482 



Profs. Liveing and Dewar. 



[Dec. 11, 



Preboiled and cooled 20 ozs. of distilled water, and divided into two 

 exactly eqnal portions. Took two portions of 50 grs. each of partly 

 dehydrated sodic carbonate ; heated one portion to 180° F. for fifteen 

 minutes until it ceased to evolve moisture (it lost 12*4 grs.), and 

 cooled it. Dissolved the two separate portions of salt in the separate 

 quantities of water, and tested the solutions against each other in the 

 two apparatuses. No difference of strength of current occurred at any 

 temperature up to that of 180° F, 



In another experiment of this kind, 200 grs. of ammonia alum were 

 divided and treated as above. The solution of the non-preheated por- 

 tion gave the strongest current; deflection '75° at 180° F. 



Remarks. — The conclusion I draw from these last two experiments 

 and results, combined with those of the previous ones, is that pre- 

 heating the salt either before or after solution usually weakens the 

 current, both in thermo-electropositive and in thermo-electronegative 

 liquids. 



In the several classes of cases described in this paper, it is evident 

 that the currents are not due to chemical action, but to heat acting 

 upon and altering particular molecular structures, and that the differ- 

 ence of electromotive power in different liquids is not due to differences 

 of chemical action, but probably to differences of molecular arrange- 

 ment of the solutions. Also that in the class of experiments where the 

 liquids compared had the same chemical composition, but had been 

 differently treated, the differences of molecular arrangement were less, 

 .and the currents obtained were consequently more feeble. A difference 

 of strength of current caused by repetition of experiments, or by con- 

 tinuance of heat, also renders manifest the change of molecular 

 arrangement ; and the above method may be employed for detecting 

 molecular differences in conducting liquids having the same chemical 

 composition. 



In the class of cases in which the differences of molecular arrange- 

 ment were the least, and the currents the most feeble, the direction of 

 the currents was the most uniform. This is in accordance with the 

 common truth in science that the smallest phenomena are the most 

 constant. n> 



II. " Quantitative Spectroscopic Experiments/' By Professor 

 G. D. Liveing, M.A., F.R.S., and Professor James Dewar. 

 M.A., F.R.S. Received November 27, 1879. 



The well-known expansion of the lines of sodium and other metals, 

 when increased quantities of the metals are introduced into the flame 

 or arc, seems to afford a means of attaining a quantitative spectrum 

 .analysis in certain cases, if the relations between the width of 



