•tw. 
 
Ou 1/ 
 
 ^nuW Uttim^itg J 
 
 ^Mmi 
 
 BOUGHT WITH THE INCOME | 
 
 FROM THE 
 
 
 SAGE ENDOWMENT 
 
 FUND 
 
 THE GIFT OF 
 
 
 Sienrg ^m. Sage 
 
 
 189X 
 
 ■W-^Jn 
 
 ..AJA53SJ. 
 

Cornell University 
 Library 
 
 The original of tliis book is in 
 tine Cornell University Library. 
 
 There are no known copyright restrictions in 
 the United States on the use of the text. 
 
 http://www.archive.org/details/cu31924031219367 
 
[THREE SHILLINGS AND SIXPENCE.] 
 
 ■ 
 
 HANDBOOKS FOR PHOTOGRAPHERS. 
 
 EVENING TALKS 
 
 AT THE -'- 
 
 Camera ffilub 
 
 ON 
 
 THE ACTION OF 
 
 LIGHT IN PHOTOGRAPHY 
 
 BY 
 
 CAPTAIN ABNEY, c.b., dc.l., f.r.s. 
 
 ILVUSTRATED. 
 
 LONDON 
 SAMPSON LOW, MARSTON & COMPANY 
 
 ^ f!l m 
 
s •*;j' 
 
 CO 
 
 D 
 
 
 
 
 
 1^1 
 
 
 LU 
 
 _i , 
 
 < 
 
 I- 
 
 O 
 a. 
 
 h-- 
 o 
 
 ■ni It 'Z *M '■ 11 OS 
 'e Sm -|»iq ?9 'fS SI '8 "31 J 
 ■•<^\n 'e -Sia— f9iq?8-nl fit 
 
 's 'Sm ! njSB'fiSii:— giqoi 
 
 '9di/ ;sod ii9B 
 
 siiiiao aixiJKoa 
 
 C4 
 
 
 |-a.|s 
 
 'ssouj ^spi 
 
 fc saivij iaa 
 
 ^ pnB 'jayCamiiBa 'bsoh Xq 
 
 ,§ sasNai 
 
 I. 
 
 CO . 
 
 0, { 
 
 -°°M, 
 
 C8 CO (n ■ 
 
 *» » 
 o" - 
 
 « ® I 
 
 a. o 5 
 
 •s SI 
 
 5 ^? 
 
 Ri g "" 
 
 « « d 
 
 s ^ 8 
 
 ^ - a 
 
 £• i 5 
 
 •a i -5 
 
 5 » 2 
 
 » C^ "^ 
 
 s -. S 
 
 "< ft s 
 
 S <§ b 
 
 o c w 
 
 i; I ft 
 
 £ I £ 
 
 p, h 
 
 IS 
 
 rS- S 
 
 s & 
 
 Oil o 
 
 « a 
 O 
 
 I I 
 
 •a 
 § ^ I 
 
 ^ ^ S 
 
 & I R 
 1 a 
 
 o 
 
 S 
 
 o 
 
 o 
 
 o 
 
 cc 
 
 s 
 
 a. 
 
 B 
 
 CO 
 
 JB 
 
 3 
 
 O 
 
 CO 
 
 c>< 
 
 o 
 
 3 
 C 
 
Uonthly, Price 3d. ; Post Free, 4id. 
 
 A. JOXJB,:^T A-X. FOR FUOTOO- RA-FKCEI^S- 
 
 EDITORS: 
 H. SNOWDEN WAED. CATHARINE WEED WARD. 
 
 Has the Largest Circulation per issue of any Fhotographio Journal. 
 
 Tlia Optician. — " Far and away ahead of its photographic contemporaries." 
 
 Vows of the HfTorld. — " The Photogram is a most excellent magazinei with 
 
 charmingljr reproduced photograms, and excellent practical articles." 
 
 Beax^ and Home. — ** Crammed with vaiied and interesting matter . . . and a 
 
 vast amount of information to the lover of photography, while the illustrations are excellent." 
 
 Wii^^VM 
 
 Kontlily, 
 
 prioe ed. ; 
 
 post free, 8d. 
 
 
 Per Annnm, 
 post free, 88. 
 
 >^3SrX> ILLTJSTn-A-TOIt. 
 
 The Oldest Journal Specially Devoted to tlie Interests of Fhoto-Uechanical Workers. 
 Tbe Inland Ftlnter says:— "From across the water we receive Tbe Fxoeeis 
 
 Fbotogram. a Marvel of JSxcellenee, 
 Frequent Valuable Supple ments. Wo Process Worker pan keep up to Sate vithout it. 
 
 The Photogram for 1896 and The Process Photogram for 1896. Handsomely 
 bound in cloth, blocked in gold and colour, and crammed with illustrations and 
 supplements. Price 78. 6d. each, post free 8s. (A few copies still remain of 
 Vol. I. of The Photogram, 1894, and may be had at the same price.) 
 
 " PHOTOGRAMS OF '96." 
 
 A Pictorial and Literary record of the best Photographic Work of the Year. 
 
 Price, handsome cloth, 2/- ; paper covers, 1/- nett ; postage 6d. extra. 
 
 KBA-IDTZ" IIT OCTOBEH. 
 
 A few eopiea of "FUotOBTams of '96 " vnoy 'tm ie had at same price at above, 
 
 THE ABOVE ABE PUBLISHED POR 
 
 THE PHOTOGRAM, LIMITED, 
 
 By DAWBABN & WARD, Ltd., 6, Faningdon Avenue, E.O. 
 
1 ^ 
 
 'fgvlS 
 
 m V * 1 ^ 
 
 HAND- 
 CAMERA 
 
 £2 18 8 
 
 TO 
 
 £12 12 
 
 CARRIES FORTY FLAT FILMS " LIKE A PACK OF CARDS." 
 
 The FRENA CAUEBA can be successfully used by the absolute novice 
 
 or the expert photographer. 
 It is a neat little leather- covered instrument no larger than a small hand-bag. 
 The photographs are taken upon flat celluloid films, which are put into the 
 
 camera like a pack of cards. 
 Forty films can be put into the camera at one charge. 
 One or all of the exposed films can be taken out at any lame without interfering 
 
 with those that remain. 
 Three operations only are required in taking a photograph: — 
 
 1. Set off the shutter by pressing the bolt. 
 
 2. Wind the shutter ready for the next exposure. 
 
 3. Change the film by turning a handle. 
 
 The LENS is a Beck's Autograph Lens, giving superb definition, provided 
 
 with diaphragms. 
 The SHT7XTEK is arranged to give either time exposures or instantaneous 
 
 of h A. TO. A, -io second. 
 The Camera is provided with a swing-back and level, and can be used on a 
 
 tripod if desired. 
 The Finders give a view corresponding with the image on the film. 
 An Indicator registers the number of photographs taken. 
 Every Camera is suppUed with a handbook so complete that no further 
 
 instruction is necessary. 
 
 68, Cornliill, London. 'K.O., 
 
 Umufaeturer^ of L^n^es, Cameras, Microscopes, Telescopes, and Optical Instruments. 
 
 DEAiiBBs nr aiA fbotogbafbic affabatvs. 
 
EVENING TALKS 
 
 AT THE 
 
 Camera Club 
 
 ON 
 
 THE ACTION OF 
 
 LIGHT IN PHOTOGRAPHY 
 
 BY 
 
 CAPTAIN ABNEY, c.b., d.c.l., f.r.s. 
 
 ILLUSTRATED. 
 
 LONDON 
 SAMPSON LOW, MARSTON & COMPANY 
 
 1897. 
 
K \o's^^\ 
 
 LONDON : 
 CARTER AND CO., PRINTERS, 5, FURNIVAL STREET, HOLBORN, E.G. 
 
PREFACE. 
 
 In th^ autumn of the year 1 895 I was requested 
 by the Committee of the Camera Club to give 
 some evening talks on "The Action of Light 
 in Photography." They were given from notes, 
 and illustrated by experiments, and I am led 
 to believe they created a certain amount of 
 interest. 
 
 At the request of several members of the 
 Camera Club I have in the first chapters 
 endeavoured to put into a shape for publication 
 what I believe I said, or, at all events, what 
 I intended to say. 
 
 It seemed, however, that the work would be 
 incomplete unless it included an account of 
 the experiments I had made on the action of 
 feeble and intense light on a photographic plate. 
 
11. PREFACE. 
 
 These, except in one case, had formed the 
 subject of papers read before the Camera Club 
 on different Thursday evenings, and so have 
 been added to the more informal talks, with 
 the hope that they will prove suggestive to 
 the reader of further experiments. 
 
 The photographer, as a rule, is not too 
 fond of the theoretical considerations that 
 underlie his craft. My aim always has been, 
 and I hope always will be, to try and arouse 
 an interest in the scientific side of it, as it is 
 from it chiefly that future advances in photo- 
 graphy can be built up. 
 
 That there are many faults in the treatment 
 of the subject no one is more aware than 
 myself, but I have thought it better to keep 
 to the original line I took in my evening talks 
 than to travel in another. 
 
 W. DE W. Abney. 
 South Kensington, October, 1896. 
 
CONTENTS. 
 
 — • — 
 
 CHAPTER I. 
 Motions of Molecules and Atoms . . 
 
 CHAPTER n. 
 Lengths of Waves in the Spectrum 
 
 CHAPTER in. 
 Absorption 
 
 CHAPTER IV. 
 Eifect of Small Partigles in the Path of Light . . 34 
 
 CHAPTER V. 
 Light in the Dark Room . . . . . . . . 46 
 
 CHAPTER VI. 
 Light Scattered by Large Particles . . . . . . 53 
 
 CHAPTER VII. 
 Interference and Diffraction . . . . . . • • 59 
 
 CHAPTER VIII. 
 Pinhole Images . . . . . . . . • • 7 1 
 
CONTENTS. 
 
 CHAPTER IX. 
 The Colours of Thin Films 8i 
 
 CHAPTER X. 
 Modes of Measuring Grey Surfaces and Opacities 91 
 
 CHAPTER XI. 
 Measurement of the Luminosity of Colours . . 107 
 
 CHAPTER XII. 
 Eifect of Temperature of Sensitiveness .. .. izo 
 
 CHAPTER XIII. 
 Effect of Intermittence of Exposure .. ■• '37 
 
 CHAPTER XIV. 
 Effect of Small Intensities of Light. . .. .. 148 
 
 CHAPTER XV. 
 Action of Very Intense Light on Plates .. ..158 
 
 CHAPTER XVI. 
 Measurement of Spectra, &c. . . . . ..171 
 
 APPENDIX. 
 
 Tables of Temperature on Sensitiveness .. •• 179 
 
 Tables of Effect of Intermittent Exposure . . . . 1 84 
 
 Tables of Effect of Small Intensities of Light . . 188 
 
 Tables of Action with Intense Light . . . . 193 
 
 A Theory of Alkaline Development . . . . 198 
 
THE ACTION 
 
 OF 
 
 LIGHT IN PHOTOGRAPHY. 
 
 CHAPTER I. 
 
 Motions of Molecules and Atoms. 
 
 When we have to study a subject, it is well to start 
 with first principles, so far as we know them, and to 
 call in analogy to aid us in making our conceptions. 
 For instance, when we begin to consider the reason 
 why light can act upon matter, it is not sufficient to 
 khow that the ultimate subdivision of matter is a 
 molecule or an atom. We must take into account all 
 we have good grounds for believing about such atoms 
 or molecules, and endeavour to form a picture in our 
 minds as to what is taking place, and what they are 
 doing when light is causing a change. 
 ~ We have the best of evidence, which we need' 
 not enter into here, for believing that every atom and 
 molecule at ordinary temperatures is in a state of 
 unrest, or movement. If we are dealing with a gas, we 
 know from experiment that the molecules move about 
 in paths of their own, darting away in one direction till 
 they come into collision with another molecule, and 
 then darting off in a different direction, only to collide 
 once more. By applying fairly simple mathematics to 
 
 B 
 
2 LIGHT IN PHOTOGRAPHY. 
 
 the result of experiments made on the densities of gases, 
 the average velocity with which they move has been 
 calculated. We thus find that at atmospheric pressure 
 the molecules of hydrogen travel at a mean rate of 1,842 
 metres a second. When the number of molecules in a 
 given space is very much larger, as they are in liquids or 
 solids, they are not free to move to the sa;me extent. As the 
 mutual attractions hold them comparatively close together, 
 the paths in which they move are very much shorter, and 
 they come into collision with other molecules almost as 
 soon as they start on their journeys, and have to return 
 whence they started, and that thus they may take up a 
 rhythmic motion of some kind. As the temperature rises, 
 the body, as a rule, expands and becomes larger, and thus 
 the small enclosed space in which the motion of each 
 individual molecule takes place becomes enlarged, which 
 means that the length of path which it can follow 
 is increased. A molecule, however, is composed 
 of elementary atoms. For instance, a molecule of 
 silver chloride, we have excellent reason for believing, 
 is composed of, at least, two atoms of silver and two 
 atoms of chlorine. If we place silver chloride in a 
 crucible, and mix carbonate of soda with it, we find that 
 on applying moderate heat both swell ; but that if we 
 increase the heat, the mass liquefies, and the atoms of 
 chlorine leave the chloride of silver and combine with 
 the soda salt, leaving the silver atoms free to coalesce 
 with one another, to form, on cooling, a solid silver mass. 
 This indicates that, though when at a low temperature 
 the molecules of the soda and silver salts are separated, 
 yet when the temperature is increased the molecular 
 motion is so great that the atoms themselves take larger 
 
ATOMS AND MOLECULKS. 3 
 
 swings. For we must conceive in our minds that these 
 atoms, though kept together to form the molecule by 
 a central attraction, are also in motion among them- 
 selves, oscillating backwards and forwards towards one 
 another, or revolving round one another, such motions 
 being probably more rhythmical than those of the mole- 
 cules themselves. We can readily account for the atoms 
 of silver and chlorine, which make up the silver chloride, 
 leaving one another, if we suppose that the increased 
 space caused by the increased temperature, in which 
 the molecules are free to move, gives more room — 
 very minute though it be — in which the atoms can 
 swing, and that the increased energy of the impacts of 
 the molecules against each other are sufficient to set up 
 the longer swings. The attractive force which, at an 
 ordinary temperature, and without other influences at 
 work, compels the atoms to cling together, will be so 
 diminished by the increased distance which they travel 
 from the centre of attraction, that some of them (in this 
 case the chlorine) may come under the influence of some 
 other external and greater attraction existing in the 
 molecules of the carbonate of soda, and that therefore 
 they will leave the silver atoms and cause a decompo- 
 sition of an original molecule of silver chloride. If this 
 occur, the molecule, before it is actually decomposed, 
 will occupy a larger space, which, added to their 
 increased distance apart, may cause an increase in the 
 volume of the heated substance. We have now 
 pictured to ourselves both the molecules, and the 
 atoms composing them, in a state of increased motion 
 when the temperature of the body is raised. 
 As to the sizes of these molecules or atoms, and the space 
 
4 LIGHT IN PHOTOGRAPHY. 
 
 which separates them one from another, we need not enter 
 into ; suffice it to say that both the molecules and their 
 distances apart are ultra-microscopic ; they have never 
 been seen, and probably never will be. Their approxi- 
 mate dimensions and distances apart are known within 
 limits from circumstantial evidence of a most convincing 
 kind, but for our purpose it is quite unnecessary to enter 
 into the evidence or the arguments founded upon it. 
 All we wish to impress is that both atoms and molecules 
 in solids have vibratory motions, and that probably the 
 oscillations of the former are more regular and rhythmic 
 than those of the latter, the one being what might be 
 termed free vibration, and the other vibration due to 
 collisions. 
 
 But we have further to conceive that atoms and 
 molecules are imbedded in a medium which is infinitely 
 elastic, and yet somewhat jelly-like. Let us take a jelly 
 made with ordinary gelatine, moulded into a cylinder, 
 and let several small weights suspended by strings press 
 against its sides. Now let us remove one weight from 
 the side, still keeping it suspended from the string, and 
 allow it to swing against the jelly. We shall see that 
 all the other weights spring away from the jelly, showing 
 that the blow by the one sets the others in motion, 
 the jelly being the medium of communication, owing to 
 its elasticity. If we continued to give a series of rhythmic 
 blows by the one weight, the others (if the strings were 
 of the same length) would take up the same motion. 
 We can apply this example to the motion of the mole- 
 cules and atoms in the jelly-like ether which surrounds 
 them. Let them vibrate or oscillate as they will, the 
 ether will carry the blows it receives as impulses through 
 
PENDULUM ANALOGIES. S 
 
 it, and may set in motion (or increase the motion of) other 
 atoms or molecules. These carried impulses we call 
 radiation, for they radiate from the source of molecular 
 motion. 
 
 We have, owing to the transmission of these effects 
 of motion, a means of gauging to some extent — we may 
 say to a very large extent — the amount and character of 
 the original motion. Suppose we have a certain rhythmic 
 motion in the molecule, we should have the same kind of 
 impulses conveyed through the ether, and an atom or 
 molecule whose swing can keep time with these im- 
 pulses would be impressed with the same kind of motion 
 as that which started it. For purposes of illustration 
 we may take away any intervening conductor of motion, 
 and communicate such motion almost direct. For 
 instance, let us have a pendulum constructed as follows : — 
 
 B 
 
 A 
 
 Fig. i. 
 
 Fig. I is a double pendulum arranged as shown. The 
 pendulum A is heavily weighted, whilst B is light, 
 being only a string with a small weight attached. This 
 difference in weight is made to prevent any great effect 
 
O LIGHT IN PHOTOGRAPHY. 
 
 of the movement on B being shown on A. Set A in 
 motion. If the lengths of the pendulum are the same 
 A will begin to swing, and soon it will take up a violent 
 oscillation. If B be longer or shorter than A, B will 
 still swing, but as the times of their oscillations will be 
 different, the two swings will be eventually in opposite 
 directions, B's swing will then diminish, and gradually 
 cease, but will start anew. 
 
 
 Fig. 2. 
 
 Fig. 2 shows graphically the motion of the pendulum 
 B, it being made on the same principles that Lissagou's 
 figures are made. Pendulum B, with a pen attached to 
 its top, was started by a pendulum corresponding to A, 
 which had a diflferent period, and the amplitude of the 
 former registered itself on paper, moved by clockwork 
 round the axis of suspension. It will be noticed that 
 the amplitude or length of swing increased rapidly at 
 first, and then very gradually attained a maximum. It 
 then diminished very slowly, and then came rather 
 rapidly to a zero, and the pendulum was for an instant 
 at rest. 
 
 Now let us apply this: If instead of the small 
 pendulum we have a molecule oscillating at a distance 
 from another molecule, and find that the latter assumes 
 a more energetic motion, we can but suppose that the 
 
PHOTOGRAPHIC ACTION. 7 
 
 ether has been the means of conveying it. The increased 
 violence of motion may be known by a rise in tempera- 
 ture of the body to which the molecule belongs. But 
 we can also conceive that the vibrations conveyed by 
 the ether are so timed that they may be capable ot 
 increasing the motion of the atoms of a molecule in 
 preference to that of the molecule itself. As already 
 stated, the increase in the extent of the vibrations of 
 the latter may be such that they can swing out of the 
 molecule of which they originally formed a part ; and 
 we may take it that whenever their swing oversteps some 
 fixed amount this must surely happen, more especially if 
 there be some other molecules present which can annex 
 them. The breaking away of an atom and its transfer 
 to some neighbouring molecule is a destruction of the 
 original molecule and the formation of a new one, and 
 this we call photographic action or chemical decomposi- 
 tion caused by radiation. 
 
 In the experiments conducted with the pendulum we 
 have seen that the lighter pendulum need by no means 
 be exactly in tune with the large pendulum for the 
 swing of the former to reach a large maximum 
 value before it begins to diminish. Reasoning from 
 this, we may take it that a vibrator}- motion, though it 
 may not be in exact synchronism with the motion of the 
 atom, may still be sufficient to cause it to swing beyond 
 the sphere of attraction of the molecule of which it 
 forms a part, and so to be the cause of chemical 
 decomposition. We must also bear in mind that 
 there is no reason why the ether should only transmit 
 one set of rhythmic motions or vibrations, any more 
 than that air should do so as in the case of sound. 
 
o LIGHT IN PHOTOGRAPHY. 
 
 The latter will transmit an apparently unlimited number 
 at the same time. Thus in a similar way an unlimited 
 number of series of vibrations in the ether may be 
 transmitted, all starting at the same time from the 
 same point. If a great number of series be started, not 
 only those vibrations which are in absolute tune with 
 those of the atoms, but also those which are not quite 
 in agreement will be effective in causing chemical 
 decomposition to take place. 
 
 Looking at the subject in the light of what we have 
 stated, it can scarcely be conceived that any series of 
 vibrations, which are in accord with those of the molecule 
 as a whole, can be effective in causing a chemical change 
 in its constituents except in extreme cases, but that 
 vibrations which most nearly approach in period those 
 of the atoms may do so. 
 
 Increasing the motion of a molecule only means 
 an increase in the temperature of the body of which 
 it forms a part. Unless the radiation be so powerful 
 that the temperature of the body on which it falls 
 becomes extremely high (as was the case when we 
 considered the decomposition of chloride of silver), 
 thus giving the molecules greater freedom of movement 
 through the liquid or gaseous state being assumed, the 
 vibrations necessary to loose the atoms will be absent. 
 
 The atom being smaller than the molecule — a 
 collection of atoms being naturally smaller than the 
 single atom — it is quite conceivable that the mean period 
 of vibration of the first will be smaller than that of the last. 
 Consequently we may reason that those vibrations which 
 induce chemical action will be, as a rule, of shorter 
 period than those which induce a rise in temperature. 
 
CHAPTER II. 
 
 Lengths of Waves in the Spectrum. 
 
 We must now picture to ourselves that the motion of a 
 vibrating molecule may give rise to an almost infinite 
 series of vibrations. Suppose we have a pea moving 
 with uniform velocity inside a hollow shell with a rough 
 boundary. By some artifice we may imagine that it 
 never stops. It will first dart from one side, to strike 
 the opposite, then glance off perhaps on a short path 
 up, then on a longer path down, striking at some time 
 or another every part of the inside surface of the shell.* 
 It may finally come on its original path and follow the 
 same course as before. The length of the paths will be 
 a measure of the time between each of the collisions 
 that take place. Now if we take a sufficient time during 
 which the collisions are happening, we can sort out, as 
 it were, the paths which are all equal in length, the 
 shorter perhaps first, and then those increasing in length, 
 and so on. For the pea let us substitute a molecule in 
 
 * If it were a smooth shell the paths would be all equal, and 
 would be made in a great circle of the sphere; but it might be 
 arranged that every part of the circumference would be struck 
 within limits. This would only account of a single series of waves. 
 
10 LIGHT IN PHOTOGRAPHY. 
 
 its enclosure. If we consider only the motion in the 
 ether due to one length of path, we should arrive at 
 vibrations of one period, that due to another length of 
 path to another, and so on. The numerous varying 
 lengths of paths could thus give rise to an orderly series 
 of vibrations of varying periods. In some such way we 
 can imagine that the motions of molecules in their 
 confined space give rise to a very large number of 
 series of vibrations in the ether. 
 
 We have so far refrained from using the word " light," 
 for light is really an effect of radiation, and not radiation 
 itself. It happens that certain radiations or rhythmic 
 motions of the ether are of such a character that they 
 affect some matter in the eye, not by causing a rise in 
 its temperature, but probably by a chemical decomposi- 
 tion. The sentient being has thus a knowledge of the 
 existence of these particular radiations through the 
 medium of the eye. The existence of other series may 
 be known by the sensation of warmth or rise in tempera- 
 ture which they produce. For our purpose, however, 
 we may adopt the word " light " as synonymous with 
 radiation, being careful to remember that in its strict 
 sense it is an effect of radiation, and not the thing itself. 
 Bearing this in mind, we shall use either word 
 indifferently. 
 
 Can we sort out these different series of vibrations 
 and present them to ourselves in any recognisable 
 manner.? We can in several ways. Let us take that 
 one which is the most common. When a series of 
 vibrations travel across free space, and come in contact 
 with a transparent body, they will travel through the 
 latter apparently unaffected if it happen that the 
 
RETARDATION OF LIGHT WAVES. II 
 
 surfaces of ingress and egress are parallel to one 
 another, as they are in a pane of glass. If, however, 
 the series fall on a surface inclined at an angle to its 
 path, its character is altered, and the greater the 
 number of vibrations in a given time (which is the 
 same as saying the shorter the length of the vibrations, 
 or waves), the more they are deflected. The ether 
 which is in the transparent body (for it permeates 
 everything) is in a diiferent state ; it is more condensed, 
 to put it in a popular way, and the shorter the waves 
 the slower they must travel through it. Thus long waves 
 will travel more rapidly through it than shorter waves, 
 and when they fall at an angle on a denser transparent 
 body the latter series will be less bent away from the 
 perpendicular to the surface of the dense medium than 
 the former. On emergence they will again take their 
 usual velocities. This might be shown by means of a 
 diagram, which can be found in most text books, and 
 need not be repeated here. These different retardations, 
 and consequent diiferent amounts of bending in a 
 transparent medium, and the subsequent recovery of 
 the speed on quitting it, make it possible, by means of 
 a prism, to sort out the different series of radiations. 
 They spread out in a fan-like form if we only allow 
 a thin slice of ordinary light to fall upon the surface at 
 an angle. The diagram shows what we mean. 
 
 The narrow beam of white light, which can be secured 
 by passing a beam of light through a narrow slit, falls 
 upon the surface A B of the transparent prism, and is 
 split up into the different series of radiations, and emerges 
 from B C as a band of colour, three being most distinct 
 — viz., red, green, and violet— but these three gradually 
 
LIGHT IN PHOTOGRAPHY. 
 
 fade into one another, so that we have a large assort- 
 ment of different colours. The violet ray is most bent, 
 from the perpendicular to the surface B C, and the red 
 
 ..-•^' 
 
 .^^^ 
 
 B 
 
 I 
 
 ^c 
 
 - ft t y 
 
 
 ^ 
 
 
 ^'£-- 3- 
 
 the least, the others being intermediately bent. This 
 tells us that the oscillations which produce Wolet are 
 more rapid than the rest, and those that produce red 
 the least rapid, and as we know, from experiment, that all 
 the component rays of white light travel in space at the 
 same speed, it follows that the wave-length of the violet 
 is the shortest, and of the red the longest. Unless the 
 air be thick with motes, these rays would travel on in 
 space unperceived, but directly we place a white screen 
 in their path they become distinct and visible to the 
 eye ; or if the eye itself be placed in their paths, as 
 many as fall on the small pupil will be recognised as 
 existent. By placing a lens in the path of the rays, and 
 bringing the screen to such a distance that an image of 
 the slit would be in focus with white light, we have 
 successive images of it in varying colours, slightly 
 overlapping one another. If the source of light be the 
 white-hot carbon points of an electric light there will 
 
THE SPECTRUM. 1 3 
 
 be found no break in the succession of colours, and 
 one can scarcely realise that the band is truly an infinite 
 series of images of the slit ; but if instead of this white* 
 hot solid being used as a source, we vaporise in the 
 arc of the same source some body such as lithium, we 
 at once see that such is the case ; for, instead of a 
 continuous band, we shall have a few fine coloured 
 lines placed apparently irregularly along the space pre- 
 viously occupied by the continuous band. Each of these 
 lines is a coloured image of the slit. In this case there 
 are evidently but a few series of waves given out by the 
 vibrating molecules of the metallic vapour, and not an 
 almost infinite number, as in the case of the white solid. 
 This coloured band, into which white light is decomposed, 
 is called the spectrum, and the lines are the spectrum 
 of the material turned into vapour in the arc. As stated 
 before, the path of a molecule in a gas (or vapour) is 
 comparatively free from collisions, but the atoms still 
 vibrate, and it is the less complicated vibration of the 
 atoms which give rise to the coloured lines which are 
 thus, sorted out. 
 
 It is not at all necessary to have an elaborate apparatus 
 to see a spectrum. If we block up a window, and place 
 a slit (say) iVth of an inch in it, and look at the sky 
 through a prism, we shall see the line of white light 
 expanded into the spectrum, showing every colour ; and 
 further, we shall notice a certain number of black lines, 
 parallel to the slit, crossing the rainbow colours, and if 
 a more elaborate apparatus were used we should increase 
 the number manifold. These black lines, known as 
 Fraunhofer lines, occupy fixed positions in the spectrum, 
 and the most conspicuous are useful, amongst other 
 
H 
 
 LIGHT IN PHOTOGRAPHY. 
 
 things, as colour reference points, much in the same 
 way that we use milestones along a road. 
 
 It must not be imagined that the spectrum is con- 
 fined to the length we see with the eyes. The eye takes 
 cognisance of those rays only 
 which can do work on the retina. 
 There are other rays which play 
 an important part in nature, and 
 none more so than those which 
 lie beyond the red. As the red 
 are the least bent of the visible 
 rays, because they are the longest, 
 these dark rays must be still 
 longer if they lie beyond the 
 red. Being longer, the vibrations 
 must take place at longerintervals 
 than the waves which give rise 
 to the visible spectrum, and, as 
 said before, are more likely to 
 aifect the molecules than the 
 atoms of which they are made. 
 They are thus more likely to be 
 efficient in causing a rise in 
 temperature than a chemical 
 decomposition in the matter on 
 which they fall. This is experi- 
 mentally found to be the case, 
 and for this reason they have been 
 mis-named heat rays. No ray 
 can be called a heat ray. It 
 depends entirely on the kind of 
 body on which a ray falls as to ^vhat its action will be, 
 
 f^vf . 
 
 
DARK RAYS. 
 
 IS 
 
 whether as heat or as chemical decomposition. If the 
 vibrations of atoms are nearly in tune with these wave 
 oscillations, chemical decomposition may take place if 
 they be not bound together too strongly. On the other 
 hand, if they synchronise better with the molecule, we may 
 expect a heating effect. In proof of this we can show a 
 photograph taken with the dark rays of the solar spectrum 
 (Fig. 4). All to the right of the line marked A are • 
 impressions due to these dark rays, and they cover a 
 length equal to that of the visible spectrum. Again we 
 have a facsimile of a photograph of the carbon points of 
 the arc light taken through a sheet of ebonite. The 
 ebonite only transmitted dark rays below the red of 
 the spectrum, yet we have an image as well defined as- 
 it would be if taken with visible rays. 
 
 These photographs were taken on a form of silver 
 bromide discovered by the writer, and it is sensitive ta 
 a far larger range of the spectrum than is the ordinary 
 
i6 
 
 LIGHT IN PHOTOGRAPHY. 
 
 silver bromide. These dark rays were first discovered by 
 Sir W. Herschel by their heating effect. He obtained 
 a spectrum from an image of the sun, and placed 
 thermometers in the different parts of it. He found 
 that the heating effect was greater as the red was 
 approached from the blue, and that when he placed one 
 beyond the red it was still higher, and that it continued 
 to rise for some considerable distance beyond the visible 
 limit, till it attained a maximum, when it fell as the 
 distance from red was further increased. It finally 
 reached a distance where the temperature was no higher 
 than the surrounding air. 
 
 Lamansky measured the heating effect of the solar 
 spectrum on lampblack with a thermopile, and found 
 that when graphically represented it exhibited peaks of 
 increased temperature beyond the extreme limit of the 
 visible red. His diagram is shown in Fig. 6. 
 
 Herschel's and Lamansky's curves and the photo- 
 graphs thus not only indicate the existence of these 
 dark rays, but also the two-fold effect they may produce. 
 
PHOSPHORESCENCE. 
 
 17 
 
 We can also show their existence to the eye by an 
 artifice. If we take a phosphorescent plate (a glass 
 painted with Balmain's paint will do), and expose it to 
 •white light till it glows brilliantly, and then place it in 
 a short bright spectrum for some seconds, we shall, on 
 withdrawing it, see that the visible part of the spectrum, 
 from the violet to the yellow, is very brilliant, but that 
 the yellow, red, green, and dark rays below the red have 
 apparently destroyed the phosphorescence, the back- 
 ground being of the intermediate brightness due to the 
 gradual fall in phosphorescence caused by lapse of time 
 when in darkness. 
 
 Fig. 7. 
 
 The band III. shows the rays that excite the phos- 
 phorescence, and these naturally excited the plate when 
 placed in the spectrum, and the bottom band (IV.) shows 
 the destruction of the phosphorescence, the background 
 being shown as grey, to indicate the feeble light which 
 was not extinguished. It will be noted that there is 
 one small band in the dark part of the spectrum which 
 does not destroy the phosphorescence. The bands 
 I. and II. are the visual and photographic spectra of the 
 phosphorescent light. 
 
 We have now to deal with another part of the spectrum, 
 which is also invisible, but lies beyond the violet. And 
 
 c 
 
1 8 LIGHT IN PHOTOGRAPHY. 
 
 here we can call to our aid the beautiful discovery of 
 Sir George Stokes, who found that waves of shorter 
 length than those which are visible could be made 
 apparent by their effect on other matter. Thus, to take 
 the commonest example, if when a spectrum is on the 
 screen we place a strip of white paper on which quinine 
 dissolved in slightly acidulated water has been brushed, 
 light which was before invisible springs out into 
 visibility, and we can see the extent to which the 
 spectrum extends beyond the violet. We may diversify 
 the experiment by placing in the spectrum paper oiled 
 with ordinary machine oil. In this case the light 
 emitted will be yellowish instead of the pale lavender 
 which is so distinctive of quinine. 
 
 Though we are photographers, we must ask ourselves 
 why invisible rays can thus be made visible. At first 
 sight it looks as if the light itself had been altered so 
 as to reduce it to waves of such a length that they 
 could affect the eye ; this is, however, not quite the 
 case. If we remember that the light which caused our 
 phosphorescent plate to glow was not of the same colour 
 as the phosphorescent light itself, we can understand 
 that invisible rays of light may excite a substance to 
 emit rays which are visible. It is not the original light 
 which affects the eye ; it is the light which the invisible 
 rays call forth in the quinine or oil which is active. The 
 incident light shakes up the molecules of these bodies 
 and makes them vibrate, but in a diiferent period to the 
 waves which fall upon them. A substance which acts 
 in this way is called a fluorescent substance, the 
 word being coined from fluor-spar, which exhibits this 
 phenomenon. 
 
FLUORESCENCE. 1 9 
 
 The most ordinary way of showing the existence of 
 these ultra-violet rays is by means of a photograph. If 
 we expose a plate or sensitive paper or film to the action 
 of the spectrum, we at once recognise them as being 
 present by the action they have on the sensitive salt, an 
 action which is in no ways different from that caused by 
 the visible rays which are effective (see Fig. 9, p. 25). 
 
 When the substance of this chapter was communicated 
 as a lecture to the Camera Club, a very pertinent question 
 was asked as to whether the invisible rays beyond the 
 violet, when made visible by fluorescence, were more 
 photographically active than when acting unchanged. 
 This question was answered by throwing in a darkened 
 room a bright spectrum on a card, half of which had 
 been brushed over with quinine. The division between 
 the two parts was made to divide the width of the 
 spectrum, so that in half its width the ultra-violet rays 
 were invisible, and in the other half they were shown as 
 fluorescent rays. The spectrum thus cast was photo- 
 graphed by a camera and lens in the ordinary way, when 
 it was found that the fluorescing rays gave a less dense 
 image than the invisible rays reflected from the card, 
 indicating that no advantage was gained by causing 
 them to become fluorescent. 
 
