,,curiosities ,,( ,,light ,,& ,,sight4 ,,curiosities ,,( ,,light ,,& ,,sight ,,by ,,%el=d ,,bidwell1 ,m4,a41 ,,ll4,b41 ,f4,r4,s4 ,,) ,,fifty ,,illu/rations ,,london3 ,,swan ,,sonn5s*e9 ,,co41 ,,limit$ ,,pat]no/] ,,squ>e #ahii ,,preface4 ,! foll[+ *apt]s are bas$ ^u notes ( sev]al unconnect$ lectures address$ to audi5ces ( v diff]5t classes 9 ! !atres ( ! ,royal ,in/itution1 ! ,london ,in/itution1 ! ,le$s ,philosophical & ,lit]>y ,society1 & ,caius ,h\se1 ,batt]sea4 ,9 prep>+ ! notes = publication ! matt] has be5 re-arrang$ ) ! object ( pres5t+ x1 z f> z might be1 9 me?odical ord]2 additions & omissions h be5 freely made1 & num]\s diagrams1 illu/rative ( ! app>atus & exp]im5ts describ$1 h be5 provid$4 ,i d n "k t any apology is ne$$ = (f]+ ! collection z ?us re-modell$ to a l>g] public4 ,?\gh ! essays are1 = ! mo/ "p1 ( a popul> & in=mal "*1 !y t\* ^u a numb] ( curi\s matt]s ( :i* no readily accessible acc\nt has yet appe>$1 :ile1 ev5 9 ! mo/ elem5t>y p>ts1 an attempt has be5 made to h&le ! subject ) "s degree ( fre%ness4 ,! int]e/+ subjective ph5om5a :i* are associat$ ) ! s5se ( vision d n appe> to h receiv$ 9 ?is c\ntry ! att5tion !y des]ve4 ,?is little book may p]haps be ( "s slight s]vice 9 sugge/+ to exp]im5tali/s1 bo? pr(essional & amateur1 an attractive field ( rese>* :i* has hi!rto be5 only p>tially explor$4 ,,cont5ts4 ,,page4 ,,*apt] ,i4 ,light & ! ,eye #a ,,*apt] ,,ii4 ,col\r & its ,p]ception #ci ,,*apt] ,,iii4 ,"s ,optical ,defects ( ! ,eye #hd ,,*apt] ,,iv4 ,"s ,optical ,illusions #acj ,,*apt] ,v4 ,curiosities ( ,vision #afe ,,li/ ,,( ,,diagrams4 ,,fig4 ,,page4 #a4 ,image ( ,slit & ,spectrum #ab #b4 ,diagram ( ! ,eye #bd #c4 ,abney's ,col\r-pat* ,app>atus #de #d4 ,p>tially ,int]cept$ ,spectrum #di #e4 ,/5cil ,c>ds #eb #f4 ,helmholtz's ,curves ( ,col\r ,s5sations #gb #g4 ,knig's ,curves #gc #h4 ,/5cil ,c>d = ,complem5t>y ,col\rs #gg #i4 ,ano!r =m #gi #aj4 ,slide = ,mix+ any two ,spectral ,col\rs #hj #aa4 ,refraction ( ,mono*romatic ,light by ,l5s #hg #ab4 ,refraction ( ,di*romatic ,light #hi #ac4 ,n>r[ ,spectrum z se5 f a ,di/ance #ig #ad4 ,spectrum =m$ ) ,v-%ap$ ,slit #ajc #ae4 ,bezold's ,device = ,demon/rat+ ,non-a*romatism ( ! ,eye #ajh #af4 ,cross$ ,l9es %[+ ! ,effect ( ,a/igmatism #aac #ag4 ,ano!r ,design %[+ ! same #aad #ah4 ,/>-l ,images ( ,lum9\s ,po9ts #aaf #ai4 ,sutures ( ! ,cry/all9e ,l5s #aag #bj4 ,multiple ,images ( a ,lum9\s ,po9t #abj #ba4 ,! same1 %[+ an increas$ numb] ( ,images #abb #bb4 ,! same :5 a ,slit is held be=e ! ,eye #abc #bc4 ,multiple ,images ( an ,electric ,lamp ,filam5t #abe #bd4 ,! same se5 ?r\gh a ,slit #abf-abh #be4 ,illusion ( ,l5g? #acb #bf4 ,ano!r =m #ace #bg4 ,ano!r =m #acf #bh4 ,ano!r =m #acg #bi4 ,ano!r =m #ach #cj4 ,illusion ( ,incl9ation #adc #ca4 ,zlln]'s ,l9es #add #cb4 ,slide = %[+ ,illusions ( ,motions #adg #cc4 ,illusion ( ,motion #adi #cd4 ,illusion ( ,lum9osity #aeb #ce4 ,illusion ( ,col\r #aee #cf4 ,recurr5t ,vision demon/rat$ ) a ,vacuum ,tube #agf #cg4 ,! same ) a ,rotat+ ,disk #agh #ch4 ,app>atus = %[+ ,recurr5t ,vision ) ,spectral ,col\rs #aha #ci4 ,*>p5ti]'s ,d>k ,b& #ahg #dj4 ,*>p5ti]'s ,effect %[n ) ! ,h& #ahi #da4 ,multiple ,d>k ,b&s #aib #db4 ,tempor>y ,ins5sitiv5ess ( ! ,eye aft] ,illum9ation #aid #dc4 ,visual ,s5sations att5d+ a ,p]iod ( ,illum9ation #aii #dd4 ,b5ham's ,artificial ,spectrum ,top #bjj #de4 ,demon/ration ( ,r$ ,col\r-bord]s #bje #df4 ,black & ,:ite ,scre5s = ! same #bji #dg4 ,rotat+ ,disk = ! same #baj #dh4 ,demon/ration ( ,blue ,col\r-bord]s #bae #di4 ,disk = ,exp]im5ts on ! ,orig9 ( ! ,col\r-bord]s #bag #ej4 ,disk = ! ,subjective ,trans=mation ( ,col\rs #bbd ,,*apt] ,i4 ,,light ,,& ,,! ,,eye4 ,9 ! pres5t sci5tific age e "o kn[s t light is transmitt$ across space ?r\gh ! m$ium ( ! lum9if]\s e!r4 ,?is e!r fills ! :ole ( ! kn[n univ]se1 z f> at lea/ z ! remote/ /> visible 9 ! mo/ p[]ful telescopes1 & is (t5 said to be possess$ ( prop]ties ( s p>adoxical a "* t _! unres]v$ acceptance has always be5 a matt] ( consid]able difficulty4 ,! e!r is a ?+ ( immeasurable t5uity1 be+ _m millions ( times r>] ?an ! mo/ p]fect vacuum ( :i* we h any exp]i5ce3 x (f]s no s5sible ob/ruction to ! movem5ts ( ! cele/ial bodies1 & ev5 ! flimsie/ ( mat]ial sub/ances c pass ?r\gh x z if x 7 no?+4 ,yet we h be5 taught t ?is same e!r is an ela/ic solid ) a great degree ( rigidity1 its resi/ance to di/ortion be+1 9 comp>ison ) ! d5sity1 ne>ly t5 ?\s& million times great] ?an t ( /eel3 ?us 0 expla9$ ! prodigi\s spe$ ) :i* x propagates transv]se vibrations4 ,a few ye>s ago1 a di/+ui%$ lead] 9 sci5ce endeav\r$ 9 ! c\rse ( a lecture to illu/rate ^! app>5tly incompatible prop]ties ) ! aid ( a l>ge slab ( ,burgundy pit*4 ,he %[$ t ! pit* 0 h>d & brittle1 yet1 z he said1 a bullet laid ^u ! slab w\ld1 9 ! c\rse ( a few mon?s1 s9k into & p5etrate ?r\gh x1 ! h>d brittle mass be+ really a v visc\s fluid4 ,! e!r1 x 0 sugge/$1 resembl$ ! pit* 9 hav+ ! rigidity ( a solid & yet gradually yield+2 x 01 9 fact1 a rigid solid = lum9if]\s vibrations execut$ 9 ab\t a hundr$-billion? "p ( a second1 & at ! same "t highly mobile to bodies l ! e>? go+ ?r\gh x at ! rate ( tw5ty miles 9 a second4 ,?is illu/ration1 felicit\s z x is1 w\ld1 h[ev]1 sc>cely avail to =ce conviction ^u an unwill+ m9d1 ev5 if x 7 admitt$ t ! p]iod ( an e!r wave is necess>ily no m ?an a hundr$-billion? ( a second or !reab\ts1 :i* is