270 Dr. G. Gore. Relations of Heat to Voltaic and 



series (Tables I and IV, pp. 253 and 259) were not always due to a 

 film of insoluble matter (pp. 261, 270, and 278). I have not examined 

 whether this phenomenon is related to the great thermal capacity and 

 expansibility by heat of those metals. The liquid in which the metals 

 generally were least altered in position by heating one of them was 

 the weak solution of potassic chloride, it Buffered also very few voltaic 

 reversals by rise of temperature (see Table X, p. 266). 



By making similar comparisons of the thermo series of Table I with 

 the " hot " chemico series of Table X, it was observed, that whilst 

 heating only one metal of a voltaic pair to 160 F. depressed the 

 relative positions of the magnesium group of metals a total of 354 

 places, and raised them only 9 (see p. 269) ; heating subsequently the 

 second one in addition, raised them from the latter position a total of 

 845 places, and depressed them only 5, and therefore restored them 

 nearly 98 per cent, towards their former positions. This appears to 

 be in harmony with the fact that the orders of the voltaic series at 

 60 and 160 F. are not widely different (see Table X, p. 260. By 

 similar tabular comparisons and reckonings, the electrical effect of 

 heating the first electrode was found to be usually a little greater 

 than that of heating the second one. 



The extremely negative position of magnesium in many thermo- 

 electric series, both with weak and strong solutions, does not appear to 

 be explicable either by its strongly volta- positive character, its own 

 thermo-positive character amongst metals, or by the thermo-electric 

 properties of the liquids, but requires the assumption of some other 

 cause operating. The thermo-electric positions of several other metals, 

 such as aluminium, nickel, and palladium, &c., in particular liquids, 

 are also, though in a less degree, inexplicable by these causes alone. 



Many of the metals in Table X (p. 260) were reversed in their 

 order of position in the series by rise of temperature ; the total num- 

 ber not reversed was 174, and of reversed 112, or 1*55 to I'O = 64*51 

 per cent. The reversals varied in number in different liquids ; most 

 occurred with nitric and sulphuric acids and potassic cyanide (8 in 

 each). Every metal suffered reversals ; the number was greatest 

 with aluminium (30) and lead (23), and least with zinc and nickel (4). 

 The metals which most frequently crossed each other were aluminium 

 with lead and with iron, each seven times. Reversals were less fre- 

 quently caused in chemico-electric series by a rise of temperature of 

 100 F. than in a thermo-electric series by a difference of strength of 

 liquid, the proportion in the former case being 174 out of 286, and in 

 the latter 250 out of 286 (see p. 258). 



Regarding the effect of heating two different metals simultaneous!] 

 in the chemico-electric experimenta (shown in Table X, p. 260), 

 being chiefly composed of the two effects of heating the same metal 

 separately (as shown in the thermo-electric series of Table I), I have 



