Wireless Transmission of Photographs / Second Edition, Revised and Enlarged 1919

The Receiver.—The magnified current from the relay U is taken to a special telephone receiver, the construction of which is given in Fig. 40. The diaphragm F is about 2¹/₂ inches diameter, and should be fairly thin but very resilient. Only one Fig. 41. Fig. 41a. coil is provided, and this should be wound with No. 47 S.S.C. copper wire for a resistance of about 2000 ohms. By using only one coil and therefore only one core, the movement of the diaphragm is centralised. To the centre of the diaphragm a light steel point is fastened, about ¹/₂ inch long, and provided with a projecting hook H. An enlarged view of this pin is given in Fig. 41. The movement of the diaphragm and consequently of the steel point P is communicated to a pivoted rod R, which is of special construction. A piece of aluminium tube 3³/₄ inches long, and of the section given at B, is bushed at one end with a piece of brass of the shape shown in Fig. 41a. A stiff steel wire T about 1 inch long (20 gauge) is screwed into the end of Z, and carries a counterbalance weight C. A hardened steel spindle, pointed at both ends, is fastened at D, and runs between two coned bearings, one of which is adjustable. The underside of Z is flattened, and a small coned depression is made for the reception of the pointed end of the pin. By means of the spring J the two pieces, Z and P, are held firmly together, at the same time allowing perfect freedom of movement. The bridge G is made from a piece of sheet aluminium placed in a slot cut in the tube R, the end of the tube being pressed tight upon G, and secured by means of a small rivet.

The optical arrangements are as follows. By means of the Nernst lamp L, and the lenses B and B', Figs. 42 and 43, a magnified shadow of G is thrown upon the screen J. When the shutter G is in its normal position (i.e. at rest), its shadow is just above the small hole in J, and light from L reaches the photographic film wrapped round the drum V of the machine.

When, however, signals are sent out from the transmitting apparatus, the magnified current from the relay U energises the coil of the special telephone S, attracting the diaphragm F, and consequently giving movement to the pivoted rod R. As by means of the optical arrangements a magnified movement as well as a magnified image of G is thrown upon the screen J, the shadow of G will, when the telephone S is actuated, cover the hole in the screen, and prevent any light from reaching the film on V, until current from the relay U ceases to flow. Therefore, when the stylus of the transmitter traces over a conducting strip on the metal print, no light reaches the film on V, but when tracing over an insulating strip the shadow of G on the screen J rises, and the light from L reaches the film. By this means a positive picture is received, which is a great advantage where the photographs are required for reproduction. Atmospherics would be represented by irregular transparent marks on the film after development, and these can be easily eradicated by retouching.

The drum of the machine moves laterally ¹/₇₅th of an inch per revolution, and the hole in the screen is ¹/₉₀th of an inch in diameter. As the screen J is not in direct contact with the film, the slight diffusion of the light that takes place will produce a mark of about the right thickness. With a movement of the diaphragm of only ¹/₄₀₀₀₀th of an inch, the actual movement of G will be ¹/₄₀₀₀th of an inch. If the optical arrangements have a magnifying power of 100, then the movement of the shadow upon the screen will be ¹/₄₀th of an inch, which will be ample to cover the aperture.

The aluminium rod R, minus the counter-weight, can be made to weigh not more than 12 grains. It is necessary to enclose the optical parts in a light tight box, indicated by the dotted lines in Fig. 43, in order to prevent any extraneous light from reaching the film.

