,,! ,,fifte5 ,,watt ,,tung/5 ,,lamp ,,by ,,clair ,,elmore ,,&]son ,b4 ,s41 ,univ]sity ( ,ill9ois1 #aiaa ,,!sis ,submitt$ 9 ,p>tial ,fulfillm5t ( ! ,requirem5ts = ! ,degree ( ,,ma/] ,,( ,,sci5ce ,,9 ,,electrical ,,eng9e]+ ,,9 ,,! ,,graduate ,,s*ool ,,( ,,! ,,univ]sity ,,( ,,ill9ois #aiab ,illu/ration3 ,,univ]sity ,,( ,,ill9ois ,,! ,,graduate ,,s*ool ,may #ca1 #aiab ,i ,,h]eby ,,recomm5d ,,t ,,! ,,!sis ,,prep>$ ,,"u ,,my ,,sup]vision ,,by ,,clair ,,elmore ,,&]son ,,entitl$ ,,! ,,fifte5 ,,watt ,,tung/5 ,,lamp ,,be ,,accept$ ,,z ,,fulfill+ ,,?is ,,"p ,,( ,,! ,,requirem5ts ,,= ,,! ,,degree ,,( ,,ma/] ,,( ,,sci5ce ,,9 ,,electrical ,,eng9e]+ ,ern/ ,b]g ,9 ,*>ge ( ,major ,"w ,ern/ ,b]g ,head ( ,dep>tm5t ,recomm5dation concurr$ 93 ,morgan ,brooks ,ell]y ,b ,pa9e ,j ,m ,bryant ,committee on ,f9al ,exam9ation ,,cont5ts ,page ,i4 ,introduction1 #a ,,ii4 ,description ( ,lamps & ,te/s1 #b-d ,,iii4 ,*>act]i/ic ,curves1 #e-ab ,,iv4 ,sph]ical ,c&le ,p[]1 #ac-ae ,v4 ,ph5om5a ( 8,ov]%oot+01 #af-ah ,,vi4 ,!ories ( 8,ov]%oot+01 #ai-bj ,,vii4 ,am\nt ( 8,ov]%oot+01 #ba-bb ,,viii4 ,curves ( 8,ov]%oot+01 #bc-bf ,,ix4 ,conclusions1 #bg ---------------------------------------- -------------------------------- ,,! #ae ,,watt ,,tung/5 ,,lamp ,i4 ,,introduction4 ,s9ce ! introduction ( ! tung/5 lamp "s five ye>s ago1 ! manufactur]s h attempt$ cont9ually to produce small] & small] units 9 ! /&>d voltages4 ,! late/ lamp (f]$ today is ! #aae volt1 #ae watt1 tung/51 & x is ! purpose ( ?is pap] to %[ ! *>act]i/ics ( ?is lamp1 h[ x comp>es ) ! l>g] units z to life "u diff]5t conditions & its behavior 9 g5]al4 ,fir/ ( all1 x mu/ be borne 9 m9d t ^! te/s h be5 made ^u a comp>atively small numb] ( lamps1 & = t reason ! results %\ld n be tak5 z absolutely conclusive4 ,= ! life te/s1 at lea/ #ajj lamps %\ld h be5 us$ "u ea* condition1 b ?is 0 impossible because ( ! exp5se4 ,special att5tion has be5 giv5 to ! ph5om5on ( 8ov]%oot+04 ,an entire ye> c\ld h easily be5 sp5t inve/igat+ ?is subject1 & ! writ] regrets t lack ( "t has prev5t$ m elaborate & compreh5sive te/s ( ?is /range ph5om5on4 ,,ii4 ,,description ,,( ,,lamps ,,& ,,te/s4 ,! total numb] ( #ae watt lamps te/$ 0 #bd1 "o half ( :i* 0 obta9$ directly f ! manufactur] & ! o!r half b\ght 9 op5 m>ket4 ,x is well to m5tion at ?is "t t ?is may h be5 ! cause ( ! diff]5t qualities z br\ght out by ! life te/s4 ,! lamps 7 rat$ at #a4ca watts p] horizontal c&le p[] & 7 suppos$ to h a useful life ( #ajjj h\rs4 ,! voltage rat+s ( ^? obta9$ f ! factory 7 #aad - #aab - #aaj & ^? b\ght 9 op5 m>ket 7 #aae - #aac - #aaa4 ,! correct effici5cy ( ! lamps z f\nd by te/ 0 #a4cd watts p] c&le p[]4 ,a %ot diagram foll[s :i* %[s ! actual rat+ ( ! lamps at high effici5cy4 ,all read+s 7 made by a ,lumm]-,brodhun photomet] & ! voltmet]s & ammet]s us$ 7 c>efully /&>diz$4 ,! ammet] 0 plac$ beyond ! voltmet] 9 ord] to get ! true curr5t tak5 by ! lamp4 ,! drop across ! ammet] 0 tak5 into acc\nt 9 ! voltmet] read+s4 ,life te/s 7 made "u two conditions1 namely1 a %ock te/ :]e ! lamps receiv$ sev]e vibrations & a te/ "u ideal conditions1 i4e4 no j> & con/ant voltage4 ,9 ord] to obta9 vibrations = ! lamps ^u ! %ock te/1 a small motor1 ) its %aft pulley (f set1 0 screw$ rigidly to a table4 ,! lamps 7 plac$ 9 a normal position ^u ! table by means ( wood5 frames4 ,! result 0 t :5 ! motor 0 runn+ x _h a p\nd+ effect1 ?us putt+ ! table1 consequ5tly ! lamps1 9 a /ate ( sev]e vibration4 ,! filam5ts ( ! lamps c\ld be se5 viol5tly %ak+ = "s di/ance4 ,! te/ 0 inde$ a h>d "o1 & "o t w\ld n be f\nd 9 _m actual cases4 ,x is v d\btful if railway lamps are subject$ to su* a /ra9 & !y are ( ! heavy filam5t l[ voltage type4 ,! foll[+ photograph %[s ! arrangem5t above describ$4 ,t5 #ae watt lamps 7 us$ on ?is te/1 ! rema9d] %[n be+ #bj & #be watt & c>bons4 ,illu/ration3 photograph1 arrangem5t ( tung/5 lamps ,illu/ration3 ,%ot diagram = #ae watt lamps ,,iii4 ,,*>act]i/ic ,,curves4 ,figure #a1 ,page #h1 %[s ! v>iations ( ! c&le p[] ) ! voltage1 curr5t & watts4 ,figure ,,ii %[s ! relation betwe5 c&le p[] & ! effici5cy1 watts p] horizontal c&le p[]1 & also ! v>iation ( ! c&le p[] ) ! resi/ance4 ,an empirical =mula = ! c&le p[] express$ z a function ( ! watts is cp ,,kwx :]e ,k is a con/ant ( ! lamp & ,w d5otes ! watts4 ,f ! curve :5 cp #e1 watts #aa4a & :5 cp #ae1 watts #ag4e divid+ cp#a_/;cpa ,,kw#ax_/,,kwax sub/itut+ #e_/ae #aa4ax_/#ag4ex & log #c x log #aa4a x log #ag4e 4#dgga #a4jdecx #a4bdcjx 4#aihx 4#dgga x #b4da solv+ = ! con/ant ,k #e ,k #aa4ab4da #e #ccb ,k ,k 4#jaej & ! f9al equation = ! c&le p[] is cp 4#jaej w#b4da ,9 ! same way1 ! c&le p[] may be express$ 9 t]ms ( ! voltage & ?is is f\nd to be cp #ccd #aj-i ,e#c4fh ,?is =mula *ecks precisely ) ! "o us$ 9 ! eng9e]+ dep>tm5t ( ! ,g5]al ,electric ,company at _! lamp works1 ,h>rison1 ,n4,j4 ,! horizontal di/ribution curve ( a lamp ) its filam5t m\nt$ z is ! mod]n tung/5 is ne>ly a circle4 ,?is is n true1 h[ev]1 9 ! case ( v]tical di/ribution & ?is curve is %[n1 ,figure ,,iii4 ,z w be not$1 ! tip c&le p[] is only ab\t #bc p] c5t ( ! horizontal4 ,life ,te/s4 ,! results ( ! life te/s 7 v surpris+4 ,! lamps ^u ! te/ "u ideal conditions1 namely1 no vibrations & con/ant voltage1 _h only an av]age life ( #dfj h\rs1 :ile e "o ( ^? ^u ! %ock te/ are /ill burn+ at ! pres5t "t1 hav+ be5 burn$ #cjj h\rs4 ,9 ord] to make ! te/ /ill m sev]e1 ! lamps 7 subject$ to vibrations )\t voltage be+ impress$1 & z yet1 n a filam5t has brok51 ! total "t be+ #djj h\rs4 ,x 0 impossible to give m "t to ^! lamps z 0 done = ^? "u ideal conditions1 = ! reason x 0 ?