468 Dr. Carl Barns on the Absorption of 



water on glass to an actual period of subsidence, should this 

 he possible. The aqueous silicate obtained was a clear hard 

 core, plainly demarcated from the igneous glass envelope by 

 difference of refraction, but otherwise not readily distin- 

 guishable from it. The bubbles contained no water. They 

 were probably partial vacua with traces of steam. A fine 

 axial canal, less than ^ Q millim. in diameter, ran apparently 

 through the water-glass core, threading the bubbles. The 

 column of water shortened at 210° to its original length in 

 the cold tube, after about 20 minutes of exposure. The 

 transparent stage was reached in this tube earlier than usual, 

 even at 210°. Indeed, the solvent action was energetic 

 throughout, as is evidenced by the rapid volume-contraction. 

 The chart gives the successive values of volume and compres- 

 sibility (/3) in the lapse of time. The quantity /3 is an 

 estimate, since observations in triplets could not be obtained, 

 increase of pressure corresponding to an increment in the 

 continued volume-contraction. 



Boiling began at 2 h 40 m , and constant temperature"^ was 

 reached at 2 h 55 m . Pronounced volume-contraction had set 

 in before this. The lower meniscus was lost as such before 

 2 h 60 m . On increasing pressure the column ascended in form 

 of a succession of ovoid drops with constrictions between 

 them, the mercury diameters being smaller than the bore. 

 Removal of pressure frequently separated these drops alto- 

 gether, the viscous glass flowing in between them. Indeed 

 the mercury often retreated in such a way as to leave the cavity 

 half empty. In figure 2 (p. 464) I have drawn an interesting- 

 case observed at 3 h 3 m . The mercury fills only the top half 

 of the cavity left on reducing pressure. Very soon after the 

 metal was seen falling to the bottom in a rain of very small 

 drops. Since the diameter of the cavity did not exceed j 

 millim., the extreme fineness of these drops is remarkable even 

 at 210°. On slightly increasing pressure again everything 

 was seen to move upwards in a regular current, meniscus 

 (below), detached mercury drops and threads, and the cavities. 

 At 3 h 20 m , the mercury still showed increased advance on 

 raising the pressure to about 100 atrn. ; but on withdrawing 

 pressure the mercury did not retreat. I measured the large 

 compressibility ft = '000440 for these conditions. Even at 

 3 h 30 m cavities half filled with mercury were much in evidence. 

 At 3 h 50 m the column of aqueous silicate was nearly clear and 

 continuous, and the compressibility now measured was found 

 enormously reduced, perhaps less than the normal compres- 



* See Chart (p. 470), curve No. 4, for time and length increments, &c. 



