^: fe, t; t; ^' §u *M' i^ "'^ "■■ ^' ^

' ;k' *i .1? P> 3

■#' «, 'i' f i '*

ijf tr«i

mmMmm

■■■#'• t'"# 1'^ t B ii' ii/ '

AT 148

Lindsay, Curtis Morgan A comparative life test of five types of incandescent

A COMPARATIVE LIFE TEST OF FIVE TYPES OF INCANDESCENT LMIPS

A THESIS

PRESENTED BY

CURTIS MORGAN LINDSAY

TO THE

PRESIDENT AND FACULTY

OF

ARMOUR INSTITUTE OF TECHNOLOGY

FOR THE DEGREE OF

BACHELOR OF SCIENCE IN ELECTRICAL ENGINEERING

HAVING COMPLETED THE PRESCRIBED COURSE OF STUDY IN

ELECTRICAL ENGINEERING

ILLINOIS INSTITUTE OF TECHNOLOGY

PAUL V. GALVIN LIBRARY May 18,1909.

35 WEST 33RD STREET X^ / y

CHICAGO. IL 60616 - ^^^^fu^^-.^

PREFACE,

-1-

It ia the purpose of this theaia to make a comparative life test of five of the different typea of incandescent lamps on both alternating and direct current. There have, up to the present time, been numerous tests on the life of these several typea but as far as the author has been able to ascertain, there have been no teats made in which the lamps were tested simultaneously on both kinds of current. Life teats have been made on direct current alone and on alternating current alone but an effort to teat incandescent a under the same conditions of operation, on both kinds of cur- rent and at the same time has not been attemp%- ed.

The author, therefore, desired to make the purpose of this thesia a study of the life of these several typea of lamps when operated under the conditions as before mentioned: i.e., to test the life of these different types so that every condition of operation would be ex- actly the same throughout their life on both

direct and alternating current.

O.M.L.

-II-

BIBLIOGRAPHY.

Stine,W.M.

Photometerio Maasurnienta. Palaz,A.

A Treatise of Industrial Photometery. Barrows, W.E,

Electrical Illxaminating Engineering. National Electric Lajwp Association.

Bullatins from Engineering Department. Eustice,A.L,

The Westinghouse Nemst Laanp. University of Illinois Bulletin # 19.

Comparative Life Teats of Oarbon, Metallized,

Carbon and Tantal;am Filament Lamps. Bureau of Stemdarda Bulletins.

A Comparative Study of Plain and Frosted Lamps

On the Determination of the Mean Horizontal Intensity of Incandescant Lamps.

On the Detemnination of the Mean Horizontal Intensity of Incandescant Lamps by the Rotating Lamp Method. Numerous Articles in the leading Techincal Papers of the day.

m.

Q

*-*«ei

i

d

u

2J.

ASKotra

TABLE OF CONTENTS.

Page

Preface I

Bibliography --. — -^ -.,,-11

Illustrations -ill

Table of Contents --^^ 1

PART I.

THE LAMPS.

Introduction -^ »^- 4

Selection or Lampa . - 4

Rating . 4

Pilamenta . 5

Bulbs «_, ,_ 7

The Nernst Lamp 8

The Standard Lamp H

PART II. THE SELECTION OF THE APPARATUS FOR OPERATION.

The Machines ■ 13

The Recording Device 17

Variable Conditions for Operation ■- — 19

Page.

Classification and Numbering — ■ 80

Photometerlng •- ■ .,^_-- 21

Photometer Scale ^--- ^^^_^,_. 23

Photometer Screens •- = S5

PART III. DATA AND CALCULATED RESULTS.

Photometerio Data of the Gem Lamps — --— 27

Photometeric Data of the Tantalwa Lamps ■- 36

Photometerio Data of the Tungsten Lamps — -- — 41

Photometerio Data of the Nernst Lamps 60

Photometetic Data of the Carbon Lamps b8

G.P. during Life of the Gem Lamps •--- 67

C.P. duringLife of the Tantalum Lamps 39

C. P. during Life of the Tungsten Lamps • 71

C.P. during Life of the Nernst Lamps ■ — 73

C.P. during Life of the Carbon Lamps — 74

Running Log ■- — 75

PART IV. CURVES AND DIAGRAMS.

PART V.

Conclusion — ■ 79

PART I,

**•;.'■?{-:;•

THE LAMPS .

INTRODUCTION. The importanoe of the suTbjQct " A Com - parative Life Teat of Five Types of Incandescent Lamps" , involves a diacuasion of the selection of the lamps, their rating, the kind of filaments, the bulbs, the Nernst lamp, and the standard lamp.

SELaCTION OF LAMPS.

Four of the five types which are used in this test were procured on the open marks t. These consisted of the Gem metalized, the tantalura, the tungsten and the Nernst lamps. The fifth type, the carbon lamp, was obtained from the general supply department of the Institute and were sel- ected at random from a lot of one hundred lamps.

RATING.

The carbon lamp is of the "Monarch" type

which is rated at 16 c.p. at 110 volts. The Gem

e

metalized lamp is a "Banner" 50 watt which may

be operated at from 106 to 110 volts. The tanta-

-5-

lum ia also a "Banner" and is labelsd as 25 watt 12 1/2 c.p. at 110 volts. The tungsten is a "Ban- ner" 25 watt lamp when run at 110 volts. The met- al^ized and the tungsten made no mention of any rated candle power on the labels. The Nernst lamps are of the Weatinghouae Nernst screw base type which contain but one glower and consumes .6 amperes at 110 volts,

FILAMENTS.

The carbon lamps contain the familiar double loop filament which is anchored in the middle of the loop. The filament had a very smooth appearance and was of a steel-gray color which is a characteristic of all good carbon lamps .

The metallized carbon lamps had a fila- ment differing in shape somewhat from the carbon. Instead of having a drop anchor, as in the prev- ious lamp, the filament of this lamp was anchored close to the sealing tube. In this way, the fila-

•6-

inent conaiated practically of two separate loops. The color was vary much the same as that of the carbon but the surface was not as smooth.

The filament of the tantalum lamp was mounted in the usual zigzag manner Tv-hich condi- tion requires a comparative longer length of the filament than the other lamps, which fact is due to the low specific resistance offered by the metal tantalum. There were nine supporting spires at both the top and the bottom of the glass sup- porting rod. The same color prevailed in these filaments as in the previous types with the ex- ception that a metallic luster was observed in the case of the tantalum.

The tungsten filament consists of a sin- gle length which was shaped into four loops. The filament was continuous over eight supporting spires at the top and anchored at the bottom of the loop with a metal eye which is in turn fused into a glass upright support. The reason for this

-7-

method of support is that when heated, the tungs- ten becomea very soft and by the aaaiatance of the looped anchor, the lamp may be burned in any posi- tion.

BULBS.

The Mr3t photograph shows the style, the comparative size, and the charaoteristis shape of the different types of bulbs. The carbon, metal- lized and the tantalum lamps are similar in design and are about the same size. These bulbs are five inches over all in length with a maximum diameter of two and three eights inches and a minimum dia- meter of one inch. The tungsten bulb is six and three eights inches long and is two and seven six teenths inches in diameter at the largest point and one and nine-sixtaenths inches at the small- eat. The Bides are straight and taper from the maximum to the minimum diameter. The Nernat lamp has a length of nine and three eights inches over

-8-

all.

Tha bulbs ara all of clear glass with the exception of the globe on the Nernat lamp. Thia ia made of an opaleaoant glaaa and ia spherical in ahape being about four inchea in diameter.

THE NERNST LAMP.

From A.L.Euatice* paper "Tha yTeatinghouae Nernat Lamp" .

"The holder of the Weatinghouae mult- iple -onita preaenta a radical change in deeign. The old two-piece holder porcelain ia replaced by a one-piece holder baae,to which ia attach- ed the characteriatic terminal pronga. Two pronga are brought through the holder baae and are aeourad in auch a manner that they lie in a plane parallel to the glowara and at right anglea to them. The uae of two or more heater tubea ia auperaaded by a 'Wafer Heater'; a

heater consisting of a amall platinum- wound and refractory cement-ooated rod bent so that several sections lie parallel to the glowers and securely ..lounted on a flat porcelain. The wafer slided on the heater prongs when inser- ted in the holder; the heated terminals being in the form of a sleeve contact. Hence, it will be noted that the heaters can be readily chang- ed without tools and without disturbing any other member of the lamp.

An improved method of supporting the globe is employed so that the glassware can be removed instantly and at the same tine is lock- ed to the lamp body, thus minimizing both labor in cleaning and breakage due to careless hand- ling.

The various aizes of the single glowsr lamps are of the Edison base type and present a similar appearance to the now popular 110 watt lamp, although the construction ia a unique

10-

departure from the forTner praotica. The cutout ia located within the Sdison base, from which the pronga lead to a baae porcelain on whose lower side ia a acrew reoapticle, while on the upper aide ia a ballast.

The holder oonsiata of a glower and a wafer heater permanently connected on a araall porcelain provided with a standard acrew baae, with an additional contact pin in the center. By an asaortment of diameters and lengtha of contact pin3,it ia impossible to insert any other than the proper holder in the lamp body, thereby insuring the consumer againat troubles incident to the use of lamps of various sizes and voltage.

This form of renewal ia popularly termed the 'Screw Burner' and .-should supply the demand for the high-efficiency incandes- cent lamp,30 rugged in its design that the lamp can be maintained by anyone. The burner ia

11-

furni3hed complete with glassware when small balls are desired and without glassware when the standard size of ball is used on the lamp"

THE STANDARD LAMP.

The standard larap was seasoned accord- ing to the proper method and was standardized by repeated comparison with the standards of candle power at the Electrical Testing Labora- tories at New York City. The constancy of the lamps can be rslied upon during a limited per- iod of use or as long as the ratio between the current and the voltage remains unchanged. The total number of hours burned during the periods for photometry did not exceed twenty-five hours at the greatest, and great care was taken 3 0 that the voltage on the standard did not exceed that specified for the lamp.

