In test No. 5 a rectangular box, in which were placed two vertical sheet iron plates, was filled with tap water. The distance between the plates was 5/8 inch, and with a difference of potential of 414 at start and 397 at end of the run, a current of 35 amperes was kept flowing for 35 minutes. Cold tap water was kept running in between the electrodes at the rate of 6.11 pounds per minute (about 1/10 cubic foot) by means of a small rubber tube about 1/4 inch inside diameter. This test is very interesting in comparison with the preceding. The current carrying capacity, 0.3 ampere per square inch, was more than double, and the energy absorbed 183 watts per cubic inch, more than six times as great as in case where running water was not used. The temperature in some places between the plates occasionally rose as high as 205 deg. F., and it was necessary, in order to avoid too violent ebullition, to keep the inflowing stream of water directed along the water surface between the two plates. Less water would not have been sufficient, and, of course, by using more water, the temperature could have been kept lower, or with the same temperature the watts absorbed could have been increased. When a large current density is used, there is considerable decomposition of the iron electrodes when either salt or pure water is used, and in the case of horizontal electrodes, the under surface of the top plate may become covered with bubbles of gas, making the resistance between the plates quite variable. For large current density a horizontal top plate is not advisable, unless a large number of holes are drilled through it. A better form for the top electrode would be a hollow cylinder long enough to give sufficient surface. Washing soda is often a convenient substance to use instead of salt. If, from experience, the size of a liquid rheostat for absorbing a given amount of energy cannot be estimated, the dimensions may be calculated approximately as follows: Suppose, for instance, it is desired to absorb 60 amperes at 40 volts difference of potential between the electrodes. Now, it is inconvenient to obtain a saturated solution of salt, and to use tap water would require too large a cross section--especially if a barrel or trough is to be used--in order to have the resistance with the plates at a safe distance apart, small enough to give 60 amperes with 40 volts.