492 Mr. Ri Sabine on the Motions produced 



on mercury were made 



before and after 



deoxidati 



means : — 







d. 



d 1 . 



d 



1*85 centim. 



1*7 centim. 



1-09 



2*0 centime. 



1-75 „ 



1-14 



1*95 centim. 



1-8 „ 



1-08 



2*4 centims. 



2-19 centims. 



1-10 



2-73 „ 



2-52 „ 



1-08 



3*83 „ 



3-55 „ 



1-08 





Mean . 



. 1-095 



Allowing for the difficulty of measuring liquid drops*, 

 which are not always quite circular, and for a cause of error 

 arising from circulation in the mercury, these values are con- 

 firmatory of the view I have stated. With greater care and 

 better means of measuring the area of the drop, and of regula- 

 ting the tension of the mercury surface outside the drop, these 

 values would, I have no doubt, be found to correspond strictly 

 with the changed atomic volume of the molecules at the base, and 

 might even be employed as a means of deducing the specific 

 gravity of the suboxide, which is at present rather difficult on 

 account of its instability. 



When a battery of higher power is used for the deoxidation 

 of the mercury, the drop of water or acid water is observed at 

 first to contract and soon afterwards slightly to expand again. 

 On closely observing the water, its interior is seen in rapid 

 circulation ; and fixing the attention upon any little floating 

 particles, it is seen that close to the mercury they travel away 

 from the immersed positive wire for a short distance, then rise 

 up towards the surface, and return in the other direction. 

 The distance between a travelling particle and its reflected 

 image in the mercury allows the direction of circulation to be 

 very exactly observed. This circulation, which is communi- 

 cated to the water-drop by a similar circulation in the mer- 

 cury underneath, causes the drop to become slightly expanded. 

 Therefore whenever the battery is strong enough to establish 

 any circulation, the diameter of the drop cannot be found to 

 correspond with its deoxidized state. This source of error is 

 difficult altogether to avoid. 



In order to be able better to observe the effect of this circu- 

 lation, which appeared to be due to unequal chemical action 

 on the surface of the mercury, the following experiment was 

 arranged : — 



A glass trough was constructed with a groove about 0*6 



* Photography suggested itself for this purpose. 