The Contact Breaker.—The contact breaker (L, Fig. 35), as will be seen from Fig. 44, consists of an electro-magnet N, the windings of which are connected with the battery B and the polarised relay K. The armature which is supported by the spring G carries a contact arm A, which in its normal position makes permanent contact with the contact screw T, and completes the circuit between the relay K and the telephone relay U (Fig. 35). As soon as the transmitter sends out the first signal, the magnified current from the telephone relay actuates the relay K, which in turn completes the circuit of the contact breaker. Directly the armature M has been attracted, the contact with T is broken, and A makes fresh contact with the screw H, by means of the spring Z fastened to the underside of A. The armature, once it has been attracted, is held in permanent contact with H by the catch S, independent of the magnets N. As soon as contact is made with H, the clutch (F, Fig. 35) circuit is completed, and the circuit of the relay K is broken. When the circuit of the clutch F is broken by means of the circuit breaker C on the machine (Fig. 36), the stop S is pulled back by hand, allowing the contact arm A to rise, and again make fresh contact with the contact screw T.

Driving Apparatus

The Friction Brake.—This consists of a steel disc A, Fig. 45, about 2¹/₂ inches diameter and ³/₈ inch or ¹/₂ inch wide on the face, secured to the main shaft of the driving motor. The arm H, pivoted at C, carries at one end the curved block B, which is faced with a pad of tow F. The other extremity is pivoted to the steel rod P, which slides Fig. 45. in holes bored in the standards J. One end of the rod P is screwed with a fine thread, about 75 to the inch, and is fitted with a regulating wheel T, by means of which the block B can be made to press upon the disc A with any required degree of pressure. A fairly stiff steel spring R is placed upon the rod P, between one standard J and the collar N. As the speed of the driving motor is slightly in excess of that required by the machine, the block B, by means of the wheel, is made to press upon the disc A, setting up friction which reduces the motor speed until the isochroniser indicates that the correct working speed has been attained.

The Clutch.—The details of this will be seen from Figs. 46 and 47. It consists of a steel shaft coned at both ends running between two countersunk bearings, one of which is adjustable. This shaft carries the two portions of the clutch A and B, the portion A being a fixture on the shaft, and the portion B running free upon it. The portion B is a gun-metal casting bored to run accurately upon the steel shaft. A soft iron annular ring is fastened to the face.

The portion A consists of a gun-metal casting Fig. 47. bored a tight fit for the shaft E, secured by means of a set screw. The two magnet cores P are screwed into the front plate V, which is also of gun-metal, and after the bobbins R have been slipped on, the shanks of the cores are passed through holes drilled in the flange N of the main casting and held in place with nuts. The faces of both A and B must be turned perfectly square with the shaft, so that they run accurately together. The portion B is kept in contact with A by means of a spring S, the pressure being regulated by the collar H. Current is taken to the magnets by means of the two insulated copper rings D mounted upon the body of A. The gear-wheels on both portions have teeth of very fine pitch, the number of teeth on each being regulated by the speed of the driving motor and the required machine speed. Connection with the circuit breaker L and the battery B² is made with the collecting rings D by the brushes T. The complete connections are given in the diagram Fig. 51.

The Isochroniser.—This is a device for ensuring the correct speed regulation of the driving motors, and is shown in detail in Fig. 48. It comprises two portions, one portion being rotated at a definite speed by electrical means, and the other portion rotated by the driving motor.

The main portion consists of a metal tube N, bushed at both ends, the bottom end of the tube being arranged to work on ball-bearings. An ebonite bush C carries three copper rings T, T¹, T², and the brushes R, R¹, R² are in electrical contact with them. The ebonite plate J, 3¹/₂ inches diameter, is secured to the top end of N, and carries a contact piece Q, shown separate at E. As will be seen this is a block of ebonite with three contacts arranged on the top surface. The middle contact P is ¹/₆₄th of an inch wide, and the contacts P¹ and P² are placed on either side at a distance of ¹/₁₆ inch; the contact strips P¹, P² carry the brass pins D, which are about ¹/₁₆ inch diameter, and spaced ³/₈ inch apart. A connecting wire is carried from the contact P to the copper ring T, another from P¹ to T¹, and one from P² to T².