The standard which was used in these tests was # 2638 as tested by the Electrical Testing Laboratories. It gives sixteen candle

18-

power ?;h8n operated at 113.3 volt a and .551 amperes and which is oorrent to within one- tenth of one per cent of the absolute stand- ard.

This standard lamp has four fiducial lines etched in quadrature positions on the bulb. The position marked "1" was placed in the photometric axis and toward the screen and always kept in this position when the other lamps were photometered.

PART II.

■<:■*

THE SELECTION OP THE APPARATUS 70R OPERATION.

13-

THS IvlACHINES,

It waa first thought that it would be boat to place the lamps for the direct current teat directly on the lighting circuit of the Institute and then run one of the larger alter- nating current generators in the main labora- tory to supply the current for the lamps which were to be tested on alternating current. This situation, however, did not prove acceptable for three reasons .'first, because it waa found that, on accovint of the regular work in the labora- tory, it would be impossible to tie up the use of the generator for a period of two months or more as would be the case; second, because the direct current supply of the Institute fluctu- ated 30,due to the throwing on and off of the large machines in the laboratory and due also to the starting of the elevator motor in Mach- inery Hall; third, because this method of opera- tion would not permit the adoption of the main

14-

purport of the teat which was to run both cir- cuits ao that there would be the same relative changes in both and at the same time.

Another idea was to run the two circuits from power purchased from the Edison system in the city. It would be possible to obtain this power at practically a constant potential due to the fact that the Edison system is so com- prehensive that there are very few fluctua- tions and quite small ones at that. There are, however, three disadvantages to this scheme: first, because it would be necessary to rvin the test at some distance from the school, since the Edison wires do not lead into the Institute and this would necessitate carrying the lamps to and from the school to photometer which condition would provide a greater chance to loss due to breakage in handling;second, because it would be hard to find a suitable place to run the test where it would be known that

15-

9 very thing was antirely safe; third, and the most important , because the department lid not consid- er the expense of purchasing the extra power aa a necessary factor in the carrying out of the test.

Then as a third alternative, it was sug- gested that the power necessary for the teat be generated on the spot. To carry out this thought then, a 5 K.W. "Lincoln" variable speed notor was direct connected to a 5 K.W. "Wood" converter. The motor was supplied with power from the line of the Institute which, at the motor terminals, was about 109 volts. The converter was operated as a double current generator at a speed of 1200 R.P.M, This machine supplied direct cur- rent for the test at 110 volts and alternating current at 76 volts, the alternating current hav- ing a frequency of 60 cycles. This voltage was stepped up to 110 volts by means of a small auto-transformer placed in the circuit. The

16-

coiwerter could "be excitad by either one of two ways by uaing a double-pole double-throw switch, one aide of which was connected to the leads to the motor and the other to the bua-bara of the storage battery. The idea in providing this dou- ble means of excitation of the converter was to try to keep the generated voltage as constant as possible. Early in the morning, when the elevator in Machinery Hall was running and when there was a section in the dynamo laboratory throwing on and off the different machines and their respec- tive loads, the voltage always fluctuated. If the field was connected to the line, this fluctuation produced not only an increase or decrease in the speed of the motor but also a strengthening or weakening of the field of the double-current generator at the same time. Thus an increase of speed and field excitation and vica versa pro- duced a fluctuating voltage on the load. But on the other hand, if the field was connected to the battery at these particular times, the speed

■17.

would increase or decrease slightly due to the fluctuation of the impressed voltage and the field on the generator remained constant, thus producing a comparatively small change in the generated voltage.

The object of every consideration of the generator was to produce as constant volt- age as possible on the lamps. It was hoped that with this combination of motor and converter run as a double- current generator, that the var- iaUion in the voltage across the lamps would not exceed 111 volts as a maximum limit and that it would not fall below 108 volts as a min- imum. Every affort was made to prevent any fluct uation which would be greater than that already suggested and if possible keep it even within a closer range.

THE RECORDING DEVICE.

A recording voltmeter of the Westing- house type was connected across the direct cur-

18-

rent lamp circuit so that the time of running could be graphically recorded. In the case then of any accident to the machines or if the supply voltage should at any time be shut off, the meter would then give accurate imformation as to the exact time of such trouble. Not only was this the purpose of the recording voltmeter, but also for another and really primary reason. By means of this instrument, the author was able to keep a fairly close record of the voltage across the direct current lamps for every hour of the test. Since the voltage across the alternating current lamps varied directly with that across the dir- ect current, it was unnecessary to record the alternating voltage, since connected in as it was, the meter read at all times the voltage across both circuits.

At the time of installing the meter, it was observed by calibration with a calibrated Weston instrviment, that it had a constant error

-19-

of reading four volts too high. Instead of at- tempting to make the necessary changes in the setting of the instrument which were very deli- cate and which might make the difference even greater, it was decided to insert a high resis- tance rheostat in series with the voltmeter. This proved to dispel the trouble and with a certain resistance cwt in the rheostat, It was found that the recording voltmeter read cor- rectly as checked with the calibrated voltmeter,

VARIABLE CONDITIONS IN OPERATION.

Any variation in the speed that should be found necessary, which variation might be due to the temperature of the motor fields and the motor itself, could be very easily made by turn- ing the wheel on the motor. By turning this wheel, the effective length of the armature could be varied, since a turn in one direction pulled the armature out from under the poles a

-so-

certain definite amount and a turn in the oppo- site direction would give the opposite result. In this manner then, the speed of the motor could "be easily and definitely adjusted to auit the condi- tions of operation.

In order to take care of any variation in load that would occur due to the burning out of the lamps and to also take care of the slight variation in voltage due to change of speed, it was deemed advisable to place a low resistance rheostat in the field circuit of the double-oiar- rent generator. With the assistance of this aux- iliary apparatus, it was hoped that all variations and fluctuations could be easily and speedily remedied.

CLASSIFICATION AND NUMBERING OF THE LAMPS.

Having everything in readiness for the generation of the current, attention was then turned to getting the lamps ready and photometer* ing them before the test. In order to facilitate

-21-

the method of recording the photometric readings, the different types were each given series letter as A,B,C,etc. and the different lamps in each series were numbered as 1,2, 3, etc. Series "A"con- cisted of the Gem metallized lamps, series "B" of the tantalum, series "C of the tungsten and each of these series having lamps numbered from one to twelve inclusive. Series "D" consisted of the Westinghouse Nemst lamps and series '*£'* of the carbon filament lamps, each of these two series having lamps numbered from one to six inclusive.

PHOTOMETERING.

The idea of photometering was not so much to obtain the photometeric analysis of the several lamps at the different stages in the life test, but rather to obtain comparative data of the lamps at the different periods of the life. Very complete photometric tests have been made on every one of these five types of lamps to obtain the mean horizontal and the mean spherical candle powers and also the reduction factors for the

•22-

different hours of life, so that to repeat these investigationa would be unneoeaaary.

In order then, that the object of this thesis work might be beat carried out, it was con- sidered best that the lamps be photometered in onlji one position throughout the teat. This method would give the desired meana of compariaon of the larapa for the life test but, of course, the candle power only for that one position of the lamp and in that one direction. Consequently, to insure that the lamp be always photometered in the same position, a mark was cut at the bottom of each globe with a diamond point and then this scratch always put in line with the axis and toward the screen when ever photometered. Great care was taken to place this scratch exactly in this posi- tion every time. Care was also taken to place this scratch in the aame relative position with regard to the filament on each bulb and in each series.

•23-

PHOTOMETER SCALE,

To use the scale which ia on the photo- meter bench for measuring the distances between the standard and tested lamps and the screen to determine the candle power, is a very tedious method. Consequently, other methods were resorted to in order to save time in the calculations if possible. It was suggested that a scale be con- structed to read the candle power of the lamp tested directly. To carry out this idea, the lamp to be tested is placed at the zero position of the scale on the bar, and the standard lamp is placed in the carriage on the opposite side of the screen as shown in sketch (1).

t

-/OOO

Sketch 1.

The two carriages which carry the screen and the standard lamp are connected by an iron bar so

■24-

that the diatance between the center of the screen and the center of the standard lamp is exactly 1000 mm. Calling the distance between the zero scale reading and the position of the screen (x), then from the law of inverse squares, we arrivei at the equation.

1°°°' -A = S50 A,

/ 16

in which "A" is the unknown candle power. Then by substituting different values of candle power, from seven to twenty-five inclusive, for "A" in the above formula, the setting of the screen nec- essary to give the balance for each candle power, when using the sixteen candle power standard, could be calculated. Then with an accurate meter scale, these different values for ''X" were laid off and marked their corresponding candle power. The distances between each candle power reading was divided into ten proportional divisions so that readings could be made very easily to hun-

-as-

dredtha of a candle power. These divisiona were then traced on a piece of tracing cloth which could be fastened to the photometer bar with stickers. The points of nine, sixteen and twenty- five candle power fell at even divisions on the old scale or at 750, 1000 and 1250 mm. marks, so that at each time of using, the accuracy of this derived scale could be easily checked.

PHOTOMETER SCREENS..,

In the case of the carbon filament and the Gem metallized lamps, the Lummer-Brodhun pho- tometer screen was used. It is considered that a more accurate balance can be obtained with this

screen for lamps of the same color, and since these

same lamps and the standard are the^^ color, it was con- sidered best to use it with these lamps. However, in comparing the white light of the Nernst, tan- talum and the timgsten lamps, a balance could not be as easily or as accurately obtained with the

-26-

LuEimer-Brodhun as with the Fliober photometer. Consequently, the Flicker was used with these three last mentioned types. In the case of the Nernat lamp, the tip candle power was obtained hy swin^the lamp into the horizontal position after it had burst into full light.

