One of the most successful modern cells is that devised by Prof. Thirring of Vienna. In this type the bridges are built up on what is called the "condenser" plan, and consist of a number of thin strips of brass and mica arranged alternately, and so cut that the mica is slightly narrower than the brass. If such a pack of strips be clamped together and stood on its edge upon a flat surface the brass will stand slightly higher than the mica, and selenium melted over the surface, sinks into the tiny channels between the upper part of the brass strips and thus forms a series of bridges. The selenium is then carefully annealed and finally ground off so as to expose an extremely thin layer to the action of light. The alternate brass strips are connected together in two groups after the manner of an electrical condenser.
A thin layer of selenium is now generally admitted to be essen tial where sensitivity and greater reliability are desired. It should also be pointed out that for many years the necessity of exclud ing damp air from selenium cells has been realised, and an im portant contribution to the subject, recently made by Prof. Ran kine (Proceedings of the Physical Society of London, v. 39, Part 3, April z 5th 1927) and Mr. Avery, shows conclusively that the variable behaviour of selenium bridges is to a large extent attrib utable to the presence of moisture. Polarization occurs and causes irregularities and inability to reproduce readings. It is therefore important to dry and effectively seal up all selenium bridges re quired for experiment. This may be partially done by coating them with paraffin wax, shellac, resin, etc., but the best plan is to seal the cells into glass tubes after thorough drying by exposure to All forms of selenium appear to absorb moisture slightly. Selenium melts at 217° C and when cooled rapidly remains amorphous. On reheating to 18o° C it changes to one of the grey varieties, all of which are more or less sensitive to light. Close examination discloses a mass of minute crystals. Their shapes are largely influenced by the temperature at which they are formed. A thin layer of selenium for instance, which appears deep red by transmitted light, will gradually develop a crop of light-sensitive spherical crystals, the growth of which can be observed on the stage of a microscope if the temperature is main tained at about ioo° C (fig. 2). At temperatures in the neigh bourhood of 26o° C a transparent highly hydroscopic type of crystal is produced which is practically insensitive to light. This
and other forms of selenium crystals, excepting the spherical type, may most conveniently be obtained by condensation. A flat brass plate is placed upon a tripod and heated till selenium, rubbed over the central area of it, melts easily. This should then be smoothed off with the edge of a glass strip and a separating ring of mica laid on the molten selenium. A sheet of glass is then placed over and in contact with the mica, and the vapour rising from the selenium condenses on its under surface. At first a red deposit appears which a lens shows to consist of minute beads of transparent amorphous selenium. After a few minutes of con tinued heating however, the col our changes to grey and with experience, crops of very beauti ful crystals can be obtained (see fig. 3). An exposure of twelve hours, with the plate held an inch above the molten selen ium will, under suitable condi tions, produce quantities of needle-like crystals 2 or 3 mm.
long. If this be replaced by a slate tablet or gold coated plate on which a line, as previously described, has been cut, the selenium is soon deposited in sufficient quantity to constitute a light-sensitive layer. It is comparatively easy, in half an hour, to prepare a selenium cell in this way that will carry 5,000 micro-amperes with an illumina tion of ioo candle-feet and a potential difference of so volts.
The sensitivity of a selenium bridge is often expressed by stat ing the ratio of the "dark current" to that which passes when the cell is illuminated by a light of i oo candle-feet. In the case of a well-made cell this ratio may be i to ioo.
The value of the current should, on shutting off the light, return to within 5% of its initial value in one second. If a longer time lag is observed it is generally believed to be due to secondary effects. The time-lag is frequently spoken of as "inertia." Al though the above conditions have been actually realised, the usual practice is to work with a i to 6 or i to io ratio for the dark to light current, and to be satisfied with a return to within 5% of zero within five seconds.
It will be found that selenium melted on to ordinary glass will not always adhere well on conversion to the grey or metallic variety and cooling. By using finely ground glass however, and heating the selenium till it freely fumes, it may be stroked with a mica spatula to a smooth surface as it cools, and will then adhere to the base sufficiently well for most practical purposes.