The system I have adopted is the Wheatstone's bridge arrangement, with equal sides, never using multipliers except for some experimental purpose. In each multiplier wire I have 500 ohms resistance. When the bridge is balanced, one-half of the current flows through the cell and acts upon the selenium. Between the bridge and the cell is a reversing switch, so that the current can be reversed through the cell without changing its course through the bridge. A Bradley tangent galvanometer is used, employing the coil of 160 ohms resistance. The Leclanche battery is exclusively used in measurements for comparison.

2. The kind of battery employed has a marked effect upon the sensitiveness to light, which is largely reduced or entirely destroyed when the bichromate battery is used. The same cells again become extremely sensitive with the Leclanche battery. We might expect that a change in the current employed would cause a change in the resistance of a cell, but it is not clear how or why it should affect the sensitiveness of selenium to light.

"If one kind of battery current destroys its sensitiveness, may we not suppose that another kind might increase its sensitiveness? Although the Leclanche has operated well, some other may operate still better, and by its special fitness for use on selenium cells may intensify their actions, and so bring to light other properties yet unthought of. Is not here a promising field for experiment, in testing the various forms of battery already known, or even devising some new form especially adapted to the needs and peculiarities of selenium cells?"

One year ago I made the foregoing suggestion in a paper on A New Form of Selenium Cell, presented before this Association at Minneapolis. I am now at liberty to state that my photo-electric battery, presently to be described, marks an advance in the direction indicated. The current from this battery increases the sensitiveness of the cells to light, and also to reversal of current. One cell whose highest ratio in light was about 83 to 1, with the Leclanche battery, when measured with my battery gave a ratio of 120 to 1. It seems to make the resistance of the cell both higher in dark and lower in sunlight than with the Leclanche battery. But the field is yet open to others, for the discovery of a battery which may be still better for use with selenium cells.

3. The two surfaces of the selenium act differently toward currents sent into them from the contiguous conductors. One surface offers a higher resistance to the current than the other. The former I utilize as the anode surface, as I have found that the cell is more sensitive to light when the current enters at that surface, which is ordinarily the one covered by the gold or other transparent conductor. Some cells have this property but feebly developed; but in one instance the resistance offered to the current by the anode surface was 256 times as high as that offered by the cathode surface to the same current. In the majority of cases, however, the ratio does not exceed ten times. Table B gives some recent results.

 -------------------------------------------------------------------

TABLE B.

SENSITIVENESS TO REVERSAL OF DIRECTION OF CURRENT.

-----------------------+--------------+------------------+---------

| | Resistance |

No. of cell. | Battery. | "gold | "gold | Ratio

| | anode."|cathode."|

-----------------------+--------------+--------+---------+---------

| | ohms. | ohms. |

3/8 inch square. No. 4 | 5 elements. | 20,000 | 1,000 | 20 to 1

" " " 3 | Se. cell. | 6,500 | 400 | 16.2 "

Full size, No. 13 | 1 element. | 9,000 | 800 | 11.2 "

" " " 14 | 5 " | 2,440 | 130 | 18 "

" " " 15 | 5 " | 4,640 | 210 | 22 "

" " " 27 | 5 " | 6,900 | 440 | 16 "

" " " 126 | 1 " | 5,000 | 330 | 15 "

-----------------------+--------------+--------+---------+--------- 

The direction of the current is always indicated by stating the position of the gold electrode, by the terms "gold anode" and "gold cathode." The above measurements were made in dark.

4. Sensitiveness to change of battery power. - My cells are extremely sensitive to any change in the strength or character of the current flowing through them, which is shown by a corresponding change in the resistance of the cell. I can, therefore, vary the resistance of one of my cells in many ways, and the following may be specified -

(a) By changing the potential or electromotive force of the current through the cell.

(b) By changing the "quantity" of the battery or current.

(c) By putting more or less resistance in the circuit.

(d) By dividing the current, by one or more branch circuits or shunts around the cell.

(e) By varying the resistance in any or all of said circuits.

A cell whose resistance becomes greater as the battery power becomes greater, and vice versa, I call an "L B cell" signifying Like the Battery power. A "U B cell" is one whose resistance becomes greater as the battery power (or strength of current) becomes less, and vice versa, being Unlike the Battery power, or current strength.

These changes of resistance are not due to heating of the conductor or the selenium, and the following instance will illustrate this. I have one cell in which the selenium has about one-fourth inch square of surface melted on a brass block one inch thick. This cell measured, with 25 elements of Leclanche, 40,000 ohms. On changing the battery to 5 elements the resistance fell instantly to 30 ohms, and there remained. On again using the current from 25 elements, the resistance instantly returned to 40,000 ohms. Had these results been due in any degree to heating, the resistance would have changed gradually as the heat became communicated to the brass, whereas no such change occurred, the resistances being absolutely steady. Moreover, even the fusion of the selenium would not produce any such change.

The "U B" property does not ordinarily change the resistance of the cell to exceed ten times, i.e., the resistance with a weak current will not be over ten times as high as with a strong one. But I have developed the "L B" property to a far higher degree. Table C gives some recent results obtained with L B cells, including one whose resistance, with 25 elements Leclanche, was 11,381 times as high as with 8 elements, and which, after standing steadily at 123 ohms (and then at 325 ohms with 1 element), on receiving the current from 25 elements again returned to its previous figure of 1,400,000 ohms.