116 Mr. T. Tate’s Experimental Researches on the 
that the filtering power of the paper had sensibly decreased ; but, 
owing to the rapidity of the process, the effect arising from this 
cause had not materially interfered with the normal law of 
filtration. 
Upon repeating the process, it was found that the velocity of 
ascent followed, for the most part, the law expressed by the 
general equation (7). 
Hence it would appear that the normal law is, that the rate 
of filtration varies directly as the pressure upon the filter, the 
deviations from this law bemg due to the change which takes 
place in the molecules of the filter during the process. 
The following experiment shows that by alternating the direc- 
tion of the current of filtration, the original power of the filter 
may be maintained nearly unimpaired. 
Experiment XXV. 
The filter in this experiment was coke, presenting an interior 
surface of half an inch in contact with the water, and an exterior 
surface of about two inches, other things being the same as in 
the last experiment. | 
By upward filtration, the law of ascent of the liquid was 
accurately expressed by the formula © 
h 
144’ 
where v is the velocity per second, 2 being the column of liquid 
pressure on the filter expressed in units of the graduation of 
the tube. 
The water thus filtered into the tube was allowed to be dis- 
charged by downward filtration. In this case the liquid followed 
nearly the same law of descent ; but upon repeating this process 
of downward filtration, the velocity of descent was found to fol- 
low the law expressed by equation (7). Upon reversing the 
direction of the current of filtration, the law of ascent was found 
C= 
to be accurately expressed by the formula = And so on 
to other alternations. 
Similar results were obtained with wood-charcoal filters. 
The following experiment was made to determine the rate of 
change which takes place in the filter during the process of 
filtration under a constant pressure. 
Experiment XXVI. 
This experiment was made with a close coke filter about an 
inch and a quarter in depth, and presenting three-fourths of an 
inch of surface in contact with the water, which had been care- 