probably v f> f ! tru?4 ,b1 inde$1 ! ela/ic solid !ory ( ! e!r has fail$ to give a consi/5t explanation ( "s ( ! mo/ important po9ts 9 obs]vational optics2 &1 9 spite ( ! exalt$ position :i* x has held1 x c n[ h>dly be reg>d$ z repres5t+ a physical reality4 ,! fam\s rese>*es ( ,h]tz h e/abli%$ ^u a secure exp]im5tal basis ! hypo!sis ( ,maxwell t light is an electro-magnetic ph5om5on4 ,su* electrical radiations z c be produc$ by suitable in/rum5ts are f\nd to behave 9 exactly ! same mann] z ^? to :i* light is due4 ,!y travel ?r\gh space ) ! same spe$2 !y c be reflect$1 refract$1 pol>is$1 & made to exhibit int]f]5ce effects4 ,no fact 9 physics c be mu* m firmly e/abli%$ ?an t ( ! ess5tial id5tity ( light & electricity4 ,x foll[s !n t ! displacem5ts ( ! e!r :i* con/itute light-waves are n necess>ily ( ! same gross me*anical nature z ^? :i* we see on ! surface ( wat]1 or :i* occur 9 ! air :5 s\nd is transmitt$ ?r\gh x4 ,! displacem5ts :i* ! e!r und]goes are n me*anical--prim>ily at all ev5ts--b electrical4 ,e "o kn[s :at a simple me*anical displacem5t is4 ,if we pu% aside ! bob ( a susp5d$ p5dulum1 t is a me*anical displacem5t4 ,b if we electrify a /ick ( seal+ wax by rubb+ x ) flannel1 ! surr\nd+ e!r und]goes electric displacem5t1 & no "o und]/&s :at electric displacem5t really is4 ,ultimately1 no d\bt1 x w turn out to be ( a me*anical nature1 b x is almo/ c]ta9ly n a simple bodily di/ortion su* z is caus$1 = example1 :5 "o presses a jelly ) ! f+]4 ,s9ce1 !n1 x is no long] necess>y to assume t ! exce$+ly r>e & subtile e!r is a jelly-l solid 9 ord] to acc\nt = ! mann] 9 :i* x transmits light1 "o ( ! mo/ s]i\s difficulties 9 ! way ( its acceptance is remov$4 ,x is true t no?+ is def9itely kn[n conc]n+ ! me*anism :i* takes ! place ( ! simple transv]se vibrations =m]ly po/ulat$1 b e "o w admit t x is f> easi] to believe 9 :at we "k no?+ ab\t ?an 9 :at we "k to be impossible4 ,all sci5tific m5 are 9 fact agre$ 9 recognis+ ! real & g5u9e exi/5ce ?r\gh\t space ( an e!r capable1 among o!r ?+s1 ( transmitt+ at ! spe$ ( #ahf1jjj miles p] second di/urbances :i*1 :atev] _! precise nature1 are ( ! k9d :i* ma!maticians are accu/om$ to call waves4 ,h[ an e!r wave is con/itut$ w probably be kn[n :5 we h f\nd out exactly :at electricity is3 & t may be nev]4 ,! s5sation ( light results f ! action ( e!r waves ^u ! organism ( ! eye1 b ! old belief t ! s5sation 0 prim>ily due to a s]ies ( m]e me*anical impulses or beats1 j z t ( s\nd results f ! me*anical impact ( air-waves ^u ! drum ( ! e>1 _c any long] be upheld4 ,! ess5tial nature ( ! action ex]t$ by e!r waves is /ill undet]m9$1 ?