PABJI III.

■;:•-•'•*-- —

DATA AND CALCULATED RESULTS,

PHOTOMETERIC DATA AT THE SUpUT OF THE RUN. GEM METALLIZED LAiiP.

jrisB.	No.	Styla.	Candle ?ows; 1. 2,	Ave .
A	1	A.C.	14.90	14.84	14.870
A	8	A.C.	14,13	14.14	14.135
A	3	A.C.	14.43	14.47	14.475
A	4	A.C.	14. 15	14.11	14. 130
A.	5	A.C.	14.58	14.55	14.565
A	6	A.C.	13.70	13.72	13.710
A	7	D.C.	15.03	15.05	15.040
A	8	D.C.	14.02	14.00	14.010
A	9	D.C.	14.50	14.48	14.490
A	10	D.C.	15.57	15.65	15.630
A	11	D.C.	15.(31	15.48	15.555
A	12	D.C.	15.19	15.21	15.200

Average ~ 14.653

PKOTOi'ETERIO DATA AT END OF 50 HOURS RUN, GSIvl METALLIZE!^ LPJ-'iP .

Candle power. Series. No. Style. 1. 2. Ave.

A	1	A.C.	15.61	15.57	15.590
A	S	A.C.	14.89	15.95	15.930
A	3	A.C.	14.71	14.70	14.705
A	4	A.C.	14.95	14.99	14.970
A	5	A.C.	15.57	15.61	15.590
A	6	A.C.	14.39	14.43	14.435
A,	7	B.C.	15,14	15.17	15.155
A	g	B.C.	14.60	14.62	14.610
A	9	D.C.	15.54	15.62	15.5S0
A	10	B.C.	16.01	16.03	15.020
A	11	B.C.	16.12	16.16	16. 140
A	12	B.C.	16.49	16 . 45	16.470

Average - 16.433

PH0T0MET3RIC DATA AT END OP 100 HOURS RUN, GEM LIETALLIZED LAMPS.

		_-.-	-.**5J.5J	-
Sariea,	. No.	Style.	Candl3 power. 1. 2. Ave.
A	1	A.C.	15,73	15.80	15.755
A	2	A.C.	15.36	15.40	15.380
A	3	A.C.	15.37	15.36	15.365
A	4	A.C.	16.10	16.08	16.090
A	5	A.C.	16.65	16.67	16.650
A	6	A.C.	15.99	15.97	15.930
A	7	D.C.	15.14	16.18	16 . 160
A	8	D.C.	16.28	16.29	16.285
A	9	D.C.	16.41	16.43	16.420
A	10	D.C.	18.30	16.36	16.330
A	11	D.C.	17. IS	17.18	17.170
A	12	D.C.	17.10	17.08	17.090

Average - 16.224

-?Q-

PHOTOMETERIC DATA AT END OF 200 HOURS RUN. GEM METALLIZED LAIAPS.

-— **-*4.-

Oandle Power.

iriea.	No.	Style.	1.	2.	Ave.
A	1	A.C.	15.21	15.23	15.820
A	2	A.C.	14.67	14.65	14.660
A	3	A.C.	14.29	14.33	14.310
A	4	A.C.	15.19	15.13	15.165
A	5	A.C.	16.02	16.04	16.030
A	6	A.C.	13.91	13.90	13.920
A	7	B.C.	15.^2	15.59	15.605
A	8	D.C.	15.49	15.53	15.510
A	9	B.C.	15.52	15.50	15.510
A	10	D.C.	15.42	15.39	15.405
A	11	D.C.	16.27	16.31	16.290
A	12	D.C.	16.09	16.13	16.110

Averaga a 15.311

.31-

PHOTOMETERIG DATA AT END OF 3S0 HOURS RUN. GEM METALLIZED LAIvIPS.

Candle Power. Ssri93. No. Style. 1. 2. Ave.

A	1	A.C.	14.56	14.50	14.530
A	2	A.C.	13.76	13.78	13.770
A	3	A.C.	13.80	13.22	13.210
A	4	A.C.	14.60	14.54	14.570
A	5	A.C.	15.25	15.19	15.220
A	6	A.C.	12.87	12.85	12.860
A	7	D.O.	14.70	14.75	14.725
A	8	B.C.	14.38	14.40	14.390
A	9	D.C.
A	10	D.C.	14.73	14.79	14.760
A	11	D.C.	15.70	15.64	15.670
A	12	D.C.	15.30	15.35	15.325

Average - 14.457

PHOTOMETERIC DATA AT END OF 400 HOURS RUN.

GEM METALLIZED LAI^PS .

-»«-:;•« — ^-

Candle Power. Seriea, No. Style. 1. 2. Av9.

A	1	A.C.
A	2	A.C.	1^.45	13.48	13.465
A	3	A.C.	13.14	13.17	13.155
A	4	A.C.
A	5	A.C.	13.13	13 . 15	13 . 140
A	6	A.C.	11.37	11.32	11.345
A	7	D.C.	13.12	13.09	13.105
A	8	D.C.	12.88	12.90	12.890
A	9	D.C.
A	10	D.C.	13.32	13.30	13.310
A	11	D.C.	13.90	13.92	13 . 9 10
A	12	D.C.	IS. 94	12.96	12.950

Avarag© - 15.030

PHOTCl^ETERIC DATA AT END OF 500 HOURS RUN. GEii METALLIZED LAIvIPS .

Candle Powar. Series. No. Styla. 1. 2. Ave.

A	1	A.C.
A	2	A.C.
A	3	A.C.
A	4	A.C.
A	5	A.C.
A	6	A.C.
A	7	D.C.
A	8	B.C.
A	9	D.C.
A	10	D.C.
A	11	D.C.
A	12	D.C.

13.07	13.04	13.055
13.09	13.13	13.110
12.87	12.85	12.850
12.02	12.00	12.010
12.89	12.85	12 . 370
12.47	12.49	12.480
13.22	13.19	13.205
13.90	13.87	13.885
12.52	12.56	12.540
Avoraee -	12.891

■34-

PHOTOMETSRIO DATA AT END OF 600 H0DR3 RUN. GEM METALLIZED LAMPS.

Oandls Power. Series. No. Style. 1. 2. Ave.

A	1	A.C.
A	2	A.C.	12.54	12.57	12.555
A	3	A.C.	12.40	12.42	12.410
A	4	A.C.
A	0	A.C.	12.12	12 . 15	12.135
A	6	A.C.	12.24	12.20	12.220
A	7	D.C.	12.52	12.58	12.600
A	8	D.C.	12.10	12.11	12.105
A	9	D.C.
A	10	D.C.	13.14	13.17	13.155
A	11	D.C.	13.82	13.78	13.800
A	12	D.C.	12.22	12.19	12.205

Average - 12.576

■35-

PHOTOMETERIG DATA AT END OF 700 HOURS RUN. GEM METALLIZED LAi.IPS.

Candle Powar. Series. No. Style. 1. 2, Ave.

A	1	A.C.
A	S	A.C.
A	3	A.O.
A	4	A.C.
A	5	A.C.
A	5	A.C.
A	7	D.C.
A	8	D.O.
A	9	D.C.
A	10	D.C.
A	11	ouc.
A	12	D.O.

11.27 11.29 11.280 11.17 11.15 11.160

11.01	11.04	11.025
10 . 47	10 . 49	10 . 480
11.57	11.62	11.595
11.09	11.13	11.110

11.24	11.27	11.255
11.57	11.55	11.560
10.26	11.22	11.240
	Average	- 11.179

-36-

PHOTCMETERIC DATA AT START OF THE RUN. TANTALUM LALIPS.

Candla Power. Series. No. Style. 1. 8. Ave.

B	1	A.G.	10.06	10.11	11.085
B	2	A.O.	8.81	8.84	8.825
B	3	A.G.	10.72	10.69	10.705
B	4	A.O.	10.46	10.47	10.465
B	5	A.G.	9.98	9.89	9.885
B	6	A.G.	9.68	9.64	9.660
B	7	D.G.	10.39	10.41	10.400
B	8	D.G.	10. SB	10.32	10 . 300
B	9	D.G.	9.40	9.42	9.410
B	10	D.G.	10.49	10.50	10.495
B	11	D.G.	10.20	10.13	10 . 170

Average - 10.127

.r57-

PHOTOMETERIC DATA AT END OP 50 HOURS RUN, TAJITALUM LAIvlPS.

Candle Power. >eri©3. No, Style. 1. 2. Ave.

B	1	A.C.	12,97	12.91	12.040
B	2	A.C.	10.91	10.85	10 . 880
B	3	A.C.	11.72	11.75	11,735
B	4	A.C.	10.87	10.90	10.885
B	5	A.C.	10.31	10.64	10.625
B	6	A.C.	10.21	10.18	10.195
B	7	D.C.	11.74	11.79	11.775
B	8	D.O.	10.67	10.65	10.660
B	9	D.C.	9.72	9.76	9.7'iO
B	10	D.C.	10.63	10.56	10.595
B	11	D.C.	10.69	10.71	10.700

Averag© - 10,894

-38-

PHOTOMSTERIC DATA AT END OF 100 HOURS RUN. TANTALUM L.^P3. — -^^:.^.—

Candle Power.

isrias.	, No.	style.	1.	2.	Av(
B	1	A.O.
B	2	A.O.
3	3	A.O.	12.78	12.80	12.790
B	4	A.G.	11.67	11.69	11.580
B	5	A.O.	11.18	11.12	11.150
B	6	A.C.	11.02	11.09	11.070
B	7	D.O.	12.46	12.48	12.470
B	e	D.O.	11.61	11.65	11.330
B	9	D.C.	10.34	10.37	10.355
B	10	D.O.	10.87	10.92	10.895
B	11	D.C.	11.88	11.86	11.870

Avsrags - 11.545

-59-

PHOTOMSTERIC DATA AT SNB OF 20C HOURS RUN, TANTALUM LAMPS .