\ght unadvisable to leave ! motor1 :i* gave ! vibrations1 runn+ ov] night4 ,! curves h ! same g5]al =m = ! two conditions b ! v>iations are f> m great = ! lamps :i* 7 ^u ! %ock te/4 ,! reason = ?is is t ! vibrations 7 s sev]e z to %ake p>ts ( ! filam5t toge!r ?us giv+ a p>tial %ort circuit1 caus+ great v>iations 9 c&le p[]4 ,illu/ration3 ,fig ,i ,*>act]i/ic ,curves = #ae watt tung/5 lamps ,lamp ( av]age rat+ us$ ,illu/ration3 ,fig ,,ii ,*>act]i/ic ,curves = #ae watt tung/5 lamps ,lamp ( av]age rat+ us$ ,illu/ration3 ,fig ,,iii ,v]tical ,di/ribution = #ae watt #aae volt ,tung/5 lamp ,illu/ration3 ,life ,te/s #ae ,watt #aae ,volts ,tung/5s ,conditions ,ideal o--,burn$ ,out ,illu/ration3 ,life ,te/s #ae watt #aad volts tung/5 ,%ock ,conditions ,all lamps /ill ,burn+ ,,iv4 ,,sph]ical ,,c&le ,,p[]4 ,ow+ to ! abs5ce ( an integrat+ photomet]1 ! mean sph]ical c&le p[] has be5 f\nd by ,k5nelley's graphical me?od4 ,?is me?od is v simple z comp>$ ) ,r\sseau's & has ! advantage ( yield+ ! mean sph]ical int5sity z a "o dim5sional quantity4 ,?is disp5ses ) ! use ( a planimet] or equival5t measur+ surface device4 ,x consi/s ess5tially 9 det]m9+ graphically f ! giv5 pol> curve an evolute & ! involute ( ! same & !n project+ ?is involute ^u a v]tical l9e4 ,half ! l5g? ( ! projection is equal to ! mean sph]ical int5sity to ! same scale z ! orig9al pol> curve4 ,figure ,,iv %[s ! me?od & is expla9$ ?us3 ,! pol> curve ,o ,a ,h ,b corresponds to ! di/ribution ( int5sity f an inv]t$ inc&esc5t lamp hav+ its base at ,v & tip at ,v4 ,?is is n precisely true b ! only v>iation to speak ( is t ! tip c&le p[] does n fall (f s mu* z ! curve betwe5 ,b & ,o4 ,?is v>iation is slight1 h[ev]4 ,! mean horizontal int5sity is ,,oh1 ! diamet] ( ! circle1 & 9 ?is case is equal to #ab4d c&le p[]4 ,9 ! diagram ! con/ruction is adapt$ to zones ( #cj & ! radii ( ! midzones f\nd i4e4 at #ge1 #de1 #ae1 -#ae1 -#de -#ge4 ,^! are m>k$ by dott$ l9es ,o t1 ,o s1 ,o r1 ,o r1 ,o s1 ,o t respectively4 ,) radius ,o r & c5t] ,o1 ! arc hra is describ$ ?r\gh an angle ( #cj4 ,! radius ,oa is drawn at ! end ( ! arc4 ,a di/ance ,ab is measur$ f a along a ,o equal to ,o s1 ! second midzone radius4 ,) a c5t] b & radius ,o s1 ! arc ac is describ$ ?r\gh an angle ( #cj s t bc makes an angle ( #fj ) ! horizontal ,,oh4 ,! l9e bc is drawn at ! end ( ?is arc4 ,f c t[>ds b1 a di/ance cd is m>k$ (f equal to ,ot1 ! ?ird midzone radius4 ,) c5t] d & radius ,o t1 ! arc ce is describ$ ?r\gh an angle ( #cj s t de makes an