The bushes S are bored a running fit for the steel rod W (shown separate at A), which is coned at both ends, and runs between two countersunk bearings, the bottom bearing E being fixed while the top bearing (not shown) is adjustable. A needle K is fastened near the end of the rod W, and attached to this needle is the spring Z, which presses lightly but firmly upon the contact block Q. To provide a level surface for Z to work over, the spaces between the contact pieces are filled in with an insulating material, and the whole surface finished off perfectly smooth. The spring Z is ¹/₈ inch wide for portion of its length, but at the point where it presses upon Q it is reduced in width to ¹/₆₄th of an inch (see Fig. 48). The driving arrangements are as follows. A counter-shaft Q, Fig. 51, fitted with a grooved pulley, is run in bearings parallel with the shaft W, and is connected by suitable gearing to the shaft of the driving motor, so that the needle K makes one revolution in about 2¹/₂ seconds. A belt passing over the pulleys connects the two shafts, and the tension of the belt is regulated by means of an adjustable jockey pulley.

The tube N, carrying the disc J, must be rotated at a fixed speed, and this is accomplished in the following manner. An ordinary electric clock impulse dial, actuated from a master clock, is connected by suitable gearing H, so that the tube N makes exactly one revolution in 2 seconds; it being possible to adjust an electric clock of the "Synchronome" type, so that it only gains or loses about 1 second in 24 hours, and this provides an accuracy sufficient for all practical purposes. The connections are given in Fig. 49, and the face of the instrument in Fig. 50. It will be seen that a connecting wire is run from the steel spindle W to one terminal each of the lamps L, L¹, L², and from the other terminal of the lamps to one terminal of the batteries J, the battery comprising a set of three 4-volt accumulators. The other terminals of the batteries are joined one to each of the brushes R, R¹, R².

The lamps are coloured, the lamp L being white, and the lamps L¹ and L² blue and red respectively, and care must be taken in connecting up that when the needle K makes contact with the stud P the white lamp L is in circuit. When the machines are working, the operator, by means of the brake (already described), reduces the speed of the driving motor until the needle K travels in unison with the disc J, making permanent contact with P on the contact block Q, which is evidenced by the lamp L remaining alight. If, however, the needle travels faster than the disc J, contact with P is broken and fresh contact is made with P², the lamp L is extinguished and the red lamp L² lights up, and remains alight until the operator reduces the speed. Similarly, too, if the needle travels slower than J, contact is made with P¹, and the circuit of the blue lamp L¹ is completed. When the speed is either above or below the normal, the needle K engages with one or the other of the pins D, and as the tension of the driving belt is only such as is required to drive the needle, the belt slips on the pulleys until the normal speed is regained.

Method of Working

The clockwork motor M, Fig. 51, should be capable of running for several hours with one winding, and powerful enough to take up the work of driving the machine without any appreciable effort. The main spindle of the motor is so arranged that it makes one revolution in two minutes, and the reduction in speed between the motor shaft and the shaft to which the coupling A is attached is 30:1. The metal line print having been wrapped round the drum of the machine, the stylus is put into position, at the edge of the lap, and with the needle resting about half-way on the margin of the bare foil left at the commencing edge of the print. Now, when the two stations are in perfect readiness for work, the motors are started and the speed adjusted; the speed of the machine being just under one revolution in four seconds.

The switch D is then closed, and the arm of the switch D¹ placed on the contact stud (1), at the transmitting station only. As soon as the switches are closed the clutch F comes into action, and the transmitting machine begins to revolve. When the whole of the line print wrapped round the drum of the machine has passed under the stylus, the end of the shaft D, Fig. 36, engages with the spring m, breaking the clutch circuit and allowing the motor to run free. As soon as the machine stops, the switch D is opened and the machine run back to its starting position by hand.

At the receiving station the switch D is also closed, and the arm of the switch D¹ placed on the contact stud (2). The closing of these switches does not bring the clutch F into operation until current from the telephone relay U connected to the wireless receiving apparatus works the sensitive polarised relay K, which in turn completes the circuit of the circuit-breaker L. When the armature of L is attracted, the circuit of the relay K is broken, the circuit of the clutch F is completed, and the machine starts revolving.