Canils Power. Series. No. Style. 1. 2. Ave.

B	1	A.C.
B
B	2	A.C.
B	3	A.C.
B	4	A.C.
B	. 5	A.C.
B	6	A.C.	10.94	10.93	10.935
B	7	D.C.	11.24	11.27	11.255
B	8	D.C.	11.02	11.00	11.010
B	9	D.C.
B	10	D.C.
B	11	D.C.

Avorage - 11.066

■40-

PHOTOMETERIC DATA AT END OF 300 HOURS RUN. TANTALUM LAMPS.

Series. No. Stlye

1 2 3 4 5 6 7 g

9 10 11

A.C A.G A.C A.G A.C A.C D.C D.C D.C D.C D.C

Candle Power. 1. 2. Avw.

10.78 10.84 10.810

PHOTOMETKRIC DATA AT END OP 400 HOURS RUN.

B

6

'.22 9.20

9.210

■41-

PHOTOMETERIG DATA AT THE START OF THE RUN. TUNGSTEN LMIPS.

			Candle Power.
■ies.	NO.	Stylo.	1.	2.	Ave.
C	1	A.C.	18.69	18.71	18.700
'■J	2	A.G.	18 . 15	18 . 27	18.210
0	3	A.C.	17.70	17.65	17.675
c	4	A.C.	18.62	18.47	18 . 545
Q	5	A.C.	18.98	19.17	19.075
c	6	A.C.	18.23	18.58	18.305
c	7n	CO.	17.67	17.12	17.395
c	8	C.C.
c	9	D.C.	18.06	18.11	18.085
c	10	D.C.	18.53	16.44	18.485
c	11	D.C.	l^.SO	18.44	18.370
c	12	D.C.	18.70	18.92	16 . S 10

Average - 18.33J

-42-

PHOTOMETERIO DATA AT END OF 50 HOURS RUN. TUl^GSTEN LAMPS.

Candle Power, Series. No. Style. 1. 2. Ave.

0	1	A.C.	19.01	ie.98	18.995
0	2	A.O.	19.12	19.09	19.105
c	3	A.C.	18.12	18.17	18,145
c	4	A.C.	18.87	18.89	18 . SSO
0	5	A.C.	19.51	19.53	19.520
c	6	A.C.	19.47	39.51	19.490
c	7	D.C.	18.21	18.23	18.220
0	8	D.C.
c	9	D.C.	19.01	19.04	19.025
0	10	D.C.	19.05	19.08	19.065
c	11	D.C.	18.84	18.81	18.825
c	18	D.C.	19.14	19.11	19.125
			Av©	rage -	18.945

PH0TCMET3RIC DATA AT END OF 100 HOURS RUN. TUNGSTEN LALIPo.

Candle Power. Series. No. Style. 1. S. Ave,

0	1	A.C.	21.38	21.40	21.390
c	2	A.C.	20.74	20.77	20.775
0	3	A.C.	20.40	20.44	20.420
c	4	A.C.	19.12	19.18	19 . 150
G	5	A.C.	20.51	20.52	20.515
0	6	A.C.	21.45	21.47	21.460
0	7	D.C.	19.19	19.25	19.220
c	8	D.C.
c	9	D.C.	19.75	19.74	19.745
0	10	D.C.	19.90	19.86	19.880
0	11	D.C.	19.03	18.99	19.010
c	12	D.C.	20. 2S	20.25	20.235

Average - 20.164

-44-

PHOTOMETERIO DATA AT END OF 200 HOURS RUN. TUNGSTEN LAMPS.

^ies.	No.	Style.	Candle Power. 1. 2. Avi
C	1	A.C.	21.42	21.47	21.445
C	2	A.C.	20.70	20.69	20.695
C	3	A.C.	20 . 6'f	20.65	20.660
c	4	A.C.	19.27	19.22	19 . 245
c	5	A.C.	20.04	20.07	20.055
c	6	A.C.	SI. 29	21.32	21.305
0	7	D.C.	20.57	20.52	20.455
c	8	D.C.
c	9	D.C.	19.70	19.73	19.7 25
c	10	D.C.	20.10	20.07	20.085
c	11	D.C.	19.07	19.04	19.055
0	12	D.C.	19.93	19.97	19.950
			Average -	20.245

■45-

PHOTOMETERIC DATA AT END OF 500 HOURS RUN. TUNGSTEN LAMPS.

Caai.dle Power. Series. No, Style. 1. 2. Ave,

0	1	A.C.	21.71	21.72	21.715
c	2	A.C.	20.68	20.64	20.660
c	3	A.C.	20.92	20.87	20.895
c	4	A.C.	19.32	19.28	19 . 300
c	5	A.C.	19.67	19.72	19.695
c	6	A.C.	21.28	21.21	21.245
c	7	D.C.	21.45	21.38	21.465
0	8	D.C.
0	9	D.C.	19.72	19.77	19.745
0	10	D.C.	19.74	19.69	19.715
0	11	D.C.	19.08	19.10	If .090
c	12	D.C.	19.87	19.90	19.885

Average - 20.301

FHOTOMETERIC DATA AT END OF 400 HOURS RUN,

TUNGSTEN LAMPS .

Candle Power.

Series.	, No.	Style.	1.	2.	Ave.
C	1	A.C.	50.27	20.31	20.290
c	s	A.C.	19.62	19.57	19.595
c	7,	A.C.	20.65	20.63	20 . 640
c	4	A.C.	18.00	17.60	17.800
0	5	A.C.	19.15	19.20	19.175
0	6	A.C.	20.25	2C.25	20.240
c	7	D.C.	19.86	19.89	19.875
c	8	D.G.
G	9	D.C.
0	10	D.C.
c	11	D.C.	18.50	13.54	13.520
c	12	D.C.	18.92	18.90	18.910

Average -19.228

■47-

PHOTOMETERIC DATA AT END OF 500 HOURS RUN. TUNGSTEN LAilPS.

Candle Power. Series. No. Style. 1. 2. Ave,

0	1	A.C.	19.47	19.42	19.455
0	2	A.C.	19.22	19.27	19 . 245
c	3	A.C.	20.02	20.00	20.010
0	4	H.C.	20.07	20.04	20.055
0	5	A.O.	19.00	18.99	18.995
c	6	A.C.	19.24	19,27	19.255
c	7	D.C.	19 . 34	19.31	19.325
0	8	D.O.
0	9	D.C.
G	10	D.C.
c	11	D.C.	18.72	18.69	18.705
0	12	D.C.	18.67	18.59	18.630

Average - 19.295

I

•48-

FHOTOMETERIC DATA AT 5ND OF 600 HOURS RUN.

TUNGSTEN LAMPS,

---*•;!■*

Candls Poorer. Series. No. Style. 1. 2. Ave,

c	1	A.C.	18.98	19,00	16,990
c	2	A.C.	18.42	18,38	18 . 400
c	3	A.C.	19.21	10 . 17	19 , 190
0	4	A.C.	18.61	18,59	18.600
c	5	A.C.	18.80	18,83	18,815
0	6	A.C.	18.64	18,63	18.635
c	7	D.C.	18,87	18,91	18,990
c	p	D.C.
n	9	D.C.
c	10	D.C.
c	11	D.C.	18 . 90	18,91	18.905
c	12	D.C.	17.9?	17.97	17.945
			Av:	3 rage	18,720

-49-

PHOTOl'ETERIO DATA AT END OF 700 HOURS RUN. TUNGSTEN LAiilPS.

Candla Power. Sariea. No. Stlye. 1. 2. Avs

17.80 17.84- 17.8S0

c	1	A.G.
c	2	A.G.
--1	3	A.G.
0	4	A.C.
c	c	A.O.
0	6	A.C.
c	7	D.G.
r%	B	D.O.
u	9	D.C.
»>	10	D.G.
c	11	D.O.
0	12	D.O.

17.55 17.59 17.570

17.67 17.63 17.650 17.42 17.40 17.410

Average 17.608

-50-

PHOTOMETT^Rin DATA AT THE START OF THE RUN,

NERNST LAlaPS.

«-** — -

Screen Setting.

)ri©e. NO. Style. 1. 2. Ave. C.P.

D 1 A.C. 90.04 92.70 91.370 55.00

D 2 A.C. 92.85 92.34 92.595 54.98

D ? A.C. 90.80 91.10 90.950 54.00

D 4 D.C. 92.80 92.9 5 92.875 53.7; 1

D 5 D.C. 92.70 92.50 92.600 54.92

D 6 D.C. 92.01 91.80 91.900 52.90

Average -•- 54. 18

■51-

PHOTCMETERIO DATA AT END OF 50 HOURS RUN.

Series. No. Style,

Screen Setting. 1. 2. Ave.

O.F.

A.C A.C A.I D.O D.C D.G

80.42 80.57 80.4P5 41.30

82.26 as, 14 BP.200 43.20

PI. 49 82.21 81.850 42.84

82.52 80.72 81.620 42.21

80.42 80.79 80.605 41.62

79.97 80.12 80.045 40.59

Average

41.96

PHOTOMETERIC DATA AT END OF ICO HOURS RUN. NERNST LMIPG,

Screen Setting. Series. No. Style. 1. 2. Ave. C.P,

D	1	A.C.	68.8	68,4	68.70	32.10
D	2	A.C.	68.6	68 . 4	65.50	32.00
D	3	A.C.	38.1	68.0	68.05	31.80
D	4	D.C.	75.1	73.?.	73.20	34.20
D	5	D.C.	73.6	73.1	73. 35	34.80
D	6	D.C.	71.8	71.2	71.50	33.40

Average - 33.13

■53-

FHOTOMETERIG DaTA A1 END OF 200 HOURC RUN. NERNST Lx'^PS.

