■ 
 
Cornell University Library 
 TH 7703.C89 
 
 Practical Illumination. 
 
 3 1924 004 595 264 
 
Cornell University 
 Library 
 
 The original of this book is in 
 the Cornell University Library. 
 
 There are no known copyright restrictions in 
 the United States on the use of the text. 
 
 http://www.archive.org/details/cu31924004595264 
 
y. 
 
PRACTICAL 
 ILLUMINATION 
 
 BY 
 
 James Raley Cravath 
 
 AND 
 
 Van Rensselaer Lansingh 
 
 NEW YORK 
 McGRAW PUBLISHING COMPANY 
 
 1907 
 
Copyrighted, 1907, 
 
 by the 
 
 McGraw Publishing Company, 
 
 New York. 
 
PREFACE. 
 
 It is a matter of common knowledge among experts that a 
 large amount of the artificial Hght that is now being produced 
 is unnecessarily lost. Cases where this loss reaches one-half or 
 two-thirds are matters of common observation. This does not 
 include the waste of light by users who forget to turn it off 
 when not needed. It refers simply to the failure of the average 
 user to get anywhere near the full benefit possible from the light 
 produced. Even more important than the economical side of 
 the subject is the disastrous effect on the eyes caused by numer- 
 ous common artificial lighting arrangements. 
 
 The broad general principles which should be followed in the 
 design of artificial illumination have been already well laid down 
 by eminent authorities. The present work has been undertaken 
 by the authors in order to supply much needed specific data and 
 suggestions as to how to carry out these principles in practice. 
 
 This common failure to get the practical results that should 
 be obtained from the light purchased, as well as much of the- 
 ruin of eyesight now common with artificial light, is due to the 
 fact that so few people understand the' importance of the proper 
 placing, reflecting and shading of artificial lights.- More atten- 
 tion is being paid to these things than formerly, as is evidenced 
 by the recent formation of the Illuminating Engineering So- 
 ciety ; but there is still a great lack of easily available information 
 on the many practical and technical points involved in the use 
 of artificial light. 
 
 In what follows, the authors attempt to present exact prac- 
 tical information of every-day use on many points that come up 
 in arranging artificial lighting. Most of the information is in 
 such shape that, while being of value to the technical man, it 
 can be read with profit and applied by the non-technical public 
 as well. A spreading of this kind of knowledge is to the inter- 
 Pbis both of electric light and gas companies and the users of 
 such light. 
 
vi PRACTICAL ILLUMINATION. 
 
 No attempt has been made to treat of the apparatus for pro- 
 duction of light, but rather to confine the w€)rk to the much 
 neglected subject of how best to use the light after it is produced. 
 
 A large number of tests are shown giving information on the 
 light distribution of various illuminants with different globes, 
 reflectors, and shades. Most of this information has not before 
 been available to the general user. All of the tests in the book 
 with the exception of a few which are noted in the text, were 
 made by the Electrical Testing Laboratories of New York City. 
 They have thus the advantage of being from a uniform, unbiased 
 source, this source being a laboratory which has behind it long: 
 experience and an established reputation in this class of work. 
 No tests made by manufacturers have been used. 
 
 Interior illumination only has been taken up, no attempt 
 being made to deal with the important but much neglected 
 subject of scientific street lighting. 
 
 Since it is in many cases of practical convenience or im- 
 portance to know the exact make and trade designation of 
 articles tested, an Appendix has been prepared in which opposite 
 the Fig. number of each test will be found the manufacturer, 
 name and trade number or designation of the article tested. 
 
 Part 2 is largely given to illustrations of rooms of various 
 kinds with good and bad lighting arrangements. Their desir- 
 able and undesirable features are discussed. The method of 
 treatment here as in the rest of the book has been to illustrate 
 principles first by specific practical examples and then from 
 these to point general conclusions. 
 
- CON^TEJSTTS. 
 
 Chapter Page 
 
 I. Introduction 1 
 
 II. The Laws and Measurement of Light 3 
 
 III. Light and the Eye 9 
 
 IV. Calculation of Illumination 13 
 
 V. Individual Incandescent Electric Lamps and Their Re- 
 flectors, Shades and Globes 25 
 
 VI. Individual Gas Burners and Their Reflectors, Shades and 
 
 Globes 76 
 
 VII. Acetylene Gas Burners and Their Shades 96 
 
 VIII. Incandescent Electric Lamp Clusters and Bowls 101 
 
 IX. Nemst Lamps and Their Globes 115 
 
 X. Electric Arc Lamps 120 
 
 XI. Gas Arc Lamps 128 
 
 XII. Vapor Lamps 134 
 
 XIII. Demonstration Room Tests 136 
 
 XIV. Comparison of lUuminants 142 
 
 XV. Introduction to Practical Examples 147 
 
 XVI. Residence Lighting 151 
 
 XVII. Desk, Drafting Room and Office Lighting 215 
 
 XVIII. The Lighting of Public Halls and Lodge Rooms 227 
 
 XIX. The Lighting of Large Dining and Banquet Rooms. . . 232 
 XX. The Lighting of Large Public Rooms, Depots, 'Lobbies, 
 
 etc 241 
 
 XXI. The Lighting of Hall and Corridors of Large Office Build- 
 ings 255 
 
 XXII. The Lighting of Theaters 262 
 
 XXIII. The Lighting of Churches 268 
 
 XXIV. The Lighting of Libraries, Reading and School Rooms. . 280 
 XXV. Car Lighting 288 
 
 XXVI. Store Lighting 293 
 
 XXVII. Show Window Lighting 324 
 
 XXVIII. Shop and Factory Lighting 332 
 
 XXIX. Miscellaneous Examples of Lighting 337 
 
 Appendix 346 
 
 Index 349 
 
CHAPTER I. 
 INTRODUCTION. 
 
 Before going into the details of the various points involved 
 in artificial lighting, a short statement of some of the principal 
 reasons why present attempts at efficient illumination are so 
 often found faulty will aid to an understanding of the subject. 
 
 One reason why much light is wasted is that frequently too 
 much of the light goes in directions where it is not needed and 
 too little in directions where it is needed. If light coming from 
 a lamp were a tangible thing, as, for example, water coming 
 from a hose, failure to direct it where needed would be more 
 apparent than it is. No one would 'think of trying to wash 
 windows, flush off pavements or do similar work calling for a 
 concentrated stream of water with a hose fitted with a lawn 
 sprinkler which throws a gentle spray in all directions. Nor 
 would one think of using a nozzle giving a powerful stream on a 
 flower bed or a soft, newly-made lawn. The light thrown in 
 different directions around a lamp can be modified with globes 
 and reflectors just as the direction and character of a stream 
 of water issuing from a hose can be modified by the nozzle. 
 The important point is to pick the right kind of nozzle and the 
 right kind of reflector or globe for the particular work in hand. 
 Although a certain amount of light on ceilings and walls is a 
 good thing, it is quite likely to be the case that a much larger 
 proportion of the total hght goes in those directions than is at 
 all necessary, while a smaller proportion than is necessary is 
 thrown in places where it can be utihzed. On the other hand, 
 there are cases where the reverse of this is true. 
 
 It is of first importance to know what effects various reflectors, 
 shades and globes will produce. Much attention is therefore 
 given to this in following chapters. 
 
 Another common mistake in Ughting is to place the lamps 
 in the line of ordinary vision and to fail to cover them with dif- 
 
 1 
 
2 PRACTICAL ILLUMINATION. 
 
 fusing globes or shades so that the direct rays from the unshaded 
 lamp fall directly into the eyes of the users or occupants of a 
 room. This produces two effects which may not be realized. 
 It is not only trjdng on the eyes but it is more difficult to see 
 well with dazzling light in the line of vision than if the lamps 
 were shaded or placed out of ordinary view. The rays of an 
 intense light falling directly into the eyes cause the pupils to 
 contract in self-defense, to cut down the light entering. The 
 result is that so much light is shut out that objects in the room 
 cannot be seen nearly as well and do not appear nearly as well 
 lighted as if the intense source of light were kept out of the 
 line of ordinary vision. We do not think of trying to look at 
 the sun, as its temporary blinding effect is too pronounced and 
 painful to make one wish to attempt it. The same effects, to 
 a very much less degree, are produced when we look at intense 
 artificial light such as that from the arc, the Nernst, the gas 
 mantle burner or the incandescent lamp. The effect may be 
 so small that it is not noticed at first, but is the cause of much 
 unconscious straining of the eyes and reduces the practical value 
 of the illumination of any room where bare arc, incandescent or 
 gas lamps are not kept out of the ordinary line of vision. Such 
 light shining in the eyes many hours a day in extreme cases 
 causes an irritation that will produce partial blindness and en- 
 force absolute rest. The remedy, of course, is to place the 
 lamps high enough to avoid this if possible, and if that cannot 
 be done, to protect the eyes by placing globes, shades, or re- 
 flectors on the lamps which will reduce the intensity without 
 absorbing and wasting an undue amount of light. In this 
 connection also the selection of proper globes and shades is im- 
 portant. 
 
 The position of the lamps is frequently not what it should be 
 to get the best results. Some general rules about this can be 
 laid down, but this phase of the subject can best be studied by 
 pointing out the defects and good points of specific examples. 
 Such examples will be given in Part 2 with the aid of numerous 
 illustrations from photographs of rooms with lamps variously 
 placed. Thus clearer ideas of the things to be sought and 
 avoided can be obtained than by a study of abstract general 
 principles. 
 
CHAPTER II. 
 THE LAWS AND MEASUREMENT OF LIGHT. 
 
 It is desirable before considering practical examples of arti- 
 ficial lighting to consider some of the fundamental laws under- 
 lying the subject of light and the method of measuring the 
 values of different illuminants whether with or without their 
 globes, shades or reflectors. It is the intention here, for the 
 sake of ready reference, to give only the fundamental laws in 
 an extremely condensed form. Fuller discussions of such laws 
 can be obtained from any text-book on phvsics or photometry. 
 
 From a purely physical standpoint, rapid A'ibrations of the 
 ether striking the retina of the eye produce the sensation com- 
 monly known as light. Different rates of vibration produce 
 different sensations which we distinguish as different colors; 
 thus a low rate of vibration produces the color sensation of red 
 while a higher rate of vibration produces the sensation of violet. 
 Between these lie the other colors. Such vibrations of the 
 ether, commonly known as rays, always travel in straight lines 
 unless interfered with by some external cause. They may be 
 changed in direction, however, as will be seen later in considering 
 the questions of refraction and reflecti:n. 
 
 It is customary in this country to refer to all sources of light 
 in terms of the light given in a horizontal direction b}^ a standard 
 candle. 
 
 Photometry. 
 
 A photometer is an instrument for measuring the light from 
 any given source. All photometers are based on the law of 
 inverse squares, i.e., that the intensity of light varies inversely 
 with the square of the distance from the source of light. Fig. 1 
 illustrates this graphically. It is evident from an inspection 
 of this that the surface C, which is twice the distance of B from 
 the source of light A, will have only one -fourth of the light rays 
 falling on each point that fall on B, and that D will only have 
 
 3 
 
4 PRACTICAL ILLUMINATION. 
 
 one-ninth of that faUing on B since the areas illuminated in both 
 cases by the same rays are respectively four and nine times 
 that of B. 
 
 Fig. 2 shows diagrammatically a Bunsen photometer. Briefly 
 it consists of a standard source of Ught, A, the source of light 
 to be measured, B, and a screen, S, which is composed of heavy 
 white paper with a grease spot in the center. This is viewed 
 
 Fig. 1. — Illustrating Law of Inverse Squares. 
 
 from both sides at once by two mirrors M and A''. When seen 
 by reflected light the center will appear darker than the outside, 
 since part of the light is transmitted and therefore less is re- 
 flected. If viewed by transmitted light, i.e., light from the 
 opposite, side the center will appear lighter than the outside, 
 as the outside does not allow the light to pass through. In 
 operation the screen is moved back and forth until the center 
 spot is equally bright with the outside portion. By means of 
 
 Fig. 2. — Showing principle of the photometer. 
 
 the two mirrors, placed at an angle, both sides of the screen 
 can be seen at the same time and at the same angle. When 
 they are equally bright the strengths of the two sources of light are 
 inversely proportional to the squares of their distances fromthe 
 screen. Thus if, when the screen is balanced as before explained, 
 the distance. from .4 to S is one foot, and from 5 to 5 four feet, 
 light B will be sixteen times the strength of light ,4, 
 
LAWS AND MEASUREMENT OF LIGHT 5 
 
 This would measure only the horizontal candle power. To 
 get the candle power at other angles light B could be tipped, 
 but in the case of measurements with a gas lamp and measure- 
 ments of other sources of light where it is undesirable or incon- 
 venient to tip them, there is actually employed in practice a 
 series of mirrors which reflect the light at different angles into 
 the photometer without necessitating the movement of the source 
 
 of light. 
 
 Reflection. 
 
 It is customary to consider reflection as being of two different 
 kinds, which while in themselves identical produce never- 
 theless entirely different results. Thus we have regular reflec- 
 tion in which a ray of light striking a reflecting surface, such 
 as a mirror, is reflected as a sharp beam, the angle at which it 
 strikes the mirror being equal to the angle at which it is re- 
 flected. 
 
 Fig. 3. — Illustrating regular reflection. 
 
 In Fig. 3, if A be the beam of Ught striking M N, a. re- 
 flecting surface, and C is the reflected ray, the angles between 
 the mirror M N, and such rays will be equal. That is to say 
 the angle AON equals the angle COM. It will also be noted 
 that the rays lie in the same plane. A common example of 
 such reflection is that of an ordinary mirror. It should be 
 noted in such cases that the reflecting surface, namely, the 
 mirror, is not visible but only the object reflected. Such re- 
 flected rays are as a rule unchanged except in intensity and in 
 some cases in color. The color as a rule varies with the nature 
 of the reflecting surface. Thus burnished silver which has the 
 highest reflecting power reflects 92 per cent, of the light thrown 
 upon it while in the case of burnished copper the reflection 
 falls to 50 per cent, or less and the rays are tinged with th^ 
 color of the copper. As, however, we have to deal largely 
 with mirrored surfaces such as for example mirrored troughs. 
 
6 PRACTICAL ILLUMIXATION . 
 
 mirrored cone reflectors, etc., and as some of the ordinary forms 
 of reflectors are made with such material it will be desirable to 
 consider their case more in detail which will be done in the 
 chapters on reflectors. 
 
 In the case of a rough surface, whether polished or unpolished, 
 every individual ray of light striking such a surface is regularly 
 reflected, but as each part of such surface hes at a different 
 angle and in a different plane, the light is reflected in all direc- 
 tions. Thus in Fig. 4, we have a number of parallel rays of 
 light striking on an irregular reflecting surface. It will be 
 seen at once that while each ray is regularly reflected, the dif- 
 ferent rays of light are reflected in all directions and in different 
 planes. This phenomenon is known as irregular or diffuse re- 
 flection. In some cases it will be noticed that the rays escape 
 after a single reflection while in other cases the light is reflected 
 
 Fig. 4. — Illustrating irregular or diffuse reflection. 
 
 one or more times before escaping from such a surface. Diffuse 
 reflection, therefore, scatters the rays in all directions, although 
 in the case of certain materials such as glazed paper the larger 
 number of rays will be regularly reflected. Thus in the case 
 of a piece of glazed writing paper resting on a desk we often 
 find a glare from the paper which is extremely trying to the 
 eye. This is due to the fact that the light striking on such a 
 surface is largelj^ reflected directly into the eyes with the result 
 that one cannot read what is on the paper and it is necessary 
 to so move the paper that the light is no longer regularly re- 
 flected to the eye and the light received by the eye is from 
 diffuse reflection only. If on the other hand a white piece 
 of blotting paper were placed on the desk instead of the glazed 
 writing paper it would be found that the regular reflection had 
 almost entirely vanished. The white blotting paper is a good 
 example of almost perfect diff'use or irregular reflection. 
 
LAWS AND MEASUREMENT OF LIGHT. 7 
 
 The amount of light reflected from different surfaces whether 
 such surfaces give either regular or irregular reflection or a 
 combination of both plays an extremely important part in the 
 subject of practical illumination. Thus in the case of an or- 
 dinary living room with light colored paper the total amount of 
 light as actually measured may be twice that we would get 
 without considering the reflection of the walls, ceiling and objects 
 in the room. 
 
 The following table taken from tests by different observers 
 will give some idea as to percentage of light reflected from 
 different standard wall papers, etc. : 
 
 Per Cent. 
 Reflection. 
 
 White blotting paper 82 
 
 White cartridge paper 80 
 
 Ordinary fools-cap 70 
 
 Ordinary newspaper 50 to 70 
 
 Chrome yellow paper 62 
 
 Orange paper 50 
 
 Plain boards (clean) 45 
 
 Plain boards (dirty) 20 
 
 Yellow wall paper 40 
 
 Tracing cloth 35 
 
 Tracing paper 22 
 
 Light pink paper 36 
 
 Blue paper '■^5 
 
 Yellow painted wall (clean) 40 
 
 Yellow painted wall (dirty) 20 
 
 Emerald green 18 
 
 Dark brown paper ; 13 
 
 Vermilion paper 1- 
 
 Blue green paper 12 
 
 Cobalt blue paper 12 
 
 Black paper 0.5 
 
 Deep chocolate paper 0.4 
 
 French ultramarine blue paper 3.5 
 
 Black cloth 1-2 
 
 Black velvet 0.4 
 
 Refraction and Total Reflection. 
 Whenever a ray of light passes from one medium to another, 
 such as from air to glass, it is bent out of its course, except 
 when the ray is at right angles to the surface. Thus in Fig. 5 
 the ray, A B, passes from air into a plate of glass at the point 
 0, where it is sHghtly bent. Where it passes out of the glass 
 
8 PRACTICAL ILLUMINATION. 
 
 at P, it is again bent an equal amount in the opposite direction 
 so that the leaving ray, P B, is parallel with the enteiing 
 ray, A 0. 
 
 However, when the Hght strikes at less than a given angle 
 ( the critical angle) it will not enter the glass but will be reflected 
 
 Fig. 
 
 Refraction and total reflection. 
 
 as shown by the ray, DOC, in Fig. 5. This is called total 
 reflection. Total reflection also occurs in the case shown in 
 Fig. 6, where a ray of light enters a piece of prismatic glass at A 
 is reflected at B and reflected a second time at C back in a direc- 
 tion parallel to that in which it came. This is what takes place 
 in prismatic glass reflectors. In such reflectors, however. 
 
 
 
 
 E 
 
 G 
 
 " B 
 
 /^ 
 
 C / 
 
 \ 
 
 XX 
 
 / 
 
 
 
 "^ 
 
 ^ ^ 
 
 A 
 
 
 
 
 
 
 
 
 D 
 
 F 
 
 Fig. 6. — Reflection in prismatic reflectors. 
 
 some light passes straight through at the tops and bottoms 
 of the prisms as shown by the rays D E and F G, in Fig. 6. 
 This is because the tops and bottoms of the prisms are slightly 
 rounded, so that the rays do not strike them at less than the 
 critical angle. 
 
CHAPTER III. 
 
 LIGHT AND THE EYE. 
 
 The effect of light on the eye can perhaps best be understood 
 by a study of the structure of the eye. Fig. 7 shows a sketch 
 of the eye where i is the iris or diaphragm, or what is usually 
 known as the colored part, as for example when we say a person 
 has gray or blue eyes. The iris contracts and expands to regu- 
 late to a certain amount, the light which passes through the 
 opening called the pupil, p, which appears as the black spot 
 in the eye. Back of the iris is a lens, /, which is controlled by 
 muscles which automatically alter its shape so as to focus the 
 picture of the object seen on the retina, r, at the back of the 
 eye-ball. The retina is a very sensitive layer on which the light 
 
 Fig. 7.— The Eye. 
 
 falls, producing a picture or image of the object seen. On the 
 back of the retina are a large number of small nerves which com- 
 bining, form the optic nerve leading to the base of the brain. 
 The center of the eye-ball consists of a watery fluid, v, known 
 as the vitreous humor. The retina is the sensitive part of the 
 eye and while it is wonderful in its abiHty to recuperate it is 
 not advisable to strain it. In order, therefore, not to injure 
 or fatigue the eye the following points should be noted. 
 
 1. A flickering light should be carefully guarded against as 
 the iris is unable to keep pace with the rapid fluctuations and 
 consequently too much and too little Ught alternately falling 
 on the retina very quickly produces fatigue. Examples of this 
 
 9 
 
10 PRACTICAL ILLUMINATION. 
 
 are a flickering open gas jet; an incandescent lamp supplied 
 with current from a dynamo driven by the ordinary single 
 cylinder gas engine ; an arc lamp operated with 25-cycle 
 alternating current, etc. 
 
 2. Lights should not be placed so that the rays from them 
 falling on a glazed or polished surface, such as writing paper, 
 are regularlv reflected into the eve as in this case a picture of 
 the light itself is found on the retina which obhterates the 
 images of the writing or printing on the paper. This is com- 
 monly called glare from the paper and is not only extremely 
 trying on the eyes but also renders almost futile the effort to 
 see clearty. 
 
 3. Any part of the body which is subject to more violent 
 strain than other parts is immediately noticed by contrast. 
 Thus in the case of a severe blow on the arm, all attention is 
 focused on it rather than on other muscles which may at the 
 same time be onlv moderately injured. In the same way if a 
 bright light is seen a picture of it is formed on the retina along 
 with the picture of the surrounding objects. So strong, how- 
 ever, will be the effect of the bright light on the eye that in 
 comparison other surrounding objects cannot be noticed. 
 This accounts for our inability to see past a bright light. For 
 example when walking down a street lighted with arc lights 
 hung low we are unable to see bevond the arcs although there 
 may be considerable light falling on objects beyond. This also 
 accounts for the fact that if one wants to display goods in a 
 window the wrong way to do it is to put strong glaring lights 
 in view. Although there may be sufficient illumination on the 
 goods displayed the effect on the retina in comparison with 
 the glare from the source of light is small. The ability to see- 
 well means that no part of the retina should be overstim- 
 ulated by means of a bright light. This necessitates keeping 
 lights out of the field of vision and when this is not possible, 
 keeping down their brightness by diffusing globes. The ability 
 to see objects past or in the vicinity of a bright light is also 
 reduced because of the fact that the iris contracts to cut down 
 the amount of light entering the eye so as to protect the retina. 
 This virtually makes the objects appear so dimly lighted that 
 it is only with difficulty that we see them. 
 
 4. Light coming in large quantities from an unusual angle 
 
LIGHT AND THE EYE. U 
 
 should be avoided. Certain parts of the retina are, through long 
 use, trained and accustomed to receive light while other parts 
 are not prepared to receive strong light. In dayhght we re- 
 ceive a gentle light from a great many directions mainly from 
 above. If, therefore, a strong light falls on the retina in 
 certain places it is strained. We have an example of this in 
 snow blindness where a strong reflected light comes from below 
 and not only produces fatigue but may also result in temporary 
 blindness. This is also noticed where rays from a bare lamp 
 of high candle power are allowed to fall in the eye even though 
 the angle be considerably out of the ordinary range of vision 
 and one is not looking at the light. For example a powerful 
 chandelier light in the center of a church may cause discomfort 
 to one facing it even if one does not look at the lights. Many 
 people suffer from this even though they may not be aware of 
 the cause of discomfort. 
 
 5. Too little light should be avoided as the retina is strained 
 in its effort to see clearly. This is a matter of common knowl- 
 edge and experience. 
 
 6. Too much light is to be avoided as the iris can contract 
 only a certain amount and if in spite of its contraction too 
 much light enters the pupil and falls on the retina the latter 
 is strained just as our muscles are strained when too much 
 effort is made. Abuse of the eyes by too much light is not as 
 frequently found as the other abuses mentioned. However, 
 the eye frequently receives too much light due to the causes 
 mentioned in 2, 3 and 4. 
 
 7. Avoid sources of light which cast streaks on the work. 
 These are bad because the source of light frequently vibrates 
 slightly, producing the same general effect on the eye as a 
 flickering light. Streaks are also bad, even when steady, 
 because of the uneven illumination produced which makes it 
 difficult for the eye to adjust itself to the conditions caused by 
 the varying degrees of brightness. An example of this is the 
 light cast by an ordinary green plated opal desk shade or a 
 polished metal reflector such as tin or aluminum with a clear 
 bulb incandescent lamp inside. This subject is taken up 
 further in the chapters on reflectors. 
 
 8. Care should be taken to avoid a sharp contrast between a 
 light surface on which one is working and dark surroundings 
 
12 PRACTICAL ILLUMIXATIOX. 
 
 such as, for example, a brilliantly lighted desk with the rest of 
 the room in darkness. Too sharp a contrast is injurious for 
 the reason that the eye does not have an opportunity to adjust 
 itself to the different conditions when glancing up from the work. 
 On the other hand, a moderate change in the illumination is 
 good. The eye is rested thereby as when attention is focused 
 strongly and constantly on one thing the eye requires a change 
 both in the illumination and in the image on the i-etina in order 
 to give it an opportunity to recuperate. 
 
 Oculists are constantly coming across cases of temporary or 
 permanent injury to the eyes through neglect of one or more 
 of the precautions enumerated. It is likely that many others 
 who are not forced to resort to the oculist realize considerable 
 discomfort and annoj-ance. Those who have been working 
 almost entirely by artificial light surrounded on all sides by 
 glaring undiffused lights have sometimes reached the point 
 where the light is unendurable and have had to give their eyes 
 complete rest after which it is impossible for them to work con- 
 tinuously by artificial light and it is only possible for them to 
 use artificial light with comfort by avoiding all the mistakes 
 which have just been enumerated. 
 
CHAPTER IV. 
 
 CALCULATION OF ILLUMINATION. 
 
 Before being able to calculate the illumination obtained from 
 any source of light it is necessary to know the candle power of 
 the rays thrown in various directions by this source. Fig. 8 
 shows the strength of the light thrown in different directions in 
 a vertical plane about a bare 16 candle-power incandescent elec- 
 tric lamp. The candle powers at the various angles as marked 
 on this diagram were obtained by means of a photometer. The 
 dotted curve surrounding the lamp is called a photometnc'curve, 
 
 Candle Power 
 Fig. 8- — Photometric curve of bare incandescent lamp. 
 
 because its distance from the center where the lamp is located 
 is proportional to the candle power at any given angle as deter- 
 mined by a photometer. In other words, the length of the 
 radial line represents the candle power. We see that the 
 candle power varies considerably at different angles. At the 
 horizontal it is 16. Directly underneath it is only 6.6. The 
 distribution of light is varied greatly by the use of reflectors as 
 can be seen by the large number of tests given later in the 
 chapters on reflectors, shades and globes. 
 
 13 
 
1-i PRACTICAL ILLUMINATION. 
 
 Some illuminants do not give the same amount of light when 
 viewed from different directions in a horizontal plane. In 
 such cases, it is therefore customary to make horizontal dis- 
 tribution curves showing these variations. With most sources 
 of light, such curves are usually circular as they give practically 
 the same amount of light in different horizontal directions. 
 Thus with the incandescent electric lamp, the open flame gas 
 lamp, indiA'idual mantle burners either upright or inverted, and 
 aceytlene lamps, this holds true, but it does not with electric 
 arc lamps, gas arc lamps, Nernst lamps and most vapor lamps 
 so that in such cases it is customary to take the mean of read- 
 ings made in different vertical planes and assume this as giving 
 the vertical distribution curve of the source of light. This lack 
 of uniformity in horizontal distribution is sometimes of great 
 practical importance and will be considered in detail when con- 
 sidering such sources of light. 
 
 -l-Foot- 
 
 KiG 9. — Foot candle. 
 
 In the United States intensity of illumination on any surface 
 upon which light falls is usually expressed in foot-candles. 
 
 If a standard candle is set one foot from a perpendicular wall 
 as in Fig. 9, the illumination on the wall at the point A will be 
 one foot-candle. If the candle were two feet from the wall the 
 intensity at the point A would be 0.25 foot-candle which is ob- 
 tained by dividing 1, the candle power of the source of light, 
 by 4, the square of the distance in feet, as was shown in Fig. 1. 
 We then have the rule: 
 
 To find the foot-candle intensity of light on a surface at right 
 angles to the ray of light, divide the candle power intensity of 
 the ray by the square of the distance in feet. 
 
 Expressed mathematically, this would give the formula: 
 
CALCULATION OF ILLUMINATION. 15 
 
 where 
 
 / = intensity of illumination in foot-candles 
 
 cp = candle power in the given direction 
 
 d = distance in feet. 
 
 This rule does not hold true for surfaces not at right angles 
 to the ray of light. This will be taken up later. Surfaces at 
 right angles to a ray of light are usually spoken of as normal 
 to the ray and in figuring the illumination of such surfaces the 
 term normal illumination is used. 
 
 To take a further example of normal illumination, suppose 
 that it is wished to determine the illumination on a reading page 
 held at right angles or normal to the ray of light thrown by a 
 16 candle power lamp in the fluted opal cone reflector which 
 is shown in Figs. 19 and 20. Wc will suppose further that 
 
 C^ Reading Page 
 Fig. 10. — Normal illumination. 
 
 the page is held 30 degrees from a ^'ertical line drawn through 
 the lamp, and is three feet from the lamp as in Fig. 10. From 
 Fig. 19 it is seen that the candle power intensity of that ray 
 30 degrees from the vertical is 24.5. The intensity in foot- 
 candles of the light falling on the reading page will be 24.5 
 candle power divided by the square of three feet which gives 
 as a result 2.7 foot-candles. 
 
 If the foot-candle intensity of light on a surface not at right 
 angles to the ray of light is wanted, the foot-candles can be 
 calculated for a surface at right angles to the ray as before and 
 this result multiplied by the cosine of the angle between the ray 
 of light and a perpendicular drawn from the lamp to the illu- 
 minated surface. Thus in Fig. 11 with the lamp at 5 and the 
 illumination to be determined at point B, the normal illumina- 
 
16 PRACTICAL ILLUMINATIOK. 
 
 tion in foot-candles on point B must be multiplied by the 
 cosine of the angle a, this being the angle between the ray S B, 
 and the perpendicular 5 D. 
 
 Expressed mathematically, this would give the fonnula: 
 
 cp 
 Horizontal I = -^ cos a. 
 d- 
 
 where the letters have the same meaning as in the last formula. 
 The cosine of any angle can be found in a table of cosines such 
 as is used in trigonometry but those who are not familiar with 
 trigonometry and the use of such tables as well as those who 
 are familiar with them can get the desired results more quickly 
 by using the following rule the method of obtaining which is 
 based on a formula which it is not necessary to discuss here. 
 This rule is: To determine the illumination on a horizontal sur- 
 
 D B 
 
 Fig. 11. — Horizontal illumination. 
 
 face (such as a table or floor) multiply the candle power inten- 
 sity of the ray by the cube of the cosine of the angle between 
 the ray and the perpendicular (this cube of the cosine of any 
 needed angle being found in the accompanying table) and divide 
 this result by the square of the height, in feet, of the lamp 
 above the surface. The result is the horizontal illumination in 
 foot-candles. 
 
 Expressed mathematically this would give the formula: 
 
 cp 
 Horizontal I = ■— cos^* a. 
 W- 
 
 where 
 
 Horizontal / = Foot-candle illumination on surface, 
 
 h = perpendicular distance, in feet, from lamp to 
 surface . 
 
CALCULATION OF ILLUMINATION. 
 
 17 
 
 a = angle between ray and perpendicular, 
 
 cp = candle power in the given direction. 
 
 Example: — Calculate the horizontal illumination on a dining 
 
 room table four feet square, illuminated by a 16-c.p. lamp 
 
 equipped with the opal dome reflector shown in Figs. 21 and 22, 
 
 hung on a pendant three feet above the center of the table. 
 
 Table op Cosines Cubed. 
 
 Angle. 
 
 Cos^ 
 
 Angle. 
 
 Cos' 
 
 1° 
 
 .. ..1.000 
 
 29° 
 
 668 
 
 •> 
 
 998 
 
 30 
 
 649 
 
 3 
 
 998 
 
 993 
 
 31 
 
 630 
 
 4 
 
 32 
 
 610 
 
 5 
 
 988 
 
 33 
 
 ... .590 
 
 6 
 
 983 
 
 34 
 
 ... .570 
 
 7 
 
 978 
 
 35 
 
 ... .550 
 
 8 
 
 . . .971 
 
 36 
 
 529 
 
 9 
 
 963 
 
 37 
 
 ... .509 
 
 10 
 
 955 
 
 38 
 
 ... .489 
 
 11 
 
 945 
 
 39 
 
 469 
 
 12 
 
 935 
 
 40 
 
 . . .449 
 
 13 
 
 925 
 
 41 
 
 ... .429 
 
 14 
 
 913 
 
 42 
 
 ... .410 
 
 15 
 
 901 
 
 43 
 
 ... .391 
 
 16 
 
 888 
 
 44 
 
 ... .372 
 
 17 
 
 874 
 
 45 
 
 ... .353 
 
 18 
 
 860 
 
 46 
 
 .... 335 
 
 19 
 
 845 
 
 47 
 
 ... .317 
 
 20 
 
 829 
 
 48 
 
 ... .300 
 
 21 
 
 ... .813 
 
 49 
 
 ... .282 
 
 22 
 
 ... .797 
 
 50 
 
 ... .265 
 
 23 
 
 780 
 
 ... .762 
 
 51 
 
 ... .249 
 
 24 
 
 52 
 
 ... .233 
 
 25 
 
 ... .744 
 
 53 
 
 ... .218 
 
 26 
 
 ... .726 
 ... .707 
 
 54 
 
 . 203 
 
 27 
 
 55 
 
 ... .189 
 
 28 
 
 ... .688 
 
 56 .' 
 
 ... .175 
 
 Angle. Cos' 
 
 57° 161 
 
 58 149 
 
 59 137 
 
 60 125 
 
 61 114 
 
 62 103 
 
 63 0936 
 
 64 0842 
 
 65 0754 
 
 66 0671 
 
 67 0596 
 
 68 0526 
 
 69 0460 
 
 70 0400 
 
 71 0345 
 
 72 0295 
 
 73 0250 
 
 74 0209 
 
 75 0173 
 
 76 0142 
 
 77 0114 
 
 78 00900 
 
 79 00695 
 
 80 00523 
 
 81 00383 
 
 82 00270 
 
 83 00181 
 
 84 00114 
 
 The best plan is to begin by making a drawing to scale, as in 
 Fig. 12. With the lamp as a center, draw lines A B, A C and 
 A D every 10 degrees. The illumination directly under the lamp 
 will be the candle power 29.3 divided by 9, the square of the 
 height, multiplied by 1, the cube of the cosine of degrees, 
 which gives 3.2 candle feet illumination at point B. To find 
 the illumination at point C take the candle power of the ray 10 
 
18 
 
 PRACTICAL ILLUMINATION. 
 
 degrees from, the vertical which is seen from Fig. 21 to be 29. 
 Then multiply this candle power, 29, by .955 which is the cube 
 of the cosine of 10 degrees found in the table. This product 
 divided by 9, the square of the height, gives 3.0 which is the 
 illumination in candle feet at point C. The illumination at 
 point D is 25.5 candle power multiplied by 0.829 and divided 
 by 9, giving 2.3 candle feet. This process is repeated for each 
 angle at which the illumination is desired. 
 
 In order to bring out the foregoing facts more clearly, take 
 the case of a hall (the plan is shown in Fig. 13), which is 80 feet 
 long, 50 feet wide and twenty feet high, with the lamps located 
 as shown in the drawing. This is Hghted by six gas arc lamps 
 with alabaster globes, hung 15 feet from the floor. In the upper 
 
 Fig. 12. — Calculation of horizontal illumination. 
 
 left hand corner of Fig. 14 is a photometric curve of such a 
 gas arc lamp only one quarter of the curve being shown. The 
 complete curve is shown in Fig. 196 in Chapter XL This can 
 be drawn to any convenient scale. Assume that the plane to be 
 illuminated is three feet above the floor which would place the 
 lamps 12 feet above the plane to be illuminated. Then, draw a 
 horizontal line to scale 12 feet below the position of the lamp 
 in the photometric curve to represent the plane to be illuminated 
 and on this measure off different distances from the foot of the 
 perpendicular all, of course, on the same scale, as shown in 
 Fig. 14. Then draw lines from these different distances toward 
 the source of light until the photometric curve is intersected, 
 as shown in Fig. 14. Near the perpendicular the distances 
 
CALCULATION OF ILLUMLXATIOX. 
 
 19 
 
 should be laid off rather close together, but further away they 
 can be more widely spaced, inasmuch as the change in the 
 value of illumination takes place slowly as we get further from 
 the lamp, and intermediate values can easily be estimated. 
 We can then prepare the table as shown in the upper right hand 
 corner in Fig. 14. The angles a can be determined in one 
 operation by placing a protractor over the diagram. The values 
 of the different candle powers can be obtained by measuring 
 to the same scale as that on which the curve was laid out The 
 value of the horizontal illumination at each point after the 
 candle power and the angle is known, can be at once determined 
 by multiplying the candle power by the cube of the cosine 
 of the angle and dividing by 144, the square of the height. 
 
 I 
 
 ■J 
 
 K 
 
 L 
 
 
 A 
 
 Light 1 
 B 
 
 C 
 
 Light 2 
 
 o 
 
 D 
 
 Light 3 
 
 • 
 
 E 
 
 F 
 
 G 
 
 H 
 
 
 
 Light i 
 
 
 Light 5 
 
 Light 6 
 
 Fig. 13.— Plan of hall 50 x 80 ft. 
 
 This can be done most rapidly by the use of a slide rule. As a 
 check to the work, also as an aid to the eye to show how the 
 illumination varies at different distances from the light, it will 
 be desirable to draw the illumination curve shown on the lower 
 part of Fig. 14. The height of this curve from the base line at 
 any point represents the foot-candles. The values of the 
 horizontal illumination can be plotted to any convenient scale. 
 After obtaining these values we can then determine the illu- 
 mination at different points in the room. In Fig. 13 twelve 
 points have been selected as indicated by the letters A, B, etc. 
 These twelve points give practically all the variation in illu- 
 mination possible. It will be convenient in determining the 
 illumination at the different points to prepare a table as shown 
 
20 
 
 PRACTICA L ILL I 'MIX A TION . 
 
 herewith which shows the distance of the points from the lamps 
 and the illumination therefrom. The different distances of the 
 sources of light from any given point can be at once obtained 
 by measuring them with a scale rule, on the plan, Fig. 13. 
 When the value of the illumination at any point is less than five 
 per cent, of the maximum illumination falling on the same from 
 any one lamp, such values can be neglected. In the table 
 where a dash occurs it means that the illumination from the 
 
 Values of Horizontal Illumination at the 12 Points Indicated 
 
 IN Fig. 13. 
 
 Ft, 
 
 Light 1 
 
 3 
 
 4 
 5 
 6 
 
 Total 
 
 Ft. I B 
 
 Ft. 
 
 570| 01.11 
 .046 30 .093 
 
 15 
 15 
 
 — ' U5 
 
 .084 30' .093 ■ 33 
 
 — ' 42 .040 33 
 
 .70 
 
 1.34 
 
 .362 
 .362 
 .030 
 
 .072 
 .072 
 
 ,90 
 
 Ft. I D 
 
 Ft. E 
 
 30 
 
 .093 ' 18 .274 
 
 Ft. 
 
 15 .362 
 
 11.110 42 .040 '34 
 
 30 ! .093 — ' — ll — 
 
 42 ! .040 18 '.274 '15 
 
 30 j .093 42 .040 il34 
 
 42 I .040 — — ii~- 
 
 i 1 . 47 
 
 .63 
 
 ,067 
 
 362 
 067 
 
 86 
 
 Ft.' 
 
 G iFt. 
 
 H 
 
 Ft. 
 
 I 
 
 Ft.| 
 
 J 
 
 Ft. 
 
 K 
 
 Ft. 
 
 L 
 
 Light 1 1 21 
 
 .208i| 34 
 
 .067 
 
 14 
 
 .396 , 10 
 
 .570 
 
 18 
 
 .274 
 
 32 
 
 .076 
 
 ■■ 2 , 21 
 
 .208,1 15 
 
 .362 
 
 41 
 
 .043:' 32 
 
 .076 
 
 18 
 
 .274 
 
 10 
 
 .570 
 
 " 3 i - 
 
 - 34 
 .208 34 
 
 .067 
 
 — 
 
 — -_ - 
 
 
 46 
 
 .028 
 
 32 
 
 .076 
 
 " 4 21 
 
 .067 
 
 -il 
 
 . 043 40 
 
 .046 
 
 43 
 
 .037 
 
 — 
 
 — 
 
 " 5 , 21 
 " 6l- 
 
 .208 ' 15 
 
 — i 3t 
 
 .362 
 .067 
 
 — 
 
 
 
 43 
 
 .037 
 
 40 
 
 .046 
 
 Total 
 
 .83 !l 
 
 .99 
 
 
 .48 ! 
 
 .69 
 
 
 .64 
 
 
 .77 
 
 lamp is so small that in comparison with the effect of the other 
 lamps, it can be neglected. Having obtained the value of the 
 illumination at different points as shown in the table, it will 
 sometimes be found convenient to plot the resultant illumination 
 obtained by adding the illumination from the different lamps 
 falling on each point to show graphically the variations. 
 
 Fig. 15 shows the value of the illumination at the different 
 points in three planes lengthwise of the hall. Similar curves 
 could be plotted to show the variation of the illumination cross- 
 
CALCULATION OF ILLUMIXATION. 
 
 21 
 
 wise of the hall. It will be noted that the illumination along 
 the sides of the hall is considerably lower than that through a 
 row of lamps or in the center. This value, however, along the 
 sides will be considerably raised by the reflection from the walls 
 and ceiling provided these are light in color. It is probably 
 
 cos3 a 
 
 Hori- 
 zontal I 
 
 
 
 0. 
 
 1.000 
 
 160 
 
 1.11 
 
 1 
 
 4.5 
 
 .990 
 
 165 
 
 1.13 
 
 2 
 
 9. 
 
 .963 
 
 174 
 
 1.16 
 
 3 
 
 14. 
 
 .913 
 
 185 
 
 1.17 
 
 4 
 
 IS. 
 
 .860 
 
 195 
 
 1.16 
 
 5 
 
 22. 
 
 .797 
 
 196 
 
 1.10 
 
 6 
 
 26. 
 
 .726 
 
 188 
 
 .950 
 
 7 
 
 .30. 
 
 .649 
 
 183 
 
 .830 
 
 8 
 
 33.5 
 
 .580 
 
 182 
 
 .734 
 
 9 
 
 36.5 
 
 .519 
 
 182 
 
 .657 
 
 10 
 
 39.5 
 
 .459 
 
 183 
 
 .570 
 
 12 
 
 45. 
 
 .353 
 
 189 
 
 .463 
 
 14 
 
 49. 
 
 .282 
 
 202 
 
 .396 
 
 16 
 
 53. 
 
 .218 
 
 216 
 
 .327 
 
 18 
 
 .56. 
 
 .175 
 
 225 
 
 .274 
 
 20 
 
 59. 
 
 .137 
 
 235 
 
 .224 
 
 24 
 
 63. 
 
 .0936 
 
 244 
 
 .159 
 
 28 
 
 66.5 
 
 .0633 
 
 250 
 
 .110 
 
 .32 
 
 69.5 
 
 .0430 
 
 256 
 
 .0765 
 
 36 
 
 71.5 
 
 .0320 
 
 260 
 
 .0596 
 
 40 
 
 73. 
 
 .0250 
 
 267 
 
 .0464 
 
 Horizontal I. 
 
 Fig. 14. — Photometric and illumination curves of "Gas Arc.'' 
 
 safe to assume that the values of the illumination in this room 
 will be raised at least 50 per cent., providing the colors are 
 light. In this connection it is well to note that the amount 
 to be allowed for the reflection of the walls and ceiling depends 
 upon the amount of light which is thrown on them. If, for 
 example, reflectors were used oh the lamps so that practically 
 
22 PRACTICAL ILLUMINATION. 
 
 no light was allowed to fall on the ceiling and but little on the 
 walls, but very little could be allowed for reflection even though 
 the color effects were light. If on the other hand, as in this 
 case, half of the light is above the horizontal as will be noted 
 by referring to Fig. 196, a large margin can be allowed for re- 
 flection. It has been shown by Mr, Preston S. Millar in a paper 
 before the Illuminating Engineering Society in Nov. 1906, that 
 the reflection from walls and ceilings may increase the illumina- 
 tion as much as 166 per cent, with light buff walls and ceilings 
 and lamps at the ceiling without reflectors. 
 
 Primarily in calculat:ng the illumination one should consider 
 whether or not horizontal or normal illumination should be used. 
 When a book or paper is held in such a position that the maxi- 
 mum amount of light from any given source falls upon it, one 
 should figure on normal illumination. It is evident, however, 
 that if the book were held in this position one could not figure 
 on its receiving normal illumination from some other source 
 
 Fig. 15. — Illumination on three planes of hall. 
 
 of light that might be in the room. If it is desired to obtain 
 a definite illumination from the two sources of light the book 
 would have to lie horizontally. If then, as is ordinarily the 
 case, it is desired to determine the illumination at any point 
 in the room from a number of sources of light, one should use 
 the horizontal illumination rather than the normal. 
 
 If it is desired to calculate the illumination on the walls, or 
 what is known as vertical illumination in one vertical plane, 
 it is simply necessary to turn the photometric curve through 
 90° and consider instead of the height above the floor the hori- 
 zontal distance from the source of light to the wall. 
 
 The authors have omitted here consideration of the case of 
 lamps hanging at an angle inasmuch as this problem is rather 
 involved and too complicated for ordinary calculation. 
 
 Quantity of Light. 
 In laying out lighting for any purpose it is first best to deter- 
 mine, at least approximately, the quantity of light to be used. 
 
CALCULATION OF ILLUMINATION. 23 
 
 The approximation can later be revised after calculations are 
 made, in case too much or too little light has been assumed. 
 
 The amount of light necessary for good illumination depends 
 upon the purpose for which the place to be lighted is to be used 
 and also on the individual's ideas on the subject. Thus the 
 requirements for street, house and office lighting present great 
 difEerences. Further the amount of light necessary depends 
 largely upon the character of the surrounding surfaces, such as 
 walls, ceiling, etc., as well as upon the absorption of the shades, 
 if such are used, as well as their directing power. 
 
 It is safe to say that the increase in illumination by reflection 
 from walls and ceilings may be anywhere from 5 to 200 per 
 cent, depending on the color of th= walls, the size of room and 
 the location and equipment of lamps. At present writing actual 
 tests on this subject are few. 
 
 There are four methods in general use by which the quantity 
 of light necessary can be approximated. 
 
 1st. The number of candle power per square foot. 
 
 2d. The number of candle power per cubic foot. 
 
 3d. The watts per square foot (used of course only in the 
 case of electric light). 
 
 4th. The number of foot-candles. 
 
 The best practice where actual work is concerned is to use 
 all four methods if the rooms are large. 
 
 The mean spherical candle power of the lamps to be used 
 should preferably be made the basis of calculations rather than 
 the rated candle power, as the latter figure in some cases means 
 very little. To correctly judge the efficiency of a source of 
 light not only the rated (which is generally the maximum) 
 light but also the light given in other directions, must be taken 
 into account; that is, the mean spherical candle power. In 
 other words, if the source of light in question should give the 
 same total amount of light, but equally in all directions we would 
 term this the mean spherical candle power. In an oval anchored 
 filament incandescent lamp the mean spherical candle power is 
 about 85 per cent, of the rated horizontal candle power. With 
 the upright mantle gas burner it is about 76 per cent. With 
 the open gas jet the percentage is nearly 100. 
 
 The amount of light per square foot, as before noted, varies 
 widely, being in some cases as high as one candle power per 
 
24 PRACTICAL ILLUMINATION. 
 
 square foot, where a very strong light is desired, and again as 
 low as one candle power per five square feet in a large room 
 where simply a general illumination is required. The mean of 
 these two or about one candle power per three square feet, 
 when the allowance for reflection from walls and ceilings is about 
 50 per cent., is a good working basis for ordinary illumination. 
 
 Thus in the case of the room Figs. 13, 14 and 15, we have 
 2'^ square feet per mean spherical candle power; the mean spher- 
 ical candle power of the gas arc lamp used being given in con- 
 nection with Fig. 196. This also checks up with the other 
 methods inasmuch as an allowance of 55 cubic feet per candle 
 power has been made and as, allowing 50 per cent, for reflection 
 from walls and ceiling the average foot candles on the plane 
 illuminated is about 1.3. 
 
 For the cubic foot basis the height of the lamps from the 
 floor plays an extremely important part. In practice, however, 
 it is safe to allow one candle power for from 25 to 50 cubic feet. 
 The latter is sufficient in places where the walls are of light 
 color, but where dark, would not be satisfactor3^ 
 
 Under the fourth method 0.5 of a horizontal foot-candle will 
 give a fair illumination. For reading purposes this should be 
 raised to a minimum of one foot-candle, and for lighting desks 
 this should be increased to from two to five foot-candles for best 
 results. Drafting tables require from five toten foot-candles. 
 
CHAPTER V. 
 
 INDIVIDUAL INCANDESCENT ELECTRIC LAMPS AND THEIR 
 REFLECTORS, SHADES AND GLOBES. 
 
 In designing illumination it is usually important to consider 
 not only the location of the lamps but the candle power given in 
 various directions by each light. In, other words, each source 
 of light, together with the reflector, globe or shade which ac- 
 companies it, must be considered as a unit before arriving at 
 any conclusion as to the illumination. In the majority of 
 cases where the resultant efficiency is any object, it is desirable 
 to have some means of directing the rays from the sources of 
 light, so that the strongest light will fall where most wanted. 
 Manifestly, it is of first importance to know what various types 
 of reflectors, shades and globes will accomplish in this respect, 
 as there is an immense difference in the distribution of light 
 with different forms of glassware. 
 
 Reflectors, shades and globes are used for three purposes: 
 (1) Softening and diffusing the light, (2) directing the rays of 
 light in certain useful directions, (3) for ornament. A reflector 
 shade or globe may combine in one all these three character- 
 istics, or it may have only one of them. It is too often the case 
 that the one chosen is not the right thing for the place, or is 
 chosen for its general appearance without any attempt to learn 
 what it practically accomplishes in the way of altering the 
 natural distribution of light from the lamp. 
 
 The reader can best be brought to realize the importance of 
 a proper choice of reflectors and the great difference between 
 reflectors, shades and globes by a study of tests made on lamps- 
 variously equipped, showing the amount of light thrown in 
 different directions above and below the lamp. All of the fol- 
 lowing diagrams or photometric curves were taken from tests. 
 made by the Electrical Testing Laboratories of New York^ 
 except where otherwise specified. 
 
 25 
 
26 
 
 PRACTICAL ILLUMIKATlOy. 
 
 Fig. 16 shows the candle power or intensity of the light thrown 
 in different directions, in a vertical plane from a bare 16-candle 
 power, oval anchored filament incandescent lamp with a clear 
 glass bulb suspended vertically, as indicated in the diagram, 
 and without any reflector or globe. It is seen that the lamp 
 gives 16-candle power in a horizontal direction. In other 
 directions the light is less. Directly below the lamp, the candle 
 power is 6.6. It is evident that a bare lamp like this, hung in 
 a room, will throw more light in the directions of the walls than 
 anywhere else. Nearly one-half will also go toward the high 
 side walls and the ceiling. There are certain places where this 
 is desirable, as in warehouses and storage rooms. 
 
 ^- ^ 4?-- 
 
 Candle Power 
 Fig. 16. — Light about a bare incandescent lamp. 
 
 Fig. 17 gives the light distribution about an incandescent 
 lamp with a bulb frosted by being etched with acid. This test 
 was made after frosting the same lamp on which the test (Fig. 16) 
 was made. By the term frosting is here meant the usual pro- 
 cess of producing a rough surface on the glass itself by the use 
 of acid, and not the coating or painting of the surface of the 
 glass with white frosting. The latter process, known as dipping, 
 is not as satisfactory, because unless a thick enough coat is 
 applied to seriously cut down the light the glare from the bare 
 filament is but imperfectly diffused. This test showed the 
 absorption or loss of light due to frosting amounted to 11.5 
 per cent. The mean of other tests given at the end of this 
 chapter show only ten per cent, loss, In Fig. 16 the mean 
 
INDIVIDUAL IXCAXDESCEXT ELECTRIC LAMPS. 
 
 27 
 
 spherical candle power from the clear glass bulb incandescent 
 lamp was 13.2. After frosting the bulb as in Fig. 17 the mean 
 spherical candle power was 11.7. The frosting of the bulb, it 
 is seen, does not change the distribution of light about the 
 lamp materially, but simply reduces the amount of light thrown 
 in all directions, except directly below the tip of the lamp, 
 where the candle power is increased from about 6.6 to 8.9. 
 The loss of 11.5 per cent of the light by frosting of the bulb is 
 more than made up for in all locations where the lamps are 
 placed within the line of ordinary vision; that is, when their 
 rays fall directly in the eyes. The frosting of the glass diffuses 
 the light so that it does not have the blinding effect that the 
 bare incandescent filament has upon the eves. Thus, although 
 
 14.3 
 
 Candle Power 
 Fig. 17. — Light about a frosted incandescent lamp. 
 
 the actual amount of light may be reduced 11.5 per cent, or more, 
 one can really see better, and therefore the illumination is better, 
 with the frosted bulb lamps wherever the lamps are so located 
 that the rays fall directly in the eye. For this reason it is de- 
 sirable to use a frosted bulb in most locations where an incan- 
 descent lamp is to be used without reflectors or globes, which 
 completely hide the filament. It is much easier to see beyond 
 a frosted bulb lamp which is nearly in the line of vision than to 
 see beyond a lamp with a clear glass bulb, for reasons explained 
 in the chapter on Light and the Eye. In some places where 
 it is desirable to use frosted bulbs, there is so much dirt that 
 their use is inadvisable, because a rough frosted glass when 
 it is once dirty is difficult to clean. 
 
•-'8 PRACTICAL ILLUMINATION. 
 
 A point open to argument is whether, if a diffusing cover is 
 wanted, it is better to use a frosted bulb lamp rather than a 
 surrounding globe of frosted, ground or sand-blasted glass. 
 This is taken up later in the close of this chapter. 
 
 One thing decidedly in favor of frosted as against clear glass 
 bulbs is their artistic appearance. The light from them is much 
 more pleasing than from the clear bulb, which exposes the 
 filament. 
 
 Frosted or sand-blasted bulbs so diffuse the light as to over- 
 come the streaks which are quite pronounced with the clear 
 glass bulb. These streaks are very objectionable where reading 
 writing, draughting or any kind of close work is to be done. 
 They cause a strain on the eyes which the user of the light is 
 generally not aware of, and for that reason are especially to be 
 avoided. 
 
 
 32.0 'C 
 
 Fig. is. — Light about a tantalum filament lamp. 
 
 Where there is a mixture of daylight and artificial light, that 
 from an incandescent lamp with a clear glass bulb lamp is likely 
 to present a sickly yellow appearance. This sickly appearance 
 can be very much reduced, although not entirely eliminated 
 by the use of frosted or sand-blasted glass. This should be 
 remembered in the lighting of stores, halls of hotels and office 
 buildings, and all places where considerable light from incan- 
 descent lamps must be used during daylight hours in combina- 
 tion with daylight. Even at night the frosted glass greatly 
 improves the appearance of incandescent electric lamps. 
 
 Fig. 18 shows the distribution of light about a tantalum 
 metallic filament lamp, this test being one made by Dr. Louis 
 Bell and Prof. W. L. Puffer, at the Massachusetts Institute of 
 Technology, and given in the Electrical World and Engineer, 
 June 3, 1905. The filament of this lamp is supported on a 
 
INDIVIDUAL INCANDESCENT ELECTRIC LAMPS. 29 
 
 Spider in such a way that the horizontal candle power is greater 
 in proportion to the mean spherical than in the ordinary 16- 
 candle power, carbon, oval anchored filament lamp. The 
 mean spherical candle power of this tantalum lamp is 17.4. The 
 watts consumed per mean spherical candle power are about 
 2.53, the watts per horizontal candle power being about 1.85. 
 These lamps, as made in America, require about two watts 
 per horizontal candle power. This lamp calls for the same 
 treatment as regards reflectors, shades and globes as the regular 
 carbon filament incandescent lamps. 
 
 Opal Glass Reflectors. 
 
 Fig. 19 gives the light distribution about a 16-candle-powerincan- 
 descent lamp covered with what is commonly known to the trade 
 as a fluted opal cone reflector 6.5 inches in diameter, the form 
 of which is shown in Fig. 20. This is one of the best and most 
 efficient of the common cheap types of reflectors. It is useful 
 wherever a large amount of light from the lamp is needed below 
 the horizontal. It so changes the distribution of light about 
 the incandescent lamp as to throw from 23 to 25 candle power 
 in all downward directions below 45° from the horizontal. 
 The performance of opal reflectors of this kind depends somewhat 
 upon the density of the opal glass used. The more dense the 
 opal the greater the amount of light reflected and the less that 
 passing through the reflector. This particular reflector is 
 fairly dense and hence lets comparatively little light escape 
 above the horizontal. Enough, however, goes through the 
 reflector to keep the walls and ceiling of a room from being in 
 darkness. Such a reflector can properly be used in all cases 
 where it is not objectionable from an artistic standpoint, and 
 it is desirable to direct the light downward and at the same 
 time allow a small amount of light to go to the walls and ceilings. 
 
 This opal cone reflector does not give anywhere near as great 
 a maximum downward candle power as some of the reflectors 
 which will be shown later, but it illustrates very well what a 
 great difference can be made in the amount of light thrown down 
 from a lamp placed in a pendant position by the use of the 
 proper reflector. Thus, for example, this reflector gi^^es a 
 candle power directly below the lamp (that is, a tip candle 
 power) of 23.6 and at the angle of 40 degrees from the vertical, 
 
30 
 
 PRACTICAL ILLVMIXATIO.Y. 
 
 a candle power of 25. The bare lamp would give only 6.6 
 candle power at the tip and less than 12 candle power at the 
 angle of 40 degrees from the vertical. As lamps are so frequently 
 placed high up in a room on chandeliers or pendants or some- 
 times at the ceiling, the importance of directing the rays down 
 
 Candle Power 
 Fig. 19. — Light about a fluted opal cone reflector. 
 
 so that they will be of use to the occupants of the room is ap- 
 parent. 
 
 The ceilings and walls if light in color have considerable 
 value as reflectors and diffusers. "\Miere maximum efficiency 
 is desired, it must be remembered, however, that a reflector to 
 
 J 
 
 Fig. 20. — Fluted opal cone reflector. 
 
 be efficient must fit over a lamp and not be some distance from 
 it, and that further, without reflectors much of the light is re- 
 flected back and forth from wall to wall two or three times with 
 much loss each time before it reaches the place that specially 
 needs illumination. 
 
INDIVIDUAL IXCANDESCENT ELECTRIC LAMPS. 
 