CHAPTER III. 
 
 Absorption. 
 
 ^\'£ have now to deal with radiation when it traverses 
 a body. What can we suppose should happen when 
 Tadiatio:! traverses a gas, for instance } Let us argue 
 the matter, and then see if our reasoning is correct. 
 Its molecules are free to move in any direction, 
 and their path must be so long that we can scarcely 
 expect that, unless the thickness traversed be great, 
 these should take up much of the radiation transmitted. 
 On the other hand, in a gas which is a compound, the 
 atoms which compose the molecules may be caused to 
 take increased amplitude, or swing, and in this case we 
 should expect that the radiation transmitted would lose 
 something by the transfer of the energy required to give 
 such increased amplitude. Let us put this to the test. 
 We place some flake crystals of iodine, which, though 
 thin, are to all appearance perfectly opaque, in a test- 
 tube. The test-tube we plunge into boiling water, and 
 almost immediately a purple vapour begins to rise, and 
 by degrees fills the whole of it. On allowing the test -tube 
 and its contents to cool, we find dark spangles of solid 
 
ABSORPTION SPECTRA. 21 
 
 iodine on the sides of the glass. We re-vaporise these 
 crystals, and allow the beam of light to pass through the 
 vapour, and this filtered light is now decomposed by a slit 
 and couple of prisms, and the spectrum is examined on 
 the screen. We fin J two phenomena: first, there is slight 
 
 Li 
 
 D Li 
 
 
 ^^B|H 
 
 u 
 
 1 
 
 IH 
 
 I. 
 
 ^^^^p^^^^^ 
 
 lii 
 
 ^^^^^^^^^H 
 
 II. 
 
 
 '' --■■■'■t\-^ '-.-^T ^^^^fe^^^M 
 
 
 1 
 
 ■^^^^^^^t^B 
 
 III. 
 
 ^^E'M' "' 
 
 ^^¥^^^B 
 
 
 
 hhI 
 
 P'.i'oftfg^rap'is of abscrt^ribh t:ikcn on a Cadett spectrum plate. 
 
 /"/jr. s. 
 
 general absorption or cutting-off of the light in the green ; 
 and second, the orange, yellow, and remains of the green 
 are traversed by fine black lines close to one another 
 (III., Fig. 8). The wave-lengths of light represented 
 by these black lines are in tune with the vibrations of the 
 iodine atoms, and are therefore absent after they traverse 
 
2 2 LIGHT IN PHOTOGRAPHY. 
 
 the test-tube, as they have been taken up by the vapour. 
 There are a great many lines appearing rhythmical, and 
 this rhythm forces us to believe that these atoms have 
 rhythmic swings. The slight general absorption in the 
 green is probably due to the greediness with which 
 these vibrations are absorbed, for when the vapour is 
 attenuated, lines appear in this part also, though it is 
 quite possible that the atomic swings not only take up 
 the waves which are exactly in tune with them, but 
 those which are very near!}' in tune as well, exactly as 
 our pendulum behaved ; or it may be also that the paths 
 of the molecules are curtailed, and that it is due to a 
 certain want of freedom of their swings. 
 
 The colour of the iodine is due to the absence of these 
 vibrations, for, by placing a lens behind the spectrum to 
 collect the rays, and to form an image of the prism on 
 the screen, we find the colour of the iodine vapour 
 reproduced. We may repeat this with nitric oxide. We 
 shall find that the pattern of the lines in the different 
 parts of the spectrum is more complicated than with the 
 iodine, as it consists of broad bands as well as of lines 
 (II., Fig. 8). Arguing from first principles, we may 
 gather that, as the nitric oxide is a gas in which more 
 than one atom probably swing together, the complica- 
 tion is to be expected. 
 
 When we dissolve iodine in bisulphide of carbon, or 
 in water in which a little iodide of potassium has been 
 dissolved, we see that the spectrum no longer is split up 
 into lines, but that a large part is blotted out. In the case 
 of the bisulphide solution the green disappears (III., 
 Tig. 8), and in the aqueous solution the blue and the 
 violet also go. Without entering into any discussion as 
 
ABSORPTION IN SOLIDS. 23 
 
 to the reason of the different behaviour of the iodine 
 molecules in the two cases, we will turn our attention 
 simply to the fact that there are bands of absorption, and 
 no lines. In this case we have bands due to the con- 
 finement of the molecules. They can no longer traverse 
 distances measurable by ordinary means, but their motions 
 are restricted, and give rise to waves of periods which are 
 more or less in tune with a great deal of the visible 
 spectrum, and, for aught we can tell with the eye, with 
 some of the invisible portions also. This being so, the 
 molecules of the iodine are set in increased motion 
 by the radiation whose energy they take up, and so 
 only the remainder of the rays pass through them in 
 sufficient quantity to act upon the apparatus in the eye. 
 In those parts of the spectrum whose waves are com- 
 pletely taken up by the molecular motion, of course total 
 darkness ensues. 
 
 If we turn our attention to transparent solids we find 
 the same thing occurs. For instance, yellow glass cuts 
 off all the blue and violet, but allows a great deal of the 
 green to pass, and all, or nearly all, of the yellow and 
 red. The molecules of the 3'ellow colouring matter are 
 in tune with the violet and blue, but not with the green, 
 yellow, and red. The former rays, therefore, are 
 stopped, whilst the latter are allowed to pass, and, when 
 recombined, these latter give the yellow of the glass. 
 And here attention may be called to the fact that the 
 rays which the fluorescing bodies emit have nearly 
 always a greater wave-length than the rays which 
 excite them. If we place a yellow glass in front of the 
 slit of a spectroscope, and examine the spectrum thrown 
 on a screen on which has been brushed sulphate of 
 
24 LIGHT IN PHOTOGRAPHY. 
 
 quinine, we can see no trace of the invisible rays, for 
 they have been stopped by the yellow glass before they 
 reach the screen; If, however, we throw the spectrum 
 on the same screen without the intervention of the 
 yellow glass, but examine it through the yellow glass by 
 simply holding the glass in front of the eye, we find that 
 the spectrum is the same in both cases, as far as the 
 exhibition of the series of colours is concerned, towards 
 the red end of the spectrum, but that it differs at the ultra- 
 violet end. At that part the fluorescence caused by the 
 invisible rays will be seen through yellow glass, showing 
 that these latter contain, at all events, some rays which 
 oelong to the green, yellow, or red of the spectrum. 
 Perhaps no better proof of Stokes's law — which enunciates 
 that the wave-lengths of the rays emitted from the 
 fluorescent bodies are greater than of those which excite 
 the fluorescence— can be given, and it is commended to 
 the reader's notice. 
 
 We may now return to the behaviour in the spectrum 
 of some other solid bodies which are very interesting to 
 the photographer. For instance, we have solid chloride 
 of silver melted into a slab. It looks colourless, and 
 when we make the white light pass through it before the 
 spectrum is formed, we see no appreciable alteration in 
 the range of colours from the ordinary spectrum. But 
 if we place a card over which sulphate of quinine haS' 
 been brushed, in the position where the invisible rays- 
 should flash out, we find that they are altogether absent. 
 We can remove the slab of chloride from before the slit, 
 and they immediately appear. From this we gather that 
 the chloride of silver has taken up the vibrations of the 
 invisible rays. From what we have said before, we can 
 
CHLORIDE AND BROMIDE OF SILVER. 25 
 
 picture that these rays cause the atoms of the chlorine 
 to vibrate with greater vigour than before, and the result 
 should be that photographic action would take place in 
 this region of the spectrum, some of the atoms of 
 chlorine swinging off from the grouped atoms of silver 
 and chlorine. This we can at once prove by placing 
 
 in. 
 
 Fig. 9. 
 
 a plate prepared with ordinary silver chloride in the 
 spectrum, and we find that its action is principally 
 confined to these invisible rays, for it is in that part of 
 the plate on which they fall where the greatest chemical 
 decomposition is found, as shown on development (III., 
 Fig. 9). We shall allude to I., Fig. 9, later. 
 
 Again, we have a slab of fused bromide of silver, and 
 examine it in the same way. To the eye, when examined 
 by white light, it possesses a yellow colour, which is not 
 
•26 LIGHT IN PHOTOGRAPHY. 
 
 Tinlike that of the yellow glass we previously examined, 
 •and the rays it cuts off are very closely the same. Here 
 Ave have, then, the invisible rays, the violet, the blue, 
 and part of the green, cut off or absorbed by this 
 material, and the energy which they carry is taken up 
 by the molecules of bromide, and in this case the swings 
 of the atoms are increased, indicating a possible photo- 
 graphic action. We can test the matter, as before, by 
 placing a plate prepared with bromide of silver in the 
 spectrum, and noticing at what parts photographic action 
 is induced (II., Fig. 8). 
 
 Now we must not run away with the idea that, because 
 action takes place at this region, it therefore follows, 
 if we were to use this for a screen over a light during 
 •developing operations, we should not run any danger of 
 fogging a plate prepared with the same substance. 
 We might use it if all the wave motion were taken 
 up at every part where photographic action takes 
 place. As a matter of fact, there is such a large 
 portion of the rays of the spectrum whose waves are 
 not altogether in tune with the vibrations of the atoms 
 of the bromide molecule, that a good proportion pass 
 through, and would be available to affect a bromide plate 
 exposed to such light. We can lay a slab of fused bromide 
 on a dry plate, and expose to light through it, and we shall 
 find that a certain proportion of those rays which are 
 active in producing the photographic image do pass 
 through. The wiping-out of the wave motion — or, as 
 it is more usually called, the absorption — is not by any 
 means complete in some parts. 
 
 The appearance of the absorption by coloured sub- 
 stances is now familiar to you, but we wish to show 
 
ABSORPTION BY THE INFRA-RED RAYS. 27 
 
 that colourless bodies likewise absorb. By means of 
 the particular form of bromide of silver to which we 
 
 Fig. 10. 
 ihave alreadyreferred, this dan bs demor.strated. In the 
 
2 8 . LIGHT IN PHOTOGRAPHY. 
 
 figure we have copies of the absorptions in the infra-recT 
 of the spectrum of various colourless liquids. Th& 
 character of the absorptions is very different, but it is- 
 beyond our scope to endeavour to attempt to explaia 
 the reasons for the difference. 
 
 It may not be without interest, however, to show^ 
 that by heat measurement and photographs of the 
 absorption of water we gain a knowledge of the 
 condition of the atmosphere of which we might other- 
 wise be uncertain, and that the thermopile measures ai:d 
 photographs confirm one another. 
 
 In Fig. 1 1 ^^•e have in the outside curve a diagram of 
 the heating effect on lampblack, as measured by a 
 thermopile, of the arc electric light. In the inside 
 curve II. of the same shape, we have the diagram of the 
 heating radiation after passing through a glass cell. 
 Curves II., III., IV., and V., show the heating power 
 of the radiation after it has passed through various 
 thicknesses of water. The dotted curves show the 
 same when free from the absorption due to the glass 
 cell. Beneath are representations of the water spectrum 
 and the solar spectrum. The spectra shown below are 
 from photographs. All to the right of the A line are 
 invisible radiations. The scale of the spectrum is the 
 same as that of the base of the thermogram. In this 
 last the height of each part of the curves is a measure 
 of the current produced by the heating of the junctions 
 on the narrow face of the thermopile when placed in 
 the different parts of the spectrum. The current 
 is a measure of the various temperatures produced 
 by the different rays in the lampblack with which 
 the surface of the thermopile was coated. (The: 
 
THERMOGRAMS. 29 
 
 <:urves were illustrations to a paper read before the 
 E.oyal Society by the writer and General Festing.) The 
 scale is given in " turns of the screw." The meaning 
 
 Fig. II. 
 
 ■of this is, that the thermopile was caused to traverse the 
 spectrum by a screw motion, and the distance traversed 
 was therefore conveniently given by a reference to the 
 •screw thread. 
 
 A comparison of this with the next figure, the scale 
 -of which is the Same as Fig. 1 1 , will show that all the 
 dips due to water in the heat curves made with the arc 
 light arc to be found in the dips in the solar spectrum, 
 ■"ivhencc we deduce that the absorption by the water in 
 
30 LIGHT IN PHOTOGRAPHY. 
 
 the atmosphere has a marked influence on the radiation, 
 reaching the earth. The curves in this figure are twO' 
 curves made from thermopile measures with sunlight oa 
 
 Fig. 12. 
 
 separate days by the writer. This figure should also be: 
 compared with (Fig. 6) Lamansky's curve. 
 
 There are certain substances used in photography to, 
 as it is called, orthochromatise plates — that is, to render 
 them sensitive to colours to which they are ordinarily 
 insensitive. Amongst others, we have derivatives of 
 
FLUORESCENCE OF EOSIN. 3 1 
 
 fluorescin, which, as its very name implies, is highly 
 fluorescent. The kind of fluorescence it gives is very 
 beautifully illustrated by throwing a beam of light on a 
 jar of water in which a solution of this substance is 
 allowed to drop. The water appears quite colourless, 
 but where the drops penetrate we have a burst of 
 beautiful greenish-yellow light reflected from them, 
 though looking through them we see that they are of 
 a dark muddy colour. 
 
 Let us examine one of the derivatives of this substance, 
 eosin, which is very generally employed in photography. 
 Placing a cell containing a solution of it in front of the 
 slit of the spectroscope, we find that there is a great gap 
 in the green part of the spectrum ; but if we throw the 
 spectrum upon a piece of paper brushed over with the 
 solution, instead of upon the white screen, we find that 
 at the very part where the absorption takes place in the 
 spectrum when the light is transmitted through the solu- 
 tion, the dyed paper shines with a peculiar olive green 
 lustre. Looking through ruby glass at the spectrum thus 
 cast on the coloured surface, we can again verify the 
 truth of Stokes's law, for we find that the green part of 
 the spectrum is visible (which it would not be without 
 the fluorescence), but shines with a ruddy tint. The ruby 
 glass practically cuts off" all rays except the red and a little 
 bit of the yellow. The fluorescent rays must therefore 
 consist partially of red rays which are of greater wave: 
 length than the green, which excite the fluorescence. 
 
 Now let us consider once more the subject of fluor- 
 escence. If the wave-lengths of the exciting rays were 
 themselves merely altered, they should be reflected 
 from a surface coated with the fluorescent substance. 
 
32 LIGHT IN PHOTOGRAPHY. 
 
 as are the other rays. If, however, they are extinguished, 
 and simply excite new radiation in the fluorescent 
 body, that body practically becomes a source of light, 
 and it is evident they cannot be reflected, but 
 will radiate in all directions. We can put this to a 
 practical test. Let us coat a thick glass plate with 
 collodion to which a certain amount of this dye has 
 been added. We can receive an ordinary spectrum on 
 this film through the glass, and reflect it upwards on to 
 a white screen ; we shall find that we have two spectra 
 reflected — one from the uncoated, and the other from the 
 coated surface of the plate. The first spectrum will 
 contain all the rays, enfeebled, of course, by the fact 
 that only a small proportion can suffer reflection. The 
 second spectrum will be the reflection almost wholly 
 from the colouring matter, since most, if not all, of the 
 ordinary reflection from the glass surface is done away 
 with by its contact with the solid or viscous collodion. 
 This spectrum shows a complete absence of all the 
 green fluorescent rays, showing that these rays are not 
 reflected ; they will be seen, however, shining brightly 
 on the film. It is only those rays which are not fluor- 
 escent which are reflected. This is a very instructive 
 experiment, and worthy of repetition, as the requisite 
 appliances are so very simple. It is also in those very parts 
 •where the fluorescence occurs that the spectrum colours 
 are made sensitive. The rays where this fluorescence is 
 excited are, as a rule, those which have no photographic 
 action ; but when the fluorescent body is in contact with 
 the silver compound, it becomes sensitive to this ray. 
 
 Whilst perusing these pages it will have become 
 devient that we simply regard the ether waves as energy 
 
AMPLITUDES. 33 
 
 carriers, the energy carried being that of the molecules 
 which set the ether in motion. Energy we may look 
 upon as the capacity of doing work. If we have a ship 
 riding near shore at anchor, and waves roll in towards 
 the shore, these last will raise the ship, and do work. As 
 a consequence, we shall find that the height of the 
 waves on the shore side is much diminished ; that is 
 to say, the work done by the waves in raising the vessel 
 reduces their amplitude, or height. Let us see what 
 the amount of work depends upon. Throw a stone in 
 a pond, and we shall see ripples circling out, growing 
 less and less in height, till they are lost in the calm of 
 the main surface. We may take it that, apart from 
 friction of the water, the volume of each circle of wave 
 is the same, for there is no reason why it should change. 
 If such be the case, then the section of the wave will 
 be inversely as the diameter of the circle, and if the 
 wave-lengfth of the wave remains the same, the height 
 of the wave will approximately vary also inversely 
 as the diameter or distance from the centre of dis^ 
 turbance. In a similar way we may argue about waves 
 of radiation, but we have a practical experimental proof 
 of this in another way. If we illuminate a screen with a 
 candle at i ft. distance, then at 2 ft., then at 3 ft., &c., we 
 find that if the illumination be unity at i ft., then at 
 2 ft. it is i, or {iY, at 3 ft. i, or (i)=, and so on ; that 
 is, the energy diminishes inversely as the square of the 
 distance, and we know that the energy is proportional 
 to. the square of the amplitude or extent of swing of the 
 wave. That brings us to the same fact, that the amplitude 
 of a wave diminishes inversely as the distance from the 
 source of radiation. 
 
 D 
 
CHAPTER IV. 
 
 Effect of Small Particles in the Path of Light. 
 
 We must next dwell on another phenomenon connected 
 with radiation when it falls on matter. What we have 
 to consider is the effect on the waves when the matter 
 consists of very small, almost ultra-microscopic particles. 
 We can form some opinion as to what will happen if we 
 again take water waves as an analogy of radiation. 
 
 Suppose there is a very small piece of rock standing 
 above the surface of the watei : waves which are long 
 from crest to crest, however shallow they may be, will 
 pass it, and the existence of such an obstacle will be 
 scarcely felt, except close to the rock. If, however, 
 instead of the waves being long, they are merely ripples, 
 and that the same rock lies in their path, a very 
 different state of things will occur : the ripples will be 
 diverted by the obstacle, a comparative calm being 
 behind it. The new direction taken by the ripples 
 will be partly at right angles to that of the original 
 direction. With waves of intermediate length they will 
 be more or less scattered by the rock. Thus we may take 
 it that the longer the wave, the less it will be deflected. 
 
TURBID MEDIA. 35 
 
 We may apply the same reasoning to waves of light, and 
 may presume that the smaller the wave-length, the 
 greater will be the proportion that can be deflected at 
 right angles to the path. We should thus come to the 
 conclusion that the component rays of a beam of white 
 light, meeting small particles, would be differently 
 affected ; the loss of intensity they would suffer in the 
 direction in which the beam was travelling would vary 
 inversely to some function of the wave-length ; thus the 
 blue would lose more than the green, the green than the 
 yellow, and the yellow than the red. The outcome of this 
 varying loss would be to make white light, after passing 
 through a number of fine particles, yellow or even red, 
 and the light scattered at right angles bluish. This we 
 can prove for ourselves by a very simple experiment. Let 
 us have a glass trough, say six inches in length, and in it 
 pour turbid water prepared by allowing a weak spirituous 
 solution of ordinary negative varnish to fall, drop by 
 drop, into a large jar of water, with plentiful stirring. 
 If we now allow a narrow beam of light fiom the electric 
 lamp to traverse this turbid medium, and receive it on 
 a white screen, we shall see that it has a yellow colour, 
 indicating the truth of our first assertion. We shall also 
 notice that the turbid medium is itself rendered luminous 
 with a bluish light where the beam enters, but that it 
 gradually becomes yellower as the end is approached 
 where the beam makes its exit. This change in colour 
 is a necessity. The first small length of the turbid 
 medium encountered by the beam robs it of some of the 
 violet rays, rather less of the blue, and so on. Con- 
 sequently, when the beam arrives at the next small 
 length of the medium, there is only a smaller portion 
 
36 LIGHT IN PHOTOGRAPHY. 
 
 of the violet and blue, and so on, to be deflected. At 
 the next small length this diminution occurs again, till 
 finally the violet and blue may be so diminished th^t 
 there is but little to deflect, and we then have a yellowish 
 light illuminating the far end of the cell. In a beautiful 
 mathematical investigation on the subject of this scatter- 
 ing of lightfrom small particles,Lord Rayleigh has shown 
 that the co-efficient of the light penetrating is inversely 
 proportional to the fourth power of the wave-length- 
 (I' = le ' * *■"• where I and I' are the original and trans- 
 mitted intensities, k a constant, and x the thickness of 
 the medium, and \ the wave-lengths.) If, therefore, 
 one wave-length is double that of the other, the 
 exponential co-efficient of the latter is sixteen times 
 greater than of the former ; if it were three times that of 
 the . other, it would be eighty-one times. The wave- 
 length of the extreme red is about twice that of the 
 extreme violet, therefore it will be seen that if only 
 -iVth of the red is cut off" by fine particles, the intensity 
 of the violet will be diminished to less than ith of itg 
 original brightness, and for the intermediate colours 
 the intensities transmitted will vary between these two 
 limits. 
 
 Lord Rayleigh found mathematically that the light 
 deflected at right angles to the direction of the. beam 
 should have another remarkable property — viz., that it 
 should be "polarised," as it is called, in a certain 
 direction, a fact which Tyndall had already proved 
 experimentally. If we have a pendulum, O A, 
 oscillating in a circular path, A C B D, it can be 
 conceived as made up of two motions, one along 
 A .B, and the other along C D, which, combined 
 
POLARISED LIGHT, 37 
 
 together, make up the circular motion along A C B D. 
 If this be so in theory, we ought, if we can quench 
 the motion along C D, to get only the motion along A B. 
 In somewhat the same way we can conceive a wave of 
 Hght as it travels in space to have an up and down 
 motion, and also a right and left motion at right angles 
 to the direction it is taking, and if one of these motions 
 
 Fig. 13. 
 
 is annihilated whilst the other is continued, we have 
 the ray of light polarised. The simplest way of 
 showing that this can be eflfected is withj two plates 
 of tourmaline cut out of the same slab. When 
 one is placed in a beam of light, it tells us nothing 
 except that a certain percentage of light is abstracted 
 by absorption ; if we place the second piece over it, 
 the two lengths coinciding, a little more light is cut 
 off by absorption ; if, however, we turn the two at right 
 angles to one another, we find that the beam of light is 
 quenched. Only one side of what we may call the two- 
 sidedness of the waves of light can pass through the 
 crystals placed in one direction, and having quenched 
 orie side by the first tourmaline, we quench the other 
 
3 8 LIGHT IN PHOTOGRAPHY. 
 
 by turning the second crystal at right angles to the 
 first. It is more convenient to use what is known as 
 a Nicol prism to effect this reduction of light to one- 
 sidedness. We need not describe this " Nicol," as it is 
 called for short, but may simply use one for our purpose. 
 Let the beam of light pass through the trough as before, 
 but let it pass first of all through a Nicol, turned in such a 
 direction that the full blue light is seen through the side 
 of the trough (Fig. 14). Now, keeping the eye steady, 
 if we turn the Nicol a quarter round we find that the 
 bluish light has disappeared — that is, the track of the 
 beam of light alone, without its halo, is seen (Fig. 15). 
 
 Fig. 14. 
 
 Turning the Nicol slowly round another quarter turn, 
 we find that the halo gradually returns till it attains its 
 full brightness. By this simple experiment we fully 
 prove that the light is polarised. Figs. 14 and 15 being 
 from photographs, the blue light is alone impressed. 
 
SMALL PARTICLES IN A FLAME. 39, 
 
 and the main track is almost invisible owing to its 
 yellower colour. If we were to look at the trough 
 from a direction not at right angles to the direction of 
 the beam of light passing through it, we should find 
 that the illumination of the trough was not completely, 
 
 Pig- IS- 
 
 but only partially quenched. Lord Rayleigh tells us 
 that this is in accordance with mathematical theory, the 
 polarisation being only complete in a direction at right 
 angles to the direction of the beam. 
 
 Whilst we are discussing the question of the action 
 of fine particles on light, it may be of interest to 
 give a telling proof that the luminous part of an 
 ordinary candle flame is made up of fine incandescent 
 particles. This is usually supposed to be the case, 
 one of the most cogent reasons being that its 
 spectrum exhibits no breaks. But we may further 
 confirm the presumption by the very test we have 
 
+0 LIGHT IN PHOTOGRAPHY. 
 
 used with the turbid medium. We can project a very 
 intense beam of the electric light as a small point on 
 a candle or gas flame, and examine it through a Nicol, 
 from a standpoint at right angles to the beam. We shall 
 be able to trace the path of this comparatively blue light 
 when we look at it with the Nicol turned in one 
 direction, but it will disappear when it is turned a 
 quarter turn. This is in itself evidence of the presence 
 
 of fine particles in the flame. Figs. i6 and 17 show 
 photographs taken of a candle flame and of a gas jet, with 
 the Nicol turned in the two directions. As the light 
 from a flame is polarised in every direction, it is 
 probable that the light given from, say, a batswing gas 
 burner, will be found to be more photographic in one 
 direction than in the other, though the diflference must 
 be small. 
 To test the quality or colour of the light which we 
 
SUNSET COLOURS. 4' 
 
 get after its passage through increasing turbidity, we can 
 make a very pretty experiment which anyone can repeat 
 for himself. We throw on to a screen an image of a 
 small hole placed in front of a cell containing hypo- 
 sulphite of soda, and through which a concentrated beam 
 of the electric light passes. The round patch of light we 
 may suppose to be an image of the sun. Into the cell of 
 hyposulphite we may now drop a small quantity of a 
 
 Fig. 17. 
 
 dilute solution of hydrochloric acid in water. The 
 effect of this acid solution is to gradually decompose 
 the hyposulphite with a liberation of very fine particles 
 of sulphur. The particles are few in number at first, 
 and so scarcely alter the colour of the light. As they 
 increase in number the light gets scattered, as already 
 shown, and we have a yellowing of the image, and 
 when the decomposition of the hyposulphite is still 
 further complete, the image becomes cherry-coloured, 
 
42 LIGHT IN PHOTOGRAPHY. 
 
 and finally dusky red. Our mock sun has passed 
 through every phase of colour which the real sun 
 appears to take, from the mid-day to sunset, and if fine 
 particles are suspended in our atmosphere, we need not 
 seek any further cause for these changes in colour with 
 which we are so familiar. We shall learn more of the 
 cause of these alterations in colour if we throw the 
 image of this mock sun upon the slit of the spectro- 
 scope, and watch the alteration in the spectrum 
 that takes place. We shall find that the violet and 
 blue fade away most rapidly, then the green, till 
 finally, at sunset, when the sun is really red, only a 
 red band of the spectrum is left. Now we can apply 
 this to photography. Towards sunset a landscape may 
 be brilliantly illuminated with sunlight, more or less 
 ruddy or orange, and yet, on taking a photograph of 
 the same, we shall find that there is not a trace of sun- 
 light effect on the negative. The brilliantly illuminated 
 high-lights have had no effect at all. The reason of this 
 will now be apparent. The molecules composing the 
 sensitive substance of a photographic plate, as we have 
 found, respond chiefly to the violet, blue, and green 
 of the spectrum, and these very rays are nearly totally 
 absent in the light from such an evening sun. The 
 effective illumination for the photograph under these 
 circumstances is practically due to light from the sky 
 which is reflected on to the landscape. 
 
 Perhaps we have rather delayed what should 
 have been said earlier. The scattered rays from a 
 beam of sunlight must go somewhere. They are 
 scattered in all directions upwards and downwards, and 
 they reach other particles, and come to the eye as blue 
 
POLARISATION OF THE SKY. 4J 
 
 light, which we call sky. We now see why we have a 
 sky at all. When looking through a Nicol prism at the 
 sky at places at right angles to the direction of the 
 beams which reach us from the sun, we find that at 
 those parts the light is most polarised, indicating that 
 a portion of the light from the sky is due to light 
 scattered by fine particles. We could scarcely expect it 
 to be entirely polarised, as not only are rays which are 
 incompletely polarised reflected, but there are coarser 
 particles which scatter all colours alike, so that we have 
 a mixture of polarised and unpolarised light. 
 
 The writer has beside him a small instrument with 
 which he has been able to measure very accurately the 
 polarisation of light from the sky in any direction. 
 It consists of a couple of square tubes placed side by 
 side ; at one end of each tube is a lens so arranged 
 that they each throw an image on a screen inclined 
 at 45 " to their axis. These images can be observed 
 through an eye -piece.. In front of both these 
 lenses are aluminium shutters, one of which can 
 be closed to any desired amount by slow motion 
 screws. Behind the other is a Nicol prism, which can 
 be rotated on its axis. An arrangement is adopted by 
 which the axis of the lens can be accurately pointed to 
 the sky at right angles to the direction of the sun. 
 When the Nicol is turned in one direction, the shutter 
 in front of the other lens is closed until the image of 
 the same piece of sky appears of the same brightness. 
 The Nicol is then turned at right angles to its former 
 position, when equality of brightness is again estab- 
 lished. The ratio between the two apertures employed 
 gives the ratio of light polarised to that unpolarised. 
 
44 LIGHT IN PHOTOGRAPHY. 
 
 If we look at a white cloud through this apparatus 
 we find that very little, if any, change in the brightness 
 of the image occurs, in whatever direction the Nicol 
 is turned; whereas the neighbouring sky, if it is at 
 all near the plane at right angles to the direction 
 of the sun, appears dark when the Nicol is turned 
 in one direction. We thus see that for rendering 
 cloud and sky the use of a Nicol prism, combined with 
 a lens as advocated by Messrs. Dibdin and Wellington, 
 must be advantageous. 
 
 Now do not let us have the idea that the sky is only 
 above us. It is around us, and we see it in that blue 
 haze of the distance which almost entirely eliminates 
 local colouring. Distant mountains appear blue, and 
 differing but little from the sky near them. They 
 appear slightly darker, with perhaps just a faint trace 
 of local colour to differentiate them from it. Who is 
 there that has tried to photograph such distant hills, 
 and has succeeded in rendering them as he wished ? 
 We prophesy that, in all probability, not one in a 
 hundred, if he used the ordinary photographic 
 apparatus. Let a Nicol prism be used, however, 
 combined with the lens, and we shall see that success 
 is by no means impossible, or perhaps we should say, 
 impracticable. The Nicol can be turned till the blue 
 sky is cut off, and the local colouring will appear, and 
 be "photographable." A yellow glass may also be used, 
 for the same purpose, but not so successfully. The sky 
 between us and the distance is caused by the presence 
 of the innumerable small particles which exist in 
 the atmosphere, and they are most numerous near 
 the earth's surface. As a result of several years of 
 
PARTICLES IN THE ATMOSPHERE. 45 
 
 careful measurement, the writer has found that the 
 number of particles varies at any altitude as the height 
 of the barometer. If the barometer stands at any place, 
 say at twenty-five inches, and at another at thirty 
 inches, the number of particles in a thin layer at these 
 two places are as 25 to 30. A few miles along the 
 earth's surface are thus equal to a good many atmo- 
 spheres in thickness. In fact, at the horizon a horizontal 
 beam of light has to traverse what is equivalent to 
 about thirtyfive thicknesses of atmosphere, so far as 
 the number of particles is concerned. 
 
CHAPTER V. 
 
 Light in the Dakk-Room. 
 
 One reverts to the question of light in the dark-room 
 with very mingled feelings, for the suggestion naturally 
 forces itself upon one that everything that can be 
 possibly said about it has been written, and the author 
 acknowledges that he has been a considerable oilender 
 on the subject from time to time. It is a subject, 
 however, of such importance to the photographer that 
 in articles on " Light in Photography" it must not be 
 passed over. It has been shown in a previous chapter 
 the extent in the spectrum to which ordinary bromide 
 of silver is sensitive, and reasons have further been 
 given why the light coming through bromide of silver 
 would not be safe to use in developing. Fused 
 bromide of silver, as stated, possesses a yellow colour, 
 whilst in photographic plates it appears to vary 
 between red and blue, and in the case of that specimen 
 sensitive to the infra-red, to sap green. The question 
 as to whether these are really the colours of the 
 bromide is a difficult one to answer. When bromide is 
 precipitated as an emulsion the particles are evidently 
 
COLOUR OF AN EMULSION. 47 
 
 built up by innumerable infinitesimally smaller particles, 
 and these, though brought together and made into the 
 large particles, are yet separated, and light struggling 
 through them must be subjected very largely to the 
 same scattering that takes place when sunlight traverses 
 the atmosphere. When the particles are first precipitated 
 they would have what may be called a certain looseness 
 about them. In such a case we should expect white 
 light to appear ruddy after passing through them, the 
 blue and green rays being most scattered at right 
 angles to the beam of light. Now those who have 
 watched the effect on light transmitted through a 
 turbid medium will have seen that the liquid becomes 
 opalescent throughout, and that the more fine particles 
 there are, the more pronounced does this blue 
 opalescence become. Finally, when the turbid medium 
 becomes so dense that a strong beam of white light 
 fails to penetrate it, or to penetrate it very slightly, the 
 opalescence becomes whiter. At this last phase of the 
 phenomenon no doubt the small particles become very 
 thickly intermingled one with the other, and they 
 form larger aggregations through the smaller ones 
 coalescing, but they may still be separate one from the 
 other. The light which, after issuing through the film, 
 would be ruddy under ordinary circumstances, is cut off 
 and reflected internally by these larger aggregations, 
 and mix with the scattered light. By repeated reflec- 
 tions it may get absorbed to a large extent, so that the 
 colour of the film may appear to be green, or even 
 blue, owing to their absence. We can take a plate 
 coated with gelatine in which some alkaline sulphate 
 has been dissolved, and when dry immerse it in, say, 
 
4? LIGHT IN PHOTOGRAPHY. 
 
 chloride of barium. The film may be so dense in the 
 moist state that white light fails to penetrate it as a 
 beam, but it appears white ; on the other hand, when 
 dry, a gaslight examined through it will appear ruddy 
 or orange. These two states represent the same 
 amount of suspended matter in the gelatine, but they 
 are not suspended under the same conditions. The 
 bearing of this experiment will be apparent. Again, 
 when bromide of silver is heated in gelatine it passes 
 from the red stage to the blue stage, but in either 
 stage, if the film be thin, and the plate be laid on a 
 piece of white paper with writing on it, no trace of the 
 red or blue colour will be visible, showing that we 
 have not the true colour of the bromide in the ligl^ 
 transmitted through it. The colour of a bromide film 
 under these conditions must not, therefore, be assumed 
 as showing any molecular change of the bromide itself, 
 but rather as a change in the structure or aggregation 
 of the particles. 
 