Screen Setting. Series. No. Style. 1. 8. ^ Ave. C.P.

64,5 65.0 G4.75 S6.P5

66, Z 65.7 66.00 f57.92

65.7 64.9 65.30 P7.S5

D.	1	A.C.
D	2	A.C.
D	3	A.C.
D	4	n.c.
D	5	D.C.
B	6	D.C.

i5.2 65.7 65. 4& 27.51

Average - 27.38

FHOTCMETERIC DATA AT END OF 300 HOURS RUN. TJSRIT3T L/JIPS.

Screen Setting. Series. No. Style. 1. 2. Ave. C.P,

D	1	A.G.	59.-1	58.8	59.10	22.35
D	2	A.C.	57.8	57.9	57.85	21.4-C
D	l^;	A.C.	57.9	58.1	5B.00	21.00
D	4	D.C.
D	5	D.C.
D	6	D.C.	59.6	59.1	59. ?5	22.47

Average - 21.9?;

•55-

PHOTUMETERIC DAra AT ENlJ OF 400 HOURS RUN. NERNST LAM^S.

Series.	JMO.	Style
D	.1	A.O.
D	2	A.C.
D	3	A.C.
D	4	D.C.
B	5	D.C.
D	6	D.C.

Screen Setting. 1. 2. Ave. C.P,

55.4 55.1 55.25 IV. 60

55.7 55.2 55.45 20.10

56.2 56.0 56.10 21.20

i.8 55.50 20.22

Average - 20.28

■56-

HHOTOMSTERIC DATA AT END OF 500 HOURS RUN, NERN3T LAMPS.

SerieB. No. Style. 1. 2. Ave. C.P.

18.50

18.10

D	1	A.C.	53.7	53.2	53.45
D	2	A.O.	53.1	53.6	53.35
D	3	A.C.
D	4	D.C.	— —
D	5	D.C.
D	6	D.C.	53.1	53.5	53.30

18.40 Average - 18.33

■57

PHOTOMETERIG DATA AT END OF 600 HOURS RUN. NERNST LAMPS. ■»-;{••:;——

Screen Setting. Series. No. Style. 1. 2. Ave. C.P.

D 1 A.C. 51. P 51.6 51.40 16.90

B 6 D.C. 5S.? 51.8 52.00 17.10

Average - 17.00

PHOTOMETERIC DATA AT SlTb OF 700 HOURS RUN. D 1 A.O. 50.4 50.8 50.60 16.40

-58-

fhotci,:eteric data at the start of the run. g.1rb0n lamps.

Series. No. Style.

Candle Powar. 1. 2. AvG.

E	1	A.C	12.01	12.01	12.010
E	9.	A.G	13.40	13.43	13.415
E	7,	A.C	10.79	10.77	10.780
E	4	D.C	10.75	10.81	10.780
E	5	D.C	10.90	10.97	10.935
E	6	D.C	10.52	10 . 50	10.510

Average

11.408

-59-

PHOTOMETIiRIC DATA AT END OF 5C HOURS RUN, C/iRBON LAIvIPS.

Candls Power.

2. Ave.

Series.	No >	Style
E	1	A.O.
E	o	A.C.
E	rt	A.C.
E	4	D.C.
E	5	D.C.
E	6	D.C.

12.31	12.33	12.320
15 . 67	13. o9	13.680
11.17	11.19	11.180
11.21	11.19	11.200
11.49	11.52	11.505
11.85	11.89	11.870
	Average -	11.959

-60-

PHOTOHETSRIC DATA AT iNH P? 100 HOURS RUN. CARBON LAN^PS.

Candle Powsr. Series. No. Style. 1. 2. A

E	1	A.C.	12.73	12.80	12.900
E	2	A.C.	14.22	14.25	14.235
E	3	A.C.	12.20	12.30	12.250
E	4	D.O.	11.69	11.72	11.705
E	5	D.C.	12.01	12.02	12.015
E	6	D.C.	12.48	12.46	12.470

-v&rage - 12.596

il-

PHOTCMETERIC DATA AT END OF 200 HOURS KUN, CARBON LMIPS.

Candle Powsr. Series. No. Style. 1. 2. Ave.

E	1	A.C.	12.23	12.27	12.250
E	2	A.O.	12.42	12.39	12.405
E	3	A.C.	11.87	11.91	11.890
E	4	D.C.	11.21	11.19	11.200
E	r	D.C.	11.87	11.84	11.855
E	6	D.C.	12.21	12.20 Average	12.205 11.967

.6S-

FHOTCMEKRIC DATA AT ZND OF 300 HOURS RUN. CiVRBON ■ LAIvIPS.

leriea. No. Style.

E	1	A.C.
E	2	A.C.
E	3	A.C.
E	4	D.C.
E	5	D.O.
E	6	D.C.

	Candle Power.
1.	2.	Ave.
11.10	11. OS	11.090
1^.98	13.03	13.005
11.00	11.0?	11.010
10.94	10.95	10.945
11.16	11.12	11.140
11.91	11.94	11.925

Average - 11.519

FHOTOMETERIC DATA AT END OF 400 HOURS RUH, CARBON LAMPS.

Candle Power.

Series.	No.	Style.	1.	2.	Ave.
E	1	A.C.	10.87	10.89	10.880
E	c	A.C.	11.28	11.30	11.290
E	3	A.C.	10.40	10.42	10.410
E	4	D.C.	10,52	10.56	10.540
E	5	D.C.	10.24	10.21	10 . 225
E	g	D.C.	10 . 46	10 . 49	10.475

Average - 10.637

•64-

PHOTOMETZRIC DATA AT END OF 500 HOURS RUN, CARBON LAI/IPS.

Candle Power. Series. No. Style. 1. 2. /

E	1	A.C.	10.32	10.37	10.345
E	2	A.C.	10.94	10.92	10.930
E	3	A.C.	10.07	10.03	10.050
E	4	D.C.	10.21	10 . 17	10 . 190
E	5	D.C.	9.94	9.91	9.9S5
E	6	D.C.	10.00	9.97	9.985

Average -10,237

■65-

PHOTCMETERIC DATA AT END OF 600 HOURS RUN. CARBON LAIvIPS.

-K-^f-?!-?!— --

Candle Power. Series. No. Style. 1. 2. Ave,

E	1	A.C.	9.94	9.96	9.950
E	2	A.G.	10.77	10.79	10.780
E	3	A.C.	9.61	9.58	9.595
E	4	D.C.	9.92	9.90	9.9 10
E	5	D.C.	9.31	9.29	9.300
E	6	D.C.	9 .57	9. 54	9.555

Average - 9.848

-66-

PHCTCMETERIO DATA AT %]ND 0? 700 HOURS RUN.

OAREON Li^J,lPS.

-:;-**

Candl.9 Power. ;ori@3. No. Styls. 1. 2. Ave,

E	1	A.C.	9.87	9.90	9.885
E	2	A.C.	10.01	10.04	10.025
E	3	A.C.	9.42	9.48	9 . 450
E	4	D.C.	9.61	9.59	9. 600
E	5	D.C.	9.27	9.24	9.255
E	6	D.C.	9.34	9.32	9.330

Average - 9.570

.67-

CANDLE POWER DURING LIFE OF GEM METALLIZEU LAMPS,

.— -_ j;.:tt.A

Lamoa,

HOTJrs. l;. ?^ 3/ 4^ 5_j_ 6^

0 14.870 ; 14. 1:55 14.4-75 14.130 14.565 13.710

50 15.590 14.930 14.705 14.970 15.590 14.435

100 15.765 15.380 15,365 16.090 16.350 15.930

200 15.??0 14.660 14.310 15. 165 16.030 13.920

300 14.530 13.770 13.210 15.2h'G 12.S50

400 12.435 13.155 13.140 11.345

500 13.055 13.110 — ^ ~ 12.860 12.010

300 — 12.555 12.410 12.135 12.120

700 11.280 11.160 11.025 10.480

CANDLE POWER DURING LIFE OF GEM METALLIZED LAMPS.

Lamps .

Hours. 7_j_ 8. 9. 10. 11. 12.

0 15.040 14.01C 14.490 15.660 15.555 15.200

50 15.155 14.610 15.580 16.020 16.140 16.470

100 16.160 16.285 16.420 16.330 17.170 17.090

POO 15.605 15.510 15.510 15.405 16.290 16.110

300 14.725 14.390 14.760 15.670 15.325

400 13.105 12.890 13.310 13.910 12.950

500 12.870 12.480 13.205 13.885 12.540

600 12.600 12.1C5 13.155 13 . BOO 12.205

700 11.595 11.110 11.255 11.560 11.220

-69.

CANDLE POWER DURING LIFE CF TANTALUM LMSFS,

_---*•»•?;■«——.«.

Lamps.

Hour 8 . 1. ?. 3. 4. 5. 6.

0 11.085 8.825 10.705 10.465 9.885 9.660

50 15.990 10.880 11.735 10.885 10.6S5 10.195

100 12.790 11.680 11.150 11.070

200 - __-— — -^^« 10.935

300 — 10.810

400 9.210

500 "- -— »— -_„^--

600 — „-^^

700 —

-70.

CANDLE POWER DURING LIFE OF TANTALUM LAMPS.

.**•«••»—-

Larapa. Hours. 7. 8. 9. 10. 11.

0 10.400 10.300 9.410 10.495 10.170

50 11.775 10.660 9.740 10.595 10.700

100 12.470 11.630 10.355 10.895 11.P70

200 11.255 11.010 z:^

300 '—

400

500

500 •>-

700 ^-

-71-

CANDLE POWER DURING LIFE OF TUNGSTEN LAIvIPS.

Lanps. Eours. 1^ 2. 3. 4. 5. 6.