The current from the relay U, due to the transmitting stylus passing over one contact strip on the metal print, is too brief to actuate the heavier mechanism of the relay K, hence the need of the margin of bare foil at the commencing edge of the metal print, so that a practically continuous current will flow to the relay K until the armature is attracted. As, however, the relay is not actuated at the receipt of the first signal, and as it is necessary for the machine to start recording at a certain point on the film, viz. at the edge of the lap—the reason for this was given in Chapter IV.—the starting position of the receiving drum will be similar to that given in the diagram Fig. 52, where X indicates the lap of the photographic film, and the arrow the direction of rotation.

It is, of course, obvious that a somewhat similar adjustment must be made with regard to the position of the stylus on the metal print at the transmitting machine.

In the present system, as in almost every photographic method of receiving that has been described, the Nernst lamp is invariably mentioned as the source of illumination. Since the advent of the high-voltage metal-filament lamps the Nernst lamp has fallen somewhat into disuse for commercial purposes, but it possesses certain characteristics that render it eminently suitable for the purpose under discussion.

The main principle of this type of lamp depends upon the discovery made by Professor Nernst in 1898, after whom the lamp is named, that filaments of certain earthy bodies when raised to a red heat became conductive sufficiently well to pass a current which raised it to a white heat, and furthermore that the glowing filament emitted a brighter light for a given amount of current than carbon filaments.

Nernst lamps are made in two sizes, the larger being intended for the same work as usually done by arc lamps, and the smaller to replace incandescent lamps; the smaller type being made to fit into the ordinary bayonet lampholders. The principal parts of a Nernst lamp consist of the filament, the heater, the automatic cut-out, and the resistance, and their arrangement in the smaller type of lamp is given in the diagram, Fig. 52a. The current enters at the positive terminal, passes through the heater M, and out through the negative terminal. The filament B, which consists of a short length of an infusible earth made of the oxides of several rare minerals, of which zirconia is one, is a non-conductor at first, but becomes a conductor upon being raised to a high temperature by means of the heater M. As soon as the filament becomes conductive the current then passes through the automatic cut-out H, and the armature D is attracted, thus breaking the heater circuit. The current then flows from the positive terminal Fig. 52b. through the cut-out H, resistance J, and filament B, and from thence out of the lamp. Since the resistance of the filament decreases the hotter it gets, it is necessary to insert a ballasting resistance in series with it which has the opposite property of increasing its resistance as it gets hotter, to prevent the filament taking too much current and destroying itself. Such a resistance, J, consists of a filament of fine iron wire, which, to prevent oxidation from exposure to the air, is enclosed in a glass bulb filled with hydrogen gas. Fig. 52b shows the form of ballast resistance used in the small and large type of lamp respectively.

Either direct or alternating current can be used with these lamps, and with direct current the polarity must be strictly observed, and that the positive wire is connected to the positive and the negative wire to the negative terminal. With the smaller type of lamp once it has been correctly placed in its holder it is essential that it should not be turned, as a change in the direction of the current will rapidly destroy the filament.

The arrangement of the larger type of Nernst lamp can be readily seen from the drawing, Fig. 52c.

Care must be taken to see that the voltage required by the burner and resistance equals the voltage of the supply circuit, and that only parts of the same amperage are used together on the same lamp. No advantage is obtained by over-running a Nernst lamp, this only shortening its life without increasing the light. Under normal conditions the average life of the burner is about 700 hours.