 But what has this to do with light admitted to the 
 dark-room ? A good deal. It must not be assumed 
 that because we have a difference in colour of the two 
 films when examined by transmitted light, that there- 
 fore they have different spectrum sensitiveness. That 
 there is a difference in sensitiveness to the spectrum 
 cannot be gainsaid, but it is probably due to the fact 
 that, as the general sensitiveness io white light is increased, 
 so the range of the spectrum which is capable of 
 increasing the amplitude of the atom's motions to the 
 extent necessary to break up the molecule is larger than 
 it was before. This readily accounts for the fact that 
 whilst only some plates are sensitive as far as the 
 
ILLUMINATION AND RED AND YELLOW GLASS. 49 
 
 blue-green, others prepared with the same silver salt 
 are sensitive as far as the yellow. To judge of the 
 suitability of a light in which to manipulate any plates, 
 it is safe to reckon with the spectrum sensitiveness of 
 those whose range goes furthest towards the red. We 
 have, however, another consideration to bear in mind, 
 and that is the physiological peculiarity of the human eye. 
 The eye is less sensitive to feeble red light than to any 
 other. Red light does not "carry" far. A red light may 
 properly illuminate a plate close to it, but remove it a 
 couple of feet, and almost all the image on a developing 
 plate will become indistinct. Orange light, on the 
 other hand, has much more "carrying" power, and a 
 light of this colour, and of an equal luminosity or bright- 
 ness as a red light, will enable a plate to be examined 
 at a distance considerably further than can be done 
 by the last-named light. When we come to a green 
 light we find this still more true. Evidently, then, 
 the light that most nearly approaches a green light is 
 far better than any other. A green glass, however, 
 as a rule, admits a large amount of blue with it, and 
 hence we are bound to fall back on a yellow or orange 
 glass, as the light transmitted through it allows but 
 little, if any, blue to be carried with it. 
 
 Now we may go a step further, and show to what 
 extent we may overstep theoretical perfection in 
 practice. If the most sensitive plate in the market 
 at the present time be exposed to the light of a 
 candle at forty feet distance for ten seconds, no action 
 of such light can be traced on the plate. If a yellow 
 glass be placed in front of the candle, this time may 
 be increased at least twenty-five times, or the ten 
 
 E 
 
so LIGHT IN PHOTOGRAPHY. 
 
 seconds, or be increased to four minutes and more. 
 It is quite evident, then, that we may place a candle 
 shaded with this screen at eight feet from a plate 
 without any ill-effect resulting from its employment 
 during ten seconds at least. The best practical way, 
 then, to test what light may be used, is to place different 
 pieces of glass, paper, or other material, which it is 
 proposed to employ, in contact with a plate in a dark 
 slide, and expose the plate through these coloured media 
 to the white light, which would come in through the 
 window, for a time which should allow developing 
 action to commence. Probably half a minute would 
 be ample. All media should be rejected which showed 
 any but a very slight action on developing the plate. 
 That medium amongst those selected which gave 
 evidence of passing the most yellow light should be 
 accepted on physiological grounds. This is a practical 
 way of carrying out a test. One of the safest media is a 
 solution of iodine in alcohol, or iodine dissolved in an 
 aqueous solution of iodide of potassium. When con- 
 centrated, it cuts off all colour except the extreme red, 
 whilst when more dilute it allows orange as well to be 
 transmitted. A liquid screen is somewhat awkward to 
 deal with, but it can be managed with a little care. We 
 may, however, for ordinary plates which are non-ortho- 
 chromatic, use with safety a coloured screen which 
 consists of stained red, and an orange paper to diffuse 
 the light entering into the room. We cannot do better 
 than reproduce a diagram (Fig. i8), which is to be 
 found in " Instruction in Photography," to show what 
 are the absorptions of different media. Fig. 19 will 
 give an idea of the loss of illumination caused by using 
 stained orange and ruby glass. The outside curve 
 
ABSORPTION SPECTRA 
 
 s» 
 
 shows the relative brightness of each spectrum ray in 
 good white light. The two inner curves show the 
 
 amount of the different rays that are cut oflf. A feeble 
 white light outside the window glazed with either of 
 
52 
 
 LIGHT IN PHOTOGRAPHY. 
 
 these media means almost darkness within the room. 
 If we could find a medium which would admit all the 
 rays to the light of the scale No. 4Si» we should have a 
 luxurious light at all times. 
 
 Changing plates is often a difficulty to some amateurs 
 who are on tour if they have forgotten their lantern. 
 There need be none at night. If we recollect that a 
 candle when at forty feet off from a sensitive plate has no 
 effect in ten seconds, the way to utilize such a light at 
 night is easy. Place the candle beneath a table in a room, 
 
 Iiig. 19. 
 
 and change the plates on the top of the table. This can 
 be effected in a few seconds, and no evil effect need be 
 feared, particularly if the walls are of a yellowish colour. 
 
CHAPTER VI. 
 
 Light Scattered by Large Particles. 
 
 In Chapter IV. we have called attention to the 
 scattering of light by fine particles, and we shall now 
 devote ourselves to the consideration of the scattering 
 of light by larger particles — particles which are not 
 ultra-microscopic, but which can be seen. It has 
 already been stated that we may have the scattered 
 light from both kinds of particles existing together, as is 
 found in the sky, for example. There are some effects in 
 photography which are entirely due to the light falling 
 on these coarser particles. If a vertical beam of light 
 falls on a bright spherical ball placed on a sheet of 
 paper, the rays will be reflected differently from each 
 part of the surface on which it falls ; the most intense 
 ring of light which reaches the paper will be that just 
 grazing it ; the light reflected from the other circles of 
 the sphere will form less bright rings, since the reflected 
 beam will reach the flat surface at an angle, and thus 
 the quantity of light reflected from such circle will be 
 spread over a larger surface. 
 
 For a spherical ball let us substitute a particle of 
 sensitive salt, and for the white paper a sheet of glass. 
 The rays specularly reflected from the small particle 
 
54 LIGHT IN PHOTOGRAPHY, 
 
 will reach the first surface of the glass plate, and instead 
 of all being again reflected, will only be partially 
 reflected, as some will pass through the glass and reach 
 the bottom surface of the plate. When they reach this 
 polished surface they may all be reflected back, or some 
 may again pass through, and only a portion be reflected. 
 The case where the reflection from the bottom surface of 
 the plate is complete is that which is peculiarlyinteresting. 
 
 Common optics tell us that when a beam of light is 
 incident on a surface which divides a light from a 
 dense medium, no part of it will pass out of the latter 
 if it fall on such surface at a certain angle, and 
 that for this, and for all greater angles, the beam will be 
 wholly reflected. The angle at which this total reflection 
 takes place is called the critical angle, and is dependent 
 on the indices of refraction of the two media. 
 
 When a beam of light in air falls at an angle 9 
 (measured from the vertical) on the surface bounding 
 two media, the sine of that angle multiplied by a constant 
 number gives the sine of the angle which the beam will 
 make with the vertical when it enters the denser medium 
 (if we call the last angle g', we express it mathematically 
 as li sin s' ^ sin e, i and n being the constants or indices 
 of refraction, as they are called). If the beam of light falls 
 in a dense medium into air, the reverse takes place : the 
 angle at which it emerges is greater than that at which 
 it strikes the surface, and the sine of the angle of the 
 former has to be multiplied by the same constant. In 
 the case of glass, when the index of refraction of air 
 is taken as unity, n is about i"s. If, then, the angle 
 at which a beam in the glass strikes the surface is 
 rather less than 42*", the sine of the angle which the 
 
HALATION. SS 
 
 beam would make would be i, or 90°, or it would graze 
 the surface, and a beam striking at any angle greater 
 than that (since the sine of an angle can never be 
 greater than i) would fail to get through at all, but 
 would be wholly reflected back, and at the same angle 
 as that at which it struck the surface. 
 
 Now we may apply this to the light scattered by our 
 particle. Of the scattered rays which enter the glass those 
 which struck the bottom surface at a greater angle than 
 about 42 " would be reflected back tp the first surface, where 
 they would, if it were a bare surface, be again reflected 
 back to the second surface, and so on. In a photographic 
 plate, however, the first surface is not bare, but is 
 covered with gelatine or collodion containing sensitive 
 salts. The fact that the density of the gelatine is 
 comparable with that of the glass allows the first 
 reflected beam from the back surface to pass into the 
 gelatine, and to strike the particles in its path. If, then, 
 we have a small beam of light (such as that which is 
 obtained through a pinhole in a piece of foil placed in 
 contact with film) striking a few particles, we should 
 expect to find on development not only an image of 
 the hole, but also a ring having a radius of nearly twice 
 the thickness of the plate (due to the total reflection 
 of the light at the critical angle) surrounding it. 
 Outside this ring the density would shade off gradually, 
 as the totally reflected light would be spread over a 
 larger area, but inside it the shading towards the hole 
 would be much more abrupt, as a large portion of 
 the light would pass through the plate, the remainder 
 only being reflected back. We have here an explanation 
 of " halation." 
 
S6 
 
 LIGHT IN PHOTOGRAPHY. 
 
 An experiment proves the worth of this reasoning in a 
 very convincing way. Punch out a small hole in the foil, 
 and at a small distance from it cut a line, and place the 
 foil in contact with a plate, and expose. No. i , of Fig. 20, 
 
 No. I. 
 
 No. 2. 
 
 No. 3. 
 
 No. 4. 
 
 Fig. 20. 
 
 gives the result. On development the image of the hole 
 is surrounded by a ring, as described, and the deposit 
 round the line can readily be seen to be made up of a 
 series of rings due to each small part of the line. 
 If the glass had been thicker, the ring round the dot 
 would have been larger, as would have been the boundary 
 of the halation round the line. If the glass were thin, 
 or if a celloidin film had been employed, it is evident 
 that the ring would be practically coincident with the 
 point, and the line would show no practical extension of 
 breadth. This accounts for the fact that with plates 
 
BACKING A PLATE. 57 
 
 we have halation, whilst with films it is practically 
 absent. 
 
 It is rather an instructive experiment to place the foil 
 in contact with one plate, and behind the first plate to 
 place a second, separated from it by the thickness of a 
 sheet of paper, so as to prevent contact between the 
 two. The top plate will show by its rings the light 
 reflected from its back surface, and the second the 
 amount of light which passes through the. first. Of 
 course, by very prolonged exposure, a ring of halation 
 may be traced in the second plate. No. 2 in Fig. 20 
 shows what is obtained, and if " applied " to No. i would 
 fill up the darker spaces within the lighter boundaries. 
 
 We now come to another phase of the question. 
 Suppose we coat the back of the plate with some 
 substance such as asphaltum whose index of refraction 
 is not far off that of the glass. All, or very nearly 
 all, rays which come into the glass plate from the film 
 side will penetrate into the asphaltum, and will not be 
 reflected from the back of the glass, but would have to be 
 reflected from the back of the asphaltum. But in passing 
 through the asphaltum they are absorbed, and there are 
 practically none which find their way to the surface ; 
 hence no halation should take place. No. 4, Fig. zo, is 
 an example of this ; the halation is completely absent. 
 With a very strong light a very small halation would be 
 visible, owing to the fact that the asphaltum is not quite 
 so dense as the glass, and a very small percentage of 
 light would be reflected from the surface bounding the 
 two. If the back of the plate be coated with a film of 
 collodion, halation will be found of a similar strength 
 as if it were bare — the diameter of the circle will be 
 
S8 LIGHT IN PHOTOGRAPHY. 
 
 very slightly different; but if the collodion be mixed 
 with an orange dye, the halation will be absent with 
 ordinary plates, though with an orthochromatic plate 
 it will be plainly visible. Halation may be mitigated 
 by brushing over the plate gum mixed with a pigment, 
 such as sienna or chrome yellow, but the writer has not 
 found that it is so effective as if the colouring matter 
 were soluble. The light can struggle to the back of 
 the layer, and be reflected back, a certain portion 
 reaching the glass (see No. 3, Fig. 20). 
 
 In backing plates, the great point to attend to is that 
 the two surfaces are in absolute optical contact — that is, 
 that no film of air is between them at any part, and that 
 the material absorbs the photographically active rays. 
 Carbon tissue, pitch plaster, and other adhesive 
 materials will be found effective for this reason. 
 
 The absence of halation iihen a sensitive surface is 
 used for the measurement of " photographic" light is a 
 point that should be attended to, othenvise a slightly 
 erroneous result may be anticipated. 
 
 If, instead of studying halation by means of a photo- 
 graph, it be desired to study it with the eye, there is 
 nothing simpler than to place the tinfoil in contact 
 with the film on a plate, and, having illuminated it with 
 a strong light, to examine the plate from the back. 
 The ring and the shadings of light will be distinctly 
 seen. If the back of the plate be cemented by Canada 
 balsam, or even by a film of water, to a thick plate, the 
 same phenomena will be observed, but the rings will be 
 of greater diameter, since it is dependent on the 
 thickness of the glass, and the two plates become one 
 to all intents and purposes. 
 
CHAPTER VII. 
 
 Interference and Diffraction. 
 
 We are now leaving the effects of light, with which 
 photographers usually have to deal, and are entering on 
 a somewhat more difficult part of the subject, but as it 
 leads to the explanation of the formation of spectra by 
 diffraction, and of pinhole images, we shall not scruple to 
 discuss it. We have all along treated light as being due 
 to wave motion, and in the explanation of inter- 
 ference and diffraction this treatment is absolutely 
 necessary. We may once more go to a water wave as 
 an example of wave motion, since the waves which 
 cause the sensation of light are supposed to be transverse 
 waves with vibrations at right angles to the direction of 
 the beam. If two stones be thrown in a pond at the 
 same instant and far apart, the two separate sets of 
 waves will move in circles, and at some point these 
 circles will intersect. At that particular part where the 
 crest of one wavelet meets the crest of another, the 
 water will be raised higher than before, and where 
 the trough of one wave meets the trough of the other 
 
6o LIGHT IN PHOTOGRAPHY. 
 
 the water will be more depressed, but where the 
 trough of one wave meets the crest of the other the 
 water will remain at the ordinary level, and stillness 
 result. Yet, though the waves may disappear at the 
 spot where this stillness is seen, the wave motion is 
 only masked, for each wave will reappear and go on its 
 way as before. It must not for a moment be conceived 
 that the wave carries with it the water of which it was first 
 formed. There are successive motions of each part of the 
 water. A cork on the surface, for instance, is not 
 carried on with a wave ; it simply rises and falls as the 
 successive waves reach it and pass away from it. Waves 
 merely carry the energj' which is generated by the 
 fall of the stones. Light waves act in somewhat the 
 same way, but with a small diiFerence. With these the 
 two sets of waves of the same length must start from the 
 
 Fig. 21. 
 
 same point, but by different routes. When one arrives 
 at some point exactly a wave-length behind the other, 
 the effect will be to increase the distance through which 
 the wave oscillates, or, as we have called it before, the 
 amplitude. If the original amplitudes are equal, the 
 
COMBINATION OF WAVES 6 1 
 
 amplitude of the compound wave will be doubled. 
 Fig. 2 1 shows this : a a a and 6 b b are two sets of waves 
 starting from the same point. Separately they have the 
 amplitudes shown by the continuous lines, but when 
 compounded they have the double amplitude shown by 
 the dotted line. Suppose, however, one set of waves 
 arrives only slightly behind the other, as in Fig. 22, the 
 two will be compounded in the following way : Take 
 any point of the path X^. At this point one wave has 
 
 Ftg. 22. 
 
 a positive amplitude X' A, and the other a negative one 
 X' B. The difference between the two is X C, and 
 this will be a point in the motion of the combined 
 motions. By this plan of combination the wave motion 
 will assume the form shown in the dotted line c c c c. 
 
 If the one set of waves was exactly half a wave-length 
 behind the other, the positive (above the line X X) 
 motion and the negative motion (below X X) would 
 exactly neutralise one another, a:nd no motion would be 
 seen along X X; one wave amplitude would destroy 
 the other, and there would be no light. Such waves are 
 said to interfere with one another. Now, suppose two 
 series of waves of the same length start from the same 
 source ; these sets will be in the same phase — that is, their 
 
62 LIGHT IN PHOTOGRAPHY. 
 
 crests and hollows will occur at the same instant — when 
 they have travelled through the same distance in the same 
 medium. Now it is perfectly possible that both sets of 
 waves may be compelled by circumstances to arrive at 
 the same point on a screen, but by different routes. If 
 it happen that at this point the difference in length 
 of the paths makes one series of waves arrive exactly 
 half a wave-length behind the other, then at this point 
 there will be interference, and darkness will result. For 
 instance, a point of light will radiate waves which will be 
 spherical — that is, all the crests and hollows will be in 
 concentric spheres, and if it be far enough off from an 
 obstacle lying in the path of the waves we may consider 
 that the surface of the sphere close to it is a plane. Now a 
 wave at any part of its path may be considered to be made 
 up of a large number of very small waves, each spreading 
 out in a hemispherical form, and this we may take to 
 be the case where the wave meets the obstacle. This 
 being granted, we may reason that all the small waves 
 that pass by the obstacle will spread out in every 
 direction, and that those which graze the obstacle, or 
 are near to it, will spread round the edge of the obstacle, 
 and instead of a sharp shadow resulting, as there would 
 be if there were no waves having a tranverse motion, we 
 ought to have light shown on a screen within the 
 geometrical shadow of the obstacle. Experiment shows 
 this is absolutely true. When we have light issuing 
 from a small point, such as from the image of the sun 
 formed by a lens of short focus passing over an edge 
 of an opaque object, the edge is not defined on the 
 screen, but is shaded off. But we also find that there 
 are alternate bands of partial darkness and light on the 
 
FRINGES. 
 
 63 
 
 bright part 0/ the screen near where the shadow of the 
 edge falls. These alternations of light and darkness 
 are called fringes. The darker bands are due to the 
 fact that at these parts of the screen the wavelets near 
 the edge are exactly half a wave-length behind the 
 waves which are proceeding direct to these points, and 
 
 Fig- 23- 
 
 this causes the partial darkening on the screen. As 
 many wavelets which pass unimpeded to the screen 
 combine together to give the illumination, and but a 
 few near the edge of the obstacle are half a wave- 
 length behind, it follows that the interference can be 
 by no means complete, but only partial. A little 
 reflection will show that there will be an unlimited 
 number of these darkened bands, but that as they recede 
 from the edge they will become fainter and fainter, till 
 they finally are lost to the cognisance of the eye, for 
 the further away they are from the geometrical shadow 
 the less will be the reduction of the amplitude. The 
 
64 LIGHT IN PHOTOGRAPHY. 
 
 light bands are accounted for in exactly the same way. 
 On the shadow side of the edge the light of the wavelets 
 spreads outwards, and each contributes something to 
 the illumination ; but, as will be seen, the effect of the 
 combination must rapidly die out. Fig. 23 shows the 
 fringes due to an obstacle. The obstacle is a thick 
 wire crossed at right angles with a thinner wire. The 
 light lines shown in the centre of the wires are due to 
 the "interference" (a turn which will be explained 
 directly) of the light passing from each edge into the 
 geometrical shadow. 
 
 We now come to a more interesting phenomenon 
 based on the same principles, and that is the diffraction 
 of light coming from a point after it has passed through a 
 slit. Let us consider we are dealing with monochromatic 
 
 Fig. 24. 
 
 light. The light reaches the slit S S in the same phase 
 from a distant point of light, and the question comes, 
 what will be seen on a screen, A B, placed some little 
 distance from it } We may consider what will happen at 
 
LIGHT COMING THROUGH A SLIT. 65 
 
 a point P on the screen A B. The light arrives at every 
 part of the slit S S in the same phase — that is, the 
 crests and hollows arrive severally at the same time at 
 S S. We can divide S S into small strips, so that the 
 light travelling to P from each subdivision is just half a 
 wave-length behind that travelling from the next one to 
 it. If the number of these divisions be an even number, 
 it is manifest that each of these small wavelets will 
 interfere with one another, and there will be darkness 
 at P. On the other hand, if it be an odd number, 
 the point P will be illuminated. At other points in 
 the screen we shall find darkness or brightness from 
 the same cause. We shall thus have stripes of light 
 divided by darkness, since we are using monochromatic 
 light. If, however, we were using white light, we 
 should have no bands of absolute darkness, but the 
 screen would show bands of different colours, for the 
 positions of the dark bands depend upon the wave- 
 length of the light employed, and as in white light the 
 wave-lengths are numerous, evidently the light band 
 due to some one wave-length will overlap the dark band 
 due to some other, and so we get the coloured bands. 
 When monochromatic light is used, the distance apart of 
 the dark bands evidently depends on the width of the slit. 
 For let us suppose the breadth of the slit can be divided 
 into strips giving six half wave-lengths of difference on 
 arriving at P ; then P will be in darkness. If we halve 
 the slit there will be only three such strips, and being 
 an odd number it will be in brightness. By calculating 
 in this way we shall find that the narrower the slit the 
 further apart are the bright and dark bands. There 
 is, of course, a central image as well as these bright 
 
 F 
 
^^ LIGHT IN PHOTOGRAPHY. 
 
 and dark bands outside it, and if the screen be 
 sufficiently far oflf it will be a uniformly bright stripe, 
 but if it be so close to the point of light that the fringes 
 . formed in the way we have already mentioned, by each 
 edge falling within the image, it will be traversed by dark 
 lines. 
 
 This reproduction of a photograph shows what will be 
 seen when a fairly wide slit is employed. The photo- 
 graph has been enlarged in order that the alternate 
 bands may be seen more readily. In this case the 
 screen was placed at such a distance that the effect of 
 the fringes is absent, as they lie outside the geometrical 
 image. When we have a large number of equal apertures 
 side by side, and separated by equal opaque intervals, 
 and allow a beam of light coming from a distant point 
 to fall upon them, each separate aperture will have its 
 own bands of light thrown on a screen, and they produce 
 a confused appearance where they are cast; but if a 
 lens be placed in their path, then they can be 
 collected and brought to a focus in the ordinary 
 way, and they form fairly bright lines at regular 
 intervals ; bui this depends on the intervals being regular. 
 (We are, of course, talking of monochromatic light.) 
 
IRREGULAR RULING. 
 
 67 
 
 The figure is a photograph of the bands taken from a 
 surface with rather irregularly spaced apertures. Lines 
 were ruled with a point on a glass plate coated with 
 asphaltum dissolved in benzine, to which a little olive 
 oil was added to keep the surface from 
 being brittle whilst it was being ruled. 
 The lines were about iVth of an inch 
 apart. This ruling was reduced by photo- 
 graphy, so that the photograph contained 
 about 300 spaces to the inch. Instead 
 of a point of light, a slit parallel to the 
 apertures was employed, through which 
 monochromatic light was sent. An ordinary 
 photographic lens brought the various 
 images to a focus, and the result shown 
 in the figure was obtained. The top line 
 Pig- 26. jg ^jj image of the slit, whilst the lines 
 below show the central image and the bands. It will be 
 seen that the brightest lines of light are visibly duplicated, 
 owing to a slight irregularity in the ruling. 
 
 To obtain the next figure a fairly wide slit was 
 employed, and the three sets of bands are due to 
 monochromatic light of different colours coming through 
 the slit and falling on a surface on which some 3,000 
 lines to the inch were ruled. The figure is an absolute 
 reproduction of a photograph. Successive strips of the 
 plates were used for the exposures to the different coloured 
 bands. The monochromatic rays used were orange, 
 green, and violet. The orange rays, being due to longer 
 waves than the green, are more widely separated, as are 
 the green than the violet, and for a similar cause. If 
 these three monochromatic lights were all sent through 
 
68 
 
 LIGHT IN PHOTOGRAPHY. 
 
 the slit at the same time, the plate would have shown all 
 the sets of hnes side by side. We have only to imagine 
 a very large number of rays of varying wave-length to 
 come through the slit, and the lines would coalesce, 
 and we should have a complete series of spectra, but 
 
 I I I 
 i I I I 
 
 Mill 
 
 Fig. 27. 
 
 only the first on each side of the central image would 
 be pure and unmixed. For, as will be seen, the 3rd 
 of violet bands lies between the 2nd and 3rd orange 
 bands, and would therefore mix with the intermediate 
 colours. The 2nd, 3rd, and other spectra would 
 therefore be impure. 
 
 Such spectra are diffraction spectra, and white light 
 is decomposed into the same colours by this plan as 
 it is by prisms. The violet is least and the red most 
 diflfracted, owing to the fact that the waves of the former 
 are the shortest, and those of the latter the longest, 
 which can give an impression on the eye. The ultra- 
 violet and the infra-red rays are also diffracted. With the 
 former, the wave-lengths being shorter, the bands of light 
 would be closer together ; with the latter, further apart. 
 
 The wave-lengfths of th'e different colours have been 
 determined with spectra of this kind by measuring 
 
CALCULATION OF INTERVALS. 69 
 
 the angles which they make with the central image, 
 or by measurements from photographs. If the spacing 
 of the intervals is known accurately, formulae will 
 enable the wave-lengths to be determined. Conversely, 
 knowing the focal length of the lens, if we measure 
 accurately the distance apart of the images of the lines 
 produced by monochromatic light of some known wave- 
 length, the distance apart of the lines can be determined. 
 This may be useful when very fine ruled surfaces are used 
 for producing typographical half-tone blocks. For 
 instance, if we use a lens of i o inch focus, and find that 
 the bands produced by a wave-length of "0000 1 6 inch are 
 •048 inch apart, we know that very approximately the ruled 
 
 lines are ^-—2 or — of an inch apart. In practice 
 
 •000016 333 '^ '^ 
 
 one would measure the distance apart of the 3rd or 4th 
 
 band on each side of the central image. Say we measured 
 
 the distance apart of the 4th bands from each other, and 
 
 found it to be "38 inch, then the distance apart of the 
 
 ist bands from the central image would be \^, or "475, 
 
 which would make the lines -riy apart, instead of t»t, as 
 
 above — a difference which would be inconsiderable for 
 
 the purpose which such lines had to serve. 
 
 The following is a table of wave-lengths (in 
 
 io;aao-^^7r,ost of a metre) of the principal Fraunhofer 
 
 lines in the solar spectrum, together with the bright 
 
 lines of lithium. The lithium lines, the D line (due to 
 
 sodium), the 6 line (due to magnesium), and the H line 
 
 (due to calcium), are most useful in determining the 
 
 different parts of the spectrum in the arc electric light. 
 
 In subsequent chapters these lines will be shown as 
 
 reference lines for photographs of spectra. 
 
70 
 
 LIGHT IN PHOTOGRAPHY. 
 
 A .. 
 
 . 7.S96 • 
 
 limit of visibility in the red 
 
 B .. 
 
 . 6,867 • 
 
 dark red. 
 
 Lithium 
 
 . 6,705 . 
 
 dark red. 
 
 C .. 
 
 . 6,562 . 
 
 bright red. 
 
 D .. 
 
 . 5,892 ■ 
 
 orange (centre of pair). 
 
 E .. 
 
 • 5,269 . 
 
 . green. 
 
 b .. 
 
 . 5,183 . 
 
 green. 
 
 F ,. 
 
 . 4,860 . 
 
 blue green. 
 
 Lithium 
 
 . 4,603 . 
 
 blue. 
 
 G .. 
 
 • 4.307 ■ 
 
 violet. 
 
 H .. 
 
 . 3,968 . 
 
 limit of the violet 
 
CHAPTER VIII. 
 Pinhole Images. 
 
 When light is treated from a geometrical optical point 
 of view, we can trace in a very simple manner the 
 course in which a beam will travel through any trans- 
 parent medium ; but there are phenomena which cannot 
 be thus accounted for. If we treat it on the wave 
 theory, as we have already done, any difficulty disappears. 
 If, for instance, light from a distant point passes through 
 a minute hole on to a screen, we should conclude we 
 ought (according to geometrical optics) to have an 
 image of such a hole, with a sharp-cut edge. The 
 figure on the screen ought to be a facsimile of such 
 a hole, the dimensions being increased in the ratio of 
 the distance of the point of light from the screen to 
 that of its distance from the hole. Facts — or, what is 
 better, observations — are against this assumption ; the 
 image will be found to be confused, and if the hole be 
 one with angles, the sharp corners will be found to be 
 shaded off. The cause of this is explained in a 
 satisfactory manner if we take into account what we 
 have described in Chapter VII. 
 
7* LIGHT IN PHOTOGRAPHY. 
 
 We may deduce theoretically the image we ought to get, 
 and then see experimentally if it is correct. First of all, 
 we will suppose that we are dealing with light of one wave- 
 length only. Let B C be the hole in a plate A D, and 
 
 i 
 
 I-ig. 28. 
 
 let L G be a straight line passing through the centre in 
 the direction of the beam of light. Then, evidently, as 
 the source may be considered to be very far distant, 
 E F would mark the diameter of the geometrical image 
 on the screen S S. Now let us see what really occurs at 
 the point G. Let us first divide the radius 3 B of the 
 hole into such intervals that the distance from G to 
 each is just a half wave-length (which we will call J x) 
 greater than it is from the preceding one. Thus G V will 
 be greater than b G by i a, G I" will be greater than 
 I G by A, and so on. If we call a the distance I G, 
 then U G will equal ^fl2+(4' bf, and this is very nearly 
 equal to a + -^^, since r, the radius of the hole, is 
 supposed to be very small compared with the distance of 
 the screen. But 3 G = <? + -• so that a 4- - — ^ = « j. -, 
 
 2 ' 2a ^ 2 
 
 that is, i' 5 = yoA. In the same way b b" = V2 a \, 
 and b b'" = .yja~k. If we draw circles through these 
 
IMAGES GIVEN BY PINHOLES. 73 
 
 points thus found, it will be readily seen that the areas 
 enclosed between them will be each equal to one 
 another — namely ir o a. 
 
 This is evidently a somewhat important point, for 
 it means that the light coming through each of these 
 areas is equal in quantity. [It may here be stated 
 that if the point of light be near to the pinhole, and 
 c is the distance between them, that then the points 
 into which the radius of the hole are divided are 
 
 &c.. 
 
 "Va + c Vo+i: Va + , 
 
 and the successive annuli are all equal to ^ . ^ ]. Leaving 
 mathematical formula, we thus arrive at the conclusion 
 that something which would be unexpected from 
 geometrical considerations will happen. If two sets of 
 waves, each starting from the same point at the same 
 instant, arrive at a point on a screen, but by slightly 
 different paths, they may not be in the same phase, as we 
 already know — that is, the crests of each of the two sets 
 of waves may not strike the screen together. To repeat 
 what we have already said, if the hollow of one wave 
 arrives at the screen when the crest of the other wave 
 arrives there, the one will fill up, as it were, the other, 
 and there will be no motion, and, as light is simply wave 
 motion, there will be darkness. Taking this view of the 
 matter, we come to the conclusion that the light coming 
 through the central area of the pinhole will destroy the 
 light at G coming through the first ring, that of the second 
 will destroy that of the third, the fourth of the fifth, and 
 so on ; and if it happens that there is an even number 
 of rings (including, of course, the central area), then 
 there will be a dark spot at G. If there is an odd 
 
7+ LIGHT IN PHOTOGRAPHY. 
 
 number, there will be a light spot. If the outside ring 
 
 be not equal to the others, there will be a grey spot 
 
 at G, since the motion will not be completely destroyed. 
 
 It is thus evident that, as we move the screen nearer to, 
 
 or further from, the pinhole, there will be alternations of 
 
 bright and dark spots at G, and between these positions 
 
 partial brightness only. The distance of the screen for 
 
 . /"* . . 
 
 these positions is given by ^ ; when « is odd it gives 
 
 the position for a bright central spot, and when even 
 for a dark central spot. 
 
 We have next to see what will happen round G. Now 
 we shall give an approximation only. 
 
 Fig. 29. 
 
 What will happen at P, for instance ? The circular 
 hole will not appear circular from the position P, but 
 elliptical ; but P being not far distant from G, we 
 shall neglect this. But now we must see what 
 difference there is in the lengths P C, P B. Let us call 
 YG X. The distance P G = Va^ + (a: + r)», or very 
 
 nearly a ■\ jj— • 
 
 The distance P B = Va' + (x — rf, or very nearly 
 
DARK AND BRIGHT RINGS. 75 
 
 « -i J^- The difference in length between P c and 
 
 P B is therefore '" + ">'-'"- '"' -that is, ^. If 
 
 this difference is an odd number of half wave-lengths (j), 
 P will be a point in a bright ring, and if it be an even 
 number of half wave-lengths (j), it will be a point in a 
 dark ring. 
 
 We can now readily find where the dark and light 
 rays will be found, for we have then x — —,x=. -— , 
 
 S^^o .,,.,. , a\ 2aK 
 
 X =~Z7. &c-> for the bright rings, and x =-^, x= ~, 
 
 X = —,&c., for the dark rings. The points of maximum 
 brightness will not be quite at the above intervals, but 
 for our purpose this is near enough. [It will be evident 
 why these formulae are but approximations, as we have 
 only considered a single vertical section of the hole. 
 If a horizontal section were taken with P, as repre- 
 sented, the joint phases of the waves would not be 
 quite the same as given. The complete solution of the 
 accurate position of the rings is one of extreme difficulty.] 
 If P be a bright point, and situated within what would be 
 the geometrical image of the pinhole, it is evident that 
 it will be very much brighter than if situated beyond it, 
 for most of the light passing through the pinhole would 
 be in that area, and will be condensed into these rings. 
 If P be beyond, only a small proportion of light will be 
 in such a ring. We have only to imagine the hole 
 much larger to see this, for evidently then the direct 
 light would be strong in the geometrical area. The 
 rings outside the geometrical image will not be exactly 
 given by the formula, nor will they be so nearly correct as 
 
76 LIGHT IN PHOTOGRAPHY. 
 
 those "inside it. To illustrate the appearance of an 
 image given by a point of light, the accompanying is 
 an enlarged facsimile of' pinhole images taken of a 
 point of light -005 inch diameter, placed at 24 inches 
 from the pinhole, which had a diameter of '042 inches, 
 monochromatic light of a wave-length of '00001 83 inches 
 
 Fig. 30. 
 
 being used. No. 8 is the enlarged photograph of the 
 pinhole itself. 
 
 No. I was taken at 3 inches. 
 
 No. 3 
 
 „ 4'2 ., 
 
 No. 4 
 
 ,. 5'2 ., 
 
 No. 5 
 
 ,, 6-8 „ 
 
 No. 6 
 
 ,, 9'S „ 
 
 No. 7 
 
 ,, i6*o ,, 
 
 It will be seen that Nos. i, 3, 5, and 7, have bright 
 
LARGE AND SMALL PINHOLES. 77 
 
 centres, and that Nos. 2, 4, and 6, have dark centres. 
 No. 1 has four bright rings enclosing a bright centre, 
 No. 3 has three rings. No. 5 two rings, and No. 7 has one 
 ring. These correspond approximately to the results 
 obtained by the two different formulae given, as do 
 also the other Nos. with dark centres. No. 7 is where 
 the last bright centre possible is beginning to be formed. 
 It is evident that if the pinhole be so small that at a 
 distance from the screen the aperture cannot be divided 
 into rings, so that one destroys the other, the central 
 image will always remain bright, and that there will be 
 only rings outside the geometric image. 
 
 What we have shown as the results of light coming 
 through a pinhole naturally leads to the subject of 
 pinhole photography. What is the best size of pinhole 
 to use } We can only answer the question by a reference 
 to what has just been said. It will be seen that practically 
 a pinhole forms an image of a point of light of the same 
 size, or a little larger, than itself, when the aperture 
 can be divided into rings, the light coming through 
 which can mutually interfere with one another. The 
 pinhole which is of such a diameter that we have only 
 one bright central disc, and surrounded by a black ring 
 the size of the pinhole, is evidently that which should give 
 better results than any others which give more rings, for 
 then the total area of light will be larger. The question 
 is, why not make the pinhole smaller still } We can answer 
 the query by calculation and by experiment. Suppose 
 we place a screen at ten inches from a pinhole, and can 
 alter the size of the hole. The radius of the hole 
 which will give the first white centre is r = Viox -000018, 
 or r equals '0134 inches — that is, the diameter is about 
 
78 
 
 LIGHT IN PHOTOGRAPHY. 
 
