of Edinburgh, Session 1880-81. 
121 
again. At that particular point further heating should he avoided, 
and the cell allowed to cool slowly. During the cooling the resist- 
ance rises, and does not appear to reach its final value till the cell 
has been some time in use. 
It is obvious that what is wanted, to make a selenium cell as 
sensitive as possible, is to have a long thin layer of it between two 
parallel conductors forming the electrodes, and so to arrange it, that 
as great a length of that layer as possible can be brought at once 
under the influence of the source of light. These conditions are 
well fulfilled in what may be called the radial cell, which consists 
of a series of brass sectors with very acute angles fixed upon the 
end of a wooden cylinder, and having narrow radial slits between 
them. This form is easily made by first fixing a disc of brass 
about 3 inches in diameter upon the wood, and then cutting it right 
through by a saw along a series of diameters equally inclined 
to each other. The odd number sectors are connected by one wire 
which forms the one electrode, and the even numbers by another 
wire which serves for the other electrode. 
My first supply of selenium running done, I ordered a second, 
and proceeded to make another cell exactly in the same way as I 
had made the previous one. This cell, however, when finished, 
presented some very troublesome and perplexing peculiarities, 
inasmuch as its resistance during annealing fell very low — to 
something like 120 ohms — and yet it w r as almost entirely in- 
sensitive to the action of light. This effect is, without doubt, 
due to some impurity in the selenium, which I hope to detect by 
a chemical analysis which is being made for me. 
The failure to get good selenium led me to think if anything else 
would serve as a substitute, and the first thing I determined to try 
was amorphous phosphorus. I took a radial cell and packed the 
slits full of the phosphorus, and then proceeded to test its resist- 
ance in the ordinary way, but was surprised to find that I could 
not get a balance. This led me to see that the cell itself was 
acting as a battery, and on trying it directly through the galvano- 
meter, the needle was deflected off the scale. I also observed that the 
current varied with pressure on the phosphorus ; and this suggested 
to me that it might be made to serve the double purpose of battery 
and loose contact in the ordinary form of microphone transmitter. 