0 16.700 18.S10 17.675 16.545 19.075 18.305

50 IS. 995 19.005 18.145 18.880 19.520 19.490

100 21.390 20.755 20.420 19.150 20.515 21.460

200 21.445 20.695 20.660 19.245 20.055 21.305

300 21.715 20.660 20.995 19.300 19.695 21.245

400 20.290 19.595 20.640 17.S00 19.175 20.240

500 19.445 19.245 20.010 20.065 18.995 19.255

600 18.990 18.400 1919.0 18.600 18.815 18.635

700 17.820 17.570

72-

CANDLE POWER DURING LIFE OF TUNGSTEN LMIPS.

— -«•!<•«•*-—

Lanps.

Hours. 1. 8. 9. 22* 11' l^.

0 17_.3G5 18.085 19.485 IP. 370 18.810

50 18.220 ^^- 19.025 19.065 18.925 19.125

100 19.220 19.745 19.880 19.010 20.235

200 20.455 — - 19.715 20.085 19.055 19.950

300 21.465 19.745 19.715 19.090 19.885

400 19.875 18.520 1.8.910

500 19.325 "- 18.705 18.630

600 18.990 • IS. 905 17.945

700 ■ 17.650 17.410

■73-

CANDLE POWER DURING LIFE OF THE NERNST LALIPS,

HourB	• 1-	p	^'	Lamps. 4.	c •	6.
0	53.31	54.92	52.90	55.00	54. 9B	54.00
50	41.30	43 . 20	42.84	42.21	41.62	40.59
100	32.10	32.10	31.80	34.20	3-. PC	33.40
??00	26.85	27.92	27.25			27.51
500	eg JR	21.40	21.50			22.47
400	19.70	20 . 10	21.20			20.22
500	18 . 50	18 . 10				IP. 40
600	16. ?0					17 . 10
700	16 . 40

-74-

;andle power during life op GAREON LAJ.!PS.

-««-**,•{.*—.-«

Lamps . Foura. 1. 2. 3. 4,

0 12.010 13.415 10.780 10.780 10.935 10.510

50 12.320 13.680 11.180 11.200 11.505 11.870

100 12.900 14.235 12.250 11.705 12.015 12.4-70

SOO 12.250 13.405 11.890 11.200 11.855 12.205

300 11.090 13.005 11.010 10.945 11.140 11.925

400 10.880 11.290 10.410 10.540 10.225 10.475

500 10.345 10.930 10.050 10.190 9.925 09.985

600 9.950 10.780 9.595 9.910 9.300 9.555

700 9.885 10.025 9.450 9.600 9.255 9.330

RUNNING- LOG.

Mc.	Day.	Start.	Finish.	Hrs.i-Uii.	Total,
Feb.	5	9; 30 A.M.	12:00 P.M.	14.5	14.5
	8	8:00 " "	12:00 " "	16.0	30.5
	9	S:00 " "	12:00 ? "	16.0	46 . 5
	10	8500 " "	11:30 A.M.	3.5	50.0
	10	4:30 P.M.	12:00 P.M.	8.5	58 . 5
	11	8:00 A.«.	12:00 " "	16. C	72.5
	12	4:30 P.M.	S:00 " "	1,5	74.0
	13	4:45 " "	12:00 " "	7.75	81.75
	15	8:00 A.M.	8.15 " "	12.25	94.0
	16	8:00 " ""	12:00 " "	16.0	110.0
	17	8:00 " "	3:45 " "	7.75	117.75
	18	B:00 " "	12:00 " "	16.0	133.75
	19	8:00 " "	12:00 " "	IS.O	149.75
	20	8:00 •• "	8:45 A.M.	.75	150.5
	9.'5	9.15 " "	6:45 P.M.	9.5	150.0
Mar.	1	8:00 " "	12:00 " "	16.0	176.0
	8	8:00 " ••	12:00 " "	16.0	1.92
	3	8:00 " "	12:00 " "	16.0	208.0
	4	8:00 " "	12:00 " "	16.0	224.0

-76-

Mo. Day. Start.

Mar. 5 8: CO A. I

5 8:00 " '

6 5.30 P.] a 8:00 "

9 8:00 "

10 8:00 "

11 8:00 "

15 8:00 " 15. 10:00 "

16 8:00 "

17 3:00 "

18 8:00 "

19 8:00 "

20 9:00 "

22 8:00 "

23 8:00 "

24 5:00 P.

25 3:00 A.

26 8:00 "

Finish.

9:00 P.M. 12:00 M. 12:00 P.M. 12:00 " " 12:00 " " 12:00 " " 12:00 " "

2:00 " " 12:00 " "

9:30 " " 12:00 " " 12:00 •* " 12:00 " " 12:00 M. 12:00 P.M.

9:00 " " 12:00 " "

7:30 " " 12:00 " "

Hra.ri-m. Total.

13.0 237.0

4.0 241.0

10 . 5 247 . 5

16.0 263.5

16.0 279.5

16.0 295.5

16.0 311.5

6.0 3 17 . 5

14.0 331.5

15.5 345.0

16.0 361.0

16.0 377.0

16.0 393.0

4.0 397.0

16.0 413.0

15.0 423.0

7.0 433,0

11.5 444.5

16.0 460.5

-77.

Mo. Day.	Start.		Finish.	Hrs.run.	Total
Mar. 30	8:00 A.	M.	12:00P.M.	16.0	476.5
31	8:00 "	"	12:00 ? "	16.0	492.5
Apr. 1	3:00 "	"	7:30 " "	11.5	504.0
2	8:00 "	»	7:00 " "	11.0	515.0
6	10:00 ••		12:00 " "	14.0	529.0
7	8:00 "	tt	12:00 " "	16.0	545.0
8	8:00 "	fi	12:00 " "	16.0	561.0
10	8:00 "	"	12:00 a."	4.0	555.0
13	8:00 "	"	12:00 P.M.	16.0	581.0
13	3:00 "	"	12:00 " "	16.0	597.0
14.	8:00 "	"	12:00 " "	16.0	613.0
15	8:00 "	"	3:00 " "	7.0	620.0
19	11:00 "•	rt	12:00 " "	13.0	633.0
20	8:00 "	"	12:00 " "	16 .t)	649.0
21	8:00 "	ft	12:00 " "	16.0	665.0
22	8:00 "	"	12:00 " "	16.0	681.0
23	8:00 "	"	12:00 " "	16.0	697.0
24	9:00 "	••	12:00 M.	3.0	700.0

.78-

ANQLS PO-rSR DURING LIFE OF DIFFERENT L;.IT?3 TESTED, VALUES ;JIE AVERAGE OF EACH 3ERIE3. *;!-»*

Roura	.		Ser/es.
	A	B	C	JJ	£.
0	14.653	10 . 127	13.332	54.180	ll.-iC8
50	15.450	10.394	18.945	41.960	11.959
100	16.223	11.545	20.164	33.130	12.596
??00	15.311	11.066	20.242	27.380	11.967
300	14. 457	10 . 8 10	20.301	21.930	11.519
400	12.030	9.210	19.228	20.280	10.637
500	12.891		19 . 298	18.330	10 . 237
600	12.576		18.720	17 . 000	9.848
700	11. 179		17.608	16.4000	9.570

PART IV.

CURVES AND DIAGRAilS.

■i^l

I

'7KMir,MIT:^KKMI,

AKKotra

	m^
				SI	■Ml	IH	BM|	^
	m			■	H		HH	■
	^^^HB
	■^	i			V		n	1
	■	I I			ill		RH	■
		5					MHHB

utMova

i^'^ '

^&«^^^/

li|HHi

1 .;	■|					■1
	■1					1
	■1		S	MHI	— ^,	^^,	Hi
	V		i				)U:
	SI		"^	i^-			.._; 1 _: /
	■I			K-
	■1				r '
	^^m	F5|			b
1 1- .	«c	1				L^.	1
	J				^!	. ■ 1
S3 ^	\ — 3			:.:>:f!SS5S	R^R	i
^ ft)	J
					L		«0
f "					^ ^
					^
						^
				i !J
	> 1		1 J 1 /
f			i
	1 — 1 J-	"1 ■	( /				1
;							• ^
					!		'
			\
				\
			;	\	1 "^^	7M0c/	1 31 any 0

mB& Wf'M

AEKOXnS

- '- |--:;- :	1					mz:^'
~; '' '	1					1
t-. !	1	m^m	jiiKjyfe?!:	w		,.ji
			i	1		m\ :
				■		^
				m		1
_ . _ 1... . i--- -				m	!■!
				m	iBl
j ^ ^	i_		i	i		■i -
1		1	i		W^L-
-1 - -t	1			1		wM
	f " "	1		1
\ : 1 ;._ __	1		■^fl	^tmmkm	^
" ""i"^"V|-^	t				/
— -f . -^^ -					y
		i		.	/	$
!			7	1
i Uj			<	„/
- — — ^ 51				/	\
-'h^- " q	I			/	1
	1		j y			1 --
1 1			\/
	V	\
		_^	' ^ '			'
		\^^		^____
		'	\
			\		I
1		■		K -	iSMOJ.	3inN>/0

	J ..... ,		! __				I	■	V	I
			1				I	■	K	\ — -
								I	■	■
					1			■	■	P"
					1 . 1				Ul
					1
	1					i	[		1/
		_- _ u,		:		1	[		'		1
		5					1 1
		T		[.	I . . . -				i
		~ - - >sl	S;	!	r 1		_ j		L ft.
	i J	- - N						/^			1 ||C.
	1	N					/				: 5
		i3				7	r				rr^:
	i	u.	^	1	. i	/					1 ^
	j		?		j	/					i ^ 1 SJ
/
JS-,		' '■ 1	r		j			1
		k !<	^		T			1 1
	^	- ^		; 1	/			1 i
	i
					/ ^
					/
				j	/				i	j
				-4
				1	\ i	1
			j	\
				1
		i		1	i/SMC	w ^4	/TA/v:)

vMjfCf.Jitm-

] L 1	: -:- '^■							m	~1^
- :" ]									Wk
1		. ''	^^	-^-j	■g.						■1
				1	1						\
^__ i			- : -1	m
	!			M
				m
				1	^	^	. .'r.; - '	^^		■iMH&n
		0.		1 1	■ !				^	1 — ^ — l_ _	-^5
		J	__^QCL_ .	i					t	i . ! 1 i
_j		^	«t		1			/	1 ]	,
" ■ ■■ 1		(5	SI	J i	r" -']			/		^^
			^					f		^ ^ Q
			:J:.	j						^
			.		^ ^-^				1.. \	T
		J	^							u* S
		l<i	- (4		\ i					^^
j
1		•>	«S -	i 1 i	/
1	•r		/		1 ;			h-— §
		^1		i	/					' N
					/
				. :/ ;
			/						<a
			N	V.	x		j			^
			^
1 1					\	i -^^^
					A	MOd	\37aN	'^^ O

txMorra

■iJBMOd 31lI/VVi

mwK^^

	. ^.
1	!
m
T " j
	1 <o
]	■»
	i
	^
	i^
]. z	1 M-
i
I
i
HA/VD
^- ,	-

) ^K • V, , 's

1

.]

