476 E. J. LUND 



hollowed out to meet the needs in various experiments. The 

 sea-water was adjusted by means of the stopcock, s, to flow at a 

 constant and equal rate through the two troughs, m and n. The 

 pieces of internode, i, were fixed in position on a cork plate, 

 k, under glass rods, h. The troughs were then covered with a 

 glass plate to make the area of cross-section of trough constant. 

 A cross-section of the trough is shown in B, The pieces of inter- 

 node, i, were held in position by notches cut in the cork, k, and a 

 piece of capillary glass rod, h. C shows the arrangement for 

 holding the internode in a definite position. 



In order to speak of the effects of different intensities of the 

 electric current, its density should be known. The unit of 

 density, 8, is usually taken to be 1/1000 of a milliampere per 

 square millimeter of cross-section of the conductor. The current 

 density in the trough would therefore be determined by measur- 

 ing the area of the cross-section of the column of sea-water, m, in 

 B, neglecting the conductance of the internodes. The conduc- 

 tance of the cork plate may be practically reduced to zero by 

 boiling in soft paraffin. Since the area of the cross-section of m 

 is a constant, the current density at that section will be constant 

 throughout an experiment, other conditions remaining the same. 

 The electrical energy flowing through the internodes depends, 

 of course, upon the relative specific conductance of the internode 

 and the sea-water, and therefore may conceivably vary from time 

 to time with the progress of regenerative changes in the inter- 

 node, all other conditions being the same. As a matter of fact, 

 under the conditions of the experiments and for the purposes 

 of the more important conclusions in the present paper, it is 

 not necessary to know the absolute current density. 



For the purpose of comparing the effects of different current 

 densities upon polarity and regeneration, a gradient trough 

 was constructed as shown in D. The only difference between 

 this one and that shown in A is that the width of the column 

 of water under the glass plate, /, was made greater toward one 

 end. Each of the cork strips holding the rows of internodes 

 were placed in such a position as to give densities of the electric 

 current in the proportion 1: ^: f: |. The arrangement in A 