 31 
 
 Fig. 21 shows the distribution of Hght about a 16-candle 
 power incandescent lamp in a white opal dome reflector, the 
 form of which is shown in Fig. 22. This is an excellent reflector 
 for many common uses. Being 3.75 inches deep and 7.25 
 
 Candle power. 
 Fig. 21. — Light about a 7-in. opal dome reflector. 
 
 inches in diameter, it covers the lamp so as to act as a diffusing 
 globe, to a certain extent, while at the same time it gives an 
 excellent distribution of light in all downward directions, the 
 maximum being somewhat over 29 candle power. It is a good 
 
 Fig. 22. — 7-in. opal dome reflector. 
 
 type of reflector to place pointing straight downward over the 
 center of a dining-room table, one lamp giving illumination over 
 the entire top of a fairly large table. (See Fig. 12.) It can also 
 be used' on chandeliers for the general lighting of all classes of 
 
32 
 
 PRACTICAL ILLUMINATION. 
 
 rooms. When so used it should be pointed straight down, 
 because if placed at an angle it will give more light than is 
 necessary on the high side walls. It is interesting to compare 
 this with the opal reflectors, Fig. 19 and 23. The maximum 
 candle power of this opal dome reflector is considerably greater 
 than from the fluted opal cone, shown in Fig. 19. This candle 
 power extends about ten degrees each side of the vertical, and 
 
 Candle Power 
 Fig. 23. — Light about an opal bell reflector. 
 
 the balance of the distribution is very similar to that of the 
 fluted opal cone. This would be a good reflector for reading 
 and desk use, but for the fact that having a smooth reflecting 
 surface, the light is somewhat streaked. The streaks can be 
 overcome by the use of a frosted bulb lamp, as in other types 
 of smooth reflectors. Although not specially ornamental in 
 its usual form, as found on the market, it could easily be made 
 
 Fig. 24. — Opal bell reflector. 
 
 so. It is another good example illustrating how little attention 
 is commonly paid to combining the useful and artistic in glass- 
 ware and how much room there is for improvement and develop- 
 ment along this line. 
 
 In Fig. 23 is shown the light distribution about a 16-candie 
 power incandescent lamp equipped with a five-inch opal bell 
 reflector of the type shown in Fig. 24. This is a very common 
 
INDIVIDUAL INCANDESCENT ELECTRIC LAMPS. 33 
 
 type of reflector and is excellent for securing general distri- 
 bution of light in all downward directions where it is not desired 
 to concentrate a great amount of light at any one point. As the 
 lamp bulb projects a considerable distance out of this reflector, 
 
 Cnnclle power. 
 Fig. 25. — Light about a green plated opal desk shade. 
 
 it is frequently necessary to use it with a frosted bulb lamp on 
 account of the large amount of exposed filament. If such a 
 reflector is to be used on a chandelier of ordinary height in a small 
 
 Fig. 26. — Green plated opal desk shade. 
 
 room, the chandelier- should preferably be equipped with sockets 
 pointing straight down or nearly so in order that an unnecessary 
 amount of light be not thrown upward. Such a reflector put 
 on a chandelier with lamp sockets at an angle of 45 degrees 
 
34 
 
 PRACTICAL ILLUMINATION. 
 
 will evidently throw mucli more light upward than is necessary 
 unless there happen to be pictures or other objects high in the 
 room which are to be illuminated. 
 
 Fig. 25 gives the light distribution about a 16-candle power 
 incandescent lamp equipped with an opal desk shade seven 
 inches in diameter with green backing, shown in Fig. 26. This 
 
 
 Candle power. 
 Fig. 27. — Light about a flat opal reflector. 
 
 is the type of shade most commonly found in offices to-day. 
 It is, .of course, intended to concentrate light below the tip of 
 the lamp for office desk purposes and let but little light through 
 where it can reach the eyes of the users. The maximum candle 
 power is about 34. The candle power tests on this shade do 
 not show the very objectionable feature of it, namely, that it 
 
 Fig. 28.— Flat opal reflector. 
 
 throws a very streaked light. A frosted bulb lamp should 
 invariably be used with such a shade, as this would eliminate 
 the streaks and make its use tolerable. Otherwise it should 
 be prohibited by law. 
 
 Fig. 27 gives the light distribution about a flat opal reflector 
 ten inches in diameter, the form of which is shown in Fig. 28. 
 
INDIVIDUAL INCANDESCENT ELECTRIC LAMPS. 35 
 
 The reflector so changes the natural distribution of Hght from 
 an incandescent lamp as to give an increase of light in all direc- 
 tions below the horizontal and very little above the horizontal. 
 It is good to use wherever it is desired to light dimly a large area 
 from one lamp without concentrating much light below the 
 lamp. Its most extensive application is in basements, where 
 it should be placed near the ceiling to keep the lamp out of the 
 line of vision, and also in halls of the cheaper class of buildings. 
 It does not throw enough of its light down to fit it for use on 
 high ceilings. One of its chief disadvantages is that it leaves 
 the lamp absolutely unshaded so that frosted lamps should 
 be used with it both to diffuse the glare and obviate the streaked 
 light from the bare filament. 
 
 Prismatic Glass Reflectors. 
 
 The principle of the prismatic glass reflector was shown in 
 Fig. 6, Chapter II. In this the reflection is accomplished by 
 means of prisms pressed in the glass. The popularity of this 
 class of reflectors lies in the fact that they can be designed so 
 as to give almost any kind of a downward distribution of light 
 desired while at the same time enough light passes through the 
 glass without reflection so that the ceilings and upper walls of 
 a room are not left in darkness. Some of these reflectors are 
 designed to be very powerful concentrators of light, while others 
 distribute the light over a wide angle below the lamp, giving 
 an even illumination over a large area. In the practical appli- 
 cation of light, A-ery many cases arise where it is desired to have 
 the greater part of the light concentrated under the lamp, and 
 yet where it is not desired to have the reflector entirely opaque. 
 The prisms break up the light rays so that the reflected light 
 with these types is almost free from streaks. They have also 
 the advantage over opal and ground glass that the dirt which 
 inevitably collects on reflectors in some locations does not show 
 as badly as on opal or ground glass. That collecting on top of 
 the reflector does not interfere with its reflecting power. 
 
 Not every glass reflector that has corrugations upon it is to 
 be considered a true prismatic reflector. The prisms must be 
 accurately made and the reflector correctly shaped. Such 
 reflectors should not be selected on appearances or shape, but 
 rather from photometric tests which show just what light dis- 
 
36 PRACTICAL ILLUMINATION. 
 
 tribution they give. In the absence of such tests or their equi- 
 valent one is working absolutely in the dark. 
 
 Fig. 29 shows the distribution of light about a 120-watt 
 Meridian incandescent lamp with a prismatic reflector. This 
 lamp had the old style common carbon filament, not the new, high 
 efficiency Gem filament. This lamp, which is shown in Fig. 30, 
 has a large spherical frosted bulb. The upper part of the bulb 
 coming next to the reflector is not frosted. It will be seen that 
 
 -0.0 ^.g <- ^ 
 
 .^J-^^^, 
 
 Candle powei*. 
 
 Fig. 29. — Light aboitt 120-watt meridian lamp with prismatic 
 reflector. 
 
 this gives a fairly concentrated downward light. It should 
 not, therefore, be attempted to make this lamp cover too great 
 an area. On account of its artistic appearance, the lamp is 
 well adapted to use on a ceiling to be studded with a number of 
 lights, or for the lighting of a small room with one central light 
 at the ceiling. It is also suited for use on chandeliers where a 
 concentrated light is desired below the chandelier. As a light 
 producer this lamp is very similar to a standard frosted bulb 
 32 candle power incandescent lamp equipped with a similar 
 
IXDIVIDUAL IXCAXDESCEXT ELECTRIC LAMPS. 37 
 
 reflector. Its cost is somewhat higher than the standard lamp, 
 but its appearance is by some considered enough better to justify 
 the expense. 
 
 We will now take up several representative types of prismatic 
 glass reflectors designed for use over standard incandescent 
 lamps with standard shade holders. As many of the photo- 
 metric tests on these reflectors are obtainable from their manu- 
 facturers upon application, it will not be necessary here to 
 give anything further than the tests of a few representatiA^e 
 types. Only enough tests will be given to afford the reader a 
 good idea of what the various types accomplish and how they 
 compare with other kinds of reflectors. 
 
 Fig. 30. — 120-watt meridian lamp. 
 
 Fig. 31 shows the distribution of light about a 16-candle 
 power incandescent lamp in the prismatic reflector, shown in 
 Fig. 32, which is in extensii-e use, and concentrates a large 
 amount of light below the lamp. The maximum candle power 
 for an angle of a few degrees is as high as 57, while it gives more 
 than would the bare incandescent lamp at all angles below 
 30 degrees from the horizontal. This reflector is suited to all 
 locations where concentrated light is desired at one point, without 
 distribution over a wide area. This includes desk lamps and 
 reading lamps on downwardly pointed pockets on chandeliers. 
 It would be a mistake to attempt the general illumination of a 
 
38 PRACTICAL ILLUMIXATION. 
 
 large room with this particular reflector unless a number of 
 lamps pointing in various directions were used on a chandelier, 
 or unless they were studded over the ceiling. 
 
 Candle power. 
 Fig. 31. — Light about a prismatic reflector. 
 
 Fig. 32. — Prismatic reflector. 
 
 Fig. 33 shows the distribution of light about a 16-candle power 
 incandescent lamp in the prismatic reflector shown in Fig. 34. 
 It does not concentrate the light as much as the reflector shown 
 in Fig. 31, and for that reason will be found of more general 
 
INDIVIDUAL INCANDESCENT ELECTRIC LAMPS. 
 
 39 
 
 application than the former, as there are many cases where 
 the light from the other would be too much concentrated at 
 one point. 
 
 Fig. 35 shows the distribution of light about a 32-candle 
 power incandescent lamp in the prismatic reflector shown in 
 
 Candle power. 
 Fig. 33. — Light about a prismatic reflector. 
 
 Fig. 36. It represents an extreme case of concentrating light 
 over a small area, yet at the same time a limited amount escapes 
 through the reflector, so that it is not entirely opaque. The 
 same reflector with a 16-candle power lamp gives about 120- 
 
 FiG. 34. — Prismatic reflector. 
 
 candle power at the tip. A reflector of this kind is manifestly 
 suited only to places where a small but very concentrated beam 
 of light is wanted, as in lighting store windows, draughting, 
 lighting of small desks, and pendant reading lights used by one 
 person only. 
 
40 
 
 PRACTICAL ILLUMINATION. 
 
 Fig. 37 shows the distribution of light about a 16-candle 
 power incandescent lamp in the prismatic reflector shown in 
 Fig. 38. The distribution of light is in marked contrast to that 
 in Fig. 35, as this reflector is designed to throw an even illumina- 
 tion over a considerable area below it. This reflector comes 
 down over the lamp, so as to protect the eyes from the glare 
 of the bare filament when it is hanging pendant, even more 
 than would the opal dome reflector, test on which has been 
 
 Candle power. 
 Fig. 35. — Light about a prismatic concentrating reflector. 
 
 shown in Fig. 21. The contrast between Figs. 35 and 37 makes 
 it clear how important it is to select the proper reflector for a 
 given piece of work. It would evidently be a great mistake to 
 select the concentrating reflector shown in Fig. 35 for use on 
 a chandelier where it is desired to make one or two lamps furnish 
 general illumination for an entire small room. If this were done, 
 the result would be very strong light immediately under the 
 lamp if the lamp were in a pendant position (pointing straight 
 
INDIVIDUAL INCANDESCENT ELECTRIC LAMPS. 41 
 
 down), and the rest of the room would receive only a small 
 amount of Hght. The reflector shown in Fig. 38 would throw 
 the light evenly over the entire floor of a small room, but would 
 not make a very intense light at any one place. It would, on 
 the other hand, be a great mistake to put the reflector shown 
 in Fig 37 at the top of a store window for lighting the goods in 
 the window, or on a chandelier for reading purposes immedi- 
 ately under the chandelier. 
 
 Prismatic reflectors are also made which give a distribution 
 of light which is a compromise between the concentration shown 
 in Fig-. 35 and the distribution shown in Fig. 37. 
 
 Fig. 39 shows the distribution of Hght about a 16-candle 
 power incandescent lamp in the prismatic reflector shown in 
 Fig. 40. This reflector is covered either with white cardboard or 
 
 Fig. 36. — Very concentrating prismatic reflector. 
 
 celluloid to prevent light from passing through as it is intended 
 for desk use where it is placed within the ordinary range of vision. 
 
 Another incandescent lighting unit using prismatic glass 
 reflectors is the new high efliciency graphitized fllament Gem 
 incandescent lamp. These are sold with substantial shade 
 holders, which are designed to hold the shade in proper focus. 
 They should always be used with such holders; otherwise they 
 will not give the desired distribution. The lamps are usually 
 equipped with frosted tip bulbs, as they gi^^e a very powerful 
 light, and if some means of diffusion were not provided the 
 blinding effect due to the glare of the filament would be very great. 
 
 The following tests, Figs. 41 to 48, were made on lamps of 
 this type by Prof. H. E.- Clifford, of Massachusetts Institute of 
 Technology and reported to the National Electric Light Asso- 
 ciation in 1906. 
 
42 
 
 PR AC TIC A L ILL U.\ II X A TIOX . 
 
 Fig. 41 shows the light about a 125-watt clear bulb, double 
 loop, graphitized filament, high efficienc3', Gem incandescent 
 lamp without reflector. The mean spherical candle power is 
 40.7, making the consumption per mean spherical candle power 
 3.1 watts. 
 
 Candle power. 
 
 Fig. 37. — Light about a distributing 
 prismatic reflector. 
 
 Fig. 38. — Prismatic bowl 
 reflector. 
 
 Fig. 42 shows the light about a lamp of the same type frosted 
 only on the tip. The loss due to frosting is about three per 
 cent. 
 
 Fig. 43 shows the light about a clear bulb 125-watt lamp 
 
 I'' ' ' \ 
 
 <^i / ! \ ;* 
 
 \ w 
 
 \ 7 
 
 
 Hj 
 
 ^^^^Hn^^' 'llfMBsm 
 
 piM 
 
 
 K?«jn.\\\/ .aIBS^^H 
 
 
 BBHm\V^^flBI 
 
 
 HMhH 
 
 Fig. 39. — Light about a prismatic 
 desk shade. 
 
 Fig. 40. — Prismatic desk shade 
 \Tithout cardboard cover. 
 
 equipped with a prismatic reflector of the distributing type, 
 shown in Fig. 44. 
 
 Fig. 45 shows the light distribution about the same lamp and 
 reflector as in Fig. 44, except that the tip of the lamp is frosted. 
 
 Fig. 46 shows the light about a clear bulb lamp of the same 
 
INDIVIDUAL IXCAXDESCEXT ELECTRIC LAMPS. 
 
 4.3 
 
 type equipped with a concentrating prismatic reflector shown 
 in Fig. 47. 
 
 Fig. 35 shows the light about the same lamp and reflector as 
 in Fig. 46, except that the tip of the lamp is frosted. 
 
 This type of lamp and reflector is undoubtedly suitable for a 
 large number of cases, but care should be exercised in its use ; 
 
 4^.& 
 
 Fig. 41. — Light about a clear bulb 12.5 watt, Gem lamp. 
 
 one advantage of this type of lamp is that it can and should be 
 placed close to the ceiling, where it is out of the ordinary range 
 of vision, and yet give a good distribution of light for general 
 lighting purposes. Thus, in the case of an office where simply 
 general illumination is wanted and the lamp is placed at the 
 ceiling, this would be suitable, although the desks would prob- 
 ably have to be lighted bv their own individual lamps, because 
 
 Fig. 42. — Light about a frosted tip 12.3 watt, Gem Lamp. 
 
 the writers would probably be directly in their own shadows. 
 In the case of stores, halls, and similar places where a large 
 volume of light is wanted this type can be used to good advant- 
 age to replace ordinary clusters wherever a larger unit than 
 16-candle power is desired. Owing to the high temperature 
 at which they are operated they give a whiter Hght than the 
 ordinary incandescent lamp. Their size calls for care in prevent- 
 
44 
 
 PRACTICAL. ILLVAIIN A TION. 
 
 ing the light from a bare filament striking the eye. 
 where concealed they should be used with frosted tip. 
 
 Except 
 
 Opaque Metal and Mirror Reflectors. 
 Fig. 49 gives the light distribution about a 16-candle power 
 incandescent lamp with an eight-inch metal reflector painted 
 white inside , the form of wliicia' is shown in Fig. '50. This is the 
 
 Fig. 43. — Light about a clear 125-watt 
 Gem lamp w-ith distributing reflector. 
 
 Fig. 44. — Gem lamp and dis- 
 tributing reflector. 
 
 Fig. id. — Light about frosted Gem 125 watt lamp with distributing 
 
 reflector. 
 
 kind of reflector so often seen in workshops, draughting-rooms 
 and offices. It is designed to concentrate considerable light 
 below the lamp, and succeeds fairly well in this, as can be seen. 
 For a space of ten degrees each side of the vertical, it gives 
 over 36-candle power. Being of metal and entirely opaque, of 
 course no light is thrown above the edges of the shade. The 
 serious objection to such a shade as it stands is the extremely 
 
INDIVIDUAL INCANDESCENT ELECTRIC LAMPS. 45 
 
 Fig. 46. — Light about clear bulb Gem 
 125-watt lamp with concentrating re- 
 flector. 
 
 Fig. 47. — Gem lamp and con- 
 centrating reflector. 
 
 28) ^._-^ , ._\23 
 
 Fig. 48. — Light about a frosted tip Gem 125-watt lamp with 
 concentrating reflector. 
 
46 
 
 PRACTICAL ILLUMIXA TIOX. 
 
 streaked light which it casts. In other words, it does not 
 ehminate the streaks which are inherent in the incandescent 
 ]amp which is the source of hght. Such a shade, if painted 
 inside with aluminum paint, will give a well-diffused light, 
 
 Fig. 49.- 
 
 Candle Power 
 -Light about a metal cone reflector. 
 
 practically free from streaks. Another way of getting over 
 these streaks is to use a frosted bulb lamp. Although the 
 authors have not secured tests on the relative efhciency of the 
 reflector coated with white paint inside as against the reflector 
 
 Fig. 50. — Metal, Sin. reflector. 
 
 coated with aluminum paint, some rough, practical trials have 
 conA'inced them that there is not enough difl^erence between 
 the white paint and the aluminum paint to be worth considering 
 as far as the amount of light is concerned, while the quality of 
 the light from the aluminum paint is vastly superior because 
 
W DIVIDUAL INCANDESCENT ELECTRIC LAMPS. 47 
 
 of the freedom from streaks. These cheap metal reflectors are 
 in very common use in draughting rooms and offices, where 
 it is of great importance to the welfare of the users that they be 
 not forced to strain their eyes with a streaked light of this kind. 
 If the owners of such shades do not feel able to replace them 
 with some of the more efficient types mentioned later, they can 
 
 Fig. 51. 
 
 Candle power. 
 -Light about an aluminum cone reflector. 
 
 at least afford to have them coated inside with aluminum paint. 
 Fig. 51 gives the light distribution about a 16-candle power 
 incandescent lamp in an aluminum cone reflector with frosted 
 aluminum finish inside. This reflector, which is shown in Fi-g. 
 52 is a fairly powerful concentrator giving about 47-candle 
 power below the tip of the lamp. On the whole it is much 
 preferable to the green opal reflector shown in Fig. 26, although 
 
4S PRACTICAL ILLUMIXATIOX'. 
 
 perhaps not as pleasing in appearance when lighted, as it is 
 opaque. It gives a well diffused light free from streaks on 
 account of its frosted reflecting surface. A point to be remem- 
 bered in purchasing aluminum reflectors for office and draught- 
 ing room use is that they should have a frosted rather than a 
 polished aluminum surface. A polished surface will invariably 
 give a streaked light, which is trying to the eyes if the light is 
 perfectly stationary and still worse when the light swings or 
 vibrates slightly. Besides its excellent light-reflecting quali- 
 ties, the aluminum cone is very light and not easily broken. 
 This fits it well for use on portable and adjustable fixtures in 
 offices and shops. It should be considered along with other 
 efficient opaque reflectors mentioned later. 
 
 Fig. 53 shows the distribution about a 16-candle power lamp 
 equipped with the shade shown in Fig. 54. The light distri- 
 
 FiG. 52. — Aluminum cone reflector. 
 
 bution from a 32-candle power lamp with the same reflector 
 is shown in Fig. 55. This is a shade with a corrugated mirror 
 reflecting surface, the reflecting surface being applied under the 
 paint which covers the outer surface. It has a tremendous con- 
 centrating power, giving over 108-candle power at the tip of the 
 16-candle power lamp. With a 32-candle power lamp, the 
 maximum candle power covers a much larger angle than with 
 the 16-candle power. Streaks of light which are likely to be 
 present with mirror reflectors are in this case partially wiped 
 out b}^ the corrugations, so that, except when the work is very 
 near the lamp, this reflector can be used without a frosted bulb 
 lamp, although the latter is recommended. This reflector is an 
 excellent one to use wherever good, powerful, concentrated light 
 is desired, and where an opaque reflector is not objectionable. 
 It is well suited to show-window lighting, office desk lighting 
 
INDIVIDUAL INCANDESCENT ELECTRIC LAMPS. 49 
 
 and billiard-table lighting wherever other lamps are present for 
 the general illumination of a room. It is considerably more 
 expensive, heavier and more likely to be injured than the 
 
 Candle power. 
 Fig. .53. — Light about a corrugated mirror reflector. 
 
 frosted aluminum cone, Fig. 52, but it gives a more powerful 
 light and is more easily kept clean on account of its smooth 
 inner surface. 
 
 Fig. 54. — Corrugated mirror reflector. 
 
 Fig. 56 shows the light distribution about a seven-inch desk 
 shade, shown in Fig. 57. This is a shade which has a silvered 
 mirror reflecting surface, but instead of being open, as are most 
 
50 
 
 PRACTICAL ILLUMINATION. 
 
 reflectors, it has a ground-glass bottom through which the Hght 
 must pass and which eUminates the streaks. It is, of course, 
 intended to give a concentrating Hght for desk use and draught- 
 ing. The maximum candle-power below the lamp is about 48. 
 Its performance on the whole is very similar to that of the 
 frosted aluminum cone, but this type of reflector has the ad- 
 vantage that no matter at what angle it may be placed, the 
 bare lamp is not visible, as all the light comes through ground 
 glass. For desk work, where it is almost always desirable to 
 place the reflector at an angle, this is a good feature as the light 
 
 Candle power. 
 Fig. .1.1. — Lisjht about corrugated mirror reflector with 32 c. p. lamp. 
 
 practically comes from the entire surface of the ground glass and 
 the character of the light from it makes it more comfortable 
 to work by than when an open reflector is used. This shade 
 should never be used without a mica or asbestos gasket in the 
 shade holder to prevent dirt collecting inside of the shade. 
 
 Another shade similar to the above has its reflecting surface 
 opal with green backing like the common desk shade. It is, in 
 fact, the common desk shade with a ground glass bottom. 
 It is not subject to deterioration of the reflecting surface like 
 
INDIVIDUAL LYCAXDESCEXr ELECTRIC LAMPS. 
 
 51 
 
 those which depend on mirrors for their reflecting value, but 
 it does not give as strong a hght as the others. 
 
 Fig. 58 shows the distribution of light in two planes about 
 
 Candle power. 
 Fig. 56. — Light about McCreary desk shade. 
 
 a 16-candle power incandescent lamp in a parabolic aluminum 
 reflector with frosted aluminum finish inside. This reflector 
 is shown in Fig. 59. The candle power in the vertical plane 
 
 Fig. 57. — McCreary desk shade. 
 
 of the lamp axis is shown by the dotted lines, as in the^other 
 diagrams, while the candle power in a vertical plane at right 
 angles to the lamp axis, technically known, as normal to the 
 
52 PRACTICAL ILLUMINATION. 
 
 lamp axis, is indicated by a heavy line on the same diagram. 
 This is a reflector which is coming into common use for desk and 
 office work as well as for shops and all places where a light, easily 
 adjusted, opaque reflector is needed. This test shows the re- 
 
 Candle power. 
 Fig. 58. — Light about parabolic aluminum reflector. 
 
 flector to be an excellent one for this purpose. It gives over 
 50-candle power at the angles where the light is most wanted 
 in desk lighting, and gives over 40 candle power through an 
 angle of about 70 degrees. The reflector is Hght and is usually 
 
 Fig. 59. — Parabolic aluminum reflector. 
 
 made so that it can be quickly turned at. any desired angle. 
 The easy way in which it can be adjusted, as well as the excellent 
 distribution of light from this reflector, are strong points in its 
 favor. Reflectors of this kind can be obtained with various 
 outside finishes so that thev will harmonize with almost anv 
 
INDIVIDUAL INCAXDESCENT ELECTRIC LAMPS. 
 
 53 
 
 surroundings. Not being patented articles, they are low in 
 price. For most cases, the streaked light from the poUshed 
 aluminum will be objectionable. Generally, the frosted alumi- 
 num finish is the one to be chosen and it Ts preferable to use a 
 frosted lamp in it. 
 
 Candle power. 
 Fig. 60. — Light about cylindrical desk reflector. 
 
 Fig. 60 shows the distribution of light in two planes about 
 a, 16-candle power incandescent lamp placed in a cylindrical 
 desk reflector with aluminum paint finish inside, This reflector 
 is shown in Fig. 61. The solid lines show the light given in the 
 
 Fig. 61. — Cylindrical desk reflector. 
 
 plane of the lamp axis, while the dotted line shows the light 
 given in the plane at right angles or normal to the lamp axis. 
 Cylindrical desk shades of this kind are in common use, arid 
 may be compared with the parabolic shade shown in Figs. 
 
54 
 
 PRACTICAL ILLUMIXATION. 
 
 58 and 59. The light obtained from the cylindrical desk shade 
 is considerably less than from the parabolic reflector. Not 
 only is the maximum candle power approximately half that 
 of the parabolic shade, but it does not extend over as great an 
 
 Candle power. 
 Fig. 62. — Light about a window trough reflector. 
 
 angle. The cylindrical form of shade with aluminum paint 
 finish is, however, much to be preferred to the same shade with 
 white paint finish, as the latter causes streaks in the light which 
 are extremel)^ objectionable and necessitate the use of frosted 
 lamps. 
 
 Fig. 63. — Window trough reflector. 
 
 Fig. 62 shows the distribution of light about a trough reflector, 
 sho'wn in Fig. 63, with two lamps of 16-candle power. The 
 reflecting surfaces are glass mirrors. Such trough reflectors 
 are made for store window lighting purposes, being placed at 
 
INDIVIDUAL INCANDESCENT ELECTRIC LAMPS. 
 
 55 
 
 the top of the window out of sight from the street. Fig. 62 
 shows the hght on a vertical plane normal or at right angles to 
 the reflector. This can be compared with the light thrown from 
 the reflector shown in Figs. 55 and 72. Both of these types of 
 reflectors are much used for window lighting. The conical 
 
 Fig. 64. — Light about an 8-in. flat metal reflector. 
 
 reflector has much the best of it in this test for the majority of 
 windows. 
 
 For the average small window the reflector shown in Figs. 
 35 and 36 is to be preferred to Fig. 63, because of its higher 
 candle power at desirable angles and greater durability. The 
 
 Fig. 65. — Flat enameled metal reflector. 
 
 reflector shown in Fig. 72 gives an even greater amount of light 
 at desirable angles, although more subject to deterioration. 
 
 Fig. 64 gives the light distribution about a 16-candle power 
 incandescent lamp with an eight-inch flat metal reflector painted 
 white inside, shown in Fig. 65. This is similar to the light 
 obtained from the flat opal reflector in Fig. 27, and the sugges- 
 tions made with reference to that apply equally here. 
 
'56 
 
 PRACTICAL ILLUMINATION. 
 
 Fig. 66 shows the distribution of light about a six-inch orna- 
 mental mirror reflector illustrated in Fig. 67. This gives a 
 candle power of from 20 to 30 at the most useful angles below 
 the horizontal. The principal application of this reflector is 
 in connection with clusters and on ceilings where the fact that the 
 reflector is opaque is not a drawback. It is sometimes used by 
 mistake where the more concentrating type of mirror reflector, 
 shown in Fig. 54, should be used. The use of this or any other 
 reflector on a chandelier or cluster at an angle and hung low 
 enough to be in the ordinary line of vision is objectionable on 
 
 Candle power. 
 Fig. 66. — Light about an orna- 
 : mental mirror reflector. 
 
 .-t.-'\ 
 
 Candle power. 
 
 Fig. 68. — Light about a flat corru- 
 gated mirror reflector. 
 
 Fig. 67. — Ornamental mirror 
 reflector. 
 
 Fig. 69. — Flat corrugated mirror 
 reflector. 
 
 account of the glare receiA^ed in the eyes. The remedy is the use 
 of frosted bulbs or the placing of the cluster high. 
 
 Fig. 68 shows the distribution of light about a 16-candle power 
 incandescent lamp in a ten-inch flat fluted mirror reflector, 
 shown in Fig. 69. Although a reflector of this form would 
 naturally be expected to give a rather wide distribution of light 
 below the horizontal, it will be seen that the candle power 
 directly below the larap is considerably higher than with other 
 flat reflectors shown in this series of tests,* while the candle 
 power a little below the horizontal is nearh^ as great as with 
 
INDIVIDUAL INCANDESCENT ELECTRIC LAMPS. 57 
 
 some of the other flat reflectors. The proper place to use a 
 reflector of this kind is where it is desirable to illuminate a 
 considerable area and at the same time to obtain much Hght 
 immediately under the reflector. Like all opaque reflectors 
 
 Fig. 70. — Light about a 10-in. mirrored cone reflector with 16 c.p. lamp. 
 
 Fig. 71. — Ten inch mirrored cone reflector. 
 
 it cannot be used where darkening of the ceiling is objectionable. 
 Fig. 70 shows the distribution of light about a 16-candle 
 power lamp in a metal cone reflector with plate mirrors for 
 reflecting surface. The reflector is shown in Fig. 71. 
 
58 
 
 PRACTICAL ILL UMINA TIOX . 
 
 Fig. 72 shows the Hght about the same reflector with a 32- 
 candle power lamp. These are the most powerful of the con- 
 centrating reflectors when new. 
 
 Frosted Reflectors. 
 Reflectors of frosted, roughed or sand-blasted glass are not com- 
 mon, though some of the effects of a reflector are sometimes 
 obtained with common bell-shaped shades as explained 'later. 
 
 Fig. 72. — Light about a 10-in. mirrored cone reflector with 32 c.p. lamp. 
 
 Sand-blasted domes are used occasionally, and a test on one 
 with a 16-candle power lamp, sand-blasted inside is shown in 
 Fig. 73. This is of the same general shape as the seven-inch 
 opal dome shown in Fig. 22, except that it is 3.25 inches instead 
 of 3.75 inches in depth. It does not alter the distribution of 
 light from the lamp as rauch as the opal dome, although some- 
 what more pleasing in appearance because of the diffused light 
 which shows through it. 
 
INDIVIDUAL INCANDESCENT ELECTRIC LAMPS. 59 
 
 Enclosing Globes. 
 Fig. 74 gives the light distribution about a 16-candle power 
 incandescent lamp placed in a glass ball six inches in diameter, 
 sand blasted on the inside. This ball is shown in Fig. 75 
 Sand-blasted globes of this kind are very common at the present, 
 time. It will be seen that the distribution of light is not altered 
 materially from that of the bare incandescent lamp. Such a 
 sand-blasted globe is of considerable value in diffusing the 
 light, as mentioned in a paragraph on the frosted bulb incan- 
 descent lamp. In practice, when such a bulb is put in a pen- 
 dant position on a chandelier or fixture, the open space at its 
 neck should be closed with an asbestos gasket, to prevent dirt 
 collecting in the globe. If used in an upright position the 
 gasket can be omitted. 
 
 ^.>^% 
 
 '•»/ \ ^ 
 
 
 12.0/ - / \ -\l3.0 
 
 / 
 
 .J — ^^^<. % 
 
 Fig. 73. — Light about a sand blasted dome. 
 
 Fig. 76 shows the distribution of light about a 1 6-candle power 
 lamp enclosed in a pressed glass stalactite globe shown in Fig. 77, 
 This globe absorbs considerable light, and does not alter the 
 direction of the rays materially. The absorption or loss of 
 light in passing through the thick glass of the stalactite amounts 
 approximately to 25 per cent. This is another example of a 
 globe which is purely ornamental and of very little use as 
 diffuser of light. The glare of the filament received through a 
 globe of this kind is much more than would be received through 
 a frosted globe or bulb or a Holophane globe either one of which 
 absorbs less light. 
 
 The globes known as Holophane are designed with two objects 
 in view. One of these objects is to redirect the rays of light, 
 
60 
 
 PRACTICAL ILLUMINA TICK. 
 
 thus accomplishing some of the things accompHshed by a re- 
 flector in taking Hght which would otherwise be useless and 
 directing it in useful directions. This redirecting of the light 
 rays as they pass through the globe is accomplished by re- 
 flection and refraction of the rays by means of prisms pressed 
 on the outer surface of the globe. They are made in three 
 
 Candle power. 
 Fig. 74. — Light about a sand blasted ball. 
 
 classes. Class A globes concentrate light immediately under 
 the lamp. Class B globes give a general distribution below the 
 horizontal. Class C globes give the most light at a little below 
 the horizontal. The other object accompUshed by this type 
 of globe is the diffusion of light to avoid the glare of the arc, 
 incandescent filament, Nernst glower or gas flame, as the case 
 
 Fig. 75. — Six inch sand blasted ball. 
 
 may be. That is, it accomplishes the same thing as the frosted 
 bulb or globe onh- to a much greater degree. The rays of 
 light are broken tip by prisms on the inside of the globe so that 
 the eye does not meet the direct light from the filament of the 
 lamp. The rays are so successfully broken by passing through 
 these inner prisms that the whole globe appears to glow with a 
 
INDIVIDUAL INCANDESCENT ELECTRIC LAMPS. 
 
 61 
 
 mild light. The light, therefore, instead of coming from a small, 
 intensely lighted filament, comes from the whole surface of the 
 globe. This reduces materially the blinding effect which has been 
 spoken of frequently before, and enables a person to see better 
 because the pupil or iris of the eye is not contracted to such an 
 
 Candle power. 
 Fig. 76. — Light about a. pressed glass stalactite. 
 
 extent in an involuntary effort to shut out the intense blinding- 
 light. The absorption or loss of light in passing through Holo- 
 phane globes is remarkably low considering what they accom- 
 plish. Probably the most extensive tests ever made on this 
 glassware were conducted by Professor W L. Smith of the 
 
 Fig. 
 
 -Pressed glass stalactite. 
 
 Massachusetts Institute of Technology, who found that the loss 
 of Hght by absorption in this type of globe amounted to from 
 10.0 to 16.8 per cent, with an average of 12.3 per cent. On 
 account of the deep corrugations which form the prisms .on 
 these globes, unnoticed dirt is hkely to collect on them and 
 
62 
 
 PRACTICAL ILLVMIXA TIOX. 
 
 they should be brushed ofE at frequent intervals with a dry, 
 stifE brush, brushing lengthwise of the prisms. At longer inter- 
 vals they should be taken down and scrubbed with a brush in 
 soap and water. Like all other enclosed globes, the opening 
 
 Fig. 78. — Light about a Holophane Class A Stalactite. 
 
 around the shade holder should be filled with a mica or asbestos 
 gasket to keep out dirt. The amount of light likely to be ab- 
 sorbed by the collection of dust on the prisms is not as great 
 
 Fig. 79. — Holophane Class A Stalactite. 
 
 as might be anticipated, amounting in one test made at the 
 Electrical Testing Laboratories with a Holophane globe thickly 
 coated with dust, to about 13 per cent. 
 
 Fig. 78 shows the distribution of light about a 16-can die power 
 
INDIVIDUAL IXCAXDESCENT ELECTRIC LAMPS. 63 
 
 incandescent lamp enclosed in a Holophane globe shown in Fig. 
 79. As seen from the test, this globe is one designed to direct 
 the rays of light so as. to give a maximum candle power imme- 
 diately under the lamp. For a distance of 10 degrees each side 
 of the vertical, the candle power is over 20. This is, of course, 
 much less than is obtained with most types of reflectors, and 
 for this reason a globe of this pattern is not to be recommended 
 where a large amount of downward light is wanted unless a 
 diffusing globe is necessary. ,'There are many locations, how- 
 ever, where it is necessary to use a diffusing globe in order to 
 avoid the glare- of the bare filament, and at the same time to 
 direct the light in downward directions. The principal appli- 
 cation of this globe is on chandeliers for the general lighting of 
 rooms where the sockets on the chandeliers are placed at an 
 angle of 45 degrees as is very common. On such a chandelier 
 this globe will throw the- light down into the various parts of 
 the room instead of letting it escape toward the ceiling, and 
 it will, at the same time, do away entirely with the glare from 
 bare lamp filaments. Wei;e reflectors of any kind used for this 
 same purpose, frosted bulb lamps would be necessary and the 
 diffusion or softening of the light could not be as well accom- 
 plished as with the Holophane globe. This type of globe, 
 which is designed to give a concentrated downward light, 
 should not be used on a chandelier having the sockets pointed 
 straight down if the general, lighting of the room is the thing 
 desired; but if it is desired, to create a light space for reading 
 immediately under the chandelier and at the same time obtain 
 a fairly good light in other parts of the room, use of these globes 
 on downward pointed sockets will give good results. It is also 
 good for studded lights on ceilings. 
 
 Fig. 80 gives the distribution of light about a 16-candle power 
 incandescent lamp enclosed in a Holophane globe the general 
 shape of which is that shown in Fig. 81, the only difference 
 between this and Fig. 79 being in the angles of the prisms and 
 the size. This is the smaller of the two globes. This type is 
 designed to distribute light in all directions below the hori- 
 zontal, giving a good general illumination of a room when 
 placed in a pendant position. 
 
 Fig. 82 shows the distribution of light about a 16-candle power 
 incandescent lamp enclosed in a spherical pendant Holophane 
 
C4 
 
 PRACTICAL ILLUMIXATIOX. 
 
 globe, 5.5 inches in diameter, shown in Fig. 83. In this case 
 the Hght below the horizontal is not greatly increased as com- 
 pared to the bare lamp except immediately below the lamp. 
 Aside from this increase in candle power immediately below 
 the lamp, the principal effect of this particular globe is to 
 
 
 Fig. 80.- 
 
 Candle power 
 -Light about a Holophane Class B Stalactite. 
 
 diffuse and soften the light. It can be compared to the dis- 
 tribution of light around the ground glass ball which is also 
 a good diffuser and softener of light. 
 
 Fig. 84 gives the distribution of light about a 16-candle power 
 incandescent lamp enclosed in a Holophane spherical globe. 
 
 Fig. 81. — Holophane Class B Stalactite. 
 
 placed upright, shown in Fig. 85. This should be compared 
 with the illumination obtained from a bare incandescent lamp 
 placed upright, also from an incandescent lamp enclosed in the 
 various diffusing globes before shown. The distribution of 
 light from lamps and globes placed upright can, of course, be 
 
INDIVIDUAL INCANDESCENT ELECTRIC LAMPS. 
 
 65 
 
 easily obtained by inverting any of the diagrams before given. 
 This globe offers a valuable means of increasing the illumination 
 in downward directions from all kinds of fixtures where the 
 incandescent lamps are placed upright. The majority of bell- 
 
 * -o 
 
 CP ?■ °. =o 
 
 Candle power 
 Fig. 82. — Light about a Holophane Class B Ball. 
 
 shaped shades when placed on such upright fixtures have the 
 effect of robbing the lower part of the room of hght and throwing 
 even more light toward the ceiling than would be there if the 
 lamp were bare. This globe distributes the light below the 
 
 Fig. 83. — Holophane Class B Ball. 
 
 horizontal in such a way as to give a fairly even illumination 
 all over the lower part of a small room. 
 
 The"prisms in Holophane globes have to be calculated and the 
 molds cut accurately and a good grade of glass employed. Other- 
 wise the absorption of light by the glass is very high and no useful 
 
66 
 
 PRACTICAL ILLUMINATION. 
 
 change made in the distribution of the light. This is well shown 
 by Fig. 86, which is from a test made with a 16-candle power 
 lamp in the stalactite globe shown in Fig. 87. The globe re- 
 sembles closely some with calculated prisms, but absorbs a much 
 greater amount of Hght. It also has little or no value in re- 
 
 ■tpps CO U5 aO-H 
 
 Candle povver 
 Fig. 84. — Light aboiit a Holophane Upright Class B Ball. 
 
 directing the light in useful directions and does not give the 
 same diflfusion as a globe with accurate^ cut prisms. 
 
 Fig. 88 shows the distribution of light around a prismatic 
 reflecting glass ball, the top of which is formed into a prismatic 
 reflector, as shown in Fig. 89. The bottom is plain frosted glass. 
 This ball combines in one the characteristics of a diffusing 
 
 Fig. S.'). — Holophane Upright Class B Ball. 
 
 globe and reflector. The distribution of hght from such balls 
 can be varied by varjdng the position of the lamp inside of the 
 ball. The lower the lamp is placed the wider the distribution. 
 It will be seen from the foregoing that not every globe that 
 has prisms on it is to be considered as a true prismatic globe. 
 
INDIVIDUAL INCANDESCEXT ELECTRIC LAMPS. 67 
 
 Such globes should not be selected on appearance but rather 
 from photometric tests which show just what light distributions 
 they give. In the absence of such tests or their equivalent one 
 is working absolutely in the dark. 
 
 Cnntlle Power. 
 
 Fig. 86. — Light about an uncalcu- 
 
 lated prismatic stalactite. 
 
 11, 
 I'i.s 
 
 o 
 
 
 ^f 
 
 
 u 
 
 
 Fig. 88. — Light about a prismatic 
 reflecting ball. 
 
 & 
 
 % 
 
 Fig. 87. — Uncalculated prismatic 
 stalactite. 
 
 Fig. 89. — Prismatic reflecting ball 
 
 Open Shades. 
 Fig. 90 gives the light distribution of a 16-candle power 
 incandescent lamp in an ornamental glass bell-shaped shade 
 or reflector, the form of which is shown in Fig. 91 . This reflector 
 is of a very common shape. The outside is lightly frosted and 
 the ornamentation is cut into it. It is seen that this glass bell 
 causes some increase in the candle power in downward directions 
 and a decrease in horizontal directions. While it makes some 
 
68 
 
 PRACTICAL ILLUMINATION. 
 
 improvement over the distribution of light from the bare lamp 
 it does not make any such marked change as do the reflectors 
 shown in Figs. 22 and 24. A reflector of this bell type is usually 
 chosen simply as a matter of ornament, without any attempt 
 to redirect the rays of Ught. This particular reflector probably 
 changes the natural distribution of light from the lamp more 
 
 12.0 
 
 Candle power. 
 Fig. 90. — Light about an ornamental frosted shade. 
 
 than would many others of the same general shape, because 
 of the frosting on its outside, which tends to make the bell 
 mouth a reflector. Frequently a glass bell such as this is used 
 on a bracket or chandelier in an upright position. In such a 
 case it is evident that instead of adding to the useful downward 
 
 Fig. 91. — Ornamental frosted shade. 
 
 light, it detracts considerably from it and throws much light 
 towards the ceiling. 
 
 Fig. 92 shows the light about the shade shown in Fig. 93 
 with a 16-candle power incandescent lamp. This shade was 
 roughed or sand-blasted on the inside. Its general character- 
 istics as to light distribution are very similar to those of Fig. 90. 
 
INDIVIDUAL INCAXDESCENT ELECTRIC LAMPS. 
 
 69 
 
 It would be useless to report here on many tests with this gen- 
 eral type of bell shade, as they are made in such immense variety, 
 and designs change from year to year. They must always be 
 considered mainly as ornamental, as they usually absorb and 
 waste considerable Hght, and are as a rule of little value in 
 
 VlS.O 
 
 Fig. 92. — Light about a glass shade roughed inside. 
 
 redirecting the light rays. For locations where expense is no 
 consideration globes of this kind can be freely used. It in 
 nevertheless, true that by certain changes in shape and desigo- 
 much of the glassware now manufactured purely for decoratiAm. 
 purposes might be made useful as well, and on the other hancL 
 some of the staple glassware which is now to be considered as 
 
 Fig. 93. — Glass shade roughed inside. 
 
 mainly useful rather than ornamental could be made ornamental 
 as well. The manufacturers of unpatented types of glassware 
 might well give attention to these matters. 
 
 Fig. 94 shows the distribution of light about a 16-candle 
 power incandescent lamp placed in a Holophane open shade, 
 designed for use as a pendant, and shown in Fig. 95. As seen 
 
70 
 
 PRACTICAL ILLUMINATION. 
 
 from the test, this globe is of the type which increases the illu- 
 mination in downward directions. When such a shade is used 
 on a chandelier, it should be placed on sockets which are pointing 
 straight down or at an angle of not greater than 30 degrees from 
 
 "^ Candle power 
 
 Fig. 04. — Light abovit a Holophane Class B Pendant Shade. 
 
 the vertical, as otherwise it will not only throw more light up 
 than is necessary but will expose the glaring lamp filament. 
 Where used at an angle frosted tipped lamps should be employed. 
 Fig. 96 shows the distribution of light about a 16-candle 
 power upright incandescent lamp in a Holophane open shade 
 
 Fig. 95. — Holophane Class B Pendant Shade. 
 
 shown in Fig. 97. This shade is mainly useful as a diffuser, 
 although it increases the useful illumination immediately below 
 the lamp. 
 
 Fig. 98 shows the distribution of light about a 16-candle 
 power incandescent lamp upright in a Holophane open shade, 
 
INDIVIDUAL ISCAXDESCEXT ELECTRIC LAMPS. 
 
 71 
 
 shown in Fig. 99. This shade is designed to make a marked 
 increase in the illumination immediately below the lamp, in 
 which it succeeds very well, as within a narrow angle below the 
 
 
 Candle power 
 Fig. 96. — Light about a Holophane Class B Upright shade. 
 
 lamp it gives about 20 candle power. This shade can be used 
 to advantage where it is desired to secure good hght for reading 
 or writing under chandeUers or brackets equipped with upright 
 incandescent lamps. 
 
 Fig. 97. — Holophane Class B Upright shade. 
 
 Fig. 100 shows the distribution of Ught around a 16-candle 
 power incandescent lamp, placed in a combination Holophane 
 diffusing and reflecting shade shown in Fig. 101. It is designed 
 
72 PRACTICAL ILLUMINATION. 
 
 Specially for use on wall bracket fixtures. This globe consists 
 of a prismatic reflector on one side and a Holophane globe on 
 the other, the two features being combined in the one globe. 
 In the use of such globes it is manifestly advisable to take 
 
 Candle power 
 Fig. 08 — Light about a Holophane Class A Upright Shade. 
 
 Fig. 99. — Holophane Class A Upright Shade. 
 
 unusual precautions to prevent their being turned around so as 
 to throw the maximum light toward the wall instead of away 
 from it. This globe makes a decided increase in the amount of 
 illumination thrown away from the wall, if properly used. 
 
INDIVIDUAL INCANDESCENT ELECTRIC LAMPS. 73 
 
 Life of Frosted Lamps. 
 During the last few years there has rightly been a tendency 
 in incandescent lighting to lower the intensity of the source of 
 light, either by using enclosed diffusing globes or else by frosting 
 the lamp bulb itself. The diffusion obtained by these methods 
 has not only resulted in more artistic effects, but has also made 
 the light much easier on the eyes. In this connection the ques- 
 
 18.6 
 
 Candle power 
 Fig. 100. — Light about a Holophane diffusing and reflecting shade. 
 
 Fig. 101. — Holophane diffusing and reflecting shade. 
 
 tion has arisen as to the relative efficiency of the two available 
 methods, which are : 
 
 1. Frosting the lamps themselves and using them without 
 enclosing globes. 
 
 2. The use of the bare lamp with a diffusing enclosing globe, 
 either ground glass, opal, Holophane or other types. 
 
 It has generally been conceded that a frosted lamp, as far 
 
74 PRACTICAL ILLUMINATION. 
 
 as efficiency is concerned, is more economical than a plain lamp 
 with a diffusing globe. Generally speaking, the absorption by 
 such frosting is in the neighborhood of 10 per cent. , while with the 
 use of ground glass, and other types of diffusing globes, the ab- 
 sorption runs from as high as 75 per cent, down to about 12 per 
 cent. 
 
 The authors have advocated the use of some sort of cover 
 or gasket for the neck of an enclosing globe, such a cover being 
 formed of asbestos, mica or some similar material which will 
 keep out all dust and, in the case of white asbestos, increase the 
 light below the horizontal. In order to determine the effect on 
 the hfe of the lamp, with these various conditions, tests were 
 made at the Electrical Testing Laboratories on the following: 
 
 1. Frosted lamp, bare. 
 
 2. Clear lamp and diffusing globe. 
 
 3. Clear lamp, diffusing globe and asbestos cover. 
 
 Ten lamps were taken for each test. The enclosing globes 
 used were of the Holophane type, being spheres, 6.5 inches in 
 diameter, with 3.25-inch holders. Ten of these were used with- 
 out any cover, while ten others were protected with an asbestos 
 disk or washer which practically prevented any dust from enter- 
 ing. 
 
 The loss of light due to the frosting of the lamps averaged 
 nine per cent., while the loss due to the absorption of the globes 
 was about 16.5 per cent. 
 
 When, however, the question of life is considered, it is found 
 that the life with a frosted lamp is only 216 hours, whereas the 
 globe under test gives 428 hours, or nearly twice the life. If 
 the average mean spherical candle power is multiplied by the 
 number of hours of useful life, it is found, in the first case, that 
 the life is 2,383 candle hours with frosted lamps, while in the 
 second place it is 4,401 candle hours, an increase of 85 per cent. 
 Care should be taken not to confuse this with horizontal candle 
 hours commonly used in considering the life of incandescent 
 lamps. 
 
 Coming now to the question of asbestos disks in the neck 
 of enclosing globes, we find that the total flux of light is practi- 
 cally the same in either case, but that the light in the lower 
 hemisphere is increased about two per cent., and that the loss 
 of life, due to the disks is about one per cent. Inasmuch as 
 
INDIVIDUAL INCANDESCENT ELECTRIC LAMPS. 75 
 
 there is a great gain due to the fact that dust is prevented from 
 faUing on the inside of the globe, and also on the lamp itself, it 
 is beyond question a desirable thing to equip all pendant enclos- 
 ing globes with some form of cover to prevent the collection 
 of dust or dirt. Dirt collecting in the bottom of a diffusing 
 globe is not only a cause of loss of light, but is inartistic and 
 slovenly. It shows much more plainly in a diffusing globe than 
 in any other and should be avoided both on account of the 
 efficiency and appearance. 
 
 The object of frosting lamps is to cut down the intense glare 
 of the bare filament with the least possible loss of light, render- 
 ing the lamp more agreeable to the eye and giving a more 
 artistic effect. In many instances, the same results can be 
 obtained by the use of ground glass diffusing globes with a cor- 
 responding increase in the life of the lamp. Where a reflector 
 is needed and at the same time diffusion of the light from the 
 lamp filament is advisable, a frosted bulb lamp is still the most 
 easily available device for securing diffusion. 
 
 The usual guaranteed life of bare lamps of this kind with 
 clear glass bulbs is 450 hours. The loss of Hfe by enclosing 
 them in globes is, therefore, about 5.5 per cent., while the loss 
 of life by frosting is about 52 per cent. 
 
 The authors have the results of a large number of tests on 
 the loss of light by absorption by different globes. These tests 
 show such a lack of consistency that ihej have not been given 
 here since the distribution of light with a shade or reflector is 
 information of much mere practical value than the total absorp- 
 tion. When such figures are given they are in general only 
 approximate. 
 
CHAPTER VI. 
 
 INDIVIDUAL GAS BURNERS AND THEIR REFLECTORS, SHADES 
 
 AND GLOBES. 
 
 Before considering the question of light distribution obtained 
 from different forms of gas Hghting appliances it will be well 
 to review briefly some of the difficulties which attend the subject- 
 In making any of the tests one of the primary considerations 
 is to know the quahty of the gas used. This is of most import- 
 ance when open flame burners are emploj^ed, but it is also of 
 importance with mantle burners. Naturally it was impossible 
 to show tests on different qualities of gas, so the uniform prac- 
 tice was observed of having all the tests given herein made on 
 New York Cit}'' gas. 
 
 Another variable is the gas pressure. As will be seen from 
 the curve in Fig. 102 the candle power of a mantle burner in- 
 creases quite rapidly with an increase in pressure. It should 
 be noted in this test, and also in the two following, shown in 
 Figs. 103 and 104, that the burner was in every case adjusted 
 for maximum incandescence. Fig. 103 shows that the candle 
 power increases with the gas consumed nearly at the same 
 rate as the pressure increases. Fig. 104 shows that the con- 
 sumption and the pressure increase practically at the same rate 
 and that, therefore (at least in this series of tests, which were 
 very carefully made), when the mantle is adjusted for maxi- 
 mum incandescence the efficiency of the burner, that is to say, 
 the candle power per cubic foot of gas, is very nearly inde- 
 pendent of the pressure. 
 
 Among other variables may be mentioned different kinds of 
 burners, the effect of dift'erent kinds and lengths of chimneys, 
 quality and size of mantles, adjustment of mantles to have 
 them incandesce equally on all sides, effect of side rod or center 
 supports, etc. It is therefore manifest that in order to arrive 
 at absolute A'alues all of the above factors should be known. 
 
 76 
 
INDIVIDUAL GAS BURNERS. 
 
 77 
 
 In the following tests on mantle burners the pressure and con- 
 sumption are stated and in the appendix trade data on all tests 
 are given. 
 
 What the authors have aimed at is to get as near as possible 
 to conditions as they actually exist in commercial work. Thus 
 although the pressures given are possibly below the average 
 throughout the country, they give the actual conditions as they 
 exist in New York City to-day. On the other hand, the tests 
 probably show up better than would exist ordinarily, inasmuch 
 
 115 
 
 110 
 
 105 
 
 felQO 
 
 9S 
 
 90 
 
 75 
 
 TO 
 
 
 1.5 
 
 2.0 2.5 
 
 Inches Water Pressure 
 
 3.0 
 
 3.5 
 
 Fig. 102. — Variation of candle power with pressure. 
 
 as they were all made in a laborator^r with ideal conditions as 
 regards constant pressure, consumption, adjustment, cleanliness, 
 new burners, mantles, glassware, etc. All told, then, the tests are 
 probably fairly representative of what may be expected when 
 gas apphances are properly cared for, and are in excess of what 
 often happens in commercial practice. 
 
 With regard to open flame burners, Prof. C. P Matthews* 
 calls attention to the fact that, as the result of a number of tests 
 on tips supposed to range from one to four cubic feet of gas per 
 
 *Proceedings National Electrical Light Association, May, 190-3. 
 
78 
 
 PRACTICAL ILLUMINATIOX. 
 
 lib 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 no 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 ^ 
 
 y 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 y 
 
 
 
 105 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 / 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 / 
 
 / 
 
 
 
 
 
 © 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 / 
 
 < 
 
 
 
 
 
 
 gioo 
 
 Ph 
 
 
 
 
 
 
 
 
 
 
 
 
 
 •/ 
 
 / 
 
 
 
 
 
 
 
 3 95 
 
 c 
 
 
 
 
 
 
 
 
 
 
 
 
 •/ 
 
 / 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 / 
 
 / 
 
 
 
 
 
 
 
 
 
 o 
 
 g9o 
 
 |85 
 o 
 
 
 
 
 
 
 
 
 
 
 
 k 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 / 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 / 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 •/ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 tn 
 
 80 
 
 
 
 
 
 
 
 / 
 
 / 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 / 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 75 
 
 
 
 
 
 
 y 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 lyn 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 4.0 4.5 o.O 
 
 CuJ>ie Feet of Gas per Hour 
 
 5.5 
 
 Fig. 103. — Variation of candle pOAver with consumption. 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 20' 
 
 ^ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 -^ 
 
 ', 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 o 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 ^ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 ftlO 
 
 1~> 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 G 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 P^ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 -^3 
 
 o 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1.5 
 
 2.0 2.5 
 
 Inches Water Pressure 
 
 3.0 
 
 3.5 
 
 Fig. 104. — Variation of candle power per cu. ft. with pressure. 
 
INDIVIDUAL GAS BURNERS. 79 
 
 hour, in some cases a one-foot tip gave as high a candle power 
 as a four-foot tip with no greater gas consumption. So great 
 are the variations that apparently the nominal ratings of tips 
 are practically valueless. 
 
 The introductory remarks of Chapter V, with reference to 
 the necessity of studying the distribution of any light source 
 in conjunction with its reflecting globe or shade should be care- 
 fully noted here, especially by those who have not read the 
 previous chapter. 
 
 Open Gas Flames. 
 The first thing one notices in studying the distribution of 
 light about an open flame burner is the remarkable fact that 
 the flame gives very nearly the same amount of light whether 
 
 'P^ 
 
 \<LUi.\y'^^ 
 
 Fig. 105. — Open gas-horizontal distribution. 
 
 viewed either in the plane of the flame or at right angles to it. 
 This is probably due to the fact that the hot particles of carbon 
 do not greatly tend to screen the light from each other, although 
 there is a slight tendency that way. 
 
 In Fig. 105 is shown the horizontal distribution curve of a 
 flat flame burner consuming five cubic feet of gas per hour at 
 a pressure equal to that of a column of one inch of water. This 
 curve is made by measuring the candle power of the flame 
 when viewed from different angles but all taken in the same 
 horizontal plane. 
 
 Fig. 106 shows the distribution in a vertical plane about the 
 same open gas jet as shown in Fig. 105. The mean spherical 
 candle power is 21. This was the gas jet used in making all the 
 
80 
 
 PRACTICAL ILLUlMLXATIOX. 
 
 following tests on open gas flame shades. It will be noticed 
 that the distribution of light is remarkably uniform in all 
 directions with the exception of directly underneath where a 
 
 Fig. 106. — Open gas-vertical distribution. 
 
 small shadow is made by the pillar and the tip. As in the case 
 of the mantle burner shown later practically one-half of the 
 light is above and one-half below the horizontal. 
 
 ^WW:- 
 
 Fig. 107. — Light about an open 
 gas etched globe. 
 
 Fig. 108. — Etched globe for open 
 
 Fig. 107 shows the distribution of light about an etched glass 
 globe, shown in Fig. 108, representing perhaps the great majority 
 of open flame globes. The gas consumption was five cubic 
 feet per hour at a pressure equal to that of a one-inch column 
 
INDIVIDUAL GAS BURNERS. 
 
 81 
 
 of water. It should be noticed that a large amount of the 
 light is thrown upward, the globe acting somewhat as a re- 
 flector, there being a decrease of 30 per cent, below the hori- 
 zontal as compared with the bare flame. 
 
 The real efficiency of a globe should not be measured, generally 
 speaking, by the amount of light it gives forth, but rather by 
 the amount of light which it throws below the horizontal, or in 
 some particular direction where the Hght is wanted. Thus in 
 the case of the globe shown in Fig. 107, although its absorption 
 
 Fig. 109. — Light about an open gas Class A Holophane globe. 
 
 is comparatively small, the loss of light below the horizontal is 
 rather large, and as this is the light which is generally used, it 
 is seen that efforts to obtain an artistic effect and at the same 
 time shield the eyes, succeed at a considerable loss of effective 
 illumination. 
 
