more time [6 to 8 minutes] to attain a uniform temperature 

 throughout its Avhole mäss than before, Therefore the time of 

 each observation must be considerably prolonged [20 to 25 



minutes]. 



.f 



f, 



When the vokimetric determinations are completely fin- 

 ished, the instrument is taken from the mercury bath and 

 placed in melting snoAv for a couple of hours. The superfluous 

 mercury in the reser voir is removed, so that the dilatometer 

 is filled exactly to the stopcocks a and c with mercury (and 

 water) of 0^ C. Afterwards it is cleaned and weighed at ordi- 

 nary temperature, The mercury is collected and weighed 

 separatel3^ 



The results are calculated in the following way. ^ From A, 

 B and dow^nwards the instrument is immersed in the mercur}''- 

 bath and partakes of its variations of temperature. The volume 



^ See the preceding fignre. 