The efficiency of the Nernst lamp is fairly high, being only 1.45 to 1.75 watts per c.p. The light given is remarkably steady, and the lamps are adaptable for all voltages from 100 to 300. In one of the large type of lamps for use on a 235-volt circuit the burner takes 0.5 ampere at 215 volts, and the resistance 0.5 ampere at 20 volts, while one of the smaller lamps for use on the same circuit takes 0.25 ampere at 215 volts and 0.25 ampere at 20 volts for the burner and resistance respectively. The burner and heater are very fragile, and should never be handled except by the porcelain plate to which they are attached. The lamps burn in air and emit a brilliant white light of high actinic power, the intrinsic brilliancy (c.p./square inch) varying from 1000 to 2500, as compared with 1000 to 1200 for ordinary metal filament lamps, and 300 to 500 for carbon filament lamps.

The chief advantage of the Nernst lamp from a photographic point of view lies in the fact that it produces abundantly the blue and violet rays which have the greatest chemical effect upon a photographic plate or film. These rays are known as chemical or actinic rays, and are only slightly produced in some types of incandescent electric lamps. Carbon-filament lamps are very poor in this respect.

Because a light is visually brilliant it must by no means be assumed that it is the best to use for purposes of photography, and this is a point over which many photographers stumble when using artificial light. Many sources of light, while excellent for illumination, have very low actinic powers, while others may have low illuminating but high actinic powers. A lamp giving a light yellowish in colour has usually low actinic power, while all those lamps giving a soft white light are generally found to be highly actinic.

In addition to the actinic value of the source of illumination, the photographic film used must be very carefully chosen, as the chemical inertia of the sensitised film plays an important part in the successful reproduction of the picture, and also, to a certain extent, affects the speed of transmission. The length of exposure, the amount of light admitted to the film, and the characteristics of the film itself, are all factors which have a decided bearing upon the quality of the results obtained, and the film found to be most suitable in one case will perhaps give very unsatisfactory results in another.

In photo-telegraphy the length of exposure is determined by the time taken by the transmitting stylus to trace over a conducting strip on the metal print, and this time, of course, varies with the density of the image and also with the speed of transmission.

The film in ordinary photography is chosen with regard to the subject and the existing light conditions, and the amount of light admitted to the film and the length of exposure are regulated accordingly. No such latitude is, however, possible in photo-telegraphy. With each set of apparatus the various factors, such as the light value, the amount of light admitted to the film, and the length of exposure, will be practically fixed quantities, and the film that will give the most satisfactory results under these fixed conditions can only be found by the rough-and-ready method of "trial and error."

The films in common use are manufactured in four qualities, namely, ordinary, studio, rapid, and extra rapid. These terms should really relate to the light sensitiveness of the film (or, as it is technically termed, the speed), but at the best they are a rough and very unsatisfactory guide, for the reason that some unscrupulous makers, purely for business purposes, do not hesitate to label their films and plates as slow, rapid, etc., without troubling to make any tests for correct classification.

The speed of photographic films and plates is generally indicated by a number, and the system of standardisation adopted by the majority of makers in this country is that originated by Messrs. Hurter & Driffield, abbreviated H. & D. In their system the speed of the film and the exposure varies in geometrical proportion, a film marked H. & D. 50 requiring double the exposure of one marked H. & D. 100. The highest number always denotes the highest speed, and the exposure varies inversely with the speed.

Besides the Hurter & Driffield method of obtaining the speed numbers of plates and films adopted by a large number of makers in this country, there are also two standard English systems known as the W.P. No. (Watkin's power number) and Wynne F. No., both of which are used to a fair extent.

The "Actinograph" number or speed number of a plate in the H. & D. system is found by dividing 34 by a number known as the Inertia, the Inertia, which is a measure of the insensitiveness of the plate, being determined according to the directions laid down by Hurter & Driffield—that is, by using pyro-soda developer and the straight portion only of the density curve. If, for instance, the Inertia was found to be one-fifth, then the speed number would be 34 ÷ ¹/₅ = 170, and the plate is H. & D. 170. The W.P. No. is found by dividing 50 by the Inertia. Thus 50 ÷ ¹/₅ = 250, and the plate is W.P. 250, but for all practical purposes the W.P. No. can be taken as one and a half times H. & D. The Wynne F. numbers may be found by multiplying the square root of the Watkins number by 6.4. Thus

For those photographers who are in the habit of using an actinometer giving the plate speeds in H. & D. numbers, the following table, taken from the Photographer's Daily Companion, is given, which shows at a glance the relative speed numbers for the various systems. The Watkins and Wynne numbers only hold good, however, when the inertia has been found by the H. & D. method.