angle ( #ij ) ! horizontal ,,oh4 ,! l9e de is drawn4 ,! arc hace is ext5d$ f ! horizontal to ! v]tical b5ea? 9 ! same mann] z above by /eps ( #cj ) c5t]s ,o1 b1 & d1 & radii ,or1 ,,os & ,ot respectively4 ,! curve ec>race is n[ cont9u\s & complete4 ,a v]tical l9e ,,qq is drawn ?r\gh ! conv5i5t po9t ,h & ! po9ts e c a a c e are project$ ^u ! same4 ,! l5g? ,,hq1 is ! upp] hemisph]ical int5sity & ! l5g? ,,hq ! l[] hemisph]ical int5sity4 ,_! ari?metical mean is ! mean sph]ical int5sity4 ,s9ce 9 ?is case ! upp] & l[] hemisph]es are symmetrical ,,hq ,,hq ,,qq_/#b mean sph]ical int5sity4 ,by measurem5t ?is half l5g? is f\nd to be #c4abe in*es & f ! scale us$ ?is corresponds to #i4fg c&le p[]4 ,! sph]ical r$uction factor = ^! lamps is1 !n #i4fg_/ab4d #gh ,illu/ration3 ,fig ,,iv ,k5nelly's ,diagram = ,sph]ical ,c4 ,p4 ,v4 ,,ph5om5a ,,( 8,,ov]%oot+04 ,! s+ul> prop]ty ( ! tung/5 lamp to 8ov]%oot0 or to give tempor>ily a high] initial ?an normal c&le p[]1 0 fir/ discov]$ by ,john ,b4 ,taylor & is expla9$ 9 ! foll[+ mann]3--,! filam5t ( ! c>bon inc&esc5t lamp possesses a negative temp]ature coeffici5t2 t is to say1 a rise 9 voltage causes a m ?an correspond+ rise 9 curr5t & :5 ! lamp is connect$ to a s\rce ( con/ant pot5tial1 ! curr5t />ts at a comp>atively small value & increases to a maximum :5 ! lamp has atta9$ full c&le p[]4 ,9 ! case ( ! tung/5 lamp1 ! situation is j ! rev]se1 s9ce tung/5 has a positive temp]ature coeffici5t4 ,:5 ! lamp is connect$ to a con/ant pot5tial supply ! curr5t is a maximum :5 ! lamp is cold & decreases to a f9al value :5 ! lamp rea*es full brilliancy4 ,! mo/ important diff]5ce betwe5 ! two lamps due to ^! diff]5t *>act]i/ics is t :ile a tung/5 lamp rea*es full c&le p[] ! in/ant ! curr5t is turn$ on a c>bon inc&esc5t lamp comes up to full c&le p[] only aft] a p]ceptible p]iod ( "t4 ,! app>5t tempor>y increase 9 ! c&le p[] ( a tung/5 lamp 0 obs]v$ e>ly aft] ! lamp 0 inv5t$ b x 0 g5]ally ascrib$ to "s possible physiological action due to ! sl[ contraction ( ! pupil ( ! eye4 ,! foll[+ curve 0 obta9$ by means ( ! oscillograph & %[s cle>ly ! ru% ( curr5t = ! fir/ in/ant aft] ! lamp is turn$ on4 ,! break 9 ! curve is due to an imp]fection 9 "o ( ! op]at+ swit*es & has no?+ to d ) any *>act]i/ic ( ! lamp4 ,! cycle wave 0 put on m]ely to obta9 ! "t4 ,illu/ration3 photograph1 curr5t aft] lamp is turn$ on ,9 ord] to prove t ?is ov]%oot+ occurs1 an actual photograph ( ! int5sity has be5 made4 ,?is 0 obta9$ by mak+ a box #a #a #c absolutely light pro( & arrang+ a lamp inside s t x c\ld be turn$ on & (f at w4 ,a slit1 fitt$ ) a %utt]1 0 cut 9 "o end ( ! box :i* p]mitt$ ! light to fall ^u a revolv+ oscillograph film4 ,! film hold] 0 atta*$ to ! box ) ?9 metal /rips & revolv$ by means ( a small motor4 ,z is se5 ! :ole arrangem5t 0 no?