 It should also be noted that in the case of open flame burners, 
 the necessity for diffusion is not nearl}'' as great as with mantle 
 burners, as both the intensity and quantity of light are low. 
 
 Fig. 109 shows the light about a Class A Holophane globe, 
 shown in Fig. 110, designed to diffuse the light and at the 
 
82 PRACTICAL ILLUMINATION. 
 
 same time throw a strong light downward. Its absorption is 
 about the same as the other open gas globes but its distribution 
 curve is entirely different. 
 
 It is also possible to enclose the open flame gas jet in opal 
 dome reflectors similar to those used on electric incandescent 
 lamps. The relative distribution of Hght with such opal dome 
 
 Fig. 110. — Open gas Class A Holophane globe. 
 
 reflectors will be similar to that obtained with incandescent 
 electric lamps as given in a previous test. 
 
 Individual Mantle Burners. 
 Fig. Ill shows the distribution of light about a bare upright 
 mantle burner consuming 3 7 cubic feet of gas per hour at a 
 pressure equal to that due to a water column of 1.6 inches. The 
 
 Fig. 111. — Light about a bare vtpright gas mantle burner. 
 
 mean spherical candle power is 63. Fifty-five per cent, of the 
 light is above the horizontal and 45 per cent, below. It is 
 therefore evident that for economical lighting as well as for 
 artistic reasons it is inadvisable to use bare mantles in this 
 way. There is a noticeable lack of light directly under and 
 near the burner due to the shadow cast by the burner itself. 
 Fig. 112 shows the curve of the same burner under identical 
 
INDIVIDUAL GAS BURXERS. 
 
 S3 
 
 conditions to those in Fig. Ill but equipped with a flat fluted 
 porcelain reflector shown in Fig. 113. 
 
 While such a shade does not hide the bare mantle and there- 
 fore does not aid in the diffusion, it does give an increase of 
 light at all angles below the horizontal and as such, is to be 
 preferred to the bare mantle. 
 
 Fig. 112. — Light about a mantle burner with S-in. fluted opal re- 
 flector. 
 
 Fig. 114 shows the distribution of light about the common 
 air hole or opal Q globe shown in Fig. 115, the lamp burning 5.0 
 cubic feet per hour at a pressure equal to that due to a 1.5 inch 
 water column. Inasmuch as its intensity is less than the bare 
 mantle, it is to be preferred to the small flat fluted porcelain 
 
 Fig. 113. — Fluted opal 8-in. reflector. 
 
 shade although its efficiency is much less. In all compar sons, 
 both diffusion and distribution must be borne in mind, and inas- 
 much as it is seldom that both factors are present in one globe 
 one must select that which is most necessary for the case in 
 hand. 
 
84 
 
 PRACTICAL ILLUMINATION. 
 
 Fig. 116 shows the distribution of light about the ordinary 
 ground glass globe shown in Fig. 117, the lamp burning 3.1 cubic 
 feet per hour at a pressure equal to that due to a 1.7 inch water 
 column. This globe simply diffuses and softens the Ught with- 
 
 Fig. 114. — Light about a mantle burner with opal Q chimney. 
 
 out materially altering the distribution. The use of a ground 
 glass globe or some other diffusing glass around a mantle burner 
 is desirable in most cases because the undifEused light from the 
 gas mantle burner is so intense as to have a blinding effect on 
 
 Fiu. 115. — Opal Q chimnej'. 
 
 the eyes of all looking in its direction and so reducing the prac- 
 tical efficiency of the illumination. 
 
 Fig. 118 shows the light about a fancy globe, shown in Fig. 
 119, the lamp burning 3.1 cubic feet per hour at a pressure 
 equal to that due to a 1.7 inch water column. Its absorption 
 
INDIVIDUAL GAS BURNERS. 
 
 S5 
 
 ^ ^ <^ 
 
 ^ * 
 
 Fig. 116. — Light about a mantle btirner 
 with ground glass chimney. 
 
 Fig. 117. — Ground glass 
 chimney. 
 
 Fig. 118. — Light about fancy opal chimney with mantle burner. 
 
 Fig. 119. — Fancy opal chimney. 
 
86 
 
 PRACTICAL ILLUMINATION. 
 
 is ver)^ large. Such globes give very artistic results and good 
 diffusion, and where economy is no object and a sufficient 
 number can be used to get the required illumination they are 
 to be highly recommended. 
 
 
 \ /•^> 
 
 Fig. 120. — Light about a mantle burner in Class A Holophane 
 
 shadf . 
 
 Fig. 121. — Class A Holophane shade for mantle burner. 
 
 Fig. 120 shows the light about what is known as a Class A 
 Holophane globe, shown in Fig. 121, the lamp burning 3.55 
 cubic feet per hour, at a pressure equal to that due to a 1.25 
 inch water column. It is designed to throw the Hght directly 
 
INDIVIDUAL GAS BURNERS. 
 
 87 
 
 downward but of course at the expense of the Hght at other 
 angles. The diffusion obtained by such globes is almost equal 
 to that of opal without however the large absorption, inasmuch 
 
 V 
 
 \.-;> 
 
 
 ~±- as 
 
 / 
 
 '•^' 'b^ 'i-'^'~if' " ~"i."s'~'~>~'* ■' 
 
 Fig. 122.— Light about a Class B Holo- Fig. 123.— Class B Holophane 
 phane shade with mantle burner. shade for mantle burner. 
 
 Fig. 124. — Light about a Class C Holophane shade with mantle burner. 
 
 Fig. 125. — Holophane Class C shade. 
 
 as the'globes are made of perfectly clear glass and depend upon 
 their accurate design for their diffusing and redirecting quaU- 
 ties. Such globes owing to their prismatic faces are apt to 
 collect the dust and should therefore be kept clean. The dust, 
 
88 
 
 PRACTICAL ILLUMINATION. 
 
 however, affects their appearance more than their efficiency, 
 the additional absorption of a dusty globe in one test that was 
 made being about 13 per cent. They should never be used 
 in places where the air is greasy, as over stoves, as the grease is 
 apt to settle on the prisms and is extremely difficult to remove. 
 When such globes, however, are used intelligently they form 
 an important adjunct to the engineer's work, as they combine 
 both qualities of diffusion and redirecting powers with small 
 absorption. 
 
 Fig. 126. — Light about a mantle burner; opal dome and opal bobesche. 
 
 Fig. 122 shows the light about a Class B Holophane globe, 
 shown in Fig. 123, the lamp burning 3.6 cubic feet per hour at 
 1.5 inch pressure, designed to give a general rather than con- 
 centrated distribution. 
 
 Fig. 124 shows the light about a Class C Holophane globe, 
 shown in Fig. 125, designed to throw the light sideways just 
 below the horizontal in order to cover a wide area. 
 
 Fig. 126 shows the Hght about an opal dome and opal bobesche, 
 or cup-shaped eye protector, shown in Fig. 127, the lamp burn- 
 
IXDIVIDUAL GAS BURNERS. 
 
 89 
 
 ing 3.6 cubic feet per hour at a pressure equal to that due to 
 a 1.6 inch water column. Such a combination is good for read- 
 ing when the light is directlj^ above the work. A bobesche should 
 always be used with a dome shade whenever there is a possi- 
 bility of the eye being able to see the mantle. As a matter of 
 fact, bobesches should be made about half an inch deeper, as 
 at present most of them do not come high enough to cut out all 
 view of the bare mantle. 
 
 Fig. 127. — Opal dome and opal bobesche. 
 
 Fig. 128 shows the curve of a green dome shade and opal 
 bobesche, burning 4.1 cubic feet per hour at 1.6 inch pressure. 
 This makes a good reading light if directly above the work, but 
 concentrates it too much if used on a low portable with the 
 reader sitting alongside the table. A green dome with green 
 bobesche is practically out of the question, owing to the large 
 absorption of the cup. Such a combination depends entirely 
 on the reflection from the white lining of the green dome and 
 is very inefficient. 
 
90 
 
 PRACTICAL ILLUMIXATIOX. 
 
 Fig. 129 shows the curve of a green dome and Class A Holo- 
 phane bobesche, the lamp burning 3.8 cubic feet of gas per hour 
 at a pressure equal to that due to a 1.6 inch water column. 
 This bobesche, being higher, cuts off all view of the mantle and 
 gives a strong light at all angles. A comparison of this with 
 Figs. 126 and 128 is instructive and shows how easily illumina- 
 tion may be sometimes improved by slight changes. 
 
 Fig. 128. — Light about a mantle 
 burner; green dome and opal bob- 
 esche. 
 
 Fig. 129. — Light about a green 
 dome and Holophane bobesche; 
 mantle burner. 
 
 Inverted Burners. 
 
 The question of the efficiency of inverted burners compared 
 with other types is taken up in Chapter XIV. 
 
 Fig. 130 shows the distribution in a vertical plane about an 
 inverted burner equipped with clear air hole chimney and 
 without an outer globe. It was operated at a pressure equal 
 to that of a 1.5 inch water column and 3.0 cubic feet per hour 
 consumption. It will be noted that this gives a good Hght at 
 all angles below the horizontal, also that 67 per cent, of the 
 light is below and only 33 per cent above the horizontal plane. 
 
IX DIVIDUAL GAS BURNERS. 
 
 91 
 
 This should be compared with the distribution of hght about 
 an upright burner where 55 per cent, is above and only 45 per 
 cent, below the horizontal. The mean spherical candle power 
 is 54. 
 
 Jo ^ 
 
 O 
 00 
 
 Fig. 130. — Light about a bare inverted mantle burner. 
 
 Fig. 131. — Light about an inverted burner with ground glass ball. 
 
 Owing to the high intrinsic brilliancy of the inverted mantle, 
 which is considerably higher than the upright, such a burner 
 should never be used without a diffusing outer or a frosted 
 inner globe, especially as such burners generally hang lower than 
 the upright type. 
 
92 
 
 PRA CTICA L ILL UMLXA TIOX 
 
 Fig. 131 shows the same burner equipped with an ordinary 
 .sround glass or sand blasted ball, operating at a pressure equal 
 to that of a 1.5 inch water column and 3.0 cubic feet consump- 
 tion. This cuts down the intensity at all angles except directly 
 underneath where the Hght is increased owing to the small hole 
 
 V 
 
 Fig. 1.32. — Light about an inverted burner with opal ball. 
 
 in the bottom to provide a draft. The upper part of the ground 
 glass globe acts somewhat as a reflector thus increasing the 
 light through the hole at the bottom. 
 
 Fig. 132 shows another burner without inner chimney but 
 with a very dense round opal globe. The consumption of this 
 
 
 \ 'x 
 
 -X r- 
 
 ■- .1 
 / 
 
 Fig. 133. — Light about an inverted burner with 9-in. flat opal reflector. 
 
 burner is not given here inasmuch as the test was purely ex- 
 perimental. The curve is shown in order to give an idea of the 
 character of the distribution when using an opal globe. 
 
 Fig. 133 shows the distribution of light about the burner 
 used in Figs. 130 and 131 equipped with a 9-inch flat porcelain 
 
IXDIVIDUAL GAS BURNERS. 
 
 93 
 
 reflector and with clear inner air hole chimney. This burner 
 took 3.1 cubic feet of gas per hour at a pressure equal to that 
 of a 1.5 inch water column. This should be compared with 
 Fig. 130. It shows that the light is increased at all angles 
 under 15 degrees below the horizontal. However this com- 
 bination should not be used where it is apt to come in the line 
 of vision unless the inner chimney is frosted as the glare is 
 altogether too intense. On the other hand, if it is desired to 
 get a strong light down, the reflectors shown in Figs. 136 and 
 138 would be preferable. 
 
 Fig. 135. — Light about inverted 
 burner with 8-in. deep opal cone. 
 
 Fig. 134. — 9-in. flat opal 
 reflector. 
 
 Fig. 136. — S-in. opal cone 
 for inverted gas burner. 
 
 Fig. 135 shows the light about the same burner taking 2.8 
 cubic feet of gas at a pressure equal to that of a 1.5 inch water 
 column equipped with clear inner air hole chimney and a deep 
 eight-inch porcelain reflector shown in Fig. 136. As will be 
 noticed, this gives a strong increase in Ught through a wide angle 
 and at the same time is deep enough to shade the eyes from 
 the bare mantle at most ordinary angles. 
 
 Fig. 137 shows the light about a prismatic reflector shown 
 in Fig. 138, on the same burner with clear inner air-hole chimney 
 burning 3.1 cubic feet of gas per hour at a pressure equal to that 
 of a 1.5 inch water column. This reflector gives a very strong 
 
94 
 
 PRACTICAL ILLUMINA TION. 
 
 light, being greater at all angles than the deep cone shown in 
 Fig. 135. It gives an increase over the bare mantle at all angles 
 under 25 degrees below the horizontal and is deep enough to 
 
 11 11 
 
 IS 2 12 
 
 njl'^ p^^ IGrij 
 
 Fig. 137. — Light about inverted mantle burner with prismatic re- 
 flector. 
 
 Fig. 138. — Prismatic reflector for inverted gas. 
 
 hide the mantle. It is well adapted for show windows and 
 other places where a strong light is wanted downward. 
 
 It will be seen from the representative photometric curves given 
 here that generally speaking not only is the natural distribution 
 
INDIVIDUAL GAS BURNERS. 95 
 
 of the light from an inverted burner superior to the upright for 
 most cases, but also the light can be more easily altered in dis- 
 tribution. It also lends itself better to artistic treatment- than 
 the upright tj^pe, both as regards position and glassware. On 
 the other hand, it has its disadvantages such as tarnishing the 
 finish of fixtures, danger from falling particles unless a globe 
 with a bottom is provided, etc. At the present time it can be 
 stated that the inverted burner is not yet by any means per- 
 fected but that such rapid advances are being- made that 
 we can safely count on its being an important factor owing to 
 its appearance, efficiency and distribution. 
 
CHAPTER VII. 
 
 ACETYLENE GAS BURNERS AND THEIR SHADES. 
 
 The flame of an acetylene burner is in general very similar 
 to the ordinary open gas flame as regards its distribution of 
 light. UnUke open gas, however, acetylene is extremely intense, 
 giving from 75 to 100 candle power per square inch of flame, 
 whereas the open gas gives in the neighborhood of five-candle 
 power per square inch. This means that in order to use acety-' 
 
 Fig. 139. — Light in horizontal plane about an acetylene gas flame. 
 
 lene when in the line of vision with any degree of comfort it 
 becomes necessary to use diffusing shades to cut down the 
 intense glare, and that, therefore, outside of the question of 
 ornament the diffusion of different types of shades is of great 
 importance. As several different types of shades for open gas 
 were shown previously in this chapter and as the distribution 
 from them with acetylene is practically the same as with open 
 gas only a few kinds will be considered here. 
 
 96 
 
ACETYLENE GAS BURNERS AND THEIR SHADES. 97 
 
 Unfortunately, in these tests the gas consumption was not 
 measured, but was assumed to be the rated consumption of the 
 
 Fig. 140.— Light in vertical plane about open acetylene gas flame. 
 
 Fig. 141. — Light about acetylene gas flame in etched shade. 
 
 tip , namely, one half cubic foot of gas per hour. The tests must be 
 considered therefore as indicative of what may be expected with 
 different shades when used with acetylene, rather than of abso- 
 
98 
 
 PRACTICAL ILLUMINATIOX. 
 
 lute values. The tests were made with 2.5 inches of water 
 pressure. 
 
 Fig. 139 shows the distribution of hght in a horizontal plane 
 about a bare acetylene flame. It will be noticed that the curve 
 
 Fig. 142. — Light about acetylene gas flame in opal shade. 
 
 Fig. 14.3. — Opal acetylene shade. 
 
 is nearly circular which means that we get practically as much 
 light whether we look at right angles or parallel to the plane 
 of the flame. This fact (taking also into consideration that 
 the vertical distribution, as shown in Fig. 140, is nearly uniform) 
 
ACETYLENE GAS BURNERS AND THEIR SHADES. 
 
 99 
 
 shows that the hght is emanated equally in all directions and 
 that the mean spherical candle power is practicall}'- the same 
 as the horizontal or rated candle power. This, of course, applies 
 to the bare flame only. Fig. 140 also shows that as much light 
 is thrown above the horizontal as below and that unless some 
 means be taken to direct this light downward the ceiling and 
 . high side walls must be depended upon to obtain any useful illumi- 
 nation from this upward light, a most inefficient process. With 
 ordinary shades which act more or less as reflectors this condition 
 is made still worse, and even with the Holophane type there 
 is a large opportunity for improvement with respect to acetylene, 
 shades. 
 
 Fig. 144. — Light about a Holophane acetylene globe. 
 
 Fig. 141 shows the distribution of light about an etched globe, 
 shown in Fig. 108. A comparison with Fig. 107, giving the 
 test on the same globe with ordinary gas shows about the same 
 shape of curve. This globe gives some diffusion and is much 
 preferable to the bare light both from the standpoint of ease 
 on the eyes and artistic results. It gives 61 per cent of the 
 light above and only 39 per cent, below the horizontal. 
 
 Fig. 142 shows the distribution of light about an opal globe, 
 shown in Fig. 143. This globe gives 72 per cent, of the hght 
 above the horizontal and only 28 per cent, below, and is there- 
 fore ordinarily very wasteful. It does, however, give such fine 
 diffusion that it greatly enhances the ability to see well and 
 with low ceilings of a hght tint is to be recommended. 
 
100 PRACTICAL ILLUMINATION. 
 
 Fig. 144 shows the hght about a Holophane globe. This 
 throws 43 per cent, above and 57 per cent, below the horizontal, 
 and at the same time gives good diffusion. This shade is given 
 as representative of this type. 
 
 At present the ordinary forms of glassware are not well suited 
 for acetylene gas and there is a good field for the glassmakers 
 to make shades which will be both efficient in their distribution, 
 give good diffusion and be artistic. 
 
CHAPTER VIII. 
 
 INCANDESCENT ELECTRIC LAMP CLUSTERS AND BOWLS. 
 
 Fig. 145 shows the mean distribution of light in a vertical 
 plane about a cluster consisting of six 60-watt spherical frosted 
 bulb lamps, and one 120-watt larap of the same type, consuming 
 a total of 480 watts. This cluster has above it a white enameled 
 metal reflector as shown in Fig. 146. The central lamp has a 
 white enameled prismatic reflector. As will be noticed, it gives 
 a good general distribution of light ancl can be used to advantage 
 
 Fig. 145. — Light about a 4S0-watt incandescent cluster. 
 
 where a good light is required sideways as well as at other 
 angles. The use of frosted lamps gives a soft mellow light, and 
 this frosting is to be commended even at the expense of the life of 
 the lamps. 
 
 Fig. 147 shows the distribution of light about a cluster with 
 prismatic reflectors. This cluster contains six 16-candle power 
 standard lamps, each lamp being in an individual prismatic 
 reflector, and the whole being surmounted by a prismatic 
 reflector, as shown in Fig. 148. This cluster has the advantage 
 
 101 
 
102 PRACTICAL ILLUMIXATION. 
 
 of using standard lamps of any desired candle power and effi- 
 ciency. In the curve given six 16-candle power, 3.1 watts per 
 candle power lamps were used, giving a total of 300 watts. 
 Similar clusters of this type can also be obtained for lamps of 
 32 to 50 candle power. This type of cluster gives a strong 
 downward light as well as a fair amount at other angles. It 
 should be used with frosted tip lamps, to cut off all view of the 
 filament. It gains considerably in appearance by the use of all 
 frosted lamps. 
 
 Fig. 149 shows the distribution of light in a vertical plane 
 drawn through the lamps from a three-light cluster with con- 
 
 FiG. 14(). — 4S()-\vatt cluster. 
 
 centrating prismatic reflectors. This cluster is shown in Fig. 
 150. It is specially suited to lighting corridors and similar 
 places where light is to be thrown lengthwise of the corridor. 
 It can be used with the bottom lamp omitted if it is not desirable 
 to have as much light immediately under the fixture as in 
 Fig 149. 
 
 Fig. 151 shows the mean distribution of light in a vertical 
 plane about a cluster of four 16-candle power incandescent 
 lamps under a flat opal reflector, as shown in Fig. 152. This is 
 a very common type of cluster, and gives a very good general 
 distribution of light below the horizontal, but casts a shadow 
 
IXCAXf>E.-<CEXr EI.F.CTKIC LAMP flJJSTERy 
 
 10-: 
 
 
 ^^ " 
 
 \*/- 
 
 // IV 
 
 T 
 
 ■7* 
 
 .o*\ 
 
 
 % 
 
 Fir,. 14,. — Light alxait a oOO-watt 
 cluster in prismatic reflect nrs. 
 
 U^- 
 
 XJ4 
 
 / 
 
 
 
 /j\ 
 
 
 
 \ 
 
 5^- 
 
 "*-?'l 
 *' 
 
 
 ~8' 
 
 
 
 — >■■ 
 
 Fvf^. 14<t--Li,uht al.out a r; 
 li,t;lit cluster in prismatic re 
 tlectcirs. 
 
 Fig. 14s. — Prismatic cluster. 
 
 Fig. l.'iO. — o-liglit cluster with 
 prismatic reflect' irs. 
 
104 
 
 PRACTICAL ILLUMINATION. 
 
 on the ceiling. It should be used mainly for covering a large 
 area. For securing light under the cluster or when placed close 
 to high ceilings, especially in corridors, some of the other types 
 shown are preferable. As the lamps are entirely exposed such 
 
 .=& •♦ •£*l^-?-?#-S:s>-~s>- 
 
 Fig. 151. — Light about 4-light cluster with 19 in, opal reflectors. 
 
 clusters should have frosted bulb lamps where hung low in the 
 ordinary range of vision. Its chief objection for use in hand- 
 some stores, etc., is its entire lack of artistic appearance. It 
 should only be used where appearance is of little value. 
 
 Fig. 152 — 4-light cluster with opal reflector. 
 
 Fig. 153 shows the mean vertical distribution of light about 
 a cluster of four 16-candle power incandescent lamps under a 
 metal reflector, shown in Fig. 154. The arrangement of lamps 
 is such that the amount escaping above and near the horizontal 
 is much larger than is consistent with economy. It moreover 
 
IXCAXl'ESCEXr ELECTRIC LAMP CLUSTERS. 
 
 105 
 
 has the objections of casting a shadow on the ceiHng and being 
 inartistic in appearance. 
 
 Fig. 155 shows the mean distnlxition of hght in a vertical 
 plane about a cluster of four 16-candle power incandescent 
 lamps under a prismatic glass reflector, as shown m Fig. 156- 
 
 ;t--*-_ *- 
 
 ^ ■^z ^r; 
 
 
 
 \62. 
 
 ~~'~~—~—.^^ 
 
 —— 
 
 \ 
 
 
 \ 
 
 "" i'S- 
 
 Fig. 153. — Light alioiit 4-liglit clu.'^te^ \vith I'.i-in. cnarmled nit-tal ro- 
 
 flcclor. 
 
 This reflector is used by the United States (Government in its 
 buildings 'where clusters are needed, one of tlie principal reasons 
 vdiy it Avas chosen in -[jreferencc to the oi^al reflector being that 
 it does not darken the ceiling as much as tlie opal and gives 
 more light immediateh" underneath o\"er an area 3(1 degrees 
 
 Fii;. 154. — I'.idn. metal reflector and 4-li,t;ht-cluster. 
 
 each waA* from tlie A'crtical. It is also somewhat better in ap- 
 pearance. Unless placed liigh and out of the line of \-ision it 
 should be used with frosted lamps. 
 
 The general apphcation of clusters is m the lighting of places 
 which have a limited number of outlets, where it is not practi- 
 cable to distribute individual lamps over the area to be lighted. 
 
106 
 
 PRACTICAL JLLUMIXATIOX. 
 
 They are not to be recommended for efficiency as generally the 
 same number of lamps distributed over the area to be lighted 
 would give better results for a given operating cost, although 
 costing more to install. Some types_,of clusters can be installed 
 at a lower cost than chandeliers of equal appearance. 
 
 Fig. 155. — Light about 4-light cluster with prismatic reflector. 
 
 Fig. 156. — 4 light cluster, prismatic reflector. 
 
 Bowls or Hemispheres. 
 Fig. 157 gives the light distribution obtained from a sand- 
 blasted glass bowl with a 16-candle power incandescent lamp, 
 shown in Fig. 158. Such bowls or hemispheres are frequently 
 used for ceiling lights and sometimes on basket fixtures. The 
 
I^'CA^'DESCENT ELECTRIC LAMP CLUSTERS. 107 
 
 sand-blasted bowl evidently acts somewhat as a reflector, as 
 much light is thrown up. In practice this upward hght would, 
 of course, be thrown against the ceiling, and most of it would 
 probably be absorbed unless a good reflecting surface was pro- 
 vided inside the bowl. 
 
 Fig. 159 shows the distribution of light about a 16-candle 
 
 .S>* c^* «D (a 
 
 CO cp ^ >^ 'O 
 
 Candle power. 
 Fig. 157. — Light about a sand blasted hemisphere, 16 candle power lamp. 
 
 power incandescent lamp placed in a Holophane hemisphere 
 designed for use on ceilings and shown in Fig. 3 60. This test 
 was made without a backing of white asbestos and consequently 
 is comparable with Fig. 157. 
 
 As hemispheres and bowls are in such common use, it becomes 
 
 Fig. 158. — Sand blasted hemisphere. 
 
 of much importance to study the best arrangement oi lights 
 inside of such bowls. The old plan of placing clusters of lamps 
 at indiscriminate angles inside of these bowls is usualty not 
 to be recommended. A single lamp of high candle power gen- 
 erally gives better results than a number of, smaller units, as 
 such a lamp can be equipped with a reflector inside the bowl_ 
 
108 
 
 PRACTICAL ILLUMINATION. 
 
 When a cluster of lamps is used and each lamp is not equipped 
 with its individual reflector a piece of white asbestos, being a 
 good reflector, should be employed. This should not be placed 
 against the ceiling but between the supporting rim and the glass, 
 so as to bring the reflecting surface as close as possible to the 
 
 CauJle power 
 Fig. 159. — Light about a Holophane hemisphere, 16 candle power lamp. 
 
 lamps and avoid absorption by the rim, which would occur if 
 the asbestos were placed against the ceiling. Such a reflector 
 should be perforated with small holes which will allow ventila- 
 tion but effectually prevent light escaping between the ceiling 
 
 Fig. 160. — Holophane hemisphere. 
 
 and the rim, which often occurs if no reflector is used. The 
 best arrangement, however, is to use one lamp of high candle 
 power equipped with either a concentrating or distributing 
 reflector, depending on the way it is desired to throw the light. 
 While the diffuising bowl undoubtedly alters the distribution 
 
INCANDESCENT ELECTRIC LAMP CLUSTERS. 
 
 109 
 
 from the lamp and reflector, nevertheless the general character- 
 istics are preserved. It is generally desirable to use an opal or 
 prismatic rather than an opaque reflector if the reflector comes 
 down far enough on the lamp to cause shadows on the bowl. 
 
 so.o< 
 
 ■19.0 
 
 50.0 
 
 Fig. 161. — Light about 50 candle power lamp (lower hemisphere). 
 
 i33.0 
 
 c\i cc *o "^ 
 
 Fig. 162. — Light about ground glass bowl with 50 candle power lamp. 
 
 Fig. 163. — Light about ground glass bowl with 50 candle power lamp 
 
 and reflector. 
 
 Fig. 161 shows the light below the horizontal from a bare 
 50-candle power oval anchored filament lamp which was used 
 in the following tests on bowls. The tests without reflectors 
 are not exactly what would occur in actual practice, inasmuch 
 
no 
 
 PRA CTICA L ILL UML\'A TIOX. 
 
 as a certain amount of the upward light would be reflected if 
 the ceiling were light in color. As, however, even a light ceiling 
 quickly grows dirt}', owing to the particles of dust carried 
 upward bv the warm currents of air, the tests referred to give 
 
 20-0/ -' I v20.0 
 
 •>- ,y :^ "i %'--^ 
 
 Fig. 164. — Light about a beaded bowl, 50 candle power lamp (no reflector). 
 
 a good idea of what actually happens in practice if no special 
 attention is given to reflecting down the upward light. 
 
 Fig. 162 shows the light distribution from a 50-candle power 
 incandescent lamp in a 12-inch ground glass bowl, no reflector 
 being provided on the lamp. The great increase in useful 
 
 Fig. 165. — Beaded bowl. 
 
 illumination by the use of a reflector over the lamp in the bowl 
 is shown in Fig. 163. 
 
 Fig. 164 shows the light distribution from a 50-candle power 
 lamp in a 12-inch glass bowl covered with strings of glass beads, 
 
INCANDESCENT ELECTRIC LAMP CLUSTERS. 
 
 Ill 
 
 shown in Fig. 165. Such a bowl absorbs a great amount of 
 Ught, since the beads act as prisms to turn the hght in various 
 directions, where it is lost by multiple reflection and refraction. 
 Where cut glass beads are used around lamps in this manner, 
 
 Fig. 166. — Light about a beaded bowl with 50 candle power lamp and 
 
 reflector. 
 
 due allowance must be made (in this case about 50 per cent.) 
 for their high absorption by providing much larger candle 
 power in lamps than would otherwise be necessary. 
 
 The illumination from this beaded bowl, however, can be 
 very much increased by the use of a prismatic reflector over 
 
 Fig. 167. — Light about an opaline bowl with 50 candle power lamp (no 
 
 reflector). 
 
 the lamp, as is shown in Fig. 166, which gives the light distribu- 
 tion about the same beaded bowl with a reflector over the lamp. 
 The increase in the total amount of light below the horizontal 
 is 34 per cent, and in the zone 60 degrees each side of the vertical, 
 
112 
 
 PRACTICAL ILL DM I IX A TIOX . 
 
 which in hemisphere lighting is generally the most important, 
 no less than 60 per cent. 
 
 Fig. 167 shows the light about a 50-candle power lamp in 
 an opaline 12-inch bowl, while Fig. 168 shows the light distri- 
 
 FiG. 168. — Light about an opaline bowl with 50 candle power lamp and 
 
 reflector. 
 
 bution when using a reflector over the lamp in the same opaline 
 bowl. 
 
 Fig. 169 shows the Hght distribution from a 50-candle power 
 lamp in a 12.5-inch Holophane bowl. While the downward 
 28- ■ - aj.u 
 
 Fig. 169. — Light about Holophane bowl with 50 candle power lamp 
 
 (no reflector). 
 
 distribution of light in this case is considerably greater than 
 with any of the other bowls without reflectors on the lamps and 
 is almost as great as when reflectors are used on the other bowls, 
 the amount of downward light with the Holophane bowl can 
 
INCANDESCENT EBECTRIC LAMP CLUSTERS. 113 
 
 33.0 
 
 82.0 
 
 Candle Pov.'er 
 
 Fig. 170. — Light about Holophane bowl with 50 candle power lamp and 
 
 reflector. 
 
 'i''^ 
 
 Candle Povrer 
 Fig. 171. — Light about a 13-in. sand blasted ball with three lamps 
 
 (no reflectors). 
 
 Fig. 172. — Light about 13-in. sand blasted ball with 3 lamps and reflectors. 
 
114 PRACTICAL ILLUMINATION. 
 
 be greatly increased by the use of a reflector over the lamp, as is 
 shown by the light distribution in Fig. 170. 
 
 It is sometimes possible when hemispheres, large balls and 
 spheres already installed are placed high with a number of 
 lamps in them, to put small reflectors on the lamps and so in- 
 crease the downward light materially. This is not as good as 
 having one large lamp and reflector, but it is to be considered 
 as a method of making the best of a bad arrangement. 
 
 Fig. 171 shows the mean distribution of li.ght in a vertical 
 plane around a ground glass ball 13 inches in diameter with 
 three 16-candle power lamps hung vertically in it without re- 
 flectors. The effects of adding small prismatic reflectors (3.5 
 inches in diameter) over these lamps is shown in Fig. 172. It 
 is probably possible to design smaller reflectors specially built 
 for this class of work, which would give better results. 
 
CHAPTER IX. 
 NERNST LAMPS AND THEIR GLOBES. 
 
 Fig. 173 shows the mean distribution of light in a vertical 
 plane about a three-glower, 220-volt, 264-watt Nernst lamp 
 fitted with a clear six-inch glass ball globe. The mean spherical 
 candle power is 81. 
 
 Fig. 174 shows the mean distribution of light in a vertical 
 plane about a three-glower Nernst lamp with a six-inch sand- 
 blasted or roughed globe. 
 
 Fig. 173. — Light abottt 3 gloAver Nernst lamp, clear globe. 
 
 Fig. 175 shows the mean distribution of light in a vertical 
 plane about a three-glower Nernst lamp with a glass ball globe 
 having a slightly bluish tinge. 
 
 Fig. 176 shows the mean distribution of light in a vertical 
 plane about a three-glower Nernst lamp with an opalescent 
 ball globe. 
 
 The natural distribution of light from the Nernst lamp is 
 very similar to that of an incandescent lamp with certain 
 i-eflectors, the greater portion of the light being thrown down 
 so that for the best results it should be hung over the area to 
 
 115 
 
116 
 
 PRACTICAL ILLUMINATION. 
 
 be lighted and it should not be expected to light too large an 
 area. 
 
 Fig. 177 shows the hght about a three-glower Nemst lamp 
 with a Holophane distributing class B globe and indicates how 
 
 Fig. 174. — Light about 3 glower Nernst lamp, sand blasted globe. 
 
 the Nernst lamp can be made to give a wider distribution of 
 Hght. 
 
 Fig. 178 shows the light in a horizontal plane 30 degrees 
 below the horizontal with the same sand-blasted globe and lamp 
 as shown in Fig. 174. This shows the candle power is not 
 
 I - 
 
 
 /J ■& .^ V> JJ 
 
 /!\ 
 
 nSi 
 
 
 N I 
 , / 
 
 Fig. 175. — Light about 3 glower Nernst lamp, bluish tinged globe. 
 
 uniform in all directions. The effect of the sand-blasted globe 
 is to make it more uniform than with a clear globe. With the 
 Holophane globe it is even more uniform than with the sand- 
 blasted globe. 
 
NERNST LAMPS AXD THEIR GLOBES. 
 
 117 
 
 The distribution of light about the one, four and six-glower 
 lamps is very similar to that about the three-glower lamp, for 
 which tests have been shown. The efficiency increases with 
 an increase in the number of glowers. Taking the amount of 
 useful downward light (mean lower hemispherical candle power) 
 per glower as 100 per cent, for the three-glower lamp, the amount 
 of light per glower with the various sizes of lamps is about as 
 follows : 
 
 1 Glower 87 per cent. 
 
 3 " - -- --- 100 per cent. 
 
 4 " - 113 per cent. 
 
 6 " - 119 per cent. 
 
 ^^^SlnHl^^.o 
 
 Fig. 1/6. — Light about 3 gloAver Xernst lamp, opalascent globe. 
 
 Although the Nernst lamp naturally gives most of its light 
 downward a large increase in light at all angles below 20 degrees 
 from the horizontal can be secured by the use of opal, prismatic 
 glass or metal reflectors, an enameled metal reflector giving as 
 much as 70 per cent, increase directly down. 
 
 The Nernst lamp is well suited to the lighting of stores, halls 
 and churches, where it can be placed high above the area to 
 be lighted. It is important to place it high, out of the line of 
 ordinary vision, because a unit of such high candle power hung 
 low has a blinding effect on the eye. Because of the downward 
 distribution of the light as seen from the various tests given, 
 the hanging of Nernst lamps low is Hkely to result in too much 
 
118 
 
 PRACTICAL ILL UMINA TlO.y. 
 
 light immediatel)^ under each lamp and insufficient light midway 
 between the lamps. 
 
 Because of the intense brightness of the light from the glowers 
 it is never advisable to use these lamps with globes of clear or 
 very lightlv tinged opalescent glass, except where the lamps are 
 
 Fig. 177. — Light about 3 glower Xernst lamp, Holophane Class B globe. 
 
 placed so high that they are practically never seen or are hung 
 above sky lights which act as diffusers, Sand-blasted ball globes 
 are much to be preferred to the globes with slight bluish tinge 
 which are so frequently used with these lamps. The diffusion 
 is excellent with the sand blasted globe, so that the whole globe ap- 
 
 FiG. 178. — Light abotit 3 glower Nernst lamp, Horizontal plane, 30" 
 
 pears to be aglow, whereas with the blue tinged ball the intensely 
 bright glowers can be seen, which is not only bad because of 
 the blinding effect on the eye, but also because of its inartistic 
 appearance. Where the lamp is used for lighting streets and 
 corridors it must be remembered that the horizontal distribution 
 
NERNST LAMPS AND THEIR GLOBES. 119 
 
 about the light is not uniform and that the lamp should be hung 
 so as to throw the maximum light lengthwise of the corridor. 
 The lamp produces excellent artistic results when the complete 
 unit is placed inside of large ground glass or Holophane spheres. 
 The Nernst lamp is frequently used in locations where other- 
 wise an arc lamp would be used. Its steadiness is a strong point 
 in its favor when compared with arc lamps. It also gives nat- 
 urally more useful light below the horizontal than an alternating 
 current arc lamp taking the same amount of energy. As the 
 Nejnst lamp is made for alternating current only it is to be 
 considered only where such a current supply is available. 
 
CHAPTER X. 
 
 ELECTRIC ARC LAMPS. 
 
 The measurement of the Hght about arc lamps with a photo- 
 meter is a matter of considerable difficult}?. The light at any 
 one point varies considerably owing to the wandering of the arc 
 from one point to another on the carbon tips and variations in 
 the length of arc. Other things causing variation in results are 
 the size and quality of carbons used and the glassware. The 
 results obtained by different observers at different times are apt 
 to vary considerably. It is alwaj^s necessary to take the aver- 
 
 FiG. 179. — Light about 5 ampere d. c, 110 volt arc, clear inner globe alone. 
 
 age of a large number of observations to secure approximately 
 correct average results. The photometry of electric arc lamps 
 is in fact alm^ost as troublesome as that of gas lamps, as men- 
 tioned at the beginning of Chapter VI. These things should 
 always be remembered when controversies over the candle 
 power of arc lamps arise. 
 
 Fig. 179 shows the distribution of light about an enclosed 
 arc lamp operated from a constant potential 120-volt circuit 
 with about 82 volts at the arc and a current of about five amperes, 
 the total power consumption being about 600 watts. The arc 
 was equipped with a clear glass inner globe alone without any 
 other glassware. The maximum candle power is given at about 
 
 120 
 
ELECTRIC ARC LAMPS. 
 
 121 
 
 45 degrees below the horizontal, after which the light falls off 
 rapidh^. The use of nothing but a clear glass inner globe is 
 almost never permissible in interior lighting, because of the 
 great brilliancy of the arc and the dazzling effect which its 
 un diffused light would have on the eye. The mean spherical 
 candle power in this test was 213. 
 
 Fig. 180. — Light about 5 ampere d. c. 110 volt arc, clear inner and outer 
 
 globes. 
 
 The distribution and amount of light is affected considerably 
 by simply adding a clear glass outer globe over the clear glass 
 inner globe, the result of which is shown in Fig. 180. Practically 
 nothing in distribution is gained by this procedure and the glare 
 is nearly as bad as before. 
 
 Fig. 181 shows the same lamp with an opal inner globe alone. 
 The effect of the opal inner globe is to cause more light to be 
 
 Fig. 181. — Light about 5 ampere d. c. 110 volt arc, opal inner globe alone. 
 
 thrown above the horizontal, which in itself is not usually a 
 desirable thing to accomplish in interior lighting. The opal 
 inner globe, however, diffuses the light so much better than 
 the clear glass that the illuminating effect is much better, since 
 the light is not so blinding to persons in the room and they can 
 therefore see objects in the room better as explained before. 
 
122 
 
 PRA CTICAL ILL UMINA TIOX . 
 
 This is an arrangement in common use. With the aid of a 
 reflector it becomes ver^^ efficient. The small inner globe, how- 
 ever, does not offer as much diffusing surface as could be desired 
 and the arc should therefore not be placed in the line of vision. 
 Fig. 182 shows the effect of adding a clear outer globe over 
 
 
 ^,,fy-^ 
 
 
 Fig. 182. — Light about 5 ampere d. c, 110 volt arc, opal inner and clear 
 
 outer globe 
 
 an opal inner globe, this being to further cut down slightly the 
 light and to som.ewhat change the distribution. The clear outer 
 globe does away with the shadow or dark area immediately 
 under the lamp, which appears when the opal inner globe is 
 used. However, a clear inner and opal outer globe (see Fig. 
 
 
 
 ^^•■^^ ^'/-^.. 
 
 -\148 
 
 '-7'//\\\ -^>.%. 
 
 V75 
 
 Pig. 183. — Light about 5 ampere d. c, 110 volt arc opal inner and 
 
 outer globe. 
 
 184) gives far better diffusion and about the same loss in the 
 glass, and is much to be preferred. 
 
 Fig. 183 shows the light from the same lamp enclosed in both 
 an opal inner and an opal outer globe. The diffusion in this 
 case is excellent, but of course the hght is materially reduced. 
 There is also considerable strengthening of the light directly 
 
ELECTRIC ARC LAMPS. 
 
 123 
 
 below the arc which is sometimes of a good deal of A^alue. Fur- 
 ther, the violet rays are largety filtered out, producing a white 
 light of better quality. The inequality of light due to the 
 wandering of the arc is considerably reduced by the opal globes. 
 All told, for general interior work this is probably the best 
 combination. 
 
 183-' ' \ l93 
 
 Fig. 184. — Light about 5 ampere d. c, 110 volt arc, clear inner and opal 
 
 outer globe. 
 
 Fig. 184 shows the same direct current lamp with a clear inner 
 and opal outer globe. As far as diffusion is concerned, this is 
 probabl}'' the second best combination for interior lighting, as 
 the size of the opal outer globe causes better diffusion than the 
 small opal inner globe. It does not lend itself well to the use 
 of a reflector, so that some efficiency is usually sacrificed as 
 
 *Jfe^ 
 
 Fig. 185. — Light about o ampere d. c., 110 volt arc, opalescent inner 
 globe and |-in. carbons. 
 
 compared to the use of an opal inner globe alone. With the 
 latter reflectors can be used to increase the downward light, 
 but in such case the lamp must be hung high out of the range 
 of vision. 
 
 Fig. 185 shows the result of a test also made b)?- the Electrical 
 Testing Laboratories and presented in a paper by Mr. George 
 
124 
 
 PRACTICAL ILLUMINATION. 
 
 N. Eastman before the American Institute of Electrical Engi- 
 neers at Chicago on March 28, 1905. This test is on a five- 
 ampere direct current constant potential arc lamp operated 
 with 80 volts at the arc, an opalescent inner globe being used; 
 the lamp was provided with the usual 0.5-inch carbons. With 
 
 ^fi. t?. 
 
 316,' 
 
 
 Fig. 186. — Light about 5 ampere d. c, 110 volt arc, opalescent inner 
 globe and j-j- in, carbons. 
 
 the same current and voltage and the same inner globe, but with 
 5/16-inch carbons, the distribution of light shown in Fig. 186 
 was obtained, the total increase in mean spherical candle power 
 in Fig. 186 with small carbons over Fig. 185 with large carbons 
 being 38.2 per cent. The gain in efhciencjr by the use of small 
 
 Fig. 187. — Light abovit 3i ampere d. c, 110 volt arc, opalescent inner 
 globe and i in. carbons. 
 
 carbons is even more marked with arcs of smaller current. In 
 this connection it should be remembered that arcs employing 
 a small current are not as efficient as those using a larger current. 
 Fig. 187 shows the light about a 3.5-ampere arc lamp with 
 i-inch carbons, while Fig. 127 sho-^vs the light from the same 
 3.5-ampere arc lamp equipped with 5/16-inch carbon, the total 
 
ELECTRIC ARC LAMPS. 
 
 125 
 
 increase in light in this case being 52.8 per cent, due to the use 
 of the smaller carbons. Besides the great increase in efficiency 
 by the use of smaller carbons the light is much steadier because 
 the arc does not have the same opportunity to wander. This 
 in itself is an advantage, as one of the greatest objections to the 
 
 Fig. 188. — Light about 3| ampere d. c, 110 volt arc, opalescent inner 
 globe and yg^ in. carbons. 
 
 ordinarjr arc lamp with large carbons is its unsteadiness. The 
 small carbon arc also has less preponderance of violet in the 
 color of its light. In other words, it approaches nearer to daj^- 
 light than the large carbon lamp. While the small carbon 
 lamps require somewhat more frequent trimming than large 
 
 .. ^^ 
 
 
 
 # 
 
 %-~\ 
 
 
 
 /-;^ 
 
 ^>>' 
 
 A 
 
 
 
 ^«5\\. 
 
 ^ '\ 
 
 / 
 
 :V^<* 
 
 228-;::: 
 
 ^N: '^ 
 
 . / > 
 
 i::^/!^ 
 
 9»\ 
 
 ..— 
 
 ■■• — 
 
 -t91 
 
 97/. 
 
 _^. 
 
 .~-- 
 
 ::\97 
 
 
 
 
 
 Fig. 189. — Light abotit 5.7 ampere a. c. 
 
 globe. 
 
 *:-. 
 
 % 
 
 110 volt enclosed arc clear inner 
 
 carbon lamps, this is compensated for by the increase in effi- 
 ciency. The small carbon arc maintains its efficiency better 
 between trimmings because there is not such a blackening of the 
 inner globe as with large carbons. The cost of carbons and 
 trimming labor is, of course, higher the smaller the carbons. 
 Fig. 189 shows the distribution of Ught about an en- 
 
126 PRACTICAL ILLUMIXATIO.Y. 
 
 closed alternating current arc lamp operated from a 1 18- volt 
 constant potential circuit of 60-c3'cle frequency. It was ad- 
 justed so as to take about 5.7 amperes with 77 \'olts at the arc. 
 The power factor under these conditions was approximately 
 S7 per cent, for the arc and 62 per cent, for the lamp. The 
 power consumption was, therefore, about 416 watts. In Fig. 
 189 the arc was equipped with a clear glass inner glohie and had 
 the usual small white enameled reflector immediately at the 
 top of the enclosing globe, as used on G. I. arc lamps. This 
 reflector is 4.5 inches in diameter and is so placed that it inter- 
 cepts all rays of light above 58 degrees above the horizontal. 
 As the hght-giving crater forms on the ends of both upper and 
 lower carbons on the alternating current arc lamp, a large amount 
 of light is thrown above the horizontal in spite of the small 
 
 ^ ■ ^:^ -;i98 
 
 =^ /'l83 
 
 Fig. 190. — Light about .)." ampere a. c., 110 volt enclosed arc opal inner 
 
 globe alone. 
 
 reflector which is used. The mean spherical candle power is 152. 
 
 Fig. 190 shows the effect of substituting an opal inner globe, 
 which gives the conditions usually found in interior lighting 
 except that usually a large opal reflector is used in addition to 
 reflect down some of the fays thrown above the horizontal, where 
 they are of little use. 
 
 The equipment of an alternating current arc lamp with a 
 proper large reflector adds greatly to the useful downward light. 
 Thus a shallow opal dome-shaped reflector, 16 inches in diameter 
 by four inches deep, will increase the amount of light below the 
 horizontal 50 per cent., according to the tests made for the 
 National Electric Light Association b}^ Prof. C. P. Matthews 
 and reported in 1900. Without proper reflectors an alternating 
 current arc is likely to be a very inefficient means of securing 
 interior illumination. With such reflectors the alternating arc 
 
ELECTRIC ARC LAMPS. 
 
 127 
 
 can be made to approach the direct current arc in practical 
 illuminating efficiency. 
 
 The Flamixg Arc. 
 The flaming arc lamp is coming into use in America for out- 
 door street lighting and advertising purposes, and is used con- 
 siderably for outdoor work abroad. The efficiency of the arc 
 itself in some of the types is as much as five times that of the 
 five-ampere enclosed direct current arc, using 0.5-inch carbons. 
 At present writing it is not to be considered for interior lighting 
 except in such places as foundries, factories, and large exhibition 
 buildings where the fumes given off from the flaming carbons 
 are not objectionable, and where it can be hung high. For 
 such places it should give good results. 
 
 Fig. 191. — Light about 8 ampere flaming arc. 
 
 Fig. 191 shows the light about one type of eight-ampere 
 flaming arc equipped with a clear globe and requiring 45 volts 
 at the arc. The mean spherical candle power is 1020. The 
 distribution is mainly downward, so it is important to hang 
 these lamps high, both to secure proper diffusion and avoid such 
 a powerful source of light in the ordinary range of vision. The 
 light is so great in volume and intensity that large diffusing 
 globes should alwaj^s be used. The color of the light depends 
 on the chemicals which are used in the carbons. The highest 
 efficiency is obtained with carbons giving a reddish yellow light. 
 A white Hght can be obtained with a sacrifice of from 25 to 50 
 per cent, in efficiency. Care should be taken not to place such 
 arcs near electric signs, as such powerful iJluminants will spoil 
 the effect of any less intense lights placed in their vicinity. 
 
CHAPTER XI. 
 
 GAS ARC LAMPS. 
 
 Before taking up the study and performance of gas arc lamps 
 the reader is referred to the introductory remarks of Chapter VI 
 relative to the various factors which enter into gas testing. 
 The question as to the relative efficiency and value of individual 
 burners and gas arcs is considered in the Chapter on Comparison 
 of Illuminants. Two gas arc lamps of well known makers were 
 selected for test as being representative of gas arc lamps in 
 general. 
 
 Fig. 192. — i-light gas arc horizontal distribution, clear globe. 
 
 Fig. 192 shows the distribution of Hght in a horizontal plane 
 about the gas arc lamp shown in Fig. 194, equipped with clear 
 glass globe and flat porcelain reflector. It will be noted that 
 the distribution is unsymmetrical, the maximum exceeding the 
 minimum by 35 per cent. This is due to the shading of the 
 mantles. In one position only two mantles are visible, while 
 by shifting the point of view 30 degrees three of the mantels are 
 entirely visible and the fourth partly so. 
 
 128 
 
GAS ARC LAMPS. 129 
 
 It should, therefore, be noted that when instaUing such lamps 
 they should be p)laced so as to secure their maximum candle 
 power m the proper direction, as it is thus possible to materially 
 
 I -—-- - -S : -IL^ 1406 
 
 Fig. 193,— Light about a -i-light ga,s arc, clear globe, flat porcelain reflector. 
 
 increase the illumination in gn-en directions without any addi- 
 tional cost. This is a point of considerable practical use, but 
 seldom taken ad^-antage of. 
 
 Fig. 194. — Four-light gas arc, clear globe, flat porcelain reflector. 
 
 While no horizontal distribution tests were made with ala- 
 baster globes, as the lamps so tested was equipped with clear 
 globe and flat porcelain reflector, results of tests on electric 
 lamps giving unsymmetncal horizontal distribution show clearly 
 
130 
 
 PRACTICAL ILLUMINATION. 
 
 that a dense diffusing globe like alabaster tends to greatly offset 
 this lack of uniformity. 
 
 Fig. 193 shows the distribution of light in a vertical plane 
 about the same lam.p, as shown in Fig. 194. 
 
 -?i>,; 
 
 SlO' 
 
 320'- 
 
 
 h320 
 
 290\ 
 
 ^2% 
 
 ffi^ cd o ^ 'g 
 Fig. 195. — Light about a 4-light gas arc alabaster globe no reflector. 
 
 FiG. 196. Four-light gas arc, alabaster globe, no reflector. 
 
 Tests were made in two vertical planes, the first, normal to 
 the common vertical plane of two adjacent mantles. In this 
 case only two mantles were visible. The second vertical plane 
 investigated was 45 degrees removed from the first. In this 
 plane three mantles were visible. The indicated mean spherical 
 
GAS ARC LAMPS. 
 
 131 
 
 candle power and the indicated lower mean hemispherical candle 
 power values were computed from the intensities determined 
 by measurements in these vertical planes. As the distribution 
 of light in the horizontal plane about the lamp is non-uniform 
 
 .265 'r 
 
 Fig. 197. — Light about a four-light 
 gas arc, clear globe, no reflector. 
 
 Fig. 198. — Four-hght gas^arc 
 clear globe no reflector, fi 
 
 J 198 
 
 Fig. 200. — Four-light gas arc, 
 alabaster globe, no reflector. 
 
 ^'^^ 
 
 Fig. 199. — Light about a four-light 
 gas arc, alabaster globe, no reflector. 
 
 neither of the values given represents the true mean spherical 
 candle power, which will lie somewhere between them. This 
 can be verified at once by reference to the horizontal distribution 
 curve, shown in Fig. 192. 
 
132 
 
 PRACTICAL ILLUMINATION. 
 
 Fig. 195 shows the distribution in the second vertical plane 
 about the same lamp equipped with alabaster globe and without 
 reflector, as shown in Fig. 196. It will be noted that, except 
 directly underneath, there is a material increase in the light due 
 to the use of the clear globe, but on account of the diffusion 
 obtained by the alabaster globe, it is generally speaking, better 
 to sacrifice the increase of light, whenever the mantles are likely 
 to fall directly in the line of vision. The results of the test on 
 this arc are as follows: 
 
 Clear With Opal Alabaster 
 Globe Reflector Globe 
 First Second Second 
 Plane. Plane. Plane. 
 
 Pressure in inches of water 1.5 1.5 1.5 
 
 Consumption in cubic feet of gas per hour.. 21.0 21.0 22.5 
 
 Indicated mean spherical candle power. . . 230. 267 245. 
 Indicated mean lower hemispherical candle 
 
 power 281. 345. 240. 
 
 Mean spherical candle power per cubic foot 11. 12.7 10.9 
 Mean lower hemispherical candle power per 
 
 cubic foot of gas per hour 13.4 16.4 10.7 
 
 Fig. 197 shows the distribution in the second vertical plane of a 
 gas arc lamp, shown in Fig. 198, equipped with clear glass globe 
 and without reflector. With this arc tests were made in three 
 vertical planes, the third being half way between the first two 
 mentioned in the case of the previous test. In the third plane, 
 it is possible to see part of the fourth mantle. 
 
 The results of the tests on this arc are as follows : 
 
 Pressure in inches of water 
 
 Consumption in cubic feet of gas per hour 
 Indicated mean spherical candle power. . 
 Indicated mean lower hemispherical candle 
 
 power 168 
 
 Mean spherical candle power per cubic foot 
 Mean lower hemispherical candle power per 
 
 cubic foot of gas per hour 9.3 
 
 Clear Globe 
 
 Alabaster 
 
 without 
 
 Reflector. 
 
 Globe. 
 
 First 
 
 Second 
 
 Third 
 
 Plane. 
 
 Plane. 
 
 Plane. 
 
 1.6 
 
 1.6 
 
 1.6 
 
 . 18.0 
 
 18.0 
 
 19.0 
 
 177. 
 
 200. 
 
 162. 
 
 168. 
 
 185. 
 
 166. 
 
 ; 9.8 
 
 11.1 
 
 8.5 
 
 10.3 
 
 8.7 
 
GAS ARC LAMPS. 133 
 
 The remarks with reference to the mean spherical candle 
 power values of the first gas arc lamp appty also to this test. 
 
 Fig. 199 shows the distribution in the third vertical plane 
 about the same lamp equipped with alabaster globe but no re- 
 flector, as shown in Fig. 200. 
 
 Relative to this arc lamp it should be noted that it was not 
 provided with adjustments which permitted it to be operated 
 on the low pressure as satisfactorily as the first one mentioned. 
 Both lamps, however, were operated on the pressure which 
 obtains in New York City. 
 
CHAPTER XII. 
 
 VAPOR LAMPS. 
 
 The mercury vapor arc lamp is the principal vapor electric 
 lamp in commercial use. The light is bluish green in color 
 and since it gives off no red rays its use is out of the question 
 where it is desirable to see objects in their ordinary colors as 
 we know themi. There are a large number of places, however, 
 where color is of no importance and where the mercury vapor 
 lamp gives excellent results, as, for example, in factories, draught- 
 ing rooms, warehouses and work rooms of all kinds where color 
 values are not needed and where the important thing is a good 
 
 Fig. 201. — Light about mercury vapor lamp without reflector. 
 
 diffused Light to work b}-. The two points strongly in its favor 
 are its high efficiency and its excellent diffusion of light. As 
 to efficiency it is undoubtedly next after the flaming arc lamp, 
 the most efficient electric illuminant in commercial use at the 
 present writing. On account of the color of the mercury vapor 
 lamp, the length of the tube and other unusual factors the 
 photometric testing of these lamps offers grave difficulties and 
 no accurate absolute figures are available at present writing. 
 It is safe to say that these lamps will not require more than about 
 one watt per mean spherical candle power when operated at an 
 
 134 
 
VAPOR LAMPS. 
 
 135 
 
 efficient point. The general form of the distribution of light 
 in a vertical plane about a horizontal mercury vapor tube with- 
 out any reflector over the lamp is shown in Fig. 201. Absolute 
 candle power values are not given for reasons before stated. 
 The effect of putting a white enameled reflector over the tube 
 is shown in Fig. 202, where the values are given in percentages 
 of the maximum value obtained without reflector. The reflector 
 increases the light immediately below the lamp 213 per cent, 
 and also increases it at all other angles up to 20 degrees below 
 the horizontal. For most positions the reflector is advisable. 
 
 Besides being very efficient the fact that the light comes 
 from a long tube gives excellent diffusion, so that the light is 
 almost as free from sharp shadows as daylight. This makes 
 it a good light to work by. Besides being free from shadows 
 the light is similar to daylight, in that there is not as much 
 
 Fig. 202. — Light about mercury vapor lamp with reflector. 
 
 glare and glisten from paper, tools and machine work as with 
 most artificial light sources which are smaller and more intense. 
 This fact even more than the absence of sharp shadows is the 
 cause of its being comfortable to work under. These lamps are 
 made only in large sizes, one taking 385 and another 193 watts. 
 For this reason the general practice when using them is to pro- 
 duce enough general illumination all over the room to be lighted 
 so that close work can be done anywhere in the room rather 
 than to concentrate light at special points for special purposes, 
 as with smaller and less efficient illuminants. 
 
 The Moore vacuum tube light is hardly to be considered a 
 commercial proposition at present writing, and as reliable data 
 on it is not yet available the subject is not taken up. 
 
CHAPTER XIII. 
 
 DEMONSTRATION ROOM TESTS. 
 
 In Chapter V are given a large number of tests on reflectors, 
 shades and globes for individual incandescent electric lamps. 
 These tests show the amount of light given out in various 
 directions from an incandescent lamp as equipped with various 
 glassware. Although these tests as given tell the experienced 
 eye just what may be expected from reflectors and globes in 
 practice, it is, nevertheless, instructive to he able to note the 
 practical effect of different reflectors, shades and globes in the 
 lighting of a room. For this reason there are reproduced here 
 a number of photographic demonstration room tests. To make 
 these tests, two small rooms were constructed side by side. 
 The dimensions of each room were about one-third those of a 
 room 13 feet square by 9 feet high, thus giving conditions com- 
 parable to the average small residence or of&ce room. The side 
 of the room toward the observer was left open so that the light- 
 ing effects on the floor, walls and ceiling could be noted. The 
 same color was used on the floors, walls and ceilings of the two 
 rooms. The ceilings were gray and the walls light green. By 
 equipping the lamp in the right hand room with one kind of 
 reflector shade, or globe and the lamp in the left-hand room with 
 another kind, the comparative practical results in the lighting 
 of the room can be seen; the source of light being shaded from 
 the eye by a piece of black cardboard which prevents the direct 
 rays from shining into the eye or the camera and interfering 
 with a careful comparison. By using the same kind of incan- 
 descent lamps in both rooms the light is practically the same 
 color in both, and hence will have the same value on the photo- 
 graphic plate. The two rooms can therefore be photographed 
 as a pair, as seen in the tests v>rhich follow, and a record of the 
 tests can thus be preserved. This method of testing is not to be 
 considered as accurate as testing with a photometer and the 
 
 136 
 
DEMONSTRATIOX ROOM TESTS. 137 
 
 results given are purely comparative. Nevertheless it has the 
 advantage that it shows the practical effect of using different 
 reflectors, shades and globes in the average room. It also takes 
 into account the effect of reflection from the walls of a room 
 which photometric tests do not. In all of the following tests^ 
 unless otherwise specified, one 16-candle power incandescent 
 lamp was used in each room. The total amount of light and 
 electrical energy in each room being the same the differences 
 shown are, therefore, due entirely to the different effects of the 
 reflectors, shades and globes used. 
 