AKWotra i

^Uf^'^r

AXOKOVa

l^ 'i

^^K^MWtKtBBt^

AStWOTrB

AHKotra

MtSSETTUMS VJt rSUIUIOJMM»k

i

S^5^5 S^^SSS^BBB^Baa^BsgS^BmagggSSSSaiaasaa ■

ABKotra

.'J'JWV")1><«W»

m-H

AKMOtrU

ppi:

		r.
	1 • t •■ 1	■
j
	!	kq	^	._
			2
		/^Mff\r\ 3	7n:Nvo

ABtKOtni

It k'

*

'Jiem^tt^

wwp%Mm

	1 IMIHI^^^I^Ih
	1 -^^MMM^B^B^BB
	|.,,..,.,^ ,.. ,.^ . . -^SKm^K^^^BM
	wsttsssS^B^B^B^B^B^
	^^^^BiHfllHHiJIHiiiliiHiiH
	l^^^^B^^^^M|^^^^^H^HI
	H^HhI
	■H
	I^^HIHH
	^^^^^^^^Sh^^^^BHHMH|b
	I^^^^^H

AJEUtOtns

AJunovnt

lK?v ^#f1yr^

~1

i I

*ts««>a>i«»

,^M

i

m

lLutl:,JhiiiJa

...

1

I--

PFi

1

__,. 4

J

i

_,_^ -J

1 1 .

\

-S

r

.-

i

^Jk ^

OA.

1 " -

1

■^ •^

'<4

1

1

I

^

1

111

^

'

' i

r

:i

^J

1

O-

^

j

^<i

k,

; 1^

"1

\~

i

k

U4

1

._^

Vi '

1-

i"

f—

r

i

i

\

r — t

:_-

i i

1 —

\

:\

J7CA'

J.

f'^M

VJ

"" — ~i			] ■■	1
--- , i
1			1 '		j		1
1			! 1				1
					i
			i		1
							1
		, 1		i
		! I (A						\
			cJ		1				i
'. ' . ]			[ ^
' ' - j	i	_J:									.1
\:.'- '-\	L		1 ^
	1		r v^j		1						^
		9									^
		^	L— *»_ -					\J-"			L.._..__ tj_.
			•A4	1			I
		k	t	i^_^ _j	L .-	^
			r v^' ... 1
					/
				1 ■				1 1
				\ ,_.				1 - - '
	1		i		1 1				1 ' ,
				\		\ 1		t '
						\
									1
						!^	31	IINVJ
	___			^^MC

rM^i:r■^ ^

^r^^Sl	11	m
	■	L-
	■	li	H	■H	Hi	K
	■	■	■	m	■|	i
	H	■	■			i.
	■	■	■	m	m	■■		■■

I

i	V
«o ,
i	:3
	o ^
	1	•< M
	i
1	j	1
1	I	t
	1
■I7ff^/\	'J

m^mM-'

wMt^r-' "ix

AJtwocra

■At •ii'aoiULjr'Wju»«*»

		1
	1

mmm:^ '

AEMotra

«J£MOtr»

1	■H5	1
1	1	^^S		1

iSHiis::;;

;s;R::;x::si:Rs::r

lynHOIl

h:srssb:s

»:r:ii:n»S::H»s|s

:»»:S

•m!!!!!»

i«i^&a:^:3t7SiMa

i?la'!j(yjj3jpO

usMovra

SmminiiUK^mSi&iSt

^^iuft-l

	j^^HH
	m^^^n
		^^^^^^^^^^^^^^^■bw
		^^^^^^^^^^^^^^^^^^^^^^^BB*af H
		^^^^B^^^fe^SBB^B
		I^^^H
		^^^^^^^mi^^g^g^^^
	^^^^^^^^^^^^^^^^^^^BSffiS
	B^^^^^^HhHirI^^s
	H^H^nilillMilfimlllU

tEKOTra

nsffr^^irs

O'.* acKcjUjMoi^oj

	■■i

eiz

AfiMomr

ASKOUa

AKMOtrx

		^V
^^	^^	^
m ^..._,i	1
iM		% _ _^_^.._^.-^..^ -__
J- -•
^^	^^	^ ^^^^

.t^

^^\

PART V.

**•»

30NCLUSI0N,

-79-

Th© thres photographs which have boen plac- ed at the beginning of this theaia,3hov7 the appar- atus as it was set up in the laboratory for the purpose of carrying on the test. The first view gives an idea of the kind of apparatus lised to gen- erate the power for the lamps, and the position of the lamps as tested, together with the recording apparatus. The wiring diagram, showing all the con- nections for this part of the apparatus, ia shown in Fig, 47. The second view is that of the photome- ter table and all the apparatus as used to measure the candle power of the lamps, while the wiring dia- gram for this is given in Fig. 48, The third view shows the five types of lamps tested. They are pla- ced from left to right according to the serial classification which has been carried throughout this thesis. The lamp at the extreme left is the Gem metallized, aeries ''A";the second is the tanta- Ixim lamp, serie3"B" ; the third is the tungsten, aeries "Cjthe fourth is the Nernst, series "D"; and the

■80-

fifth ia the carbon filament lauap, series "E". This picture will give a clearer and more comprehensive idea of the comparative size and shapes of the bulbs than the description as given under Part I of this thesis.

Pig. 50 shows the curve of the voltage as tra- ced by the recording voltmeter on Tuesday, April 20. This particular curve was selected at random from the number that were taken, not to show any partic- ularly good features but merely as indicating the average conditions of the voltage on the lamps.

It will be noticed that the voltage on start- ing at 8:00 A.M. does not reach 110 at once and that it took a half hour or so to do it. This was due to the following condition. At the beginning of the test, it seemed desirable to run the lamps twenty-four hours a day, sixteen on the line and eight on the storage battery but in trying this scheme, it was found that the load was too heavy for the battery and that four hours run was all that it could possibly stand. Consequently, the test could be

11-

run but sixteen hours a day at the moat. This rest of eight hours gave the machines a chance to cool down regardless of the fact that the fields were kept warm by connecting them to the battery during this period. And therefore, when starting in the moiling, it would take about this half hour for the machines to heat up to the temperature for contin- uous operation and of constant speed.

Occasionally, during the day, the power would be shut down due to a short-circuit or overload on the circuit-breaker in the engine room so that the machines would stop and the lamps would not be go- ing for a few moments. But in nearly every case, this interval was short. Then the fluctuations, too, may be partially explained as due, in the early morn- ings, to the starting of the elevator motor in the Machinery Hall, and in the afternoons until 5:00 o'clock and in the evenings until 9:30 o'clock, to the throwing on and off of the machines and their respective loads in the dynaaio laboratory. The author also observed voltage fluctuations due to

■82-

t roubles with the machines. The motor brush holders were so designed that the slightest wear on one si side or the other of the brush, \7hich condition arose because of irregular pressure on them, produc- ed excessive sparking thus causing the supply cur- rwnt to vary and therefore producing fluctuating voltage on the lamps.

After 9:30 P.M., most of the load of the Institute was shut off so that all the power that was supplied by the generator in the engine room was for the varying lighting load of the Armour flats. Since there was no demand on the engineer for constant voltage for these lamps in the flats, he made no attempt to obtain leas thain about five or six per cent variation. Consequently, the volt- age on the test lamps varied considerably between 9:30 and 12:00 P.M. The voltage during this period however, did not exceed 110 volts but was always less, being usually between 105 and 110 volts at the test lamps.

•83-

Very seldom during the entire teat, did the voltage exceed 111 volts. On a few ocoaslona, the voltage did rise momentarily to 113 and 114 volts but the interval of such rise was very short. In fact, it cam be safely stated that the variation on the lamps from the tine of starting in the morning until 9:30 P.M. was between 111 and 107 volts and that after 9:30 P.M. has previously been stated.

It was not deemed necessary to include the voltage records for the entire test in this thesis but if there be any who desire to see them, they may be found accompanying the copy of this thesis which has been presented to the library of the Department of Electrical Engineering of the Institute.

Part III of this thesis contains the data which has been taken throughout the teat and Orlao the curves as plotted from it. As has already been stated, the lamps were photometered before starting, after fifty hours run, after one hundred hours run, and every succeeding hundred hours up to the limit of seven hundred. Consequently, for the sake of

■84-

clearnsaa and reading the data intellagantly, the author has tabulated the results of each series of lamps for each photometeric setting on a separate page. Beginning then with page 27, there is the phot- ometerio data at the start of the riin for the Gem metallized lamps ;on the next page (28) is the data for the ssune series after the fifty hours run; and on page 29, the data at the end of one hundred hours run. Each succeeding page shows the data for this series for every one hundred hours test. The photo- lueterio data for the Gem metallized lamps may there fore be observed as above outlined on pages 27-35 inclusive. Pages 36-40 inclusive, show the data for the tantalxim lamps arranged in the same manner as for the metallized. In a similar way, the succeeding pages up to and including page 67, show the data for the tungsten, Nernst, and carbon filament lamps as arranged in serial order.