+ m ?an a l>ge cam]a4 ,! foll[+ photograph %[s ! ph5om5on q cle>ly1 po9t ,a d5ot+ :]e ! lamp 0 turn$ on4 ,illu/ration3 photograph1 int5sity ( light1 ov]%oot+ ,,vi4 ,,!ories ,,( 8,,ov]%oot+40 ,! !ory giv5 by ,taylor to acc\nt = ! 8ov]%oot+0 ( tung/5 lamps is bas$ on ! fact t !re is a small am\nt ( residual gas 9 ! lamp1 :i* is attract$ to ! walls ( ! lamp :5 x is cold2 & :5 ! lamp is light$ & w>ms up1 ?is residual gas is driv5 (f l[]+ ! vacuum4 ,) a high vacuum1 practically all ! en]gy mu/ be radiat$ f ! filam5t2 conv]sely1 on a l[]+ ( ! vacuum1 "s ( ! heat is c>ri$ away by convection & conduction4 ,:5 all ! heat is c>ri$ away by radiation ! filam5t runs at a high] temp]ature & w give m light4 ,ano!r !ory is t a cold tung/5 filam5t lamp absorbs & occludes c]ta9 gase\s sub/ances f ! l[ pressure space )9 ! *amb]4 ,ow+ to ! pres5ce ( ^! gases ! filam5t %9es m brightly :5 fir/ br\ght quickly to inc&esc5ce1 b aft] ! gases h be5 driv5 (f by ! heat1 ! extra lum9esc5ce disappe>s & c be rega9$ only by prolong$ cool+ & re/4 ,/ill ano!r !ory1 & ! "o t seems ! mo/ logical to ! writ]1 is t ! increase ( resi/ance accompany+ ! rise ( temp]ature takes a c]ta9 small int]val ( "t s t :5 ! temp]ature is ris+ at ! rate ( ?\s&s ( degrees p] second1 ! resi/ance lags p]ceptibly4 ,! resi/ance does n suppress ! curr5t z quickly z x %\ld & an extra ru% ( curr5t & heat en]gy goes ?r\gh ! filam5t1 rais+ ! temp]ature above normal1 ) a correspond+ increase 9 brilliancy4 ,,vii4 ,,am\nt ,,( 8,,ov]%oot+04 ,9 ord] to det]m9e ! am\nt ( ov] %oot+1 ! writ] has made photographs z %[n bel[4 ,numb] "o 0 made ) ! lamp "u voltage1 numb] two by us+ normal voltage & sudd5ly turn+ ! lamp on by means ( a snap swit*1 !reby obta9+ ! ov]%oot+1 & f9ally numb] ?ree 0 made by impress+ voltage above normal4 ,! pictures 7 obta9$ by us+ ! photographic arrangem5t z be=e describ$4 ,all ?ree pr9ts 7 made f ! same film1 t is ! ?ree pictures 7 made ^u "o film ?us insur+ ! same developm5t & pr9t+ = all4 ,! pr9t has be5 cut m]ely to all[ a clos] comp>ison ( ! int5sities4 ,illu/ration3 am\nt ( ov] %oot+ ,x is se5 t numb] ?ree comp>es favorably ) ! ov]%oot+ z %[n by numb] two & ! c&le p[] correspond+ 0 f\nd to be approximately #ej great] ?an normal4 ,x is n claim$ t e lamp w ov]%oot ?is am\nt z ! degree ( vacuum or o!r factors ( individual lamps may play an important "p 9 ?is ph5om5on4 ,!re is no d\bt1 h[ev]1 t ?is /range fact really occurs & is n due to physiological reasons4 ,,viii4 ,,curves ,,( 8,,ov]%oot+04 ,9 ord] to prove t ! law ( resi/ances1 namely1 ,r ,ro7#a t7 does n hold = ! fir/ in/ance aft] clos+ ! swit* on a tung/5 lamp1 ! foll[+ curves h be5 plott$4 ,numb] ,v4 has be5 tak5 f ! oscillograph record %[n 9 ! fir/ "p ( ?is pap] & %[s t ab\t 4#jbd second elapses be=e ! curr5t becomes normal4 ,kn[+ ! curr5t at any in/ant z giv5 by ?is curve1 x is easy to f9d ! resi/ance at ! same in/ant by ,ohm's law1 ! electromotive =ce be+ a con/ant & kn[n value4 ,curve ,,vi %[s ?is relation4 ,curves ,,vii & ,,viii are approximate values & n absolute4 ,n[ f ! temp]ature curve1 values are tak5 & sub/itut$ 9 ! =mula = resi/ance1 ,r ,ro7#a t71 ! result+ curve be+ ,figure ,,ix4 ,x is se5 t curves ,,vi & ,,ix d n take ! same values at all until aft] a brief int]val ( "t has elaps$4 ,curve ,,vi is absolutely correct1 h[ev]1 z ^! values h be5 obta9$ f ! oscillograph record4 ,consequ5tly1 ! assumption ^u :i* curve ,,ix is bas$ mu/ be incorrect = ! fir/ 4#jbd? ( a second & ! conclusion is t ! law ( resi/ances does n hold4 ,?is result t5ds to /r5g!n ! !ory ( ! lag ( resi/ivity = ! 8ov]%oot+0 ( a tung/5 lamp4 ,illu/ration3 ,fig ,v ,illu/ration3 ,fig ,,vi ,illu/ration3 ,fig ,,vii ,illu/ration3 ,fig ,,viii ,illu/ration3 ,fig ,,ix ,r ,ro#a t ,,ix4 ,,conclusions4 ,! foll[+ conclusions may be drawn f ! results ( ! te/s4 #a/1 ,t ! quality ( ! two sets ( lamps 0 greatly diff]5t4 #b;d1 ,t x is d\btful if ! #ae watt1 #aae volt1 tung/5 lamp z fir/ put ^u ! m>ket met ! gu>antee z to life4 ,?is conclusion is rea*$ by te/s 9 ! laboratory & exp]i5ce ) lamps in/all$ 9 resid5ces4 #c;d1 ,t ! tung/5 lamp is subject to ov]%oot+4 #d?1 ,t dur+ ?is p]iod ! initial c&le p[] may be z mu* z #ej above ! normal4 #e?1 ,t ! mo/ probable !ory ( ov]%oot+ is ! lag ( resi/ance beh9d ! temp]ature4 ---------------------------------------- -------------------------------- ,,transcrib]',s ,,note ,italiciz$ phrases are pres5t$ by surr\nd+ ! text ) und]scores4 ,sup]script$ *>act]s are repres5t$ by prec$+ ) a c>et 77 symbol4 ,:5 long] ?an "o digit1 !y are also surr\nd$ 9 curl$ brackets4 ,subscript$ *>act]s are repres5t$ by prec$+ ) an und]score & enclos+ 9 curl$ brackets4 ,fractions are repres5t$ horizontally us+ ! sla% _/ symbol4 ,! orig9al sep>at$ ! num]ator & d5om9ator horizontally1 sep>at$ by a l9e4 ,punctuation has be5 normaliz$4 ,except 9 ! foll[+ cases1 v>iations 9 hyph5ation1 spell+ & spac+ h be5 reta9$ z !y 7 9 ! orig9al publication4 ,*apt] #b3 ,lum]-,brodhum photomet] -- ,lumm]-,brodhun photomet] ,*apt] #d3 ,:at appe> to be h&-drawn primes 9 ! text h be5 confirm$ z3 ,! arc hace -- ,! arc hace4 e c a a c e -- e c a a c e ,figure ,,iv ,!re are two po9ts b radiat+ f po9t ,o4 ,! l[] is likely to repres5t b4 ,*apt] #e3 inportant diff]5ce betwe5 -- important diff]5ce betwe5