 In test No. 1, which is here shown, the left-hand room has 
 a bare 16-candle power incandescent lamp, hanging at the usual 
 height of a lamp on a pendant. In the right-hand room is a 
 lamp at the same height equipped with the fluted opal cone 
 reflector shown in Fig. 20. This reflector adds considerably 
 to the light thrown down toward the floor, this light being 
 secured at the expense of the high side walls and ceiling, which 
 are considerably darker than where the bare lamp with no 
 reflector is used. 
 
 In test No. 2 the left-hand room has a bare lamp on a pendant. 
 In the right is a lamp equipped with the prismatic reflector 
 shown in Fig. 34. This reflector concentrates considerable light 
 immediately under the lamp without materially darkening the 
 ceilings and side walls. 
 
 Test No. 3 has in the left room a lamp in a very concentrating 
 form of prismatic reflector, shown in Fig. 36. In the right is 
 a bare lamp. A great amount of light is concentrated under 
 the lamp in the left-hand room. This represents an extreme 
 case of changing the distribution of light about the bare lamp 
 and concentrating it at one point with the aid of a reflector. 
 
 Test No. 4 has in the left room a lamp in the corrugated 
 mirror desk shade, shown in Fig. 54, as against a bare lamp in 
 the right-hand room. This is also an extreme case of directing 
 the light downward at the expense of the walls and ceilings. 
 In this case the upper walls and ceiUngs in the left-hand room 
 are lighted entirely by reflection from the intensely lighted floor 
 and lower side walls. 
 
 Test No. 5 has in the left room a ground glass hemisphere 
 at the ceiling, shown in Fig. 158, and at the right a Holophane 
 hemisphere at the ceiling, shown in Fig. 160. The amount of 
 
138 PRACTICAL ILLUMINATION. 
 
 light thrown on the floor in the latter case being notably greater. 
 
 Test No. 6 has in the left room a Holophane Class A globe, 
 designed to throw the light downward on a pendant, 
 shown in Fig. 79, and at the right a prismatic reflector, at 
 the ceiling, shown in Fig. 34. The floor is considerably 
 better lighted with the prismatic reflector at the ceiling than 
 with the Class A Holophane diffusing globe on the pendant. 
 The idea in making this test was that when lamps are placed 
 on pendants or on chandeliers they should be equipped with 
 difliusing globes to avoid the glare of the lamps in the ordinary 
 range of vision of persons in the room. This is not so necessary 
 when lamps are placed at the ceiling, although even in the latter 
 case lamps with frosted bulbs should be used. The question 
 naturally comes up whether the lamp at the ceiling in a reflector 
 is as efficient a method of lighting a room as a lamp placed lower 
 on a pendant in a diffusing and directing globe. This test 
 settles this point in favor of the ceiling arrangement. For 
 another test similar to this see Test No. 13. 
 
 Test No. 7 has in the left room a lamp placed upright in the 
 Holophane open shade, shown in Fig. 97. In the right is a bare 
 incandescent lamp upright. The illumination on the floor is 
 considerably better in the left-hand room, indicating a method 
 bjr which the down ward illumination can be increased on fixtures 
 having upright lamps. 
 
 Test No. 8 has in the left room a lamp in a cvlindrical desk 
 shade, shown in Fig. 61, and in the right an aluminum., parabolic 
 half reflector, shown in Fig. 59. The amount of light with the 
 parabolic reflector is greater both in area and intensity than 
 with the cylindrical reflector. 
 
 Test No. 9 shows in the left-hand room the prismatic reflector 
 shown in Fig. 36, and in the right-hand room the prismatic 
 reflector shown in Fig. 34. This test shows the miarked differ- 
 ence in light distribution with different types of prismatic 
 I'eflectors and emphasizes the necessity of selecting the reflector 
 which will give the proper distribution in any given case. 
 
 Test No. 10 shows in the left-hand room an opal bell reflector 
 (Fig. 24), pointed at an angle of 45 degrees to the left of vertical, 
 and in the right-hand room a bare lamp pointed at an angle of 
 45 degrees to the right of vertical. In both cases it is seen that 
 considerably more light than is necessary is going to the ceiling. 
 
Left Nil. 10 Rik'lit Xo. 17 
 
 Bare lami>. Fluted opal cone. 
 
 Showing great difference in the amount of hght thrown down hy the 
 use iif the opal eone reflecti'r as compared to no reflector. 
 
 Tert X:.. 1. 
 
 Left Xi.. 10 Right X.I. 1 1 
 
 Bare lamp. l\igoda reflector Xo. 2:501. 
 
 Shewing how light from a hare lamp can be modified to thmw much 
 more useful hght downward without seriously darkening walls and ceilings. 
 
 Test X... L>. 
 
 Left Xu. lil Right Xo. 10 
 
 Pagoda reflector Xo. 2().'>1. Bare lamp. 
 
 Showing way in which amount of light given out in dift'erent directions 
 can be changed V)\- use of a reflector. This is a rather extreme case. 
 
 Test Xo. :-!. 
 
^ „.aai 
 
 Left No, 15 Riu-ht Nu. 10 
 
 X-ray desk sli;ide Xn. ()9("i. Bare lamp. 
 
 Showing an extreme ease of dire:'ting light in a eertain direetion f^r 
 S]ieeial purposes. 
 
 Test Xo. 4. 
 
 Left Xm. 37 Right Xi> :is 
 
 Ground glass hemis])here i m ceiling LLilfiphane hemisphere Xd. 7.50 on 
 
 eeiling, 
 Sho^^"ing relati\'e efheiene\' of t\\"o eommnn me'lhods ot lUummatiem. 
 
 Test No. 5. 
 
 Left Xo. 2:-! Right Xo. 11 
 
 Holophane Class A globe X..i. 33(i:i Pagoda reflector ("lass B Xo. ^IriOl at 
 on pendant. ceiling. 
 
 Showing reflector at eeiling to give more light underneath the lamp 
 than even a Class A Holc^phane, and also to give Letter illumination of the 
 room. 
 
 Test Xo. G. 
 
Left No. 31 Right No. 30 
 
 Holopliane open shade No 2805 Class Bare incandescent lamp upright. 
 B upright. 
 Showing easy way of improving general downward distribution of 
 light on fixtures with upright lamps and also on gas bvimers. Bare in- 
 candescent or gas lamp throws most light to the ceiling and upper walls. 
 
 Test No. 7. 
 
 Left No. 27 
 Cylindrical desk shade, coated with 
 aluminum, f)ainted inside, placed 
 horizontally as in practice. 
 
 Right No. 26 
 Aluminum parabolic half reflector, 
 frosted, placed horizontal, as in 
 practice. 
 
 Showing general sujieriority of parabolic alummunt reflector. 
 Test No. 8 
 
 Left No. 21 Right No. 11 
 
 Pagoda reflector Class A No. 26.51. Pagoda reflector Class B No. 2301. 
 
 Showing difference in two classes of reflectors and the necessity of care 
 in selection for any given p)urpose. 
 
 Test No. 9. 
 
Lt-ft Xn. 12 
 
 ( )]>;il bell at angle 4.') deg. 
 
 Right Xij. i;-; 
 
 Bart lamp at angle 4') deg. 
 
 Test X.- 
 
 10. 
 
 Left Xu. 11 Right Xd. .'):> 
 
 I'ag'.da retleetc^r Xo, l!:-!01 '.i ft. Stand laniii with flecnrated ■ipal dmne 
 
 alxixe fl'H>r. almut 11 inelies diameter. 
 
 The area occupied Vjv an ordinary table is indicated by the lines drawn 
 "11 the floor of the right-hand room. This illustrates how much larger 
 an area of useful reading light is obtained with a good reflector, pointing 
 down on a chandelier than by a table lamp, unless ttie liook to be read 
 IS placed on the table almost against the lamp. The lamji in the pris- 
 matic reflector on the chandelier illuminates the whole room much better 
 than the stand lamp/. 
 
 Test Xo, 11, 
 
Left Xo. .')(> 
 Id-inch sand blasted dome 
 
 Right No, 52 
 10-inch opa! dome. 
 
 Test Xo. 12. 
 
 ■^■^^ 
 
 irrJHH 
 
 
 E_.^ 
 
 
 
 '"" 
 
 '^"^^ 
 
 ^BH 
 
 Hi 
 
 m 
 
 
 P 
 
 1 
 
 
 
 
 ■ 
 
 
 
 
 
 
 '^im 
 
 
 
 
 ^mg^^gg 
 
 wKSkiii.-^. , 
 
 
 
 
 
 
 '^ 
 
 '^m 
 
 ^^^H 
 
 mipi 
 
 
 
 
 
 
 ^' 
 
 M 
 
 ^^^KffKm 
 
 ppppjura-^.Ji 
 
 
 
 ■* ' 
 
 
 
 
 ^te^H 
 
 BB«:: 
 
 «..,.^.^_ .. 
 
 w^-ai^- .^.^ 
 
 
 
 
 
 
 T^a 
 
 Left Xo. 11 Riglit Xo, 4V) 
 
 Pagoda reflector Xo. 2::!()1 at ceiling. Pagoda reflector Xo, 7381 on a ]ien- 
 
 dant. 
 This shows Pagoda Xo, 7:->Sl on a pendant to be superior for the gen- 
 eral lighting of the lower part of the room. Pagoda Xo. 7381 is deep 
 enough so that when on a pendant but little of a 16-c,p, lamp fjlament 
 comes in the ordinary hne of \-ision of a person seated anvwhere in the 
 room , 
 
 Test No. 13. 
 
^^^^^^^Hi^P^^^^^a 
 
 
 ^^M 
 
 Left X... .')(,) Right So. 49 
 
 Opal 7-inch diinie at Cfihng. Paguda reflector Xo. 7381 at ceiling. 
 
 The corners of the rnom are better lighted 1)\- the X^i. 7381 Pagoda, 
 although the light immediatelv under tlie o]ial dome is lietter. 
 
 Test Xo. 14 
 
 Left Xo. ,-,:-; 
 
 Holii],li;inc Class B stalactite 
 
 Right Xo. ,-,4 
 l'nm]ied opal bell with frosted bullj 
 lamp. 
 This test was made ti'i determine the relati\"e ethciency of two equip- 
 ments A\"liich gi\-e a \"er\' similar distriliution of light with both of which 
 the light from the bare filament is dift'used and softened. 
 
 Test Xo. 15. 
 
DEMONSTRATION ROOM TESTS. 
 
 139 
 
 It is evident that with such an opal bell reflector the socket 
 should be pointed straight down instead of being placed at an 
 angle of 45 degrees, as is common on chandeliers. 
 
 Test No. 11 has in the left room the prismatic reflector shown 
 in Fig. 34, and in the right a table lamp with a decorative opal 
 dome reflector, 11 inches in diameter. This table lamp is shown 
 in Fig. 203, and is of a more efficient type for reading purposes 
 than the average electric portable table lamp, for the reason that 
 the opal dome reflector tends to throw light sideways beyond 
 the edges of the table. The lamp in the reflector in the left- 
 hand room is at a height above the floor that the ordinary 
 chandelier lamp would be above the reading page of a person 
 
 Fig. 203. 
 
 seated under it. In such a case there would be no table under, 
 the chandelier and persons reading would sit in the well-lighted 
 area, getting the full benefit of the light. The table lamp in 
 the right-hand room being on the floor is, of course, directly 
 comparable to the stand lamp on a table, if we consider the floor 
 as representing the height of a reading page of a person seated 
 around the table. The space occupied by the table is indicated 
 by lines drawn about the lamp. It is seen that a much larger 
 area of useful reading light is obtained with a good reflector 
 pointed straight down on a chandelier than by a table lamp, 
 unless the book to be read is placed on the table almost against 
 the lamp. The lamp in the prismatic reflector on the chandelier 
 
140 PRACTICAL ILLUMINATION. 
 
 also illuminates the whole room much better than the stand 
 lamp. 
 
 Test No. 12 has in the left-hand room a ten-inch sand-blasted 
 dome reflector, and in the right-hand room a ten-inch opal 
 dome. The general shape of these reflectors is similar to that 
 seen in Fig. 22, although the reflectors in this test were larger 
 in diameter. This test shows the relative reflecting powers of 
 sand-blasted and opal glass. The opal dome throws much more 
 light on the floor than the sand-blasted dome, but of course 
 robs the side walls and ceilings. In some cases the distribution 
 from the sand-blasted dome would be preferable and in other 
 cases that from the opal dome. 
 
 Test No. 13 has in the left-hand room the prismatic reflector, 
 shown in Fig. 34, and in the right-hand room the prismatic 
 reflector shown in Fig. 38. The reflector in the right-hand room 
 is deep enough to hide the lamp filament when put on a pendant, 
 and consequently was so placed. The other reflector in the left- 
 hand room being shallow was placed at the ceiling to get it out 
 of the ordinary line of vision. The whole floor area is better 
 lighted in the right-hand room since the reflector used gives 
 a wide distribution. The results of this test should also be con- 
 sidered in connection with test No. 6, where the same reflector 
 was used on the ceiling and the conclusion was drawn that 
 with the glassware used in that test better efficiency could be 
 secured by the reflector on the ceiling than by the diffusing globe 
 on a pendant. In this case, however, we see that this particular 
 type of deep reflector on a pendant is more efficient than the 
 reflector at the ceiling, as far as the lighting in the lower part 
 of the room is concerned, but it must be remembered that when 
 the general illumination of a room is wanted rather than specific 
 lighting for reading purposes, it is desirable to throw consider- 
 able light around the lower side walls for lighting pianos, book- 
 cases, etc. In most cases the reflector in the right-hand room 
 (Test No. 13) will not give sufficient light at such points. The 
 same can be said of the reflector in the left-hand room. For 
 this reason it is usually better to use a diffusing globe on a pend- 
 ant for such general lighting or else to place at the ceiling a 
 reflector giving a distribution similar to that shown in the right- 
 hand room. 
 
 Test No. 14 has in the left-hand room at the ceiling a seven- 
 
DEMONSTRATION ROOM TESTS. 141 
 
 inch opal dome, shown in Fig. 22, and in the right-hand room 
 at the ceiling the prismatic reflector shown in Fig. 38. Both 
 these reflectors, when at the ceiling, give good results for the 
 general illumination of a room, including both the center and 
 side walls. The opal dome gives a little more immediately 
 underneath, while the prismatic reflector gives more around 
 the lower side walls. Both these reflectors are good for the 
 purposes just outlined. 
 
 Test No. 15 has in the left room a Holophane Class B stalacite, 
 designed to give a general illumination below the horizontal 
 seen in Fig. 81, and at the right the crimped opal bell, shown in 
 Fig. 24, with a frosted bulb lamp. This test shows what might 
 be expected from the photometric tests of these two different 
 pieces of glassware, namely, that the distribution of light is very 
 similar. Both these arrangements are well suited to the general 
 lighting of small rooms where light is wanted in all directions 
 below the horizontal. In both cases the light is well diffused, 
 although the diffusion secured with the globe is better than that 
 with the frosted bulb lamp. 
 
 In looking over these tests no attempt should be made to 
 compare the light shown in a certain room in one test with the 
 light shown in a certain room in another test, for the reason that 
 there are always unavoidable differences in photographs which 
 cause variations in the apparent amount of light in one test as 
 compared to another. Comparisons should be made only be- 
 tween rooms side by side in the sarae test. 
 
CHAPTER XIV 
 COMPARISON OF ILLUMINANTS. 
 
 In comparing different illuminants one must consider the quest- 
 ions of economy, convenience, steadiness, cleanliness, safety and 
 color. As to economy, the following table for electric light gives 
 the total amount of light obtained or mean spherical candle power 
 hours per kilowatt hour for various electric illuminants. This 
 figure is obtained by dividing the mean spherical candle power 
 of the illuminant by the kilowatts required to operate it. The 
 mean spherical candle power of the illuminant was taken from 
 tests made with clear glassware without reflectors in every case. 
 The same is true of the second table, which gives the mean 
 spherical candle power hours per cubic foot of gas per hour for 
 various gas burners. 
 
 In considering the tables allowance mu«t be made tor a 
 number of things. The mean spherical candle power without 
 diffusing globes or reflectors represents simply the total amount 
 of light produced. The amount that is delivered where it is 
 wanted is another matter. In order to deliver the light where 
 it is wanted and to soften and diffuse it, reflectors, shades or 
 diffusing globes are necessary in almost every case, and there 
 is a certain amount of loss in so reflecting and diffusing the 
 light. This loss will of course be less with illuminants which 
 naturally give a large percentage of their light below the hori- 
 zontal (where it is usually wanted) than with those which do not. 
 Thus the Nernst lamp and the inverted gas burner have an 
 advantage at the start over other illuminants, as can be seen 
 from the light distribution curves shown in the chapters treating 
 them. Likewise also the direct current arc lamp has an initial 
 advantage over the alternating current arc lamp. These differ- 
 ences in distribution can be more or less corrected by reflectors, 
 as has been shown, but the nearer the desired distribution we 
 have to start with the less light must be reflected and conse- 
 quently the less will be the total loss in the process of reflection 
 
 142 
 
COMPARISON OF ILLUMINAJSTS. 
 
 1.43 
 
 The initial efficiency of the- larger sizes of electric arc and 
 Nemst lamps is higher than for the srrialler sizes, but this fact 
 is likely to deceive the novice as to the possible practical efficiency 
 of the various sizes. All of the smaller lighting units, because 
 they can be placed nearer or over the areas to be lighted, even 
 though of lower efficiency than the larger ones, can, if properly 
 
 Comparison of Electric Illuminants without Reflectors and with 
 Clear Glass Globes. 
 
 Amount 
 of Light 
 per Kilo- 
 Watts watt Hour 
 T. *■ or Mean 
 
 Mean per Mean Splierical 
 
 Spherical Spherical Cand^« 
 
 Candle Candle Hours per 
 
 Kind of Lamp. Power. Power. Hour. 
 
 4.24 
 
 3.78 
 
 236. 
 
 264. 
 
 Common 56 watt carbon filament incan- 
 descent lamp rated at 3.5 watts per can- 
 dle ; 16 horizontal candle power 13.2 
 
 Common 50 watt carbon filament incan- 
 descent lamp rated at 3.1 watts per can- 
 dle ; 16 horizontal candle power 13.2 
 
 Direct current 3.5-ampere enclosed arc c^n 
 
 110 volt circuit, 0.5 inch carbons 110. 
 
 3-glower 264 watt Nemst lamp 81.0 
 
 High-efficiency, Gem, 125-watt graphitized 
 carbon filament lamp of 50 horizontal 
 candle power. 40 . 7 
 
 44 watt tantalum lamp 22 rated horizontal 
 
 candle power 16.0 
 
 Alternating current enclosed 5.7-ampere 
 arc taking 388 watts on 110-volt circuit, 
 0.5 inch carbons 152 . 
 
 Direct current 5.1 ampere enclosed arc on 
 
 110 volt circuit, 0.5 inch carbons 213. 
 
 Luminous 8 ampere arc, 440 watts, 2 in 
 
 series on 110-volt circuit 1020. 
 
 3.50 
 
 285 
 
 3.26 
 
 307 
 
 3.07 
 
 326 
 
 2.75 
 
 364 
 
 2.55 
 
 392 
 
 2.63 
 
 380 
 
 0.431 
 
 2320 
 
 placed and properly equipped with reflectors, give fully as much 
 useful illumination for a given expenditure for current as larger 
 lamps. It is not always feasible to distribute them as thoroughly 
 as would be necessary to obtain these results, hence if the outlets 
 must be few for any reason, the larger lamps will give better 
 
144 PRACTICAL ILLUMINATION. 
 
 efficiency. The advertising value of the larger units in a store 
 must also be considered. 
 
 In comparing the gas arc with individual burners it must first 
 be noted that the efficiency of the gas arc lamp is somewhat 
 less, as was shown by the table. We should also note here 
 again that, generally speaking, better distribution and therefore 
 more efficient lighting can be obtained with small units, because 
 when such small units are equipped with the proper shade or 
 reflector we can obtain more uniform lighting at a less expendi- 
 ture of energy. However, it is not always desirable to do this, 
 
 Comparison of Gas Illuminants with Clear Glass Globes. 
 
 Cubic Amount 
 Pfpt nf °f Light 
 
 I'eet oi pg^ Cubic 
 
 Gas per Foot of 
 
 TT Gas or 
 
 Hour Mean 
 
 Mean per Mean ?;a^dle 
 
 Spherical Spherical Po'^'er 
 
 Candle Candle Cubic Ft. 
 
 Kind of Lamp. Power. Power. "^Hou?." 
 
 Open gas, 5 cubic feet per hour at 1.5 inch 
 
 water column pressure 20 . . 250 4 . 
 
 Gas arc, 21 cubic feet per hour at 1.5 inch 
 
 water column pressure 325 . . 065 15.5 
 
 Upright standard mantle burner 3.7 cubic 
 
 feet per hour at 1.6 inch water column 
 
 pressure 63 . . 0588 17.0 
 
 Inverted burner, 3.0 cubic feet per hour at 
 
 1.5 inch water column pressure 54. 0.0556 18.0 
 
 inasmuch as a large source of light, such as a gas arc, is often 
 an advertising feature and when limited to a few outlets the 
 gas arc may be better than a chandelier equipped with a number 
 of burners, on account of space occupied and cost of installation. 
 Generally speaking, however, it is safe to say that the gas arc 
 is a much more expensive proposition to use than the individual 
 burners properly placed. There are so many factors, however, 
 entering into the problem, such as maintenance and first cost 
 (as in many cases the gas company installs and maintains the 
 arc free of charge), that each case must be decided on its own 
 merits. A gas arc is apt to receive much better care than a 
 
COMPARISON OF ILLUMINANTS. 145 
 
 number if individual burners and the lower efficiency of the arc 
 may be offset by the superior condition in which it is kept. . ; 
 As to color of light, the direct current electric arc lamp with 
 small carbons enclosed in dense opal inner and outer globes, 
 approaches most nearly to daylight of any of the illuminantS 
 enumerated ; that is, it gives very nearly white light. The en- 
 closed arc as ordinarily used has an excess of violet rays in its 
 Ught. The gas mantle burners, some of which approach very 
 near to white light, have a slight preponderance of green, which 
 causes the ghastly effect sometimes complained of with these 
 burners. The Nernst lamp has considerably more yellow in its 
 rays than daylight, although it is whiter than the ordinary in- 
 candescent lamp. The ordinary electric incandescent, the open 
 gas jet and the kerosene lamp are very similar in color, the light 
 being decidedly yellow as compared to daylight. A short dis- 
 cussion of the question of the color of artificial lights is taken up 
 in connection with store lighting. 
 
 It is well to bear in mind in the case of all these illuminants 
 that the tests given in this book represent approximately the 
 best working conditions with new lamps. The tendency of all 
 is to fall off in efficiency with age. 
 
 Gas mantle burners in practice, in the hands of ordinary users, 
 are frequently not adjusted to give the best results. 
 
 Incandescent electric lamps are frequently used after their 
 efficiency has fallen off 30 or 40 per cent, from blackening and 
 other causes. 
 
 Steadiness is an important quality to secure in an illuminant. 
 Those most likely to be unsteady are open gas jets and arc lamps. 
 These will usually flicker some no matter how steady the source 
 of supply. Other electric and gas lamps are steady as long as 
 the supply is steady, as it is in the majority of cases. 
 
 The consumption of the mercury vapor lamp is not obtainable 
 in reliable figures at present writing, but as stated in the chapter 
 on that subject, the best lamps of this type probably consume 
 as low as one watt per mean spherical candle power. The 
 light from mercury vapor lamps as it comes from a long tube 
 surface does not produce dark shadows or as much annoying 
 glare from papers or tools as would carbon arc lamps of sim- 
 ilar candle pov/er unless the latter had very large reflectors and 
 diffusing globes. The absence of red rays and the greenish blue 
 
■146 PRACTICAL -ILLUMINATION. 
 
 color of the light would be extremely objectionable in some 
 places, but in places where it is not necessary to judge of color 
 and where a good light to work by is the principal object, the 
 color. apparently is no drawback to its use after once the users 
 become accustomed to the strange light. 
 
CHAPTER XV. 
 INTRODUCTION TO PRACTICAL EXAMPLES. 
 
 In the following chapters of this book the discussion is taken 
 up of a large number of practical examples of lighting arrange- 
 ments, both good and bad. The authors believe that by this 
 method a much clearer idea of what constitutes good and bad 
 design can be obtained than by a discussion of abstract general 
 principles. Attention is called to the fact that nearly all of the 
 photographs illustrating these practical examples were necessarily 
 taken by daylight for the reason that the halation of a photo- 
 graphic plate by artificial light shining in the camera would 
 prevent photographs being taken which would show the position, 
 design and equipment of fixtures. The photographs from which 
 the engravings were made are therefore illustrations of lighting 
 'design and arrangement and are not illustrations of lighting 
 effects produced. ' ; : 
 
 ■j Before going into a discussion 'of' these 'practical examples 
 in detail, it will be of advantage to state "some of the general 
 principles which govern the design -of good artificial lighting. 
 
 The difference between light and illumination should be care- 
 fully noted. Light is a cause, illumination an effect. An object 
 on which light falls becomes illuminated. Illumination, there- 
 fore, is the result produced by light. The object of artificial 
 illumination, except where it is purely decorative, is to enable 
 us to see things. An object is well illuminated when it can be 
 easily seen without fatigue or strain on the eyes. As the eye 
 is the organ with which we are enabled to see when there is 
 proper illumination of the objects we are looking at, it is of first 
 importance to consider all artificial illumination with reference 
 to its effect on the eye, and it is well to consider carefully the 
 essentials of good illumination and the things to be avoided, 
 as noted in the chapter on Light and the Eye. 
 ' ' Those desiring information on the" amount of light to be pro- 
 
 147 
 
148 PRACTICAL ILLUMINATION. 
 
 vided for the illtimination of large rooms should consider not 
 only the data given in the following chapters on Practical 
 Examples, but also the information on the Calculation of Illu- 
 mination to be found in Chapter IV. It should be remembered 
 that the small rooms in small and moderate sized residences 
 come in an entirely different class from large rooms as regards 
 the amount of light that must be provided. In residences the 
 reflection from, ceilings and walls plays . a much more important 
 part than in large rooms. If the suggestions made by the 
 authors as to the location and equipment of fixtures for resi- 
 dences are followed, it is an easy matter to increase or decrease 
 the candle power of the lamps to meet the personal tastes of the 
 owners or occupants. In the calculation of illumination for 
 large halls and public buildings it is always well to remember 
 that at some time the building may be decorated in very 
 dark colors, and provision should be raade to take care of the 
 very worst conditions that could arise. 
 
 In presenting these practical examples the authors have at- 
 tempted as far as possible to cover all classes of lighting from. 
 the cheapest to the most expensive. That some of the arrange- 
 ments illustrated are crude is fully realized,. but they represent 
 the best that can be done with a limited amount of money 
 to be expended in lighting fixtures. 
 
 The authors wish to put in a plea for more rational fixture 
 design. The plan in the past has been to first design the fixture 
 and afterwards put the lamp on to accommodate the fixture 
 design. The exact reverse of this should be the practice. The 
 desirable location, angle and equipment of lamps should first 
 be determined and the fixture should be designed or selected 
 accordingly. A fixture should be first of all useful, and after 
 that as ornamental as the pocketbook of the purchaser cares to 
 make it. In this connection, attention is called to the fact that 
 in the great majority of cases for which electric chandeliers are 
 used the time-honored custom of placing lamp sockets at an 
 angle of about 45 degrees is not as desirable as placing the sockets 
 pendant; that is, pointed straight down. With sockets pendant 
 there is opportunity for much greater range in the selection of 
 satisfactory glassware and glassware can be selected which will 
 give a uniform distribution in all directions from a chandelier. 
 With sockets at an angle the choice of glassware to j give good 
 
INTRODUCTION TO PRACTICAL EXAMPLES. 149 
 
 results is very much limited and in some cases, thoroughly good 
 results are impossible. There is, furthermore, the very serious 
 mechanical objection to sockets pointed at an angle, namely, 
 that the weight of shades or globes is such that frequently the 
 shade holders bend and sag and the sockets come to pieces, due 
 to inherent weaknesses of sockets and shade holders. This 
 produces a very bad appearance, to say nothing of its bad effect 
 on the illumination. With pendant sockets practically all 
 these troubles are avoided. Wherever sockets are used at an 
 angle, it is well to select the most rigid shade holders obtainable. 
 
 Another practical point that it is well to remember in plan- 
 ning the illumination of all classes of buildings is that switches 
 should be located in sight of the lamps they are to control, thus 
 saving confusion as well as making it more likely that lamps will 
 not be forgotten and left burning when persons leave the room. 
 A switch out of sight of a lamp is a prolific source of waste and 
 •dissatisfaction. 
 
 Most of the examples shown are of electrically lighted rooms, 
 for the reason that electric light offers the greatest opportunity 
 for proper and improper use, because of the great variety of 
 positions in which it may be located. A sufficient number of 
 examples of gas lighting are given to indicate the way in which 
 gas should properly be used. Gasoline and acetylene gas 
 lighting offer practically the same problems as common gas. 
 
 The authors have endeavored as far as possible to avoid 
 useless discussion of matters which are largely those of individual 
 taste and to confine their discussion to questions of illuminating 
 ■engineering. 
 
 It is always to be remembered that artificial lighting is a 
 •compromise between appearances or artistic effect and economy. 
 It is the province of the engineers and architects who understand 
 these matters to present them to their clients in such a way 
 that clients will understand the effect of different designs on 
 economy. It is then for the client to determine how much 
 economy he is willing to sacrifice for the sake of artistic effect. 
 
 Acknowledgement is due to the Electric City, the Chicago 
 central station publication, for a number of excellent photographs 
 of some of the prominent buildings in that city which are here 
 used. 
 
 In discussing these practical examples frequent references 
 
150 PRACTICAL ILLUMINATION. 
 
 are made to globes, shades and reflectors upon which tests are 
 given in the preceding chapters of this book. In the Appendix 
 will be found the manufacturers' names and trade numbers of 
 the various reflectors, shades and globes tested. 
 
CHAPTER XVI. 
 RESIDENCE LIGHTING. 
 
 Living Rooms and Small Parlors. 
 
 As the living room in the average home is the one most occu- 
 pied during the evening and hence the one about the lighting of 
 which the occupants are most particular, it naturally comes 
 first in a consideration of residence lighting. 
 
 The general lighting of the living rooms of small or moderate, 
 sized residences and the provision for reading lamps in these 
 rooms are so closely connected that they must necessarily be 
 considered together. Reading light is likely to be used con- 
 stantly and has so much to do with the comfort and eyesight 
 of members of the family that it needs earnest consideration. 
 There are three common ways of obtaining light which is suitable 
 for reading. One way is to illuminate the whole room so brightly 
 that reading is easy in any part of it. The cost of lighting a 
 room so brilliantly as this makes it out of the question in the 
 majority of rooms, and for living rooms is not usually considered 
 good artistically. We will, therefore, leave out of consideration 
 this method of securing reading light and consider the second 
 and third methods. The second m.ethod is to use a lamp placed 
 near the reader and fitted with a reflector which will concentrate 
 nearly all the light on the book or paper. If this is done the 
 reading lamp is likely to be of little' value in the general lighting 
 of the room, and other lamps besides the reading lamp will have 
 to be added if the greater part of the room is not to be in com- 
 parative darkness. The third method, which is the one usually 
 to be preferred where the greatest economy is an object, is to 
 use the same lamp both for reading and for the general lighting 
 of the room by equipping the reading lamp with a globe or re- 
 flector which will concentrate a considerable portion of the light 
 within the area in which it is desired to read, and at the same 
 time allow enough Hght to radiate in all directions to give fairly 
 
 151 
 
152 PRACTICAL ILLUMINATION. 
 
 good illumination over the rest of the room. There is a popular 
 impression that the portable table lamp is better adapted to 
 reading than a lamp on a chandelier or bracket. Such table 
 lamps are an inheritance from the days of the candle and oil 
 lamp. With electric lamps or with mantle burners, usually 
 much more satisfactory results can be obtained with a properly 
 equipped reading lamp on the chandelier or on a bracket on one 
 of the side walls. The reason for this is that it is impossible 
 with a table lamp to secure a reflector which will throw as large 
 a proportion of the light where it is needed for reading purposes 
 as can be obtained from the proper reflectors on chandeliers 
 and brackets. Most of the portable stand lamps throw the 
 greater part of the light down on the table around the base of 
 the lamp, so that the readers sitting around the table get only 
 a small percentage of the light. (See test No. 11 in the chapter 
 on Demonstration Room Tests.) Taking any of the shades or 
 reflectors which are at all suited for use on a portable table 
 lamp, it will be seen that this is true. The opal dome used on 
 electric lamps, Fig. 22, or the opal dome and bobesche used on gas 
 mantle burners, Fig. 126, are good examples of this. If the 
 portable stand lamp is used on a library table where the readers 
 place their books on the table each side of the lamp to read its 
 use is permissible, but such is not the usual condition. 
 
 There is also considerable misconception as to the compara- 
 tive distance from the reader of a lamp located six feet above 
 the floor on a chandelier as against a lamp located on a table. 
 The average person who has never measured these distances 
 is under the impression that the lamp on the reading table is 
 much nearer. As a matter of fact the distances are nearly the 
 same. There is really for the majority of cases only about one 
 thing to be said in favor of an electric table lamp for reading as 
 against a properly equipped lamp on a chandelier or bracket. 
 This is, that with the table lamp it is not as likely that a reader 
 will get the regular reflection commonly known as the " glare " 
 from a page of white paper because the light comes so much 
 from one side. In reading underneath a lamp on a chandelier 
 or bracket the reader must turn the page at such an angle that 
 he does not receive this glare from the paper. This is easily 
 done, but many people undoubtedly suffer from this without 
 knowing what is the trouble or taking the pains to find out. 
 
RESIDENCE LIGHTING. 163 
 
 Since this glare is likely to be more pronounced with electric 
 lamps and mantle burners than with kerosene lamps, it is probable 
 this that is partly responsible for the preference that some people 
 have for oil reading lamps, even though electric light is available 
 in the rest of the house. 
 
 The light intensity per square inch of the incandescent lamp 
 filament is much higher than that of the kerosene lamp flame, 
 hence the glare from the paper is more concentrated and intense. 
 The electric lamp can be reduced to an equality with the kero- 
 sene flame in this respect by using a frosted bulb. 
 
 In what has been said as to the advantages of reading lamps 
 placed on chandeliers and brackets as against an electric table 
 lamp, it must be carefully noted that the authors do not advocate 
 trying to read with light from chandeliers which are arranged 
 entirely for the general illumination of the room. Chandeliers 
 should be equipped with reference to using them for reading, 
 if they are to be superior to electric or gas portable table lamps. 
 
 The authors have also neglected here the consideration of a 
 portable reading lamp equipped with an art glass or other fancy 
 shade, the primary use of which is for ornamental purposes, 
 and secondarily for reading. Such shades are taken up later 
 in this chapter. 
 
 The following illustrations are from photographs takefi by 
 daylight to show the arrangements of fixtures, etc., and do not 
 attempt to show the illumination of the room by artificial light. 
 
 In Fig. 204 is shown one of the best arrangements of chan- 
 delier for a small living i;ooxn where a good reading lamp and 
 lamps for the general illumination of the room are provided 
 on the same fixture. On this chandelier a middle socket 
 which points straight down is equipped with a prismatic 
 glass reflector, which concentrates light under the chan- 
 delier for reading purposes and at the same time allows enough 
 light to escape through the reflector to illuminate the walls and 
 ceilings ifioderately so that the single reading lamp is sufficient 
 ordinarily, both for reading and general illumination of the 
 room. This arrangement giving moderately illuminated walls 
 and ceilings will be satisfactory to most people, but there are 
 some who for artistic reasons or because of weak eyes will 
 prefer to have the walls and comers of the room darkened more 
 than this. It is easy to accomplish this darkening by placing 
 
1.54 
 
 I'RACTICA L IIJ.UMIXATU )\. 
 
 an ornamental silk coverins^ over the reflector or by using a 
 decorated opal reflector which obstructs most of the light that 
 it does not reflect. In this way the brightly lighted reading area 
 and darkened room corners that some think so cosy and cheer- 
 ful a feature of table lamps can be obtained. As to whether this 
 
 Fig. 204.- (Jne nf the be.st arrangements for small livin.e; room. 
 
 darkcnniy- of the corners is desiraljlc, is largely a matter of per- 
 sonal taste. There is no doubt such a thing as carrying it too far 
 for the gouil of the e}'e sight. AVhile it is desirable to rest the eye 
 occasionally while readin.g with artificial light by glancing from 
 the biighth' liglited page to parts of the room not so brilliantlv 
 
RESIDENCE ^LIGHTING. 155 
 
 illuminated, it is not good to have this change too violent. 
 Thus, suppose, for example, a person is reading under an opaque 
 concentrating reflector. The light on the reading page will be 
 very strong, and the balance of the room, especially if it has 
 dark rugs and walls, will be almost in darkness. A change from 
 the brightly lighted page to the dark corners of the room is ex- 
 tremely violent and is a strain on the eye, although a change 
 less violent would be a helpful rest to the eye. It is also un- 
 doubtedly true that the eye is more comfortable when receiving 
 a moderate amount of light from all directions, as it does in day- 
 light than when getting all its light from a bright page in a dark 
 room. The particular reflector shown on the center socket in 
 Fig. 204 (the light distribution of which is shown in Fig. 33) 
 concentrates its light over a floor area sufficiently large to per- 
 mit four persons to read comfortably under it. 
 
 Larger or smaller reading circles can be obtained by using 
 more or less concentrating reflectors. If only two persons use 
 this chandelier for reading a deeper and more concentrating 
 reflector covering the lamp more completely is advisable, thus 
 accomplishing the double purpose of getting a stronger light 
 immediately under the chandelier and shading the bare lamp 
 filament from any who may be seated around the room looking 
 toward the chandelier while the center lamp is turned on. 
 For a very large reading circle, the center socket should have 
 a reflector giving a distribution like that shown in Fig. 37 and 
 if that particular reflector be used it will also shade the lamp from 
 those seated around the sides of the room. On special occas- 
 ions or when there are visitors, when the general illumination 
 of the room rather than a concentrated light is wanted, the two 
 lamps on the arms of the chandelier are lighted and the center 
 lamp put out. These chandelier arms are equipped with globes 
 which give a good general distribution of light in all directions 
 below the horizontal similar to that shown in Fig. 80 and also 
 diffuse the glare from the lamps to make the light comfortable 
 to visitors and others who may sit facing it. The same general 
 scheme could manifestly be used with a chandelier having three 
 or four arms carrying lamps for the general illumination of the 
 room. The chandelier as it stands in Fig. 204 is in a room 14 
 feet square with dark walls and rugs and is equipped with eight 
 candle power lamps in the arm sockets and a 16-candle power 
 
156 PRACTICAL ILLUMINATION. 
 
 in the center socket for reading. This candle power can, of 
 course, be increased at the desire of the user. In a light coloi-ed 
 room of this size, lamps of this candle power would be more 
 than sufficient, but may not suit all tastes in a room with dark 
 walls. The centre or reading lamp is not used when general 
 illumination is desired, and vice versa, the lamps for general 
 illumination may usually be turned off when the reading lamp 
 is on. If an especially strong concentrated light is wanted 
 an attachment plug with a 16-inch drop cord can be hung 
 from the middle socket so as to bring the lamp about one foot 
 above the head of a person seated under it. With a good con- 
 centrating type of reflector on the end of the drop cord, this 
 makes an excellent arrangement for sewing on dark goods 
 at night or any work requiring an extra strong light for one per- 
 son. Many modifications of Fig 204 suggest themselves. For 
 the general lighting of the room Class B Holophane globes were 
 chosen in this case, as before explained, because they give the 
 proper distribution of light below the horizontal- and at the 
 same time diffuse the light so that persons sitting around the 
 room in general conversation are not annoyed by the glare 
 from the bare filaments of the lamp. The same purpose could 
 be accomplished in a different way by selecting some type of 
 reflector which would give a good intensity in all directions 
 below the lamp and using lamps with frosted bulbs inside 
 the reflectors. (See Test 15 in the chapter on Demonstration 
 Room Tests). The opal bell with crimped edge, Fig. 23, prob- 
 ably comes the nearest to these requirements as far as distribu- 
 tion of light is concerned. In such a case, the center reading 
 socket should preferably be an opal dome or cone to harmonize. 
 It is important to remember in the general lighting of a living 
 or reception room where a number of people are likely to be 
 sitting around in positions facing the chandelier to avoid any- 
 thing which would allow the undiffused rays from a large portion 
 of the bare filament of a lamp to fall in the eyes. For this 
 reason, whenever open reflectors are used, they should be 
 equipped with frosted bulb lamps. On a chandelier like that 
 shown in Fig. 204, if the reading lamp does not have a frosted 
 bulb, it ought to be turned off when general illumination 
 on special occasions is desired. It will be noticed that the 
 sockets on the arms of the chandelier. Fig. 204, are placed pen- 
 
RESIDEXCE LIGHTING. 157 
 
 dant, or pointing straight down. With a globe or reflector 
 giving a similar distribution of light to that in Fig. 23 and 80 an 
 angle of 10 degrees from the Vertical should not be exceeded 
 with these sockets, otherwise art unnecessary amount of light 
 will be thrown up toward the ceiling and high side walls. 
 
 Another good way to provide for reading and general light- 
 ing in a small living room is to place a hemisphere in the middle 
 of the ceiling for general lighting and for reading to put in;one 
 or two wall' brackets. The hemisphere should be equipped with 
 a reflector inside as recom.ntended in the chapter on clusters and 
 bowls for incandescent lamps. The reading bracket should have 
 its lamp pointed at from- 30 to 45 degrees from the vertical and 
 should be eqtiipjied with, a reflector according to suggestions 
 made for the reading larap socket on the chandelier in Fig. 204. 
 
 In Fig. 205 ,ig -^how-pL'a good example of how not to arrange 
 a reading lamp.-;:- Here the portable table lamp is attached by 
 an extension cotd to one" of the sockets on a chandelier. The 
 table lamp like most' of its kind, throws the better part of its 
 light on the table, while the persons reading are supposed' to 
 sit around the table at some distance from the lamp. The 
 extension cord from the chandelier is a nuisance and un- 
 sightly. This chandelier could be easily converted to afford an 
 efficient reading light by replacing the wastful opal ball 
 on the bottom of the chandelier by a reflector or globe for 
 throwing light down. If a number of persons were to 
 read under the chandelier, an opal dome reflector (Fig. 21) 
 with a frosted lamp could be used to advantage, as it would 
 throw light over a considerable area, supposing that the 
 other glassware on it were changed to harmonize. Re- 
 turning to Fig. 205, the opal bell reflectors on the three arms of 
 the chandelier are excellent for the particular purpose of general 
 illumination in this particular room. When placed at an angle 
 of 45 degrees, as on this chandelier, they throw considerable light 
 to the high side walls and illuminate the pictures which are 
 placed high. They should be used with frosted bulb lamps. In 
 the majority of cases, the arms of this chandelier. Fig. 205, if 
 equipped with opal bell reflectors of this kind should have 
 sockets pointed nearly straight down, otherwise too much 
 light goes to the high side walls On this chandelier, as it is 
 shown, the opal ball on the bottom absorbs a large percentage 
 of the light and does not direct the rays in useful directions. 
 
158 
 
 fRACnCAL llJJWIIXATliiX. 
 
 V]ii. 206 is an exanipk- of a L-ummon arrangement of chandelier 
 for parlor and living rofjm ligliting. Its greatest defect is that 
 no pro\'ision is made f(jr readnig lamps unless they be attached 
 m place of the lamps m one of the sockets, which is an 
 unsightly and unsatisfactory arrangement. Wdiere such a 
 
 Fig 20.1 
 
 -P'Kir arrangement of rearlinc lamp. 
 
 chandelier is used in a lix'ing room, pro\'isioii should be 
 made for a reading lamp mi a side bracket, but it is 
 too frequently the case that it is not made. The chande- 
 lier in Fig. 206, when equipped Avith fluted opal cone re- 
 flectors of tlie t_\-pe shown in Fig. 19, placed at an angle 
 of 45 degrees, will, on a three light chandelier m the middle 
 
RESWEXCE LIGHTIXG. 
 
 159 
 
 of a square room, as in this case give an excellent distribution 
 of light for the general illumination of the room. Frosted 
 bulbs should always be used in them when so placed 
 in order to avoid the glare in the e3^es of the occupants. A 
 general illumination of a room is usually wanted only at times 
 
 Fig. 206. — Common chandelier for small parlor. 
 
 when a number of people are using a room, and it is at such times 
 that some are sure to be sitting facing the light. To throw the 
 light from a strong reflector in the ej-es of the guests, in addition to 
 the undifEused light of the bare incandescent lamp filament 
 is anything but hospitable treatment, and in the course of an 
 
KiO 
 
 /'RACTICAL ILLi'.\]IXATI(>-\ 
 
 ex'ening will cause discomfort, although its cause may not be 
 ay)i)arent to the person effected. 
 
 Fig, 207 shoAVS how good reading light was obtained from a 
 chandelier like that m Fig, 206, A short length of drop cord 
 Avas attached to one of the sockets and a reflector harmonizing 
 
 
 i^^'l -^ 
 
 
 HI '^!^^feSf<d«. 
 
 ■ 
 
 '' ' ■'-'. 
 
 ■r ■ ^ ^^^ 
 
 ^|wgg«gia^^^H 
 
 
 ^■S^aar- ~ ■;— _^^^^^^S|^M^M^HB^^^^fcj|^^^^B 
 
 -■-.!:,.,.;.;■-' 
 
 
 HBlt- ^jjflte^^ w& 
 
 '^';'' -''l^'" 
 
 'Mb "^".'"^gii^ " «'■*■#' 
 
 1 
 
 % .^^- 
 
 ~~ 'S „ii^ 
 
 i ^'^^li 
 
 f "KfijB 
 
 -^^TMH fl 
 
 
 
 WMl 
 
 
 
 t- 1 ' : - 
 
 p-'- J^- '^.^'wPl 
 
 
 l'':v«' 
 
 /- : 
 
 ip)^h 
 
 ~'tt 
 
 
 ¥^ 
 
 
 f 
 
 
 
 ■ '^^^^■- iu 
 
 't 
 
 , ;'■ 1 I'^ffiS*-"^ ' 
 
 
 ',/'..l.L ■■;- , , ■■mil''t- -^ :< ' _, ■ 
 
 
 11^^ 
 
 wm 
 
 i*--'^^- 
 
 
 « 
 
 i^-s* 
 
 1^^^- . 1 
 
 Fig, 2(J/ , —Making the best of bad design twr readm^' jmrpnscs. 
 
 with those on the chandelier was hung pendant, as seen, so 
 that it throAvs its light directly down. This, of course, is a 
 makeshift, hut is intended simph- as a suggestion as to how to 
 make many existing arrangements tolerably satisfactory. The 
 chandelier Avas equipped A\ith etched glass shades. The reading 
 
RESIDENCE LIGH TIXG. 
 
 161 
 
 lamp was placed in a prismatic glass reflector givinjj; a distri- 
 bution similar to that shown m Fig. 35, and the light for read- 
 ing under the chandelier since the change is far better with one 
 16 candle power lamp, than it was formerly with three lamps 
 
 Fu 
 
 -ll'S. — ArrantjeiiiL'Tit fur small r( 
 
 on the chandelier arms, although, of course, the general illumin- 
 ation is not as good. 
 
 Fig. 2().S sho\\'s a good chandelier arrangement for both read- 
 ing and general illumination in a room where, as in a small 
 library tlie light is used almost entirely for reading and persons 
 
ir.2 PRACTICAL ILLC M IXATIOX . 
 
 arc rai'ely .'^eated anN'Avlicre but under the chandelier. This is 
 simph' a two-lamp chandelier with both sockets A'ertical, equipped 
 with the prisnratic reflectors shown in Fig. o'.]. This concentrates 
 enou^'h light for readinsj" in the center of a room while gi\"ing good 
 
 ^ ■ „ /?•■; , 
 
 Fic. L>()0. — Arraii,L:cment \. 
 
 -n, ..luu.u .1 illiiminalii <n. 
 
 enough illummation for ordmary purposes in the far c(jrners. 
 These reflectors gi^'e their concentrated light over a fairly wide 
 area as compared Avith many others of tliis type. If a more 
 concentrated light is desired, an attachment cord and plug can be 
 put on as in Fig. 209. This attachment equipped with a power- 
 
RESIDEXCE UGH TIXG. 
 
 163 
 
 ful concentratiiiLj reflector can be used either hung up on the chan- 
 delier, as shown, or let clown for sewing or Avork recjuiring intense 
 light. This chandelier is not as good an all-around arrange- 
 ment as that of Fig. 204 as it does not provide as well for the 
 general illumination of the room. The reflectors nearly cover 
 
 -MIJ — Poor arraiiLjeiiient tfjr rcadiii.i,' lamp 
 
 the lamp bulbs so that not much of the bare lamp filament 
 is visible to a person seated m one corner of the room. Even 
 this little is objectionable, and it would be better to haA-e it 
 covered by using a frosted bulb lamp. 
 
 Fig. 210 is another good example of how not to do it. The read- 
 
I(i4 
 
 J'KAC'flirAL ILLI'MIXATliKX 
 
 11114- lamp located on the table equipped with a do\vn\vardl_\' pointed 
 Socket and opaline fluted cone shade throws nearly all its lit^^ht on 
 the table. The absurdit\' of tr\-ing to read In' such an arran,t;ement 
 can be seen by consultmt,'' the test tjiA-en m Fit,'. 1',) showing the 
 liL;ht distribution from a somewhat similar reflector. The 
 
 I"lG. 21 f. — Eftcctnc reading light un bracket. 
 
 absurdit}- of the arrangement is further emphasized bv tlie 
 presence on the wall immediately back of the reading tal>le 
 of a l.)racket with a socket pointed at just the riglit angle to 
 throw a good light for reading purposes. 
 
 Fig. 211 shows how the arrangement in Fig. L'lO was eon\-erted 
 
RESIDEXCE Lie 1 1 TIXi, 
 
 1 i5o 
 
 into a very efficient reading arrangement b_\' simply putting a 
 good concentrating type of reflector on tlie wall bracket and 
 moving the large table awa\' so that those desiring to read 
 could sit within the range of the concentrated light from the 
 reflector on the bracket. 
 
 Fig. '2\'2. — Gihm:1 arrant;enient for small nioiii. 
 
 Kg. 212 shows a good wav of lightmg a small parl(.)r with a 
 chandelier ha\'ing sockets at an angle of 45 degrees. This is a 
 similar case to that shown in Fig. 2(l(j. Xo concentrated light 
 was wanted in any portion Avith the possible exception of the 
 piano, and for this, sufficient light is ol.)tainei.l from the chandelier. 
 
] (Ki 
 
 FKAi'TfCAL JIJJ/MIXATIOX. 
 
 The chandelier is equipped Avith three gloljcs, sucli as shown in 
 Fig. 79, these being designed to concentrate much of the Hght 
 in the direction ni wliich the tip of the lamp points. This chan- 
 delier in Fig. 212 -a as pre^'iouslv equi]iped with opaline bell 
 reflectors, which allowed so much of the light to escape towar(.ls 
 the ceilmg tliat thcrc> was not enough at the piano and m the 
 lower part of the room. It is a mistake to equip a chandelier 
 of this kind Avith an\- t\'pe of globe or reflector Avhich will gi\"e 
 a general distribution r)f light at a Avide angle below tlie horizon- 
 
 Dec'ir.'itix'u im-ftic'icnt talilc 'lamp 
 
 tal unless the lani]! is j.ilaeed in a pemlmt or A'ertical position. 
 With laniiis at an angle and a distrrriuting reflector too much 
 light is directed u]i. 
 
 Fig. 213 is a good example of a handsome table lam|) AA'hich is 
 useful jirincipally for decoratixe purposes. It has the same 
 shortcomings as all electric table lamps, nairu'ly, the inability to 
 throw a large proportion of its Hght m directions most needed, 
 unless the light is needed on the table near the lamjj. It does, 
 liOAvcA'cr, add considcrafilv to the artistic a])]:iearance of a room 
 
RESIDEXCE LIGHTIXG 
 
 167 
 
 and the authors do not wish to be understood as condemning 
 artistic ornamental lamps of this kind. When they are used, 
 however, it should be with a full understanding of their ineffi- 
 ciency for all purposes except those mentioned. There are cases 
 where expense is of secondary importance and by equipping 
 decorative lamps with sufficiently high candle power lamps they 
 can be made to do satisfactory useful work. As a general thing 
 an ornamental lamp of this kind might better be provided with a 
 low candle power lamp and used simph' for decorative purposes 
 depending on something else for the useful illumination. Such 
 
 Fig. 214. — E.Kpen.'^ive X'arli.ir with makeshift reading lamp. 
 
 a large lamp as that shown in Fig. 'IV-i might ha\-e the two lamps 
 in it fitted with small opal or prismatic glass reflectors so pointed 
 as to throw the light where needed. 
 
 LiGHTixG OF Large Parlors. 
 
 The parlors illustrated under this head are mainly those of 
 the larger class where expense is a secondary consideration. 
 
 Fig. 214 teaches an interesting lesson in the lighting of a living 
 room. Although an abundance of outlets for electric lamps have 
 been prOA'ided, and expense has not been spared in fixtures, the 
 
ir,s 
 
 I'KACTICAL JLLfMIXATloX 
 
 lamps lia\(_' rxiilrntlv Ijeen placed in opal enclosing; globes which 
 absorljed so much li,i,'ht that the owntT has l.een obliged to resort 
 to a makeshift tor reading purposes and has jjurchased a desk 
 lamp with the t\-i)e of desk shade shown in Fig. 57. This shade, 
 when placed un this t\-pe of portable hxture is \-ery good for 
 one ])erson to read by, but in a room of the character of that 
 shown m Fig. 214 a makeshift like this should nrjt be necessary. 
 A central socket should have l)een pro\-ided on the chandelier 
 on Avhieh a reflector in keeping with the fixture could haxe been 
 placed. It w<iuld haAX- been better, in any e\'cnt, to use sand 
 
 ■^^S 
 
 
 ' 
 
 
 
 
 
 i^^HI 
 
 
 
 i 
 
 
 
 
 
 
 
 ^Hl 
 
 ^ 
 
 1 
 
 
 
 ^' 
 
 
 >4^^B^^H 
 
 
 1 
 
 H 
 
 l^''i^ 
 
 ^j^ 
 
 n 
 
 gl 
 
 
 iIb 
 
 s 
 
 9 
 
 f 
 
 ^sh^spV 
 
 ■1 
 
 ■ 
 
 mm 
 
 
 H 
 
 Mr 
 
 
 i«g!SS'^^^S 
 
 ■ 
 
 aL 
 
 M 
 
 § 
 
 
 
 W 
 
 •I'^fflK 
 
 
 ■B 
 
 i 
 
 Pl^ 
 
 ^m 
 
 i 
 
 1 
 
 Sa^^^ 
 
 9 
 
 I'^i^ LM.'). — Living' ri)( nn with u|>riL,'lil lanip.s. 
 
 blasted or frostec.l lialls on the chandelier instead of opal. If this 
 had been done, it would probably ha\'C l.)een unnecessar}' to 
 provide any special reading lamps unless eeononiA' of operation 
 were an object. 
 
 Fig. LMT) shovv's a large parlor equipi)ed with upright electric 
 lamps in frosted globes. The hxture designers ha\-e been care- 
 ful to place a cujj under each lam]j as they ha\-e under each 
 imitation caii'lle gas jet. an absurditx" in design both as to looks 
 and utilit\'. This tallow cup cuts oft considerable downward 
 light. The use of upright lamps here mav be artistic, but 
 could not be urged as an economical design. Fulh' as 
 
RESIDEXCE LIGH TIXG. 
 
 Iii9 
 
 artistic designs can be obtained m i.lownwardl}' pointed lamp 
 sockets and the results in lighting the room will be much better. 
 Where upright electric lamps are tised, the only way of increas- 
 ing the do\\'nward illumination and at the same time adhermg 
 to the use of enclosing globes is to use a diffusing and directing 
 globe such as shown in Fig. So. If the owner much prefers 
 upright to pendant lamps it is permissible to use them pro\'id- 
 ing thev are used with the full understanding that more light 
 must be pro\'ided than with pendant lamps and that the cost 
 of satisfactorily lighting a room will be greater. 
 
 Fir,. 215. — Livint; roimi with dnnie fi.xture. 
 
 Fig. 21(3 shows a method of liAdng room lighting which is most 
 commonly found in dining rooms. There is an art glass dome 
 in the center for lighting immediately around the table for 
 reading purposes. The general lighting of the room is provided 
 for h\ art glass fixtures hanging from the beams and by the 
 round frosted bulb candle lamps on the the art glass dome. To 
 proA'ide a satisfactory illumination around the table for reading, 
 the art glass dome in such a case should preferably be equipped 
 Avith one frosted lamp pointed straight doAvn, this lamp being 
 from 32 to 50 rated horizontal candle power. Over the lamp 
 inside of the art glass dome should be used an inA-erted boAvl 
 
17LI 
 
 fKACTlCAL iLL'/MlXATh <X 
 
 reflector gniiiy" a ilistriljuti'^n cf light similar to that shown in 
 Fig. 37. Such a rctlectcjr \\ill also have the eftect of obviating 
 the streaks around the sirk- walls vhich would otherwise be 
 present, due to the light s'lining through the frhige on the 
 art glass donte. The art glass ilonie will be more uniformly ■ 
 and pleasantly lighted bi}' one large lamp in a reflector than by 
 se\'eral lamps of lower cauiile pciwer. 
 
 Fig. 217 IS an example ui the hghtmg of a large parlor which 
 could be considerabh- improved. The tirst objection to be 
 
 Fic. 
 
 -LarQc jKirl">r witli tun much ijlare. 
 
 urL^cd against the arrangement sh;:)wn in this figiu'e is the great 
 amount of glare from the lamj) filaments wdiich anA' one sitting in 
 the room receiA'es. Idie oi:>aline shades on these chandeliers are ot 
 little value as reflectors and are not of a shape to act as diffusers 
 and softeners of the light. An opal or ground glass reflector of 
 this shape, used with frosted bulb lamps, would give much better 
 results. The location of the fixtures in this case is very good 
 for a parlor of this kind. Both the chandeliers and the brackets 
 along the wall are placed high, although the chandeliers might 
 AA'ell be placed still higher. In general, it can be said that 
 
RESIDENCE LIGHTIXG. 
 