 Vo inch, and the dark ring will be at the circumference 
 
 of the geometrical image. Suppose we make it -^ of 
 
 an inch in diameter, or "0083 inch in radius, then the 
 
 «,,,,. ■„ ■ « X 10 X 000018 , . , . . 
 first black ring will be — . or — ;^^g^ — , which is "02 1 
 
 — that is, the image will be larger than when the pinhole 
 is '0134. The central disc in the latter case would, 
 however, fade off more gently than in the former, so the 
 practical difference would not be so great. For our 
 own part the diameter of pinhole which appears best 
 suited for photography is that which the writer has 
 already given, which is /- = Va x ■ooooF?, where a is 
 the distance of the plate from the pinhole, and '000017 
 the wave-length of the ray of maximum photographic 
 activity. The accompanying diagram gives an idea of 
 
 jlenA 
 
 sol 
 
 ▲ ▲ 
 
 A ▲ 
 
 [if 
 
 
 ^'g- 31- 
 
 the kind of image of a point of light formed by pinholes 
 of aV, «V, and to^ inch in diameters, on a plate ten 
 inches distant. The triangle was photographed at ten 
 
ZONE PLATES. 7-9 
 
 feet through these holes, and an image by a lens of nearly 
 the same size is shown with them for comparison. 
 
 We have so far only spoken of the effect of a pinhole 
 with monochromatic light. One of the most beautiful 
 phenomena is to examine the light coming through a 
 pinhole by means of an eye-piece or focussing-glass. 
 The eye will see, as the magnifier is gradually drawn 
 further away from the pinhole, a succession of coloured 
 centres. This is due to the fact that at every distance 
 there is one colour which is quenched entirely by 
 interference, and others which are more or less 
 quenched. What is left of the remaining colours is 
 therefore seen blended together. Similarly, the black 
 rings are fringed with colour due to the same cause. 
 If we put a solution of ammonia sulphate of copper in 
 front of the pinhole, we get what should be approximately 
 photographic light, and we can notice the sharpness or 
 otherwise of the rings. They are not so sharp as with 
 monochromatic light. 
 
 We next may call attention to a method of increasing 
 the rapidity of a pinhole exposure by an artifice. If we 
 divide a hole into concentric rings in such a way that 
 the path of the rays to the point exactly opposite to the 
 centres of the hole be exactly half a wave-length behind 
 each other, the rays coming through the alternate 
 rings destroy one another. If, however, we block out 
 the alternate rings — say the second, fourth, and sixth, 
 and so on — the rays coming through the remaining 
 rings will arrive at the screen all in the same phase, and 
 will not destroy one another, but will increase the 
 illumination. If we draw rings with a diameter of i, ■y~2, 
 Jl, v'4, &c., on a piece of paper, and blot out every 
 
8o LIGHT IN PHOTOGRAPHY. 
 
 alternate ring with black paint, and photograph the 
 target-like drawing, and then make a transparent 
 negative by means of photography, reducing the centre 
 ring to (say) -^o- of an inch, and use this photograph 
 instead of the pinhole, we shall get an aperture which 
 will give a brighter image at a distance of ten inches 
 than with the simple pinhole. Such an aperture is 
 called a zone plate, and with it the focus of the red is 
 nearer the screen than the violet, a result exactly 
 the reverse to that obtained with an uncorrected lens. A 
 zone plate, to be properly tested, should be used with 
 monochromatic light, which can be obtained in the way 
 we shall shortly describe. 
 
CHAPTER IX. 
 
 The Colours of Thin Films. 
 
 Who of our readers has not in his childhood, and 
 perhaps in his manhood, amused himself by blowing soap 
 bubbles, and found himself delighted at the rainbow hues 
 they exhibit, or in watching the endless variety of colours 
 which rest on the lather floating in a wash-hand basin ? 
 We may ask, further, how many have traced the cause of 
 these colours to their source ? For the sake of those 
 who have not, we will endeavour to give a simple answer. 
 We have seen if two rays of light of the same wave- 
 length start from the same point, and happen to meet 
 again, one being half a wave-length behind the other, 
 the wave-motion is destroyed where this occurs. This 
 is exactly what happens when a ray of light strikes on a 
 soap film. Part of the ray is reflected from the front 
 surface, the larger portion goes through, and a part of 
 this residue is reflected from the back surface, and 
 follows the track of the first reflected beam ; and if, the 
 ray being monochromatic, on striking a screen or the 
 eye, one beam is an even number of half wave-lengths 
 behind the other, there will be darkness. When the 
 
 G 
 
82 LIGHT IN PHOTOGRAPHY. 
 
 number of half wave-lengths is odd, there will be light. 
 Suppose the thickness of the film be such that, say, the 
 waves from the back arrive at a screen nine half wave- 
 lengths behind those reflected from the front, then we 
 shall have brightness. Let us see, however, what would 
 happen to a ray set of whose wave-length was tV, say, 
 that of the other. Evidently, in the thickness of the 
 film there would be ^-9^ of 4^, or ten half waves — that 
 is, the last ray would be annihilated. If there were 
 rays with wave-lengths intermediate between the two, 
 it is also evident that the colours whose rays were 
 short would be dimmed, and those whose rays were 
 approaching the longer set would be increased in 
 intensity. Thus, with white light falling on a soap 
 bubble at each point there would be an intensification 
 of some rays, and a reduction in intensity of others, and 
 the combination of them all would result in a colour 
 which was a mixture in varying proportions of the 
 spectrum colours. What colours were dimmed and 
 what intensified would depend on the thickness of the 
 film. Now let us reason out whether the colours ought 
 to be more distinct with a thick or with a thin film. 
 Suppose, instead of nine half waves, we had ninety-one 
 half waves in the thickness of the film, and we can see, 
 as before, what would happen with the rays that were 
 -1% as long. Evidently, in the thickness we should have 
 1 01 (and a small fraction) of half waves within the film, 
 and we should have nearly maximum brightness 
 with both rays. There would be some wave- 
 length intermediate, when complete interference would 
 take place. Hence the thicker the film the closer 
 together would be the parts of the spectrum which 
 
INTEKFEKEN'CE BANDS. 83 
 
 would be darkened or intensified. Finally, we should 
 come to such a thickness that the parts would be so 
 close together, and melt so one into the other, that we 
 should fail to distinguish any loss or increase in 
 intensity of the various parts, and we should by their 
 mixture on the screen get white light. Thus, the thinner 
 the film the less mixed with white light the colour 
 would be. This agrees with observation, for a large 
 bubble with a thin film takes more brilliant colours 
 than a smaller bubble with a thick film. 
 
 Now a soap film need not be in the form of a bubble 
 to be examined. It is much more amenable to physical 
 observation if it is merely a flat film, and this can 
 readily be secured by dipping a piece of card or metal 
 in the soap solution, and drawing the moistened end 
 across an aperture such as that in a large stop of a lens. 
 The film can be placed in a white or monochromatic 
 beam, and the image of the aperture be then thrown 
 on a white screen or card. What we shall see at first 
 will be an image of a colourless film, for the reasons 
 just stated, as the thickness will be comparatively 
 great. As we will suppose tl%e film to be vertical, 
 the excess of liquid will be found drawing away from 
 the top, and thinning the film from the top down- 
 wards ; each horizontal section will practically have 
 the same thickness. The thicker part of the film will 
 be downwards. The beam of light will now have to 
 traverse varying thicknesses before reflection from the 
 back surface, with a result that, when using mono- 
 chromatic light, we shall have bands of blackness 
 close together at the bottom, and gradually getting 
 further apart as the top of the film is approached. 
 
84 
 
 LIGHT IN PHOTOGRAPHY. 
 
 With white light we shall have the same number of 
 coloured fringes as we have black bars. This^ as wc 
 have just explained, will be due to the fact that for 
 every thickness of film there will be a different mixed 
 colour. 
 
 The accompanying figures are photographs of a film 
 illuminated with green monochromatic light. Fig. 32 
 evidently shows a thicker film than Fig. 33, since the 
 bands of darkness are closer together. It must be 
 
 noticed the white portions (which were really green 
 in the film) are not of one uniform tint. They shade 
 off gradually towards the dark bands. This simpi)' 
 shows that the varying thickness of the film. This will 
 be evident if we consider what would happen if the rays 
 were only, say, i of a wave-length behind one another. 
 Then the trough of one would not fill up the crest of the 
 other, but would only do so partially. This will be seen 
 
SPECTRUJr OF LIGHT FROM A SOAP FILM. 
 
 85 
 
 by reference to Fig. 22, page 61. These reproduced 
 photographs tell us that the colours reflected by a soap 
 film are not merely visual effects, but act on a sensitive 
 surface like colours produced in other ways. 
 
 We can learn more about these colours, however, and 
 make an interesting experiment, by viewing the spectrum 
 of the light reflected from a narrow section of a film, 
 and more still if we secure a permanent record of the 
 spectrum by means of photography. We can throw a 
 
 Fk- 33- 
 beam of sunlight or the electric light on to the soap 
 film, and then, by means of a lens of suitable focal length, 
 form an image of the film on the slit of a spectroscope 
 furnished with a camera. On looking at the focussing 
 screen we shall see a continuous spectrum traversed 
 obliquely in its length by black bands. The slit, it will 
 be remembered, allows only the light from a narrow 
 vertical section of the film to be decomposed into its 
 spectrum colours, the width of the spectrum answering 
 
86 LIGHT IN PHOTOGRAPHY. 
 
 to the length of the slit. As through each part of the 
 slit we admit light of different mixed colours, we may 
 expect to see blank bands in the spectrum, which show 
 us what colours are absent, and also what residue is 
 left to form the mixtures. Fig. 34 shows such spectra 
 
 P'g- 34- 
 
 as photographed on a Cadett orthochromatic spectrum 
 plate. In one part these bands are close to one another, 
 and in another they are separated by wider intervals. 
 This tells us two things— if we take a horizontal line 
 along the spectrum, i st, we can tell what wave-lengths 
 are absent, and 2nd, knowing this, we can calculate the 
 distance apart of the bands, and the thickness of the 
 film at any part. It is true that there are bands of 
 interference when the waves reflected from the back are 
 just half a wave-length behind those reflected from the 
 front, but owing to the fact that one reflection takes 
 place from the soap solution into air, and the other 
 
THICKNESS OF THE FILMS. 87 
 
 from air into soap solution, the phases of the waves are 
 to begin with retarded as regards one another by half a 
 wave-length — that is to say, if the path of the ray 
 through twice the thickness of the film (for it has to 
 traverse it twice) should be so long as to make it an 
 exact multiple of some wave-length, it will be found 
 to differ really by half a wave-length. Unless we 
 knew this from theoretical considerations we should 
 suppose the two beams would reinforce one another, 
 though they will really interfere with one another. (If 
 we call t the thickness of the film, and « the angle at 
 which the light falls on it, in order to see what the 
 issuing waves will do, we must add half a wave-length 
 to the path ; thus the length of path for any beam will 
 be 2/ cos fl -|- ix.) 
 
 We now measure the distances apart of the bands in 
 a horizontal line, and, knowing the wave-length scale of 
 the photograph, we see what wave-lengths are obliterated. 
 This is useful to know, because the path might be any 
 number of wave-lengths, and these will enable us to 
 calculate the thickness of the film by means of very 
 easy equations. All the wave-lengths of the black bands 
 in a horizontal line must be exactly multiples of the 
 thickness of the film, and thus we can easily see that 
 these bands may be made to tell us what the thickness is. 
 From the photographs it was found that in one 
 horizontal line the wave-lengths of the black bands 
 were 397, 441, 496, 566. If we call « the number of 
 wave-lengths of the first band, then there will be n — i 
 wave-lengths of the second, and [n — 2) of the third, 
 and so on, and as there is interference » must be a 
 whole number. We must first suppose the film to be 
 
88 
 
 LIGHT IN PHOTOGRAPHY. 
 
 of air, as the wave-lengths are given in this medium, 
 and subsequently convert this thickness into that of 
 soap solution, and allow for the angle at which the 
 beam strikes the film. Calling this air thickness zt we 
 find that — 
 2/ = 397» = 441 (« — i) = 496 («— 2) =566 («— 3). 
 
 From this we find that « = 10, it being so almost 
 exactly from the first and second wave-lengths, and also 
 from any other combinations. The thickness of the film 
 in the direction in which the light entered it is therefore 
 
 IbbF"^ 
 
 m 
 
 1 
 
 ■■ 
 
 r*a 
 
 wf: 
 
 1 
 
 M 
 
 
 
 anw 
 
 m 
 
 ^^^|Bi|kkL2!'''~*' 
 
 a 
 
 i 
 
 'WB^ 
 
 five times the wave-length 397 (the unit being 100,000,000 
 of a centimetre)— that is, '0000794 of an inch if the film 
 were of air, but as it entered at 45°, and is in water, we 
 must multiply this result by cos 45", or by about 7, and 
 divide by^, the co-efiicient of refraction in water, which 
 is about 1-3. The thickness of the film normally is 
 therefore "000036 inches. 
 
SOAP FILMS AND SOUND. 
 
 89 
 
 It will be seen now how photography lends itself to 
 the exact measurement of thin films, which it is difficult 
 to carry out by visual observations. 
 
 Soap films are so thin that they can be made to show 
 the nodes produced by sound. 
 
 When we have sound waves, the vibration will be taken 
 up by metal in which the aperture is cut, and the vibra- 
 tions set up will also be in the film stretched over 
 the aperture. Now, owing to its rigidity, the metal will 
 not show the vibrations of these waves, but owing to the 
 thinness of the film the vibrations set up in it will be 
 very apparent, avl the nodes will be seen in a very 
 
 Pig- 36- 
 beautiful way. Fig. 35 is the note E sounded by a 
 penny whistle. It will be noticed that not only are the 
 circular nodes shown, but also the dark bands of light 
 interference are likewise visible, and apparently un- 
 disturbed. This and the next two photographs (Fig. 36 
 
90 
 
 LIGHT IN PHOTOGRAPHY. 
 
 and 37) were taken with monochromatic light, one ray 
 of the spectrum of the electric light being spread out 
 to cover the film, and the reflection was received on a 
 lens. The image of the film being focussed, exposure 
 
 Pig- 37- 
 
 was given to a plate for about the -Ath of a second. 
 Fig. 36 is the effect of a shrill whistle sounded some 
 two feet off the film, and Fig. 37 the eflfect of a shrill 
 note sung by a human voice. In Fig. 3 6 the film showed 
 one uniform tint, and hence no bands of interference 
 are visible. The fact that the colours of thin films can 
 be photographed shows that a rainbow can also be 
 photographed, and on several occasions the writer has 
 secured a landscape picture in which a rainbow appears 
 most markedly. 
 
CHAPTER X. 
 
 i\IODES OF jMeASURING GrEY SURFACES AND 
 
 Opacities. 
 
 Having a plate, film, or other sensitive surface which 
 has been exposed, and, where necessary, developed, 
 cases occur in which we ought to be able to measure 
 the transparency or blackness of the different parts of 
 the image, for we can then apply such measurements to 
 the investigation of various problems which present 
 themselves to us, amongst them being the amount of 
 chemical action produced under varying conditions. 
 
 The oldest plan of measuring the illuminating value 
 of two lights is by no means the worst. This plan is 
 due to Count Rumford, and is called the Rumford 
 method. It consists essentially in causing two lights to 
 cast two shadows of a rod on a white screen. By 
 altering the distance of the rod from the screen, or by 
 a slight adjustment of the angle which the lights 
 make with the screen, the two shadows (each of which 
 is illuminated by the light not casting such shadow) can 
 
92 LIGHT IN PHOTOGRAPHY. 
 
 be placed side byside so as to touch one another (Fig. 38). 
 
 One of the two lights can then be drawn away from or 
 advanced towards the screen, till the eye 
 judges that the two shadows are equally 
 bright. The judgment is aided if only 
 a small portion of the shadows can be 
 seen, and this can be secured by covering 
 
 Fig. 38. the screen with black paper or velvet, in 
 
 which an aperture of, say, one inch square is cut. The 
 shadows are cast on the exposed white surface, and the 
 attention of the eye is then directed entirely to their 
 brightness, and is not distracted by the brighter white 
 light around them. The surface of a white cube against 
 a black velvet screen answers well, and in some cases is 
 more convenient. The comparative illumination of the 
 two lights is found by measuring their distances from the 
 shadows on the screen, squaring these distances, and 
 dividing one square by the other, the illuminating value 
 being inversely as the square of the distance. Thus, if 
 one light be 2 feet, and the other 4 feet, from the screen, 
 the latter is \-^, or four times more luminous than the 
 former. 
 
 If, instead of a white square on the screen, we have 
 one half of it grey (such as is produced by a moderate 
 exposure of platinum paper to light), the grey lying on 
 one side of a central vertical line, the other half 
 being white, we can throw one shadow on the grey 
 half, and the other shadow on the white half. By 
 moving the light which illuminates the white, the two 
 shadows can again be made equally bright (say we have 
 to move it six feet away to cause this equality), and if 
 we replace the grey half by a white, we can again make 
 
MEASURING BOX. 93 
 
 the illumination equal (say we only have to move it four 
 feet away from the screen). The amount of white light 
 which the grey reflects is given by the same rule ; as 
 before, it will be inversely as the square of the distances 
 of the one light on the two occasions. In the measure- 
 ments we have supposed, this would make the grey 
 
 4X4 4 
 reflect zzn = - of the white light, which the white 
 0X09 
 
 screen would reflect — that is, it would reflect 44*4 per 
 
 cent, of light. 
 
 All extraneous light should be carefully excluded 
 
 from the shadows when a grey is being measured, for 
 
 supposing that the grey absolutely reflects, say, 44^4 
 
 per cent, of white light, then it will only reflect the 
 
 same percentage of the stray light, whilst the white half 
 
 would reflect 100 per cent. A measurement made then 
 
 in diffused light would make the grey appear darker 
 than it really is. To overcome this difficult)', these 
 measurements should be made in a darkened room, and 
 
9+ LIGHT IN PHOTOGRAPHY. 
 
 the screen should be in a box such as given in the 
 figure below. 
 
 The box is B B. In the front towards the light is an 
 opening (o) through which the light passes. At one edge 
 is another opening (A) through which the screen P P' 
 can be viewed, the rod R being inside the box. By this 
 plan such darkness is secured that the blackest black 
 appears luminous by contrast with the white when the 
 light which illuminates it is entirely shut off. The figure 
 (Fig. 39) shows a slight variation of the method. Both 
 illuminations are due to one source of light, the second 
 being furnished by the reflection of the first light from a 
 
 Fig. 40. 
 
 silvered mirror M (Fig. 39). In this case the amount of 
 light fallingon the screen from either beam can be reduced 
 
ROTATING SECTORS. 95 
 
 by placing in it rapidly rotating sectors (Fig. 40), the 
 apertures in which can be opened or closed at pleasure 
 during rotation. A A, are the two sectors which move 
 across one another. B is a sleeve in which is a point 
 that moves in a screw cut in the spindle. The pins, 
 which are fixed to one of the wings in the sector, cause it 
 to turn where B turns. B is moved to and fro in the 
 sectors by the lever C, which has a join worked in the 
 grooved portion of B. D is an electro motor attached 
 by a band to pulley on the spindle. The rim of the 
 sectors is divided into degrees. The shadows can thus 
 be made to be equally illuminated. This is a refine- 
 ment, and is not absolutely requisite, though it is 
 convenient. The two lights may be used, as before 
 stated, but as each light might vary (particularly if it 
 were that of a candle), the measurements are more 
 liable to error than is the case where one light alone is 
 used, for then any variation in one shadow is equally 
 felt in the other. 
 
 In a box such as described, there is an aperture cut 
 out at P P (Fig. 39), which is not shown. This aperture 
 is oblong, the longest side being horizontal. Inside 
 the box it is seen as an open space. Filling half the 
 opening is an oblong of white. If it be desired to 
 measure a piece of grey paper, it is made to fill the 
 remainder of the aperture from the back, and we thus 
 have the grey and white alongside one another. In 
 practice it is well to have the back of very thin metal, 
 so that the junction between a piece of white paper 
 placed in the aperture, and that of the grey alongside, 
 is almost invisible. The two then appear, as in Fig. 38. 
 
 This, then, is a plan for measuring the greyness of a 
 
96 LIGHT IN PHOTOGRAPHY. 
 
 surface. But besides this, we may require to know what 
 are the relative transparencies of a deposit produced 
 photographically. Here the circumstances are com- 
 pletely changed. We might, indeed, place such deposit 
 of a negative in the path of one beam, and see how 
 much of the light was cut off by it. A deposit of this 
 kind does not cut off light by absorption, for it is not 
 continuous. It is formed of opaque granules, and the 
 light is really cut off by the interposition of these grains 
 of silver in the path of the beam. Now where we have 
 granulation a certain proportion of the light will not 
 pass through regularly, but will be scattered by reflection 
 in all directions, and thus will not proceed in the path 
 of the beam. This may be readily demonstrated. Let 
 us take a small gas flame, and balance it by the shadow 
 method against another source of light, say an amyl 
 acetate lamp. Now let us take a " whole" plate, which 
 has been exposed to a uniform light, and develop it. It 
 will have approximately a uniform deposit all over it. 
 Next cut out in a blackened card an area somewhat 
 larger than the flame, and place the aperture in the 
 path of the beam to the screen. No alteration in the 
 balance will be seen ; the shadows will appear equally 
 illuminated, because each part of the shadow illuminated 
 by it will "see" the whole flame. Now place the 
 negative in front of the black mask, and alter the other 
 light till the shadows again appear equal, and note the 
 distance the moving light is from the screen. Change 
 nothing, but remove the black mask, and at once the 
 shadow illuminated by the light coming through the 
 negative will appear lighter. The only cause of this 
 alteration is the increased area of the plate exposed. 
 
PHOTOMETER SCREEN. 97 
 
 and it is from those parts of the plate not in the direct 
 line of the beam that the increased illumination comes. 
 Moreover, we have only now to alter the distances of the 
 plate from the source of light, and the illumination of 
 the shadow will vary. The place which gives the least 
 illumination will be found about half way between the 
 light and the screen, and when the plate approaches 
 either the source of light or the screen, the illumination 
 will become greater. What we want to know is the 
 percentage of light which absolutely penetrates through 
 the deposit. For this reason the writer has devised a 
 method by which the deposit is almost in contact with 
 the screen, so that not only the rays which come directly 
 through the deposit are measured, but also those rays 
 which are scattered, since they are equally caught by 
 the screen. In this case the size of the plate presented 
 to the light is immaterial. The plan the writer has 
 adopted is as follows : — 
 
 Front l/iew. 
 
 Bach View. 
 
 Fig. 41. 
 
 Suppose we pierce a small square aperture in a thin 
 
 H 
 
9 8 LIGHT IN PHOTOGRAPHY. 
 
 (say a ferrotype) plate, or a thin black card, and cover 
 all one side with white paper, and over this place a 
 mask of black paper cut out to a double square (Fig. 41), 
 one half of which is placed over the paper covering the 
 aperture, and the other half an equal square of white 
 paper ; then a light from behind the aperture will 
 illuminate the paper covering the aperture, and a light 
 in front will illuminate both the squares of white paper. 
 If we place a rod so as to cast a shadow on the white 
 square illuminated from behind, the oblong will have 
 two illuminations, one half by transmitted light, and the 
 other by the direct light from the front. These two 
 illuminations can be equalised (i) by moving the front 
 light, (2) by placing sectors such as have been described 
 in the path of the beam of the front light, or (3) by 
 altering the distance of the screen between the two 
 lights. It is not everyone who has the means of using 
 the second method, and it is best, if possible, not to 
 resort to the first, so the writer will describe the third 
 method, which he has now adopted. Mr. Chapman 
 Jones first described a most complete method of doing 
 this, and he only used one light, the second light being 
 a reflection of the first light. The writer's plan is a 
 modification, which is simpler in some respects, but in 
 which two lights are employed. 
 
 D D D' D' is a box 2 ft. 6 ins. long by 9 ins. broad 
 and high. D' D' is a side of the box left open, but 
 which can be covered by a black cloth. A A is an 
 aperture sufficiently large to allow the whole of the light 
 from the lamp L to illuminate the covered aperture in 
 the screen. S S is another similar aperture for the light 
 of L" to traverse. Behind S S the negative, N, can be 
 
MEASUREMENT OF OPACITIES. 99 
 
 placed over the aperture in the screen. The rod, R, 
 casts a shadow over the paper covering this aperture. 
 M is a silvered mirror in which the double squares can 
 
 Fig. 42 
 
 be viewed. E E is a cradle carrying S S, R, and M, 
 which can move along on a batten, CCA scale is 
 placed along D' D', so that the distance of the screen 
 from each light can be read off. 
 
 This scale is conveniently made a logarithmic scale, 
 as first carried out by Messrs. Hurter and Driffield in 
 their measuring box, since it enables the squares of the 
 distances from each light to be readily found by 
 logarithm tables. If this scale, or an ordinary scale, be 
 used, it is convenient that the zero point should either 
 be exactly half-way between the middle of the lights, 
 or else start exactly opposite the centre of the light 
 directly illuminating the white square, in this case L" 
 The two squares can be made to appear equally 
 illuminated by moving the screen, S S, and the distances 
 (or logarithms of the distances) noted, and then the part 
 of the plate whose transparency is to be measured is placed 
 with a clip over the aperture in the screen, and equality 
 
100 LIGHT IN PHOTOGRAPHY. 
 
 of illumination again arrived at by a further movement 
 of the screen, S S. The distances are again noted, and 
 from these measures the light transmitted through the 
 negative is readily calculated. An objection has been 
 raised to this plan of placing the negative near the 
 screen, as it was supposed that the reflection of the 
 light from the white surface of the paper on to the plate, 
 and back again, made more light appear to penetrate 
 than should do if it were at a distance. That more 
 light passes is true, because the screen takes into 
 account the scattered light, but the reflection is 
 negligible. We may popularly explain it as follows : — 
 We will suppose that a light of one unit falls on the 
 white square, and that three-quarters enters the paper, 
 and gives the illumination at the front surface ; then the 
 remaining quarter is reflected back. Of the quarter 
 reflected by the paper, if glass be next it, about 2 per 
 cent, will be again reflected back by the glass, and 
 three-quarters of this will penetrate — that is, -1^ of i per 
 cent., or f per cent. — and J per cent, of J per cent. — 
 that is, ^ per cent, will again be reflected back. A 
 minute percentage of this will again pass through, and 
 so on, in diminishing terms. If the surface away from 
 the white paper be also bare glass, somewhat of the 
 same amount would penetrate, which would make about 
 I per cent, in all due to reflected light. The measure- 
 ments ought to show a diff"erent result if a black deposit 
 be next the paper, and the bare glass away from it, to 
 that which it would show if the bare glass were the 
 surface next the paper. As a matter of fact, the readings 
 are indistinguishable one from the other. Perhaps the 
 readiest plan of ascertaining whether the reflection from 
 
HURTER AND DRIFFIELD's PHOTOMETER. lOI 
 
 the paper causes an error in the true reading, is to place 
 a piece of bare glass behind the aperture, A A, and 
 equalise the illumination of these two squares, and then 
 move the glass towards the light. If the reflection 
 causes any error, it should at once show, for the light 
 reflected from the paper will spread away, and a 
 diminished reflection from the glass would ensue. When 
 the glass is over the aperture, any light that the paper 
 reflects back will be so scattered that it must be 
 negligible. 
 
 We have entered into this somewhat lengthily 
 because the reflection has been said to be fatal to 
 correct readings. We may further point out that for 
 printing purposes the screen and the plate are in 
 contact, as they are in the photometer, and thus a 
 correct value is obtained for the uses to which a negative 
 is usually put. 
 
 l\Iessrs. Hurter and Drifiield had previously con- 
 structed a photometer which is very compact, and appears 
 to be capable of measuring very exactly the light that 
 falls upon a screen. Broadly speaking, it is constructed 
 on the same lines as Fig. 42, it being a box containing 
 a movable screen, the screen being furnished with a 
 grease spot, the measurements being made after the 
 manner of a Bunsen photometer. The lights are 
 admitted through two apertures, one at each end of the 
 box, and as the flame more than covers the apertures, the 
 measurements are taken from the apertures to the screen. 
 The density to be measured is placed in contact with one 
 aperture, and the screen shifted till equality of illumina- 
 tion of the grease spot is secured. The position of the 
 opacity to be measured, it will have been gathered, is 
 
102 LIGHT IN PHOTOGRAPHY. 
 
 not that which the writer w mid choose, for the reasons 
 already given. The great point in choosing a source of 
 light is that the part opposite, and close to the aperture, 
 should be of equal brightness throughout. If not, the 
 screen at differing distances might be illuminated by 
 difforeat quantities of light, irrespective of the amount 
 dependent merely on the distances. Even a flat flame 
 will vary, as the accompanying figure of a lamp flame, 
 photographed with different exposures, shows. The 
 
 F^g- 43- 
 shortest exposure tells us where the flame is brightest. A 
 needle point placed in front of the flame is a guide as 
 to the position of the brightest part. 
 
 For measuring densities it is sometimes of advantage 
 to use a black wedge or other transparent graduated 
 opacity. There are two plans of measurino- such 
 opacities. The first is to measure it by means of the 
 screen described before, using a narrow long vertical 
 
THE OPACITY OF A WEDGE. 103 
 
 aperture, so that when covered by the wedge only a 
 small portion is employed. This gives very accurate 
 readings, the real measure being that of the portion 
 of the wedge next the junction of those parts where 
 the reflected and transmitted lights are used. By 
 adopting this plan, not only can we graduate a wedge, 
 but also can measure the opacity of negatives at different 
 points. This plan has been followed in the measuring 
 of some negatives of the spectrum, which will be alluded 
 to in a subsequent chapter. Another method, however, 
 will be described which is very effective, especially when 
 a very opaque image or negative is to be measured. In 
 the case of a negative, if we measure a scale of opacity 
 in the same way that we measure the opacity of the 
 negative, it is evident we can refer the last to the first, 
 without any appreciable error, and if we subsequently 
 measure the scale of opacity by the first (contact) plan, 
 we can convert one set of measures of the negative into 
 those which would have been given by the same plan. 
 
 The accompanying figfure will show the arrangement. 
 EL is the arc electric light ; L, a lens which throws an 
 image of the points on a vertical slit S, placed at one 
 
 /::-^ 
 
 Sectors 
 
 EL-=:::::i:::::::::-~Eiinnzir;::o^--f'-=^-v;.:::,» / 
 
 W ScfMS 
 
 Ftg. 44. 
 
 end of a tube C. The light passes through S, and 
 strikes the lens Li„ which is fixed at the other end of 
 
104 LIGHT IN PHOTOGRAPHY. 
 
 the tube. The rays pass out parallel, and are collected by 
 another lens, Lm, which forms an image of the slit on W, 
 which in this case is the wedge. The rays spread out and 
 form a circle of light on the screen, as shown. A narrow 
 vertical slit of light thus passes through the wedge, and 
 a certain proportion is cut off by the opacity of the 
 wedge. If we insert in the path of the beam a plane 
 piece (or very lightly silvered glass), M, part will be 
 reflected and be brought to a focus, where a silvered 
 mirror may be placed, throwing a slightly larger circle of 
 light on the screen than that proceeding directly to it. The 
 two should be thrown concentrically on the screen. This 
 can best be secured by covering up nearly all the lens, 
 Ln, leaving only a small central hole in the mask. The 
 circles of light of the screen will be much curtailed in 
 diameters, and the two may be made to overlap exactly. 
 When these precautions are taken, any variation in one 
 circle of light will have a similar variation in the other. 
 A rod, R, when placed in a proper position, will cast two 
 shadows touching one another, due to the direct and 
 reflected beams. In the path of the reflected beam 
 the sectors with movable apertures during rotation may 
 be inserted, and equality of illumination of the shadows 
 secured, or the sectors may be fixed to different 
 apertures, and the wedge moved till the same equality 
 is obtained. The slit at S may be curtailed in length, 
 or any shaped aperture may be given, and for the 
 wedge may be substituted a negative, all parts of which, 
 together with an impressed scale of opacity (see later 
 chapter), may be measured. The different parts of the 
 negatives of the corona taken during an eclipse can be 
 measured by this artifice, and the relative brightness of 
 
warnerke's annular sensitometer. 
 
 loS 
 
 different parts of the corona be thence deduced. Mr. 
 Warnerke has recently introduced what we may call an 
 annular wedge. It is made by pouring gelatine, mixed 
 with an extremely finely ground black pigment, into 
 a mould. The mould is a flat thick plate of steel, in 
 which is cut a circular chase, which gradually increases 
 
 in depth. The coloured gelatine is run into the mould, 
 and extracted by its adherence to a flat glass plate, 
 placed in contact with the flat surface of the steel 
 mould. When dried, it presents a very even gradation. 
 Wedges such as these have been graduated by both 
 plans, and the measurements in each have been found 
 practically to coincide, showing that the scattering by the 
 very finely divided black pigment used is inappreciable. 
 
J06 LIGHT IN PHOTOGRAPHY. 
 
 Having graduated a wedge, or this new sensitometer, 
 we have another plan of measuring densities open to us. 
 We can use the same screen as before, and use one 
 light from behind and the other from the front ; but it 
 is now better to throw an image of a flame, placed 
 edgewise to the screen, on to the wedge, and alter the 
 illumination of the part of the oblong (see Fig. 38) 
 which receives the front light by moving the wedge. 
 The position of that portion of the wedge which gives 
 equality of illumination is read off, and from a table the 
 amount of light transmitted is found. With Wamerke's 
 annular wedge it is convenient to bore a hole in the 
 centre, and rotate it round an axis placed through such 
 hole. The length of this annular wedge is conveniently 
 read off in degrees. A wedge, when accurate, gives a 
 logarithmic reading which can be converted into 
 common logarithms by multiplying by a factor, which 
 will have been found when graduating the wedge. 
 
CHAPTER XI. 
 
 Measurement of the Luminosity of Colours. 
 
 Having demonstrated the way in which the greys in a 
 print may be measured, and also the transparency of 
 a deposit, we must say a word on measuring the 
 brightness of the colour of an object, say of a pigment, 
 for we often have to deal with the correct rendering of 
 pictures when they have to be copied by means of 
 photography. We cannot enter into all the niceties 
 of the subject, as it would involve far more space than 
 can be allotted to it in a work of this description. 
 Colour is, it must be remembered, a physiological 
 effect, and depends upon the effect produced by waves 
 of light on some kind of apparatus in the retina of 
 the eye. 
 
 The kind of light which illuminates a picture has a 
 great deal to say to the brightness of the various 
 colours. If we analyse light from a white cloud, we 
 shall find that practically it is reflected sunlight, and 
 sunlight, we know, has in it a great deal more red and 
 yellow than sky-light, as the figure will show. In the 
 figure all the colours of the spectrum formed by sunlight 
 
io8 
 
 LIGHT IN PHOTOGRAPHY. 
 
 are, for convenience, represented as if they were of 
 uniform brightness, and below it we see a curve of 
 light from the sky in which the proportions of the 
 different colours are given. The blue end of the 
 spectrum evidently predominates in the colour of the 
 light coming from the sky, compared with that of sun- 
 light. Again, we have an artificial light (gas) shown 
 in the same figure, and here we see that the blue is 
 almost absent compared with sunlight, whilst the red 
 is strong. 
 
