Water in Hot Glass, 4?3 



Internal diameter (cm.) 

 Original '0526 



Corroded 1250 



Section. 

 •0022 



•0119 



Cold thread-length 



13-5 cm. at 20°. 



External diameter '68 cm. 



The march of volume-contraction is given in the chart. 

 Boiling began at 9 h 50 m and the temperature was constant 

 5 m or l0 m later ; but contraction must have set in long before 

 this. At 10 h 15 m the thread appeared indistinct and broken, 

 marking the opaque period. At 10 h 30 m , though the water- 

 glass was still granular, the mercury column, now about two- 

 thirds the diameter of the hydrated core, became clearly 

 manifest. At 10 h 55 m it stood out brilliantly; the granulations 

 had retreated toward the axis, where they shone on illumina- 

 tion like a white cloud. In fig. 13 a group of mercury threads 

 is given specially measured as to their distance apart. They 

 moved bodily across the crosshairs of the telescope, keeping 

 their position relatively to each other. This seems to be 

 positive proof that the whole internal column is moving- 

 upward in consequence of contraction and is followed by 

 the mercury meniscus moving in the direction of constant 

 pressure. In so viscous a mass, a current up in the middle 

 and down at the sides of the bore of about f millim. is out of 

 the question. 



The deformation of the top of the column of water-glass in 

 this experiment was striking, and is given in successive stages 

 in figs. 6 to 12, showing its relation to the fixed (upper) 

 mercury meniscus. In fig. 6 the water-glass is opaque and 

 granular : in fi g. 7 the granular column is shrinking • in 

 fig. 8 granulations have vanished except in the axis. In 

 fig. 9 the clear water-glass at a shows a steam (?) cavity at b, 

 above the residual water (?) at c. This deformation may be 

 regarded as an axial opening of the core due to contraction 

 outward. Figures 10 and 11 show successive gradual changes 

 of form of b, c in fig. 9, suggesting an upward inarch of 

 pressure. Fig. 12 finally shows the gradual deformation of 

 fig. 11, as the result of cooling ; but the mercury did not fall 

 down into the channel of fig. 12, which here terminates in 

 the upper meniscus. The expansion of the bubble b is the 

 marked feature. Cooled with great caution, the tube never- 

 theless broke spontaneously ; parts of it after twelve hours, 

 other parts even after months. No liquid water was detected in 

 the cold tube, and the mercury ran freely through the channel 

 in fig. 12. The large rate of reaction is an anomalous feature 

 in view of the large diameter. Cf. curves 4 ; 5, 6 of chare. 

 The final tube No. 7, containing a solution of cobaltic 