On each sheet, there are six columns of fig- ures. The first one on the left indicates the seriee

■85-

A,B,C etc. to which that sst of lamps belong. The aecond colunm gives the number of the lamp for which the candle power has been determined. The third col- umn shows which kind of current in its test-alterna- ting or direct. The fourth and fifth columns show the candle power as given by the photometer settings and the sixth is the average of the values in the two previously mentioned columns. Then below this sixth column is another average reading. This is the ave- rage of the average candle power readings for all the lamps in that series at that particular length of rxjn. From these final readings, another set of curves was plotted as shown in Fig. 49, the data for these being collected in tabular form on page 78.

Beginning with page 67, there is anotlier set

of data. This is merely arranging the average candie

aet of power readings for each^lamps so that the variation

during life could be more readily observed than it

possibly could in the previous set of data. The first

column on the left in this case, gives the different

hours of life at which the lamps were photometered

■83-

and the aucoasding colmnna give the average can- dle power of the different lamps of that series as observed at these intervals for pho tome trying. It is from this sat of data that the curves, as shown in Pigs. 1-46 inclusive, were plotted.

On pages 75-6-7 there is the riinning-log for the test. These pages show the different days of the month that the lamps were going, the time of day that the machines were started and stopped, the number of hoiirs run on each of these days together with the total nixnber of hours up to and including each day's run.

Figs. 1-46 inclusive show to a great ex- tent, the results of this test. These curves show the life of every lamp and the candle power of each as photometered at the start of the test, after fi#- ty hours rvua, after one hiindred hours and every suc- ceeding hundred up to seven hundred hours at which time the greater per cent pf each series of lamps (excepting the carbon) had broken so that it was not considered essential or of marked interest to

■87-

continue longer. The author deamed it better to plot the curve for each lamp on a separate sheet instead of drawing all the curves for each series on one sheet. This gives a better opportunity to see at a glance the performeinoe of each lamp and does not make it as confusing as would otherwise be the case.

PigB. 1 -la inclusive show the life curves for each Isuap in series "A" -the Gem metallized type. The first six of these are for those lamps which were operated on the alternating current circuit and the remaining six for those operated on the direct current circuit. The cause of the four hundred hour life of # 1 was due to the fact that it was dropped when photoraetered. The fila- iiieut of # 4 broke at two htmdred and twenty-two hours and # 7 at two hundred and sixty-seven hours. All of the remaining lamps of this series lasted the full seven hundred hours, although the useful hours of life does not figure more than about six hundred and twenty-five, after which time the candle

-88-

power i3 Issa than 80^. It will also be noticed that there is practically no difference between the curves of those lampa placed on the alternating cur- rent circuit and those placed on the direct current circuit except that the latter show more character- iatic curves than the former.

The life curves for the tantalum lamps, series "B^jare given in Pigs. 13-S3 inclusive. The first six figures of this list show the curves for the lajups which were placed on the alternating cur- rent circuit. Of this n'omber, lamps # 5 and # 6 are the only ones which did not break before one hxm- dred hours of life. Lamps # 1 and # 2 broke at about eighty-five hours and were mended but did not last more than a few hours longer in either case. Lamp # 3 broke at sixty-eight hours, was mended and burn- ed thus lantil one hundred and eighty-two hours at which time the whole filament seemed to give way. Lamp # 4 broke first at one hiindred and five hours, was repaired so that it lasted lontil one hundred and seventy- three hours, at which time, it acted the

•89-

same as the previous two. Lamp # 5 lasted on© hundrad and fifty-eight hours before it broke for the first time and after mending, did not last but a few hours. Lamp # 6 was the longest lived of any in this aeries, it having burned four hizndred suid thirteen hours be- fore inipture. This lamp acted the same as the others after being repaired. Lamps # 7,8,9,10 and 11 were on the direct current circuit and all lasted at least one hundred and asventy-fivs hours before breaking, and two of them # 7 and # 8 lasted about two hund- red and sixty-five hours while # 10 lasted only one hundred and ninty-threa hours.

Prom the life as here shown, it can be deduc- ed that the tantalum lamps which were operated on direct current had a longer life than those operated on alternating current. The short life of all the lamps may be accounted for in two ways jflrst, because the tantalum filament is more fragile than any other filamentjand second, because of its greater fragility, it would tend to rupture quicker due to the vibration caused by the machines. This vibration was eliminated

.90.

as much aa poaaibls by padding the supports but it aeema there was enough transmitted through the floors and walls to cause a slight jar of the frame supporting the lamps. This was not sufficient to af- fect the other lamps but it does seem as though it shortened the life of the tantalum lamps.

The life curves for the tiingaten lamps are given in Figs. 24-54 inclusive. As with the two pre- vious types, the first six show curves for the lamps operated on the alternating current circuit and the remaining five for those on the direct current cir- cuit. Only two of the first six lamps mentioned last- ad the seven hundred hours, these being # 1 and # 6. The other four, however, lasted over six hundred hoiora, while #2,4 and 5 lasted about six h^undred and sev- enty-five hours. Of the five lamps on the direct current circuit, # 10 only lasted three hundred and twenty-five hours at which time the leading in wires broke, # 9 lasted four hundred hours and then broke due to a severe jar when photomet9red,and # 7 last- ed six hundred and five hours, the filament breaking

■91-

a few hours after the lamp was replaced in the frsme after having been pho tome tared. Lamps # 11 and # 12, however, lasted the full seven hundred hours. On the whole, then, a sianunary would indicate that the tungs- ten lamps lasted longer on the alternating current than on the direct current. The candle power of the lamps which lasted the seven hundred hours decreased by about 17,4^. It will be noted as a peculiarity of the tungsten leimps in this teat, that their candle power rose to a certain value after about one hun- dred hours and remained fairly constant at this value for approximately two hundred hours and then fell again as the other types of lamps. This fact would tend to show that the tungsten lamps had a greater efficiency for the first four hundred hours of life than any of the other types.

The curves for the Nernat lamps which were operated on alternating current are ahown in Figs. 35 36 and 37 while those on the direct current are given in Figs. 38-39 and 40. Lamp # 1 of this series, oper- ating on alternating current is the only one which

-92-

lasted the full seven hundred hours. The glower for this lamp burned out after fourteen and one half hours life and was replaced with another. Lamp # 6 operating on direct current lasted six hundred and twenty-five hours. The other two which were on the alternating current circuit lasted five hundred and eighteen and five hundred and ninty-aix hours respectively. And the remaining two operating on the direct current circuit lasted one hundred and seventy-five and one hundred and eighty-seven hours respectively. The test on these lamps would show, then, that the Nernst lamps operated longer on sixty cycle alternating current than on the direct current. In the case of this type of lamp, it was noticed that the glower became very crystalline in structure after about one hundred and fifty hours service. When the filament was in this condition, it was much more easy for the vibration to break the glower than when it was solid. Consequently, those glowers that became crystalline after such a short life were the ones that broke first. In three of these cases, the slower broke almost in the center

■93-

axid in the fourth case, the wire 'becaine detached from the glower itself.

It will be observed from the curves in Plga 41-47 inclusive, that the carbon lamps all lasted the full seven hundred hours. The efficiency of the three operating on alternating current seems to be a little higher than for the three on the direct current cir- cuit although the difference is quite small.

Summing up the teat, then, as a whole, the author concludes that it would be preferable to oper- ate the Gem metallized lamps on direct current cir- cuits; that the tantalum lajnpa operated best on direct current; that the tungsten lamps gave a great- er efficiency on alternating current than on direct current; that the Nernst gave a longer life on alter- nating current than on direct current; and that the carbon filament lamps gave better results when used on direct current circuits.

Respectfully submitted.

7

._ . f f i -I 4 Ml ill i i t i <^ , , . ■

- : ■•> V -'^ ?■ }- M §-M t ^^ L.

^ ; :■ « :; ?: ut € ^ m ^ t ;

. - :■ •: ? :# € i iTl. f 4 i f li ,k ^

M M if M M m ^ V :

r # « H 1 i i^€ f «^-i_.t,« t : f f :| «■ * € f MM il.t 1 € * ;^i^ t S I If lit f t « -i'.*.^ « -

.r I €' 4 W I # I ^ I € € € I .i 1 ,t ■« V^: .

<<<<•:<<«<•:<♦<<'>:'.-

■ f -I t # 1 f i i i € :i f :^.^ 4 ^

; i « € M t 'i f g M € € # * m. .^

. 1^ i « I 1 J .i i. J J ,t .€ ;i :i .t ^-^

i i I ^ ,€ ,.f t f .1 .€ ,^ ,« .€ t € ,« ,-■ V :. ^ I i 4 f ^i f f. t ^'i:^ f^ - <r^ «. « # ':' ■ ^

r# € f f f II 1 - I # f -^^

■.a t ;! t W « « i .: : ^ * ■« * :.,, ,

. .■ '^.*l«K*«\*«\«imm«--- i.i t.t « f f i ..

^^^ M -t -i

i^TMK0'm'^-i^\

^. ^.. m: #:

I't/f.

1' i' #

^ Vm^y** :\: _'*%^^»■**^•ft'^

■ " :' r r r i^^ §• §1*'

, », ti t, 9

-/r /» rt*.r ;r •

. ■i-.''i:. -■■'

. t *:

I- 9 '«■ ■

«^i*«*i!*«:c»:8-«

■, « f * \i

I s « « .«:i:

; ml ''ss' « '•^' *5 '^»i .'^ .*-'^- -■•

mmmmmm