332 Frederick Guthrie on certain 
other, B, about 2\ inches long, reckoned from the inner bend a. 
This was fastened into a mass of fusible metal foot to give sta- 
bility. The U-tube was dried perfectly under the ordinary air- 
condition, and received pure dry mercury, which stood in both 
limbs at a height of about 2§ inches (reckoned from a). The 
whole was placed in a flat-bottomed vessel g containing a little 
melted paraffin, and then upon an immovable slab, to which it 
was stuck by a few drops of paraffin. The vessel g then received 
water slightly acidulated with HC1 so as to cover the mercury 
in the shorter limb, and reach about { inch above the edge of 
the glass tube on that side. A test-tube filled with similarly 
acidulated water was inverted over the shorter limb. Upon the 
surface of the mercury in A about 15 grams of the amalgam 
Am was placed; this was covered with petroleum, and the 
tube was plugged with cotton- wool. 
Immediately after introducing the sodium amalgam a kind 
of frosted appearance is seen on the immediately lower parts 
of the mercury and glass surface in A. This appearance, which 
is a blush of bubbles, creeps downwards with strange rapidity, 
reaching the bend, say 2^ inches, in a quarter of an hour. 
In about 30 hours, bubbles of hydrogen appear at the sur- 
face of the mercury in B and collect in the pneumatic tube. 
Such evolution continues sensible for about a month. After 
two months such evolution ceased, the contents were emptied 
out, thereby being of course mixed, and no further evolution 
of hydrogen could be detected. 
Such a method of experimentation is, however, far from 
quantitative, because, when the sodium has diffused down 
through A as far as a, it will, being lighter than mercury, rise 
through B and cause whirls. 
The ideal condition of such diffusion would be of course 
similar to the ideal condition of heat- or electrical transference, 
where one may have a given potential at one end of the column, 
and a given lower one, fancifully called zero, at the other. 
Perhaps this condition is to be attained with the greatest 
practical completeness by the simple long vertical column. 
§ 28. Three glass burettes were made, about a foot in 
length and an inch in internal diameter. They w r ere drawn 
out sharply at the bottom into capillary tubes, upon which 
pressure-taps were fixed in the ordinary way. These were 
nearly filled with pure mercury. A little of the mercury 
was allowed to run through so as to fill the capillary and caout- 
chouc tube. 
Upon a tube so prepared and filled, about 15 grams of the 
amalgam Am svere placed. The amalgam was thereupon 
covered liberally with petroleum; and the top of the tube was 
