332 Messrs. Wright and Rermie on the Determination of 



quent retention therein of varying quantities of gas should be 

 rendered negligible. The projecting ends of the rods were 

 passed through small india-rubber corks and amalgamated ; 

 by then fixing short pieces of wide glass tubing over these 

 smaller corks and pouring mercury into the cups thus formed, 

 connexion by mercury-cups could readily be established be- 

 tween the battery and the voltameter-plates. 



This arrangement of mercury-cups was also applied to the 

 copper Daniell-battery cells, a thick wire bent downwards and 

 amalgamated at the end being also soldered to each zinc plate so 

 as to dip into the mercury-cup of the next cell. Much trouble 

 in brightening connexions &c. was thus saved; the current 

 could readily be broken instantaneously by simply lifting one 

 of the zinc plates an inch or so, so that the wire no longer 

 dipped into the mercury ; whilst an easy means was afforded 

 of introducing more cells or shutting some out from the cir- 

 cuit when required without actually interrupting the current. 



39. A number of attempts were made to utilize Bunsen's 

 ice-calorimeter for the measurement of the heat evolved ; the 

 construction of apparatus sufficiently large to enable conside- 

 rable amounts of heat to be measured accurately, however, 

 was found to present some difficulties ; whilst with smaller 

 apparatus the errors of measurement of the amount of gas 

 evolved became considerable, even when the voltameter was 

 connected with the gas-measuring apparatus used for Frank- 

 land and Armstrong's water-analysis process ; for the occlu- 

 sion of the evolved gases by the electrodes, the absorption of 

 oxygen by the acidulated water of the voltameter, the pro- 

 duction of ozone or of hydrogen dioxide, and the removal of 

 hydrogen whilst still nascent from the one electrode by com- 

 bination with dissolved oxygen, and of oxygen from the other 

 by combination with dissolved hydrogen, present sources of 

 diminution in the amount of gas actually evolved, which mostly 

 become relatively greater with weak currents, such as would 

 have to be employed with a small ice-calorimeter. Although 

 it did not appear to be impracticable to overcome these diffi- 

 culties, yet it was found that to do so would require a consi- 

 derable amount of time and trouble; and therefore we reverted 

 to the use of an ordinary water-calorimeter, employing cur- 

 rents of sufficient magnitude to furnish upwards of 500 cub. 

 centims. of mixed gases during the time which the experiment 

 lasted (from ten to forty-five minutes, according to the battery 

 power used). Even with the strongest currents used, repre- 

 senting about 0'65 C.G.S. current-unit, the amount of sub- 

 stances formed capable of liberating iodine from potassium 

 iodide corresponded to no more than '001 to '002 gramme of 