 171 
 
 the results, as far as the amount of Ught obtained below the 
 horizontal is concerned, would be much better in this case if any 
 good reflector giving a large amount of light lielow the hori- 
 zontal was selected in place of those on the chandeliers at 
 present. 
 
 Fig, 21S is an example of lighting in a living room where ex- 
 pense is a secondarv consideration. The ground glass globes, 
 in which all the lamps are enclosed, diffuse the light so that the 
 arrangement is ^'ery good from a hygienic standpoint. With 
 
 Fig. 218. — l^ivini,' room well and e.Kpensively lighted. 
 
 the arrangement shown, an unnecessary amount of light is 
 being spent on the ceihng. The proper prismatic glass reflectors 
 over the large, spherical bulfis seen on the ceiling fixtures in the 
 center of the room would add to rather than detract from the 
 appearance of the fixture and would greatly increase the amount 
 of light secured around the table and at the piano where the 
 light is most needed. In a dark colored room of this kind it is 
 seldom that there is any too much light thrown down m the 
 lower part of the room for useful purposes. 
 
 Fig 219, the parlor of the Windcmere Hotel, Chicago, is lighted 
 
172 
 
 1 ■/<. 1 CriCA I. ILLUMIX. 1 rii >x. 
 
 with lamps in 1iall i^lcibes siicli as sliown in Fi;^-. S.'!, tlic iii-i7U-i])al 
 [mrpose of AAiiich is to diffuse and soften tlie li,i,dit so as to avoid 
 glare. Gi-ound glass balls could be used for the same inirpose 
 but would not give Cjuitc as good a diffusion. Alore downward 
 distribution of light could be obtained liy certain othe'r tyi^es 
 of globes which modifv tlie distribution of light fron^ tlie l)are 
 lamp more than do these balls. Aside from tins, tlie (Icsign of 
 the lighting of this piarlor is gooi.l. _ :' ' 
 
 1-lL 
 
 L'19.— AVcll diffused liHit. 
 
 Dixix,; Ro(.i.\is. 
 A small dining room is probabh* the easiest rouni m the house 
 to light. All that is necessary for ordinar\' inirpiises is to tlirow 
 a strcjiig light down on the table. The reflection from tlie "white 
 tablecloth to the walls assists greatlv in tlie general lighting of 
 the rfiijm. The main thing, ho\ve\-er, is to haA"c the table well 
 lighted and the balance of tlie room Avill usualh' take care of 
 itself. With electric lighting the simplest and most economical 
 arrangement possible for a small dining-room is a ]dain one 
 lanipi jjendant o\'er tlie center of tlie table, equi]j]jed witli some 
 
KESJPEXCE LIullTIXG 
 
 173 
 
 type of reflector which will thi'OAv a large proportion of its liglit 
 on the table while alloAving some to escape through to illuminate 
 the walls and ceiling. 
 
 Fig. 220 shows a simple arrangement for dining room ligliting-. 
 The lamp is placed six feet aboA-e the floor, v.'hich is slightly 
 lower than the ordinary height of chandelier.- The dining room 
 shown in Fig. 220 is about 15 feet square, and for all ordinarv 
 occasions an eight candle power lamp in the jnismatie glass re-' 
 fleeter shown is sufficient. A lamjj of larger candle power could," 
 of course, be used on special occasions if the table were lengtli.- 
 
 -Snuple but eltk'ient dining ruoni arrangement. 
 
 ened out. The particular prismatic reflector shown m Fig. 220, 
 the distribution of which is shown in Fig. 33 concentrates a 
 little too m.uch light on the center of the table at the expense of 
 the edges. One giving a wider distribution of light as shown in 
 Figs. 21 and 37 would be desirable. This one lamp pendant 
 idea can be elaborated into a ver}- artistic fixture by substituting 
 a chain for the rod and using an art glass shade as in Fig. 221. 
 With such an art glass dome the preferable arrangement is to 
 haA'C one lamp pointing straight down so that a reflector 
 giving a distribution like that shown in I'^ig. 37 can tje placed 
 
]74 
 
 PRACTICAL I LLC MIX A TIOX. 
 
 over it. The whole top of the table will then be evenly Ughted 
 and the fringe will be prevented from making unsightly streaks 
 and shadows on the walls. With a deep dome it is a good plan 
 to have a ground glass bottom on it to diffuse the light. With a 
 large ground glass bottomed dome it is also a fairly good arrange- 
 ment to put several lamps in a cluster, each lamp pointed at an 
 angle and equipped with a small opal or prismatic reflector. On 
 the whole, however, a larger single lamp of equal candle power 
 pointed straight down and equipped with the right reflector 
 
 Fig. 221. — Art glass dome for dining; room. 
 
 will give more satisfactory results. In Fig. 221 small deep 
 reflectors can also be used to increase the downward light from 
 the art glass ceiling fixtures over the corners of the room and 
 will help eliminate the streaks due to the fringe. 
 
 Fig. 222 shows how not to light a dining room with a one lamp 
 pendant. This pendant has the lamp placed at an angle of 45 
 degrees in a bell glass shade. The shade is not of the kind 
 which greatly influences the distribution of light, and what 
 influence it does have on the Hght is of no ]n-actical benefit 
 
RESIDENCE LIGHTIXG. 
 
 175 
 
 because of the angle at which the lamp is mounted. With the 
 arrangement shown in Fig. 222 a 16-candle power lamp will not 
 
 w- 
 
 Fig. 2212. — Simple but inefficient arrangement for dining room. 
 
 give as good results in any given room as an eight candle power 
 lamp with the pendant shown in Fig. 220. 
 
 Fig 223 is another example of how not to light a dining room. 
 
17G 
 
 PRACTICAL ILLUMIKATION. 
 
 It is a case of getting a large amount of light everywhere but on 
 the table, where it is'most needed. The lamps are placed at 
 
 Fig. L'23. — Arrangement to be avoided. 
 
 such an angle that the light of the bare filament strikes directly 
 in the eyes of those seated at the ends of the table. 
 
 Fig. 224 shows an excellent way of lighting a dining room 
 
RESIDENCE LIGHTING. 
 
 177 
 
 where something a Httle more elaborate than Fig. 220 is desired 
 and where the room is too large to be lighted well with a single 
 pendant. In Fig 224 the ordinary lighting of the dining room 
 table is accomplished with the lamp within the opal dome re- 
 flector located on the bottom of the chandelier. This opal dome 
 (See Fig. 21) gives an excellent distribution of light for this pur- 
 pose and will throw light over a considerable angle so that a 
 
 Fig. 224. — Good dining room arrangement. 
 
 large table can be Ughted. If additional light is wanted around 
 the sides of the room on special occasions, the three lamps 
 on the arms of the chandeUer can be turned on. These are 
 equipped with opal bell reflectors, (See Fig. 23) which, when 
 placed at this angle, throw considerable light to the high side 
 walls, Ughting up the plate racks and decorations of the room. 
 Since the occupants of the dining-room are usually seated 
 
178 
 
 PRA CTICA L ILL UMINA TIOX . 
 
 around the table, frosted bulbs are not as necessary in these 
 reflectors as they would be were the occupants of the room 
 seated where the light from the chandelier would come more 
 nearly in the line of ordinary vision, but they are desirable 
 nevertheless as they do away with glare and give a softer, better 
 
 Fig. 225. — Poor lighting of hotel private dining room. 
 
 diffused light than the clear bulb lamps and with but small loss 
 by absorption. 
 
 Fig. 225, which is a view of one of the private dining rooms 
 in Hotel Stratford, Chicago is a good example of how not to 
 light such a room. A cluster of nine bare lamps is placed near 
 the ceiling without reflectors of any kind. The ceiling acts 
 
A, 
 
 RESIDENCE LIGHTIXG. 
 
 179 
 
 somewhat as a reflector but a very inefficient one. It is safe to 
 say that the table and the lower part of the room in general would 
 be as well lighted with half the number or candle power 
 of lamps were some simple and ornamental form of reflector 
 used. With a chandelier placed as high as this, the high side 
 walls, upon which there are elaborate decorations, would re- 
 ceive sufficient light were any translucent reflectors used, while 
 the table and the lower part of the room would receive 
 very much more light and the artistic effects would be much 
 
 Fig. 226. — Dining table dome with upright lights. 
 
 improved. In this case frosted bulb lamps are not as necessary 
 to pleasant illumination as where they are placed lower over a 
 table. 
 
 Fig. 226 shows another design of art glass dome for dining 
 room lighting. The remarks made about Fig. 221 as to the 
 proper equipment of such art glass domes also apply here. 
 In this case, the general lighting of the room is provided for by 
 two frosted upright lamps, mounted on the dome. These lamps 
 are shielded by the dome from the eyes of persons seated at the 
 
180 
 
 PRACTICAL ILLUMINATION. 
 
 table, which is an important point where the general lighting 
 of the room is provided for by lamps placed so low. Another 
 absurdity in fixture design is seen here. Round bulb lamps 
 have been placed on a candle fixture on the dome. 
 
 Fig. 227 shows another design of dining room lighting with 
 an art glass dome. In this case, the general lighting is accom- 
 plished by ceiling lamps in frosted globes and by frosted bracket 
 lamps on each side of the fireplace. The design is very good. 
 The art glass dome .shields the eyes of a person seated at the 
 table from the rays of the opposite ceiling light which might 
 
 Fig. 227. — Artistic dining room lighting. 
 
 otherwise be objectionable. The bracket lamps on each side 
 of the fireplace are permissible in a dining room fireplace for 
 ornamental purposes. In a living room, where people are likely 
 to sit in front of the fireplace, lamps on each side of the fireplace 
 are not to be recoinmended except where intended entirely for 
 ornament and not for use when there is a fire, because they 
 detract from the cheerful effect of the fire and are likely to be 
 harmful to the eyes. To make a fireplace thoroughly cheerful 
 the surroundings of the fireplace should not be lighted too 
 much. 
 
RESIDEXCE LIGHT IX G. 
 
 181 
 
 Fig. 228 shows a dining room well lighted by gas. The cen- 
 ter arrangement is equipped with an argand burner, giving a 
 soft mellow light, while the opal dome acts as a good reflector. 
 The other lamps on the chandelier can be used when necessary, 
 for general illumination of the room. The argand burner might 
 be equipped with a frosted or Holophane bobesche to good 
 advantage so as to protect the eyes. 
 
 It should be noticed in such cases that where a center light 
 is used, along with other lights it is not desirable to mix the 
 color effects of mantle burners with open gas. The argand 
 
 Fig. 228. — Good dining table lighting with gas. 
 
 burner, having practically the same color as an open flame, 
 should be used in preference. 
 
 The examples given of dining room lighting so far have nearly 
 all been of rooms with ceiHng outlets. Side wall brackets for 
 general lighting are employed in addition to the center lamp for 
 the table in some modern dining rooms with good artistic effect. 
 The lamps on such brackets since they are sure to be in the 
 line of vision of some one sitting at the table must be well 
 provided with diffusing globes to avoid discomfort to the persons 
 facing them. 
 
182 PRACTICAL ILLUMINATION. 
 
 Bedrooms. 
 
 The lighting of bedrooms, especially in small residences where 
 first cost must be kept low, is the most difficult problem to be 
 met with in residence lighting. To light a bedroom properly, 
 lamps are needed at various times in a -number of places, yet 
 it is seldom that the owner is willing to pay for as many outlets 
 as would be necessarj^ to provide for all contingencies. Before 
 considering specific cases, let us see just what lamps are needed 
 in the ordinary bedroom. The first essential is usually light 
 at the dresser, so that a person standing in front of the dresser 
 can have the use of the mirror without annoying shadows. To 
 illuminate a dresser mirror perfectly three lamps are necessary 
 although they are seldom provided. One of these three lamps 
 should be directly above the mirror, and the other two should 
 be located on each side about 4.5 to 5 feet above the floor. 
 
 For the side brackets on each side of a dresser, upright lamps 
 are generally preferable for the reason that if pendants are used, 
 considerable light is thrown down on the floor, where it is 
 practically of no value, little of it being reflected. If uprights 
 are used, the light thrown upwards strikes the ceiling and walls, 
 adding to the cheerfulness and general illumination of the en- 
 tire room. 
 
 The side lamps are needed to light the sides of the face of a 
 person standing in front of the dresser and should be low enough 
 so that there will be no shadow under the chin of a man shaving. 
 These side lamps, if placed at this height, however, will not be 
 high enough to light the top of a person's head so that there will 
 be a shadow at that point, which must be obliterated by a lamp 
 placed above the mirror for use in hair dressing. Thei'e should 
 manifestly be some general illumination in a perfectly lighted 
 bed roora, for which purpose a chandelier or pendant, or a 
 lamp at the ceiling in the middle of the room can be provided. In 
 addition to this, it is desirable to provide a bracket with a chain 
 pull socket immediately over the head of the bed for reading 
 purposes and where it will be within easy reach of a person de- 
 siring to turn on the light in the night before getting out of bed. 
 Besides this, if the closet is large there should be a closet lamp^ 
 although this is hardly necessary in small closets. Thus there 
 should be five or six outlets in a perfectly lighted bedroom,, 
 which is more than in most of the other rooms in the house and 
 
RESIDENCE LIGHTING. 183 
 
 would be considered out of the question by most owners of small 
 residences. Some compromise must, therefore, be adopted in 
 most cases which will cut down the number of outlets or make 
 one or two outlets available for more purposes. The use of 
 portable brackets as described later is one of these compromises 
 which, although undesirable from the standpoint of appearance, 
 is nevertheless too practical and useful to be left out of con- 
 sideration, especially in the wiring of old houses where one out- 
 let in the center of a bedroom inust frequently suffice for all 
 purposes. 
 
 Another trouble usually met with in planning the lighting of 
 a bedroom is the uncertainty as to the location of various pieces 
 of furniture. It sometimes happens that there is only one good 
 place for the dresser in a room, and in such a case, whoever 
 plans the illumination can with safety and certainty place 
 the dresser lamps with the knowledge that they will not be in 
 the way of any future changes of furniture location the house- 
 wife may wish to make. 
 
 On the other hand it is sometimes difficult to locate brackets 
 where they will not under some circumstances prove incon- 
 venient and in the way of furniture. Some suggestions as to 
 overcoming this difficulty are given later. 
 
 Where the general lighting of a bedroom is to be provided for 
 by a central fixture, the problem is not materially different from 
 the general lighting of living rooms and parlors, and for sugges- 
 tions on this point the reader is referred to the previous remarks 
 on that subject. The points that need special consideration 
 in bedroom lighting here are, therefore, the lighting of dressers 
 and the provision of bracket lamps over the beds or at other 
 desired points. 
 
 As the lighting equipment of a bedroom depends so much on 
 the amount that the owner is willing to spend in order to secure 
 good lighting the simplest and cheapest plans will be taken up 
 first and then the more elaborate ones. 
 
 The authors fully realize that some of the cheaper arrange- 
 ments proposed are crude and of not very pleasing appearance 
 but they are not any more so than some of the existing Ughting 
 arrangements in bed rooms of the class for which they are pro- 
 posed. 
 
 In a bedroom where there is but one outlet the most satisfac- 
 
1S4 
 
 PRACTICAL ILLUMIXATIOX. 
 
 tory plan is tisually to run a portable lamp cord from this outlet 
 to a portable wall bracket. Such portable wall brackets can be 
 
 Fig. 229. — Portable bracket over dresser. 
 
 hung on ornamental brass hooks at any points they are wanted, 
 and when supplied from a central ceiling outlet can be changed 
 at a moments notice from one part of the room to another. 
 
'RESJJPJENf_E LIGHTIXG. 
 
 185 
 
 Such a bracket is shown in Fig. 229, used as a dresser lamp, 
 hung directly over the center of the mirror. It is equipped with 
 a prismatic glass reflector, such as shown in Fig. 31, which throws 
 a powerful light down on to the head of a person standing in front 
 of the dresser, hence this arrangement is excellent for hair dress- 
 ing, and in fact, almost all purposes for which a dresser lamp 
 is needed except shaving. It can be made to provide the proper 
 light for shaving iii a manner mentioned later. i 
 
 '-^v-<HKi;- 
 
 Fig. 230. — Portable bracket as reading light over bed. 
 
 Fig. 230 shows the use of the portable wall bracket as a read- 
 ing lamp over the head of a bed. In this case the key socket 
 at the attachment plug is within easy reach for turning on the 
 light in the night. With the bracket in this position and a 
 concentrating type of prismatic glass reflector, a four candle 
 power lamp is sufficient for reading purposes if the bracket is to 
 be used for that purpose alone. 
 
186 PRACTICAL ILLUMINATION. 
 
 If the ceiling of ttie room is of light color, this particular 
 kind of bracket can be used to give a fairly good, evenly- 
 diffused light over the entire room by simply turning it up as 
 shown in Fig. 231. By means of a thumb screw the arm of the 
 
 Fig. 2.31. — Portable bracket for general lighting of room. 
 
 bracket can be loosened and turned around at any angle. It 
 can also be extended several inches to adapt it to different 
 requirements. By turning the lamp towards the ceiling, the 
 ceiling acts as a reflector to light the whole room by diffuse re- 
 
RESIDENCE LIGHTING. 
 
 187 
 
 flection and the light is soft and pleasing without being intense 
 at any one point. 
 
 Fig. 232 shows how this bracket can be made to serve for 
 shaving by setting it on the dresser in front of the mirror so as 
 to throw the Hght into the mirror. The reflected light from the 
 mirror falls on the face. 
 
 Where a portable fixture of this kind is used in a bedroom for 
 furnishing most of the light and the owner does not wish to pay 
 for more than one outlet, a good arrangement is to place a three 
 or four lamp wireless cluster at the ceiling in the center of the 
 
 Fig. 232. — Portable bracket for shaving. 
 
 room. Attachment plugs for various other lighting and heat- 
 ing devices can then be inserted later in this wireless cluster 
 or a lamp on a drop cord in the center of the room can be secured 
 by attachment in a similar manner. Of course these are in the 
 nature of makeshifts, but, as said before, bedroom, Ughting 
 especially in houses already wired, is very much of a compromise 
 between what is desirable and what is feasible. 
 
 Fig. 233 shows an arrangement of lamps mounted directly 
 on a dresser. On a dresser especially designed with reference to 
 having electric lamps mounted upon it, the arrangement could 
 
188 
 
 PRACTICAL ILLUMINA TION . 
 
 be made very artistic. On the dresser shown, there are two 
 wall sockets mounted on each end of a piece of standard hard- 
 wood moulding fastened against the back of the mirror. The 
 unsightly ends of the moulding are covered with round blocks 
 of wood such as are used in mounting wall sockets, and the 
 sockets are mounted on these blocks, the wood being finished 
 to correspond with the dresser. The flexible cord and attach- 
 
 FiG. 233. — Dresser with lights upon it. 
 
 ment plug which connect with the wires in the moulding make 
 it easy to connect up these dresser lamps to any socket outlet 
 in the room. Such an outlet can either be located along the 
 wall or at the middle of the room. In the latter case a long 
 lamp cord is necessary. This arrangement has the advantage 
 of always having the lamps on the dresser wherever the dresser 
 may be moved. The lamps as shown are excellent for shaving 
 
RESIDENCE LIGHTING. 189 
 
 purposes, but must be supplemented by some higher lamp for 
 hair dressing, as these lamps are too low to illuminate the top 
 of the head. 
 
 A diffusing and directing globe giving a distribution similar 
 to Fig. 80 is placed on these side lamps for two reasons: In the 
 first place, these lamps are so nearly in the ordinar^r line of vi- 
 sion that they would have a decided blinding effect if they 
 were not provided with something to diffuse and reduce the 
 intensity of the light. As far as diffusion is concerned, the 
 same results could be secured to a less degree by the use of frosted 
 bulb lamps or by sand blasted enclosing globes. The second 
 reason for using these globes in this place was that, in addition 
 to diffusing the light, they direct the rays. The type of globe 
 used as seen from Fig. 80, gives such a distribution of light that 
 when placed as these lamps are on the side of the dresser mirror, 
 much more light will be thrown on the face of a person standing 
 in front of the mirror than with any of the other diffusing globes 
 mentioned. The arrangements shown in Figs. 229 and 233, 
 combined make an almost ideal dresser lighting arrangement be- 
 cause it furnishes light all around the face and on top of the head. 
 
 We now come to the class of bedrooms where cost does 
 not prevent the placing of a number of permanent fixtures. 
 
 Fig. 234 is a good example of the way bracket lamps are 
 frequently placed in a small bedroom without any regards to the 
 needs of the occupants and at places where they are most incon- 
 venient. In this case the architect appears to have placed the 
 outlets symmetrically one on each side of the windows without 
 considering anything else. Neither bracket is where it is needed. 
 It is impossible to make such lighting arrangements satisfactory 
 in a bedroom. It would have been much better to have placed 
 these two brackets on one side of the room at such a distance 
 apart that a dresser could have been placed between them. 
 The lamps should be frosted if used without globes or else 
 should be covered by diffusing globes. 
 
 For large bedrooms where expense does not prevent, the 
 arrangement of dresser lamps shown in Fig. 235 is desirable. 
 This dresser has a lamp at each side as well as one directly over 
 the middle of the mirror. This is the ideal arrangement as 
 far as the location of the fixtures is concerned, although there 
 is some room for criticism as to the shades used on the brackets 
 
190 
 
 PRACTICAL ILLUMIXATIO\\ 
 
 each side of the dresser. "IVith brackets in such positions, 
 where open shades of any kind are used, the lamps should 
 have frosted bulbs or be enclosed in diffusing globes. As regards 
 the lamp on the pendant directly over the dresser, it would be 
 well, (if the neck of the globe was large enough; four inches 
 or more), to put inside the ornamental globe some small reflector 
 which would be concealed by the globe but which would assist 
 greatly in throwing the light down on the head of the person 
 using the dresser. As said before, this is a case where expense is 
 of secondary importance, otherwise a different equipment both 
 
 Fig. 234. — Inconveniently placed bracket lamps. 
 
 of the center chandelier and of the pendant over the dresser 
 might be advisable. The ornamental opal globes on the center 
 chandelier absorb a large amount of light and do not distribute 
 it very well in downward directions where it is most needed. They 
 diffuse the light well, however, and so are good from every 
 standpoint except that of low cost of operation for a given 
 amount of light. 
 
 As the arrangement shown in Fig. 235 calls for more fixtures 
 and outlets than many builders care to pay for, plans calling 
 for fewer outlets will be discussed. 
 
 The question is, which of the three outlets around the dresser 
 
RESIDEXCE LIGHTING. 
 
 191 
 
 can be most easily dispensed with. The compromise usually 
 adopted is to omit the bracket or pendant over the mirror and 
 have the side brackets only. If this is done the lamps on the 
 brackets should be placed from five feet, six inches to five feet, 
 nine inches, above the floor. This will be neither high enough 
 to light the top of the head of a tall woman or under the chin 
 of a short man but will answer the requirements of most persons 
 of average height. Part of the light for the top of the head can 
 
 Fig. 235. — Bedroom fixtures properly located. 
 
 be obtained from a lamp near the ceiling in the center of the 
 room. 
 
 Another way in which the provision of three outlets for the 
 lighting of the dresser mirrors can be obviated is by placing one 
 bracket lamp directly over the mirror, and if the room is to 
 be used by gentlemen who wish to shave, providing a portable, 
 electrically lighted shaving mirror with frosted bulb lamps 
 mounted on each side of it. Such mirrors can be purchased 
 ready made in the open market or made to order to suit the 
 
192 
 
 PRACTICAL ILLUMINATION. 
 
 purchaser. They are provided with attachment plug and cord for 
 attachment to any socket and can be hung anywhere in a bed- 
 room. The same plan can be used if there are brackets placed 
 each side of the dresser but too high for shaving. 
 
 Where but a single bracket lamp is provided over the dresser 
 mirror, it may be located too high to be reached with comfort. 
 
 Fig. 236. — Hotel Astor bedroom. 
 
 In^fact, in a new house it should be located high to provide 
 for the possibility of dressers with high mirrors being placed 
 under it. For this reason a chain pull socket should be used 
 on the bracket unless it is to be controlled by a wall switch. 
 
 Fig. 236 shows one of the bedrooms in Hotel Astor, New York. 
 Here the dresser is lighted by two pendant 16-candle power 
 lamps placed on a bracket above the mirror. These lamps are 
 
RESIDEXCE LIGHTING. 193 
 
 equipped with prismatic reflectors of a concentrating type and 
 throw a very powerful light down on the person standing in 
 front of the dresser and some light is reflected from the cloth 
 on the dresser to the chin so that shaving is possible, though 
 not altogether easy because of the shadow under the chin which 
 is relatively much darker than the face. The scheme is hardly 
 practicable where less than two 16 candle power lamps are used, 
 as that amount is needed to secure sufficient reflected light from 
 the white cloth on the dresser to light under the chin. With 
 lamps on each side of a mirror, lamps as low as four candle 
 power can be used with excellent results. In Fig. 236 the 
 general lighting of the bedroom is beautifully accomplished by 
 three 16 candle power lamps at the ceiling with prismatic reflec- 
 tors similiar to that shown in Fig. 32. These lamps have 
 frosted bulbs and the room is well illuminated to its farthest 
 corners so that reading can be done in any place in the room, 
 although the light is much the strongest under the chandelier. 
 The number and candle power of lamps in this bedroom is larger 
 than would usually be demanded in a private house, since in a 
 hotel of this class, it is necessary to provide ample light to suit 
 the most exacting rather than to attempt economy. The chan- 
 delier arrangement, shown in Fig. 236, would not be satis- 
 factory for general lighting without good reflectors on the laraps, 
 and in fact, was not satisfactory before efficient reflectors were 
 used for throwing the light down where it was needed. The 
 light distribution of the bell formerly used was shown in Fig. 92. 
 Fig. 237, shows a peculiar lighting arrangement in one of the 
 bedrooms at the Claypool Hotel, Indianapolis. The arrange- 
 ment is remarkably unsatisfactory and inefficient. At the far 
 end of the room is a two light fixture suspended from the wall 
 outlet. This in itself, is bad for mechanical reasons, as it is 
 likely to sag. The fixture is of very little use where it is placed, 
 as it is almost over a radiator, so that it cannot be used to advan- 
 tage either for the dresser or for reading. The two lamps used 
 on this fixture might have been placed to much better advantage 
 on each side of the space now occupied by the picture in which 
 case the dresser should have been between them. The three- 
 lamp chandelier in the middle of the room, in order to give the 
 best general illumination in the room, should be higher and 
 equipped with reflectors. The chandeliers in this room have 
 
194 
 
 PRACTICAL ILLUMINATION. 
 
 roughed white shades, which are excellent for diffusing the light 
 but they do not cover the filaments and they let entirely too 
 
 Fig. 237. — Claypool Hotel bed room. 
 
 much light escape upwards. The room is not well lighted in 
 all parts, even though the number of lamps is sufficient to more 
 than accompHsh this. It will be instructive to compare this 
 
RESIDENCE LIGHTING. 
 
 195 
 
 room with that shown in Fig. 236, as the same amount of power 
 is used in both cases and the room shown in Fig. 236 is for 
 practical purposes much better hghted. 
 
 Fig. 238 shows one of the bedrooms in the Brown Palace Hotel, 
 Denver. Although it is not perhaps apparent from the photo- 
 graph, the room is very long and narrow. The chandelier is 
 equipped with corrugated, glass, bell shades, which act to some 
 extent as prismatic reflectors, although very inefficient ones. The 
 notable mistake made in this case is that the lamps are pointed 
 cross ways of a long narrow room, so that the maximum light 
 
 Fig. 238. — Brown Palace Hotel bed room. 
 
 is thrown on the side walls opposite the chandelier, leaving the 
 ends of the room in comparative darkness. This is a common 
 mistake, as there seems to be some sort of a tradition among 
 architects and electric wiremen that a two lamp chandelier 
 should be pointed crosswise instead of lengthwise of a room or 
 hall. If the lamps were bare and not equipped with reflectors 
 this would be the best arrangement, but when reflectors are 
 used, lamps should be pointed lengthwise of a room or hall. In 
 this bedroom no provision is made for dresser hghting so that 
 the dresser must be moved into the middle of the room if the 
 user thereof is not to stand in his own shadow. 
 
196 PRACTICAL ILLUMINATION. 
 
 Fig. 239 shows an ordinary bedroom with a three-lamp chan- 
 deher, a side bracket next to the dresser and a bracket with a 
 mantle burner in the comer. This seems to be a well lighted 
 room. The chandelier gives good general illumination and 
 the bracket next to the dresser is placed at the proper height 
 to throw light on the person standing in front of the mirror. 
 It would hardly be advisable to use a mantle burner on this 
 bracket unless a dense diffusing globe were used. Another 
 bracket should have been placed on the other side of the dresser. 
 The mantle burner in the corner which is used for reading, 
 
 Fig. 239. — Gas lighted bed room. 
 
 is well placed and well equipped, except that a bobesche might 
 be used to good advantage. This is not necessary, however, 
 in case the person sits with his back to the lamp, or the lamp 
 is directly above the table as in this instance. 
 
 Kitchen Lighting. 
 The kitchen requires the best general illumination of any 
 room in the house. In most of the rooms in the house a con- 
 centrated light at one or two places in a room is sufficient for 
 ordinary uses. In the kitchen there is usually a sink, a table, 
 
RESIDENCE LIGHTING. 
 
 197 
 
 and a stove, upon all of which work is done, which requires 
 good illumination. It is too frequently the case that the kitchen is 
 the last room in the house to be considered in planning the il- 
 lumination. If the dishes are to be clean and the food served 
 free from dirt, there must be light enough in the kitchen to show 
 the condition of dishes and food. A poor light will cover up 
 a multitude of defects in dishwashing and general cleanliness. 
 
 Fig. 240. — Common faulty kitchen lighting. 
 
 Fig 240 shows a common faulty method of kitchen lighting, 
 which consists simply of hanging a bare lamp on a drop cord 
 about six feet from the floor. Now, from Fig. 16 it is seen 
 that a bare lamp will give an unnecessary amount of light to 
 the side walls and ceiling and not enough down on the stove, 
 sink and table with such an arrangement. Further, from 
 
198 
 
 PRACTICAL ILLUMIXATIOX. 
 
 what has gone before, it is evident that a bare lamp Hke this, 
 hung in the iniddle of the kitchen, has a certain bHnding effect 
 on persons attempting to see across the room, and hence detracts 
 from the efficiency of the illumination. Further than this it 
 is objectionable because, with table, stove, and sink located 
 around the sides of the room, the users must always stand in 
 their own light or else stand very much to one side. 
 
 Fig. 241. — Good position for kitchen light. 
 
 Where a single lamp must serve for the lighting of the whole 
 kitchen, the best plan is to put it at the ceiling in the center as 
 in Fig. 241, with a reflector which will throw light over all of 
 the lower part of the room. A reflector giving the maximum 
 light at an angle of about 40 to 50 degrees below the horizontal 
 is the one to choose for such a location, in a kitchen 10 feet square 
 
RESIDENCE LIGHTING. 199 
 
 or less. Reflectors shown in Figs. 19 and 37, come in this 
 class. In a larger kitchen this maximum light should be 35 to 
 40 degrees below the horizontal as in Figs. 23 and 27. 
 
 Another better arrangement for kitchen lighting than a single 
 central ceiling light would be to place bracket or pendant lamps 
 well up over the sink, the table, and the stove, using them with 
 downwardly pointed sockets and reflectors of opal or prismatic 
 glass. With three four candle power bracket lamps of this kind 
 the power consumption will be about the same as with one 16- 
 candle power lamp and by concentrating light on the three 
 points where the most work is done will give a better light for 
 working, although the general illumination of the room will not 
 be as good. It sometimes happens that cupboard doors and 
 windows interfere with such bracket arrangement in kitchens. 
 A bracket lamp over a stove will require frequent cleaning on 
 account of the steam and grease rising from the cooking. For 
 this reason a Holophane globe should never be used in such 
 places ; either an opal or prismatic reflector with a smooth inside 
 surface being preferable. 
 
 In a small pantry the best thing that can be selected is a 
 lamp in the middle in a diffusing globe hung about six feet from 
 the floor, because what is needed here mainh^ is good light on 
 the shelves. In a large or butler's pantry, the conditions are simi- 
 lar to those in the kitchen and a reflector to direct the light down 
 should be used. 
 
 Porch Lighting. 
 
 The requirements for porch lighting are evidently to light the 
 floor of the porch and the steps to it as well as to throw light 
 on a person standing on the porch. This means, of course, a 
 strong downward light. The ordinary incandescent lamp as 
 was seen in Fig. 16, throws less than one-half its rated candle 
 power directly downward so that if this type of lamp is to 
 be used, some means should be adopted to increase the 
 downward light. For artistic reasons a closed form of globe 
 is usually chosen, as in Fig. 242, instead of a reflector, although 
 a reflector is by far the most economical thing that can be used. 
 If a good prismatic reflector is employed, a four candle power lamp 
 is usually sufflcient, although an eight candle power lamp is 
 better. Another plan sometirnes adopted is to use a turn- 
 down lamp having the one candle-power filament turned on 
 
200 PRACTICAL ILLUMINATION. 
 
 for the ordinary everyday requirements of lighting the porch 
 and steps and using the 16-candle power filament on special 
 occasions. Some favor the use of an opal or ground glass ball 
 on a porch because the house number can be painted on the 
 ball, This is of little value because there is, or should be, a 
 number on or near the door readable from the street b)^ day- 
 
 FiG. 242. — Inefficient porch light. 
 
 light. This number can be read by night if the lighting of the 
 porch is what it should be. The number that can be placed on 
 any ordinary size of opal globe is too small to be read from the 
 street. 
 
 One of the best forms of porch lighting is by hemispheres placed 
 on the ceiling, as in Fig. 243. A single lamp should be mounted 
 
RESIDENCE LIGHTING. 
 
 201 
 
 inside in the center pointing straight downward and provided 
 with a reflector inside of the hemisphere as recommended in 
 the chapter on clusters and bowls. When used with a reflector 
 almost any kind of diffusing hemisphere can be used, ground 
 glass being preferable to opal and opaline because of its higher 
 efficiency and more artistic appearance. If the lamp cannot 
 be pointed down a Holophane hemisphere will assist greatly 
 in directing the hght down as shown in Fig. 159. An effective 
 
 Fig. 243. — Good plan for long porch. 
 
 and practical arrangement for porch lighting when using hemis- 
 pheres, where expense is of secondary consideration is to place 
 one lamp in the center for common use and others surrounding 
 it and controlled by a separate switch. 
 
 When a ball effect is desired, one should avoid the typical 
 example shown in Fig. 242 where most of the light is thrown 
 sideways and on the ceihng and the minimum downward where 
 it is wanted. By the use of a ground glass ball with a four inch 
 holder permitting the use of an 8 or 16 candle power lamp with 
 
202 PRACTICAL ILLUMINATION. 
 
 a small prismatic reflector hanging directly on the lamp inside 
 the globe, a pleasing and efficient arrangement can be obtained. 
 It is also possible to use in such cases a prismatic reflecting ball 
 as shown in Fig. 88 or globes of the kind shown in Figs. 78 and 
 82. It is not possible to obtain as great efficienc}^ using an 
 
 Fig. 244. — Good bath room mirror lighting. 
 
 enclosing globe even with a reflector inside of it as can be 
 obtained with a lamp in a large reflector without the enclosing 
 globe. 
 
 Brackets are often effectively and artistically employed in 
 porch lighting. Such cases present practically the same prob- 
 lems as ordinary bracket lighting which has been treated. 
 
RESIDENCE LIGHTING. 203 
 
 Fig. 243 shows a good plan for lighting a large porch. One 
 hemisphere with a lamp and reflector inside of it is placed at the 
 end of the porch most frequently used and another hemisphere 
 for occasional use at the other end of the porch, controlled 
 by a separate switch. If only one lamp is placed on a long 
 porch where the steps and the door are at one end, it is usually 
 better to put the lamp over the end most commonly used or 
 at the front door rather than in the middle. 
 
 Bath Room Lighting. 
 The lighting of the bathroom depends somewhat on the amount 
 . of use that is to be made of the bath room mirror. In a small 
 bathroom where the mirror is not used much, a lamp in the cen- 
 ter on a pendant about 6.5 feet above the floor, and equipped 
 with a reflector which gives good general distribution below 
 the horizontal, will answer very well. A preferable arrange- 
 ment is to place it at the ceiling and control it with a wall 
 switch. A lamp of four or eight candle-power is usually suf- 
 ficient. Where mirror lighting is important, the arrangement 
 shown in Fig. 244 is good. In this figure, the upright lamps on 
 brackets on each side of the mirror are equipped with Class B 
 Holophane globes, which diffuse the light and also direct most 
 of it below the horizontal. Because of the fact that bath rooms 
 have usually light walls and ceilings, reflectors do not play 
 as important a part here as in dark colored rooms, although, 
 in the example given in Fig. 244 there is so much dark wood 
 work in the vicinity of the brackets that the per cent of reflection 
 is very low. 
 
 Hall Lighting. 
 The lighting of a long hall in a residence from one fixture can 
 best be accomphshed by placing a two-lamp fixture with lamps 
 near the ceiUng at the middle, placing frosted tip lamps in concen- 
 trating reflectors and pointing them so that the maximum intensity 
 of light is in the direction of the floor at the ends of the hall as 
 in Fig. 245. This accomplishes the lighting with one outlet. For 
 longer halls two or more outlets are advisable. In the latter 
 case the lamps ma}^ be placed on pendants either in diffusing 
 globes or in reflectors giving a general distribution of light below 
 the horizontal, such as Figs. 21, 23 and 80. 
 
204 PRACTICAL ILLUMINATION. 
 
 The use of a turn-down lamp in a hall is frequently desirable 
 in which case the cords for turning up and down must be within 
 reach of an ordinary adult from the floor, or a wall switch can 
 be provided. The use of side brackets for wall lighting is also 
 artistic and permissible. Lamps in side brackets should be 
 always frosted or enclosed in diffusing globes. The main thing 
 in hall lighting is a general moderate illumination without con- 
 centration at any one point, and without any blinding glare 
 from bare lamps placed in the ordinary line of vision. 
 
 Basement Lighting. 
 In the basement in the average residence but a limited amount 
 of light is needed at night, but what few lamps there are must 
 illuminate a large area. It is best not to provide anything which 
 will collect an unnoticed amount of dirt, and for this reason flat, 
 opal reflectors and lamps, placed near the ceiling, usually make 
 
 Fig. 245. — Diagram of hall lighting. 
 
 the best combination. If a more concentrated light is wanted 
 than is afforded by a flat, opal reflector, some conical reflector 
 can be used. Where several lamps are used in a basement they 
 may all be connected to one switch at the top of the 
 basement stairs and all sockets may be equipped with a 
 " turn-down " lamp, which has one filament of one candle 
 power, and another filament of 16 candle power. The one candle 
 power filaments will be kept turned on, so that they will 
 light whenever the switch controling the basement is thrown on. 
 After the switch is thrown on, the user has a dim light from the 
 one candle power filament with which to find his way around, 
 and whenever he desires more light at one lamp, can immediately 
 turn on the 16 candle power filament. When he leaves the base- 
 ment all lights are turned off at the one switch. It is well to 
 provide in most basements a length of portable cord and 
 an extension plug for temporary light in dark comers. 
 
RESIDENCE LIGHTING. 
 
 205 
 
 Fig. 246.— Plan of first floor. 
 
206 
 
 PRACTICAL ILLUMINATION. 
 
 Room 
 or 
 Location 
 
 CO 
 
 1 
 
 & 
 
 *:»- 
 
 ^ 
 
 s 
 
 Height 
 
 of 
 Outlet 
 
 Height 
 
 of 
 Socket 
 
 Style 
 
 of 
 Socket 
 
 Kind 
 
 of 
 
 Switch 
 
 
 ■i: 10 
 
 ■35 
 
 •a 
 
 1 
 5 
 
 
 
 Bemarks 
 
 1 
 
 2 
 
 1 
 
 S 
 ^ 
 
 S 
 
 1 
 
 1 
 
 1 
 
 
 
 .Rafipmpnf 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 , 
 
 switch at 
 
 
 , 
 
 
 
 / 
 
 ceil I no 
 
 ceiling 
 
 kei/less 
 
 
 
 
 
 '4' 
 
 / 
 
 //:■ 
 
 furnace room 
 
 / 
 
 / 
 
 / 
 
 
 s' 
 
 
 
 
 
 
 
 
 
 
 
 fzimace Toom 
 
 / 
 
 
 
 / 
 
 ceilinQ 
 
 6 'S" 
 
 null 
 
 
 droi 
 
 rora 
 
 28 
 
 H' 
 
 / 
 
 /<$ 
 
 extension , 
 cord and pluc 
 
 front room 
 
 / 
 
 
 
 / 
 
 ceilino 
 
 6' G" 
 
 null 
 
 
 irop 
 
 ■nra 
 
 26 
 
 '.r 
 
 / 
 
 16 
 
 la roe room 
 
 7. 
 
 
 
 7 
 
 ceilino 
 
 5' 
 
 keu 
 
 
 a»,f 
 
 
 sa 
 
 H" 
 
 a 
 
 /e 
 
 prouide cord 
 
 First rionr 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 front vorch 
 
 7. 
 
 s 
 
 
 , 
 
 ceding 
 
 ceiling 
 
 keyless 
 
 2 push ft." 
 
 V 
 
 
 
 jp* 
 
 2 
 
 a 
 
 "Separate circuits 
 
 front /?at/ 
 
 10 
 
 /o 
 
 
 1 
 
 ceil I Tiq 
 
 ceiling 
 
 keyless 
 
 •/.push p. 
 
 41 
 
 
 
 .-Jit" 
 
 fO 
 
 V J 
 
 on 3 switches 
 one on 3 way 
 at upper cthd 
 lowf/^halls 
 
 front hall 
 
 r 
 
 , 
 
 / 
 
 
 s's" 
 
 
 
 
 
 
 
 '."ji 
 
 
 /o' 
 
 hall closet 
 
 / 
 
 / 
 
 
 1 
 
 s's" 
 
 
 key 
 
 
 fi 
 
 
 
 
 / 
 
 * 
 
 livinQ room 
 
 z 
 
 z 
 
 
 3 
 
 ceiling 
 
 6'S" 
 
 pall 
 
 t/oushfl* 
 
 P' 
 
 
 
 ■■r 
 
 G 
 
 '^■I 
 
 on separate 
 switches 
 
 UainQ room 
 
 3 
 
 3 
 
 / 
 
 
 S'G" 
 
 
 
 
 
 
 
 
 .1 
 
 ,^'1 
 
 billiard room 
 
 , 
 
 / 
 
 / 
 
 
 S'G' 
 
 
 
 
 
 
 
 ss; 
 
 / 
 
 S' 
 
 •Ltiis biUiardtabk 
 glassware is 
 
 onli/ c^eper 
 
 billiard room 
 
 / 
 
 , 
 
 
 5 
 
 ceilino 
 
 P" 
 
 key 
 
 / pusk fl. 
 
 s 
 
 , 
 
 
 ^^' 
 
 S 
 
 /ff i 
 
 bed room No/ 
 
 
 1 
 
 
 
 ceiling 
 
 
 k.? 
 
 / push fl 
 
 p 
 
 
 32 
 
 i^' 
 
 ,1 
 
 fl 
 
 bed roomM?/ 
 
 1 
 
 t 
 
 
 / 
 
 s' 
 
 
 pull 
 
 
 p. 
 
 
 
 'f 
 
 / 
 
 ^ 
 
 B'chai 77 on socket 
 
 bed Tvom No/ 
 
 P. 
 
 
 / 
 
 / 
 
 
 
 key 
 
 
 I 
 
 
 
 
 ? 
 
 
 frosted lamps 
 
 hall 
 
 , 
 
 
 
 z 
 
 celling 
 
 a' 
 
 keyless 
 
 /joush fl. 
 
 .9" 
 
 
 32 
 
 ir 
 
 3 
 
 8 
 
 ■point sockets 
 
 'oT^li""" ■ 
 
 screen porch 
 
 1 
 
 
 
 / 
 
 ceiling 
 
 ceilino 
 
 keyless 
 
 /snap 
 
 
 
 
 H' 
 
 1 
 
 fl 
 
 screen closet 
 
 1 
 
 1 
 
 
 / 
 
 S'S' 
 
 
 /,eu 
 
 
 G 
 
 
 
 
 / 
 
 ¥ 
 
 center kitchen 
 
 
 
 
 1 
 
 celling 
 
 ceiling 
 
 he^res. 
 
 /snap 
 
 
 
 
 ,k- 
 
 / 
 
 /6 
 
 
 kitchen sink 
 
 / 
 
 1 
 
 
 1 
 
 ceiling 
 
 o' ff" 
 
 null 
 
 
 8 
 
 
 20 
 
 2V 
 
 t 
 
 fl 
 
 
 kitchen range 
 
 / 
 
 1 
 
 
 
 S'e' 
 
 
 
 
 
 
 
 
 
 
 'round frosted 
 lamps 'goes 
 inside a ff glass 
 
 pantry 
 
 / 
 
 1 
 
 
 / 
 
 ceil I ng 
 
 6-6~ 
 
 ft. 11 
 
 
 fl 
 
 
 9S 
 
 91,' 
 
 1 
 
 8 
 
 dinina room 
 
 / 
 
 / 
 
 
 S 
 
 ceiling 
 
 G' 
 
 ket/le&s 
 
 2 pushfl 
 
 3 
 
 
 
 
 V 
 
 
 dining room 
 
 S 
 
 s 
 
 / 
 
 
 S'G' 
 
 
 
 
 
 
 
 te 
 
 / 
 
 
 xtlii^ sep circuits 
 
 bed room fh/-closet 
 
 1 
 
 1 
 
 
 / 
 
 ^'6' 
 
 
 kei/less 
 
 /sna^ 
 
 2 
 
 
 
 !:^■ 
 
 / 
 
 ¥ 
 
 30' instead of 2i 
 
 bed rooTTi /Vol- bath 
 
 1 
 
 
 
 1 
 
 ceiling 
 
 a- 
 
 lie >/ less 
 
 /pushn 
 
 6 
 
 
 
 i¥ 
 
 / 
 
 >/ 
 
 
 iaecond t/nnr 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 bed room V/o 2 
 
 1 
 
 1 
 
 
 .T 
 
 ceiling 
 
 a' 
 
 iev 
 
 /push fl 
 
 ^ 
 
 
 
 If' 
 
 .1 
 
 e 
 
 
 \bed roomNoz 
 
 1 
 
 1 
 
 
 / 
 
 S'G" 
 
 
 null 
 
 
 p 
 
 
 
 '}■ 
 
 
 
 
 \bedrQomNoZ 
 
 z 
 
 i 
 
 / 
 
 / 
 
 S-6" 
 
 
 ken 
 
 
 , 
 
 
 
 
 P. 
 
 fl 
 
 
 ^hed rvoTT) No 2-closet 
 
 1 
 
 1 
 
 
 / 
 
 P'G' 
 
 
 he u less 
 
 /snaji 
 
 z 
 
 
 ZO 
 
 Zlz' 
 
 / 
 
 
 
 bed room A/0 3 
 
 1 
 
 , 
 
 
 J 
 
 ceiling 
 
 8' 
 
 kei/ 
 
 pushfl. 
 
 ?• 
 
 
 3Z 
 
 'f 
 
 3 
 
 a 
 
 
 bed room /^fo 3 
 
 1 
 
 1 
 
 
 / 
 
 S-G' 
 
 
 pull 
 
 
 2 
 
 
 32 
 
 4 
 
 / 
 
 ¥ 
 
 
 bed room No 3 
 
 z 
 
 3 
 
 / 
 
 / 
 
 S'G' 
 
 
 %eu 
 
 
 , 
 
 
 
 
 P. 
 
 8 
 
 
 bed TQomtlo3-closet 
 
 
 1 
 
 
 / 
 
 T^'e- 
 
 
 keyless 
 
 snap 
 
 P 
 
 
 20 
 
 ifir' 
 
 , 
 
 ¥ 
 
 
 bed room t/o¥ 
 
 1 
 
 1 
 
 
 3 
 
 ceilino 
 
 8' 
 
 L„ 
 
 push fl. 
 
 P 
 
 
 3Z 
 
 /if" 
 
 3 
 
 fl 
 
 
 bed roomNof 
 
 
 
 
 / 
 
 5-6' 
 
 
 pull 
 
 
 Z 
 
 
 3Z 
 
 IK" 
 
 f 
 
 
 
 bed room Nov 
 
 z 
 
 Z 
 
 / 
 
 / 
 
 5-6- 
 
 
 keu 
 
 
 1 
 
 
 
 
 2 
 
 fl 
 
 
 bed room NoV- closet 
 
 1 
 
 / 
 
 
 / 
 
 p^'e' 
 
 
 keuless 
 
 snap 
 
 z 
 
 
 so 
 
 M" 
 
 / 
 
 ¥ 
 
 3o'instead cfso 
 
 \bed roomNoS 
 
 1 
 
 I 
 
 
 S 
 
 ceilino 
 
 a' 
 
 itu 
 
 
 p. 
 
 
 
 f^' 
 
 .1 
 
 fl 
 
 
 bed roomNoS 
 
 1 
 
 , 
 
 
 , 
 
 
 
 Bull 
 
 
 ? 
 
 
 
 
 
 ¥ 
 
 ie"chain 
 
 bed raomNoS 
 
 ;; 
 
 s 
 
 / 
 
 , 
 
 S'6' 
 
 
 %eu 
 
 
 , 
 
 
 
 
 2 
 
 S 
 
 
 bed roomNoS-closel 
 
 / 
 
 1 
 
 
 1 
 
 P'6' 
 
 
 kej/less 
 
 
 p. 
 
 
 20 
 
 pi' 
 
 / 
 
 V 
 
 3o'/7/s/eado^ 
 
 bed room No 6 
 
 / 
 
 t 
 
 
 .1 
 
 ceiling 
 
 8' 
 
 keyless 
 
 push n 
 
 -^ 
 
 
 32 
 
 'fi' 
 
 3 
 
 a 
 
 
 bed room t\lo6 
 
 
 
 
 / 
 
 5-6- 
 
 
 pull 
 
 
 z 
 
 
 
 ir 
 
 1 
 
 ¥ 
 
 
 bed room No6 
 
 2 
 
 2 
 
 / 
 
 / 
 
 5'6' 
 
 
 key 
 
 
 / 
 
 
 
 
 Z 
 
 e 
 
 
 bed roomNo6-clospf 
 
 / 
 
 1 
 
 
 / 
 
 ?''6' 
 
 
 
 
 p 
 
 
 
 si' 
 
 / 
 
 t- 
 
 3o'insteadcifJid 
 
 sersants' room 
 
 / 
 
 1 
 
 
 2 
 
 ceiling 
 
 8' 
 
 key 
 
 snap 
 
 ? 
 
 
 3Z 
 
 
 Z 
 
 8 
 
 
 seruants- room 
 
 / 
 
 1 
 
 
 1 
 
 S'S' 
 
 
 pull 
 
 
 
 
 32 
 
 l^' 
 
 / 
 
 ¥ 
 
 
 seroants- room 
 
 z 
 
 
 / 
 
 1 
 
 S'S" 
 
 
 key 
 
 
 1 
 
 
 
 
 2 
 
 8 
 
 
 servants' closet 
 
 1 
 
 I 
 
 
 1 
 
 ceilino 
 
 y 
 
 keyless 
 
 snap 
 
 a 
 
 
 20 
 
 2»" 
 
 / 
 
 ¥ 
 
 
 servants- bath 
 
 1 
 
 1 
 
 
 1 
 
 ceilino 
 
 B' 
 
 keyless 
 
 snap 
 
 Fi 
 
 
 32 
 
 4- 
 
 / 
 
 V 
 
 
 servants- bath 
 
 1 
 
 
 / 
 
 
 5' a' 
 
 
 
 
 
 
 
 
 / 
 
 s' 
 
 
 main bath 
 
 1 
 
 / 
 
 
 1 
 
 ceiling 
 
 8' 
 
 keyless 
 
 Ziush fl. 
 
 a 
 
 
 32 
 
 /*' 
 
 / 
 
 t- 
 
 
 main bath 
 
 / 
 
 1 
 
 J 
 
 
 5' 6" 
 
 
 
 
 
 " 
 
 
 
 / 
 
 s\ 
 
 point sockets 
 lengihwise. 
 
 front end hall 
 
 
 
 
 Z 
 
 ceilino 
 
 8' 
 
 
 push fl. 
 
 9 
 
 
 
 
 ^ 
 
 8 • 
 
 rear end hall 
 
 
 
 
 ' 
 
 aeilino 
 
 8' 
 
 
 
 8 
 
 
 A< 
 
 3^' 
 
 / 
 
 f 
 
 
 Attic 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 ceilino 
 
 / 
 
 
 
 1 
 
 ceilino 
 
 8' 
 
 keyless 
 
 snap 
 
 
 
 
 
 / 
 
 32 
 
 extension pli& 
 and cord teem 
 /triable lamp 
 
 between i^tindaia 
 
 / 
 
 
 
 / 
 
 3' 
 
 3' 
 
 key wall 
 
 
 
 
 
 w=ny 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
RESIDENCE LIGHTING. 207 
 
 An Example of House Lighting Design. 
 
 In order to illustrate further the principles which have been set 
 forth and to offer some suggestions as to the rapid carrjdng out 
 of the work of drawing up lighting specifications the authors 
 have selected for presentation the design of the lighting for a 
 specific house where the planning of the illumination could be 
 treated with a free hand from the beginning. The planning of 
 the illumination for a house or any building should precede that 
 of planning the wiring, because illumination is the ultimate pur- 
 pose for which the wiring is to be installed. As illumination is 
 the primary object desired, we should work back from the 
 desired illumination to the outlets, installation of wires and ma- 
 chinery and not vice versa, as is too frequently done. 
 
 The accompanying plans shown here in Figs. 246, 247 and 248 
 are of a house where the owner desired to obtain excellent illumi- 
 nation with a minimum operating cost. It was recognized at 
 the outset that the first cost of installing the lighting system 
 was small when compared with the running or operating cost 
 from year to year and therefore an effort was made to reduce 
 operating expenses by securing a maximum efficiency of illumi- 
 nation rather than to cut down the cost of installation in general. 
 
 The most rapid and convenient method of drawing up lighting 
 specifications that the authors have found is to mark on the 
 architect's floor plans the location of the outlets or fixtures and 
 at the same time draw up a schedule like the one here shown. 
 Such a schedule can be made out very rapidly compared with any 
 other method of specifying the necessary items with regard to each 
 outlet or fixture. It is a quicker and easier job for the contractor 
 and workmen to get information off such a schedule than 
 from ordinary specifications. On the schedule a separate item 
 is made for each outlet except where there are a number of similar 
 outlets and fixtures in a room. The plans show the location of 
 the fixtures and switches together with the sketches of fixtures. 
 The plans taken with table or schedule furnish nearly all the 
 information necessary for the contractor so that as regards this 
 part of the work but few other specifications are necessary. It 
 should be noted on the accompanying schedule that in the case 
 of all wall brackets the height of the outlet is in every instance 
 specified and in the case of lamps suspendedfrom the ceiling the 
 height of the socket from the floor is given. From such a sche- 
 
208 
 
 PRACTICAL ILLUMINATION. 
 
 Fig. 247. — Plan of cellar. 
 
RESIDENCE LIGHTING. 209 
 
 dule, together with a plan showing the location of lamps, the 
 wireman can immediately take off all that is necessary for his 
 use and the fixture man has complete specifications except as to 
 the particular design of fixture for each case. 
 
 The sketches for the chandeliers and brackets as noted in one 
 column on the schedule and as given in Figs. 249 and 250 (sketches 
 1 to 9), are rough, free-hand drawings intended only for the 
 purpose of indicating at what angle the lamps should hang and 
 indicating only in general the style of fixtm-e; the selection of 
 fixtures being left entirely to owner and fixture dealer with the 
 sketches for guides. Should the fixtures be definitely selected from 
 a fixture dealer's catalogue when the schedule is being made out 
 the dealei's catalogue number should be used instead of the 
 sketch number, as the trouble of making the sketches is thus 
 avoided. 
 
 The specifications for the wiring are outside the province of 
 this book. 
 
 In planning the location of fixtures, especially that of wall 
 brackets in bedrooms, it is desirable to have some means of 
 quickly showing the probable location of furniture or of trying the 
 furniture in different locations. For doing this either in confer- 
 ence with the owner of the house or even if the entire matter is 
 left to the illuminating engineer, a plan adopted by Mr. J. 
 Kermode of the Cleveland Electric Illuminating Company, is 
 to be recommended. Blocks of wood representing furniture and 
 made to the same scale as the architect's plans, (0 . 25 inches repre- 
 senting one foot) , are provided so that beds, dressers, chairs and 
 tables can be arranged in the room under consideration just as 
 they would probably be after the house is finished. In this way 
 both the engineer and the owner can see what is wanted much 
 more quickly than by attempting to picture the situation mentally 
 and measure up the available spaces with a rule. 
 
 The location of fixtures is marked on the plans. Figs. 246 to 
 248, by the usual cross and circle. 1 E marked opposite a fixture 
 on the plans indicates one electric socket, 2 E, two electric sockets, 
 and so on. Gas tips are indicated by the mark 1 G opposite 
 the fixture. Switches are indicated by dotted lines leading to 
 the lamps they control whenever feasible. 
 
 In the schedule it will be seen that one column is devoted to 
 the Fig. number (in this book) of the glassware to be used on 
 
210 
 
 PRACTICAL ILLUMINATION. 
 
 Fig. 248. — Plan of second floor. 
 
RESIDENCE LIGHTING. 
 
 211 
 
 each fixture. In practice this column should be used to give the 
 manufacturer's or supply dealer's catalogue number of the shade 
 or globe required. If glassware from more than one dealer or 
 
 T~7" 
 
 P1_AM OF BJi-1-lAIE.D TAet-C- 
 
 o 
 
 o 
 
 
 9 _ 
 
 6-0- 
 
 -^ ' 
 
 AiiTQ-J^i. ^tVH FciNCfc 
 
 tlWIIiiilf 
 
 Fig. 249.— Sketches 1 to 8. 
 
 manufacturer is needed the numbers can be followed with stars, 
 and other reference marks, referring to the catalogue or maker's 
 names in foot notes at the bottom of the sheet. Such foot notes 
 would then read: 
 
212 PRACTICAL ILLUMINATION. 
 
 * Catalogue of X. Y. Z. Supply Co. 
 
 ** Catalogue of A. B. C. Glass Co., etc. 
 It is manifestly not necessary to make out the whole schedule 
 at once. Parts can be left blank and filled in later. In some 
 cases part of the columns will not need to be used. 
 
 Where much work is being done it is a good plan to have blank 
 schedules printed on tough, thin bond paper. When once the 
 original blank is filled out, blue prints can be obtained of it for 
 bidders, contractors and workmen, or carbon copies can be 
 made if pencil is used. If not enough work is being done 
 to justify this and duplicate copies of the schedule are not 
 needed, a blank schedule with column heads lettered in can be 
 drawn on a piece of tracing cloth. Blue prints can be made of 
 this and the schedule filled out in red pencil on the blue print. 
 For small jobs in small towns this will answer. 
 
