Molecular Constants. 333 
slightly corked. Instantly clouds of minute bubbles began 
to make their appearance between the mercury and the glass. 
In half an hour the whole column appeared frosted (see § 27). 
On drawing off a measure, say ^ of the whole, from the bottom 
after two or three hours, no appreciable amount of hydrogen 
was to be got from it. 
Accordingly the tube was reemptied, cleaned, dried, and 
refilled. The amalgam (A m) was then allowed to rest upon 
the top for 14 days and nights in an undisturbed and steady 
place, where the temperature ranged from 13° to 18° C. At 
the end of this time the amalgam was drawn off. The draw- 
ing off was effected as follows: — A little block of paraffin 
was hollowed so as to have a smooth cavity of the capacity of 
about j'g- of the tube in fig. 6. The edge was ground fiat, and 
a flat slab of paraffin served as a cover. The amalgam was 
drawn into this very slowly so as to stand above the edge; the 
slab being then pressed down, a unit volume was entrapped. 
This being transferred to a porcelain capsule, the few drops of 
overflow were returned to the unit measure, which was again 
filled up, and so on. The six lowest measures (each about ^) 
did not show a trace of hydrogen. The seven higher ones 
evolved hydrogen in the quantities shown in the following 
table, in which the actual weights of the mercury are re- 
duced to 100, the cub. centim. of hydrogen being recalculated 
and reduced to dry hydrogen at 0° C. and 760 millim. It 
appears that in 14 days the sodium had penetrated down a 
little more than halfway, say 7 inches, in quantity appreciable. 
I put now these results in such a form that they may be as far 
as possible immediately comparable with the results obtained 
by other metals. They come out as follows: — 
Per cent. Na. Hg. Na. 
•0035 100 and -0035 
•0178 , „ -0178 
•0665. 
•1769. 
•2034. 
•2295. 
•2414. 
•0666 
•1772 
•2038 
•230 
•242 
§ 29. A potassium amalgam prepared in a similar manner 
was found, when analyzed as in § 27, to have the composition 
l - 34 per cent, of K. About the same quantity of this was 
put into the same tube as had been used for the !Na, under, 
as far as possible, the same conditions. 
Reducing the evolved H to 0° and 760, as before, it was 
found that the 13 volumes of the column (all of which were 