Table of Comparative Speed Numbers for Plates and Films

Although theoretically the higher the speed of the film the less the duration of exposure required, there is a practical limit, as besides the intensity and actinic value of the light admitted to the film a definite time is necessary for it to overcome the chemical inertia of the sensitised coating and produce a useful effect. With every make of film it is possible to give so short an exposure that although light does fall upon the film it does no work at all—in other words, we can say that for every film there is a minimum amount of light action, and anything below this is of no use. The exposure that enables the smallest amount of light action to take place is termed the limit of the smallest useful exposure.

There is also a maximum exposure in which the light affects practically all the silver in the film, and any increased light action has no increased effect. This is the limit of the greatest useful exposure.

In photo-telegraphy the duration of exposure, as already pointed out, is determined by certain conditions connected with the transmitting apparatus, and with conditions similar to those mentioned on page 75 the length of exposure will vary roughly from 1-50th to 1-150th of a second.

The most suitable film to use for purposes of photo-telegraphy is one having a fairly slow speed in which the range of exposure required comes well within the limits of the film. There is no advantage in using a film having a speed of, say, H. & D. 300 if good results can be obtained from one with a speed of, say, H. & D. 200, as the use of the higher speed increases the risk of overexposure. With the high-speeded films the difficulties of development are also greatly increased, there being more latitude in both exposure and development with the slower speeds, and consequently a better chance of obtaining a good negative.

Another point, often puzzling to the beginner, and which increases the difficulty of choosing a suitable make of film, is that, although one make of film marked H. & D. 100 will give good results, another make, also marked H. & D. 100, will give very poor results. This is owing, not to a poor quality film, as many suppose, but to the almost insurmountable difficulty of makers being able to employ exactly the same standard of light for testing purposes, so that although various makes may all be standardised by the H. & D. method, films bearing the same speed numbers may vary in their actual speed by as much as 30 to 50 per cent.

APPENDIX A

SELENIUM CELLS

Selenium is a non-metallic element, and was first discovered by Berzelius in 1817, in the deposit from sulphuric acid chambers, which still continues the source from which it is obtained for commercial purposes, although it is found to a small extent in native sulphur. Its at. wt. is 79.2, and its sp. gr. 4.8. Symbol, Se.

In its natural state selenium is practically a non-conductor of electricity, its resistance being forty thousand million times greater than copper. Its practical value lies in the property which it possesses, that when in a prepared condition it is capable of varying its electrical resistance according to the amount of light to which it is exposed, the resistance decreasing as the light increases.

Selenium is prepared by heating it to a temperature of 120° C., keeping it there for some hours, and allowing it to cool slowly, when it assumes a crystalline form and changes from a bluish grey to a dull slate colour. A selenium cell in its simplest form consists merely of some prepared selenium placed between two or more metal electrodes, the selenium acting as a high resistance conductor between them. The form given by Bell and Tainter to the cells used in their experiments is given in Figs. 53 and 53a. It consists of a number of rectangular brass plates P, P', separated by very thin sheets of mica M, the mica sheets being slightly narrower than the brass plates, the whole being clamped together in the frame F by the two bolts B. By means of a sand-bath the cell is raised to the desired temperature, and selenium is rubbed over the surface, which melts and fills the small spaces between the brass plates. All the plates P are connected together to form one terminal, and the plates P' to form the other. By using very thin mica sheets, and a large number of elements, a very narrow transverse section of selenium, together with a large active surface, can be obtained.