 7" 
 
 \ y 
 -< 
 
 Fig. 250. — Sketch 9. 
 
 Although electric lighting is depended upon almost entirely 
 in this house, a few gas jets have been provided for use should the 
 electric service be temporarily interrupted. 
 
 The reason for adopting the plans as shown can best be seen by 
 considering each room briefly. These remarks, however, simply 
 supplement and further explain the plans and schedule. The 
 essence of the specifications is contained in the plans and 
 schedule and each item on the schedule should be studied before 
 reading the following supplementary explanations about the 
 various rooms. 
 
 Porch. — One lamp was placed in front of the door to illuminate 
 the doorway and steps. Inasmuch as the porch would probably 
 be used in the evening another lamp was desirable and placed at 
 the other end. These lamps are controlled by separate switches 
 so that only one need be turned on. A ball globe with direct- 
 ing prisms, like Fig. 89, was specified because of its handsome 
 
RESIDENCE LIGHTING. 213 
 
 appearance and its ab ility to throw the Ught strongly downward 
 although reflectors would have given a greater amount of down- 
 ward light. In case it is desired to read on the porch, 16-candle 
 power lamps are necessary, otherwise eight or four candle power 
 lamps will suffice. 
 
 Front Hall. — The lamps in the front hall or reception room 
 "were placed on three separate circuits. One lamp which can 
 be used as a night lamp is controlled both from a three way switch 
 placed close to the entrance and by another three way switch 
 xipstairs. Lamps A were placed on another switch and lamps 
 -B on a third switch. With these three switches it is therefore 
 possible to get a number of combinations of light and the arrange- 
 ment is such as not to interfere with the symmetry and artistic 
 effect of the lighting when they are not all turned on. These 
 lamps have globes which are calculated to deliver the light down 
 in the room. 
 
 Living Room. — As this room was large, two chandeliers were 
 used, equipped with cut glass prismatic reflecting balls hanging 
 straight down so as to throw the maxiin.um light on a table or 
 -on the book of a person sitting underneath. 
 
 Billiard Room. — In order to provide first, uniform table illumi- 
 nation of sufficient strength; second, to avoid shadows which 
 are unavoidable when the light is concentrated over the middle 
 of the table ; third, to keep the light as much as possible out of the 
 eyes of the players to prevent blinding; fourth, to provide 
 .sufficient general illumination in the room to make it possible 
 to count the string easily and, fifth, to keep the glassware 
 out of the way where it would not be easily broken, the arrange- 
 ment of the five lamps over the table as indicated in sketch No. 
 5, was decided upon; the dimensions on the sketch being those 
 found by experience in billiard rooms to give satisfaction. The 
 bed of a billiard table is generally 31 inches above the floor and 
 if the bottom of the shades are about four feet above the bed of 
 the table the results will be satisfactory. Generally speaking, the 
 lamps when using the shades specified (giving an even distribution 
 of illumination over an area 45 degrees each waj^ from the vertical) 
 should be for a 5 x 10 foot table 6.5 feet apart lengthwise and 
 three feet apart sidewise. For a 4.5 x 9 foot table they should 
 be six feet lengthwise and 33 inches sidewise. In both cases 
 the sockets should hang six feet seven inches from the floor 
 
•214 PRACTICAL ILLUMIXATIOX. 
 
 which would bring the reflector about four feet from the bed of 
 the table. 
 
 Bed Room. — The bed rooms were all treated practically alike. 
 The general illumination is given by a three lamp chandelier 
 placed in the center of the ceiling, eight feet above the floor. 
 The lamps are pointed downward and provided with a rather 
 concentrating form of reflector with eight candle power lamps, 
 which give a good light throughout the room and provide ample 
 concentrated illumination for writing and reading directly under- 
 neath. With very dark walls this candle power might have to 
 be increased somewhat. With light colored walls it is ample. 
 A bracket is provided on each side of the dresser, the brackets 
 being placed five feet apart and equipped with bare frosted bulb 
 lamps, the natural distribution of light about which is such as 
 to throw the maximum light on a person standing in front of 
 the mirror. If gas had not been necessary on the same fixture, 
 it would have been better to place the electric lamp upright on 
 these brackets. With the form of center chandelier used (which 
 is close to the ceiling), sufficient light is thrown on the top of 
 a person's head to light the hair and make unnecessary a bracket 
 or pendant lamp over the top of the dresser. With less light 
 from the central chandelier or with the chandelier lower, a 
 lamp over the dresser niirror would be necessary. Over the 
 head of the bed is provided a four candle power lamp in a con- 
 centrating reflector with a chain socket to provide for reading 
 in bed and easy access to one awakening in the night. Gas is 
 provided in each room simply for use in case of breakdown of 
 the electric service. 
 
 Kitchen. — One lamp was placed in the center of the ceiling 
 of the kitchen for the general lighting of the room. The 
 reflector chosen, lights evenly the entire lower portion of 
 the room. A lamp was placed on a bracket over the sink to give 
 a strong light where most needed in dish washing. 
 
 These notes, in connection with the plans and sketches will 
 enable the reader to make a study of the methods and reasons 
 for arrangements adopted. 
 
CHAPTER XVII. 
 DESK, DRAFTING ROOM AND OFFICE LIGHTING. 
 
 In the lighting of desks there are five principal requirements r 
 
 1. The lamp should be out of the line of ordinary vision of the 
 person working at the desk or should be shaded so that the rays 
 
 ■ cannot strike the eye. 
 
 2. The position of the lamp with reference to the top of the desk 
 should be such that the worker at the desk will not receive the 
 regular reflection or glare from paper on the desk. This 
 regular reflection is one of the commonest causes of trouble in 
 continued working at an artificially lighted desk. 
 
 3. The light should be free from streaks or striations . 
 
 4. Too great intensity of light should be avoided. An inten- 
 sity of from two to five foot candles (normal) on the top of the 
 desk is about right. 
 
 5. The light should be steady — in other words, the lamp should 
 not be swinging; and if it is electric light the voltage regulation 
 must be good. 
 
 A very common and popular method of lighting desks is by 
 the use of a cylindrical desk shade at the center of the desk, as 
 in Fig. 251. While this arrangement is fairly satisfactory, it is 
 capable of much improvement. The lamp is in such a position 
 that when the user of the desk attempts to read or write with a 
 paper directly in the middle of the desk in front of him, there is so 
 much reflection or glare from the paper as to make it not only very 
 trying to the eyes, but difficult and sometimes impossible to read. 
 The work must usually be shifted either to one side or the other 
 of the desk or the lamp must be shifted. The cylindrical desk 
 shade shown in Fig. 251 does not throw a light nearly as strong 
 as the parabolic aluminum refiector shown later in Figs. 256 and 
 257, as can be seen from the light distribution of these reflectors 
 given in Figs. 58 and 60. 
 
 Fig. 252 shows one of the most satisfactory, comfortable and effi- 
 cient ways of desk lighting. Here the lamp is located on the 
 
 215 
 
216 
 
 PRACTICAL ILLUMINATION. 
 
 bracket at the left hand end of the desk. The bracket is equipped 
 with a concentrating type of prismatic glass reflector giving a light 
 distribution similar to Fig. 31. The user of a desk so lighted 
 experiences practically no discomfort from the glare or regular 
 reflection from the paper, while at the same time the desk is well 
 lighted over its entire surface. The comfort of working at such 
 a desk as compared with those lighted by many of the common 
 methods is remarkable. The lamp is high enough above the user 
 of the desk to be out of the line of ordinary vision and further- 
 
 FiG. 251. — Poor arrangement of desk lamp. 
 
 more the eyebrows shield the eyes with the head in ordinary work- 
 ing position. It is not absolutely necessary to equip the lamp 
 with a frosted bulb for diffusing purposes. It is advisable so to 
 equip it, however, on account of other occupants of the room, 
 and because the light is so well diffused by the frosting that less 
 glare is received from the work on the desk and it is conse- 
 quently an easier light to work by. The lamp should be so 
 placed that a plumb line dropped from the lamp to the top of 
 the desk would strike about one foot from the front edge of the 
 desk. This is important if the best results are to be obtained. 
 
DESK, DRAFTING ROOM AND OFFICE LIGHTING. 217 
 
 The bracket in Fig. 252 would give good results with any one 
 of several other types of concentrating reflectors, such as shown in 
 Figs. 19, 21, 51, 53, and 56. The two first should always be used 
 with frosted bulb lamps, and in fact frosting is desirable for any 
 desk lamp for the reasons before explained. 
 
 
 
 |H: ' fl^i> ^ ■ 
 
 i ,■.!.„, 
 
 ..-.;.-. _ -!<".'_i-i ■ ' i ■-»^^,' 
 
 m- 1 ^mh'T mm I 
 
 
 ^^^^^H 
 
 ( 
 
 li 
 
 - ■■• ■■.\\. '~AiS 
 
 ^^B 
 
 ^ 7 — . ir^**" 
 
 ^^^F 
 
 f 
 
 .^^^^^^^^^^_ 1 
 
 ■' " _a^^^^^S 
 
 w 
 
 I : ;%r3:; :l 
 
 
 f 
 
 
 
 
 ■ 
 
 
 
 
 ■*• 
 
 I__;_ 
 
 Fig. 252. — Desk lamp in right position. 
 
 The same relative location of the lamp shown in Fig. 252, can 
 be obtained with portable desk lamps, as, for example, in Fig. 253 
 which shows a fixture of special design. This fixture, however, 
 does not bring the lamp as far out toward the front edge of the 
 desk as it should to give the best results. It should overhang 
 
218 
 
 PRACTICAL ILLUMINATION. 
 
 far enough to bring it directly above a point one foot from the 
 front edge of the desk, as before explained. 
 
 Fig. 254 shows the same position of the lamp as in Fig. 252, 
 but with an ordinary portable desk fixture equipped with a green 
 opal desk shade. A large opal shade like this, however, is too 
 heavy for the ordinary desk portable if placed in the position 
 giving the best results. An aluminum cone such as shown in Fig. 
 52 would be much lighter and better illuminate the desk. A 
 prismatic reflector hke that shown in Fig. 32, would also be hght 
 
 Fig. 253. — Desk portable in fair position. •- "i- 
 
 '.■ ■■ ■«• 
 
 enough to be safe. ', It will always be hecessar}^ to eliminate the 
 streaks which accompany the light from the opal desk shade 
 by using a frosted bulb lamp. •■aV; 
 
 Fig. 255 shows a misdirected attempt to light a desk which is 
 bad from every standpoint. A powerful prismatic reflector 
 mounted on a bracket above the middle of the desk throws the 
 light down at such an angle that the occupant gets a large amount 
 of glare from the paper by working anywhere near the middle of 
 the desk, and receives the full glare of the light from the i-eflector 
 
DESK, DRAFTING ROOM AND OFFICE LIGHTING. 219 
 
 whenever he raises his eyes. Furthermore, much of the Ught 
 does^not fall on the desk, but on the person who is using it. This 
 could be easily rectified by simply turning the desk around as 
 shown in Fig. 252. 
 
 Fig. 256 shows an excellent desk fixture misused. Here the desk 
 is small, and in placing the lamp in the parabolic aluminum re- 
 flector in the position shown it is entirely too close to the paper. 
 The intense light is concentrated on the paper at one point, the 
 edges of the desk being in darkness, and there is much glare or 
 reflection from the paper under the reflector so that the only 
 
 Q\t 
 
 
 
 ~ 
 
 f 
 
 \ 
 
 
 '' i 
 
 If 
 
 
 
 -^^.A.X*'^ 
 
 
 A 
 
 1 
 
 
 
 
 r J 
 
 
 Fig. 254. — Position for standard portable desk lamp. 
 
 portion of the desk which is well lighted can hardly be used with- 
 out ruining the eyes. On a low top desk such as this, the proper 
 position for such a fixture is as shown in Fig. 257. Here the 
 lamp is virtually placed in the same relative position to the user 
 as in Figs. 252 and 254. The fixture is a very useful one because 
 it can be adjusted in so many difEerent positions. The parabolic 
 aluminum frosted finish reflector with which it is equipped (see 
 Figs. 58 and 60) , is more efficient than the cylindrical desk shade 
 shown in Fig. 251. 
 
 Fig. 258 shows the proper position for a lamp on a flat-top 
 
220 
 
 PRACTICAL ILLUMINATION. 
 
 \~ "iti 
 
 Fig. 255. — Bad position for desk lamp. 
 
 Fig. 256. — Good desk portable misused. 
 
DESK, DRAFTING ROOM AND OFFICE LIGHTING. 221 
 
 desk or table. The stand should be placed near the left-hand 
 elbow and an opaque shade should be used, pointed at such an 
 angle that the light will fall on the work without striking the eyes 
 of the user. This position can manifestly be obtained with 
 several other types of portable desk lamps. Fig. 258 was taken 
 merely to show the proper position. 
 
 Fig. 259 shows a method by which as a makeshift, a lamp on 
 an ordinary drop cord can be made to give a fairly satisfactory 
 desk illumination. The lamp should hang over the left-hand end 
 
 Fig. 257. — Desk portable properly adjusted. 
 
 of the desk. By looping a short piece of cord around the thumb- 
 screw on the shade holder and tying the other end around the 
 lamp cord, as indicated, the lamp can be given the proper angle 
 to -light the entire surface of the desk. 
 
 With opaque desk shades other light should always be used in 
 the f oom to supplement the desk light , as otherwise there will be too 
 mudi strain on the eye in changing frequently from the brightly 
 lighted top of the desk to the dark room. With opal or pris- 
 matic glass reflectors enough light escapes through the reflector 
 
222 
 
 PRA CTICA L ILL UMINA TION. 
 
 Fig. 258.^ — Proper position for lamp on flat top desk. 
 
 Fig. 259. — Good desk illumination by a makeshift. 
 
DESK, DRAFTING ROOM AND OFFICE LIGHTING. 223 
 
 for considerable general lighting of the room, so that other light- 
 ing of the room may not be necessary. 
 
 Fig. 260 has been selected as a splendid example of how not to 
 do it. Almost all the principles of good illumination have been 
 violated in this case. In the first place, the lamps are so placed, 
 that when a man looks up from his work, he will have the bare 
 gas mantle glaring him directly in the eye. Second, the light from 
 the lamps will strike the paper at such an angle that it will be 
 regularly reflected, with a resultant glare which will soon produce 
 
 Fig. 260. — Badly lighted tables. 
 
 injurious results. Third, the light is not diffused and the intensity 
 is therefore too high. Fourth, sufficient amount of light is not 
 thrown on the work and there are annoying shadows. In fact, 
 this furnishes a ver)^ good example for a gas man to show his 
 customers how not to light their desks. It would be well in this 
 case to use an opal dome and bobesche which would render the 
 light soft to the eye and give far more light on the work than the 
 present equipment. 
 
 Fig. 261 shows an of^fice where as far as the lighting arrange- 
 ment is concerned, cheapness is evidently the primary object. The 
 
224 
 
 PRACTICAL ILLUMINATION. 
 
 lamps are placed in such a position that when working at the 
 desk one either sits in his own shadow or else as the hght comes 
 from the right, one has a constant shadow of the hand. In fact 
 the arrangement is about as poor as could be conceived. 
 
 The same general principles that apply to the Ughting.of desks 
 apply to the lighting of drafting tables with some modifications. 
 On a drafting table much higher intensity is required and an 
 illumination of ten foot-candles is none too much, since the light 
 must be sufficient to shine through heavy tracing cloth on to pencil 
 
 Fig. 261. — Absurd office lighting. 
 
 drawings underneath. The proper position for a drafting table 
 lamp is at the left of the worker near the far left-hand comer of the 
 table. As it is necessary to place the lamp in front of the worker 
 in order to avoid shadows from the rulers the only way to keep 
 the light from shining in the eyes is for the draftsman to wear a 
 shade over his eyes. A powerful reflector should be used on the 
 lamp and it should be pointed usually toward the near right hand 
 comer of the board. A frosted lamp should always be used over a 
 drafting table and for ordinairy large tables a 32-candle power 
 
DESK, DRAFTING ROOM AND OFFICE LIGHTING. 225 
 
 lamp will be required. The object of the frosted lamp is of course 
 to avoid streaks and to reduce the glare from the paper. 
 
 What has been said heretofore in this chapter as to the lighting 
 of desks has been intended to apply to cases where the general 
 illumination of a room is not sufficient to give all the light that 
 is wanted on the desks and special desk lamps are required. This 
 is a condition which exists in a great majority of offices and draft- 
 ing rooms. There is another method which is that of providing 
 
 Fig. 262. — Drafting room lighting with vapor lamps. 
 
 a great abundance of light in all parts of the room, and so making 
 individual desk lamps unnecessary. Attempts made to provide 
 enough illumination by means of arc lamps as ordinarily equipped 
 so that individual desk lamps are not needed, are likely to be un- 
 satisfactory because of the sharp shadows and the fact that many 
 of the workers at the desk are likely to be in such position as to 
 cast shadows on their own work. With the mecury vapor 
 lamp, where the light comes from a long tube and with ordinary 
 arcs in very large reflectors or concentric diff users, however , light 
 
226 PRACTICAL ILLUMINATION. 
 
 usually comes from sufficiently large surfaces so that, if the 
 walls and ceilings are light in color, shadows do not give 
 trouble, even though the light from only one or two lamps 
 falls on a desk. These large lighting units are manifestly suited 
 • only to large offices. It must be remembered that this scheme 
 of providing for desk or drafting table lighting with the same 
 lamps that are used for the general illumination of a large room 
 is only applicable when a great abundance of light is provided. 
 In many offices such an abundance of light in all parts is a good 
 investment because of the higher working efficiency of the employes 
 it secures. It is especially desirable where there are long book 
 keepers' desks and nuriierous filing cabinets to light because the 
 numerous low lamps necessary to light these by special or indi- 
 vidual lighting would be difficult to arrange so that they would 
 give a good light for the purpose intended and at the same time 
 not shine directlyin the eyes of a number of workers. 
 
 With mercury vapor lamps, office lighting of the character 
 indicated can be accomplished with an expenditure of from . 5- 
 to one watt per square foot of surface to be lighted, depending on 
 the color of the walls. With arc lamps it can be accomplished 
 with from 1.2 to 2.5 watts per square foot. For drafting 
 room lighting with mercury vapor lamps from 1 to 2 watts per 
 square foot should be allowed and with arc lamps from 2. 5 to 5' 
 watts per square foot. 
 
 Fig. 262 is the drafting room of the Link Belt Engineering Com- 
 pany at Nicetown, Pa. The drafting tables are arranged so that 
 light from the windows comes from the front and left of the 
 draftsmen during the day, consequently the mercury vapor lamps, 
 are placed so as to have the light come as nearly as possible from 
 the same direction at night. This is a rule that should be fol- 
 lowed in all drafting room lighting. 
 
■ CHAPTER XVIII. 
 
 THE LIGHTING OF PUBLIC HALLS AND LODGE ROOMS. 
 
 In the lighting of pubhc halls, lodge rooms and the like, the- 
 object to be obtained is an even, general illumination of the whole 
 room with as little eye-trying glare in the eyes of the audience 
 as possible. In audience rooms of all kinds, it is desirable either 
 to conceal the sources of light, or, if this is not practicable to place 
 them high, and so far out of the ordinary line of vision that the 
 effect on the eye will be least detrimental. The rays from a 
 source of artificial light, if that source is not very intense, may 
 not be sufficient to cause noticeable discomfort to a person with 
 strong eyes. When one sits for a long time with the rays 
 
 E 
 
 Fig. 263. — Critical angle for eyebrow shading. 
 
 falling in the eye, however, there usually follows a slight unpleas- 
 ant burning sensation, the cause of which the owner is frequently 
 not aware of. The only way to obviate this entirely is to keep 
 the sources of light where direct rays from them cannot fall 
 in the eye. Some discomfort will result from light coming from 
 above and in front of a person sitting as in an audience unless the 
 light is arranged so that the eyebrows shade the eyes from all 
 direct sources of light. This practically means that light should 
 not be placed where it can shine directly on the face of a person 
 in the audience unless it shines at an angle of less than 25 degrees 
 in front of a perpendicular drawn through the eye. The eyebrows 
 
 227 
 
L'28 
 
 PRACTICAL ILLUMINATION. 
 
 Tvill usually cut off all rays of light coming at less than 25 degrees 
 from such a perpendicular. Thus in Fig. 263 if the eye is at E and 
 the source of Hght at L, the angle a, should be 25 degrees or less 
 for the most comfortable lighting. Wherever possible the lamps 
 sho.uld be placed so that no rays shine backwards at an angle of 
 ■over 25 degrees from the vertical. As such placing is frequently 
 not practicable, the use of diffusing globes is the next best thing, 
 always keeping the lamp as far out of the ordinary line of vision 
 as possible. 
 
 Fig. 264 is an example of a well-lighted lodge ha 11. The lamps 
 
 Fig. 264.— Well lighted lodge hall. 
 
 are placed high on the ceiling and are enclosed in globes such as 
 shown in Fig. 79 which diffuse the light and at the same time 
 •give the maximum amount of light downward. A prismatic re- 
 flecting ball like Fig. 89 would produce similar results. It 
 would have been possible to have lighted this hall well by using 
 instead of these globes, a prismatic reflector with frosted bulb 
 lamps. The latter arrangement would probably have resulted 
 in a higher intensity of light on the floor, since reflectors give a 
 stronger downward candle power than the globes mentioned. 
 The general effect would have been more trying on the eyeS on 
 account of the less perfect diffusion of the light, and the cost of 
 
LIGHTING OF PUBLIC HALLS AND LODGE ROOMS. 229 
 
 lamp renewals would have been greater since frosted bulb lamps 
 have only approximately half the life of clear bulb lamps enclosed 
 in globes. The hall shown in Fig. 264, which is a lodge room 
 known as Washington Hall in Chicago, is 44 feet wide by 50 feet 
 long. There are 48 lamps of 16 candle power on the two ceiling 
 beams and 1 1 lamps in the circle at the center. The six lamps 
 back of the platform should hardly be counted as aiding 
 in the useful illumination of the room, as they detract from rather 
 than add to the general effect, and should never be turned on. 
 Counting the ceiling lamps only, therefore, there are 59 16-candle 
 power lamps to illuminate this hall of 2,200 square feet floor 
 
 Fig. 265. — A bad method of hall lighting. 
 
 space. With lamps taking 56 watts each the power consumption 
 would be 1 . 5 watts per square foot of floor space in this room. 
 The illundnation is good except when the lamps get blackened 
 with age and need renewing. 
 
 Fig. 265 shows an assembly room Ughted mainly from a cen- 
 tral chandelier. This chandelier is equipped with opal bell re- 
 flectors. The light falls in the eyes of persons seated in the back 
 rows. Although a central chandelier is the common method 
 of lighting such a room, it is hard on the eyes of that part of the 
 audience which must face it. It would have been better to 
 place the lamps at the rear and sides of the room so that the rays 
 
230 
 
 PRACTICAL ILLUMINATION. 
 
 would not fall in the eyes of persons in the audience. Somewhat 
 higher efficiency could be secured with the present chandelier 
 and reflectors by pointing the lamps straight down, because with 
 this reflector, which, as seen from Fig. 23, gives a wide distribu- 
 tion of Hght, an unnecessary amount of Hght is thrown toward 
 the ceiling when the light is placed at an angle as in this case. A 
 better way if the chandelier is retained would be to raise the 
 chandelier near to the ceiling and use powerful prismatic concen- 
 trating reflectors to throw the light down, leaving the lamps at 
 about the same angle as shown, thus getting the lamps more out 
 
 Fig. 266. — System of lighting ruinous to eyes. 
 
 •of the line of vision. The bracket lamps on either side of the stage 
 are very trying since they must be faced by the audience. 
 
 Fig. 266, the Masonic Temple lodge room in Chicago, is a glaring 
 example of how not to light such a room. The large number of 
 bare lamps placed on the arches make it impossible for persons 
 in the audience not to receive a large amount of light from these 
 lamps directly in the eyes. Furthermore, since there are no re- 
 flectors on the lamps, a large amount of light is wasted which 
 could be directed downward were reflectors used. The lighting 
 of this hall could be considerably improved by simply putting on 
 deep reflectors with frosted bulb lamps, which would increase 
 
LIGHTING OF PUBLIC HALLS AND LODGE ROOMS. 231 
 
 the amount of illumination on the floor and decrease the glare. 
 In planning the lighting of a new room of this kind the lamps should 
 be placed behind rather than upon the arches and thus be hidden 
 from the view of the audienge when the room is used for general 
 entertainment purposes. If this were done, the lamps should 
 be placed at an angle of about 45 degrees to the ceiling, pointing 
 forward toward the platform , and equipped with efficient reflec- 
 tors to direct the light forward and downward. If possible, no 
 light should be allowed to shine backward toward the audience 
 at an angle of over 25 degrees from perpendicular for reasons 
 explained at the beginning of the chapter. The question of con- 
 cealing lamps in audience rooms is taken up further in the chapters 
 on church lighting and theater lighting. 
 
CHAPTER XIX. 
 
 THE LIGHTING OF LARGE DINIKG AND BANQUET ROOMS. 
 
 In the lighting of banquet rooms custom has dictated that 
 the effect must be that of brillianc}^ and a flood of light. The 
 same has been to a large extent demanded in large dining rooms 
 and is still demanded in the majority of cases, although for cer- 
 tain rooms, (such as grill rooms and rathskellers) , in many high- 
 class hotels and restaurants recently, a subdued effect is aimed at. 
 
 Fig. 267. — Dining room of Auditorium Annex, Chicago. 
 
 Fig. 267 shows a dining room of the Auditorium Annex Chicago 
 lighted with large chandeliers having clusters of frosted bulb 
 lamps. Brackets equipped with frosted bulb lamps bring up 
 the illumination along the sides of the room. The effect is that 
 of brilliancy and plenty of light without glare. This is a case 
 where reflectors, although they would increase the efficiency, 
 
 2.32 
 
LIGHTING OF LARGE DINING AND BANQUET ROOMS. 233-, 
 
 could not be used for artistic reasons. Such cases, however, are 
 rare. In a dining room of this kind expense of operation is an 
 entirely secondary consideration. 
 
 Fig. 268 shows St. Huberts Grill room, Chicago, which is both 
 efficiently and artistically lighted by means of lamps equipped 
 with Holophane spheres. These spheres not only diffuse and 
 soften the light, but direct the rays downward so as to secure 
 good efficiency. These spheres should have been hung nearer the 
 side walls to produce the best distribution of light but since 
 that would not have looked well, side brackets were put in to- 
 bring up the illumination along the side walls. 
 
 Fig. 268. — St. Huberts Grill, Chicago. 
 
 Fig. 269 is of the Auditorium banquet room, Chicago, an excel- 
 lent example of an arrangement which is ruinous to the eyes. The 
 lamps are located on certain architectural features of the room 
 suspended from the ceiling. No one. sea ted in the room save 
 those near the ends can fail to get an immense amount of glare 
 from the array of lamps in clear bulbs. Even the use of frosted 
 bulbs which should be insisted on in a case of this kind would 
 hardly render this arrangement tolerable to a person not having 
 strong eyes. To sit an evening through facing these lamps is a 
 very uncomfortable experience and undoubtedly adds much to 
 
234 
 
 PRACTICAL ILLUMINATION. 
 
 Fig. 269. — Auditorium banquet room, Chicago. 
 
 Fig, 270, — Dining room in Auditorium., Chicago. 
 
LIGHTING OF LARGE DINING AND BANQUET ROOMS. 235 
 
 the profits of Chicago occulists. The effect of such light falling 
 in the eyes for long periods of time, as before explained, is to 
 produce discomfort either in the way of smarting, burning sen- 
 sations, or headaches. The average person with strong eyes 
 does not notice these things until too late. 
 
 Fig. 270 shows the lighting of a dining and banquet room in 
 the Auditorium Hotel, Chicago, which is almost, though not 
 
 Fig. 271. — Wasteful lighting of restaurant. 
 
 quite as bad as the main banquet room shown in Fig. 269 . Light- 
 ing is done by means of bare lamps placed on the arches. The 
 use of deep reflectors on these lamps would considerably increase 
 the efficiency and reduce the glare that proves so ruinous to the 
 eyes of those who must face it continuously. Frosted bulbs 
 should by all means be used. The effect of opal or prismatic reflec- 
 tors would be to deliver a large percentage of useful light down 
 on the tables and reduce the amount of light shining in the eyes of 
 
236 
 
 PRACTICAL ILLUMINATION. 
 
 persons looking lengthways of the room. This would materially 
 improve the conditions. The lamps low down on the arches are ob- 
 jectionable in any event as they are inevitably in the ordinary 
 Une of vision of some of the people seated at the tables. This 
 room is occasionally used as an assembly room; in which case 
 the same objections as to the present lighting would hold. If 
 used exclusively as an assembly room with the platform at the 
 far end the lamps should all be concealed behind the arches so 
 as to be out of sight of the audience. 
 
 Fig. 272. — German room Hotel Stratford, Chicago. 
 
 P"ig. 271 is an excellent example of how not to light a restau- 
 rant. The ceiling is dark. It is studded with incandescent 
 lamps over which there are no reflectors. The candle power of 
 lamps in this room is ample to light the tables fairly well and yet 
 the lighting is far from satisfactory, owing simply to the dark- 
 ness of the ceiling and the absence of reflectors. 
 
 Fig. 272 which is of the German room of the Hotel Stratford, 
 Chicago, is an example of the class of lighting before referred to, 
 in which a subdued effect rather than great brilliancy is desired. 
 
LIGHTING OF LARGE DINING AND BANQUET ROOMS. 237 
 
 The principal lighting is done by lamps placed in special box light 
 fixtures against the ceiling. These boxes have art glass bottoms 
 and open sides. A more absurd design both from an artistic and 
 illuminating standpoint would be hard to imagine. The open 
 sides permit the glaring lamps to appear in all their crudity 
 while the dense art glass bottoms absorb most of the light going 
 down. Such a box fixture should have a ground glass bottom and 
 art glass sides and should be deep enough to permit the lamps to 
 be pointed straight down. The lamps should be equipped with 
 opal or prismatic reflectors to throw the light down through the 
 
 Fig. 273. — Glare and waste in restaurant lighting. 
 
 ground glass. Such a box fixture would be excellent both in 
 efficiency and protection of the eyes. The tables would be well 
 lighted and at the same time the subdued effect desired in such 
 a room would be secured by the comparative darkness of the 
 ceilings and walls. ' 
 
 Fig. 273 shows the lighting of a large restaurant where the 
 chandeliers are placed low and equipped with opal bell reflectors 
 such as shown in Fig. 24, Chandeliers equipped with reflectors of 
 this kind should have the lamp sockets pointed straight down, 
 otherwise too much light is thrown toward the ceiling. Frosted 
 
238 
 
 PRACTICAL ILLUMINATION. 
 
 bulb lamps should be substituted for clear bulb lamps in this 
 case, because of the large amount of lamp filament exposed. 
 With these two changes, the lighting of this room would be 
 efficiently done. 
 
 Fig. 274 shows the lighting of the dining room in the Chicago 
 Atheletic Club in which bare lamps are placed on ceiling stalac- 
 tites. These lamps should be frosted. The lighting arrange- 
 ment of this room is fairly satisfactory, except to persons who 
 are suffering from overstrained eyes to whom presence of light 
 which cannot be avoided is objectionable. However, since the 
 ceihngs are high, it is easy to avoid the rays from the lamps near 
 
 Fig. 274. — Dining room of Chicago Athletic Club. 
 
 by. From an illuminating standpoint it would be better if the 
 form of the ceiling were reversed, that is, if the lamps instead of 
 being placed on the tips of stalactites were placed in recesses in 
 the ceilings so as to be invisible except to those under them. 
 
 Fig. 275 shows a restaurant lighted with lamps placed in orna- 
 mental stalactite globes. The efficiency of this lighting could 
 be much increased by the substitution either of good reflectors 
 for the stalactite globes or by the use of Holophane stalac- 
 tites which would direct the light downward. The present 
 stalactites act only to absorb the light without directing it, so 
 that a large percentage is wasted in the upper portions of the 
 
LIGHTING OF LARGE DIXIXG AND BANQUET ROOMS. 239 
 
 Fig. 275. — Crude effect with opaline stalactites. 
 
 Fig. 276. — Ef&cient ceiling lighting. 
 
240 
 
 PRACTICAL ILLUMIXATIOX 
 
 room and in the globes themselves. The effect is also crude 
 because the lamp filaments show through the opaline globes. 
 
 Fig. 276 is a view of a dining room in the Windermere 
 Hotel, Chicago, with somewhat similar placing of lamps to that 
 shown in Fig. 275 but in which the efficiency is much greater by 
 virtue of the use of Holophane globes and hemispheres for direct- 
 ing the light downwards as well as for diffusing it. 
 
 Fig. 277 shows the Cafe Martin, New York. Here the principal 
 lighting is done from ceiling fixtures equipped with prismatic 
 reflectors which deliver a large amount of light down on the tables. 
 
 Fig. 277. — Cafe Martin, New York. 
 
 These fixtures are high enough to be well out of the ordinary line 
 of vision of persons seated at the tables. Reflectors extending over 
 the'^lamps also tend to reduce the glare received by persons seated 
 sorfi§ -distance away. The decorative lighting is accqjnplished Iby 
 frosted bulb lamps placed on vines on the pillars. It is worthy 
 of note that with a high ceiling of this kind until reflectors such 
 as shown in Fig. 32 were used, very much higher candle-po#6r 
 lamps were necessary. The reflectors not only increase the econ- 
 omy but make a more subdued effect possible, and render the 
 illumination easier on the eyes of the users. 
 
CHAPTER XX, 
 THE LIGHTING OF LARGE PUBLIC ROOMS, DEPOTS, LOBBIES, ETC. 
 
 In the lighting of hotel lobbies, large railroad depots, main 
 floor halls o.f; large buildings, etc., the object to be attained 
 is a general cheerful effect, good illumination on the floor and 
 absence of glare. Glare in such places invariably looks crude and 
 cheap. Its only legitimate use is in cheap advertising by electric 
 light. The day has passed when it is to be tolerated in high- 
 class buildings. The object should be to produce a brightly 
 lighted and cheerful effect without bringing the sources of light 
 into glaring prominence. :/ 
 
 Fig. 278 which shows the interior of the Union Depot at Indian- 
 apolis, is an example of lighting in a style common a few years 
 ago, but seldom installed now. The principal lighting is by enor- 
 mous and elaborate chandeliers hung high from the middle of 
 the arch. The position of the lamps is very good, and they are 
 high enough to.be inoffensive to persons in the waiting room below. 
 With lamps placed high, it is important that more care be given 
 to equipping them with efficient opal or prismatic glass reflectors 
 than was given in this case. The lamps on the chandelier are 
 equipped with etched glass bell shades which are of little worth 
 in directing the light downward. The location of the bracket 
 lamps along the edge of the balcony floor is excellent and would 
 be very efficient were efficient translucent glass reflectors used 
 to throw the light down in the waiting room. The tendency of 
 late for railroad depots and places of this kind has been away 
 frorn elaborate chandeliers with a large number of small incandes- 
 cent lamps, and toward fewer and larger lighting units and these 
 on simple but substantial fixtures, or toward lamps placed 
 directly on the ceiling. 
 
 Fig. 279 shows the interior of the Union Depot at Pittsburg, 
 which is lighted in an unusual but very effective manner by means 
 of six-glower Nernst lamps hung in the space between the first 
 
 241 
 
242 
 
 PRACTICAL ILLUMINATION. 
 
 and second skylights. The result is an' excellent diffused light 
 coming through the skylight, giving the same effect as when the 
 room is lighted by daylight. The room is about 150 feet long, 
 65 feet wide and 40 feet high. There are 105 six-glower lamps 
 hung 4.5 feet above the skyhght. The skylight is wired clear 
 
 Fig. 278. — Indianapolis Union Depot. 
 
 glass, with a ribbed surface, and is from 1 to | in. thick. The 
 absorption of the glass is about 25 per cent. Clear glass globes 
 and white enameled metal reflectors 14 inches in diameter by 4 
 inches high are used over the lamps. The use of concentrating 
 reflectors on the lamps is necessary owing to the" height "of the 
 lamps above the floor 45 f:et below. 
 
LIGHTING OF LARGE PUBLIC ROOMS, ETC. 
 
 243 
 
 The effect of the Hghting is, on the whole, extremely good.^and 
 one can easily read anywhere, a desirable quality in a waiting 
 room. It is probable that four instead of six-glower lamps could 
 be used here and the illuminaton still be sufficient, although the 
 
 o 
 
 ft 
 
 t3 
 
 i. 
 
 present effect of brilliancy might be lost. As the skylight does 
 not extend over the entire room, there is some little difference 
 in the illumination between the sides and the center, but not 
 enough to be greatly objectionable, as the illumination is sufficient 
 
244 PRACTICAL ILLUMINATION. 
 
 at all points for the purpose for which the room is intended. The 
 •chief objection is from the artistic standpoint, as the skylight 
 is badly streaked, due to the kind of glass used in the skyUght 
 itself. 
 
 Fig. 280 shows the lobby of the Majestic Hotel in Chicago. 
 Here handsome modern fixtures are equipped with Holophane 
 globes with a hemisphere in the center. These globes which are 
 similar to that shown in Fig. 83, give excellent diffusion of light. 
 The illumination downward from the hemisphere could be much 
 increased by the use of reflectors over the lamps, as shown in the 
 chapter on clusters and bowls. / ; 
 
 Fig. 280. — Majestic Lobby, Chicago. 
 
 Fig. 281 shows the lobby of the Auditorium Hotel Chicago, as 
 originally equipped. The lighting is done by ceiling clusters below 
 small mirror reflectors of an inverted cone shape.. These reflectors 
 not only look crude, cheap, and out of keeping in such a location 
 but are too small and not of the proper si-^e or shape to be very 
 efficient. For a cluster of this kind a very much larger reflector 
 is needed. A cluster in which each individual lamp is equipped 
 with an opal or prismatic reflector and is frosted would be much 
 better looking and would throw much more light downward. 
 With clusters placed as high as this and individual reflectors on 
 each lamp, these reflectors should be of a fairly concentrating 
 
LIGHTING OF LARGE PUBLIC ROOMS, ETC. 
 
 245 
 
 type so as to deliver as much light as possible in the direction 
 the lamp is pointed. With a small inverted cone reflector such 
 as used, a large amount of light strikes the ceiling from which a 
 part of it is reflected to the side walls and from thence to the floor. 
 
 Fig. 281. — Auditorium Lobby, Chicago. 
 
 As there is loss in each reflction, but a small percentage is de- 
 livered where it is most needed. 
 
 Fig. 282 shows the lobby of the Auditorium Annex Hotel, 
 Chicago, as it appeared when originally equipped- The rows of 
 .bare lamps along the ceiling beams are not in accordance with 
 
240 
 
 PRACTICAL ILLUMINATION. 
 
 modern ideas of artistic lighting. The arrangement is certainly 
 not efficient, since by use of glass reflectors over the lamps a part 
 of the glare from the bare filaments could be avoided and much 
 more light delivered down in the room. This is a case where 
 frosted bulb lamps should be used. The lighting of this 'lobby 
 has since been changed to an entirely different scheme. 
 
 Fig. 283, which was taken in the hall of the Historical Society 
 Building, in Chicago, shows some things to be avoided in this 
 class of lighting. The lighting is accomplished by clusters of 
 lamps on brackets placed rather low. These lamps are in opal 
 bell reflectors, such as shown in Fig. 24, which expose a large 
 
 Fig. 282. — Old arrangement of Annex Lobby, Chicago. 
 
 part of the lamp bulb. The result is a large amount of 
 glare in the ordinary line of vision of persons entering the 
 room, which detracts from the general effect of the illumina- 
 tion. These lamps should have been placed higher and more 
 out of the line of vision, and should have frosted bulbs or be 
 enclosed in diffusing globes, inasmuch as the hall only requires 
 general illumination and not any special illumination at par- 
 ticular points. 
 
 Fig. 284 shows the lighting of the main hall and stairway of 
 the Colonial Theatre, Chicago, where the hghting of the architect- 
 ural details is the principal thing to be obtained. For this reason 
 
LIGHTING OF LARGE PUBLIC ROOMS, ETC. 
 
 247 
 
 no attempt at directing the light in special directions is necessary. 
 The lighting is done by upright bracket fixtures equipped with 
 ground glass globes such as shown in Fig. 75, and basket fixtures 
 
 Fig. 283. — Glare in line of vision. 
 
 Fig. 284. — Entrance to Colonial Theater, Chicago. 
 
 hanging from chains. The greater part of the lighting is done by 
 the bracket lamps. The principal thing to avoid in a location of 
 this kind is the use of opaline globes, which do not diffuse the 
 
248 
 
 PRACTICAL ILLUMINATION. 
 
 light to any extent and which show the lamp filaments through 
 the glass, these filaments having'a sickly red hue and producing 
 a general crude effect out of keeping with the surroundings. 
 Globes selected for use in such a place should either be frosted 
 or sand blasted, or a sufficiently dense ^ppal so that the 
 lamp filamente„will not show through. If 'dense opal globes 
 are used, the total candle power in lamps provided mttstvlse 25 per 
 cent, more than if frosted or sand blasted globes are used. With 
 opaline globes which allow the lamp filaments to show through, 
 the difficulty can be partly overcome: by, wing frosted bulb lamps 
 but this is npt.advisable sincdtlie frosted lamps have about 50 per 
 
 Fig. 285. — Stratford Lobby, Chicago. 
 
 cent! of the life of clear glass lamps, and the same results with 
 less lixpenditure can be accomplished by using frosted enclosing 
 globes instead. When basket reflectors such as shown in Fig. 284 
 are used in such a location, it is to be remembered that they act 
 to a certain extent as reflectors to throw the light upwards, which 
 in the case of lighting arches and vaulted ceilings is desirable. 
 
 Fig. 285 shows the lobby of Hotel Stratford, Chicago, in which 
 the principal lighting is accomplished by lamps located in art 
 glass baskets hung by chains from the ceiling.. This class of 
 lighting is in order where expense is entirely secondary to artistic 
 considerations. In baskets of this kind it is.advisable to use either 
 
LIGHTING OF LARGE PUBLIC ROOMS, ETC. 
 
 24» 
 
 tumip-shaped reflector lamps in a cluster pointing outward at an 
 angle of about 45 degrees, or ordinary lamps equipped with some 
 good common type of reflector which will not deteriorate with 
 heat. The latter is preferable where there is room in the basket, 
 because reflector lamps with a mirrored backing on the globe 
 
 Fig. 286. — Bad mixture of light in office building. 
 
 have a much shorter life than standard lamps and fall off rapidly 
 in efficiency. 
 
 Fig. 286 shows the lighting of the lobby of a large office building 
 in which arc lamps are hung above the main area and a border 
 of incandescent lamps is placed around the edges. This mixture 
 
250 
 
 PRACTICAL ILLUMINATION. 
 
 of arc and incandescent lighting in a place like this produces a 
 bad effect as the whiteness of the arc light tends to make the 
 incandescent look sickly yellow. As a general principle a mix- 
 ure of different kinds of illuminants in this way almost never 
 produces a good effect. The arc lamps in this hall have since 
 been replaced by incandescent lamps placed in large Holopl ane 
 spheres hung by chains, considerably lower than the arc lamps. 
 The incandescent lamps in Fig. 286 should be frosted. Reflect- 
 ors are not necessary in this location because of the white marble 
 surroundings. The clear bulb incandescent lamps, however, 
 produce a crude, inartistic effect on account of their glare. 
 
 .-: ,S- 
 
 FiG. 287. — Indirect lighting in Hotel Jefferson, St. Louis. 
 
 Fig. 287 is a view in the lobby of Hotel Jefferson, St. Louis, 
 which is an interesting example of what is usually known as 
 indirect lighting. The lamps themselves are not visible, being 
 placed in a cove behind beams next to the ceiling at points marked 
 " A A " Fig. 288 shows a cross-section through one of these 
 beams and coves. The lamps are placed in trough mirror reflec- 
 tors, these reflectors being placed at such an angle that the light 
 is thrown towards the ceiling from which it is reflected down into 
 the room to accomplish the desired illumination. It is, of course, 
 essential that the ceiling be light in tint. Because of the fact 
 that all of the light must be reflected from the ceiling such a 
 
LIGHTING OF LARGE PUBLIC ROOMS, ETC. 251 
 
 scheme of illumination requires a greater candle power in lamps 
 than direct illvunination. With indirect illumination the light 
 comes entirely from the ceiling. The result is an absence of 
 shadows and also an absence of glare either from the lamps or 
 from the objects illuminated, which makes both indirect illumin- 
 ation and daylight easier on the eyes than any scheme of artificial 
 illumination where the light comes directly from intense sources 
 of light. 
 
 In the ladies' parlor of this same hotel which is lighted the same 
 way, there are 65 lamps of 16-candle power. The room is oval 
 in shape, being 34 feet 8 inches long by 26 feet 6 inches wide. 
 The area is 645 square feet . If 56-watt lamps are used, the cur- 
 rent required would, therefore, be about 5 .6 watts per square 
 
 Fig. 288. — Section of cove lighting plan. 
 
 foot as against about 2 .5 which would ordinarily be ample for a 
 room of this character with direct illumination. 
 
 Indirect illuminating arrangements are not to be recommended 
 for locations in dirty cities as their effectiveness is so rapidly 
 reduced by dirt darkening the ceiling and collecting in the reflecting 
 trough under the lamps. In no case should it be used unless at- 
 tention is frequently given to keeping the reflecting surface and 
 the ceiling clean. 
 
 A modification of this indirect lighting scheme sometimes used 
 is to have a few bracket lamps in diffusing globes so as to have 
 a little light coming from a definite source and so avoid the almost 
 entire absence of shadow and contrast that some object to where 
 all the lighting is indirect. 
 
 Fig. 289 shows a lobby in which most of the lighting is done 
 
252 PRACTICAL ILLUMINATION. 
 
 by bracket lamps, located high on pillars. The lamps are 
 enclosed in sand blasted globes open at the bottom. If a globe 
 of this shape is used, as it is here, for the sole purpose of enclosing 
 the lamp, it is better not to select one which is open at the bot- 
 tom as a closed globe will look better and prevent the collection 
 of dirt on the globe if closed with asbestos at the top. The ar- 
 rangement shown is fairly good and artistic . In a lobby of this kind 
 where the ceilings are fairly low, it would be possible to design 
 special ceiling fixtures which would conceal the sources of light 
 from those not sitting directty under them. Deep art glass 
 domes or boxes with ground glass bottoms, and lamps equipped 
 
 Fig. 289. — Good lighting of a lobby. 
 
 with reflectors inside them, the reflectors being of such shape 
 as to distribute light over the square spaces between four posts 
 would accomplish this well. A distributing rather than a concen- 
 trating reflector should be used . The light would not be offensive 
 to those seated under it because it would be so far out of the line 
 of vision that the rays would not strike the eye to any great 
 extent while to those seated in an adjoining section of the room 
 the lights would not be visible. 
 
 Fig. 290 shows the rotunda of the Rookery Building, Chicago, 
 recently equipped with artistic fixtures. The central spheres 
 are Holophane glass and similar globes are placed on each of 
 
LIGHTING OF LARGE PUBLIC ROOMS, ETC. 
 
 253 
 
 the four corners. These globes direct the Hght downward, and 
 at the same time diffuse it so as to avoid the glare. 
 
 Fig. 291 shows the lobby of Hotel Vendome, Boston. The cen- 
 
 FiG. 290. — Rotunda of Rooker)', Chicago. 
 
 Fig. 291. — Lobby of Vendome Hotel, Boston. 
 
 tral basket of art glass is surrounded with a group of lamps in 
 Holophane stalactite globes which accompHsh most of the useful 
 lighting. The arrangement is an excellent one in so far as it 
 
254 PRACTICAL ILLUMINATION. 
 
 avoids a mistake sometimes made in connection with this type 
 of fixture, namely, that of using reflectors on the lamps outside 
 the baskets which allow the clear glass lamp bulb to be exposed 
 in all its crudity, thus spoiling the artistic effect of the basket. 
 The globes used in this case increase the illumination at all angles 
 below the horizontal. 
 
CHAPTER XXI 
 
 THE LIGHTING OF HALL AND CORRIDORS OF LARGE OFFICE 
 
 BUILDINGS. 
 
 The lighting of halls of large office buildings, hotels and public 
 buildings is frequently one of the most difficult tasks in illumina- 
 tion for the reason that the artificial light must be used during 
 the daytime when the halls are also partially lighted by daylight. 
 The light from electric incandescent lamps which must be used 
 as a rule in such places is likely to appear yellow and sickly by 
 contrast with white daylight. The object to be attained in the 
 lighting of long halls and corridors is to produce a general cheer- 
 ful well lighted effect. Here is a case where the lighting of the 
 ceilings and side walls as well as that of the floor is important. 
 Nevertheless, the color of the floor plays a prominent part in the 
 general effect and it is almost impossible to make a hall with a 
 very dark floor appear well lighted. It is important to have the 
 lamps well distributed along the length of a long hall rather than 
 to place large lamps or clusters at less frequent intervals. The rea- 
 son for this is very simple. With chandeliers or clusters at in- 
 frequent intervals, the spaces midway between clusters appear 
 poorly lighted by contrast with those portions of the hall near 
 the clusters. The result is that this alternation of light and dark 
 places malces the hall appear gloomy and poorly lighted even 
 though the total candle power of lamps may be sufficient. 
 Therefore to light a long corridor well, it may be laid down as 
 a first principle that small lighting units at frequent intervals 
 should be employed. These units may be equipped in a variety 
 of ways as explained later. 
 
 The next problem is to so equip the lighting units as to avoid 
 the glare of a row of bare incandescent lamp filaments as one 
 looks the length of a hall. This glare, in addition to looking 
 crude, cheap and inartistic in the evening, has the further ob- 
 jection that it looks sickly and ineffective when a hall is partially 
 
 255 
 
'J.-)f, PRACTICAL ILLUMINATION. 
 
 lighted by daylight. Enclosing globes of sand blasted glass or 
 lamps with frosted bulbs have the effect of making the electric 
 incandescent lamps appear to much better advantage when 
 there is a mixture of daylight and artificial Hght besides looking 
 the more refined and artistic at all times. Holophane globes 
 while they accomplish the purpose of diffusing and reducing the 
 intensity of hght and at the same time redirecting its rays, do 
 not appear to as good advantage when there is a mixture of day- 
 hght as do sand blasted or frosted globes. 
 
 In the lighting of hotel corridors off which there are sleeping 
 rooms the problem is somewhat different from that in office 
 and public buildings and corridors off the lobb}'' of a hotel be- 
 cause the upper part of the hotel corridor must be kept dark to 
 prevent light shining through the transoms and annoying 
 guests who have retired for the night and wish an absolutely 
 dark roorri. 
 
 Fig. 292 shows a corridor in the State Mutual Building, Boston, 
 in which correct principles are carried out in the location of 
 lighting pendants at frequent intervals down the length of the 
 hall. These pendants are in this case equipped with Holophane 
 spheres which for the reasons before mentioned are more , per- 
 missible in a dark hall of this kind than in one where there 
 is a considerable mixture of daylight. One lamp with a flat 
 reflector as explained in Chapter VIII, is the best equipment to 
 be placed inside of spheres in such a location. In a hall where 
 there is much daylight an arrangement like that in Fig. 292 might 
 be employed with the substitution of ground glass spheres in place 
 of the Holophane spheres and the use of a flat opal or prismatic 
 reflector over the lamp inside each sphere. With an arrangement 
 such as shown in Fig. 292 it is possible although of questionable 
 utility (as explained later) , to equip each sphere with two lamps 
 on separate switches so that either one or two lamps can be turned 
 on as may be required, thus giving two different intensities of 
 illumination in the hall. If two intensities are desired, this is 
 a much better arrangement than that of turning off the light on 
 every other pendant as is sometimes done. The effect of turn- 
 ing off every other lamp is to produce light and dark areas 
 down the length of the hall making it appear gloomy as well as 
 spoiling the appearance of the row of lamps. 
 
 Another way of providing for lamps on separate switches at 
 
HALL AND CORRIDOR LIGHTING. 257 
 
 each outlet in corridor lighting is to install hemispheres as shown 
 in Fig. 293. 
 
 As to whether two intensities of light are practically needed in 
 the ordinary office building, there is room for argument, although 
 it is to be seriously doubted whether such provision is necessary. 
 
 Fig. 292. — Corridor of State Mutual Building, Boston. 
 
 Enough light must be provided to light the halls well during the 
 hours of darkness. What would be a very good intensity of 
 light in a hall in the hours of darkness would seem rather gloomy 
 during the hours when certain portions of the hall were lighted 
 with daylight. It will therefore usually be necessary if any 
 artificial light is eniployed during the daytime to use as much 
 
258 
 
 PRACTICAL ILLUMINATION. 
 
 then as is needed at night in order thus to avoid a gloomy ap- 
 pearance in the hall. 
 
 There is room for some argument as to whether reflectors are 
 necessary or advisable where a row of pendants are located as in 
 Fig. 292. It may be argued that since the object of hall lighting 
 is to make the whole appear cheerful, it is as necessary to light 
 
 Fig. 293. — Corridor of Board of Trade, Boston. 
 
 the ceilings and high side walls as the floors. It must be re- 
 membered, however, that the lamps themselves are located \'ery 
 near the ceiling and high side walls, while they are some dis- 
 tance from the floor and that if no attempt is made to alter the 
 natural distribution of light, the ceilings and high side walls 
 will receive much more light than the floors with the result 
 that the floors may appear rather dark and spoil the whole effect. 
 
HALL AND CORRIDOR LIGHTING. 
 
 259- 
 
 For this reason reflectors are usually advisable in the ordinary 
 office building hall such as shown in Figs. 292 and 293 since by 
 their use a fairly even illumination of ceilings, walls and floor 
 can be secured. In casR enclosing globes with 4 inch holders or 
 larger are used, small reflectors which slip directly on the lamp 
 
 Fig. 294. — Main floor H. W. Hillman Building, Los Angeles. 
 
 bulb can be secured which add to the downward illumination 
 very much. Some such reflectors are of the prismatic type and 
 others are of metal with either bright metal or white enamel 
 reflecting surface. A ver}'' small opaque reflector is permiss- 
 ible inside of a ground glass diffusing globe for the reason that 
 
260 
 
 PRACTICA L ILL UMINA TION. 
 
 the diffusion in the globe with a small reflector will disguise the 
 fact that the reflector is opaque and the ceiling above will not be 
 seriously darkened. 
 
 There are certain classes of public building halls which are 
 in the nature of lobbies where it is sometimes desirable to throw 
 light upward to illuminate architectural or artistic features of 
 high vaulted ceilings^ Such-eases are taken up under the head 
 of lighting of lobbies. 
 
 Fig. 295. — Corridor Hotel Somerset, Boston. 
 
 A case somewhat of this character is presented in Fig. 294 
 which is a corridor on the first floor of the H. W. Hillman Build- 
 ing, Los Angeles. Here the lighting is by means of lamps enclosed 
 in sand blasted globes placed on high brackets along the side 
 walls in a position where they illuminate floors and ceilings 
 about equally. In a hall of this kind which has white marble 
 finish and light floors it is easy to produce a cheerful appearance 
 and no special attempts to direct the rays of light are necessary 
 with side brackets located as shown. 
 
HALL AXD CORRIDOR LIGHTING. 261 
 
 Another example of this is the corridor of the Hotel Somerset, 
 Boston, Fig. 295. Here the lighting from the brackets is sup- 
 plemented by lamps located in a cove above the cornice and 
 throwing their light to the ceiling. This indirect lighting, although 
 pleasing, is an expensive method of producing results for the 
 reasons mentioned in Chapter XX in connection with the 
 Hotel Jefferson, St. Louis. 
 
CHAPTER XXII. 
 
 THE LIGHTING OF THEATERS. 
 
 Under this head it is proposed to take up simply the lighting 
 ■of that portion of the theater occupied by the audience. Stage 
 lighting is a special branch which has received due attention 
 from specialists in that line. In theater lighting the lamps must 
 be kept as much out of the sight of persons in the audience as 
 possible; that is, they must be placed so that the r&js from the 
 
 Fig. 296. — Auditorium Theater, Chicago, from balcony. 
 
 lamps will not enter the eyes of persons seated in the audience 
 who are looking toward the stage. Of course to conceal them 
 absolutely would be an impossibility. Next after keeping the 
 lamps out of the line of ordinary vision the most important thing 
 in theater lighting is to deliver as much light as possible on the 
 audience from lamps located very high above the audience. 
 
 262 
 
THE LIGHTING OF THEATERS. 
 
 26.? 
 
 For this reason efficient reflectors can be used to great advantage 
 although in a great many cases they are not used ; which results 
 in a great loss of light. 
 
 The statement made in Chapter XVIII as to the angle with 
 reference to the eye that lamps must be placed if the eyebrow 
 is to shade the eye can be considered here with advantage. 
 
 Fig. 296 shows the Auditorium Theatre in Chicago which is 
 famous as a ruiner of eye sight. A prominent Chicago occulist 
 states that the lighting of this theater has brought him thousands 
 
 Fig. 297. — Side view, Auditorium Theater. 
 
 of dollars worth of business because of the disastrous effect it 
 has on the eyes of the persons going there habitually. As seen 
 from Fig. 296 the principal lighting is done by bare lamps placed 
 on the arches so that the persons in the audience in the balconies 
 must take the glare from this entire array of lamps. Even persons 
 with strong eyesight experience some discomfort if obliged to 
 face these lamps for a whole evening. This is used considerably 
 as a music hall and consequently the lamps are not alwa\ s 
 turned out during a performance as in a theater. 
 
 Fig. 297 is another view of the auditorium theater taken 
 
264 
 
 PRACTICA L ILL UMINA TION. 
 
 Fig. 298. — Majestic Theater, Boston. 
 
 Firs. 299. — Theater with concealed lights. 
 
THE LIGHTING OF THEATERS. 
 
 265 
 
 from one side. It shows how easy it would have been to locate 
 the lamps behind the arches so that the}'' would have been out of 
 sight of the audience and at the same time would effectually Ught 
 the theater. No reflectors are used on these lamps. This is a 
 case where a good reflector would greatly increase the illumination 
 on the audience. With bare lamps the ceilings must be depended 
 upon to reflect down the large amount of light which strikes- 
 them . Much of this light must be reflected back and forth between 
 various surfaces several times before it reaches the audience and 
 
 Fig. 300. — Orchestra Hall, Chicago. 
 
 is virtually lost in the process. With lamps placed so far above 
 the audience as is necessary in a theater, reflectors which will 
 converge the rays of light and throw them down are absolutely 
 essential to economy. 
 
 Fig. 298 shows the Majestic Theatre, Boston, where Holophane 
 hemispheres are located on the arches. In this case the arches 
 are not as nearly in the ordinary line of vision as in the Chicago 
 Auditorium, and the arch lamps are further rendered as inoffensive 
 as any exposed lamps can be by covering with the diffusing hemis- 
 
■266 
 
 PRACTICAL ILLUMINATION. 
 
 pheres. This is as pleasing and artistic an arrangement as can 
 be found for lamps placed in view on the arches. As theater 
 owners so frequently clamor for plenty of exposed lamps for 
 decorative effect, this is an easy way to give it to them with a 
 minimum of injury to the eyes of theater goers. 
 
