108 Dr. 0. Barus on the Aqueous Fusion of Glass. 



of contact of both bodies. For a given length f thread this 

 surface decreases as the radius, r, of the tube. On the other 

 hand, the volume of water decreases as the square of the radius, 

 r, of the capillary tube. In fact, if V be the volume and S 

 the surface for a given length of thread, S/V = 2/r. Let a be 

 the rate of absorption of water in glass, i. e., the number of 

 cub. centim. of water absorbed per square centim. of surface 

 of contact, per minute. If v is the volume absorbed by S 

 cm. 2 per minute, v = a S and therefore v/V = 2 ajr. Hence, if 

 r is large, the apparent effect of absorption vanishes ; but in 

 proportion as r is smaller, or as the tube becomes more finely 

 capillary, the effect of absorption will become more obvious 

 to the eye. In other words, the length of the column of water 

 included between the two terminal threads of mercury will 

 decrease faster for small values of the capillary radius. 

 In the above results v/V taken directly from the table is 

 about *003 cub. centim. per minute. The diameter of the tube 

 measured microscopically was found to be about '045 centim. 

 Therefore a = '000034 cub. centim. is the volume of water 

 absorbed per square centim. of surface of contact, per minute, 

 at 185°. This is about 180 kg., per sq. metre, per year, at 185°. 



True the phenomenon is not quite so simple as here com- 

 puted, for as the action proceeds the water holds more body 

 in solution, the area of unchanged glass increases, and 

 possibly the liquid must diffuse or percolate through the 

 layer of opalescent accretion to reach it. As against the 

 seriousness of this consideration, one may allude to the 

 regularity of the above results in the lapse of time and the 

 occurrence of a reaction rather accelerated as time increases. 



In view of the large surfaces of reaction available even in 

 small bulks of porous or triturated rock and the fact that the 

 intensity of the reaction increases rapidly with temperature, 

 I cannot but regard this result as important. Direct experi- 

 ments * have been made with care to detect a possible thermal 

 effect (rise of temperature) of the action of water on hot glass, 

 but thus far without positive results. The difficulties of such 

 experiments are very great. To insure chemical reaction, 

 they must be made with superheated water under pressure, 

 with allowances for heat conduction &c, all of wdiich make the 

 measurement of small increments of temperature very 

 uncertain. If, however, rise of temperature may be associated 

 with the marked contraction of volume in the system water- 

 glass specified, one may note, in the first place, that for a 



* G. F. Becker : Monographs U. S. Geolog. Survey, No. III., 1882. 

 I have since made similar experiments with superheated water (200°). 