The cell used for commercial purposes is usually constructed as follows. A small rectangular piece of porcelain, slate, mica, or other insulator, is wound with many turns of fine platinum wire. The wire is wound double, as shown in Fig. 54, the spaces between the turns being filled with prepared selenium. A thin glass cover is sometimes placed over the cell to protect the surface from injury.

A strong light falling upon a cell lowers its resistance, and vice versa, the resistance of a cell being at its highest when unexposed to light; the light is apparently absorbed and made to do work by varying the electrical resistance of the selenium. Selenium cells vary very considerably as regards their quality as well as in their electrical resistance, it being possible to obtain cells of the same size for any resistance between 10 and 1,000,000 ohms, and also, a cell may remain in good working condition for several months, while another will become useless in as many weeks.

The ability of a cell to respond to very rapid changes in the illumination to which it is exposed is determined largely upon its inertia, it being taken as a general rule that the higher the resistance of a cell the less the inertia, and vice versa, and also, that the higher the resistance the greater the ratio of sensitiveness. Inertia plays an important part in the working of a cell, slightly opposing the drop in resistance when illuminated, and opposing to a Fig. 54 much greater degree the return to normal for no-illumination. The effects of inertia or "lag," as it is termed, can readily be seen by reference to Fig. 55. It will be noticed that the current value rapidly increases when the cell is first illuminated, but if after a short time t the light is cut off, the current value, instead of returning at once to normal for no-illumination, only partially rises owing to the interference of the inertia, and some time elapses before the cell returns to its normal condition; the time varying from a few seconds to several minutes, depending upon the characteristics of the cell and the amount of light to which it is exposed. An actual curve is given in Fig. 55a. The inertia or "lag" of a cell produces upon an intermittent current an effect similar to that produced by the capacity Fig. 55 of a line, as was noted in Chapter I., preventing the incoming signals from being recorded separately, and distinctly. To obtain the best results in photo-telegraphy, the resistance of a cell should only be decreased to an extent sufficient to pass the current required to operate the recording apparatus, and the illumination should be regulated so that this condition of the cell takes place.

The comparative slowness of selenium in responding to any great changes in the illumination offers a serious difficulty to its use in photo-telegraphy, but various methods have been devised whereby the effects of inertia can be counteracted. In the system of De' Bernochi (see Chapter I.) the changes in the illumination are neither very rapid nor very great, and the inertia effects would therefore be very slight; but in any photo-telegraphic system in which a metal line print is used for transmitting, where the source of illumination is constant and the resistance of the cell is required to drop to a definite value and return to normal instantly, many times in succession, the inertia effects are very pronounced. The most successful method of counteracting the inertia is that adopted by Professor Korn of always keeping the cell sufficiently illuminated to overcome it, so that any additional light acts very rapidly. Another method worked out and patented by Professor Korn, and known as the "compensating cell" method, gives a practically dead beat action, the resistance returning to its normal condition as soon as the illumination ceases. The arrangement is given in the diagram Fig. 56.

Light from the transmitting or receiving apparatus, as the case may be, falls upon the selenium cell S¹, which is placed on one arm of a Wheatstone bridge, a second cell S² being placed on the opposite arm. The selenium cell S¹ should have great sensitiveness and small inertia, the compensating cell S² having proportionally small sensitiveness and large inertia. Two batteries B, B', of about 100 volts, are connected as shown, B being provided with a compensating variable resistance W; W' is also a regulating resistance. When no light is falling upon the cell S¹, light from L is prevented from reaching the second cell S² by a small shutter which is fastened to the strings of the Einthoven galvanometer (described in Chapter III.), and the piece of apparatus C—relay or galvanometer as the case may be—remains in a normal condition. When, however, light falls upon the cell S¹, the balance of the bridge is upset, and light from L falls a fraction of a second later upon the second cell S², and the current flowing through C completes the circuit. Needless to say it is necessary that the two cells be well matched, as it is very easy to have over-compensation, in which case the current is brought below zero.