 If we had three pigments — a red, a green, and a 
 blue — which all appeared to be equally bright to the 
 
 
 
 ' 
 
 
 " 
 
 
 
 
 " 
 
 
 sL ulght 
 
 
 
 
 ■ 
 
 
 rm 
 
 
 
 
 
 r^ 
 
 ^ 
 
 
 
 s 
 
 \ 
 
 
 
 
 
 
 
 
 / 
 
 TO 
 
 
 
 y 
 
 ^ 
 
 
 
 
 
 
 \ 
 
 S 
 
 
 
 
 
 
 
 / 
 
 80 
 70 
 
 
 y' 
 
 
 
 
 
 
 
 
 
 ^ 
 
 ». 
 
 
 
 
 
 / 
 
 
 
 
 
 
 
 
 
 
 
 
 
 h 
 
 1 
 
 
 
 
 / 
 
 
 
 
 
 
 
 
 
 
 
 
 
 \ 
 
 \ 
 
 
 
 / 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 \ 
 
 
 / 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 y 
 
 
 
 
 
 30 
 
 20 
 
 
 
 
 
 
 
 
 
 
 
 
 
 y 
 
 
 \ 
 
 
 
 
 
 
 
 
 
 
 
 
 GiJ 
 
 tj. 
 
 %■ 
 
 
 
 
 
 S 
 
 \ 
 
 
 
 
 
 
 
 
 
 
 \ 
 
 
 
 E 
 
 
 
 b 
 
 
 & 
 
 S 
 
 
 
 
 1 
 
 
 
 a lei 
 
 7 te ;« 14 13 12 11 to 9 . 
 
 7 < 
 
 
 • 
 
 
 
 
 
 
 
 
 
 
 • 
 
 ;3 
 
 
 
 1 
 
 
 
 
 I 
 
 • 
 
 • 
 ( 
 
 
 
 Fig. 46. 
 
 eye in sunlight, in skylight the red would appear very 
 much darker than the other two ; whilst if we illuminated 
 them by candlelight, the blue would be much darker 
 than the red and the green, and the red would appear 
 the brightest of the three. The reason for these 
 changes in brightness would be owing to the fact that 
 the colour of the red is due to the capacity of the 
 pigment to reflect red rays most plentifully, that of the 
 
PIGMENT COLOURS. 109 
 
 green to the pigment reflecting green rays most 
 abundantly, and the blue pigment to its greater 
 reflection of the blue rays. To measure the brightness 
 of the colours under these varying conditions would be 
 an endless task, and all that we can do is to get some 
 light which will not be far off that in which we usually 
 look at pictures, and refer to it as a standard. An 
 artist will always try to paint, we believe, in any light 
 rather than one in which the blue from, the sky 
 predominates. He likes what he calls a warm light, 
 and a sky in which white clouds are present is one 
 that he delights in. In other words, he likes reflected 
 sunlight, and this is the kind of light in which pictures 
 should be examined in justice to the artist. 
 
 Now the light from the positive pole of the electric 
 arc light is a very fair imitation of sunlight, and as 
 cloud-light is reflected sunlight, it may be taken as a 
 good standard light in which to view colours, and 
 measure them, when sunlight is not available. 
 
 We say that one colour is brighter than another, or 
 more luminous, and we may generalise, and say that 
 one colour has more luminosity than another. It may, 
 perhaps, be hard to say how one may judge that a red 
 and a green have equal luminosities. It has been said 
 by the highest authority on colour (Helmholtz) that he 
 could not picture a method by which a red and a green 
 might be matched for brightness. But yet there is an 
 intuitive instinct, if one may call it so, by which the 
 brightness is judged, and it is this instinct which enables 
 us to get accurate and repeatedly concordant measures 
 of the luminosities of different colours. 
 
 We may commence our measurement of the brightness 
 
no LIGHT IN PHOTOGRAPHY. 
 
 of two colours by lamplight, and may use the apparatus 
 described at page 92. Instead of the reflected light from 
 the mirror we may use a second lamp. We must, however, 
 exclude as far as possible all scattered light, so that 
 the screen should be placed in a box which cuts off all 
 light except that proposed to be used (see Fig. 39). 
 Instead of a square of grey, we may insert a square of 
 any colour — say green — and use the sectors as described, 
 or else move one lamp. The white may be made first 
 evidently too bright, and then too dark, and then, by 
 gentle oscillation, we find the place where they appear 
 equally bright. 
 
 The aperture of the sector, or the distance of the 
 lamp which illuminates the colour, is then read off, and 
 noted. We can then replace the green colour by a red, 
 and we go through the same operation. Evidently, 
 then, we can tell the relative brightness of the red to 
 the green by a simple reference of both to the white. 
 But here, again, we are met with our first difficulty. 
 Supposing we reduce equally the light falling on 
 each, we shall see that the brightnesses no longer hold. 
 For instance, if we illumine the two colours, placed side 
 by side, with a lamp, the red may be much the brighter 
 of the two ; we can cut off the same amount of light 
 from each by the sectors, or by moving the lamp very far 
 back, and we shall soon see that when the light is 
 feeble the red becomes the darker of the two colours, 
 showing that the intensity of the light falling on them 
 has much to do with the luminosity. To judge, then, 
 of the luminosity of the colours, we must know the 
 intensity of the light which calls them into being. 
 Thus, a picture in which there was red and green 
 
MEASURING COLOURS. Ill 
 
 might, under bright illumination, show a particular 
 red as brighter than a particular green, whereas, in a 
 sombre light, the reverse might be the case. It is 
 well known that towards evening, the last colours in 
 a painting which appear vivid to the eye in a picture 
 gallery are the blues, the reds disappearing early. One 
 thing must be said, however, that after a certain 
 illumination is attained, the relative brightness of 
 colours does not alter, so that if we use a very bright 
 light for our measurement, or even a light equal to 
 one candle power, placed at, say, six feet distance, we 
 are not doing an injustice to any colour. 
 
 To know diflSculty is to surmount it, but another 
 crops up, which is also physiological. If we place a 
 blue-green and a red in juxtaposition in large masses, 
 the blue-green may appear much brighter than when 
 the same colours are presented in small portions (we 
 are only taking these as typical colours), and there 
 must be some reason for the change. The reason is 
 to be found in the fact that we have a yellow dye 
 staining a small central part of the retina, and that this 
 dye cuts oif a large proportion of the blue when the 
 light forming the image has to pass through it, as is 
 the case when the masses are small. When the masses 
 are large, the image partly falls on the retina beyond 
 this yellow spot, and the full effect of the colour is felt in 
 this outside portion. In a picture, then, it is evident that 
 the luminosity of a small patch of blue will not be the 
 same as it would be if the patch were large. We can 
 imitate what is seen by the retina beneath the yellow 
 spot by placing a pale yellow glass over half a patch of 
 blue. The blue is very largely cut off by the yellow 
 
112 LIGHT IN PHOTOGRAPHY. 
 
 glass, as it is by the yellow spot. When we examine a 
 picture, only a small area will be in focus on the retina, 
 and this area will fall on its centre, where the yellow 
 pigment is situated. The picture is practically taken 
 in by a movement of the eyes, so that each area in 
 succession is examined, and it is the sum of these 
 impressions which gives us the mental image of the 
 picture. No one can doubt that if we are to measure 
 the brightness of a colour in a picture, we ought then 
 to receive the colour so that practically it will fall upon 
 the yellow spot ; and if we do so, we shall be on the right 
 road to obtain measurements which are practically useful. 
 If we wish to reproduce by photography the various 
 luminosities of colour in a picture — of course, in black- 
 and-white — we should seek to render them as nearly 
 as possible as they appear on that area. Now it may 
 be considered that a measurement of the brightness of 
 the pigments as used in a picture is unnecessary, and so 
 it is if we measure the brightness of, say, four standard 
 pigments in the same light, and are then able to repro- 
 duce their luminosities accurately in black-and-white by 
 means of photography; for if these can be reproduced, 
 then all other colours will also be accurately reproduced. 
 Most fortunately, as we have already said, the light 
 from the positive pole of the electric arc light is a very 
 close approximation to sunlight of ordinary character, 
 and if we use this as the light illuminating the pigments 
 we shall not go far wrong in our measurements, and 
 hence we use this quality of light as a standard light. 
 The standard pigments are not diiBcult to choose. If 
 we take a vermilion, an emerald green, and a French 
 ultra-marine, together with chrome yellow, as a further 
 
LUMINOSITY OF PIGMENTS. II3 
 
 test, and, having measured their brightness, can repro- 
 duce them faithfully in black-and-white by means of 
 photography, we shall also be able to reproduce any 
 other colours correctly on the same kind of plates. 
 What these plates are to be we leave for the present ; 
 all we wish to demonstrate is the way in which the 
 brightness can be measured in such a standard light. 
 
 We may practically adopt the very same arrangement 
 as we did for measuring the colours with lamplight, 
 using the electric light for the purpose ; but we must 
 exclude the arc as far as possible, as this radiates violet 
 light to some extent. What we ought to use is light 
 from the hottest part of the positive pole, which we call 
 the crater. We can place the electric lamp in a box 
 in which an orifice is cut sufficiently large to hold a 
 lens which may throw a fairly large image of the crater 
 on an opaque screen, on which a circular hole of, 
 say, one-eighth of an inch is cut. The screen can be 
 moved so that the crater just falls upon the hole, and the 
 only light which then traverses the screen is that which 
 we require. Now we have to devise some means by 
 which this beam can be cast directly on the coloured 
 square, and also as a comparison light to .proceed 
 from another point at an angle with the white square. 
 We can effect this by placing in the path of the beam 
 a piece of plain flat glass close to the hole, and 
 reflect a portion at an angle of about 45" with the 
 original direction, and at the distance necessary to give 
 the proper parallax we may place a silvered mirror, and 
 reflect the beam on to the squares which are side by 
 side (as in Fig. 39), the direct beam also falling on 
 them. The rod placed in position, as in the same figure, 
 
 I 
 
114 LIGHT IN PHOTOGRAPHY. 
 
 will cast two shadows, one of which gives much feebler 
 illumination than the other. In the stronger one we can 
 place the sectors. As before, we shall make at first one 
 square — say the coloured — appear too dark, and then, 
 by altering the aperture of the sectors, make it appear 
 brighter than the white. Somewhere intermediate 
 between the two the brightness or luminosity will be 
 equal when this point is arrived at, and the aperture 
 noted. The coloured square also, as before, can be 
 replaced by a white square, and equality of brightness 
 again established. The ratio of the two apertures of 
 the sectors is the ratio of the brightness of the pigment 
 to that of white. Suppose in one case the aperture 
 used to make the brightness of the pigment the same 
 as that of the white was 35", and to make the two 
 whites equally illuminated 140°, then the pigment would 
 only reflect iVir, or 25 percent, of the light that the white 
 did. If sunlight be available, a similar artifice may be 
 employed, but in this case the light should be first 
 reflected from the surface of a plain glass to avoid the 
 illumination being too great. A plain mirror, as 
 described, should be placed in the path of the beam 
 which falls on the squares, and the second mirror 
 should be used as before. (There is no necessity to 
 have an image of the sun in this case, as there is no 
 arc light to be cut off.) The sizes of the squares, 
 coloured and white, should not be more than one inch, 
 and the eyes should not be less than three feet away 
 when the equality is being judged, for reasons stated at 
 the beginning of this chapter. 
 
 We can get a very fair approximation to the brightness 
 or luminosity of a pigment in daylight by the following 
 
LUMINOSITY OF THE SPECTRUM. IIS 
 
 artifice. Cut a rectangular hole, say half inch by one inch, 
 in a piece of blackened card, and place the coloured 
 square on the table, or against the wall, so that it 
 occupies half the rectangle forming a coloured square 
 of half an inch side. Bring beneath the other half greys 
 of different depths, and note one which appears lighter 
 at first glance. Then insert a deeper one, and note 
 that it appears too dark. Then try a grey of inter- 
 mediate tint. It will soon be ascertained which is the 
 grey which is the nearest match. A very suitable 
 series of greys can be made in a sensitometer such as 
 Spurge's by exposure of platinotype paper, though, of 
 course, the same can be effected by other means. The 
 white contained in these greys can be measured as 
 described in Chapter X., and thus the luminosity of the 
 different pigments can be estimated. 
 
 The subject would be incomplete if an idea were not 
 given as to how we may estimate the relative luminosity 
 of pure spectrum colours. 
 
 The accompanying diagram (Fig. 46A) will give a very 
 good idea of the method employed. The source of light 
 may be the sun reflected by a heliostat on to Li, or it 
 may be the carbon poles of the electric light. In 
 either case the image is focussed on the slit Si, which 
 can be of any desired width, say tbVo, or even jV of an 
 inch. The lens, L^, makes the rays parallel as they 
 emerge to fall on the prisms, Pi and P^. A lens, L^, 
 attached to a camera, B, with a swing-back, as shown, 
 makes a spectrum on D, in which there is a slit, Sj, 
 through which any ray may be caused to emerge by 
 moving D in the spectrum. A lens, L^, when D is 
 removed, collects all the rays of the spectrum, and by 
 
Ii6 
 
 LIGHT IN PHOTOGRAPHY. 
 
 giving it a slight tilt, as shown, a colourless image of 
 the first surface of Pi is thrown on a screen, F. When 
 D is inserted, any ray of light coming through Sj forms 
 a patch of approximately monochromatic light at F. 
 
 /// 
 //I 
 
 Fig. \(ia. 
 
 Part of the beam which strikes Pi is reflected to G, a 
 silvered mirror which reflects the light on F, and in its 
 path we place a lens, L5, which also forms an imageof the 
 first surface of Pi. The coloured and the light images 
 
LUMINOSITY OF THE SPECTRUM. ll'J 
 
 overlap at P. If we place a rod near F, a coloured 
 stripe may be made to fall alongside a white stripe, the 
 latter being made purposely the brighter. At M are 
 placed the sectors, and the two stripes are made to be 
 equally luminous to the eye, and the aperture at which 
 this occurs is noted, and so on for each part of the 
 spectrum. 
 
 There is a small part of the apparatus not shown in 
 the figure. A very small part of the beam is made to 
 pass beyond Pa, and this is caught by a mirror and 
 sent to illuminate the back of D, on which is a fine 
 scale. Projecting from the back of the camera is a 
 vertical needle, and the shadow of this is thrown by 
 the beam last mentioned on the scale. The scale 
 indicates at what position S3 is situated when the scale 
 numbers shown for the issue of, say, the lithium, and 
 sodium, and magnesium lines are known. The 
 apparatus is thus complete. With a collinator of about 
 one foot, and the lens, L3, of about thirty inch focus, a 
 spectrum more than three inches in length is obtained. 
 Fig. 47 gives the luminosity of the spectrum of the 
 crater of the positive pole of the arc light. 
 
 The luminosity is shown in the vertical heights, and 
 the maximum luminosity is taken as 100, the others 
 being reduced to that scale. The small figure is the 
 part from o to 22 of the scale, the vertical scale being 
 enlarged ten times. In the above figure we have the 
 luminosity of the spectrum as measured by the central 
 part of the retina. The wave-lengths of the scale 
 numbers of the prismatic spectrum can be readily 
 obtained, owing to the fact that very approximately the 
 square of the reciprocals of the wave-length can be 
 
ii8 
 
 LIGHT IN PHOTOGRAPHY. 
 
 found from the length of the spectrum. If the square 
 of the reciprocals of (say) the two lithium lines are set 
 up on a scale, which represent the distances apart of 
 
 Luminosity. 
 
 
 
 
 ; 
 
 
 
 O 
 
 
 o 
 
 
 
 O 
 
 
 
 o 
 
 0) 
 
 
 o 
 
 
 
 -o 
 
 
 
 
 
 
 
 
 o 
 
 
 
 
 o 
 
 
 
 0(0 
 
 - 
 
 
 
 
 I 
 
 1 ' 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 - 
 
 T 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 r 
 
 
 1 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 ■' 
 
 
 _ 
 
 
 _ 
 
 
 J 
 
 
 1 1 
 
 _ 
 
 
 _ 
 
 
 
 
 
 L^- 
 
 1 
 
 
 
 _ 
 
 i_ 
 
 
 
 f- 
 
 
 
 
 _ 
 
 _ 
 
 \-i 
 
 — 
 
 u 
 
 
 rj 
 
 
 Z! 
 
 T 
 
 - 
 
 
 T 
 
 -■ 
 
 "- 
 
 "n 
 
 " 
 
 
 - 
 
 
 
 
 
 
 -|- 
 
 - 
 
 - 
 
 
 
 - 
 
 - 
 
 
 
 ~\ 
 
 
 - 
 
 -^ 
 
 ^ 
 
 - 
 
 - 
 
 ~ 
 
 
 
 > 
 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 r^.;ir 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 r'. ' 
 
 m 
 
 I ; 1 
 
 ! 1 
 
 
 
 1— 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 — 
 
 
 
 
 
 
 
 
 T ' 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 u 
 
 
 I'll 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 U) 
 
 
 1 1 1 1 
 
 
 
 
 
 r— 
 
 ■" 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 « 
 
 - 
 
 
 
 r 
 
 ^ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 - 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 N 
 
 
 
 
 / 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 - 
 
 ^ 
 
 
 
 
 O 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 \ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 > 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 "v 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 »o 
 
 
 
 
 
 
 
 
 
 ~-. 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 ' 
 
 
 1 
 
 N 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 ""T:- 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 o 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 -^_ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 p-^l" 
 
 
 
 
 
 
 
 
 
 
 
 
 ■^ 
 
 
 
 
 
 =)^5 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 dT " 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 l« 
 
 
 
 
 
 
 
 
 
 
 ■-. 
 
 
 
 
 
 
 
 
 B— 
 
 
 
 
 
 
 
 
 N, 
 
 
 
 
 
 
 
 
 
 
 k 
 
 
 
 
 
 
 
 
 
 
 
 s 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 V 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 ^ 
 
 
 
 
 ^ 
 
 
 
 
 
 
 
 
 
 
 
 
 s 
 
 
 
 
 
 <rt 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 N 1 
 
 
 "< 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 k, 
 
 
 
 pj 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 ^ 
 
 
 "l 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 ^ 
 
 
 7^' 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 s 
 
 
 o 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 '^i 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 \ J 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 \\ 
 
 <o 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 ' 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 ' 
 
 
 
 
 
 
 ' 
 
 
 
 
 
 
 
 
 
 
 
 
 
 lO 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 1 
 
 \ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 i 
 
 V 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 ■^ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 \ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 N 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 \ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 - 
 
 f 
 
 C4 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 \ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 ] 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 o 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 — 
 
 
 
 
 
 ] ' 
 
 
 
 
 
 
 1 
 
 
 
 
 
 N 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 r 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 ' 
 
 
 '? 
 
 
 
 
 
 
 1 
 
 
 
 
 
 
 i 
 
 
 
 
 
 
 
 _ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 I 1 
 
 T 
 
 kjci 
 
 
 
 
 
 
 
 - 
 
 
 
 
 
 
 
 
 
 - 
 
 - 
 
 -J 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 .i. 
 
 
 h 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 , 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 ~r 
 
 
 J<« 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 , 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 f4^ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 l' 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 L 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 [ 
 
 
 
 
 
 
 
 
 
 
 .. 
 
 
 U 
 
 
 
 
 
 
 
 I- 
 
 
 
 
 
 
 
 
 
 
 
 
 ..i 
 
 ft -^ 
 
 these lines in the spectrum, and the tops be joined 
 by a straight line, the square of the reciprocal of the 
 
CORRECTNESS OF THE METHOD. II9 
 
 wave-length of any scale number will be found by 
 measuring the height of this line about it. This can 
 be either calculated, or measured from a diagram (see 
 Appendix for tables of luminosities). 
 
 It should be noticed that in the arrangement shown, 
 the fluctuations in the source of light are equally felt 
 in the spectrum and in the white light with which the 
 colour is compared. This is a great point when using 
 sunlight, since any small temporary passing haze is 
 thereby discounted. 
 
 It may be asked how we know that the method 
 we use in measuring luminosity is a correct one. We 
 can answer this very readily — there is a strip of red in 
 the spectrum which, combined with a blue green, gives 
 a white light of the same quality as the comparison 
 light. If we make this artificial white light by using two 
 apertures in D (Fig. 46A), and measure the luminosity of 
 each component separately, and add them together, we 
 shall find that they are together equal to the luminosity 
 of the artificial white light. No one doubts that white 
 light can be measured against white light. By taking 
 a yellow and a blue in the same way we add additional 
 proof to the correctness of the method. 
 
CHAPTER XII. 
 
 Effect of Temperature on Sensitiveness. 
 
 We have already shown that a phosphorescent plate is 
 increased in luminosity by the application of heat. 
 Let us make an experiment to impress this fact once 
 more on the mind. A plate covered with calcium 
 sulphide (Balmain's luminous paint), exposed to a fairly 
 strong light for a second or two, will be found to glow 
 when taken into a dark-room. If a warm body, such 
 as a flat iron, be applied to the back of the plate whilst 
 it is glowing, an image of the part so applied will at 
 once be seen as an increased brilliancy of phosphor- 
 escent light. On removing the iron, if we watch the 
 plate, it will be noticed that the phosphorescent surface 
 will, after a short time, become of equal brilliancy 
 throughout, but that after a further lapse of time the 
 portion which the iron warmed will become darker than 
 the rest of the plate. Let us consider the reason of 
 these changes of luminosity. The molecules of the 
 phosphorescent material have had transferred to them a 
 certain amount of energy by their exposure to light. A 
 part of this energy they store up, and gradually give out 
 
PHOSPHORESCENT LIGHT. I2f 
 
 again as phosphorescent light. At the part of the 
 plate where the flat iron is applied, the molecules or the 
 atoms were made to oscillate more violently by the 
 warmth imparted, and consequently there is an increase 
 of phosphorescence, but at the expense of the original 
 energy stored up. It follows that, when the heat is 
 withdrawn (or, indeed, the same would occur if the iron 
 was still kept in contact with the plate), the light energy 
 at that portion having been largely consumed, rapidly 
 becomes darker than the surrounding portions. It 
 might be thought that the flat iron had communicated 
 fresh energy ; but another experiment will show that any 
 energy it does give is not such as can be given out as 
 light. If a plate, covered with the same material, but not 
 exposed, be taken into the dark-room, and the hot iron 
 be applied as before, the heating of a portion of the plate 
 will have no effect whatever — the blackness of the plate 
 will be unaltered. What we are wishing to inculcate is 
 this — viz., that a heating effect may stir up the molecules 
 to greater motion, and thus allow the atoms a greater 
 swing and greater freedom of path. If this be so, then 
 sensitive salts should, under such circumstances, be more 
 sensitive when heated than when exposed at ordinary 
 temperature, for what is true for one kind of matter 
 ought to be true for another. 
 
 We may press a hot flat iron against the back 
 surface of a gelatine bromide plate, and examine 
 the film by reflected daylight. The colour of the 
 sensitive salt will be seen to be altered, becoming 
 deeper. On the plate cooling, the film resumes its 
 normal colour. Place iodide of silver in an intense 
 cold, such as produced by carbonic acid snow and 
 
I2Z LIGHT IN PHOTOGRAPHY. 
 
 ether, and the iodide will be found to have a grey 
 colour, quite different to that it assumes at a normal 
 temperature. Such changes of colour mean something, 
 and that "something" ought to show itself when 
 exposing these sensitive compounds. 
 
 Now, can we prove that this is the case ? A theoretical 
 deduction should always be capable of experimental 
 demonstration. If we heat a portion of an ordinary 
 sensitive plate, and expose the cold and hot portion to 
 the same light, and develop, we can see at once whether 
 the hot portion is more sensitive than the cold. The 
 same heated flat iron may be used as before. We can 
 lay it on black paper lying against the back of a gelatine 
 plate, and then expose to some feeble light for a second 
 or two. If we at once proceeded to develop the plate, 
 it might very correctly be said that those parts of the 
 plate which were heated would render the developer 
 more active, and that consequently we ought to get a 
 greater reduction of silver. To guard against this, the 
 plate should be put on one side till it is perfectly cool 
 all over, and then the development may be proceeded 
 with. The image of the hot iron will be perfectly 
 evident, indicating that the increased temperature of the 
 sensitive material during exposure has increased the 
 sensitiveness. But gelatine is a treacherous substance, 
 and it may be thought that the heat has in some way 
 or another altered the sensitiveness of the material 
 permanently, so we can repeat the experiment of heating 
 the plate as before, but in this instance we should allow 
 it to cool before exposing it to the feeble light. On 
 development there will be no trace of any action of 
 the heated body. We will now give an example 
 
EFFECT OF HEAT ON COLLODIO-BROMIDE. 12 J 
 
 of how photographers may fall into pitfalls, and be 
 unable to extricate themselves, unless they know 
 something of the theoretical aspect of photographic 
 action. In the days immediately preceding the advent 
 of gelatine plates the photographer worked with 
 collodio-bromide emulsion, and had he experimented 
 with plates coated with emulsion, and used the flat 
 iron to heat a part, he often might get exactly the 
 opposite results to that which he would obtain with a 
 gelatine plate. The result of such an experiment might 
 have been to cause him to believe that the increase 
 in the temperature of a plate lessened its sensitiveness, 
 and that the theoretical reasoning, which would make 
 it the reverse, was altogether wrong, and that there was a 
 screw loose somewhere. No doubt we can picture him 
 looking at his developed plate, and uttering the wise 
 saw — wise, that is to say, when properly applied — 
 that an ounce of practice is worth a pound of theory. 
 But, after all, the good old practical photographer would 
 have been wrong, as he very often is. It is lucky that 
 he is so often wrong, as if he were less frequently so it 
 may be opined that the pages of photographic literature 
 would be considerably curtailed from lack of matter. If 
 the old practical photographer had known that a certain 
 amount of moisture in the film, or an equivalent to such 
 moisture, was a necessity for sensitiveness, he would not 
 fall into the error. Suppose we had instructed him to coat 
 a plate with collodio-bromide, and to dip it all into cold 
 water, next to plunge one end only into hot water, and 
 then to expose to the feeble light, he would have found 
 that where the hot water had been in contact with the 
 sensitive surface the plate would show increased 
 
124 LIGHT IN PHOTOGRAPHY. 
 
 sensitiveness compared with that on which the cold 
 water rested, and this would have been on all fours 
 with the experiment narrated with the gelatine plate. 
 
 We may take it, then, as an axiom, that by increasing 
 the temperature of the sensitive surface we increase the 
 sensitiveness. The converse of this should also be true, 
 that if we diminish the temperature we should also 
 diminish the sensitiveness. Let us make an experiment 
 as to this : — Warm up the dark room to about go° F., 
 and instead of heating the flat iron, let us immerse it in a 
 mixture of snow and salt or pounded ice and salt. The 
 metal will acquire a temperature considerably below 
 freezing point, and when thus chilled we can place it in 
 contact with the back of a plate which has the temperature 
 of the warm room, and expose as before to the feeble 
 light. When the part of the plate which was lowered 
 in temperature has regained the heat which it had 
 lost, we can develop, and we shall find that the image 
 of the iron is less dense than the surrounding parts of 
 the exposed surface, and consequently we may assume 
 that the plate has become less sensitive by the lowering 
 of the temperature. This may not be quite conclusive ; 
 it may be said that the water in the gelatine was frozen, 
 and that therefore the conclusion might not be justified. - 
 A reference to Daguerreotype plates will settle that. 
 Sensitise two of them as we were taught to do, and freeze 
 one and not the other, and expose them side by side to 
 the light. When both become of equal temperature, 
 develop them together. We shall find that it is always 
 the cold plate that shows the feeblest deposit of 
 mercury. It may be asked, why expose two plates, 
 and why not use a portion of one plate ? The reply is. 
 
PHOSPHORESCENCE OF BROMIDE OF SILVER. 1 25 
 
 that a Daguerrotype plate, being metal, is a good 
 conductor, and that therefore the cold would be com- 
 municated throughout. 
 
 We have now shown that a low temperature is adverse 
 to the greatest speed. 
 
 Professor Dewar recently showed that the speed of a 
 film at the extremely low temperatures with which he is 
 now accustomed to work ( — i8o to — zoo Centigrade) 
 is reduced to, at least, a quarter of what it is at ordinary 
 temperatures. The writer, in a communication to the 
 Camera Club, gave evidence that between the tempera- 
 tures of 100° F. and o" F. the speed of a rapid plate was 
 reduced by even more. 
 
 In regard to a supposed phosphorescence of bromide 
 of silver at these moderate temperatures, we may say 
 that we have fair evidence that it is absent. The 
 writer has arranged a phosphorescope in a manner 
 different to that ordinarily employed. A beam of the 
 electric light of the diameter of a threepenny bit was 
 allowed to fall on a screen, and in the path was placed 
 a revolving disc in which twenty-four apertures of five 
 degrees each were cut out. At an angle with the beam 
 of light was placed a tube, through which the eye looked 
 at the screen. When the beam of light was cut off from 
 the screen by the opaque part of the sector, the eye would 
 see the screen through one of the apertures ; whereas 
 when the beam of light fell on the screen, the screen 
 was invisible to the eye, the view being cut off by 
 another portion of the sector. The screen was replaced 
 by a sensitive plate, and the disc caused to revolve i oo 
 times a second, so that an aperture passed through the 
 beam 2,400 times a second. Allowing that the screen 
 
126 LIGHT IN PHOTOGRAPHY. 
 
 was cut off from the eye, and also from the beam for 
 the equal intervals, it follows that any phosphorescence 
 which lasted a little more than — iVoth part of a second 
 would have been visible ; but not a glimmer was 
 visible. If there be any phosphorescence at all, it 
 must last less than this interval, and we can scarcely 
 credit that any will be found if the revolutions are still 
 further increased. This is, of course, by the way. 
 
 Having thus arrived at a definite conclusion by means 
 of qualitative experiments, we turned our attention to 
 devising quantitative experiments with a view of forming 
 some idea of the increase and diminution of the sensi- 
 tiveness by known heat and cold. The next few pages 
 will give the results. 
 
 The method of exposure was as follows : The apparatus 
 (Fig. 48) used was a simple box (B B), with a smaller tin 
 box (A A), when cold is required inside within it. Between 
 the box and the tin was placed a mixture of ice and salt(I), 
 which reduced the temperature of the box to o" F. about. 
 This is further facilitated by placing in the tin box a 
 similar mixture up to a level which is regulated by the 
 plate (P). The plate was suspended in a pair of grooves cut 
 in two stays (V V), which are pendant from the lid. In the 
 lid a small hole is cut to allow the bulb of a thermometer 
 (T) to be thrust to the level of the plate, and it thus 
 acquired the same temperature as the plate. A square 
 hole (A) was cut in a sliding batton which moved across a 
 slot-like aperture cut in the lid, and over this was a mirror 
 (M) supported at an angle of 45 ° to reflect light vertically 
 down from a horizontal beam of light (R R) on to the 
 plate. The source of light being at some distance from 
 the mirror, the light impressed the plate with an image 
 
EXPOSURE OF A PLATE TO A LOW TEMPERATURE. 1 27 
 
 of the hole (in this case square), and as the hole is 
 moved along horizontally, so diflferent parts of the plate 
 can be exposed for different periods. The plate was 
 exposed at the ordinary temperature of the dark-room, 
 next at — 1 8° C, and then at about 34° C. This last was 
 effected by removing A A, and heating the box, by placing 
 at the bottom a heated brick to keep the temperature 
 
 Fig. 48. 
 
 constant during exposure. The order of these exposures 
 could be varied, of course, but in all cases the same 
 length of exposure was given to each series, so as to be 
 strictly comparable. The source of light used was either 
 a small argand paraffin light, or else the amyl acetate 
 lamp which has been so much used by myself for these 
 and for other similar experiments. The square aperture 
 has a means, of course, of being closed and opened 
 when required. The plate is backed with black paper 
 to prevent the light transmitted through it from being 
 
128 LIGHT IN PHOTOGRAPHY. 
 
 reflected from the ice, and causing fog. These small 
 matters, it may be stated, were the outcome of practical 
 experience. 
 
 It would, of course, be useless to give details of every 
 experiment made, but three examples are recorded : 
 first of all where the temperature employed varied 
 between 33° C. and about — 18-5 C. These tempera- 
 tures it was difficult to obtain in every case quite 
 exactly, but the variation of a degree makes no practical 
 difference in the general result. From the diagrams 
 and tables given it will be seen that there is not 
 necessarily any variation in the gradation of the curves ; 
 that is, their general slopes are usually parallel to one 
 another, as might be expected, since the varying 
 exposures are made by altering the time, but that the 
 rapidity is altered — but not to the same degree — for 
 each kind of plate. In every case the rapidity is 
 invariably less with those cold temperatures which the 
 photographer sometimes meets with. The exposures 
 were successively doubled, and for convenience these 
 successive exposures are made the scale of abscissae. 
 Thus 10, 20, 40, 80 seconds' exposure are shown as i, 2, 
 3, 4, and the transparency of the deposit is shown on 
 the vertical scale as ordinates, 100 being total trans- 
 parency. For a lantern plate (Fig. 49), the distance 
 apart of the "hot" and "cold" curves is I'l of the 
 scale divisions, which implies, as each successive scale 
 number means double the exposure of the previous one, 
 that the chemical decomposition at — 16° C. is 2 '^•' 
 times, or 2-3 times less than that at 4-33° C. With the 
 " lightning" rapid plate (Fig. 50) it is just double, whilst 
 with a Fitch film (Fig. 5 1 ) i' is only 2- ' times, or i -6 times. 
 
KFFECT OF COLD ON SENSITIVENESS. 129 
 
 These numbers indicate that all degrees of sensitive- 
 ness are not equally affected. The " lightning " plate 
 and the slow lantern plate are not very far off one 
 another, but the film which is intermediate is so much 
 less that there must be some reason for it. Whatever 
 may be the cause, we find, however, that when time of 
 
 Mawson Lantern Plate Exposed to Paraffin Lamp at 5 ft., visually 
 equal to an Amyl Acetate at 39 inches. 
 
 Fig. 49. 
 
 exposure, multiplied by the intensity of the light, is a 
 constant quantity, the chemical action is different,- 
 and that the colder the surface the less sensitive it 
 becomes. 
 
130 LIGHT IN PHOTOGRAPHY. 
 
 A reference has been made to the effect of heat and 
 cold on a collodio-bromide plate, and experiments were 
 made with it to ascertain the diminished rapidity between 
 17" C, and + 32" C. Only two exposures were made 
 at these temperatures, 12 minutes and 6 minutes. A 
 
 Lightning Plate Exposed to Amyl Acetate at 6 feet. 
 Fig. so. 
 
 scale was made at ordinary temperature, from which to 
 determine the difference in rapidity. 
 
 From this experiment we gather that the same opacity 
 as is got with the hot exposure is only obtained when 
 the cold exposure is increased by J-|^ or i^iSi or 
 4*5 times (see Table IIIa). 
 
DEFINITION OF GRADATION. 
 
 '3' 
 
 In these and other experiments we shall use the word 
 " gradation " to express the slope of the curves in their 
 straightest part. This is a proper word to use when it 
 is considered that the transparency of deposit and time 
 of exposure are expressed by these curves. 
 
