446 
Proceedings of Royal Society of Edinburgh. [sess. 
The weiglit of the normal water drop was 0*1081. 
,, „ alcohol „ 0*0307. 
„ „ benzine „ 0*0449. 
The water was then dropped in the vapours at different rates, 
and the weights corrected so as to reduce them to what they would 
have been at zero rate so that they might he comparable : — 
Growth Time. 
Water in Vapours at 37°. 
Alcohol. 
Benzine. 
1 " 
0*0983 
0*0940 
17*5" 
0*0947 
0*0816 
50" 
0*0928 
0*0613 
We see clearly the influence of allowing time for the condensation 
of the vapour in the decreasing drop with increasing growth-time ; 
but in the case of alcohol the effect is not so marked, as the alcohol 
mixes with the water, and hence the surface is continually renewed, 
whereas the benzine, being practically insoluble, accumulates entirely 
on the surface. 
The liquids were raised to a temperature within a few degrees of 
their boiling points, and the water dropped more quickly, with the 
following results 
Growth Time 0-5". 
Water in Alcohol Vapour. Water in Benzine Vapour. 
0*0326 0*0534 
It will be seen that notwithstanding the solubility of the alcohol 
in the water, its action when there is a sufficient supply is more 
potent than the benzine. 
This is owing to its solubility. The capillary height of two non- 
miscible liquids in the same tube depends not only on the specific 
gravities of the two liquids and the capillarity of the upper liquid, 
but also on the action of the meniscus separating the two liquids ; 
whereas, when the liquids are freely miscible, the height depends 
only on the mean density and the capillary action of the upper 
surface. 
Hence with the benzine experiment the drop (had the capillarity 
of benzine controlled it) ought to have weighed *0449, whereas, 
owing to the limiting surface between the benzine and water, it 
weighed *0534, an increase of 18*9 per cent, j whereas with alcohol 
the weight was ;0326 instead of *0307, an increase of 6*1 per cent., 
