•60 Scientific Proceedings, Royal Dublin Society. 



that plaster treated in a similar manner, except that it had 

 not been exposed to boiling, also shows itself unsatisfied with 

 water, although an excess of water had been present while it 

 was setting. 



From the measurements given in Experiment ii., we may- 

 obtain an approximate estimate as to how far this absorption of 

 water by the plaster will affect Copeland's experiment. In my 

 experiment 62 c.c. of plaster of Paris absorbed 1*045 c.c. of water 

 in twenty-six days. Each of the component tubes of Copeland's 

 apparatus was 0*3 cm. in diameter and 40 cm. long. Consequently 

 each contained 2*826 c.c. of plaster; and if they behaved similarly, 

 would absorb 0*047 c.c. At the start of the experiment, the rate 

 of absorption is more rapid, so that the amount absorbed per diem 

 will be greater than ^ff^c.c., or 0*0018 c.c. 



We have already seen that the volume of water, driven 

 through by the difference of pressure observed, will not amount 

 to 'more than 0*0027 c.c. per diem ; while we now see that the 

 absorption for only 60 cm. of the tube used in the apparatus will 

 exceed this amount. And it is to be remembered that 840 cm. 

 intervened between the upper and lower manometers in Copeland's 

 experiments, rendering the equalization of pressure-conditions, by 

 the passage of water, impossible. 



Hence it follows that any flow of water due to pressure such as 

 Copeland observed, would most certainly be completely masked 

 by the absorption taking place in the plaster of Paris in the tube. 



We may now summarize our discussion of Copeland's experi- 

 ment as follows : — 



The rise of mercury in the manometers was due for the most 

 part to the reduction in volume of the water in the main tube 

 by the absorption of the plaster of Paris. External atmospheric 

 pressure lifted the mercury as space was given by this reduction 

 in volume. 



The readings of the manometers do not give any indication of 

 the pressure -conditions of the water in the plaster, for the mano- 

 meters were not so arranged as to be capable of indicating tension 

 {i.e. negative pressure) in the liquid, even if such existed. 



The differences of level in the manometers were due to 