 ■■■■ 
 
 A Fitch Film Exposed at 5 ft. to an Amyl Acetate Lamp. 
 
 Fig- 51- 
 
 From several experiments made, in which the " hot " 
 exposure was above 3 3 *" C, it appeared that the gradation 
 changed. A lantern plate was exposed at 4 feet from a 
 paraffin lamp at — iS" C, and the heat increased to 
 40". C. (see Table IV.) 
 
132 LIGHT IN PHOTOGRAPHY. 
 
 Fig. 52 shows that there is an alteration in the grada- 
 tion, although it is a time exposure. As we shall see 
 from a subsequent chapter, the gradation given by time 
 exposures under any ordinary conditions to the same plate 
 is the same — that is, the curves are always parallel to one 
 another. 
 
 J i^BLJE IV 
 
 
 9C 
 
 manB 
 
 
 8C 
 
 KSMiaa 
 
 
 V ■'C 
 
 ■Hnn 
 
 
 
 
 u 6( 
 
 ■■a— 
 
 
 < 
 
 iHBm 
 
 
 g4C 
 
 ■■rasHB 
 
 
 3t 
 2( 
 
 1 
 
 
 
 5 10 aO 40 .80 160 320 
 1 EXPOSURE , 
 
 The following will help to throw light upon the cause 
 of the different gradation : — 
 
 A Mawson lantern plate was exposed to a paraffin lamp 
 at s feet (visually equal to an amyl lamp at 33 inches). 
 The exposures were made in the following order (see 
 table V) :— Expose at + iS^ C, next at — 18° C, next 
 at + 38° C, and finally at + 18° C. 
 
ALTERATION IN GRADATION. 
 
 '35 
 
 It will be seen in Fig. 53 that in this case we have a 
 probable explanation of the different behaviour of the 
 curves. A plate was exposed at an ordinary tempera- 
 ture, then to the cold, and the curves so obtained are 
 parallel to one another. The temperature was then 
 
 Fig- Si- 
 
 raised to 38° C, and the curve is of the same character, 
 as regards gradation, as given in Fig. 52 — that is, it 
 is not parallel to other two curves. When the plate 
 was cooled down, and before it had time to absorb 
 moisture, an exposure at the ordinary temperature was 
 again given, when its cui-ve became parallel to the " Itot " 
 cur ve. In the same way a plate was exposed as follows :— 
 
134 LIGHT IN PHOTOGRAPHY. 
 
 I St to — 1 8°, 2nd to ordinary temperature, 3rd to +38°, 
 4th to — 18". The first two curves and the last two 
 were parallel to each other, but these last showed a 
 steeper gradation. Evidently by heating to 38° we 
 have something altered in the plate, which alteration 
 remains after it is cooled for some little time — long 
 enough, indeed, to allow another complete set of 
 exposures to be made after the cooling is complete, 
 and before it resumes its normal condition. From this 
 it appears' that up to 33" C, at all events, the heat has 
 no effect on the plate beyond making it more rapid. 
 
 The natural and most probably correct conclusion to 
 draw from this is, that at near 3 8 " the water which is in the 
 gelatine begins to be eliminated, and alters the condition 
 of the film. There is nothing unusual in this combina- 
 tion of water with gelatine. Professor Guthrie some 
 years ago made exhaustive researches into what he 
 called chryohydrates, and showed that water, although 
 not what is usually called chemically combined, still 
 had the power of acting almost as if it were so com- 
 bined. At — 18° C, or near zero F., it is quite possible 
 that the amount of water present would be such that it 
 would not freeze, leaving the water in the state in which 
 it can take its usual part in combining with the halogen 
 when released from the sensitive salt. 
 
 The question of whether exposure to different intensities 
 of lighty&r a fixed time was equivalent to a fixed intensity 
 and variable times, was a matter which next called for 
 attention, for it cannot be too well remembered that it 
 is the former we have to deal with in a camera exposure, 
 and not with a variation in time. There were some 
 difficulties in carrying out these experiments, but they 
 
INTENSITY AND TEMPERATURE. IJS 
 
 were gradually smoothed away. A scale consisting of 
 squares of gradation of known transparency was laid 
 upon the plate. The light was admitted through a thick- 
 ness of blotting paper, which took the place of a portion 
 of the lid of the box, and thus equal illumination of the 
 whole length of the plate was obtained. Only one 
 example need be given (see Table VI.). 
 
 inri 1 1 1 1 
 
 
 ■Rp^n 
 
 
 
 
 MHsna 
 
 
 ■1 
 
 
 .i J T 3 to / 
 
 INtETNSITIES IN POWERS OF 2 
 
 Pig- 54- 
 
 The light in the two cases was not exactly the same, 
 being rather stronger for the " cold " exposure than for 
 the " hot." But as it was not sought to get the absolute 
 values, but only the gradation, this is immaterial. 
 
 It will be seen from Fig. 54 that in this case, as in 
 
136 LIGHT IN PHOTOGRAPHY. 
 
 all Others tried, the " hot " curve is steeper than 
 the " cold " curve. To sum up these experiments, we 
 see that the effect of heat is to increase the apparent 
 rapidity of the plate with, time exposures, and to 
 increase the steepness of gradation of the plate with 
 different intensities of light. 
 
CHAPTER XIII. 
 
 Effect of Intermittance of Exposure. 
 
 The experiments which led the writer to try the experi- 
 ments described in the Chapters XIV. and XV. were 
 based on a difference which was observed in the amount 
 of chemical action obtained from a continuous and 
 an intermittent exposure, the time during which the 
 sensitive surface was exposed being the same in both 
 cases. Now let us see whether, theoretically, there 
 should be any difference, or, at all events, admitting 
 there is a difference, whether it can be accounted for. 
 
 Suppose we have the amplitude of the swing of an 
 atom of a molecule increased by the impact on it of a 
 ray of light, and further, suppose that the light is 
 suddenly shut off, the tendency of the atom will be to 
 revert to its normal swing ; if we again admit the ray of 
 light for another brief interval, the atom will again 
 increase in its swing, but perhaps just not sufficient to 
 allow it to swing away from the attraction of the molecule. 
 The swing or amplitude attained by the atom will not be 
 
138 LIGHT IN PHOTOGRAPHY. 
 
 SO great as it would be when the two exposures were 
 combined into one, for in the latter case there would be 
 no falling back towards the normal swing. In the first 
 case, then, the atom might not leave the molecule ; in 
 
 f'g- 55- 
 
 the last case it might do so. The minimum number of 
 exposures have been supposed, for simplicity's sake, but 
 the same argunients hold when the exposures are 
 increased. 
 
EXPOSURE SLIDE. 
 
 139 
 
 To make the experiments it is convenient, though 
 not necessary, to have a special slide, which will enable 
 any number (within limits) of exposures to be made on 
 the same plate. The slide, as made by Meagher with 
 his usual ability, is shown in Fig. 55. It is really an 
 ordinary back for 7 J by 5 plates, which worked between 
 
 Pig- 56- 
 
 two grooves C C, as shown, the plate when the front was 
 opened being protected by D, to which the projecting 
 piece E is attached. The front view is shown in 
 Fig. 56. A thin metal batten, F, can pass through 
 E, in which a slot is cut, as shown. In the batten a 
 
140 
 
 LIGHT IN PHOTOGRAPHY. 
 
 hole, K, is pierced, which can be moved to any part 
 of the slot. The slide itself is raised or lowered by 
 a rack and pinion, a scale being attached at the side 
 on C, to register the movement. It will be seen that a 
 large number of exposures may be given on the same 
 plate by raising the slide and moving the hole K. In 
 a later pattern F has been done away with, and six 
 holes pierced in E instead of the slot, all or any one of. 
 
 100 
 90 
 
 80 
 70 
 
 >-i 
 3 
 tn 
 
 H 
 40 
 
 30 
 20 
 
 Relative mtensibes. 
 Pig- 57- 
 which can be covered up at pleasure. Carriers for half-, 
 5 by 4, and quarter-plates are provided for the slide. 
 In the subsequent experiments this piece of apparatus 
 was employed. 
 
 [" 
 
 ^H^HI 
 
 I^B 
 
 hhhhI 
 
 IHI 
 
 i^i^hI 
 
 Ijjllj 
 
 ^h^hI 
 
 
 H 
 
 E 
 
 BH 
 
IKTERMITTENT EXPOSURES. I4I 
 
 A Striking result in intermittent exposures was 
 obtained with a gelatino-bromide lantern plate. Two, 
 sets of exposures were given at different distances from 
 the source of light. Exposures at each distance were 
 made (ist) with the naked light and (2nd) with a 
 rotating sector intervening, having twelve apertures of 
 S" each; the exposure in the second case had to be, 
 therefore, six times longer than in the first case, to 
 make the total exposures equal, for the total number of 
 apertures was only equal to 60°. From the note which 
 the apertures gave when a current of air was projected 
 against them, it was found 720 exposures were given 
 per second. 
 
 Fig. 57 shows graphically this result (see Table VII). 
 
 The apparent loss of exposure by the intervention 
 of the sectors increases as the actual intensity of the 
 light diminishes, the ratios of the single exposures 
 to that of the intermitting exposures being — 
 
 For intensity i . . . . i to "815 
 
 „ „ i . . . . I „ -500 
 
 „ -iV • • . • I ,- '423 
 
 )) >i «* • • • • ' >> 37° 
 
 (See Table VII.) What shape the curve of these 
 ratios would assume when the intensity of the light 
 is greater or less than that shown, can only be 
 guessed at. One point which this experiment makes 
 clear is that there is no error introduced by any want 
 of accuracy in the apertures of the sectors which can 
 account for the great differences in the apparent 
 exposures, for were it so the ratio would not be variable 
 as it is. 
 
142 LIGHT IN PHOTOGRAPHY. 
 
 We have thus arrived at the conclusion that the sum 
 of small exposures is not equivalent to one exposure for 
 the same length of time. 
 
 Now the question arises whether a long rest after a short 
 exposure, and then another exposure, and again a long 
 interval of rest, is conducive to minimize chemical action. 
 The reasoning is not difficult when we take the same 
 4ines as we took in our last argument. A long interval 
 of rest after a short exposure must inevitably give a 
 longer time to those atoms which are not swung out of 
 the molecules to come back to their original swing ; 
 hence, the shorter the time of exposure is to the time of 
 rest, the less chemical action there should be. This 
 was tested experimentally. 
 
 For this purpose a sector with moveable apertures 
 was employed, and the exposures were made with the 
 sectors set to different apertures. As the sectors had 
 two apertures opposite one another, the maximum 
 aperture was 2 x 90° — that is, 180°, but without the 
 intervening sectors it is 360". The plate was a slow 
 gelatino-bromide, and different small portions were 
 exposed to an amyl lamp for times depending on the 
 aperture. A time scale added as before enabled the 
 chemical action resulting to be measured. The con- 
 tinuous curve in rig.58 shows the results (seeTable VIII.). 
 
 It will be noticed that although equivalent exposures 
 were given in each case, there is a gradual reduction of 
 chemical action as the aperture is diminished. As the 
 speed of rotation was the same in all cases — viz., 48 per 
 second, this indicates that the longer the period of rest 
 between two exposures the less action there is, as our 
 theory led us to suppose would be the case. 
 
INTERVALS BETWEEN TWO EXPOSURES. 
 
 '43 
 
 Another slow plate was next tried in a similar manner, 
 but two series of exposures were given to it, one with a 
 speed of fifty revolutions of the sector per second, and 
 the other once a second. These two series of exposures 
 were given to ascertain what effect the time of revolution 
 
 40 
 
 30 s 
 
 
 10 "9 -8 7 -6 'S -4 -3 -2 -I 
 Ratio of time of exposure to time of rest. 
 
 — that is, the shortness of exposure — had. It will be . 
 seen from the dotted curves in Fig. 58 that the difference 
 in chemical action between intermittent exposures and a 
 single one was less with the slower rate than with the quick 
 rate of revolution, as theoretical considerations foretell 
 (see Table IX.). It happens that the curve given by the 
 slow rotations in this experiment almost coincides with 
 the curve given by the quick rotation in the last one. 
 
144 LIGHT IN PHOTOGRAPHY. 
 
 With the fast revolution the ratio of the chemical 
 action given by exposure through the i o" aperture to the 
 continuous exposure is 3'3J, and for the slow only 2'o. 
 With a quicker plate the ratios were "j and "S- 
 
 The effect of varying the length of each individual 
 exposure, the ratio of interval of exposure to that of rest 
 remaining the same, was found by the following 
 experiment. A rotating drum, the light to which was 
 admitted through an adjustable slit, was brought into 
 requisition. The slit was kept at a fixed width of -aV inch, 
 and the exposure was made to the light of a paraffin 
 lamp, placed 27 inches away from it. The drum was 
 covered with fairly sensitive bromide paper, and in all 
 cases was exposed for the same length of time. As the 
 drum was iz'zs inches in circumference, it follows that 
 the ratio of the time of exposure to the time of rest was 
 about I to 250 in all cases, whatever the velocity of 
 rotation might be. A scale, as usual, was made on the 
 same paper, and five exposures were given with the drum 
 rotating at different speeds (see Table X.). 
 
 Fig. 59 shows these results graphically. The theoreti- 
 cal conclusions, as epitomised at the beginning of this 
 chapter, again seem amply justified. 
 
 These results show that apart from any influence 
 that ratio of the time of exposure to the time of rest 
 may exercise, the absolute duration of exposure affects 
 the result. When, as in this case, the absolute duration 
 is 200 times greater in one than in the other, the 
 chemical action is rather more than doubled. 
 
 It has already been shown that the length of the 
 interval which elapses between successive and equal 
 exposures has an effect upon the apparent exposure, 
 
EXPOSURES AND PERIODS OF REST. 
 
 '+5 
 
 but it became necessary to trace the result when the 
 exposures were equal in length, but the intervals of rest 
 between such exposures differed in amount. To make 
 this experiment, discs of thin black paper were cut 
 but, and radiating from the centre of these discs, 
 
 40 . 
 
 30^ 
 
 20 „ 
 
 a 
 
 
 S-J 
 
 40 30 20 10 
 
 No. of rotations of the drum per second. 
 
 Pig- 59- 
 apertures having angles of 10" were cut out at equal 
 intervals round the centre. The discs used had 1, 3, 6, 
 12, and 24 apertures cut out, as above, respectively. 
 These were held in position between two glass discs, 
 and rotated in front of the sensitive surface for periods 
 which make the product of intensity and time the same. 
 A scale of density was likewise impressed on the 
 surface, for the purpose of arriving at the apparent 
 exposures. 
 
 During all the exposures uniformity of rotation of 
 
 L 
 
146 
 
 LIGHT IN PHOTOGRAPHY. 
 
 the discs was secured, and the only difference between 
 them is the alteration in the period of rest between 
 successive insolations. The number of revolutions 
 made was 43 per second. 
 
 In Fig. 60 the continuous curve shows that as the 
 number of apertures is diminished, the less does the 
 
 no" 100" 90" 80° 70° 60° 50° 40° 30" 20" 
 Intervals between exposures. 
 
 Fig. 60. 
 
 10" 5 O" 
 
 chemical action (as derived from the scale) become (see 
 Table XL). 
 
 This is again more fully shown in a slow plate. It 
 was exposed four feet away from the amyl acetate lamp, 
 and the apertures used were the same as in the last, but 
 another single aperture of 10" was added. Moreover, 
 two sets of exposures were made, one with the sectors 
 
INCREASE AND DIMINUTION OF ATOMIC SWINGS. 147 
 
 rotating at the speed of one revolution in aVth second, 
 and the other with one revolution in one second. 
 
 The results are also graphically shown in Fig. 60 by 
 the two dotted curves, with the exception of the last 
 exposure to the single aperture of 10°. The quicker 
 rotation in this case gives the greater diminution in 
 the apparent chemical action. 
 
 We have now answered the questions we proposed, 
 and it will be seen that theoretical considerations are 
 not to be despised even in a branch of applied science, 
 such as is photography. We have by these experiments 
 gained an idea of the relative times necessary to increase 
 the amplitude of a swing, and for the swing to subside. 
 
CHAPTER XIV. 
 
 Effect of Small Intensities of Light. 
 
 We have next to deal with an effect of light which is, 
 if not unexpected, at least contrary to generally received 
 opinion. When a plate is exposed for a certain time 
 to a light of given intensity, we say commonly that a 
 plate has received so much exposure. If the time is 
 altered, and the intensity also, so that the exposure (time 
 multiplied by intensity) is the same, we have been taught 
 that the same chemical effect will ensue, or, to put it in 
 other words, the exposure will be the same. Now, as 
 matter of fact, this is not true in every case, though it 
 is true under the conditions under which a photographer 
 usually works. Let us look at the matter from the 
 beginning. We have in our sensitive salt the atoms 
 composing it vibrating, and when light acts on it we 
 have the oscillations increased, and the amplitudes 
 greater. We can scarcely conceive that there should 
 be any difference of effect produced by a series of 
 impulses of a certain strength prolonged over a certain 
 time, and a series of impulses of a fraction of that 
 
ENERGY OF LIGHT. 149 
 
 strength, lasting over a proportionally longer time. 
 Yet there is a difference, as experiment shows. A 
 perfectly satisfactory explanation of this deviation from 
 what might be expected is at present wanting. We 
 may, however, look upon the matter in the following 
 way. Suppose we have a series of waves of light 
 falling on a particle, which is, of course, composed of 
 an almost infinite number of molecules, and that in the 
 oscillation of these last, some one atom feels the impact 
 of one wave of light, and the extra swing so given is just 
 able to send it off from its mother molecule, then that 
 molecule will be liable to decomposition by the developer. 
 Now supposing we have a series of waves carrying half 
 the energy, it might be supposed that it would require 
 two of these blows to send the atom away from the 
 molecule ; but it by no means follows that the same atom 
 would receive the second blow. Some other atom 
 might be in the line of the waves and get the blow 
 instead, or, if the. original atom received it, it might 
 not be in that part of the swing which would render the 
 blow most effective. In either event the atom would not 
 leave, but would be kept in the group of the mother 
 molecule. In the case cited the intensity of the 
 light in the first would be double that of the 
 second, but the time of exposure would only be half 
 that of the other. It will be noticed, of course, that 
 this explanation depends on the fact that a certain 
 amount of " light energy " has to be expended in 
 separating the atoms from the molecules. This is a 
 common experience in photography, for a very short 
 exposure to a bright light, or an exposure to a very 
 feeble intensity for a longer time, will fail to show any 
 
ISO LIGHT IN PHOTOGRAPHY. 
 
 signs of chemical decomposition on a plate. Th« 
 question is one of averages, and can be explained on 
 the doctrine of chances. The general look of the 
 curves given by the absolute measures support this 
 view, since it is somewhat of the character of the law 
 of error. 
 
 There is still another explanation into which it would 
 be difiScult to enter, unless one were to make high 
 mathematics play a considerable part, but it may be 
 briefly put in ordinary language. When Helmholtz 
 gave his explanation of anomalous dispersion he founded 
 it on the theory that when matter (molecules or atoms) 
 was in contact with the ether or wave medium, although 
 the latter might be considered as frictionless, yet when 
 the two were together, a certain amount of viscosity or 
 resistance to motion might be experienced. If, then, 
 an atomic swing was increased by the motion of the 
 ether waves, there nevertheless would be a retardation, 
 and the full amount of energy would not be expended in 
 increasing the swing, but part would be used up in over- 
 coming what may be called the inertia. If this be the 
 case, then, the increase in amplitude of swing of the 
 atoms by a large number of small shocks would not be 
 so great as that produced by a smaller number of greater 
 shocks, though the same energy might be expended in 
 the two cases. In the former, the increased swing 
 might be sufficient to send the atom out of the mother 
 molecule; in the latter, it might not. These two 
 explanations are offered, either of which, or both 
 together, seem to be applicable. 
 
 The writer had often suspected the failure of the 
 assumption made, that on the same sensitive surface, 
 
BRITISH ASSOCIATION ADDRESS. ISI 
 
 when the product of intensity of light and time were 
 equal, the chemical action would be the same, before 
 he made any quantitative experiments. Thus, in his 
 presidential address to the Mathematical Section of the 
 British Association, he stated : " One word as to a 
 problem which we may say is only qualitatively and not 
 quantitatively solved. I refer to the interchangeability 
 of length of exposure for intensity of light. Put it in 
 this way. Suppose that with a strong light L a short 
 exposure E be given, a chemical change C is obtained : 
 will the same change C be obtained if the time is only 
 an «th of the light L, but with n times the exposure } 
 Now this is a very important point, more particularly 
 when the body acted upon is fairly stable, as, for 
 instance, some of the water-colour pigments which are 
 known to fade in sunshine, but might not be supposed 
 to do so in the light of an ordinary room even with 
 prolonged exposure. Many experiments have been 
 made at South Kensington as regards this, more 
 especially with the salts of silver ; and it is found that 
 for any ordinary light, intensity and exposure are inter- 
 changeable ; but that when the intensity of light is very 
 feeble — say the n)aooo6 of daylight — the exposure is 
 rather more prolonged than it should be, supposing the 
 exact interchangeability always held good ; but it has 
 never been found that a light was so feeble that no 
 action could take place. Of course it must be borne 
 in mind that the stability of the substance acted upon 
 may have some effect ; but the same results were 
 obtained with matter which is vastly more stable than 
 the ordinary silver salts." In his address to thfe 
 Photographic Society on the 8th November, 1892, he 
 
IS* 
 
 LIGHT IN PHOTOGRAPHY. 
 
 also said, " And I would point out that up to the present 
 time it has been held that if time and intensity of light 
 gave a certain chemical change in a body, then the 
 same will be produced if the two multiplied together 
 give the same constant. I think we must put this law 
 on the same basis as that of gravity . . . When we 
 get intensities of light which are almost infinitely 
 feebler than those with which we are accustomed to 
 deal, the ' time into intensity ' law will be found to break 
 down. 1 his follows almost from physico-mathematical 
 considerations." 
 
 In the following experiments, which are culled from 
 the original paper in the Photographic Journal, vol. xviii.. 
 
 Fig. 6l. 
 
 page 56, the source of light most usually employed was 
 an amyl acetate lamp, known as the Siemens' unit light, 
 as any variation in it is excessively small. The point 
 to attend to is to light the lamp some ten minutes before 
 
EXPERIMENT WITH PAPER. I S3 
 
 it is to be used, and adjust it so that the tip of the flame 
 just reaches the gauge pointer a b (Fig. 6i). After this 
 period it will remain at the same height almost as long 
 as any amyl acetate remains in the reservoir. Every 
 part of the flame could always be seen from each part 
 of the surface exposed. Fig. 6i gives a section and 
 plan of the lamp. 
 
 The first experiments that it is proposed to describe 
 are those relating to the amount of chemical action 
 produced by different intensities of light. An obvious 
 experiment was to expose the same sensitive surfaces 
 at different distances from a source of light, increasing 
 the time of exposure as the distance was increased. 
 
 In the first instance, the sensitive surface was bromide 
 paper of medium rapidity. Small squares, on a quarter- 
 plate size, of this sensitive material were exposed at 
 different distances to the amyl acetate lamp. The times 
 were calculated so that the product of intensity and time 
 (/X /) was a constant, however much /or / varied, which, 
 in accordance with the assumption usually accepted, 
 should give equal chemical action as shown by the 
 blackness on development, / being the intensity of the 
 light, and / the time of exposure. A scale of exposures 
 was added on the same paper by exposing different 
 parts for known times, the distance of the lamp 
 remaining constant. The paper was then developed 
 carefully, and fixed. When dry, the scale of blackness 
 was measured, as also was the blackness of the small 
 squares, which had been exposed at difi"erent distances, 
 and the latter referred to the former. 
 
 The measurement of the greyness produced was made 
 by the method which the writer has described in 
 
154 LIGHT IN PHOTOGRAPHY. 
 
 Chapter X. The lowest curve in Fig. 62 gives the 
 
 results. If there were no alteration in chemical action 
 
SLOW And fast plates. 155 
 
 the curve should be a straight horizontal line (see Tables 
 XIII. and XIV.). 
 
 As a fact the energy usefully expended at twenty-four 
 feet was less than a quarter of that expended at two 
 feet on the sensitive surface, although the time of ex- 
 posure was proportionally lengfthened to make the 
 intensity and time a constant. 
 
 A lantern plate more sensitive than the foregoing 
 paper was similarly treated as in the last exposure, and 
 the top curve in Fig. 62 shows the results. The energy 
 usefully expended at 24 feet was only fth of that 
 expended at 2 feet (see Tables XV. and XVI.). 
 
 The foregoing are decidedly " slow " sensitive sutfaces. 
 The next experiment was with one of Wratten and 
 Wainwright's plates. The amyl-acetate flame was 
 reduced in height, so that the result does not abso- 
 lutely compare with the two former. Here the energy 
 usefully expended at 24 feet is only i of that expended 
 at ij- feet. The results are shown in Fig. 62 (see 
 Tables XVII. and XVIII.). 
 
 It might be imagined that these differences were due 
 to atmospheric absorption, as the decrease is so regular. 
 It will, however, be seen that such cannot be the cause, 
 since the ratios of the decrease are by no means the 
 same when the sensitive surface is altered. The more 
 rapid the plate, the less divergence from the assumed 
 law has, as a rule, been found. There are means, how- 
 ever, besides altering the distance from the sensitive 
 surface, which allow of a direct proof. 
 
 A Spurge's sensitometer, consisting of fine holes of 
 different diameters in the centre of different small square 
 diameters, and at the same distance from the plate, was 
 
'56 LIGHT IN PHOTOGRAPHY. 
 
 brought into requisition. Every third hole doubles 
 the area through which the light is admitted to the 
 sensitive surface, and the same kind of results was 
 obtained— viz., that the smaller the hole the greater was 
 the extra time required to effect the same chemical action 
 on the plate. 
 
 The next series of experiments to be recorded are 
 those which were undertaken to ascertain if there is any 
 difference in the gradation (see page 133 for the 
 definition of the term), when different intensities of 
 light were employed at fixed distances, and progressive 
 limes of exposure were given, so that time X intensity was 
 kept the same. Only two experiments out of a number 
 will be quoted as indicating the results always obtained. 
 
 We give as examples a slow lantern and a quicker 
 plate. They were exposed to a light at 2 feet and 8 feet, 
 the times of exposure being i to 16. 
 
 Fig. 63 gives the plotted curves, and each pair of 
 curves are parallel. The "gradation" is therefore 
 practically the same. (For measures see Tables XIX. 
 and XX.) It will be noticed that, though the curves 
 are parallel, they do not coincide (in a Euclidean 
 sense), as they would do if the chemical action were the 
 same for the same constant exposure. 
 
 We may sum up these results so far as to say that with 
 slow and moderately quick plates the intensity of the 
 light acting on a plate alters its apparent rapidity in a 
 marked manner, but the alteration is smaller as the 
 rapidity of the plate increases, until with a very rapid 
 plate no practical alteration is made at all. 
 
 As we shall see in the next chapter, the intensities of 
 light we are considering are moderate ones, and when 
 
GRADATION OF PLATES. 
 
 '57 
 
 we come to very intense lights the conditions and the 
 results are changed. The camera image is formed by 
 different intensities of light, and a fixed time exposure. 
 
 234 
 Scale Nos. 
 
 Pig- 63. 
 
 It follows that the gradation of some plates will be 
 different in the camera to what it would be if the image 
 was impressed by varying the times of exposure. 
 
CHAPTER XV. 
 
 Action of very Intense Light on Plates. 
 
 There is another phenomenon to which reference has 
 been made in the last chapter, which is that when the 
 light acting on a plate is very intense, the results 
 approximate to those which are found with very feeble 
 intensities — that is, that the chemical action produced 
 is lessened for a constant exposure (time multiplied by 
 intensity) when the light is very intense, and that con- 
 sequently there is some one intensity which will give 
 the greatest chemical action. We must recall our 
 former picture of the atomic swings to form an idea of 
 how this may occur. The removable atoms are swinging 
 with their ordinary swing, and a series of waves of light of 
 great amplitude fall upon them. Let us suppose that the 
 passage of two waves through the space occupied by the 
 molecule is sufficient to destroy it as before by swinging 
 out one of the atoms. Now let us double the energy 
 carried by the waves ; then the energy of one wave 
 ought to be sufficient to cause the same effect, but the 
 one wave would pass through the space in exactly half the 
 
HEAVY BLOWS AND RESISTANCE. 159 
 
 time that the two waves took. The chances that the atom 
 was at such a period of its swing that the energy could 
 be properly applied to effect this result would be half in 
 the latter case than in the former. The same kind of 
 result would obtain with a larger number of waves. 
 
 In the case of low intensities the chances of " hit or 
 miss" of an atom by a wave are the same, but with 
 them we have to use the argument that a large number 
 of blows are suflScient to increase the swing to shift the 
 atom from the molecule, whilst with high intensities the 
 number of blows required is reduced to a minimum, and 
 the times in which the blows can be given are propor- 
 tionally reduced. We thus have an investigation in 
 which the' calculation rests on the chances of a sufficient 
 number of blows being effective in a given time, the 
 strength of the blow being inversely proportional to the 
 time. An investigation of this kind will show that 
 there is some one energy due to wave motion that will 
 in a given time decompose the greatest number of 
 molecules. 
 
 We gave anotherpossible explanation of the phenomena 
 also when the intensities of the light acting were feeble. 
 We took Helmholtz's explanation of anomalous disper- 
 sion, and applied it to our molecules. In this case we 
 have a viscosity or resistance to motion to deal with. 
 If we suppose that a comparatively heavy blow given to 
 an atom will not give the full swing to it that it would 
 take supposing there were no resistance to the motion, 
 we can conceive that Avith a still heavier blow the 
 resistance would be further increased (somewhat in the 
 same way that atmospheric resistance to the passage 
 of a body increases more rapidly than the velocity). 
 
l6o LIGHT IN PHOTOGRAPHY. 
 
 and the length of swing given would be proportionately 
 curtailed. 
 
 Thus, if an atom received two blows, one double the 
 energy of the other, it is possible that the blow of double 
 energy would not increase the swing of the atom to 
 the extent it would reach from two blows of the other. 
 This being the case.-there would be some amplitude of 
 wave which would give the greatest effect, while exposure 
 (time multiplied by intensity) remained constant. 
 
 It is not proposed to give many experimental details 
 through which the above result was arrived at ; the 
 original papers in January and March 1894 of the 
 Camera Club Journal should be consulted if more are 
 required. In the experiments given in the last chapter 
 no great intensity of light was employed, and the varia- 
 tion in intensities was given by altering the distances of 
 an amyl-acetate lamp from the plate, but in no case 
 was the light closer than i J feet. With these different 
 intensities, which are feeble compared with those now 
 under consideration, it was shown that slow plates were 
 markedly different in their sensitiveness. In other 
 words, though the product of time of exposure and 
 intensity of light might be the same, yet the opacity 
 produced on development of a series of exposures on 
 the different parts of the same plate was always less as 
 the intensity became feebler. With rapid plates no 
 marked difference resulted, however. 
 
 The times of exposure, be it remarked, lasted several 
 seconds, the least being 5 seconds ; but we often take 
 a photograph in -sVth part of a second, which might 
 be supposed to mean that the same amount of chemical 
 action should be obtained in the densest parts of the 
 
USE OF A SCALE, l6l 
 
 negative in that fraction of a second as would be 
 obtained in (say) loo seconds to the feebler light. 
 That, roughly, may mean that one light is 5,000 times 
 brighter than the other. 
 
 If by any means we can get deposits produced by 
 varying intensities of light acting for such a short 
 period of time as the Ath of a second, and then produce 
 others with feebler light, but having the same pro- 
 portional variation in intensity, we can compare the 
 two gradations of transparency with great facility. We 
 can make a scale of var)'ing transparencies (». e., to allow 
 different intensities of light to pass) by exposing a plate 
 to an amyl-acetate lamp for varying times, and developing 
 it. The deposits will allow different intensities of light 
 to pass to a plate which, is in contact with it. 
 
 We can measure the light transmitted by this scale 
 of transparency by the method given before, and if the 
 method of development has caused a jet black deposit, 
 the optical value and the photographic value of the 
 density will be the same. If we wish to make sure of 
 this, we can print a positive from this scale by exposing 
 a rapid plate behind it at a fixed distance from a light, 
 and then, on the same plate, at the same distance from 
 it, make another scale by giving varying time exposures 
 to different small squares. The measurements of the two 
 will show whether the photographic and visual opacities 
 are identical. 
 
 Having got a scale, the next point to arrange is to 
 give a short exposure. In the first experiment that will 
 be described, this was accomplished by utilising an 
 extemporised drop-shutter. 
 
 The shutter was placed close to the scale, with the 
 
 M 
 
1 62 LIGHT IN PHOTOGRAPHY. 
 
 aperture parallel to its length. The plate was in contact 
 with the scale. An electric light of some i ,000 visual 
 candle power was placed 5 or more feet away from it, the 
 necessary distance being found by trial. An exposure was 
 given, the time of passage of the aperture in the shutter 
 over the scale being measured. Other exposures through 
 the same scale were given on different parts of the plate 
 to the comparatively feeble light of an amyl-acetate 
 lamp placed at 3 ft. 6 in. from the negative, for varying 
 times. It must be remembered that in all cases the 
 variations in the intensities of the light passing through 
 the scale were in the same proportion. After develop- 
 ment we have upon the same plate an electric light 
 exposure and several amyl-acetate lamp exposures. 
 (Although the visual intensity of the electric light is 
 only 1,000 caindle power, yet photographically it is at 
 least equal to 10,000 candles.) From these last we can 
 pick out two— one in which the opacity given by the 
 passage of the light through the most transparent square 
 of the positive is nearly that given by the passage of the 
 electric light through the same square, and the other 
 ii which the opacity of the most opaque squares is the 
 same. In all cases, even with the most rapid plates 
 that were tried, the gradation given by the electric 
 light exposure was much more gentle than that given 
 by the amyl-acetate. The following are two examples 
 of this, which are noteworthy — one is a Barnet plate 
 (ordinary), and the other an Edwards plate (special 
 rapid). See Fig. 64. ist. For both plates the amyl- 
 acetate lamp was placed at the same distance from the 
 standard scale of density, and in both cases one of the 
 exposures selected and measured was for 30 seconds. 
 
r63 
 
164 LIGHT IN PHOTOGRAPHY. 
 
 By the method of determining the comparative speeds of 
 plates, which was proposed by the writer at the Camera 
 Club Conference, the Barnet plate from this exposure 
 is about 2'" or rg times more rapid than the Edwards 
 plate. 2nd. The Barnet plate was exposed through 
 the scale at 4"s feet distance from the electric light, 
 whilst the Edwards plate was exposed at 8 feet. This 
 makes the intensity of illumination given to the Barnet 
 
 plate — f- , or 3 • 1 1 times more than to the Edwards plate. 
 
 The electric (arc) light was giving out approximately 
 the same amount of light in both cases, and the time of 
 exposure was also the same. Had they both received 
 the same exposure in this case, the rapidity of the 
 Barnet plate would have been almost exactly twice that 
 of the Edwards plate, but it received about three times 
 more light. The Edwards plate is therefore r^ times 
 more rapid for intense light than is the Barnet, though 
 with the low illumination the Barnet is about twice as 
 rapid as the Edwards plate. This appears to indicate 
 that the relative rapidities of plates may alter according 
 to the intensity of the light in which they are employed 
 (see Table XXL). 
 