 Fig. 299 is a view from the stage of a theater in which the 
 lamps are properly located behind the beams of the ceiling and 
 along the front of the balconies. Here also is a case where re- 
 flectors would materially increase the efficiency. The lamps if 
 
 • * 
 • 
 • 
 
 -?^"" 
 
 .v^^jidB^-*'^-. ..■■■;>.,'■ 
 
 ■ ,*■ -|i i 
 
 rT-ci .;. 
 
 — 
 
 
 ■"m-- /■ ■- 
 
 . -mm:- : 
 
 ^^^Mi**rr'*'''-iif '■"■'■■*' * 
 
 
 BB^SSSp^Br'-'-'-^ • ■ * 
 
 MpPMI 
 
 l;i \\ \\ ■\ 
 
 i ■ ' > 
 
 
 UMii&Si^Mk ■<-.: 
 
 ^ PT 
 
 ^Mk- 
 
 Fig. 301.— Orchestra Hall, Chicago. 
 
 so "equipped should be placed at an angle of about 45 degrees 
 instead of horizontal as they are shown. 
 
 Figs. 300 and 301, show the lighting of Orchestra Hall, Chicago. 
 Fig. 300 taken from a balcony shows the lamps to be almost in- 
 visible to persons in the audience except that a small amount 
 of reflection from the ceiling can be seen and the lamps along 
 the front of the balconies at each side are visible. Fig. 301 taken 
 from the stage shows better how the lamps are located behind 
 the arches so as not to annoy the audience. If the lamps along 
 
THE LIGHTING OF THEATERS- 267 
 
 the fronts of the balconies could be concealed in a cove, espec- 
 ially around the sides where they are exposed to those sitting 
 on the opposite side of the balcony, it would be another marked 
 improvement. No reflectors are used on the lamps high up on 
 the arches in this theater. The amount of light delivered to the 
 audience could be much increased by their use. 
 
 Part of the lamps in the audience room are usually turned out 
 during theater performances. Nevertheless, the lighting arrange- 
 ments should be such that if it is desired to turn all lamps on 
 during the entire performance it can be done without straining 
 the eyesight of persons in the audience. 
 
 The frontispiece shows the large audience-room in the new 
 Engineering Societies' Building in New York, which is a beauti- 
 ful example of semi-concealed lighting. The principal light comes 
 from tb.e ceiling. In the centre of each of the three spaces there 
 is a large panel of diffusing glass, behind which are placed 
 nearly 800 reflector lamps. 'Qj reflector lamps is meant the 
 turnip-shaped incandescent lamp with a silvered mirror surface 
 applied to its top, which acts as a reflector. It would have been 
 much better to have used standard lamps with some sort of a 
 reflector, but the space was too confined to allow this. The 
 objection to reflector lamps is their comparative inefificiency if 
 their life is reasonably long, and their short life if they are made 
 of reasonable efficiency. 
 
 In addition to the lighting from the ceiling in this auditorium, 
 there is a row of ceiling lamps back of the stage and there 
 are clusters of lamps around the sides of the balcony and 
 orchestra floor. These clusters should not be turned on except 
 when the audience is assembling, as the light is otherwise apt to 
 shine in the eyes of some of those in the audience. 
 
CHAPTER XXIII. 
 THE LIGHTING OF CHURCHES. 
 
 The lighting of churches is more difificult than that of theaters. 
 It is even more important to keep light from shining in the eyes of 
 the audience in a church than in a theater because in a church 
 most of the light is turned on as long as the audience is present. 
 At the same time it is more dififiult to conceal the sources of light 
 in a church than in a theater because the usual church architec- 
 ture makes it difficult. The old plan of lighting churches by 
 means of immense clusters of lamps mounted on central chande- 
 liers so that the light must fall directly in the eyes of all except 
 those seated well forward in the audience must give way to some 
 more hygienic, sensible and artistic method. To all persons but 
 those with the strongest of eyes the facing of artificial light stead- 
 il}' through church services is uncomfortable and injurious. 
 
 Fig. 302 shows the lighting of the Second Presbyterian Church 
 in Chicago. This a most remarkable example of lighting and 
 illustrates how far attempts at artistic lighting can be carried 
 to the detriment of eyesight and of satisfactory lighting. The 
 lighting fixtures which are certainly of original design have in- 
 candescent lamps placed upright in saucer shaped devices. If 
 the fixtures were judged from a purely engineering standpoint 
 one would imagine that the object aimed at was to throw as 
 much light upward as possible. This latter result must have 
 been fairly well secured because we note on a chandelier at the 
 extreme right that these saucers have been placed on top of 
 the lamps to act as reflectors for throwing the light down. There 
 is entirely too much light in the ordinary line of vision of persons 
 in the audience in this church. Not only are the lamps on the 
 chandeliers in trying positions, to those in the balcony, but the 
 whole situation is made worse by the decorative lighting around 
 the pulpit and organ, which must be looked at during the entire 
 service by everyone in the audience who makes any pretense to 
 
 268 
 
THE LIGHTING OF CHURCHES. 269 
 
 appearing attentive to what is going on. The hghting of the galler- 
 ies in this church should have been accomplished by lamps 
 located back in the arches where they would be out of sight of 
 as much of the audience as possible. The lighting of the center 
 of the main floor could be well accomplished by specially designed 
 fixtures of the type shown later in Figs. 309 to 314, these fixtures 
 being arranged so that a person sitting behind would receive no 
 more light than would shine through a very dense art glass. In 
 other words, persons in the audience should receive light from 
 above and behind rather than from the front. 
 
 Fig. 302. — Second Presbyterian Church, Chicago. 
 
 Fig. 303 shows the First Church of Christ, Scientist, Chicago, 
 which has an unusual architectural design and consequently 
 permits of unusual church lighting arrangements. The lighting 
 is done by lamps placed on the arches and around the central 
 dome. Frosted lamps should always be used in such a case. 
 The design of the lighting is fairly good from a hygienic stand- 
 point because but a small amount of light direct from the lamps 
 enters the eye of most persons in the audience. It would be 
 much better, however, were deep reflectors used on the lamps 
 on the arches, thus shading the lamps and increasing the amount 
 of light thrown down. It would be desirable also to have de- 
 
270 
 
 PRACTICAL ILLUMINATION. 
 
 signed the dome to permit of the lamps being pointed downward 
 and placed in some common type of translucent reflector to 
 direct the light down. In an audience room of such shape as 
 this there appears to be no opportunity for arranging lamps so 
 that they will be entirely out of sight of more than two-thirds 
 of the audience. The only method of treatment is therefore to 
 make liberal use of deep reflectors or diffusing globes. 
 
 Fig. 304 shows the Ughting of a Chicago cathedral in which 
 lamps are placed high up on the main arch and on the pillars. 
 The pillar lighting is objectionable for reasons before emphasized. 
 
 Fig. 303. — First Christian Science Church, Chicago. 
 
 Too much light shines into the eyes of the audience. The same 
 thing is true of the lig ting of the altar. The lamps high on the 
 arch are sufficiently above the range of ordinary vision to be un- 
 objectionable. Lamps so located should have powerful concen- 
 trating reflectors giving a light distribution like that shown in 
 Fig. 35 in order to deliver as much light as possible down on the 
 pews. In a church of this kind in which the pews all face one 
 way it would be an easy matter to conceal most of the useful 
 lighting behind the various features in the architecture so that 
 it would not injure the eyes of those in the audience. 
 
 Fig. 305 shows the lighting of a small church, this view being 
 
THE LIGHTING OF CHURCHES. 
 
 271 
 
 from the pulpit. This is not by any means to be taken as an 
 example of model lighting, but is shown to give some ideas as 
 to how lights can be arranged so as to be out of sight of the aud- 
 ience. The lighting as it stands represents simply a successful 
 attempt to remodel with a small amount of money a very unsatis- 
 factory previous arrangement. The principal lighting of this 
 church auditorium was originally designed to be accomplished 
 from the central basket form of chandelier seen. This basket 
 was of copper wire gauze and had a number of bare incandescent 
 lamps located inside of it. The copper wire gauze was designed 
 
 -• ■ ■ :"■■■•■„. I'l' i! 
 
 -1; f i 
 
 
 '''^- 11, III 
 
 
 \ II 
 
 Fig. 304. — Chicago Cathedral. 
 
 to act as a diffusing covering which it did to a certain extent but 
 with an enormous loss of light. As it was necessary to increase 
 the illumination without placing any more load on ,the circuits 
 and necessitating the expense of rewiring, the following changes 
 were made. The candle-power of lamps in the central chande- 
 lier was reduced one-third and the lamps were placed outside 
 the basket and equipped with prismatic reflectors as seen in Fig. 
 305. This failed to bring up the illumination on the sides of the 
 room sufficiently for comfortable reading. To accomplish this 
 latter purpose the lamp shown between the beams at the extreme 
 left was installed with a prismatic reflector; a corresponding 
 
272 
 
 PRACTICAL ILLUMINATION. 
 
 lamp being placed on the other side of the audience room. At 
 the pillars in the rear there was previously a lantern fixture cov- 
 ered with copper wire gauze, the absorption of which was not 
 less than 50 per cent. The illumination in the rear of the room 
 being entirely unsatisfactory, two lamps were placed in angle 
 wall sockets pointing downward at an angle of 45 degrees and 
 
 Fig. 305. — Lighting of a small Church, remodeled. 
 
 equipped with powerful prismatic reflectors so as to throw light 
 from behind on those seated in the rear of the room. The re- 
 sult of these changes was to make the amount of illumination 
 in all parts of the room fairly satisfactory. The central chandelier 
 throws too much light in the eyes of those seated in the back 
 part of the room and for this reason is objectionable. A larger 
 
THE LIGHTING OF CHURCHES. 
 
 273 
 
 proportion of the total lighting should be done from behind, 
 leaving the chandelier to serve mainly a decorative purpose. 
 Artistic ceiling fixtures placed on the walls or on the slopes of 
 the roof between beams could be used in place of the simple wall 
 socket shown in Fig. 305. Conditions as they stand could be 
 improved by substituting for the reflectors used on the chande- 
 lier a deeper type of reflector which gives a wider distribution 
 of light such as shown in Fig. 37. This would hide the bare 
 lamps from those seated in the rear seats so that they would 
 receive only diffused light through the reflector and would cause 
 the chandelier lamp to deliver more light to the sides of the room 
 and less immediatel}^ under the chandelier than is now the case. 
 
 Fig. 306. — Fifth Christian Science Church, New York. 
 
 Such reflectors were not available until after this work was done. 
 
 Fig. 306 shows another common method of church lighting 
 also open to the general objection of too much light from the 
 lamps shining in the eyes of persons in the audience. As far 
 as simple efficiency goes, the arrangement shown gives excellent 
 results, the lamps on the chandeliers being equipped with con- 
 centrating types of prismatic reflectors which deliver a large 
 percentage of the light down on the audience. Frosted lamps 
 should be used here. 
 
 Fig. 307 shows a church with bare gas jet brackets along the 
 side walls and clusters of open flame burners over the organ and 
 the choir. This is a good example of what to avoid as far as the 
 
274 
 
 PRACTICAL ILLUMINATION. 
 
 lighting of the front of the church is concerned. The lamps there 
 are placed where everybody must look directly at them during 
 the entire length of the service. They are so situated as to be 
 not only tiring to the eye, but owing to the eye working with a 
 small aperture, the abihty to see the preacher is materially de- 
 creased. A canvass among the congregation would doubtless 
 disclose the fact that many people were subject to headache 
 when in th e church and that owing to the fact that it would be easier 
 to close the eyes than to look at the light, people were put to 
 sleep, not necessarily by the sermon. 
 
 Fig. 307. — Church with bare gas jets. 
 
 Fig. 308 shows the interior of the First Church of Christ, 
 Scientist, Pittsburg, lighted by Nemst lamps hung high. Since 
 the Nemst lamp, like an incandescent lamp with a reflector, throws 
 so large a proportion of its light down asseenin Figs. 173 to 176, 
 it is well suited to being htmg high. Because the lamps are 
 placed so high the lighting of this church is the least open to 
 criticism of any of those shown. It is the best that can be done 
 without concealing the lamps. 
 
 With ordinary fixtures there are only two methods of church 
 lighting which represent even an attempt to avoid the very ob- 
 jectionable feature of light shining in the eyes of the audience. 
 
THE LIGHTING OF CHURCHES. 275 
 
 One method is to conceal the lamps behind architectural features 
 such as beams and pillars in a way that the light will shine down 
 and forward but cannot shine backwards so as to fall on the eyes. 
 This method is good, but there are many churches or parts of 
 churches where it cannot be done. Simply placing the lamps 
 very high, while it is much better than having them low where 
 they cannot be avoided by the eye nevertheless is open to serious 
 objection. As long as artificial light can fall steadily for some 
 time on the eyeball, no matter from how high an angle it will 
 produce eye strain. The light must come from such an angle that 
 
 Fig. 308. — First Christian Science Church, Pittsburg. 
 
 the eyebrow shades the eyeball. This critical angle wherL 
 a person is seated looking forward as in a church is about 25 de- 
 grees from a perpendicular drawn through the e5''eball, as explained 
 in the chapter on the lighting of public halls and lodge rooms. 
 Fig. 263. To light a church well from a hygienic standpoint 
 then, we must have no rays of light shining backwards at an 
 angle of more than 25 degrees from the perpendicular. The light- 
 ing must therefore be done by light shining down sideways and 
 forward toward the pulpit. 
 
 To accomplish this where the light cannot be shaded by archi- 
 tectural features or placed on back walls and ceiling, slopes, a. 
 
276 
 
 PRACTICAL ILLUMINATIO.X. 
 
 special design of fixture is necessary. Some suggestions along 
 this line are given here to show the general principles which must 
 govern the design of such fixtures. 
 
 Fig. 309 is a side view and Fig. 310 a rear view of a suggested 
 form of fixture. In Fig. 309 the side A, B, is placed toward the 
 pulpit and the side C, D, toward the audience. It is made up 
 in box form with the bottom open. The sides are art glass of 
 any desired design. The side, C, D, should be of very dense, 
 nearly opaque, glass, since this is the side which acts as a shield 
 to protect the eyes of the audience seated behind it. The other 
 sides need not be as dense unless uniformity for the sake of ap- 
 
 
 tc Q 
 
 A1 
 
 
 
 
 , i 
 
 
 Art Glass 
 
 
 B^ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 -2S 
 
 
 
 
 
 
 • 
 
 Fig. 309. — Side view of art glass fixture for church hghting. 
 
 pearances is wanted and that is mainly a matter of taste. A 
 sectional side view of this fixture is seen in Fig. 311, which shows 
 the arrangement of one of the lamp sockets and reflectors aS 
 viewed froin the side. Fig. 312 is a sectional view from the rear 
 showing both of the lamp sockets and reflectors as they would 
 appear from the rear if the art glass of the side C, D, were removed. 
 The reflectors selected should preferabh- be a fairly concentrating 
 type of prismatic reflector such, for example, as was shown in 
 Figs. 31 and 35. This fixture is designed for two lamps of 32 or 50 
 candle power each. The refiectors are pointed so as to throw 
 the light forward and sideways. The side C, D, keeps the 
 
THE LIGHT IX G OF CHURCHES. 
 
 277 
 
 light from shining backward at inore than the critical angle 
 of 25 degrees except the little going through the dense art glass. 
 The side A, B, and those joining it are made shorter so as not to 
 interfere with the light thrown forward and sideways. In order 
 to accomplish the shading and avoid excessive height of the side 
 C, D, the lamps should be placed close to this side as shown. 
 The dimensions are given on the drawings. This is as small as 
 the fixture should be made for two 32 candle-power lamps because 
 if the side C, D, is made less than 24 inches, it will not shade the 
 eyes from the lamps if allowances are made for ordinary inac- 
 
 1 
 
 . 
 
 
 -t- 
 
 Art Glass 
 
 1 
 
 
 5 
 
 
 
 
 
 
 
 
 
 
 
 Fig. 310. — Rear view of art glass fixture for church Hghting. 
 
 curacies of manufacture in such goods. It can, of course, be 
 made as much larger as seems desirable. The fixture arranged 
 as shown with reflectors of the kind shown in Fig. 36 will light 
 an area extending forward toward the pulpit, a distance approx- 
 imately equal to the- height of the fixture from the floor and 
 double that distance sideways. There is room for a third lamp 
 without increasing the width. It is, generally speaking, advis- 
 able to hang these fixtures rather low and use a number of them 
 to cover this area to be lighted rather than to place a few large 
 fixtures high up. There is a great gain in economy of light b}^ 
 hanging them low since the intensity of light on the pews varies 
 
278 
 
 PRA CTICA L ILL UMINA TION. 
 
 inversely as the square of the distance of the fixtures above them. 
 The usual objection to low chandeliers (that the light is brought 
 too nearly in the line of vision), does not hold here because of 
 the shading effect of the dense art glass backs of these fixtures. 
 
 r 
 
 A —I 
 
 < 23 ^ 
 
 1 
 
 / i 
 
 Lld 
 
 Fig. 311. — Sectional side view of church fixture. 
 
 The general character of the fixture shown in Figs. 309 and 310 
 is similar in external appearance to the fixtures shown in Chapter 
 XXVI on the Lighting of Stores, Fig. 343. 
 
 The same principle can be carried out in other designs, but 
 
 \ / 
 \ / 
 
 \ / 
 
 X 
 
 Fig. 312. — Sectional rear view of church fixture. 
 
 only two others will be shown here. The lantern design lends 
 itself well to the carrying out of this principle as shown in Fig. 
 313 which is a very plain lantern design and also in Fig. 314 
 which is more elaborate. 
 
THE LIGHTING OF CHURCHES. 
 
 279 
 
 Fig. 313. — Church fixture in lantern design, side view. 
 
 Fig. 314. — Church fixture in lantern design side view. 
 
CHAPTER XXIV. 
 
 THE LIGHTING OF LIBRARIES, READING AND SCHOOL ROOMS. 
 
 In lighting library reading rooms the principal light is of 
 course needed on and near the reading tables. A very common 
 method of lighting reading room tables is to use green opal desk 
 shades on desk lamps ranged along the center of a table with 
 lamps from 12 to 18 inches above the table top; the chairs for 
 reading being ranged along both sides as in Fig. 315. This ar- 
 rangement is not altogether efficient or satisfactory. Most 
 green opal desk shades concentrate light immediately under 
 the shade and when placed over the center of a reading room 
 table the result is a brilliantly lighted center (which is of no use 
 to any one), and comparatively dark edges of the table. The 
 outwardly pointed reflectors at the ends are extremely bad be- 
 cause the light shines directly in the eyes of the person seated at 
 the end of the table. In order to spread the light more evenly 
 over the table in Fig. 315 the desk shades should be raised a little 
 higher from the table so as to light it more uniformly and inside 
 of the green opal desk shades a globe or reflector should be used 
 which would distribute most of the light at and near an angle of 
 45 degrees from the vertical. This can be done by using a desk 
 shade with 3 .25 inch holder and placing inside the desk shade 
 on a 2 .25 inch holder a prismatic reflector of the type which gives 
 a raaximum candle-power at an angle between 30 and 45 degrees 
 from the vertical such as shown in Fig. 37 using frosted lamps. An- 
 other excellent plan for reading table lighting in a library is to 
 place the lamps on a chandelier above the table clear out of the 
 line of ordinary vision of the readers. This plan is likely to give 
 the readers less trouble from the glare of reflection from the paper 
 than when the low table lamps are used. If the reading tables 
 are lighted from chandeliers, however, pains must be taken that 
 the chandeliers are designed and equipped to give the desired 
 results. 
 
 2S0 
 
THE LIGHTING OF LIBRARIES. 
 
 281 
 
 For example, Fig. 316 shows a library reading room in which 
 the lighting is done by chandeliers above each table. The 
 arrangement is decidedly faulty, however, evidently due to a lack 
 of knowledge of the characteristics of the different reflectors on 
 the part of those responsible for the installation. The chandelier 
 with the bottom attachment consisting of a cluster of incandes- 
 
 FiG. 315. — Common faulty method of reading table lighting. 
 
 cent lamps under a flat opal reflector (see Fig. 151) is both in- 
 artistic and inefficient for this particular purpose. The bottom 
 cluster should be left off entirely or if used, should be placed 
 inside of a large opal dome. Such a small reflector as this over 
 a cluster of lamps allows too large a percentage of the light 
 to escape horizontally. The preferable arrangement would be to 
 have the chandelier with sockets pointing about 10 degrees 
 
■282 PRACTICAL ILLUMINATION . 
 
 from the vertical equipped with reflectors giving a distribution 
 such as shown in Figs. 19 and 21 which will concentrate most of 
 the light around and on the reading table. The opal reflectors 
 with which the chandelier arms are equipped in Fig. 316 are ex- 
 cellent for general distribution below the horizontal when placed 
 vertically as seen from Fig. 23, but do not concentrate enough 
 light below the tip of the lamp. 
 
 Fig. 317 shows a reading room which is an excellent example of 
 how not to do it as far as the lighting is concerned. The chandelier 
 over the table is equipped partially with upright incandescent 
 
 Fig. 316. — Poor design of chandelier for reading table. 
 
 lamps, and partially with lamps at an angle of 45 degrees in 
 etched glass shades. The arrangement permits about half the 
 light to escape toward the ceiling. Around the sides of the room 
 are brackets placed high. The location of the brackets is ex- 
 cellent, but their equipment is open to the same objections as 
 that of the chandelier. Too much light is allowed to escape to- 
 ward the ceiling and even if expense were no object the arrange- 
 ment is bad because it permits of too much undiffused glare 
 from incandescent filaments. The least that could be done in a 
 case like this would be to enclose all lamps in diffusing globes. 
 
THE LIGHTING OF LIBRARIES. 
 
 283 
 
 Much more light could be secured for reading purposes by point- 
 ing the lamps at a different angle, say downwardly about 10 
 degrees from the vertical, and placing them in deep reflectors 
 such as shown in Fig. 22 or 38. The area to be covered is large. 
 These reflectors would distribute light well over all angles below 
 45 degrees from the horizontal and would be deep enough so 
 that the light would not be directly visible except to those sitting 
 near the lamp. Were it undesirable to darken the ceiling as 
 much as would be done by deep opal reflectors, and if prismatic 
 
 '•"^ 
 
 
 -„] mkM. 
 
 ■-^^m'. 
 
 Fig. 317 — Wastefull)' lighted reading room. 
 
 reflectors were not used, dome, shaped reflectors of sand blasted 
 glass such as in Fig. 73, could be used which would not throw 
 so much light below the horizontal as opal reflectors, but would 
 nevertheless give considerably more than those now on the lamps 
 and at the same time be of very artistic appearance. 
 
 Fig. 318 shows a school room lighted by gas mantle burners. 
 In a school room it is desirable to avoid all unnecessary glafe 
 and provide a good working light at all desks". Chandeliers must 
 therefore be located at frequent intervals and equipped" with 
 
284 
 
 PRACTICAL ILLUMINATION. 
 
 shades or globes which will direct most of the Hght downward 
 in the immediate vicinity of the chandelier, while avoiding glare 
 as much as possible. This is fairly well accomplished in this 
 case by the use of diffusing globes which direct the light down- 
 ward in the manner shown in Fig. 120. It will be evident 
 from the spacing of the chandeliers that a better globe might 
 have been selected (such as Fig. 123) which would throw the 
 light more at an angle of 45 degrees instead of directly down- 
 ward, as is the case of the globe here used. The illumination is 
 good and the diffusion is excellent. 
 
 Fig. 318. — School room lighted by gas. 
 
 Fig. 319, showing the hbrary of the University Club, New York, 
 is an example where efficiency has been greatly sacrificed for the 
 sake of art. Chandeliers equipped with upright electric lamps 
 are much better as illuminators of the ceiling than as producers 
 of useful light. In order to deliver more light below the hori- 
 zontal, these chandeHers have been equipped with globes of the 
 kind, shown in Fig. 85 which accomplish that result as well as 
 it can be accompUshed with upright lamps on a chandelier of 
 this kind. On the whole, it is questionable whether such a 
 
THE LIGHTING OF LIBRARIES. 
 
 285 
 
 chandelier is sufficiently superior in artistic appearance to one 
 with pendant lamps to make the extra cost of illumination by 
 means of it worth while. 
 
 The tables are lighted as seen, by stand lamps placed upon the ta- 
 bles. The remarks already made about table lamps used for library 
 table lighting apply to the small stand lamps shown on the tables 
 between the book shelves at the left. As to the large ornamental 
 stand lamp shown on the central tables under the chandeliers, 
 
 Fig. 319. — Library of University Club, New York. 
 
 something further may be said. These lamps are intended both 
 for decorative and useful purposes. Where large decorative 
 shades of this kind are employed a very feasible and efficient 
 arrangement is to place inside of the shade two lamps pointing 
 in opposite directions at an angle of about 45 degrees. These 
 lamps can be equipped with concentrating reflectors, such as 
 shown in Figs. 20 or 32 and with frosted bulb lamps. When 
 placed in the center of a table the lamps should be pointed length- 
 
286 
 
 PRACTICA L ILL UMIKA TIOX. 
 
 Fig. 32(1. — Wasteful table and shelf lighting. 
 
 Fig. 321. — -Good plan for shelf lighting. 
 
THE LIGHTING OF LIBRARIES- 287 
 
 wise of the table thus securing a good illumination over most of 
 the table. 
 
 Fig. 320 shows a very inefficient and unsatisfactory chandelier 
 for the lighting of a library reading table. This chandelier is 
 one of the kind designed entirely for ornamental purposes without 
 any regard to illuminating effect. The lamp on the bottom tip 
 of the chandelier would be of considerably more use to light 
 the table if it were equipped with an opal reflector instead 
 of being enclosed in a dense opal globe. The lamps on the 
 chandelier arms, instead of having globes or reflectors for direct- 
 ing the light downward, are bare so that a great deal of the light 
 is wasted on the ceiling as shown by Fig. 16. 
 
 Fig. 321 shows an excellent plan for the lighting of the shelves in 
 a library. On the first floor the lamps are placed in large opal 
 domes near the ceiling. A characteristic of the opal dome is that 
 it gives a wide distribution of light at all angles below 45 degrees 
 as shown in Fig. 21. On the second floor of this library the shelves 
 do not extend to the ceiling, consequently the lamps are placed 
 on pendants which bring them a little above the level of the tops 
 of the shelves. The three reflectors best suited to this work are 
 the opal dome, the fluted opal cone and the prismatic reflectors 
 which give a wide distribution such as Fig. 37. 
 
CHAPTER XXV. 
 CAR LIGHTING. 
 
 The efficient electric lighting of steam railroad cars is a matter 
 which has received but little of the attention which its importance 
 deserves. Steam railroad cars, up to present writing, have been 
 almost invariably equipped without any reference to delivering 
 the most useful illumination for the money expended and this too, 
 in spite of the fact that frequently the source of current supply is 
 a storage battery with a limited capacity. If more attention were 
 paid to the efficiency after the current was delivered to the lamp, 
 it would materially help out the amount of useful service that 
 could be obtained from a given size of storage battery, as cars 
 could be well lighted with less electrical energy. In the case of 
 electric cars, it is not so much the question of efficiency as of 
 general effect and good illumination without blinding glare. 
 
 Fig. 322 shows the interior of a private electric car which can be 
 taken up for discussion to illus trate some of the points which will 
 be brought out. This car in addition to having an ample supply 
 of incandescent lamps has also an arc lamp as seen which is entirely 
 out of place in a car of this kind and will be left out of account- 
 The incandescent lighting is accomplished by a row of frosted 
 bulb incandescent lamps along each side of the car. For an elec- 
 tric car where efficiency is no object the arrangement is fairly good. 
 The frosted bulbs very much reduce the glare which would be re- 
 ceived from a row of clear bulb lamps so located and there is no 
 particular need of increasing the efficienc}^ by the use of reflectors 
 although the shading and diffusing which would be accomplished 
 by an opal or prismatic reflector would be very desirable. If 
 this were a steam railroad car drawing its electric lighting current 
 in whole or in part from a storage battery, the economy aspect 
 of the case would be quite different. These lamps should then 
 be equipped with reflectors because of the greatly increased 
 amount of light that could then be delivered at the seats as well 
 
 288 
 
CAR LIGHTING. 
 
 289 
 
 as for diffusing and shading reasons. The location of the lamps 
 which would bring them over the seats of an ordinary railroad 
 coach is excellent. The reflector selected should be one which 
 would come down well over the lamps so as to diffuse the light 
 and prevent it from blinding anyone looking the length of the 
 car, such, for example, as in Figs. 20, 36, or 38. Opaque metal 
 reflectors would manifestly be out of place in car lighting so that 
 the choice lies between opal and prismatic glass if we leave out of 
 account reflectors of sand blasted glass which do not modify the 
 distribution of light sufficiently for this particular purpose. Opal 
 
 Fig. '^'22. — Private electric car. 
 
 reflectors have the advantage of being smooth and so much more 
 easily cleaned than prismatic reflectors. They have a disadvan- 
 tage strongly urged by some railroad men who have had exper- 
 ience with them that they show very badly the dirt which collects 
 upon them between cleanings. Prismatic reflectors, while not 
 so easily cleaned, do not look as badly when dirty as opal, because 
 the dirt collects in the bottoms of the prisms where it is not notice- 
 able. After all, the cleaning of a prismatic reflector is not very 
 difficult if the cleaner is provided with a stiff brush and brushes 
 lengthwise of the prisms. 
 
290 PRACTICAL ILLUMIXATIOX. 
 
 No ordinary flimsy shade holder should be used for holding re- 
 flectors in car lighting. A brass or bronze holder fitting over the 
 socket, as do the bronze husks shown in Fig. 322, should be used. 
 This holder is fastened directly to the ceiling independently of 
 the socket so that the weight of the reflector will not pull the 
 socket to pieces. On Pullman cars the berths interfere with 
 such a location of lamps and those for the general lighting of the 
 car must be placed in the center, being usually grouped around 
 the gas light. With lamps so placed, reflectors giving a maxi- 
 mum candle-power at about 30 to 45 degrees from the vertical 
 would be best assuming the lamps to be placed vertical. If 
 placed at angles radiating from the gas fixture, more concentrat- 
 ing reflectors should be used, and the lamps should be pointed 
 at the seats ; not up and down the aisle. 
 
 Some day-coaches are provided with lamps on brackets lo- 
 cated high between each window. In this case also, reflectors 
 giving a maximum candle-power from 30 to 45 degrees from the 
 vertical should be used if the lamps ere placed vertical. With 
 such bracket lights considerable light must necessarily be lost 
 because of the failure of the woodwork to reflect part of the light 
 which goes in that direction. 
 
 In the lighting of dining cars, the principal thing is to provide 
 a lamp at each table so arranged and so equipped with reflectors 
 as to concentrate nearly all of its light on the white table cloth. 
 With well lighted tables, the balance of the car will take care of 
 itself. This table lighting can be accomplished by artistic 
 brackets extending but a short distance from the side of the car, 
 oyer each table and with lamps pointed down toward the table 
 top. If desired, the subdued effect in table lighting so much 
 sought for recently can be obtained by proper shading of the 
 lamp, always remembering, however, that the concentrated 
 beam of light from the reflector on to the top of the table must 
 not be interfered with, however much the lamps. may be shaded 
 from the eyes of a person looking the length of the car. It goes 
 without saying that all brackets and other fixtures used in car 
 lighting must be short and substantial. 
 
 Fig. 323 is a remarkable example of the small difference there 
 may be between the right and wrong way of lighting a dining car. 
 In this illustration there is a two-light bracket over each table. 
 This bracket, however, instead of being properly equipped with 
 
CAR LIGHTING 
 
 291 
 
 a lamp in a deep reflector which would concentrate the light on 
 the table and shade it from the eyes of persons seated some dis- 
 tance away, has two lamps in reflectors giving a very wide dis- 
 tribution and these lamps are pointed away from the tables. 
 The result is that only a very small per cent, of the light of the 
 lamp is directed on the table top. The reflectors do not act to- 
 diffuse or shade the light and the whole arrangement is much 
 less efficient and much more trying on the eyes than it would 
 be with the very small changes mentioned. The gas lightings 
 
 Fig. 32.3. — Dining car. 
 
 in this car is by means of the standard Pintsch gas fixtures fas- 
 tened to the center of the car deck. This is about the only 
 feasible arrangement for gas lighting in a car and is less trying 
 on the eyes than some of the attempts to place gas or other lamps 
 along the sides of the car. Electric lamps should be placed 
 low along the sides of a car only when thoroughly shaded. 
 
 Berth lamps in sleeping cars, should always be shaded in 
 such a way that persons facing them will not have the light shin- 
 ing in their eyes. This can be done by proper designing of 
 
292 PRACTICAL ILLUMINATION. 
 
 fixtures. On some electrically lighted sleeping cars this shading 
 is accomplished very well. On others these berth reading lamps 
 are entirely exposed and are very trying on the eyes of anyone 
 who must face a row of them an entire evening. This is true 
 whether the person is actually looking at the light or not 
 
CHAPTER XXVI. 
 STORE LIGHTING. 
 
 The requirements of a small store in a country town naturally 
 differ considerably from those of the immense department stores 
 of the city. The requirements of the small store will be taken 
 up first. 
 
 When incandescent electric lighting was first introduced in 
 small stores the usual plan, and one that is still in very common 
 use in country and suburban villages was to hang a row of bare 
 incandescent lamps on drop cords at regular intervals the length 
 of the store over each counter, thus making in the ordinary store 
 which is 20 to 25 feet wide two or three rows of incandescent 
 lamps the length of the store. These lamps are usually hung 
 about six or seven feet above the floor or a little above the heads 
 of clerks and customers. The same general arrangement is 
 common with gas and oil lights in such stores. The whole 
 effect is crude and the apparent illumination is nowhere near 
 what it might be with other arrangements. In the lighting of a 
 store, unlike the lighting of a warehouse, it is not simply a 
 question of how much light can be obtained on counters and 
 shelves, but it is a question of the general impression the lighting 
 of the store makes upon the public. The lighting of stores is 
 partially in the nature of advertising, as is recognized by all the 
 most progressive merchants in both large and small towns. Let 
 us therefore investigate the little store lighted with the rows 
 of bare incandescent lamps or gas burners after the manner 
 described and see what can be done with it. The first objection 
 to be found is that a person looking the length of the store is 
 partially blinded by the bare lamps which come directly in his 
 line of vision. As a result, the store appears darker than it 
 really is because. of the fact that the bright light prevents the 
 seeing of things near the light for the reasons explained in the 
 chapter on Light and the Eye. The second objection is that 
 
 293 
 
294 
 
 PRACTICAL ILLUMIXATION. 
 
 the bare lamps look undeniably crude and sickly. This is more 
 true of incandescent electric lamps and open flame gas than of 
 gas mantle burners. To simply enclose these bare electric drop 
 lights or gas jets with globes of frosted or sand blasted glass would 
 improve the general effect decidedly. The crudity of the bare 
 lamp filaments or gas jets and the blinding effect produced by 
 them can thus be relieved and the light coming from the larger 
 surface of a diffusing globe is better from the standpoint of ad- 
 vertising and general appearance. The next step in improving the 
 lighting would be to raise the lamps from a height of six or seven 
 feet to a height of from eight to ten feet. When this is done reflec- 
 
 FiG. 324. — Store with efficient lighting. 
 
 tors should be used as in Fig. 324, in order that sufficient Hght may 
 be dehvered on the counters. By using efficient reflectors over 
 the lamps, the downward illumination is increased more than 
 enough to compensate for the raising of the lamps to a higher 
 level, so that the illumination on the counter will be better than 
 it was before. At the same time the lamps will be raised out 
 of the line of. vision so that they do not interfere with seeing 
 across the store and the store will appear on the whole much bet- 
 ter lighted than with the lamps hung low without reflectors. 
 
STORE LIGHTING. 295 
 
 The distribution of light with the reflectors seen in Fig. 324 is 
 given in Fig. 45 and is probably the best that could have been 
 selected for that location. With this shape of reflector frosted 
 tip lamps shoul'd always be used. Opal reflectors, such as 
 in Figs. 19 and 21 would give a good light distribution, except 
 that they darken the ceiling more than is desirable for a cheer- 
 ful, well lighted effect. At the same time that the lamps are 
 raised the merchant may see fit to modernize his store lighting 
 still more by placing the lamps on ornamental pendants rather 
 than on the old drop cords. Fig. 324 shows a neat design of 
 chain pendant. If pendants are not used, round, reinforced 
 cord with dark silk covering offers considerable improvement 
 in appearance over ordinary lamp cord with twisted wires. 
 
 By none of the common methods of electric lighting can as 
 much good illumination be delivered to the counters of a small 
 store for a given amount of electric energy as by the method 
 just outlined, namely, by rows of pendant frosted incandescent 
 lamps equipped with proper reflectors and placed at frequent 
 intervals over the counters and about eight or ten feet above the 
 floor. The arrangement besides being efficient also looks well, 
 especially when frosted bulbs are used. The same principles 
 can be used in gas lighted stores by raising the gas burners and 
 placing reflectors or directing globes over them. The fact that 
 incandescent electric lamps by this method are placed so nearl}^ 
 above the things they are to illuminate more than compensates 
 for their lower efficiency as compared with other and larger electric 
 lighting units such as arc lamps, Nernst lamps and large high 
 efficiency incandescent lamps. In the case of gas the individual 
 burners are considerably more efficient than gas arc lamps or 
 cluster burners, as shown in Chapter XIV. As soon as 
 we begin to employ larger lighting units such as electric 
 arc lamps, Nernst lamps, gas arc lamps or high candle-power 
 incandescent lamps, the units must be raised higher above the 
 counters to avoid blinding effect. Because of their fewer 
 numbers, the large units must necessarily be placed a consider- 
 able distance apart as compared to small incandescent or gas 
 lamps and hence farther from the surfaces to be lighted between 
 lamps. This causes considerable loss in efficiency. 'There is 
 TieverthelesS a certain advertising value about large lightin'g-\inits 
 ^vhich must'-Tiot -be overlooked and which is largely responsible 
 
296 PRACTICAL ILLUMINATION. 
 
 for their popularity in store lighting. Either electric or gas arc 
 lamps in a store attract a certain amount of attention to the 
 store and light the ceilings and high side walls to such an extent 
 that the store may appear more brilliantly lighted to a passer-by 
 than if it was equally well illuminated as far as the counters are 
 concerned by incandescent lamps. Then, too, it costs less to 
 install a few large lighting units in a store than to provide enough 
 outlets for a lot of small incandescent or gas lamps. If small, 
 common carbon filament incandescent lamps are bunched on a 
 few chandeliers or clusters they are at the same disadvantage 
 as any large lighting unit, that is, they are a long distance from 
 the parts of the store between chandeliers, and the inherent low 
 efficiency of the small lamps puts the chandelier arrangement 
 at a disadvantage as compared to arc, Nernst or high candle 
 power incandescent lamps placed at similar intervals. Shadows 
 are more pronounced and annoying with a few high candle 
 power units than with a large number of small ones. If a store 
 is flooded with a great abundance of light, the inefficiency of 
 lighting by large units as compared to distributed small lamps 
 will not be felt so much as if the total candle-power of lamps 
 provided is barely sufficient to give good results. 
 
 When we come to the use of large lighting units such as arc 
 lamps and Nernst lamps, it must be recognized at the start that 
 there is considerable difference in the distribution of light from 
 these various illuminants. Both the Nernst lamp and the large 
 high candle-power incandescent lamp equipped with a prismatic 
 reflector throw most of their light downward, as seen in Chapters 
 V and IX ; being very similar to each other in this respect. The 
 direct current enclosed arc lamp (see Chapter X) gives a large 
 amount of light at from 15 to 60 degrees below the horizontal. 
 Immediately below the lamp the candle-power is comparatively 
 low. This suits it specially for lighting over a large area. With 
 incandescent lamps equipped with reflectors or with Nernst 
 lamps, each lamp should be placed as nearly as possible above 
 the portion of the store which is to be lighted by it. A direct 
 current arc lamp can be expected to light a larger area. 
 
 The alternating current arc lamp, except when equipped with 
 a large and properly shaped reflector, (and it frequently is not) , 
 delivers so much of its light at angles near the horizontal as to 
 put it at a decided disadvantage in store lighting when compared 
 
STORE LIGHTING. 297 
 
 with the Nemst and the direct current arc lamp. With only alter- 
 nating current supply available therefore and if large units are tO' 
 be used, the maximum eflficiency is to be expected in store lighting, 
 by the use of Nemst lamps rather than ordinarj'^ types of alternat- 
 ing current arcs. The Nemst lamp is also steadier. The "gas- 
 arc " lamp, which is a four burner gas mantle lamp has a natural 
 distribution of light very similar to the alternating current electric 
 arc lamp which puts it at a disadvantage in store lighting as far 
 as the percentage of its light thrown on the counters is concerned. 
 This disadvantage can only be compensated for by providing a 
 good deal more light than is necessary in the upper portions of 
 the store in order that the lower portions may be properly illum- 
 inated. The price of gas may be cheap enough as compared 
 with the price of electric power to com.pensate for this in some 
 cities. There is also the question of the relative cleanliness and 
 convenience of the two kinds of light and of the heat and products 
 of combustion from the gas lamp. The heat may be an advantage 
 in some seasons and a disadvantage in others. These are questions 
 which it is out of the province of this book to discuss. The dis- 
 tribution of light from electric, gas and Nemst lamps is given in 
 previous chapters. Users must weigh these results and draw 
 conclusions according to their particular conditions and require- 
 ments as no general principles can safely be laid down, except 
 that it is usually possible with electric lamps to deliver a larger 
 portion of the total light produced where it is needed than with 
 gas or oil lamps. 
 
 There is a great difference of opinion both among merchants 
 and illuminating experts as to the color of light which should be 
 used in stores of various kinds. It is a common opinion 
 that the light which nearest approaches daylight is best for cloth- 
 ing and dry-goods stores. In other stores it is not usually con- 
 sidered that the color of artificial light used makes much difference. 
 There are some who contend that even for dry-goods stores 
 an artificial light similar to daylight is not necessary. 
 
 That the color of a piece of goods depends upon the kind of 
 light in which it is seen and that in some cases a change in the 
 kind of light makes radical changes in the apparent color of a piece 
 of goods is a scientific fact upon which there is no room for 
 argument . Some go so far as to maintain that goods should 
 be sold under the kind of light under which they will or- 
 
208 PRACTICAL ILLUMI\'ATIOK . 
 
 dinarily be used. On this theory, goods to be used mainly in 
 dayUght should be sold under daylight or artificial light which 
 most nearly approaches white light, while goods ordinarily worn 
 in the evening under the yellow light of electric incandescent or 
 open gas flames should be sold under a similar yellow light, 
 There are others who would lay aside the question of matching 
 or selecting colors entirely and consider the question simply from 
 a psychological or advertising standpoint. Some consider that 
 yellow light like that from an electric incandescent lamp or open 
 gas flame has a certain warmth and cheerfulness about its ap- 
 pearance which gives a more pleasing effect that can be obtained 
 with a whiter light. Those who hold this theory maintain that 
 a white light is cold and cheerless in its apppearance. On the 
 other hand, some think the white light the more cheerful of the 
 two. As this is a question upon which even the authors dis- 
 agree and upon which there is room for so much variety of per- 
 sonal opinion, it will not be discussed further here except to 
 point out some of the established scientific facts in regard to 
 the color of different illuminants. ■ ■ 
 
 Of the common artificial illuminants, the old-fashioned open 
 arc lamp approaches nearest to white light, or daylight. Day- 
 light itself varies considerably. There is a greater proportion 
 of red in its rays when the sun is low thail at noon-day because 
 the rays when the sun is low must pass through many more miles 
 of air; the violet rays being absorbed by the air more than the 
 red and yellow rays. 
 
 The modem enclosed arc lamp does not give a light as nearly 
 approaching daylight as the old-fashioned open arc. The ordi- 
 nary enclosed arc has an excess of A^iolet. This excess of violet 
 can, however, be largely corrected by the use of opal enclosing 
 globes which filter out some of the excess violet rays. The use 
 of small carbons in enclosed arc lamps also tends to make the 
 light whiter and also to make it steadier, both of which objects 
 are very desirable. 
 
 Next after electric arc lamps in approach to daylight come 
 acetylene gas, mantle biimer gas lamps, tantalum incandescent 
 [^ndNemstelectriclamps. The m,antle burner gas lamp has a 
 slight excess of green in its make-up which, like the excess of 
 violet in the enclosed electric arc may interfere with color 
 selection and matching unless it is compensated for by the use 
 
STORE LIGHTING. 299 
 
 of mantles and glassware designed to accomplish this. Certain 
 mantles are said to do this to a large extent and opal enclosing 
 globes also have a tendency to filter out the green rays just as 
 they do the violet rays in the case of the enclosed electric arc. 
 
 In the case of the Nemst, high efficiency Gem carbon 
 filament, and tantalum lamps we have illuminants which are 
 midway between the yellow of the ordinary incandescent, open 
 gas, and kerosene flames and the whiteness of daylight. They 
 may be considered as in the nature of a compromise between 
 these two. Their steadiness is in their favor when compared 
 with electric arc lamps. 
 
 The ordinary carbon filament electric incandescent lamp, 
 open gas jet and kerosene lamp are about the same in color and 
 are farthest from white light of any of the common illuminants. 
 
 Taking up now some practical examples of store lighting 
 arrangements, we have in Fig. 324 already mentioned, the gen- 
 eral scheme of small store lighting which probably gives the best 
 illuminating and artistic results for the money expended. In 
 this case 187-watt Gem filament, high efficiency, frosted 
 incandescent lamps in distributing prismatic reflectors are sus- 
 pended in rows over each counter by ornamental chain pendants. 
 Lamps of 125-watts would answer very well here, but of course 
 would not flood the store with light as it is now and some of 
 the advertising value of good lighting would be lost. In this 
 store the wiring is in conduit placed on the ceiling in such a way 
 as to be neat and unobtrusive. This arrangement places the 
 lamps immediately over the places to be lighted, hence the high 
 efficiency of the arrangement. The lam.ps are hung approx- 
 imately 12 feet apart. 
 
 Fig. 325 shows a store fitted up as an automatic vaudeville 
 parlor. The principal lighting is by means of lamps studded 
 along the ceiling and by a row of brackets along the left side. 
 The idea in this case of course is to attract attention to the place 
 by the wealth of light displayed rather than to secure useful 
 illumination. The arrangement is, however, very good for 
 useful illumination. Enclosing globes, such as shown in Fig. 79, 
 are used, which direct the light downward as well as diffuse it, 
 so that it is inoffensive to the eye. The lamps are so well dis- 
 tributed that every corner of the room is well lighted. 
 
 It is sometimes the case that the lighting of a store or display 
 
300 
 
 PRACTICAL ILLUMINATIOX. 
 
 room by means of a large number of distributed incandescent 
 lamps which give the best results from a purely economical 
 standpoint is objected to by the owner on the ground that only 
 a few outlets for electric wires have been provided in the original 
 wiring of the building. The owner may dislike to go to the 
 expense of rewiring the building or does not wish to run wires 
 in exposed conduit or moulding as would be usually necessary 
 to supply a lot of scattered incandescent lamps. One neat way 
 of overcoming this difficulty is shown in Fig. 326 which is the 
 carpet salesroom of a large Los Angeles store. Here the incan- 
 
 FiG. 325. — EfBciency and decoration well combined. 
 
 descent lamps are suspended on drop cords from a handsomely 
 finished rectangular brass pipe which is hung horizontally near 
 the ceiling with its middle under each wire outlet. This gives 
 well distributed lighting, although in this case a more even 
 distribution of light over the floor could have been obtained by 
 the use of a prismatic reflector which would not concentrate 
 the light so much as the one which is seen in the illustration 
 which is a very concentrating form, similar in light distribution 
 to Fig. 35. The distribution shown in Figs. 19, 29, 33, 37, and 
 43 would be better. j. 
 
STORE LIGHTING. 
 
 301 
 
 A modification of Fig. 326 which is sometimes used is to run 
 wires in small brass tubes two or more wa3^s from each outlet, 
 these tubes being similar to those used in fixture arms, and ter- 
 minating in lamp sockets. 
 
 Fig. 327 shows a very common method of store lighting by 
 means of direct current enclosed arc lamps with opal outer globes. 
 This may be considered as a good example of light properly 
 placed. The direct current enclosed arc (see Fig. 184) 
 gives a distribution of light which will in this case cause 
 the maximum light to be thrown in the direction of the shoe 
 
 J.' ILi. -J^U.- 
 
 — i^lOLl 1 
 
 buted lighting from few outlets. 
 
 cases, which is desirable. Fro.n the shoe cases it will be re- 
 flected back to the customers' chairs, which is also desirable. 
 The lamps have opal outer globes which tend to filter out the 
 violet of the enclosed arc light and also difiiuse the light so as 
 to make it easy on the eyes of clerks and customers because the 
 .light comes from the entire surface of the globe rather than 
 from a small area. As before explained, a store lighted with 
 arc lamps as in Fig. 327 is not as economical!}' lighted as it would 
 be with rows of pendant incandescent lamps with proper re- 
 flectors well distributed over the store. There is, however, a 
 
302 
 
 PRACTICAL ILLUMINATION. 
 
 certain advertising value in the large arc lamp units, especially 
 when they are seen from the street even though they may not 
 be as good from an economical and artistic standpoint as dis- 
 tributed incandescent lamps with frosted bulbs. 
 
 Fig. 328 shows a good way of lighting a small meat market 
 by means of one direct current arc lamp. The distribution of 
 light from the direct current arc lamp is such as to throw a 
 good intensity on the counter and toward the meat hooks. 
 The lamp is equipped with a reflector and an opal inner globe 
 only. In a case of this kind, an opal outer globe is not as neces- 
 
 FiG. 327. — Arcs well placed in shoe store. 
 
 sary as in Fig. 327 because the lamp is so well out of the line 
 of ordinary vision of clerks and customers. It is nevertheless 
 desirable for the reasons mentioned in connection with Fig. 327. 
 Fig. 329 is an example of artistic and efficient lighting in the 
 city ticket office of the Chicago & Alton Railway at Chicago. 
 The principal part of the useful lighting is accomplished by. the 
 four Meridian lamps at the corners of the ceiling fixtures. The 
 distribution from these lamps which is shown in Fig. 29 is ex- 
 cellent for the purpose. Frosted art glass is used inside the 
 fixture proper. The lamps provided inside of the fixtures are 
 
STORE LIGHTIXG. 
 
 303 
 
 '^ 
 
 M 
 
 p^tf 
 
 
 HI 
 
 
 1 
 
 
 
 iS 
 
 t^^W^ ^ ^jSBgf^^m 1 
 
 w 
 
 K^ 
 
 ■J 
 
 H 
 
 ti^t^t^ 
 
 K9H' 
 
 5^^ 
 
 Wm 
 
 si9h 
 
 « 
 
 t - ' ' ^/^I^H^H 
 
 WM 
 
 ^^^^j 
 
 P^HH^Hpl^r^ y-l'- 
 
 . ^ 
 
 •^f^fir^titm/ 
 
 ^ 
 
 H 
 
 Fig. 328 — Arc well placed in meat market. 
 
 Fig. 329 — Artistic and efficient lighting in ticket office. 
 
304 
 
 PRACTICAL ILLUMINA Tli K\\ 
 
 mainly decorative in their purpose, but could be made more 
 useful by having reflectors placed over them. The lantern 
 bracket lamps on the pillars are provided with ground glass 
 which makes them at the same time pleasing and inoffensive 
 to the eye. These brackets are of course principall}' decorative 
 in purpose. 
 
 Fig. 330 shows the Hghting arrangements of a grocery store 
 where a large number of incandescent lamps have been in- 
 stalled for decorative and advertising purposes on the arches 
 
 Fig. 330 — Decorative but inefficient store lighting. 
 
 along each side of the room. This is a case where frosted bulb 
 lamps should be used to avoid the glare from so many lamps. 
 In this case the electric lamps on the chandeliers cut very little 
 figure in the total illumination as there are so many on the 
 arches. The lamps on the arches would be more effective in 
 illuminating the store if equipped with opal or prismatic re- 
 flectors, such as shown in Figs. 20, 32, and 34, although as a 
 matter of fact there is such an abundance of them that when 
 they are all lighted there is ample illumination. The same 
 
STORE LIGHTING. 
 
 305 
 
 useful illumination could be obtained with smaller candle 
 power lamps by using reflectors. 
 
 Fig. 331 is a sample of how not to attempt to light a store. 
 The lamps along the top of the medicine shelves are provided 
 with a concentrating type of prismatic reflector, such as shown 
 in Fig. 32 which is very efficient for certain purposes, but which 
 
 Fig. 331 — Wrong place for reflectors. 
 
 is the very thing which should be avoided here for two reasons. 
 In the first place, they reduce the amount of light which would 
 shine down on to the shelves and throw the light clear across 
 the store at an angle which delivers the light nowhere that it'is 
 needed. In the second place, the concentrated light from these 
 reflectors is thrown directly in the faces of customers standing 
 in the middle of the store looking toward the shelves, with the 
 
306 
 
 PRACTICAL ILLUJMINATION. 
 
 result that the shelves will appear poorly illuminated because 
 of the blinding effect of these lamps. At the upper left hand cor- 
 ner of the picture we see that there must be an ornamental chan- 
 delier high up near the ceiling provided with bare lamps. With 
 the chandelier in such a location and the lamps pointed as they 
 are, this is exactly the place to use reflectors. The proper thing 
 to do in this case would be to take the reflectors off the lamps 
 over the shelves and place them on the chandelier lamps where 
 they would be effective in throwing light down in the store from 
 the chandelier. Frosted bulb lamps without reflectors should 
 
 Pig. 332 — The wrong reflectors for the place. 
 
 be used along the top of the shelves, if no change is made in the 
 position or angle of the sockets. If the best results are wanted 
 from these shelf lamps, short bracket fixtures which will permit 
 the lamps' being pointed down about 10 degrees from vertical 
 should be used and the present reflectors retained. This will 
 throw the maximum light on shelves and counters. 
 
 Fig: 382 shows an amusing mistake in the glassware equipment 
 of a' stationery store. This store has very artistic five lamp 
 chahdeliers with lamps pendant. On the chandelier arms are 
 foiir lamps equipped with prismatic reflectors of a form which 
 
STORE LIGHTIXG. 
 
 3or 
 
 concentrates most of the light immediately under the chandelier 
 even more than shown in Fig. 31. Now, in a store of this kind 
 with the chandeliers as far apart as they are here, a concentrating 
 reflector is the very thing to be avoided since the light from each 
 chandelier must cover a large area and must also light up the 
 stock on the shelves What is needed here is some kind of 
 diffusing globe such as Fig. 83 or 81 or a type of prismatic re- 
 flector which will give a wider distribution of light, such as that 
 
 Fig. 333. — Bad mixture of colors. 
 
 shown in Figs. 37 and 43. If the reflector used here were of a 
 type which exposed the lamp bulb, as in Fig. 44, a frosted bulb 
 lamp would be decidedly advisable, even if the lamp were not 
 over 16-candle-power. 
 
 Fig. 333 is a^ picture of a Chicago lunch room originally equip- 
 ped with incandescent lamps which proved insufficient and were 
 later supplemented by arc lamps. The arc lamps are well 
 placed for the lighting of the main floor and balcony. The 
 mixture of arc and incandescent laraps or in fact, the mixture 
 
308 
 
 PRACTICAL ILLUMINATION. 
 
 of any two different illuminants does not usually give pleasing 
 results The effect is usually to make the smaller, yellower 
 lighting units look sickly and weak. Thus incandescent lamps 
 in the presence of arc lamps look yellow and ineffective although 
 if the arc lamps were removed, thus removing the contrast of 
 color, the incandescents would be very pleasing. The incandes- 
 cents might have given better results if those along the gallery 
 had been equipped with opal or prismatic reflectors. 
 
 Fig. 334 shows some of the most remarkable electric lighting 
 
 
 Fig. 334. — Absurd fixtures. 
 
 fixtures ever put in a store. The opal dome reflectors on the 
 arms of the chandeliers are placed so high that the entire lamp 
 is exposed and the effect is ridiculous, to say nothing of the fact 
 that the value of these opal domes as reflectors is thus largely 
 lost. Opal dome reflectors (see Fig. 21 )properly placed on these 
 chandeliers would be very effective, especially if frosted bulb 
 lamps were used. The opal globes on the middle of the chan- 
 deliers must be regarded as mainly ornamental since they absorb 
 probably about 40 per cent, of the light and in this case the top 
 and bottom fittings prevent very much of the useful light that 
 
STORE LIGHTING. 
 
 309 
 
 remains from escaping. The bracket lamps on the pillar placed 
 upright are also very ineffective since a large portion of their 
 light is thrown to the ceiling. Pendant bracket lamps would 
 have given considerably better results. Ground glass rather 
 than opal globes should be used on them since opal wastes so 
 much light and unless very dense, does not give a pleasing effect. 
 Fig. 335 shows an arrangement of chandehers for lighting a, store 
 which is very pleasing if employed with ground glass globes on the 
 chandeliers, but decidedly otherwise if opaline globes are employed 
 
 Fig. 335. — Fur store. 
 
 which allow the lamp filaments to show through with a sickly 
 red color. With large enclosing globes of this kind it is always 
 advisable to use pendant lamps as shown because of the chances 
 that the globe holders will sag in time, if the lamps are placed 
 at an angle. The arrangement shown in Fig. 335 is not especially 
 efficient because plain diffusing globes of this kind do not have 
 the power of directing the light downward. A more efficient 
 arrangement would be to equip this type of fixture with frosted 
 bulb lamps and prismatic reflectors giving a distribution such as in 
 Figs. 29, 37, or 45 ; or globes such as shown in Figs. 81 or 89. 
 
310 
 
 PRACTICAL ILL UMIXA TIOX. 
 
 With the latter lamps should be placed low in the globe so ;is to 
 give a wide distribution. 
 
 Fig. 336 shows the hghting arrangements in a builders hard- 
 ware salesroom. The chandeliers have lamps pointed at an angle 
 and equipped with opal bell reflectors such as in Fig. 24. Since 
 these reflectors give their maximum candle-power at an angle of 
 about 35 degrees below the horizontal, it is evident that an un- 
 necessary amount of light is being thrown to the walls above the 
 show cases. These lamps should be pointed straight down and 
 
 Fig. 336. — llanlware store. 
 
 frosted bulbs should be used, since so much of the lamp is exposed. 
 Fig. 337 shows the lighting arrangements of a picture frame 
 salesroom in which it is important to have plenty of light on the 
 high side walls because of the goods on display there. For this 
 reason, the arrangement shown is fairly good as far as the lighting 
 by incandescent lamps is concerned. Bare frosted incandescent 
 lamps are arranged around the ceiling far enough back from the 
 walls so that they do not interfere with the vision of persons 
 looking up at the samples along the high side walls. These 
 
STORE LIGHTING. 
 
 311 
 
 lamps would be even more effective if placed in reflectors giving 
 a light distribution such as in Fig. 29 or 43, as they would then 
 throw almost as much light sideways and considerably more 
 down than if bare. 
 
