ELECTRICAL CONDUCTIVITIES, ETC. 81 



It should be noted, before leaving the discussion of the temperature coefficients 

 expressed in "conductivity units," that these coefficients in general increase with 

 rise in temperature. This increase is only slight in the cases of those substances 

 which are only a little hydrated, as will be seen in Table I. Table II shows a large 

 increase in the coefficients with rise in temperature, and it will be recalled that this 

 table contains those substances that have large hydrating power. This shows that 

 the hydrates became more and more unstable the higher the temperature, there 

 being more decomposition of the hydrates between 50 and 65 than between, say, 

 20 and 35. This is what would be expected from the results already obtained in 

 this laboratory* in connection with the effect of temperature on hydrates in aqueous 

 solution. 



Certain of the temperature coefficients from 35 to 50 are not given. This is due 

 to the fact that one set of solutions was used from to 35, and an entirely different 

 set from 35 to 65. The solutions of these substances are more or less hydrolyzed, 

 and probably have an hydrolysis time factor. Since the two sets of solutions of the 

 substances in question stood for different lengths of time before using, this factor 

 would make its influence felt. 



The agreements, in general, between the two sets of results for the two sets of solu- 

 tions at 35 were very good. In those cases where the deviations were more than a 

 fraction of 1 per cent, the work, as has already been stated, was repeated. 



TEMPERATURE COEFFICIENTS OF CONDUCTIVITY IN PER CENT. 



The temperature coefficients of conductivity are also expressed in "per cent." 

 These are the temperature coefficients in conductivity units divided by the conduc- 

 tivity at the lower temperature. The relations between the coefficients expressed 

 in per cent can best be seen from the table on pages 82 and 83, which contains 

 practically all of the salts studied in this investigation. The coefficients are given 

 for two dilutions V = 8 and V = 1024, and over two ranges in temperature 25 to 35 

 and 50 to 65. This will enable us to see the effect of dilution and of temperature 

 on these coefficients. 



The most striking feature of the table is the following: Take any one column, 

 which gives the results for the different substances at the same dilution and tem- 

 perature. It will be seen that for nearly all of these different types of salts, and the 

 number is large, the temperature coefficients of conductivity in per cent is approxi- 

 mately the same; and not very widely removed from two, for V=8; and the range of 

 temperature from 25 to 35. There are some exceptions to this conclusion. 



There are two lithium salts, the nitrate and sulphate, which are, the one much less, 

 and the other much greater than two. Then there are exceptions among the com- 

 plex salts. Potassium sodium sulphate, potassium chromium sulphate, potassium 

 aluminium sulphate, and potassium ferrocyanide have values considerably less than 

 two. Ammonium acid sulphate is a marked exception, the significance of which 

 we shall try to work out in the future. Similarly, the green variety of ammonium 

 chromium sulphate has a coefficient of only 1.38. 



*Carnegie Institution of Washington Publication No. 60, 156 (1907). 