 A further development of the research was in using not 
 only the electric arc light, but the electric spark. An 
 electric spark has the advantage of emitting a very intense 
 light, but of an excessively short duration. What the 
 visual intensity may be, one can scarcely say ; but the 
 intensity must be many times greater than that of sunlight, 
 and the fact that it scarcely makes any greater impression 
 in the retina than a beam of sunlight, owing to its short 
 duration, is a proof that other sensitive matter besides 
 
METHOD OF EXPOSURE. 165 
 
 silver salts is not affected proportionately to the intensity. 
 The exposure, it was proved, could not be more than 
 the TooVstr of a second, and might be considerably 
 less. To ascertain this, a 5 inch plate was rotated 
 round its centre about 60 times in a second, and 
 iri ifront of it sharply defined small apertures were 
 placed lying 2 inches from the centre of rotation. A 
 spark illuminated the plate through the apertures — 
 I St, when at rest, and 2nd, during revolution, and no 
 trace of movement could be detected. Now, if the 
 spark had lasted a time sufficiently long to have 
 allowed the plate to move the iJu- of an inch it would 
 have been readily detected. The plate at 2 inches from 
 the centre moved about 750 inches per second, or 1 inch 
 in the yi-o second. It therefore moved tJo- inch in. 
 TTTuoo second. 
 
 In the experiments that are to be described, the-, 
 spark was taken between the terminals of a spark- 
 gap apparatus, the knobs being usually -J of an inch . 
 apart. One terminal was in connection with the inside, 
 and the other with the outside coatings of five Leyden 
 jars, which were charged by a Wimshurst machine.. 
 The machine was worked till the spark passed across.; 
 the gap, the plate being exposed, through the graduatecj 
 screen, to its action. The operation was naturally^ 
 carried out in a darkened room. Another portion of 
 the same plate was exposed to the action of the electric 
 arc light as before described, and another portion to an 
 amyl-acetate lamp (in both cases, of course, through the 
 graduated screen), and a "time scale" of density was 
 also impressed by the amyl-acetate lamp. 
 
 Fig. 65 gives the result of exposing an ordinary plate 
 
i66 
 
 LIGHT IN PHOTOGRAPHY. 
 
 to two sparks at 1 8 inches, then to the electric light at 
 5 ft. 6 in. distance for -Ju second, and, lastly, to an 
 amyl-acetate lamp for 5 minutes at 3 ft. 6 in. from the 
 plate (see Table XXIII.). 
 
 The quick plates, which were mostly experimented 
 with, were a brand of snap-shot plates, manufactured 
 by Cadett and Neall, and labelled as 133 H. and D. 
 
 I ^ J + 5 6 ^ 
 
 Scale of Intensities in Powers of 2. 
 Fig. 65. 
 
 As the lowest number for a snap-shot plate is i.oo, it 
 may be presumed that this brand of 1 3 3 is a very rapid 
 one, and thus was very interesting to experiment with. 
 With these plates the spark-gap was placed 18 inches 
 from the plate, and exposure given. The plate was 
 
ALTERATION IN GRADATION. 
 
 167 
 
 then transferred to the drop-shutter apparatus, and 
 exposed for the lio of a second to the electric light at 
 a distance of 7 feet, which was found to give about the 
 same general intensity of image as the spark, and it was 
 then exposed at 5 feet, for two minutes, to an amyl- 
 acetate lamp. Thus the exposure given to the last was 
 at least 12,000,000 times greater than the first, and 
 12,000 times that of the second. 
 
 A time scale taken with the amyl-acetate lamp at 
 10 feet distance was also impressed. 
 
 The results are seen in the following diagram (see 
 Table XXIV.) :— 
 
 8c 
 60 
 
 
 
 
 
 
 1 
 
 2C 
 
 1 
 
 H 
 
 ..■ 
 
 H 
 
 1 
 
 ^ 
 
 1 
 
 1 
 
 
 ■ 
 
 1 
 
 1 
 
 1 
 
 1234567 
 
 Intensities in Powers of 2, 
 Fig. 66. 
 
 Similar exj eriment^ were carried out with other plates, 
 and with a like result. 
 
 The figures show that the more intense the light the 
 less steep gradation within the limits used there is. 
 
J 68 LIGHT IN PHOTOGRAPHY. 
 
 ; The next set of experiments are full of interest, and give 
 the curious but not unexpected results already referred 
 to. They demonstrate that if the intensity of the light is 
 increased beyond a certain limit, the product of intensity and 
 time being kept constant, the energy used up in the work of 
 chemical decomposition diminishes. The first experiment 
 that was made was with a snap-shot plate, exposed naked 
 at 1 8 inches and 1 2 ft. distance, the exposures being i spark 
 and 64 sparks respectively. The writer was not prepared 
 for the enormous difference in results ; the one spark 
 showed but a feeble deposit, whilst the exposure at 1 2 ft. 
 distance to 64 sparks caused opaqueness. Another plate 
 (an Ilford ordinary) was exposed (without any graduated 
 screen intervening) at 12, 9, 6, 3, and i^ feet distance, 
 and the illumination of 64, 36, 16, 4, and i sparks — the 
 equivalent number of sparks— were given respectively 
 at these distances ; a time scale was also impressed on 
 the same plate, and the whole developed together. 
 Now, although the absolute equivalent exposure cannot, 
 of course, be told by applying the measures of trans- 
 parency given by the spark exposures to the measures of 
 the time scale, yet their relative values can be. Taking 
 the result of the exposure at 1 2 feet as i, that at 9 feet is 
 •87, at 6 feet "72, at 3 feet '41, at ij- foot ■22. 
 
 This shows that there-is apparently less than a quarter 
 of the energy usefully employed in producing chemical 
 action at ij ft. as compared with that at 12 ft., a result 
 which seems somewhat marvellous. It might be thought 
 that this result was due to " reversal " of Jhe image, but 
 other experiments proved that it was not so. 
 
 Another experiment made was with the same brand 
 of plate, but the exposures were made through the 
 
EXPOSURE THROUGH A GRADUATE!? SCALE. 
 
 • 6gi 
 
 Standard scale at the same distances, omitting those at 
 9 ft. and 3 ft. The number of sparks multiplied by the! 
 inverse square of the distance is the same in all cases. 
 The results of these exposures are given in Table XXVI., 
 and a similar result is to be seen in Fig. 67, and whea 
 
 I - 3 4 S 6 7 
 
 Scale of Intensities in Powers of 2. 
 
 Fig. £7. 
 
 looked at diagrammatically, show at once that there is 
 no case of reversal in it. If there had bee:i the feeble 
 intensity given by the most opaque square of the scale 
 of density would have indicated a greater intensity than 
 the most transparent one. But the best indication that 
 there is no reversal is the absolute frjcdom from halation 
 which is shown on this plate, whilst the most opaque 
 
170 LIGHT IN PHOTOGRAPHY. 
 
 portions, with the 64. sparks, show decided halation. 
 Several brands of plates were experimented with, and 
 the same results were invariably obtained. 
 
 So far the action on rapid plates has been described, 
 but to make the matter more complete, slower plates 
 were experimented with. A slow lantern plate was 
 exposed at different distances with an equivalent number 
 of sparks, and the results were instructive. Although 
 the exposure at i foot appeared a little less than at 8 ft., 
 there is far less difference than when the quicker plates 
 are used. This is what was to be expected. In the 
 preceding chapter it was shown that with these plates the 
 energy expended with feeble intensities and equivalent 
 exposures was much less than with stronger intensities, 
 whereas with the rapid plates there was not much 
 difference. The rapid plate must therefore evidently 
 arrive sooner at the point where the energy expended 
 was at its maximum than would the slow plate. 
 
 Combining these experiments, we are able to deduce 
 the fact that for every plate there is an intensity of light 
 which has a maximum effect upon it with a given 
 exposure, and that on either side of this maximum the 
 useful energy decreases. 
 
CHAPTER XVI. 
 
 Measurement of Spectra. 
 
 If we turn to page 25 and look at Fig. 9, we have there 
 a photograph of the spectrum taken on three kinds of 
 sensitive surfaces. No. II. is a plain bromide plate, 
 which apparently shows greatest action about the violet 
 marked G, whilst No. III. is taken on a bromide plate with 
 a maximum action near the extreme limit of the violet H. 
 No. I. is a spectrum taken on a new plate introduced by 
 Mr. Cadett, which is sensitive far into the red. This 
 last plate is an orthochromatic plate, and contains at 
 least two, and probably more, kinds of sensitive matter, 
 the sensitiveness in the yellow and red being due to the 
 presence of colouring matter. Photographs such as 
 these may be called qualitative, as they do not indicate 
 the exact degree of sensitiveness to each ray. But by 
 calling into use the measuring apparatus already 
 described in Chapter X., we can, if we attach a scale of 
 opacities to it — the opacities being caused by different 
 intensities of light acting for the same time — measure 
 the opacities at each part of the spectrum and that of 
 
172 
 
 LIGHT IN PHOTOGRAPHY. 
 
 the scale, and, by comparing the former with the latter, 
 deduce from the spectrum opacity the intensity of photo- 
 
 
 graphic light that was acting at each part. We will 
 give one example of this. In Fig. 68 we have three 
 
SUNLIGHT AND GASLIGHT. I 73 
 
 photographs of the spectrum. Nos. I. and II* are 
 photographs of the spectrum of sunlight, the difference 
 between them being that No.. 1. is taken with a wide slit 
 to obliterate the fine lines which traverse it when the slit 
 is narrow. No. II. was taken with a fine slit. No. I. is the 
 spectrum that was measured, and II. used as a "position- 
 finder," that is to tell us what parts of the spectrum we 
 were measuring. No. III. is the spectrum of a gas flame, 
 in which it will be seen that the blue and violet rays are 
 largely absent. No. IV, is a scale of opacities which 
 were measured, and a curve of light transmitted drawn. 
 The intensity of the light causing the deposit being 
 known, was used as the base on which the curve was 
 erected. These photographs were measured in the two 
 ways indicated in Chapter X., and the results are very 
 close to one another. 
 
 Fig. 69 gives the result of these measures, and we can 
 at once compare the effect of sunlight and gaslight on 
 the same plate. The wavy appearance of the curves is 
 somewhat remarkable, but the bends probably mean 
 either inactive absorption of some kind when there are 
 depressions, and active absorptions (that is where work 
 is done) in the rises. It will be noticed that generally 
 the depressions and rises are in the same positions in 
 the gas and solar spectra curves. In the figure a third 
 curve of the luminosity of sunlight to the eye is 
 introduced, from which it will be seen that the solar 
 spectrum would be very far short of rendering properly 
 the luminosities of the colours as seen in daylight. 
 
 Again, we give another example of a plate known as 
 Lumifere's red-yellow sensitive plate, the spectra being 
 taken of the crater of the positive pole of the arc light 
 
»7+ 
 
 LIGHT IN PHOTOGRAPHY. 
 
 ooooooooo 
 
 ^DO i^\o »or^ roN •-* 
 
 
 
 
 
 
 
 
 
 >; 
 
 ■i 
 
 
 
 
 
 
 
 
 
 
 ■ii 
 
 
 
 '■-. 
 
 --.r^- 
 
 — 
 
 — .- 
 
 LsriT." 
 
 ~^ '_ 
 
 ^'.v-*' 
 
 >^ 
 
 ■■:" 
 
 "^tlL 
 
 — . 
 
 \ 
 
 
 K 
 
 
 
 
 Q- 
 
 '•*■ 
 
 '^}y 
 
 
 ^. 
 
 ■\ 
 
 \ 
 
 
 
 
 
 
 
 s^yjj,-^ 
 
 ''"--^ 
 
 ^ol/}^ 
 
 '"i" 
 
 
 ""*) 
 
 
 (il- 
 
 
 
 
 
 
 ^ 
 
 ^ 
 
 I 
 
 
 
 
 
 
 
 ; 
 
 
 
 
 \ 
 
 
 
 
 
 /^ 
 
 .^ 
 
 
 
 
 \ 
 
 / 
 
 j 
 
 
 ^ 
 
 ^ 
 
 
 
 
 
 
 
 1 
 
 ^ 
 
 
 
 
 
 
 
 
 si 
 
 1 
 
 
 ^^ 
 
 \ 
 
 
 
 
 
 
 \ 
 
 
 
 
 
 
 
 
 
 
 
 ! 
 
 1 
 
 
 f 
 
 / 
 
 
 
 
 
 
 
 \ 
 \ 
 
 
 
 
 7 
 
 
 
 
 
 
 V 
 
 /A'-v- 
 
 
 
 
 
 
 
 
 "^ 
 
 
 
 
 
 
 
 
 
 
 
 
 \ 
 
 
 
 
 
 
 
 
 
 
 - 
 
 J 
 
 ,5 
 "3 
 
 = g 
 u 
 
 1.3 
 
 '5 
 
 
 ."^Q 
 
 s; 
 
 n cb 
 
 :i 
 
PHOTOGRAPHY OF THE SPECTRUM. 175 
 
 and of gaslight ; and as a third curve we have the visual 
 luminosity of the arc light. It will be seen that neither 
 one nor the other photographed spectra lie at all near 
 the visual curve. It follows, then, that the colours 
 cannot be photographed to give correct lumiiiosity 
 without some extraneous aid. 
 
 The curve of sensitiveness of ordinary bromide is 
 practically the same as the curve in Fig. 70 to the left 
 of E, which, it will be seen, is far and away the worst 
 for rendering colours in their true luminosity, though 
 it should be remarked it conveys the idea of other wave- 
 lengths besides those absolutely in tune with the swing 
 of the atoms being capable of increasing the amplitude. 
 
 We can, by cutting off rays which are not required, 
 produce photographs which give colours very nearly their 
 correct luminosities. The diagrams of the plates which 
 are most sensitive throughout the spectrum are seen in 
 Figs. 69 and 70. It will be noticed in them that there are 
 spaces where the sensitiveness is smaller than at places 
 on each side of them. So far as we know, there are no 
 plates which give a continuous smooth curve. The 
 difficulty of these gaps may be got over to a very large 
 extent, because the colours in nature are mixed colours, 
 occupying a considerable space in the spectrum, and 
 the gaps can thus be "bridged over." If we can obtain 
 a photograph on them which will render truly the 
 luminosities of a red, a green, and a blue, that are mixed 
 colours, we shall be able very approximately to photo- 
 graph any other colours in nature. It is not necessary that 
 the colour should be reflected. It may be transmitted, 
 so that if we can photograph correctly the luminosities 
 of the colours which are transmitted through a ruby glass, 
 
'^ft 
 
 S 8 
 
 LIGHT IN PHOTOGRAPHY. 
 ■siBuaniytu»tui 
 
COLOUR SENSITOMETER. 177 
 
 a green glass, a blue glass, and a white glass, we may 
 take it that any intermediate colours will not be very 
 incorrectly rendered, owing to the fact that each of 
 the coloured glasses transmits a considerable quantity 
 of the rays on each side of those colours of the spectrum 
 which most nearly match them. Thus the red will 
 transmit red and yellow ; the yellow — red, yellow, and a 
 little green ; the green — yellow and blue as well as g^een : 
 and the blue glass will allow some green, a little red, ajid 
 a large proportion of blue to pass. The question arises, 
 how can we best arrive at a solution of this problem in 
 a simple manner ? Suppose we reduce the luminosity of 
 the light passing through these glasses to a common 
 luminosity, say, to that of the darkest, then, if we have a 
 colour-correct plate, with equal times of exposure, all 
 the opacities in a developed plate should be identical. 
 
 The writer formed a sensitometer of these glasses by 
 placing four small squares of the different colours in a 
 row. He measured the luminosity of the light coming 
 through each small square, and then cut out sectors (which 
 can rotate in front of them) to cut off the light necessary 
 to make all appear of equal luminosity. We can place a 
 plate behind such a screen, and expose it to the same 
 kind of light as that in which the luminosity of the 
 light coming through these glasses was measured with 
 these sectors rotating in front of them. If a plate were 
 properly colour-sensitive, the impression made by the 
 light coming through these squares should show the 
 same opacity ; but in no case will they do so. Suppose 
 we find that the squares exposed through the blue 
 and white show greater opacity, evidently we must cut 
 off, by a yellow screen, some of the blue. We can cut 
 
 N 
 
178 LIGHT IN PHOTOGRAPHY. 
 
 off red by a green screen, and we can cut off green by 
 a red -screen. We can g^ess what depth of each we 
 shall require, and make an exposure with these screens 
 in front of the light, and then see what further alterations 
 are required. We can go on making alterations in the 
 screens till the densities under all squares are all the 
 same, and then use the absorbing medium so found for 
 camera exposures. 
 
 This test of the qualities of an orthochromatic plate 
 is a very searching one, and in it we have had to use 
 several of the methods of measurement we have already 
 described in this work. It is a useful application of 
 what we have endeavoured to teach in the earlier 
 chapters, and may be regarded in the light of an 
 experiment. 
 
 If the reader has travelled through these pages thus 
 far, it is hoped that the journey may not have been fruit- 
 less, and we take our leave of him with the wish that 
 what he has read may lead him to continue his study 
 by consulting other works on the subjects which have 
 been touched upon here. 
 
 FINIS. 
 
APPENDIX, 
 
 EFFECT OF TEMPERATURE ON 
 SENSITIVENESS. 
 
 Mawson Lantern Plate. 
 
 Exposed to a paraffin lamp at s feet, being visually 
 
 equal to an amyl-acetate lamp at 39 inches. 
 
 Table I. (see Fig. 49, page 129). 
 
 Temperature. 
 
 — 15° c. 
 
 ^-iz^C. 
 
 + 33» C. 
 
 Exposure in Seconds. 
 
 Readings. 
 
 
 
 100 
 
 100 
 
 100 
 
 5 
 
 99 
 
 96 
 
 94 
 
 10 
 
 97 
 
 92 
 
 89 
 
 20 
 
 90 
 
 80 
 
 7' 
 
 40 
 
 75 
 
 60 
 
 5« 
 
 80 
 
 SS 
 
 40 
 
 30 
 
 160 
 
 35 
 
 21 
 
 12-5 
 
 320 
 
 15-5 
 
 1 0*2 
 
 S"S 
 
l80 LIGHT IN PHOTOGRAPHY. 
 
 A Lightning Plate. 
 
 Exposure to amyl-acetate lamp at 5 ft. 
 
 Table II. (see Fig. 50, page 130). 
 
 Temperature. 
 
 — 16B C. 
 
 + e<'C. 
 
 + 33° C. 
 
 Exposure in Seconds. 
 
 Measures of Transparency. 
 
 
 
 100 
 
 100 
 
 100 
 
 5 
 
 98 
 
 94 
 
 86 
 
 10 
 
 86 
 
 80 
 
 70 
 
 20 
 
 69 
 
 64 
 
 S3 
 
 40 
 
 S3 
 
 49 
 
 38 
 
 80 
 
 38 
 
 36 
 
 28 
 
 160 
 
 28 
 
 26 
 
 20 
 
 320 
 
 20*5 
 
 18 
 
 i6-2 
 
 A Fitch Film — Exposed at 5 feet to an amyl-acetate lamp . 
 Table III. (see Fig. 51, page 131). 
 
 Temperature. 
 
 -i9-S°C. 
 
 + 14° c. 
 
 + 33° C. 
 
 Exposure in Seconds. 
 
 Measures of Transpa 
 
 rency. 
 
 
 
 100 
 
 100 
 
 100 
 
 s 
 
 88 
 
 87 
 
 84 
 
 10 
 
 77 
 
 74 
 
 70 
 
 20 
 
 64 
 
 59 
 
 55 
 
 40 
 
 49 
 
 45 
 
 40 
 
 80 
 
 34 
 
 30 
 
 27 
 
 160 
 
 22-5 
 
 20-5 
 
 19 
 
 320 
 
 iS-8 
 
 i4'3 
 
 13-2 
 
TEMPERATURE AND SENSITIVENESS. 
 
 l8l 
 
 Collodio-Bromide Plate. 
 
 Exposed at 3 feet from a paraffin lamp. 
 
 Table IIIa. 
 
 Temperature. 
 
 + 32° c. 
 
 — 170 C. 
 
 4- 32° C. 
 
 — 170 C. 
 
 Exposure. 
 
 Measures of 
 Transparency. 
 
 Energy used as chemi- 
 cal decomposition 
 derived from the salt. 
 
 12 min. 
 6 min. 
 min. 
 
 24" 5 
 46*5 
 100 
 
 71 
 
 86 
 
 100 
 
 149 
 
 74 
 
 32-6 
 i6-3 
 
 Scale of Opacity to the Foregoing. 
 
 Exposures. 
 
 Measures of Transparencies. 
 
 
 
 100 
 
 10 sec. 
 
 92 
 
 20 ,, 
 
 83 
 
 40 „ 
 
 62 
 
 80 „ 
 
 40 
 
 160 ,, 
 
 22 
 
 320 ,. 
 
 12-7 
 
l82 light in photography. 
 
 Lantern Plate. 
 Table IV. (see Fig. 52, page 132). 
 
 Temperature. 
 
 — i8» C. 
 
 -i- 400 c. 
 
 Exposure in Seconds. 
 
 Measures of Transparency. 
 
 
 
 100 
 
 100 
 
 S 
 
 93 
 
 86 
 
 10 
 
 85 
 
 68 
 
 20 
 
 70 
 
 40"S 
 
 40 
 
 SI 
 
 17-0 
 
 80 
 
 S3 
 
 ri 
 
 160 
 
 19 
 
 3'S 
 
 320 
 
 -•s 
 
 '■9 
 
 Mawson Lantern Plate. 
 Table V. (see Fig. 53, page 133). 
 
 Temperature. 
 
 + 18° C. 
 
 — 180 C. 
 
 + 38" c. 
 
 + 180 c. 
 
 Exposure in 
 seconds. 
 
 Measures of Transparency. 
 
 
 
 S 
 10 
 20 
 40 
 80 
 160 
 320 
 
 100 
 91 
 
 87 
 72 
 
 SO 
 27 
 ii-S 
 4-8 
 
 100 
 
 93 
 89 
 8i 
 60 
 39 
 19 
 7-2 
 
 100 
 90 
 82 
 S6 
 32 
 I3"S 
 
 S7 
 2-4 
 
 100 
 
 91 
 
 87 
 70 
 
 44 
 
 2f4 
 
 9"S 
 4-0 
 
TEMPERATURE AND SENSITIVENESS. 
 
 183 
 
 Exposure of Plate through Graduated Scale. 
 Table VI. (see Fig. 54, page 135). 
 
 Temperature. 
 
 + 34° C. 
 
 -iS'C. 
 
 Intensities in powers of 2. 
 
 Measures of Transparency. 
 
 
 
 100 
 
 100 
 
 2'2 
 
 95 
 
 92 
 
 27 
 
 9i 
 
 86 
 
 3'3 
 
 80 
 
 77 
 
 4'o 
 
 60 
 
 62 
 
 4-8 
 
 4' 
 
 44 
 
 5-8 
 
 16 
 
 24 
 
 6-3 
 
 io"S 
 
 '5 
 
 6-5 
 
 7'2 
 
 12 
 
EFFECT OF INTERMITTENT EXPOSURE. 
 
 Lantern Plate. 
 Table VII. (see Fig. 57, page 140). 
 
 Scale taken at 3 feet. 
 
 1 
 .2 
 
 8 
 
 c 
 
 s 
 5 
 
 Is 
 
 s 
 s 
 
 1 
 
 -4 
 
 hog 
 
 ^1 
 
 Comparative 
 
 chemical action 
 
 derived from 
 
 scale. 
 
 .E a 
 
 IP 
 
 Exposure. 
 
 Reading- of 
 Transparency. 
 
 ^1 
 
 sec. 
 
 
 
 
 m. a. 
 
 
 
 
 s 
 
 87-6 
 
 li 
 
 I 
 
 10 
 
 27.8 
 
 56-4 
 
 > = 
 
 10 
 
 75-3 
 
 H 
 
 z 
 
 / 
 
 jj'j 
 
 4&0 
 
 1 -815 
 
 20 
 
 55'o 
 
 3 
 
 i 
 
 40 
 
 37 
 
 40 
 
 ] , 
 
 40 
 
 37'o 
 
 3 
 
 i 
 
 4 
 
 5S 
 
 20 
 
 r soo 
 
 80 
 
 i8-o 
 
 6 
 
 i-V 
 
 2 40 
 
 S3 
 
 22'2 
 
 } 
 
 
 
 6 
 
 -A- 
 
 16 
 
 7^"5 
 
 PV 
 
 1 423 
 
 
 
 12 
 
 eV 
 
 10 40 
 
 69"S 
 
 12-7 
 
 ) 
 
 
 
 12 
 
 -e\ 
 
 <5^ 
 
 <?<?■<? 
 
 ^7 
 
 1 370 
 
 The figures in italics are 
 exposure with the sector, 
 interval of rest is i : 5. 
 
 those which relate to the 
 The ratio of exposure to 
 
INTERMITTENT EXPOSURES. 
 
 >8S 
 
 Slow Plate. 
 
 
 Table VIII. (see Fig. 5 
 
 8, page 
 
 '43)- 
 
 
 Scale. 
 
 Aperture. 
 
 II 
 
 ok 
 
 
 
 ^i 
 
 m 
 
 Exposure. 
 
 Reading of 
 Transparency. 
 
 
 s 
 
 92 
 
 360 
 
 
 m. s. 
 — 40 
 
 19 
 
 40 
 
 10 
 
 73-5 
 
 twice go" 
 
 ■50 
 
 I 20 
 
 19-8 
 
 37 
 
 20 
 
 39'o 
 
 „ 70" 
 
 ■39 
 
 I 43 
 
 2r6 
 
 35 
 
 40 
 
 21-0 
 
 „ so" 
 
 .28 
 
 2 24 
 
 22-8 
 
 33 
 
 60 
 
 irz 
 
 „ 30° 
 
 ■•7 
 
 4 
 
 26-s 
 
 29 
 
 80 
 
 5-4 
 
 „ 20° 
 
 •11 
 
 6 
 
 30'o 
 
 25 
 
 
 
 „ 10° 
 
 •06 
 
 12 
 
 3S'z 
 
 22-5 
 
 Slow Plate. 
 Table IX. (see Fig. 58, page 143). 
 
 Scale. 
 
 Aperture. 
 
 Is 
 
 II 
 
 Ex. 
 
 posure. 
 
 Readings 
 of Trans- 
 parencies. 
 
 Chemical 
 
 action 
 
 derived 
 
 from scale. 
 
 Exposure. 
 
 Reading of 
 Transparency. 
 
 
 ftJS 
 
 
 Fast. 
 
 Slow. 
 
 59 
 
 Fast. 
 
 Slow. 
 
 Bare glass 100 
 
 
 
 360 
 
 
 
 m. s. 
 — 40 
 
 59 
 
 40 
 
 40 
 
 sees. 
 
 
 twice 90 
 
 •50 
 
 I 20 
 
 63 
 
 6i 
 
 35 
 
 37 
 
 10 
 
 90 
 
 „ 70 
 
 ■39 
 
 I 43 
 
 65 
 
 63 
 
 32 
 
 35 
 
 '5 
 
 81 
 
 ,, SO 
 
 ■28 
 
 2 24 
 
 70 
 
 65 
 
 25 
 
 32 
 
 30 
 
 67 
 
 ,, 30 
 
 ■17 
 
 4 — 
 
 75 
 
 68 
 
 20 
 
 28 
 
 40 
 
 59 
 
 „ 20 
 
 "1 1 
 
 6 — 
 
 80 
 
 71 
 
 i5'5 
 
 24 
 
 50 
 
 57 
 
 „ 10 
 
 •06 
 
 12 — 
 
 86 
 
 77 12 
 
 20 
 
i86 
 
 LIGHT IN PHOTOGRAPHY. 
 
 Exposure of Negative Paper. 
 Table- X. (see Fig. 59, page 145). 
 
 Scale at 8 feet. 
 
 Exposure, 
 
 Reading of 
 greyness. 
 
 Bare glass i oo 
 96'o 
 83-0 
 64' o 
 
 48-5 
 42-0 
 
 5 
 
 sec. 
 
 10 
 
 ti 
 
 20 
 
 »j 
 
 30 
 
 »> 
 
 40 
 
 M 
 
 of each 
 exposure, 
 
 Duration Rotations of 
 
 sec. 
 
 drum 
 per second. 
 
 17* 
 26 
 
 50 
 
 Reading of 
 greyness. 
 
 Chemical action 
 
 derived 
 
 from scale. 
 
 44 
 56 
 
 S9 
 69 
 
 35 
 27 
 24 
 22 
 17 
 
 Exposure of Moderately Quick Plate. 
 Table XI. (see Fig. 60, page 146). 
 
 Scale. 
 
 No. of 
 
 intervals in 
 
 disc. 
 
 Intervals 
 between 
 exposure. 
 
 Expo- 
 sure. 
 
 Reading 
 of Trans- 
 parency. 
 
 Chemical 
 
 action 
 
 derived - 
 
 from 
 
 scale. 
 
 Exposure. 
 
 Readings of 
 Transparency. 
 
 Bare gl 
 
 sees. 
 
 ass 100 
 
 
 
 
 
 ra. 
 I 
 
 36-4 
 
 85-6 
 
 10 
 
 94 
 
 24 of 10° 
 
 5" 
 
 Ij 
 
 40'o 
 
 74" 4 
 
 20 
 
 80-5 
 
 12 „ „ 
 
 20° 
 
 3 
 
 48-5 
 
 58-4 
 
 40 
 
 57 
 
 6 „ ,, 
 
 50" 
 
 6 
 
 59'5 
 
 42-8 
 
 80 
 
 37"5 
 
 3 M »» 
 
 no" 
 
 12 
 
 67-0 
 
 3i"4 
 
INTERMITTENT EXPOSURES. 
 
 187 
 
 vO 
 tI- 
 
 s 
 
 1 
 
 ^ 
 
 a 
 
 ^ 
 
 
 
 
 
 
 J 
 
 bn 
 
 w 
 
 E 
 
 tK 
 
 
 
 
 
 H 
 
 w 
 
 Cii 
 
 
 U 
 
 HH 
 
 tr) 
 
 HH 
 
 s 
 
 X! 
 
 ><! 
 
 M 
 
 w 
 
 ri 
 
 Comparative chemi- 
 cal action derived 
 from scale. 
 
 
 N ^ f<^ « r^ 
 
 00 (^ 10 H- N 
 
 i 
 
 00 10 tJ- r^ N 
 
 a m 
 
 II 
 
 i 
 
 
 
 > 
 
 U 
 
 M 
 
 IT) »J^ »0 
 
 •^ tJ- 10 10 vo r^ 
 
 
 ■^ 10 10 ^ "O r^ 
 
 a 
 
 9 
 
 S 
 
 1 
 
 ^ eg 1 1 1 1 1 
 g 1 N t|- 00 vO 00 
 
 Intervals 
 
 between 
 
 , exposure. 
 
 10 
 
 ^1 
 
 
 ^ ;. ^ ^ 
 
 'S = r = = 
 
 S 
 
 ^1" 
 
 II 
 g s 
 
 N ON vo xj- MD 00 
 
 On r^ vo 10 Th fn 
 
 c 
 
 K 
 
 ' 
 
 doooooo'Sb 
 
EFFECT OF SMALL INTENSITIES OF 
 LIGHT. 
 
 Table XIIL (see Fig. 62, page 154). 
 Scale exposed at 4 feet. 
 
 Scale No. 
 
 Exposure. 
 
 Reading of greyness. 
 
 
 sees. 
 
 
 I 
 
 5 
 
 9S 
 
 2 
 
 10 
 
 90 
 
 3 
 
 20 
 
 69 
 
 - 3 
 
 30 
 
 58 
 
 4 
 
 40 
 
 49 
 
 s 
 
 80 
 
 29 
 
 6 
 
 160 
 
 16 
 
 White paper 
 
 100 
 
small intensities of light. 
 
 Exposure of Negative Paper. 
 Table XIV. (see Fig. 62, page 154). 
 
 189 
 
 Distance 
 of lamp 
 in feet. 
 
 Eiposure. 
 
 Reading of 
 greyness. 
 
 Amount of 
 
 chemical action 
 
 deduced from 
 
 the time scale. 
 
 Relative 
 
 amounts of 
 
 chemical action 
 
 produced. 
 
 24 
 
 m. 5. 
 24 
 
 83 
 
 13'4 
 
 •*3 
 
 20 
 
 16 40 
 
 78 
 
 is-i 
 
 •27 
 
 16 
 
 10 40 
 
 73 
 
 >7"4 
 
 •3' 
 
 12 . 
 
 6 
 
 67 
 
 20*5 
 
 •37 
 
 8 
 
 2 40 
 
 60 
 
 26-4 
 
 •47 
 
 4 
 
 40 
 
 49 
 
 400 
 
 ■71 
 
 2 
 
 10 
 White .. 
 
 38 
 . 100 
 
 564 
 
 I '00 
 
 Column 4 is derived from seeing where the greyness 
 of each square corresponded with the greyness of the 
 scale ; and if the law held good, each of the numbers in 
 it should be the same. 
 
 Exposure of Lantern Plate. 
 Table XV. shows the scale made with the light at 
 rather more than two feet distance, and Table XVI, 
 the exposures and distances of the light. 
 Table XV. 
 
 Scale No. 
 
 Exposure. 
 
 Transmitted light. 
 
 
 sec. 
 
 
 I 
 
 5 
 
 58 
 
 2 
 
 10 
 
 29 
 
 3 
 
 20 
 
 6-1 
 
 
 Bare glass 
 
 100 
 
igo 
 
 LIGHT IN PHOTOGRAPHY. 
 
 Table XVI. (see Fig. 62, page 154). 
 
 Distance. 
 
 Exposure. 
 
 Amount of 
 
 light 
 transmitted. 
 
 Amount of 
 
 chemical action 
 
 deduced from 
 
 the time scale. 
 
 Relative 
 
 amounts of 
 
 chemical action 
 
 produced. 
 
 feet. 
 
 m. s. 
 
 
 
 
 2 
 
 10 
 
 280 
 
 • '•s 
 
 I -00 
 
 4 
 
 40 
 
 3+-0 
 
 g-o 
 
 •79 
 
 8 
 
 2 40 
 
 41-2 
 
 7'S 
 
 •6S 
 
 12 
 
 6 
 
 482 
 
 6-3 
 
 •55 
 
 16 
 
 10 40 
 
 SO'4 
 
 S'9 
 
 •51 
 
 24 
 
 24 
 
 62-2 
 
 4-7 
 
 ■40 
 
 Exposure of a Wratten Plate. 
 Table XVII. (see Fig. 62, page 154). 
 
 Scale at 
 
 3 ft. from light. 
 
 Scale No. 
 
 Exposure. 
 
 Transmitted light. 
 
 
 
 sec. 
 
 
 I 
 
 
 5 
 
 40-4 
 
 2 
 
 
 10 
 
 24-6 
 
 3 
 
 
 20 
 
 17-6 
 
 4 
 
 
 40 
 
 I2"0 
 
 S 
 
 
 80 
 
 7-6 
 
 Bare glass ... 
 
 100 
 
SMALL INTENSITIES OF LI&HT. 
 