\gh _m guesses at ! tru? h be5 haz>d$4 ,x may be electrical or x may be *emical2 possibly x is bo?4 ,e!r-waves1 we "k1 are compet5t to br+ ab\t *emical *anges1 z 9 ! famili> in/ance ( ! photographic processes2 !y c also produce electric ph5om5a1 z1 = example1 :5 !y fall ^u a suitably prep>$ piece ( sel5ium2 b !re is no evid5ce t !y c ex]t any direct me*anical action ( a vibratory "*1 & inde$ x is b>ely conceivable t any portion ( our organism %\ld be adapt$ to take up vibrations ( su* enorm\s rapidity z ^? :i* *>act]ise light-waves4 ,( ! multitude ( e!r-waves :i* trav]se space x is only comp>atively few t h ! p[] ( excit+ ! s5sation ( light4 ,z reg>ds limit$ range ( s5sibility !re is a v close analogy betwe5 he>+ & see+4 ,no s5sation ( s\nd 7at lea/ ( cont9u\s s\nd7 is produc$ :5 air-waves beat ^u our e>s unless ! rate ( ! successive impulses lies )9 c]ta9 def9ite limits4 ,x is j s ) vision4 ,if e!r-waves fall ^u our eyes at a less rate ?an ab\t #djj billions p] second1 or at a great] rate ?an #gej billions p] second1 no s5sation ( light is p]ceiv$4 ,!re is ano!r & m g5]ally conv5i5t way ( /at+ ?is fact4 ,s9ce all waves f\nd 9 ! e!r travel ?r\gh space at exactly ! same spe$--#ahf1jjj miles a second--x foll[s t ! l5g?#a ( ea* ( a s]ies ( homog5e\s waves mu/ be inv]sely proportional to _! frequ5cy1 t is1 to ! rate at :i* !y /rike a fix$ object1 su* z ! eye4 ,in/ead1 !re=e1 ( specify+ waves by _! frequ5cy we may equally well specify !m by _! l5g?4 ,waves ^: frequ5cy is #djj billions p] second h a l5g? ( ab\t #a_/cdjjj in*1 ?is be+ ! "o f\r hundr$ billion? "p ( #ahf1jjj miles2 & ^? ^: frequ5cy is #gej billions h a wave-l5g? ( #a_/fdjjj in*4 ,waves1 !n1 ( a l5g? great] ?an #a_/cdjjj in* or less ?an #a_/fdjjj in* h no effect ^u our organs ( vision4#b ,9 relation to ?is important fact x w be conv5i5t to ref] to a famili> b v beautiful exp]im5t--! =mation ( a spectrum4 ,an electric lamp is enclos$ 9 an iron lant]n1 hav+ 9 its front an upright slit2 f ?is slit !re issues a n>r[ beam ( :ite light1 :i* is made up ( rays ( _m diff]5t wave-l5g?s1 all mix$ up toge!r4 ,by caus+ ! light to pass ?r\gh a prism ! mix$ rays are sort$ out side by side accord+ to _! sev]al wave-l5g?s1 =m+ a broad1 _m-hu$ b& or spectrum ^u a :ite scre5 plac$ to receive x4 7,see ,fig4 #a47 ,to ! visible rays ( ! longe/ wave-l5g? is due ! r$ col\r on ! extreme left4 ,waves ( some:at %ort] l5g? produce ! adjo9+ /ripe ( orange1 & ! succe$+ col\rs--yell[1 gre51 & blue--correspond respectively to waves ( %ort] & %ort] l5g?s4 ,la/ly !re comes a pat* ( violet due to ^? ( ! visible rays ^: wave-l5g? is ! %orte/ ( all4 ,! wave-l5g? ( ! light at ! extreme edge ( ! r$ is ab\t #a_/cdjjj in*1 & z we pass along ! spectrum ! wave-l5g? gradually dim9i%es1 until at ! extreme out] edge ( ! violet x is ab\t #a_/fdjjj in*1 or n mu* m ?an half t at ! o!r end4 ,illu/ration3 ,fig4 #a4--;,image ( ,slit & ( ,spectrum4 ,! two ends ( ! spectrum gradually fade away into d>kness1 & ! po9t t ,i wi% to insi/ ^u & make p]fectly cle> is ?is3--,! position ( ! b\nd>ies t]m9at+ ! visible spectrum does n dep5d ^u any?