from an Increase of Temperature. 369 



length of the rod. The inner case being closed to prevent the 

 evaporation of the moisture, the ultimate expansion of the green- 

 stone for 155° F., was 5.02 revolutions of the screw of the upper 

 micrometer. The pipes V V, for drying the inner case were then 

 opened, a good deal of vapour came off, and when the rod 

 had been allowed to dry for three quarters of an hour, the mi- 

 crometer read only 3.24; hence, by the effect of drying alone, 

 the rod contracted rather more than yy¥ tn P art °f an inch. I 

 then poured boiling water on the rod, and closed the drying 

 pipes, and although this could not moisten it thoroughly, yet 

 when its expansion was again at the greatest, the upper micro- 

 meter read 4.15. The drying pipes were then opened a second 

 time, and although the temperature always remained unaltered, 

 the reading of the upper micrometer was again reduced to 3.32. 

 This experiment clearly shews, that the expansibility of some 

 kinds of stone very much depends on their state of humidity. 

 An analogous experiment was tried with a rod of satin wood, 

 and the length of the wood altered with the similar changes in 

 the same manner, but to a greater extent. With the latter sub- 

 stance this might have been expected, but such an action of 

 moisture on stone is certainly very remarkable. I have there- 

 fore given two rates of expansion in the table for several of the 

 rods, the greater is when the rod contained most moisture, the 

 lesser is a mean of several experiments made when it was in a 

 dryer state, and when the results of the experiments agreed very 

 nearly among themselves. 



A very curious effect of the heat on the white marbles, was 

 that of causing a permanent increase of their length, which, in 

 the Sicilian marble, was gradually augmented by the heating for 

 each experiment. The rod of this marble was heated five times, 

 but, unfortunately, as the change is always decreasing, I at first 

 adjusted the micrometers without noting the quantity, by which 

 the line on the stud had not returned to the adjustment for the 

 previous experiment. This was the first time I had met with 