 191 
 
 Table XVIII. (see Fig 
 
 . 62, page 1 
 
 S4-)- 'i 
 
 Distance. 
 
 Exposure. 
 
 Amount of 
 
 light 
 transmitted. 
 
 Amount of 
 
 chemical action 
 
 duduced from 
 
 the time scale. 
 
 Relative 
 
 amounts of 
 
 chemical action 
 
 produced. 
 
 li 
 
 m. 3. 
 10 
 
 10 
 
 49-2 
 
 I"00_ 
 
 3 
 
 40 
 
 11-8 
 
 38-6 
 
 •78 
 
 6 
 
 2 40 
 
 14-7 
 
 28-2 
 
 •57 
 
 12 
 
 10 40 
 
 19-2 
 
 i8-6 
 
 •38 
 
 H 
 
 42 40 
 
 24 
 
 127 
 
 •26 
 
 The last column gives the equivalent in time referred 
 to the scale. 
 
 The results shown in the last columns of Tables XIV., 
 XVI., and XVIII., are shown graphically in Fig. 62, 
 page 154, the scale of the abscissae of the curves being 
 shown in powers of J intensity. 
 
 Table XIX. — Slow Lantern Plate (see Fig. 63, 
 page 157)- 
 
 Scale No. 
 
 a feet. 
 
 8 feet. 
 
 diagrram. 
 
 Exposure. 
 
 Transparency 
 of deposit. 
 
 Exposure. 
 
 Transparency 
 of deposit. 
 
 I 
 2 
 
 3 
 4 
 S 
 6 
 
 Bare glass. 
 
 sees. 
 
 5 
 10 
 20 
 40 
 80 
 160 
 
 100 
 
 88 
 
 78-5 
 
 62 
 
 44 
 27 
 12 
 
 Bare glass. 
 
 m. a. 
 
 1 20 
 
 2 40 
 5 20 
 
 10 40 
 21 20 
 42 40 
 
 100 
 
 96 
 
 90-5 
 81 
 
 6s 
 46 
 
 30 
 
192 
 
 LIGHT IN PHOTOGRAPHY. 
 
 Table XX.— Quicker Plate (see Fig. 63, page 157)- 
 
 Scale No. 
 
 2 feet. 
 
 8 feet. 
 
 diagram. 
 
 Exposure. 
 
 Transparency 
 of deposit. 
 
 Eicpbsure. 
 
 Transparency 
 of depolit. 
 
 
 
 Bare glass. 
 
 sees. 
 
 100 
 
 Bare glass. 
 
 m. 5. 
 
 100 
 
 I 
 
 s 
 
 68 
 
 I 20 
 
 75 
 
 2 
 
 10 
 
 43"S 
 
 2 40 
 
 50 
 
 3 
 
 20 
 
 2ro 
 
 S 20 
 
 25 
 
 4 
 
 40 
 
 8-S 
 
 10 40 
 
 10 
 
 S 
 
 80 
 
 + 
 
 21 20 
 
 5 
 
 6 
 
 160 
 
 2 
 
 42 40 
 
 3 
 
ACTION WITH INTENSE LIGHT. 
 
 Table XXI. (see Fig 64, page 163). 
 
 Barnet plate (ordinary), exposed through standard scale to electric 
 light at 4 ft. 6 in., with drop-shutter, and to an amyl-acetate 
 lamp at 3 ft. 6 in. 
 
 
 Reading of Transparency. 
 
 
 
 
 cies of scale in 
 powe« of 2. 
 
 Electric light 
 at 5 ft. 
 
 Amyl-acetate, 
 
 30 sees, 
 at 3 ft. 6 in. 
 
 Amyl-acetate, 
 
 90 sees, 
 at 3 ft. 6 in. 
 
 , . 
 
 100 
 
 100 
 
 100 
 
 1-44 
 
 80 
 
 95 
 
 73 
 
 '■73 
 
 77 
 
 9Z 
 
 68 
 
 2'il 
 
 70-5 
 
 89 
 
 59 
 
 2*70 
 
 64-5 
 
 83 
 
 49 
 
 3"3o 
 
 53-S 
 
 72 
 
 37 
 
 4"oo 
 
 44 
 
 S8 
 
 26-s 
 
 477 
 
 3» 
 
 4« 
 
 '7S 
 
 5-77 
 
 21-5 
 
 26 
 
 II 
 
 6-30 
 
 »7'S 
 
 18 
 
 9 
 
 6-SS 
 
 16 
 
 16 
 
 8 
 
'94 
 
 LIGHT IN PHOTOGRAPHY. 
 
 Table XXII. (see Fig. 64, page 163). 
 
 Edwards plate (special rapid), exposed tlirougli standard scale to 
 electric light at 8 ft., with djop-shutter, for yj^ sec, and to 
 an amyl-acetate lamp at 3 ft. 6 in. distant. ■ 
 
 Transparen- 
 
 Reading of Transparency. 
 
 cies of scale in 
 powers of 2. 
 
 Electric light 
 
 Amyl-acetate, 
 
 30 sees, 
 at 3 ft. 6 in. 
 
 Amyl-acetate, 
 
 60 sees, 
 at 3 ft. t) in. 
 
 
 100 
 
 100 
 
 100 
 
 1-44 
 
 92 
 
 100 
 
 96-6 
 
 173 
 
 92 
 
 98-3 
 
 92 
 
 2'2I 
 
 86 
 
 98-- 
 
 87-3 
 
 2-70 
 
 78-6 
 
 93'3 
 
 78-6 
 
 3'30 
 
 70-6 
 
 84 
 
 64-6 
 
 4'oo 
 
 61-3 
 
 75-3 
 
 SO-a 
 
 4"77 
 
 48-6 
 
 54-3 
 
 3i'3 
 
 5-77 
 
 32-0 
 
 33-3 
 
 14-0 
 
 6-30 
 
 25'3 
 
 22 
 
 9'3 
 
 6-55 
 
 21-3 
 
 18 
 
 8-0 
 
'95 
 
 Slow Plate. 
 Table XXIII. (see Fig. 65, page 166). 
 
 Comparative 
 
 Intensity of the 
 
 - Light passing 
 
 through the Scale 
 
 in Powers of 2, 
 
 Two Sparks 
 at 18 in. 
 
 Electric Light 
 at 3 ft. 6 in. 
 Drop Tin in- 
 
 Amyl-acetate 
 3 ft. 2 in. 
 
 1-44 
 
 88 
 
 97 
 
 955 
 
 '73 
 
 87 
 
 9S'S 
 
 940 
 
 2'2I 
 
 84 
 
 91 
 
 86-s 
 
 270 
 
 77 
 
 86-s 
 
 74 
 
 3-30 
 
 72'S 
 
 77 
 
 S5-S 
 
 4-00 
 
 64 
 
 68 
 
 37 
 
 4-77 
 
 S3 
 
 S2'S 
 
 212 
 
 5-77 
 
 38-S 
 
 34-4 
 
 9-5 
 
 6-30 
 
 2g'o 
 
 27-7 
 
 S'o 
 
 (•■ss 
 
 26-2 
 
 24-5 
 
 4-0 
 
 Lightning Plate. 
 Table XXIV. (see Fig. 66, page 167). 
 
 X, where 
 2 jt: is the 
 
 Spark. 
 
 Electric 
 Light. 
 
 Amyl- 
 Acetate. 
 
 intensity. 
 
 Reading. 
 
 Reading. 
 
 Reading. 
 
 1-44 
 
 66-s 
 
 6i-5 
 
 77 
 
 '•73 
 
 63 
 
 56 
 
 70 
 
 2'2I 
 
 S8 
 
 47'S 
 
 62 
 
 2*70 
 
 S' 
 
 38-S 
 
 54-5 
 
 3'30 
 
 44" 5 
 
 3i'S 
 
 44-0 
 
 4'oo 
 
 36'o 
 
 23s 
 
 30'S 
 
 4-77 
 
 25-0 
 
 i6-7 
 
 19-2 
 
 577 
 
 i6-8 
 
 lo-s 
 
 10.8 
 
 6'30 
 
 13-2 
 
 9-0 
 
 77 
 
 6-SS 
 
 11-4 
 
 8-1 
 
 6-3 
 
 Bare 
 
 I 
 
 
 
 glass 
 
 r 100 
 
 100 
 
 100 
 
 Scale Amyl-acetate 
 at 10 ft. 
 
 Time, j Reading. 
 
 sees. 
 
 
 s 
 10 
 
 77*5 
 S6 
 
 20 
 
 3i'S 
 
 40 
 80 
 
 19-5 
 II 
 
 160 
 
 6-3 
 
 320 
 
 Bare 
 
 glass" 
 
 3'S 
 r 100 
 
19^ 
 
 LIGHT IN PHOTOGRAPHY. 
 
 Table XXV. (see page i68). 
 
 Distance 
 in feet. 
 
 Number 
 of Sparks. 
 
 I Reading. 
 
 Equivalent 
 
 Value for 
 
 Scale, 
 
 12 
 
 64 
 
 20 
 
 I 
 
 9 
 
 36 
 
 22 
 
 •87 
 
 6 
 
 , 16 
 
 25 
 
 •72 
 
 3 
 
 4 
 
 35 
 
 •41 
 
 li 
 
 I 
 
 54) 
 
 
 li 
 
 I 
 
 52^53 
 
 •22 
 
 li 
 
 I 
 
 54i 
 
 
 Bare 
 
 ') 
 
 
 
 glass 
 
 _) 
 
 100 
 
 
 Scale. 
 
 Exposures 
 
 in seconds. 
 
 Reading. 
 
 Bare } 
 glass ) 
 
 100 
 
 10 
 
 91-2 
 
 20 
 
 70'0 
 
 40 
 
 47-5 
 
 80 
 
 32-5 
 
 160 
 
 20-0 
 
 320 
 
 I2'S 
 
 Table XXVI. (see page 169). 
 
 
 Readings of Transparency. 
 
 Intensities in 
 Powers of 2. 
 
 
 
 
 64 Sparks 
 at 12 ft. 
 
 16 Sparks 
 at 6 ft. 
 
 I Spark 
 at li ft. 
 
 '•44 
 
 85-5 
 
 86-2 
 
 97'S 
 
 '•73 
 
 83 
 
 85-0 
 
 97-0 
 
 z-ii 
 
 72-5 
 
 8ro 
 
 97-0 
 
 2'70 
 
 66-0 
 
 74-0 
 
 96-0 
 
 3'30 
 
 52-0 
 
 65*0 
 
 9S'o 
 
 4*00 
 
 4i"o 
 
 56-0 
 
 92'5 
 
 4-77 
 
 3o'o 
 
 42'S 
 
 87-5 
 
 577 
 
 20'0 
 
 30"o 
 
 80-5 
 
 630 
 
 iS'o 
 
 25-0 
 
 77-0 
 
 6-5S 
 
 iz-5 
 
 23-0 
 
 7S'o 
 
LUMINOSITY OF ARC-LIGHT SPECTRUM. I97 
 
 Table of Luminosities (see page ii8). 
 
 Scale No. 
 
 Wave-length. 
 
 Luminosity. 
 
 62 
 
 6957 
 
 2 
 
 60 
 
 6728 
 
 7 
 
 58 
 
 6520 
 
 21 
 
 56 
 
 6330 
 
 5° 
 
 54 
 
 6152 
 
 80 
 
 52 
 
 5996 
 
 96 
 
 50 
 
 5850 
 
 100 
 
 48 
 
 5720 
 
 97 
 
 46 
 
 5596 
 
 87 
 
 44 
 
 548 I 
 
 75 
 
 42 
 
 5373 
 
 62-s 
 
 40 
 
 5270 
 
 5o"o 
 
 38 
 
 5'72 
 
 36 
 
 36 
 
 508s 
 
 24 
 
 34 
 
 5002 
 
 14-2 
 
 32 
 
 4924 
 
 8-S 
 
 30 
 
 4848 
 
 5*5 
 
 28 
 
 4776 
 
 4-0 
 
 26 
 
 4707 
 
 2-8 
 
 24 
 
 4639 
 
 1-8 
 
 22 
 
 4578 
 
 I '4 
 
 20 
 
 4517 
 
 ro8 
 
 18 
 
 4459 
 
 •86 
 
 16 
 
 4404 
 
 •70 
 
 14 
 
 4349 
 
 •56 
 
 12 
 
 4296 
 
 ■45 
 
 10 
 
 4250 
 
 •34 
 
 8 
 
 4'97 
 
 ■26 
 
 6 
 
 4151 
 
 •18 
 
 4 
 
 4106 
 
 •14 
 
A THEORY OF ALKALINE 
 DEVELOPMENT. 
 
 So far we have treated the opacity of an image obtained 
 by development, vi^ithout reference to the principles 
 that are involved in this operation. There has been no 
 call to enter into this subject in the work, since the 
 developed image has only been taken as a measure of 
 action of light when compared with an action of light 
 that was known. It may be well, however, to show 
 briefly how development fits in with the ideas that we 
 have propounded regarding the distraction of a molecule 
 by an increase in the swings of some of the atoms. We 
 have supposed — and the supposition is made on evidence 
 which is most cogent — that an atom of bromine swings 
 itself out of the sphere of attraction of the mother 
 molecule of silver bromide, and joins some other foreign 
 molecule which is near it. The silver bromide we have 
 supposed to be formed of two atoms of silver and two 
 atoms of bromine, and that only one of these atoms of 
 bromine is removed by the action of light, leaving a 
 much more unstable molecule (which has been called 
 
ALKAXINE DEVELOPMENT. 1 99 
 
 sub-bromide of silver) behind. The action of develop- 
 ment is to bring in contact with this molecule of sub- 
 bromide a molecule of a substance which has a greater 
 attraction for the remaining atom of bromine than had 
 the two atoms of silver, although when the original 
 molecule is complete (that is, when it contains the four 
 atoms) its attraction of all its components for one 
 another is so great that the molecule of the developer 
 cannot overcome it, and no atom or atoms will be 
 separated. The flight of the atom caused by light leaves, 
 as it were, a hiatus in the molecule of sub-bromide, and 
 the mutual attractions are very much lessened, and thus 
 the sub-bromide molecule is in a fit state to be broken 
 up, and to leave the silver atoms alone uncombined 
 except with one another. [This we express in chemical 
 notation as follows : — The original molecule is AgjBrj, 
 and the sub-bromide Ag^Br. After the developer has 
 acted, Aga is left behind.] Now let us consider the state 
 of the two atoms of silver (Agj) so left. They have an 
 unsatisfied attractive force inherent in them, which they 
 endeavour to satisfy. In contact with them is a molecule 
 of silver bromide (AgaBr,), which has not been broken 
 up by the action of light. They therefore annex them- 
 selves to this molecule, and two molecules of sub- 
 bromide are thus formed (Ag.Br, 4- Ag, = 2Ag,Br). 
 These once stable molecules themselves thus become 
 unstable, and the developer can take away the bromine 
 atoms from them. A new elimination of silver takes 
 place, and those atoms of silver which are near another 
 molecule of silver bromide again continue to form fresh 
 molecules of the sub-bromide, and then again ar& 
 reduced to metallic silver, and so development takes 
 
200 LIGHT IN PHOTOGRAPHY. 
 
 place, and a visible image is formed. It will be seen 
 that for this action to take place very few— perhaps 
 only one — molecules in a particle need be shattered by 
 the action of light in order that the whole of that 
 particle may be reduced to metallic silver. Taking a 
 thinly-coated plate, and examining a small part of the 
 surface in the microscope, then allowing white light to 
 fall on it, and applying a developer, the particles can be 
 compared before and after development, when it will be 
 seen that the changes in structure that take place are 
 minute, and that the action is confined to particles. 
 But we have still stronger evidence that an action, at 
 all events, somewhat of the kind that has been described, 
 takes place. We can produce on a film which has not 
 been exposed at all an image which has been impressed 
 by light on another film, by contriving to bring the two 
 surfaces in absolute contact. For this purpose there is 
 nothing better than a film of collodion emulsion. If we 
 coat a glass plate with collodio-bromide emulsion, 
 and then expose it to a camera image, we can, when 
 the film is dry, coat it with another layer of the same 
 emulsion. The surface of the first film becomes 
 slightly dissolved, and the second film is therefore in 
 absolute contact with it, and some few particles of the 
 silver bromide in each film will coalesce with one 
 another. On applying the developer to this double 
 film, the image will begin to appear at the junction of 
 the two films, and will gradually go downwards and 
 upwards, and finally we shall have the image appearing 
 on the top film, which, it must be remembered, has 
 never seen light. We may then place a tough sheet of 
 gummed paper on the top surface, and remove the film 
 
ALKALINE DEVELOPMENT. 201 
 
 from the glass. When this is done another sheet of 
 gummed paper may be placed on the surface which was 
 next the glass, and when dried the two sheets may be 
 pulled asunder, and we shall find that the two films 
 have separated, one adhering to one sheet of paper, and 
 the other to the other, and that an image is on each. 
 This shows that the action which we have described 
 probably takes place. 
 
 Looking at development from this point of view, we 
 see that it is in strict accord with what we have stated 
 in this work. If only one molecule is decomposed, we 
 may suppose that a certain quantity of silver will be 
 isolated in a given time; if two molecules, twice— or 
 nearly twice — the amount of silver, and so on. When 
 the number of molecules so decomposed is large, it is 
 probable that the reduction will not be so great, for the 
 contact of the newly-formed silver atoms with the 
 molecules which have not been reduced by light to form 
 fresh sub-bromide must be limited, since the same 
 unaltered molecule may be equally near two or more of 
 the silver atoms, and there will consequently be a loss 
 in the total reduction in a given time compared with 
 what we might look for. There must of necessity be 
 great difficulty in finding whether such is practically the 
 case. The amount of silver deposited, as found by careful 
 chemical analysis made by Messrs. Hurler and Driffield, 
 does not indicate that such is the case, but the differences 
 that might be expected to be found are of such an 
 order that probably some more delicate plan must be 
 adopted before anything definite can be offered as 
 evidence. 
 
INDEX. 
 
 Absorption, 20 
 
 „ of Colourless Bodies, 
 
 27 
 ,, of Iodine, 21 
 
 „ of Nitric Oxide, 21 
 
 ,, of Silver Bromide, 
 
 26 
 ,, of Silver Chloride, 
 
 25 
 ,, of Yellow Glass, 23 
 
 Alkaline Development, 198 
 Amyl- Acetate Lamp, 152 
 Atoms and Molecules, i 
 
 Backing for Plates, 57 
 Barometer, The Number of 
 
 Small Particles, 45 
 Bromide, Colour of, 47 
 
 Cadett's Spectrum Plate, 174 
 Candle Flame, Small Particles 
 
 in, 40 
 Circular Motion — A Compound 
 
 Motion, 37 
 
 Colour of Bromide, 47 
 Colours, Luminosity of, 107 
 „ of Thin Films, 81 
 
 Dark-Room Light, 46 
 Dispersion, 12 
 
 Ebonite, Rays Transmitted 
 Through, 15 
 
 Energy in the Spectrum, 33 
 
 Ether, The, 4 
 
 ,, Vibrations, 9 
 
 Expansion by Heat, 3 
 
 Experiments with Phosphor- 
 escence, 121 
 
 Films, Colours of Thin, 81 
 ,, Soap, Thickness of, 84 
 
 Fluorescence, 18 
 
 ,, Photographic Value 
 
 of, 19 
 
 Fluorescence, Stokes's Law of,3I 
 
 Fringes, 63 
 
 Gas Flames, Luminosity of, 102 
 
INDEX. 
 
 Gratings, 68 
 
 Grey Surfaces, Measuring Light 
 Reflected from, 91 
 
 Halation, 56 
 Heat, Expansion by, 3 
 Heating Effect of Spectrum, 14 
 Herschel's Thermogram, 16 
 Hydrogen Molecules — Mean 
 Rate of Travelling, 2 
 
 Interference of Diffraction, 59 
 „ of Light Passing 
 
 Through a Slit, 65 
 Intense Light, Effect of, 158 
 Intermittance of Exposure, 137 
 
 Lamansky's Thermogram, 16 
 Lamp, Amyl-Acetate, 152 
 Large Particles, Scattering of 
 
 Light by, 54 
 Light, Definition of, 10 
 Lumifere's Red- Yellow Sensitive 
 
 Plate, 176 
 Luminosity of Pigment Colours, 
 
 108 
 Luminosity of Spectrum Colours, 
 
 116 
 
 Measurement of Spectra, 1 72 
 Media for Dark-RoomWindows, 
 
 5° 
 Molecules and Atoms, i 
 Motion of Pendulums, 6 
 
 Nicol Prism for Photography, 44 
 
 Orthochromatism, Test for, 177 
 
 Penduhmis,-5" 
 
 Pendulums and Atoms, 7 
 Phosphorescence, Spectrum of, 
 
 17 
 Photometer, 98 
 
 Box, 93 
 Pigments — Measurement of 
 
 Luminosity, 108 
 Pinholes, 72 
 Polarisation, 37 
 
 ,, of Light through 
 
 Turbid Jledium, 39 
 
 Radiation, 10 
 
 Red Light, Carrying Power of, 
 
 49 
 Refraction, 11 
 Ruled Surfaces, 67 
 
 Sectors, Rotating, 94 
 Sensitiveness and Temperature, 
 
 120 
 Sensitiveness, Gain of, by Heat, 
 
 127 
 Sensitiveness, Loss of, by Cold, 
 
 128 
 Silver Chloride, Heated, 2 
 Sky, 43 
 Skylight, Sunlight, and Gas- 
 
 Ught, 52 
 Slide for Exposures, 138 
 Small Intensities of Light, 148 
 
 ,, Particles, Effect of, on 
 
 Light, 34 
 Small Particles in a Candle 
 
 Flame, 40 
 Soap FUms, Thickness of, 84 
 Sound "Waves on Soap Films, 88 
 Spark^Photographs, 164 
 
INDEX. 
 
 Spectrum Colours, Luminosity 
 
 of, Ii6 
 „ Heating Effect, 14 
 ,, Infra Red, 14 
 „ of Light Reflected 
 
 from a Soap FUm, 86 
 ,, of Vapours, 13 
 „ ofWhite-HotBody,i3 
 ,, Simple Means to View 
 
 a, 13 
 „ Thermogram of Water, 
 
 29 
 Stokes's Law of Fluorescence, 3 1 
 Sunset Colours, 42 
 
 Thickness of Soap Films, 84 
 Thermogram, Herschel's, 16 
 
 Thermogram, Lamansky's, 16 
 „ of Solar Spectrum, 
 
 30 
 „ ofWater Spectrum, 
 29 
 Temperature and Sensitiveness, 
 
 120 
 Turbid Medium, 35 
 
 Wamerke's Circular Wedge, 105 
 Wave-Lengths, 70 
 Waves Compounded, 60 
 Wedges, Measurement of Light 
 Transmitted through, 103 
 
 Yellow Glass, 23 
 
 Zone Plate, 79 
 
 Carter Sc Co., Printers, 5, Fumival Street, Holborn, London, E.G. 
 
GEORGE HARE, 
 
 Manufacturer of Photographic Cameras and Apparattts, 
 
 26, CALTHORPE STREET, LONDON, W.C. 
 
 FIFTEEN PRIZE MEDALS AWARDED FOR EXCELLENCE OF 
 DESIGN AlfD WORKMANSHIP. 
 
 G. HARE'S NEW HAND-CAMERA. 
 
 This Camera has been designed for use with Double Slides in preference to 
 the Magazine form, being more certain and reliable in use, and at the same lima 
 more portable, measuring only 7^x4Jx5|. It is extremely simple, and free 
 from all complicated or intricate movements. Back adjustment for focussing. 
 Rising and falling front in each direction. 
 
 The Double Slides are new in form and design, and are extremely light and 
 portable, being f ths of an inch in thickness only. The Shutter is made of thin 
 corrugated steel, and opens inside the Camera by means of a lever from the 
 outside, which locks itself when closed. 
 
 Camera, complete, with Six Double Slides, Finder, and Shutter 
 Solid Leather Travelling Case, Spring Lock, Sling Strap, &c. 
 Bapid Rectilinear Lens, 5 J in. focus, working at//8 extra ... 
 Dallmeyer Hand-Camera Lens, f>| in. focus. Iris Diaphragm 
 
 Ditto, Sx 4 H. R. Lens, 6 in. focus. Iris Diaphragm 
 
 Extra Double Slide, each 
 
 Leather Case to hold three ditto 
 
 £6 6 
 
 
 
 16 
 
 
 
 2 2 
 
 
 
 4 2 
 
 6 
 
 4 12 
 
 6 
 
 7 
 
 
 
 4 
 
 8 
 
 ILLUSTRATED PRICE LISTS ON APPLICATION. 
 
PHOTOGRAPHIC GELATINE, 
 
 AND 
 
 3srELso:N"'s 
 
 SPECIAL X OPAQUE GELATINE. 
 
 Tlie Best for Dry-f late and otlier "Work. 
 
 MBTAIZ FROM ALL DEALhRS IN PROTOORAPEIC MATERIALS. 
 WSOLESALE FROM TEE MANUFACTURBRS- 
 
 GEORGE NELSON, DALE, & CO., Ltd., 
 14, Dowgate Hill, Loudon, B.C. 
 
 CONTAINING SEVERAL THOUSAND RECEIPTS, 
 
 Covering the latest, most impoitant, and most useful die reveries 
 
 in Chemical Technology, and their practical application 
 
 in the Arts and the Industries. 
 
 Edited cUefiy from the German of Drs. WincMer, Eisner, Meintu, MierzinM, 
 
 Jacobsen, KolUr, and Eeinzerlmg, 
 
 With additiom by William T. Brannt and William H. Wahl, Ph.D. 
 
 Illustrated by Seventy-Eigtat Engravings, Crown 8to, IOs. 6d. 
 
 London : SAMPSON LOW, MAESTOlf & COMPANY, Ltd., 
 St. Donstan's House, FEiiEa Lake, E.G. 
 
SAMPSON LOW, MAESTON& CO.'s 
 PUBLICATIONS. 
 
 — f^r^^^rV^f^^^M^k^t^t^^^t^^N'S^WNi^i^^ 
 
 HANDBOOKS-PHOTOGRAPHERS 
 
 In unitorm 8vo Volumes, fully illustrated 
 with Diagrams, &n. 
 
 Works by Capt. ABNEY, C.B., R.E., F.R.S. 
 
 THE ACTION OF LIGHT IN PHOTOGRAPHY. Ss. 6d. 
 INSTRUCTION IN PHOTOGRAPHY. Post 8vo, Ss. 6d. 
 INSTANTANEOUS PHOTOGRAPHY. Post 8vo, Is. 
 PHOTOGRAPHY WITH EMULSIONS. Post 8vo, Ss. 
 NEGATIVE MAKING. Post 8vo, Is. 
 
 PLATINOTYPE PRINTING. Written in conjunction with Ltoitel 
 
 Clabk, C.E. Post 8vo, 2s. 6d. 
 ART AND PRACTICE OF SILVER PRINTING. Written in 
 
 conjunction with Mr. H. P. RoBiNsoif. Post 8vo, 28. 6d. 
 
 COLOUR VISION. Being the Tyndall Lectures, delivered before 
 the Eoyal Institution in 1894. With Coloured Plate and 
 numerous Diagrams. 8vo, 12s. 6d. 
 
 Works by Mr. H. P. ROBINSON. 
 
 PICTORIAL EFFECT IN PHOTOGRAPHY. Post 8vo, sewed, 
 
 3s. ; cloth, 38. 6d. 
 THE STUDIO : AND WHAT TO DO IN IT. Post 8vo, sewed, 
 
 2s. ; cloth, 23. 6d. 
 
 LETTERS ON LANDSCAPE PHOTOGRAPHY. Post 8to, la. 
 
 ART AND PRACTICE OF SILVER PRINTING. Written in 
 conjunction with Captaut Abnbt, R.E. Post 8vo, 28. 6d. 
 
 London: SAMPSON LOW, MARSTON & COMPANY, Ltd., 
 St. DnNsiAN's House, Fettek Lane, E.G. 
 
PHOTOGRAPHIC PUBLICA'aONS. 
 
 BURTON'S MODEKN FHOTOOBAFHT. Comprising Practical Instrtictflts 
 in Working Gelatine Dry Plates. By Prof. "W. K. Bortok, O.K. Illustrated. Price If,, 
 per post Is. 2d. (Eleventh Edition.) \ 
 
 "The fact of this Twok going through soman; editions Bpeats for the appreciation with which itii. 
 ileAtrded. Only a short time has elapsed since the hook -was nearly entirely Te--ffritten, It is brought np"^ 
 to date. ProlesBor "W.K. Burton knows wha to write ahout, and how to do i%."—Briti§h Journal of 
 thOtography, March 8th, 1895. . . , . 
 
 "This book, 80 full of good things, has reacht-d its eleventh edition,at which none need wonder. It li 
 brought into line with to-day. The book is, as might be suppoBed, a Tcry thorough one. No phase of 
 purely photographic work of interest to the eTcry-day worker but is touched on." — PftolopropAy, 
 March ifth, 1895. 
 
 *' It is brought thoroughiy up to date, and is reliable and accurate in the information that it giyea, — 
 Amattur Photographer, March 8th, 1895. 
 
 " All the preparation and training I had when I began to try a hand at this pleasing hobby was what 
 was culled from ' Moderw Pmotoghapht,' by Mr. "W. K. liurton,'*— If. H. Jlnmo Vai ma,Princeoj Travancore. 
 
 " The book contains practical instructions on all matters the amateur need wish to know in the pursuit 
 of his hotPby. It Is up to date, additions and alterations suggfetcd by practice having been made. The 
 bookls 60 simple as to be understood by the merest tyro, and yet so full of suggestions that- the must 
 advanced may peruse it with advantage." — £a3t Zondon Advertiser , April 20th, 1895. 
 
 BTTBTON'S POCKET BOOS FOB FHOTO&BAFHEBS : Including the usual space 
 for Notes, &o., vitli Tables tor Facilitating ]£zposures. Price, cloth. Is., per post Is. Id. 
 
 OPTICS FOB PHOTOGBAPHEBS. By Professor W. K. Bubton. Profusely 
 illustrated, and giving full instructioas, based on experience, in the use of photograpliio 
 lenses, together with the theoretical consideratioos involved. Price Is., par post Is. 3d. 
 
 A CASKET OF PHOTOOBAFHIG OEKS. A Collection of 500 Dodges, Beceiets, 
 Entertaining Experiments, &c., in connection with the Art of Photography and its branches. 
 Collected, ClasBided, and Arranged for ready reference. Price Is., per post Is. 2d. 
 
 LANDSCAPE PHOTOGBAPHEB'S POCKET NOTE-BOOK. Arranged for entry 
 
 of date, light, process, lens, and stop osed, number of seconds, time of day, name of place, 
 
 and remarks. Price 6d,, per post 7d. 
 PHOTOBBAPHIC PBINTEE'S ASSISTANT. By Wiluam Heiqhwat. Price 
 
 Is., per post Is. Id. 
 FBACTICAL POBTKAIT PHOTOGBAPHY. A Hand-book for the Dark Boom, 
 
 the Skylight, and the Printing Koom. By the same Author. Price Is., per post Is. l^d. 
 ESTHETICS OF PHOTOGBAPHY. By the same Author. Being Hints on Posing 
 
 and Lighting the Sitter. Price Is., per post Is. lid. 
 HANDBOOK OF PHOTOGBAPHIC TEBMS. An Alphabetical Arrangement of 
 
 the Proee8seB,Formul8B, Applications, &o., of Pbotcgraphj for Ready Reference. Compile^ 
 
 by Wii-LiAM Hbiohwav. Price Ss. 6d., per post 28. Sd. 
 
 ELEHESTABY lESSONS IN SILVEE FEINTING. By W. M. Ashmam. 
 
 Prion Is. 6d., per post Is. 8d. 
 EIEMENTABY TEEATISE ON PHOTOGBAPHIC CHEMISTBY. By Arnold 
 
 Spilleb. Price 6d., per post 6Jd. 
 PHOTO-MICBOGBAPHY; Or, How to Photograph Microscopic Obiecta. By 
 
 I. H.jEHKllias. Also, A CHAPTEE ON PKEPARINO BAOTEBIA. Bv Dr. B. Ii. MlDDOX. 
 Price 38., per poatSs.ad. 
 ABT OF PHOTOGBAPHIC PAINTING. By A. H. BooL. Price Is., per post 
 
 ENAMELLING AND EETOITCHING. A Practical Photographic Treatise. By 
 
 p. FI4DXP8 Price 2s. 6d., per post 2s. 8d. 
 THE SPECTBOSCOPE AND ITS EELATION TO PHOTOGBAPHY. By C. Eat 
 
 'W'ooDS. Price 6d., per post 7d. 
 
 CARTER «Sc CO. (late Piper & Carter). 
 
 5, FURNIVAL STREET, HOLBORN, LONDON, EC. 
 
THE STEREOSCOPIC COMPANY'S 
 
 TWIN -LENS CARLTON 
 
 le now the Leading: Hand- Camera of the Day. 
 
 BECAUSE 
 
 1. — It fives power to focus for any distance, near 
 
 and far. 
 2. — It has a FULL size finder. 
 3. — It shows the exact size of the image, 
 i. — It has lenses working at //5'6. 
 5. — It has Iris diaphragm to working lens. 
 6, — It has u, safety shutter. 
 7. — It has a pneumatic control for all speeds, 
 8. — It gives time and instantaneous exposures. 
 9. — It holds twelve plates or films, 
 10. — It registers the number of plates taken, 
 11. — It can be used on the stand or in the hand. 
 12. — Exposed plates can be developed without dis- 
 
 ttirbing those unused. 
 1 3. — It is small, light, and perfectly constructed. 
 14. — It will stand extremes of climate. 
 15. — It is an expert's camera. 
 
 It is to simple that a beginner can use it. 
 Workmanship and advantages considered, it is 
 the cheapest Twin- Lens Camera in the Market. 
 
 Seventeen Oood Beaions 
 
 for Buying the 
 
 Twin-Lens Carlton Camera. 
 
 HAVE YOV READ 
 
 "TWIN-LENS CAMERAS, 
 
 mniat tbey are, and How 1^' 
 to Use themP" 17. 
 
 SzntFobt Fbke on Application. 
 
 PfllCES OF THE TWIlSr-LiENS CARLTON. 
 
 With EurjBCope Lenses, and all other advantages as detailed above. 
 
 To carry Twelve Plates or Twelve , films, 4i X H £18 12 
 
 „ ■ „ „ 5X4 ... .- 15 16 
 
 6J X 4| 21 
 
 FRSE LESSONS are given xoith the Camera.^ or full instructions tent with each instrument. 
 
 ciSLncrnnir? ^^Q & 108, Regent street, W. 
 STEREOSCOPIC [ J. 
 
 COMPANY, ) 64, CHEAPSIDE, E.C. 
 
BVBKY FRIDAY. OTsTE PENNY. 
 
 THE 
 
 i^osEt mp=to=msiU antf popular Wfoto^vapUt 
 ^uWtation of tf)t iSag. 
 
 VALUED BY ADVERTISERS! 
 
 APPRECIATED BY READERS! 
 
 I'S^^'^^S^^^^^^^N^^^^^'^^^ 
 
 SEE j^ ooi'ir. 
 
 22. Furnival Street, London, E.G. 
 
 i>ia:oxooB.A.i>i3:BB,s.