 Fig. 338 shows a very inefficient arrangement for jewelry 
 store lighting. Above the counters are clusters of three bare 
 incandescent lamps which are not only atrocious in appearance, 
 but offensive and injurious to the eyesight. High up near the 
 
 ^**- 
 
 
 
 
 1 
 
 
 1^ 
 
 ^;?^-:.-.;^l:^^^=- ■ :-'" 
 
 
 ill IH; iifB' ' l^kiM 
 
 tMi 
 
 1- M 
 
 
 
 ttPi^ 
 
 i 
 
 
 
 1 
 
 \ 
 
 
 ii^BfcSi 
 
 
 r 
 
 ^m, 
 
 Wm 
 
 ill 
 
 W--^ 
 
 Fig. 337. — Picture frame store. 
 
 ceiHng are rows of incandescent lamps which would be very 
 effective for the general lighting of the store if only they were 
 equipped with the proper opal or prismatic glass reflectors giving 
 a distribution such as in Figs. 29, 31 or 33. They have, however, 
 a type .of bell glass shade which is worth little for directing the 
 light downward. A concentrating type of reflector should 
 be used in this location. By doing this and either omitting the 
 three lamp clusters or changing them to downwardly pointed 
 bracket lights equipped with deep reflectors which would throw 
 
312 
 
 PRACTICAL ILLUMINATION. 
 
 Fig. 338. — Inefficient lighting in jewelry store. 
 
 Fig. 339. — Decorative lighting in candy store. 
 
STORE LIGHTING. 
 
 313 
 
 the light down on the counter, the Hghting of this store could be 
 made both agreeable and efficient. 
 
 Fig. 339 shows the lighting of a candy store by means of a 
 great ntmiber of studded ceiling lamps. The effect aimed at 
 here is fully as much that of advertising as of useful illumination. 
 A ceiling studded with so many lamps has a distinct advertising 
 value. The main criticism to be offered is that in some of the 
 sockets clear bulb lamps and in others frosted lamps have been 
 employed which considerably detracts from the pleasing and 
 artistic effect. In a case of this kind nothing but frosted bulb 
 lamps should be used. If a merchant desired to light his store 
 
 Fig. 340. — Crude equipment in hat store. 
 
 by means of studded ceiling hghting of this kind without spend- 
 ing so much for advertising effect as in this case, fewer lamps 
 could be used and by equipping these with opal or prismatic 
 glass reflectors and frosted lamps, good distribution and high 
 efficiency could be obtained. Concentrating reflectors should 
 be used. 
 
 Fig. 340 shows a high-class hat store lighted with very cheap 
 fixtures. These fixtures are clusters of incandescent lamps 
 under metallic reflectors. Metallic reflectors have no place in 
 this kind of work as they do not look well. They also darken 
 the ceiling too much. The substitution at each outlet of a plain 
 
iU 
 
 PRACTICAL ILLUMIXATION. 
 
 three-light chandelier in place of the present clusters and placing 
 on each chandelier three 32-candle-power frosted bulb lamps 
 in opal or prismatic reflectors pointed about 15 degrees from 
 vertical would improve appearances and give better efficiency 
 than the present six -light clusters without increasing the cost 
 of operation and without greatly increasing the first cost. 
 Besides the bad appearance of the clusters in such surroundings, 
 the glare received from them is likely to detract from the ap- 
 parent general illumination by the blinding effect on the eye as 
 
 Fig. 341, — Handsome and efficient equipment in hat store.. 
 
 previously explained. The}^ deliver too much light horizontally 
 as can be seen from Fig. 151. 
 
 Fig. 341 shows a very pleasing arrangement of fixtures in a 
 store of a similar character to that shown in the previous illus- 
 tration. Large Holophane spheres are suspended by chain 
 fixtures. In such a case the lamps inside the spheres should 
 be pointed down and equipped with reflectors. A similar 
 scheme might be used with ground glass spheres. The best 
 plan with spheres is usually to have but one lamp of high candle- 
 
STORE LIGHTING. 
 
 315 
 
 power with a large flat opal or prismatic reflector in the center 
 of the sphere. 
 
 Fig. 342 shows some artistic lighting arrangements in the 
 furniture department of a large store. This is a case where 
 light is wanted in all directions both horizontally to light the 
 galleries and downward, so that any type of diffusing globe is 
 permissible. In this case a globe such as shown by Fig. 82 is 
 used. The globe shown in Fig. 89 would be better yet. 
 
 Fig. 343 shows a handsome form of box fixture with art 
 
 Fig. 342. — Handsome fixtures in department store. 
 
 glass sides used in a Chicago cigar store. By equipping the lamps 
 inside this fixture with reflectors a very good arrangement 
 might be made. In addition to its artistic appearance, the strong 
 point about this fixture is that it shades the eye from the direct 
 rays of the light; only a diffused light passing through the art 
 glass. The boxes are rather deep and narrow, but this is 
 proper for covering a small area underneath the fixture as in 
 this case. The same general plan of fixture can be used in some 
 other places where the shading of the sources of light is required 
 as, for example, in church lighting as already referred to. 
 
316 
 
 PRACTICAL ILLUMIXATION. 
 
 Fig. 344 shows the lighting of a picture salesroom. The 
 problem here is very similar to that of lighting a picture gallery 
 of any kind. The general lighting is by means of hemispheres 
 placed in the middle of the ceiling. The picture lighting is 
 accomplished by incandescent lamps placed in a trough sus- 
 pended from the ceiling as shown. In the lighting of individual 
 pictures in homes and public buildings, a short trough reflector 
 
 Fig. 343. — Box fixtures in cigar salesroom. 
 
 with outside finish corresponding to the finish of the picture 
 frame can be mounted directly above the picture frame. It is 
 advisable to use frosted bulb lamps in these trough reflectors 
 to avoid streaks being cast on the picture which may interfere 
 seriousty with its appearance. 
 
 ■Fig. .34.5 shows a small store lighted by a gas arc lamp. The 
 lamp is hung high so as not to interfere with seeing across the 
 
STORE LIGHTING. 
 
 U7 
 
 Fig. 344. — Good lighting of picture salesroom. 
 
 Fig. 345. — Proper placing of gas arc lamp. 
 
31S 
 
 PRACTICAL ILLUMINATION. 
 
 store. This is a good example of how a gas arc lamp should be 
 used, except that an alabaster globe would be better. 
 
 Fig. 346 shows a hat store lighted with gas arcs equipped 
 with alabaster globes. In a store of this length, where the arcs 
 
 Fig. 346. — Bad mixture of lights. 
 
 are placed low, the use of alabaster globes is to be highly com- 
 mended. This is another example of where gas arcs are used 
 efEectivel}^ Attention is called to the fact that an examination 
 of this store will show that gas arcs, electric arc lamps, and 
 electric incandescent lamps are used. This is highly repre- 
 
STORE LIGHTING. 
 
 319 
 
 hensible if all are lighted at the same time. We should he 
 very careful at all times never to mix our color effects. Nothing 
 in the lighting line is perhaps less attractive, than a place where 
 we have the blue of an electric arc, the green of a mantle burner, 
 and the yellow of an incandescent electric lamp. 
 
 Fig. 347 shows a millinery store equipped with bare mantle 
 burners. The distribution of light is good, as will be seen by 
 referring to the distribution curve of this combination shown 
 in Fig. 112, but the glare is inexcusable. This should be com- 
 pared with the store in Fig. 349, where opal chimneys are used 
 
 Fig. .347. — Too much glare. 
 
 and the distribution is not as good as in Fig. 347, but where, 
 owing to the dense globes used, we have nearly perfect diffusion 
 and a light which is soft and easy to see by. 
 
 Fig. 348 shows a candy store equipped with combination fix- 
 tures for gas and electricity. The gas lighting is excellent as 
 far as diffusion is concerned, as Holophane shades are em- 
 ployed. The electric lamps have concentrating prismatic re- 
 flectors like those shown in Fig. 36. If frosted bulb lamps 
 were employed to reduce the glare from the lamp filaments, 
 this arrangement would be about the best one possible for store 
 lighting with chandehers of this kind. 
 
320 
 
 PRACTICA L ILLUMIXA TIOX. 
 
 Fig. 348. — Candy store with chandeliers. 
 
 Fig. 349. — Good equipment for gas chandehers. 
 
STORE LIGHTIXG. 
 
 321 
 
 Fig. 349 shows a drug store lighted by gas mantle burners 
 in opal chimneys. The opal chimneys, while absorbing consider- 
 able light, as seen in Chapter VI, nevertheless are desirable for 
 reducing the intensity and softening the light so as to make 
 it easier on the eyes. A four-lamp chandelier of this kind is 
 more efficient than a four-lamp gas arc lamp because the burners 
 do not interfere with each others light as much. 
 
 Fig. 350 illustrates the proper use of large Nernst lamps in 
 lighting a large department store. Since the light distribution 
 
 Fig. 350. — Proper use of Nernst lamaps. 
 
 from the 'Nernst lamp is mainly downward, as seen in Chapter 
 IX, they should be placed high in store lighting to produce as 
 even illumination as possible, as well as to avoid the blinding 
 effect of lamps hung low. If small Nernst lamps are used, 
 the)^ can be applied to advantage as are the incandescent lamps 
 in Fig. 324. 
 
 The quantity of light or amount of electrical energy required 
 in store lighting necessarily varies considerably according to the 
 ideas of the merchant and his customers as to what constitutes 
 
322 
 
 PRA CTICA L ILL UMINA TION . 
 
 a well lighted store. The standard is usually higher in cities 
 than in small towns. With the most efficient lighting arrange- 
 ment possible, such as for example that shown in Fig. 324, it 
 is not advisable to allow less than 0.5 watt per square foot of 
 fioor area, and it is usually best to allow considerably more. 
 
 o 
 
 O 
 
 o 
 O 
 
 a 
 
 Mr. W. D'A. Ryan gives the requirements for high class store 
 lighting b}^ arc lamps as 0.5 to ] watts per square foot for hard- 
 ware and shoe stores and 1 to 1.5 watts for department stores. 
 One of the departments of Marshall Field & Company's retail 
 store at Chicago, Fig. 351, has been the subject of (.areful 
 
STORE LIGHTING. 323 
 
 study by several engineers. This was equipped with incan- 
 descent lamps studded over the ceiling so as to give 
 a fairly even light distribution. The lamps used were 
 the Gem, graphitized carbon filament, high efficiency lamps. 
 The reflectors were " inverted bowls " of prismatic glass, deep 
 enough to cover and shade the lamp filaments from the eyes 
 of persons looking the length of the store. Their shape and 
 light distribution were similar to Figs. 37 and 38, except 
 that the reflectors were necessarily larger to fit the large 
 lamps and also gave a slightly larger proportion of their light 
 directly downward. Each ceiling panel contained four lamps, 
 of the 125-watt size and one of 187 watts. Lamps were, of 
 course, pointed straight down. The power required for each 
 panel was therefore 687 watts. The size of each panel was 
 21 by 22 feet and the ceiling height 15 feet. The watts per 
 square foot were about 1.5. Ceilings and pillars were white 
 and carpets and counters dark finish. It was found by 
 -various tests that the illumination in foot candles at the 
 counters averaged about 3. This is brilliant illumination at 
 a very reasonable expenditure. 
 
CHAPTER XXVII. 
 SHOW WINDOW LIGHTING. 
 
 Some store windows are lighted to display the goods and others 
 are lighted simply to attract attention to the place without re- 
 gard to what there may be on display there. For example, in 
 the highest class dry-goods stores the handsome display of goods 
 in the show windows is in itself the best advertisement the store 
 can offer. In other cases there may be nothing especially attrac- 
 tive to display in the window so that the main thing is to 
 provide light of sufficient quantity and kind to attract attention 
 to the place. In the latter case, the lighting can be considered as 
 purely decorative; in the former case it is strictly utilitarian. In 
 many windows these two are combined. 
 
 A common mistake in lighting show windows is to place a 
 border of bare lamps entirely around the window with perhaps 
 a lot of other lamps interspersed with the goods in the window. 
 The effect of this is to practically prevent a person who is across 
 the street from getting a very good idea of the goods on display 
 in the window because of the blinding effect of the exposed lamps. 
 We, know from what has been said previously in regard to light 
 and the eye that it is impossible .to look past a source of bright 
 light and clearly distinguish objects in the vicinity because the 
 pupil of the eye . automatically contracts in self-defense in the 
 presence of such bright light. 
 
 Fig. 352 gives a good idea of the way a show window lighted 
 with a border of incandescent lamps appears to a person across 
 the street. It is impossible to see what is around the edges of 
 the window, while the goods in the center are barely distinguish- 
 able. 
 
 In striking contrast to this is Fig. 353 which illustrates a show win- 
 dow lighted from above by concealed lamps . In this everything on 
 display is brought out to the best advantage. For the lighting 
 of such windows as shown in Fig. 354 several kinds of reflectors 
 
 324 
 
SHOW WINDOW LIGHTIXG. 
 
 325 
 
 Fig. 352. — Blinding efifect of border window lights. 
 
 Fig. 353. — Window lighted to display goods. 
 
326 
 
 PRACTICAL ILLUMINATION. 
 
 may be used. The most common reflector is trough shaped such 
 as in Fig. 63 with either glass mirror or metal reflecting surface. 
 These trough shaped reflectors give good results with deep' 
 windows; that is, windows which extend far back from the 
 street in proportion to their height. In the great majority of 
 windows which are high and shallow cone shaped reflectors over' 
 each lamp which concentrate the hght over a narrow angle as in 
 Figs. 35, 53, 55, and 70 would give far better results than trough 
 reflectors. Whatever reflector is used, it shouldinthe highest class 
 of stores be well concealed from the ordinary passer-by and care 
 should be taken to have it pointed in a direction which will throw 
 
 Fig. .'-l.i-l. — Proper placing of conical window reflector. 
 
 the light on the goods to best advantage. It is common to see 
 reflectors mounted at angles which do not give the best results. 
 
 Fig. 354 shows the cross section of a show window indicating 
 the proper position for lamps when placed in cone shaped reflectors 
 for show window lighting. An opaque black band or its equiv- 
 alent can be placed across the top of the window to hide the lamps 
 from the street, or can be omitted if it is desired to expose this 
 row of lamps to attract attention. In the latter case, an opaque re- 
 flector would not be used, of course. Where the reflectors are out 
 of sight as in most of the high-class show window lighting their 
 appearance need not be considered. The only questions to be de- 
 cided in selecting reflectors for these places are questions of 
 
SHOW WINDOW LIGHTING. 
 
 327 
 
 durability and efficiency. Cone-shaped, opal and polished or en- 
 ameled metal reflectors may be considered as out of the race for 
 this work because of their comparative inefficiency. The choice 
 ■of cone shaped reflectors will lie, therefore, between mirror and 
 prismatic glass. Prismatic glass when used alone without any 
 covering will not concentrate as much light for show window 
 purposes as good mirror reflectors. By placing white cardboard 
 or any other good reflecting surface over concentrating prismatic 
 reflectors they can be made to give a performance very similar 
 
 Fig. 355. — Combined useful and decorative window lighting. 
 
 to the best forms of mirror reflectors. Mirror surfaces are subject 
 
 to depreciation with heat, age, and handling, although there is 
 ■considerable difference in mirrors in this respect. Reflectors in 
 
 the tops of show windows get very hot. Prismatic reflectors 
 ■of course do not depreciate with heat. Their cardboard covers 
 
 will have to be renewed occasionally on account of dirt. 
 
 It is frequently thought desirable by merchants to combine 
 
 the useful illumination of a- show window with a certain amount 
 «of display or decorative effect from the lamps themselves. This 
 
328 
 
 PRACTICAL ILLUMIXA TIOX. 
 
 can he done easily by placing, the lamps near the top of the wia- 
 dows in handsome globes and fixtures designed to direct the 
 light downward. Fig. 355 is an example of this. The globe 
 used gives a light distribution similar to that in Fig. 78. Some- 
 times the visible decorative lighting is supplemented by concealed 
 lighting from above for better lighting the goods. 
 
 Fig. 356 shows a window in which the principal object is tO' 
 produce a large display of light rather than to illuminate a lot 
 of goods. So much light is here used that the goods are well 
 illuminated in spite of the fact that some of the light is so low as- 
 
 f^f -irfryrf'rriE rj .#-^ 
 
 consuime™ 
 
 Fig. 'A.'iH. — A display of lights 
 
 to interfere with this result. A concentrating t5rpe of pris- 
 matic reflector is used 
 
 Fig. 357 shows another method of lighting a show window 
 which, if properly designed, would combine the decorative and 
 goods display features in one. As it is, it is neither effective 
 nor handsome. The mirror reflectors placed in the top of the 
 window for lighting the goods, although very efficient, are not 
 handsome enough .to be exposed. The lamps around the side 
 of the window are not provided with effective glass reflectors to 
 throw hght on to the goods, although handsome reflectors' for 
 
SHOW WINDOW LIGHTING. 
 
 329' 
 
 Fig. 357. — A window easily improved. 
 
 Fig. 358. — Gas lighted window. 
 
330 
 
 PRACTICAL ILLUMINATION. 
 
 this purpose could easily be obtained, such as in Figs. 20, 32, 34, 
 36 and many others. 
 
 Fig. 358 shows the Hghting of a window by gas where it is 
 desired not only to throw the Hght directly down, illuminating 
 the pictures at the bottom, but also to throw considerable light 
 sideways as well. As the lamps are directly in the field of view 
 
 Fig. 359. — Decorative and useful window lighting. 
 
 of the passer-by, it is absolutely necessary to use diffusing globes. 
 As it is necessary also to get a strong light downward, Holophane 
 Class A shades are used (see Fig. 120), which fairly well accom- 
 plish the purpose. If the height of the ceiling were sufficient, 
 this window could be much better lighted by placing the lights 
 high enough to be practically out of the field of vision and using 
 powerful reflectors to throw the light down. 
 
SHOW WI\rDOW LIGHTING. 331 
 
 Fig. 359 shows the window of a hat store where the main illum- 
 ination is furnished by electric lamps hidden in trough reflectors, 
 which are not visible in the picture. In order, however, to assist 
 in attracting the passer-by, lamps are placed in large diffusing 
 globes, which while adding little to the illumination of the goods 
 below, nevertheless serve to attract attention. 
 
 For a very small store in a country town a very simple but 
 effective combination window and display light ca"n be obtained 
 by hanging in sight near the top of the window a frosted bulb 
 lamp of 32 candle-power or larger in a concentrating prismatic 
 reflector, such as in Figs. 30, 34, 36 or 47 or many others. The 
 use of a frosted bulb lamp is here important to make the source 
 of light look large as well as for general better appearance. 
 
CHAPTER XXVIII. 
 SHOP AND FACTORY LIGHTING. 
 
 It is coming to be recognized more and more that good arti- 
 ficial lighting in a shop or factory where night work is done has 
 an important influence on the output. There is no better way 
 to increase the output of a factory during the hours of darkness 
 than by providing good general lighting. If the conditions as 
 far as diffusion of light are concerned are made as near daylight 
 as possible, daylight results in output can be obtained. 
 
 The most common method of shop and factory lighting when 
 electric lighting was first introduced was to either place a few 
 big arc lamps at infrequent intervals or to light by means of 
 bare incandescent lamps hung promiscuously around the benches 
 and machine tools. It is hard to say which of these methods 
 is the most unsatisfactory as both are very poor. The most 
 common method to-day is probably to provide for the general 
 lighting by means of arc lamps and also to have incandescent 
 lamps on fiexible cords at each bench and machine tool. The 
 most common mistake is that of failing to provide proper re- 
 flectors on these incandescent lamps. Metal reflectors as in Fig. 
 360 are usually advisable because of the danger of breakage. 
 Tin cones (Fig. 50) painted with aluminum paint or aluminum 
 cones (Fig. 52) or parabolic reflectors with frosted aluminum 
 finish (Fig. 59) are good for this. Painted reflectors can be 
 repainted when dirty and the frosted aluminum finish can be 
 restored with sand paper. Such reflectors are important not 
 only on account of the great increase in intensity of light on the 
 work obtained by their use, but even more because they shade 
 the light from the eye of the workman so that he is not blinded 
 by it and can see what he is doing. In case the lamps are used 
 around machinery and are enclosed in lamp guards to prevent 
 their breakage, it is not practicable to use the ordinary form 
 of reflector. In such cases a piece of asbestos paper slipped 
 
 332 
 
SHOP AXD FACTORY LIGHTING. 
 
 333 
 
 over half the lamp inside the guard or a piece of tin covering 
 half the lamp is advisable. Frequently these lamps are used 
 around machinery where the eye of the workman must be close 
 to the lamp and he is working under a severe handicap if he is 
 obliged to look at a bright light as he usually is. 
 
 After all, the problem of factory hghting is in many respects 
 similar to that of lighting large offices. It may be done by means 
 of a number of small units, such as incandescent lamps placed 
 over or near the machines and benches to be Hghted or it may be 
 done by large units such as arcs or vapor tube lamps in sufficient 
 
 Fig. 360. — Shop with metal cone reflector^— incandescent lamps 
 
 numbers to flood the whole place with light. The latter method 
 approaches nearest to daylight working conditions and so gives 
 the workers the greatest freedom and permits them to do the 
 most work. For this reason it is becoming popular. 
 
 Machine shop work requires an extremely good light, as good 
 in fact as drafting where an intensity of from five to ten foot 
 candles is none too much. It is also important that there be 
 few sharp shadows around tools or work and that the glisten 
 and glare from tools and machinery be a minimum. For this 
 reason the light giving surfaces of the lamps should be as large 
 as possible and if conditions permit it the walls and ceiling should 
 
3:U 
 
 PRACTICAL ILLUMINATIOX. 
 
 be kept white. Clear globe arc lamps would be the very worst 
 thing to use. The lamps best suited to factory Hghting in view 
 of these requirements are either mercury vapor tube lamps 
 where the Hght comes from long tubes, or arc lamps with concen- 
 tric difEusers or large opal globes and large reflectors, where the 
 light comes from a large globe and reflector surface rather than 
 from a small point. With any of these lamps which have 
 large light giving surfaces there is an absence of sharp shadows 
 as well as an absence of ghsten and glare from the material that 
 
 Fig. 361. — Mercury vapor lamps in thread factory. 
 
 is being worked upon. In brief, the Hghting is more nearly 
 in character that obtained from the windows by dayHght, al- 
 though of course the shadows and ghsten referred to are not as 
 effectually avoided as they are by dayHght. The ghastly color 
 of persons under the mercury vapor light is usually objected to 
 at first by those working under it, but this feature is soon for- 
 gotten. When colors must be judged the mercury vapor lamp 
 cannot of course be used. 
 
 Another way to accompHsh well diffused lighting would be to 
 
SHOP AND FACTORY LIGHTING. 
 
 335 
 
 place large incandescent or small Nernst lamps at frequent in- 
 tervals in proper reflectors about 10 to 15 feet above the floor, 
 as in the case of some of the stores cited in Chapter XXVI. 
 This plan is seldom resorted to because of the high first cost of 
 so many outlets or because of ceiling obstructions. 
 
 Where most of the lighting of a shop is to be done by means 
 of arc or mercury vapor lamps placed high, care must be taken 
 to place the lamps so that the workmen will not stand in their 
 own light and so that the greatest light will be obtained where 
 
 Fig. 362. — Mercury vapor lamps in belting factory. 
 
 detailed work is being done. For example, in a great many 
 factories benches and small machines tools are placed around 
 near the walls in order to obtain light from the windows during 
 the day. The proper place for the lamps is, therefore, around 
 the edges of the room rather than in the center as they are fre- 
 quently placed. While this may appear like wasting too much 
 light on the side walls and robbing the center of the shop, it 
 usually gives the workmen better light than placing the lamps 
 in the center so that they are a long distance from the benches 
 and the workmen stand in their own light. If the walls are 
 
336 PRACTICAL ILLUMINATION. 
 
 whitewashed as they usually are in the progressive factor}', a 
 large per cent, of the light thrown toward the wall will be reflected 
 and not lost. 
 
 Fig. 361 shows a cotton thread factory w'th mercury vapor 
 lamps. The old method of lighting by bare gas jets is also seen. 
 The vapor lamps are located over the machines and adjacent 
 lamps are reHed upon to light under and around the machines. 
 
 Fig. 362 shows a leather belting factory with vapor tube lamps 
 placed over the machines. In both of these factories reflection 
 from the ceiling should play an important part. Where traveling 
 cranes are employed the lamps should be placed above the crane 
 and provided with reflectors. 
 
 For shop lighting about the same amount of light is required 
 as in drafting room Hghting. The power required should be 
 from one to two watts per square foot with mercury vapor 
 lamps and 2.5 to 5 watts with common carbon arcs. These 
 figures apply only to those portions where machine or bench 
 work is being done, which requires the best light. The average 
 for a whole shop or factory will rarely need to be over 
 half the foregoing figures. 
 
CHAPTER XXIX. 
 MISCELLANEOUS EXAMPLES OF LIGHTING. 
 
 Billiard Room Lighting. 
 
 A brief discussion of billiard room lighting was taken up in 
 Chapter XVI on Residence Lighting and the requirements for 
 that class of lighting were there stated together with the proper 
 height and distance apart of lamps for different sizes of tables. 
 
 The reflector used on the table in the residence referred to in 
 that chapter was a deep prismatic reflector giving a distribution 
 of light very similar to that shown in Fig. 37. With such a re- 
 flector (which gives a very wide distribution and at the same 
 time covers the lamp, so that a player will nqt receive the 
 glare from the lamps when stooping over the table to play) it is 
 good practice to place the lamps lower than if some other forms 
 of reflector were used. For example, in Fig. 363 the billiard 
 table lamps are arranged on a chaindelier in approximately the 
 same way as indicated in the chapter on Residence Lighting, 
 but the lamps are placed about 5.5 feet above the bed 
 of the table. In Fig. 363 a reflector such as shown in Fig. 
 54 is used. From Fig. 53 it will be seen that this reflector gives 
 a fairly concentrated distribution and that it can, therefore, be 
 with advantage placed higher than one giving a distribution such 
 as shown in Fig. 37. Attention is also called to the excellent 
 provisions made for the general lighting of the room aside from 
 the billiard tables in Fig. 363. This is a good example of the 
 proper use of Holophane globes in contrast to the improper use 
 so often made of them. On the chandelier in the foreground, the 
 chandelier arms which have lamps pointed at an angle of about 
 45 degrees have concentrating globes giving a distribution of 
 light such as shown in Fig. 78, this being the proper globe for use 
 on a chandelier with sockets at an angle. On the bottom of this 
 chandelier is a large ball giving a distribution at all angles below 
 the horizontal such as in Fig. 82. Around the side walls are 
 
 337 
 
338 
 
 PRACTICAL ILLUMINATION. 
 
 brackets placed very high so as to be out of the line of vision of 
 players and others and these brackets are equipped with the 
 same globe as used on the chandelier arms. It is a mistake to 
 equip sockets pointed at an angle of 45 degrees with Holophane 
 stalactites giving a general distribution below the horizontal, 
 as in Fig. 80, as it results in an unnecessary amount of light 
 being thrown toward the ceiling. 
 
 Fig. 364 shows the method of lighting a very handsome billiard 
 room, which method however is open to considerable improve- 
 
 FiG. 363. — Excellent billiard room lighting. 
 
 ment. Each billiard table is lighted with a row of four lamps in 
 opal dome reflectors. The distribution of light from a smaller 
 opal dome reflector can be seen in Fig. 21. The characteristic of 
 the opal dome is that it gives considerably more light immediately 
 tmder the lamp than at an angle of 30 degrees from the vertical. 
 The result is that where lamps are placed close together as in 
 this case with this reflector there will be a row of bright spots 
 under the lamps which is objectionable in billiard table lighting, 
 as a billiard table should be lighted evenly over its entire surface. 
 
MISCELLANEOUS EXAMPLES OF LIGHTING. 339^ 
 
 Fig. 364. — Defective billiard table lighting. 
 
 Fig. 365. — Very bad plan of billiard room lighting. 
 
340 PRACTICAL ILLUMIXATIOX. 
 
 If opal dome reflectors are used in such a place, the}' should be 
 equipped with frosted lamps to avoid' the streaks which tTiey 
 cast. They should, furthermore, be arranged somewhat as 
 shown in Fig. 363, rather than in a row close together over the 
 middle of the table. In order to shade the players from the glare 
 of the lamp filaments when playing, the lamps should be placed 
 higher. They are only a little over three feet above the bed of 
 the table in this case. 
 
 Fig. 365 is an example of how not to light- a -billiard room. 
 Over each table is a row of three Gem high efficiency incandescent 
 lamps in concentrating prismatic reflectors of the type shown in 
 Fig. 47. The glaring defect of this arrangement is that the 
 lamp is not shaded from the eyes of players and with so many 
 lamps like this hung low in a room it seriously interferes with the 
 comfort of everyone in the room. Large unshaded incandescent 
 lamps like this, even though they may be provided with frosted 
 bulbs should never be hung low, especially in a large room 
 like this. 
 
 Barber Shops. 
 
 In a barber shop the important point is to have an abundance 
 of light coming from a number of directions so as to avoid annoy- 
 ing shadows. In this case, large units of light, such as arc 
 lamps and incandescent clusters are to be avoided in small 
 barber shops because of their tendency to cast shadows. There 
 should rather be a large number of small lighting units well 
 distributed. As it is impossible to locate the lamps in a barber 
 shop so that they will be out of the line of vision of all customers 
 and barbers, they should be always provided with diffusing globes. 
 Barber shops should always be finished in light tints to aid in the 
 general diffusion of light and freedom from shadows. The 
 large mirrors also assist in this. 
 
 Fig. 366 shows the lighting of a barber shop where these princi- 
 ples are well carried out. Here much of the light is obtained 
 from lamps in ground glass balls between the mirrors. This is 
 supplemented by light from lamps in ground glass balls on the 
 ceiling fixtures. As these ceiling fixtures are so high, this would 
 be a good place to employ prismatic reflecting balls of the type 
 shown in Fig. 89 or a globe such as shown in Fig. 79, either of 
 which would assist in directing the light downward. Still more light 
 would be thrown down by using lamps in reflectors of the types 
 
MISCELLAXEOUS EXAMPLES OF LIGHTIXG. 
 
 341 
 
 Fig. 366. — Good barber shop lighting by electricity. 
 
 Fig. 367. — Good barber shop Hghting by gas. 
 
342 
 
 PRACTICAL ILLUMINATION. 
 
 shown in Figs. 30 or 44 with frosted bulb lamps but this would 
 be undesirable except from motives of economy because the 
 light would not be as well diffused and reduced in intensity as it is at 
 present with the ground glass balls and lamps of low candle- 
 power. 
 
 Fig. 367 shows a barber shop lighted with gas mantle burners 
 in ground glass chimneys and fluted opal reflectors. This 
 arrangement provides about as good diffusion and distribution as 
 can be obtained with gas. Its chief objection is its looks. 
 
 Fig. 368.- 
 
 -Bank lighting from hemisph^^. 
 
 Bank Lighting. 5* 
 
 The lighting, of banks is very similar in its requirements to the 
 lighting of large offices. There is usually the difference that more 
 attention must be paid to appearances in banks than in offices. 
 In recent years the general lighting of banks has been very 
 largely done from lamps placed at the ceiling in globes or hemi- 
 sheres, this lighting being frequently supplemented by lamps in 
 globes onitop of the partition behind which the bank emplovees 
 work. . The lighting of the desks in a bank is usually done by 
 means of desk lamps placed just over the desks to be lighted. 
 The desk lighting feature of a bank by means of such lamps in 
 
MISCELLANEOUS EXAMPLES OF LIGHTING. 343 
 
 itself introduces serious difficulties because a desk lamp properly 
 placed to light one clerk's desk or table is not right for his 
 neighbor who may be facing him. Taken all together, it is 
 much better to provide sufficient general illumination, 'jo that 
 special desk lamps will not have to be used, thus doing away 
 
 Fig. 369. — General lighting in Milwaukee Public Service Building. 
 
 with all of these difficulties at one stroke. Over the teller's- 
 windows, nevertheless it is usually advisable to place a lamp in a 
 concentrating reflector. The reflector shown in Fig. 57 is an 
 excellent one for the purpose, because it is opaque and provided 
 with a ground glass bottom. 
 
344 PRACTICAL ILLUMIXATIOX . 
 
 The deferts in most bank lighting arrangements as far as book- 
 keepers and clerks are concerned are well illustrated by the num- 
 ber of makeshifts one commonly sees employed about almost 
 any bank to secure proper shading of desk lamps from the eyes. 
 
 Fig. 368 shows a bank in which the general lighting is accom- 
 plished by Holophane hemisi.iheres on the ceiling. Such hemi- 
 spheres should always, in a case of this kind, be equipped with a 
 single lamp pointed straight down provided with a reflector as 
 explained in Chapter VIII. A very inefficient modification of 
 the plan used in Fig-. 368 is frequently seen in banks where opal 
 or opaline globes are placed on the ceiling without any means of 
 directing the light downward. ■' - * " 
 
 Fig. 369 shows the method employed in the accounting depart- 
 ment of The Milwaukee Electric Railway & Light Company and 
 in fact, all over the Public Service building of that c<)mpany. 
 The plan here has been to provide such an abundance of general 
 illumination that no special desk lighting -is necessary. One- 
 lamp pendants are placed at frequent intervals, these pendants 
 being equipped with Gem graphitized carbon filament lamps of 
 the types shown in Figs. 41 to 48. This is undoubtedly a very 
 efficient way of providing general illumination free from annoying 
 shadows and glare from papers. It also provides for lighting 
 one comer of a room thoroughly when it is not necessary to light 
 other portions, which cannot be said of schemes where very large 
 lighting units are employed, .;:-,, 
 
 Hospitals. 
 It is easy to lay down what should be the theoretical ideal 
 specifications for the lighting of a hospital room, be it large or 
 small. There should be plenty of light for the nurses to work by 
 and the light should not shine in the eyes of patients. How 
 difficult it is to live up to such specifications can be seen from a 
 brief study of' Fig. 370 which represents the average conditions 
 in a large hospital room. The lamps placed down the center on 
 chandeliers must necessarily shine in the eyes of patients in the 
 beds and the same is true of the lamps on the brackets. It 
 would seem that here, if anywhere, was a desirable place for 
 some scheme of indirect lighting for the general illumination ; 
 this to be supplemented hy lamps in opaque or nearty opaque 
 reflectors for special purposes near each bed. In the case shown 
 
MISCELLAXEOUS EXAMPLES OF LIGHT IX G. 
 
 345 
 
 in Fig. 370 the gas mantle burners have been rendered as nearly 
 inofEensive as possible by the use of Holophane globes which 
 reduce the intensity of the light, so that it is not as injurious to 
 the eye as if coming through a clear chimney. It is, nevertheless, 
 trying for patients to face it for any length of time. The prob- 
 lem of hospital lighting has never been worked out with thorough 
 satisfaction in practice and there is room for considerable im- 
 provement along the lines indicated, namely, the provision of 
 very mild general lighting, the sources of light preferably being 
 hidden, and the provision of special concentrated lighting at 
 
 Fig. 370. — Hospital lighting. 
 
 each bed for temporary use of doctors and nurses. Such con- 
 centrated lighting would probably have to be done with some 
 kind of adjustable wall bracket. Adjustable wall brackets in 
 such locations would be, to a certain extent, objectionable 
 because of the chances that they would be misused and that 
 light would be left directed in the eyes of patients. This possi- 
 bility of misuse must also be borne in mind in the design of such 
 lighting and in fact of all lighting where adjustable fixtures are 
 used. 
 
APPENDIX. 
 
 The following are the manufacturers' names and numbers, or trade 
 designations of the reflectors, shades and globes upon which tests are 
 given in the body of the book. In some cases where the articles are 
 staple trade products manufactured by a number of companies, the 
 dimensions are given. Where, however, the article tested is made by 
 one company only, its catalogue number has been considered a sufficient!)' 
 definite description. In this list, a large number of the tests given in 
 the body of the book are not included for the reason that sufficient de- 
 scription and illustration of the article tested is found in connection with 
 the test. In some cases, the name of the manufacturer is not known: 
 Fig. 16. — General Electric Co., 16 candle-power oval anchored filament 
 
 lamp. 
 -Fig. 17. — Same as Fig. 16, but frosted b^^ acid etching process. 
 Fig. 18. — Tantalum lamp made by Siemens & Halske in Germany. 
 Fig. 19. — Fluted opal cone reflector. Staple trade article. 6| inches 
 
 diameter by 3J inches high. 
 Fig. 21. — Opal dome reflector. Staple trade article. 7} inches diameter by 
 
 3f inches high. 
 Fig. 23. — Opal bell reflector. Staple trade article. 5 inches diameter by 
 
 2j inches high. Held by socket bushing. No shade holder. 
 Fig. 25. — Green plated opal desk shade. Staple trade article. 7 inches 
 
 diameter by 4| inches high. 
 Fig. 27. — Flat opal reflector. Staple trade article. 10 inches diameter. 
 Fig. 29. — 120 watt meridian lamp. General Electric Co., with Holophane 
 
 Glass Co. reflector, 257.5. 
 Fig. 31. — Prismatic reflector. Holophane Glass Co., 2621. 
 Fig. 33. — Prismatic reflector. Holophane Glass Co., 2301. 
 Fig. 35. — Prismatic reflector. Holophane Glass Co., 2651. 
 Fig. 37. — Prismatic reflector. Holophane Glass Co., 7381. 
 Fig. 39. — Prismatic reflector. Holophane Glass Co., 2510. 
 Pig. 41. — General Electric Co. " G. E. M." High efficiency, 125 watt 
 
 clear bulb lamp. 
 Fig. 42. — Same as Fig. 41, but with frosted tip. 
 
 Fig. 43. — Same as Fig 
 Fig. 45. — Same as Fig 
 Fig. 46. — Same as Fig 
 Fig. 48. — Same as Fig 
 
 41, with Holophane Glass Co. reflector, 6052. 
 
 42, with Holophane Glass Co. reflector, 6052. 
 
 41, with Holophane Glass Co. reflector, 6050. 
 
 42, with Holophane Glass Co. reflector, 6050. 
 Fig. 49. — Metal cone reflector painted white inside. 8 inches diameter by 
 
 3i inches high. 
 Fig. 51. — Aluminum cone reflector, frosted finish inside. 6J inches 
 diameter by 3J inches high. 
 
 346 
 
APPENDIX. 347 
 
 Fig. 63. — National X-Ray Reflector Co., 696. 9| inches diameter by 6 
 
 inches high. 
 Fig. 55. — Same as Fig. 53, with 32 candle power lamp. 
 Fig. 56. — McCreary desk shade. 7 inches diameter by 5\ inches high. 
 Fig. 58. — Parabolic aluminum reflector. 6J inches diameter by 3 inches 
 
 high. Frosted inside. 
 Fig. 60. — Cylindrical desk reflector. 4J inches diaineter by 6^ inches long. 
 Fig. 62. — ^Window trough reflector made up with glass plate mirrors. 
 
 Width 6J inches; height 3i inches. 
 Fig. 64. — Flat metal reflector, painted white inside. 8 inches diameter. 
 Fig. 66. — National X-Ray Reflector Co., 695. 10 inches diameter. 
 Fig. 70. — Tin cone with glass plate mirrors inside. Standard trade article. 
 
 10 incTies diameter. 
 Fig. 72. — Same as Fig. 70, with 32 candle-power lamp. 
 Fig. 73. — Sand blasted dome, similar to Fig. 21, but shallower. Diameter 
 
 7 inches, height Z\ inches. 
 Fig. 74. — Sand blasted ball. Staple trade article. 
 Fig. 78.— Holophane Glass Co., 3363. 
 Fig. 80. — Holophane Glass Co., 3358. 
 Fig. 82. — Holophane Glass Co., 3154. 
 Fig. 84. — Holophane Glass Co., 3105. 
 Fig. 88. — Holophane Glass Co., 9137. 
 Fig. 94. — Holophane Glass Co., 2450. 
 Fig. 96. — Holophane Glass Co., 2805. 
 Fig. 98. — Holophane Glass Co., 1385. 
 Fig. 100. — Holophane Glass Co., 245. 
 Fig. 107.— Welsbach Co., shade, 1687. 
 Fig. 109. — Holophane Glass Co., 4401. 
 Fig. 111. — Welsbach Co., No. 60 burner. No. 197 mantle, 8 inch clear 
 
 chimney. 
 Fig. 112. — Same as Fig. Ill, but with Welsbach Co. fluted opal reflector. 
 
 No. 74. 
 Fig. 114. — Welsbach Co., Q chimney. No. 66 burner. No. 4197 mantle. No. 
 
 302 Q globe. 
 Fig. 116. — Welsbach Co., No. 60 burner. No. 197 mantle, ^Xo. 82 chimney. 
 Fig. 118. — Welsbach Co., No. 60 burner. No. 197 mantle. No, 0472 
 
 chimney. 
 Fig. 120. — Welsbach Co., No. 61 burner. No, 197 mantle. Fig. 310 chimney, 
 
 and Holophane Glass Co,, shade, 3321. 
 Fig. 122, — Welsbach Co,, No. 66 burner. No. 197 mantle, and Holophane 
 
 Glass Co., shade, 2811. 
 Fig. 124. — Same as Fig. 122, except with Holophane Glass Co., shade, 
 
 2711. 
 Fig. 126. — Welsbach Co., No. 60 burner. No. 197_mantle, 8 inch clear 
 
 chimney. No. 10 dome and No. 43 Bobesche. 
 Fig. 128. — Same as Fig. 126, but with Welsbach Co. dome 1036 and No. 43 
 
 Bobesche. 
 Fig. 129. — Welsbach Co., No. 60 burner, No. 197 mantle, 8 inch clear 
 
348 PRACTICAL ILLUMINATION. 
 
 chimney, 10-inch green plated opal dome, and Holophane Glass Co., 
 Bobesche, 1386. 
 
 Fig. 130. — Reflex burner and reflex mantle with clear air hole chimney 
 No. 317, Welsbach Co. 
 
 Fig. 131. — Same as Fig. 130, with 5 inch French rough ball. 
 
 Fig. 133. — Same as Fig. 130, with flat opal reflector, 503. 
 
 Fig. 135. — Same as Fig. 130, but with opal cone reflector, 502. 
 
 Fig. 137. — Same as Fig. 130, but with Holophane Glass Co. reflector, 2333. 
 
 Fig. 141. — Acetylene gas flame in Welsbach Co. shade, 1687. 
 
 Fig. 142. — Opal shade. Staple trade article. 7\ inches diameter by 3J 
 inches high. 
 
 Fig. 144.— Holophane Glass Co., 2807. 
 
 Fig. 145. — Benjamin Electric Mfg. Co. " arc burst," D-47. 
 
 Fig. 147. — Holophane Glass Co., Holophane " Reflecting Arc." 
 
 Fig. 149. — Three-light Benjamin cluster with Holophane Glass Co. re- 
 flectors, 2631. 
 
 Fig. 155. — Holophane Glass Co., 2522. 
 
 Fig. 159. — Holophane Glass Co. hemisphere, 701. 
 
 Fig. 169.— Holophane Glass Co., 1201. 
 
 Fig. 172. — Three-light cluster with Holophane Glass Co. reflectors, 2535. 
 
 Fig. 177. — Holophane Glass Co., 3165. 
 
 Fig. 191. — Excello Arc Lamp Co., 8 ampere arc. 
 
 Fig. 192. — Welsbach four-light gas arc. Clear globe. 
 
 Fig. 193. — Welsbach Co., fovxr-light gas arc, clear globe and flat porcelain 
 reflector. 
 
 Fig. 195. — Welsbach Co., four-light gas arc, alabaster globe. No reflector. 
 
 Fig. 197. — Humphrey four-light gas arc, clear globe, no reflector. 
 
 Fig. 199. — Humphrey four-light gas arc, alabaster globe, no reflector. 
 
 Fig. 201. — Cooper-Hewitt Electric Co., mercury vapor lamp. 
 
PAGE. 
 
 Acetylene Gas 96 to 100 
 
 Aluminum Reflectors 48, 51, 215, 219 
 
 Arc Lamps, Electric 120 to 127 
 
 Alternating arc tests 125, 126 
 
 Difficulties of testing 120 
 
 Direct current tests 120 to 123 
 
 EfEect of small carbons 123 
 
 Efficiency of 143 
 
 Flaming are tests 127 to 143 
 
 Arc Lamps, Gas 128 to 133 
 
 Consumption of -, .• . . . . 132 
 
 Efficiency of ,•,... 132 to 144 
 
 Horizontal distribution of 128, 129 
 
 Mean spherical candle-power of 132 
 
 Vertical distribution of 130 to 132 
 
 Asbestos Reflectors for Bowls 108 
 
 Gaskets for closed globes 50, 62, 74 
 
 Bank Lighting 342 to 344 
 
 Banquet Room Lighting. See Dining-Room Lighting, Hotels. 
 
 Barber Shop Lighting 340 to 342 
 
 Basement Lighting 204 
 
 Bath- Room Lighting 203, 213 
 
 Bed-Room Lighting 182 to 196, 213 
 
 Billiard-Room Lighting 212, 337 to 340 
 
 Blinding Effects of Unshaded Lamps : 1,2 
 
 Bowls, Beaded 110 
 
 , Ground Glass . 100. 110, 137 
 
 Holophane 107, 112, 137 
 
 Opaline 112 
 
 Sand blasted 106, 110, 137 
 
 Reflectors for 108 
 
 Calculation of Illumination 13 to 24, 226, 229, 242, 299, 323 
 
 Candle-power, Mean Spherical, of 
 
 Acetylene gas flame . .". . . ■. -; . ' ■- '99 
 
 Gas, open flame ,.,. . '. ■:■ •:■■ 79 
 
 ; :' < ; 'Gas, -mantle burner .,•.:'. . .,, ; 82 
 
 349 
 
350 IXDEX. 
 
 PAGE. 
 
 Candle-power, Mean Spherical, of 
 
 Gas, inverted burner 91 
 
 Gas arcs 132 
 
 Gem electric incandescent 42 
 
 Incandescent electric, common 26, 27 
 
 Nemst lamp 115 
 
 Tantalum incandescent electric 29 
 
 Candle-power, mean spherical, definition of 23 
 
 Candle-power, mean spherical, comparisons of 143, 144 
 
 Car lighting 288 to 292 
 
 Ceiling vs. Cliandelier Lighting 138 
 
 Chandelier vs. Ceiling Lighting . 138 
 
 Church Lighting 268 to 279 
 
 Club Lighting, Dining-Room 238 
 
 See also Dining Rooms, Lighting, Hotels. 
 Clusters, Electric. 
 
 " Arc Burst " 101 
 
 Flat opal 102 
 
 Metal 104 
 
 Prismatic 101, 102, 105 
 
 Color of Illuminants 1 45, 297, 298, 299 
 
 Comparison of Illuminants 142 to 146 
 
 Concentric Diffusers 225 
 
 Contrasts, Effect on the Eye 11 
 
 Corridor Lighting, Clusters for 102 
 
 (See also Hall Lighting.) 
 
 Critical Angle for Shading Eyes 227 
 
 Demonstration-Room Tests 136 to 141 
 
 Depot Lighting 241 to 244 
 
 Design of House Lighting 207 
 
 Desk Lighting 215 to 226 
 
 Reflectors for 37, 41, 49 to 53, 138 
 
 Desk Portables 217 to 221 
 
 Dining Car Lighting 290 
 
 Dining-Room Lighting. 
 
 Disks, to keep out dirt 50, 62, 74 
 
 Hotels 178, 232 to 240 
 
 Residences 172 to 182 
 
 Drafting- Room Lighting 224 to 226 
 
 Dresser Lighting 182, 187 to 193 
 
 Efficiency of Light Sources. 
 
 Electric 143 
 
 Gas ;..".. 144 
 
 Electrical Testing Laboratories ! '. VI 
 
 Eye and Light 1, 2, 9 to 12 
 
 Factory Lighting 134, 332 to 336 
 
IXDEX. 351 
 
 PAGE. 
 
 Fixtures. De«;gE ". i4S 149, 276 to 279 
 
 Charcb liirhTing design 276 to 279 
 
 Schedule 307 
 
 Specificaticms 207 
 
 Flaming Arcs 127. 143 
 
 Flat Flame Ga* Test? . . 79 to S2 
 
 Flickering Lights 9 
 
 Foot Candles. 
 
 Amounts required 22 to 24 
 
 Deaniiion of 14 
 
 FrciJied Lamps. Advantages of 2S 
 
 Life of 73 to 75 
 
 Plwtonietric curve of 27 
 
 Use with leflectois 34. 41 
 
 Fiosied Renecttrs . .5S. 139 
 
 Gas Bumerf. Ouster li2S to 133 
 
 IndiYidnal 76 to 95 
 
 Gs5. Acetjiene 96 to 100 
 
 Ette- — c f pressure variations 76 
 
 tj— t jency of different burners 144 
 
 C^o! 3ame 79 to 82 
 
 G-- Girce; and Shades. 
 
 J^:v iisss S4 
 
 «i?tim-i giass S4 
 
 Htutpfcsjie 86 to 90 
 
 _ Oil Q clmnne y S3 
 
 Gaf 2,ai?:r :i S3. SS. 93, 130 
 
 Gasias 50, 62, 74 
 
 Ge=i I<i=nns 41 to 44 
 
 E=ciency of 143 
 
 G-iches. Entidsing Electric 59 to 67 
 
 G£= 76 to 95 
 
 Open electric 67 to 72 
 
 G7ill-S.'>jct lighting 233 
 
 HaD Lighting, Qusters for 102 
 
 H(*els 256 
 
 Office buildings 255 to 261 
 
 PuWc halk 227 to 231 
 
 Residence 203, 212. 213 
 
 Hemispbeies, see Bowls. 
 
 Hcdophane Globes, Absorption of 61 
 
 Acetylene globes 100 
 
 Bowls 107, 112. 137 
 
 Cbsses of 60 
 
 Cleaning of 61, 62 
 
 Dnst tests "5- ^^ 
 
352 INDEX. 
 
 PAGE. 
 
 Holophane Globes, Absorption of 
 
 Electric globes, closed 62 to 66, 138, 140 
 
 Electric globes, open 69 to 72, 138 
 
 Hemispheres 107, 112, 137 
 
 Mantle burner globes 86 to 90 
 
 Nemst globes 116 
 
 Open gas globes 81 
 
 Principles of , 59, 60 
 
 Holophane Reflectors, see Prismatic Reflectors. 
 
 Horizontal Illumination 16 
 
 Hospital Lighting 344, 345 
 
 Hotel Lighting. 
 
 Bed- Rooms 189 to 196 
 
 Dining Rooms 178, 232 to 240 
 
 Halls 256 
 
 Lobbies, see Lobby Lighting. 
 
 Parlors 170 to 172 
 
 House Lighting 151 to 214 
 
 Design of 207 to 214 
 
 Illuminating Engineering Society V 
 
 Illumination Curves of 21 
 
 Horizontal 16 
 
 Normal 15 
 
 Vertical 22 
 
 Incandescent Electric Lamps 25 to 75 
 
 Light distribution of carbon filament lamp 26 
 
 Light distribution of carbon filament lamp, frosted 27 
 
 Light distribution of Gem lamp 41 to 44 
 
 Reflectors, shades and globes 25 to 75 
 
 Tantalum 28 
 
 Efficiency of 143 
 
 Incandescent Gas Burners 82 to 95, 128 to 133 
 
 Indirect Lighting, Hotels 250, 251 
 
 Railroad depot 241 to 244 
 
 Injurious Effects on the Eye of 
 
 Contrasts ; . : .' 11 
 
 Flickering light .■■■>: ' 9 
 
 Glare ■!?. 10 
 
 Light frorn an Unusual Angle ■ ' . . . 10 
 
 Light in the field of vision 10 
 
 Streaks ; . . . 1-1 
 
 Too little Hght '. . . .■..:•.'. \". . v 11 
 
 Too much light ..:.....-. 11 
 
 Unsteady light 11 
 
 Inverse Squares, Law of 3, 14 
 
 Inverted Gas Burners ; •;; . 90 to 95 
 
 Efficiency of ■. 143 
 
INDEX. 353 
 
 PAGE. 
 
 iUtchen Lighting 196 to 199, 213 
 
 Large vs. Small Lighting Units 295, 296 
 
 Laws of Light 3 to 8 
 
 Library Lighting, Public Libraries 280 to 287 
 
 Residence Libraries ,, . . 161 to 163 
 
 Light and the Eye 1, 2, 9 to 12 
 
 Flickering 9 
 
 From an unusual angle 10 
 
 In the field of vision 10 
 
 Laws of 3 to 8 
 
 Laws of inverse squares 3, 14 
 
 Nature of 3 
 
 Too little 11 
 
 Too much 11 
 
 Unsteady 9 
 
 Living-Room Lighting 151 to 172, 212 
 
 Lobby Lighting 244 to 254 
 
 Location of Lamps 2 
 
 Lodge-Room Lighting 227 to 231 
 
 Mantle Burners, see Gas Burners. 
 
 McCreary Reflector 50 
 
 Mean Spherical Candle-Power, see Candle-Power. 
 
 Measurement of Light 3 to 8 
 
 Mercury Vapor Lamps 134 to 135 
 
 Efficiency of 145 
 
 Examples of 225, 226, 334 to 336 
 
 Meridian Lamps 36 
 
 Metal Reflectors, Aluminum Cone 48 
 
 Parabolic 51, 215, 219 
 
 Cluster type 104 
 
 Cylindrical desk type 53, 215, 219 
 
 Flat enameled type 55 
 
 Tin cone enameled 46 
 
 Mirror Reflectors 48 to 50, 54 to 58 
 
 Moore Tube Light 135 
 
 Moiuiting of Lamps in 
 
 Art glass domes 169, 173 
 
 Balls 114 
 
 Bowls • ■ • 108 
 
 Table lamps 167 
 
 Nemst Lamps 115 to 119 
 
 Advantages of 119 
 
 Correct placing of 118 
 
 Efficiency of 143 
 
 Efficiency of different sizes 117 
 
 Examples of use of 239 to 244, 308, 321, 335 
 
 Normal Illumination 1^ 
 
354 INDEX. 
 
 PAGE. 
 
 Office Building Lighting 
 
 Corridors 255 to 261 
 
 Lobby 249, 252 
 
 Offices 215 to 226 
 
 Opal Reflectors for 
 
 Clusters 102 
 
 Electric arc lamps 126 
 
 Gas, flat fluted 83 
 
 Gas, dome shape 88 to 90 
 
 Incandescent electric lamps. 
 
 Bell shape 32, 138 
 
 Cone shape 29, 34, 137 
 
 Dome shape 31, 139, 140 
 
 Flat 34 
 
 Mounted gas, flat 92 
 
 Cone 93 
 
 Open Flame Gas Tests 79 to 82 
 
 Pariors, Lighting of 151 to 172, 212 
 
 Photometric Curves 25 to 135 
 
 Meaning of 13 
 
 Photometers 3 to 5 
 
 Placing of lamps 2 
 
 See also Mounting of Lamps. 
 
 Porch Lighting 199 to 203, 211 
 
 Portable Lamps 139, 152, 153, 158, 163, 166 
 
 Prismatic Reflectors. 
 
 Advantages of 35 
 
 Ball type 66 
 
 Bowl type 40, 138, 140 
 
 Cleaning of 289 
 
 Cluster type 101, 102, 105 
 
 Concentrating type 37 to 39, 137 to 139 
 
 Desk type 41, 215, 216, 218 
 
 Gem type 41 to 44 
 
 " Inverted bowl " type 40, 138 140 
 
 Inverted gas 93 
 
 Meridian type 36 
 
 Wide distribution type 40, 138, 140 
 
 Public Hall Lighting 227 to 231 
 
 Quantity of Light 22 to 24, 226, 229, 251, 299, 323 
 
 Railroad, Car Lighting 288 to 292 
 
 Depot lighting 241 to 244 
 
 Reading Lamps. See Portable Lamps. 
 Reading-Room Lighting. See Library Lighting. 
 
 Reading Table Lighting 161 to 163, 280 to 287 
 
 Reflection, Coefficients of 7, 21 
 
 Diff^use 6 
 
INDEX. 355 
 
 PAGE. 
 
 Reflection, Coefficients of 
 
 Irregular g 
 
 Regular 5 
 
 Total g 
 
 Reflectors, Aluminum 48, 51, 215, 219 
 
 Frosted '. ,53, 139 
 
 Holophane. See Prismatic Reflectors. 
 Metal. See Metal Reflectors. 
 
 McCreary 5Q 
 
 Mirror 48 to 50, 54 to 58 
 
 Opal. See Opal Reflectors. 
 Prismatic. See Prismatic Reflectors. 
 
 Trough 54, 55, 326, 331 
 
 Use in art glass domes 169, 173 
 
 Balls 114 
 
 Bowls 108 
 
 Table lamps 167 
 
 X-ray 48, 56, 137 
 
 Refraction 7^ g 
 
 Residence Lighting 151 to 214 
 
 Restaurant Lighting 236, 237, 240, 307 
 
 See also Dining-Room Lighting, Hotels. 
 
 Sand Blasted Balls 59 
 
 Reflectors 58, 139 
 
 Schedule of Fixtures 206 
 
 School Lighting 283, 284 
 
 Shop Lighting 134, 332 to 336 
 
 Sleeping-car Lighting 291 
 
 Specifications for Lighting 207 
 
 Storage-room Lighting 26, 134 
 
 Store Lighting, by 293 to 323 
 
 Electric arcs 296, 298, 301, 302, 307 
 
 Gas arcs 297, 316 to 318 
 
 Gem lamps 296, 299, 322 
 
 Incandescent electric 293 to 323 
 
 Mantle burners 295, 319 to 321 
 
 Nemst lamps 296, 321 
 
 Open gas 294, 319 
 
 Store Lighting. 
 
 Color effects 297 to 299 
 
 Store Lighting, Examples of 
 
 Automatic vaudeville 299 
 
 Bakery 316 
 
 Candy store 313, 319 
 
 Carpet store 300 
 
 Cigar store 316 
 
 Dry goods store ^^ 321, 322 
 
356 IXDEX. 
 
 PAGE, 
 
 Store Lighting, Examples of 
 
 Drug store 305, 321 
 
 Fur store 309 
 
 Furniture store 315 
 
 Grocery store 304 
 
 Hardwa.re store 310 
 
 Hat store 313, 314 
 
 Jewelry store 308, 311 
 
 Meat store 302 
 
 Millinery store 318, 319 
 
 Picture store 316 
 
 Picture frame store 310 
 
 Railroad ticket office 302 
 
 Restaurant. See Restaurant Lighting. 
 
 Shoe store 301 
 
 Stationery store 306 
 
 Store Lighting, large vs. small units 295, 296 
 
 Streaks 11 , 28, 34, 46, 48 
 
 Switch Control 149 
 
 Table Lamps. See Portable Lamps. 
 
 Tantalum Lamps 28, 1 43 
 
 Theatre Lighting, Auditorium 262 to 267 
 
 Lobby 246 to 248 
 
 Trough Reflectors 54, 55, 326, 331 
 
 Unsteady Light 9 
 
 Vapor Lamps. See Mercury Vapor Lamps. 
 
 Vertical Illumination 22 
 
 Warehouse Lighting 26,134 
 
 Watts per square foot 22 to 24, 226, 229, 251, 322, 336 
 
 Window Lighting .- 39, 49, 55, 67, 94, 324 to 331 
 
 X-ray Reflectors 48, 56, 137