138 Pfoceediiif/s of ike RoyaL Society of Victoria. 



I had now to find, first, the law of variation of conduc- 

 tivitj'- with temperature, and second, its variation with 

 streno-th of solution. In working out the former, I took 

 20° C. as my standard, and in what follows, t is the excess 

 of temperature over 20°. I found that each set of observa- 

 tions was given within the limits of errors of observation 

 by the foiinula k^ = h^^ (I + at — /3 t'^) l\ and l'^^ being 

 the conductivities at 20° + f C. and 20° C. respectively. Tn 

 determine a and /3 as accurately as possible, I worked it out 

 in each case by the " method of least squares," working 

 from the conductivities at 20, 30, 40, 50, 00, 70, 80, 9('), 

 and 100° ('. found by interpolation from the results given 

 above. 



The values I found are shown in the foUowinu' table : — 