+ :atev] 9 ! nature ( light reg>d$ z a physical ph5om5on4 ,e!r waves :i* are mu* long] & mu* %ort] ?an ^? :i* illum9ate ! spectrum c]ta9ly exi/1 & evid5ce ( _! exi/5ce is easily obta9able4 ,b we _c see !m2 !y fall ^u our eyes )\t excit+ ! fa9te/ s5sation ( light4 ,! visible spectrum is limit$ solely by ! physiological con/itution ( our organs ( vision1 & ! fact t x beg9s & ends :]e x does is1 f a physical po9t ( view1 a m]e accid5t4 ,! spectrum actually project$ ^u ! scre5 is 9 tru? mu* long] ?an t portion ( x :i* any "o c see3 x ext5ds = a consid]able di/ance beyond ! violet at ! "o end & beyond ! r$ at ! o!r1 ^! invisible portions be+ kn[n z ! ultra-violet & infra-r$ regions4 ,p's eyes diff] 9 reg>d to range ( s5sibility j z _! e>s d4 ,i believe ! s5sibility ( my own eyes to be normal1 b if ,i 7 to indicate ! two po9ts :]e ! spectrum appe>s to me to beg9 & to end1 a great _m p]sons w\ld c]ta9ly be incl9$ to disagree ) me & place ! b\nd>ies some:]e else4 ,"s1 inde$1 c\ld see no?+ :atev] 9 :at appe>s to mo/ ( u to be a brilliant portion ( ! r$4 ,aga91 x is by no means probable t 9 all animals & insects ! limits ( vision are ! same z !y are 9 man4 ,we might naturally expect t l>g] & p]haps m co>sely con/ruct$ eyes ?an our own w\ld respond to waves ( great] av]age l5g?1 :ile ! visual organs ( small insects might on ! o!r h& be m s5sitive to %ort] waves4 ,! po9t is n "o t c be easily settl$1 because we are unable to cross-exam9e an animal z to :at x sees "u diff]5t conditions4 ,b ,sir ,john ,lubbock1 tak+ advantage ( ! dislike :i* ants :5 9 _! ne/s h = light1 has prov$ by a s]ies ( v exhau/ive & conclusive exp]im5ts t ^! insects are mo/ s5sitive to rays :i* our own eyes _c p]ceive at all4 ,t region ( ! spectrum :i* appe>s brighte/ to ! eye ( an ant is :at we %\ld call a p]fectly d>k "o1 ly+ outside ! violet1 :]e ! incid5t waves h a l5g? ( less ?an #a_/fdjjj in*4 ,z ,"l ,salisbury said at ,ox=d1 ! function ( ! e!r is to undulate1 &1 9 fact1 x transports en]gy f "o place to ano!r by wave-motion4 ,"s ( its waves1 su* z ^? :i* proce$ f an electric-light dynamo1 may be ?\s&s ( miles 9 l5g?1 o!rs may be %ort] ?an a million? ( an in*1 z is p]haps ! case ) ^? associat$ ) ,pr(essor ,rntg5's ,x-rays2 b all1 s f> z is kn[n1 are ( ess5tially ! same "*1 diff]+ f "o ano!r only z ! bill[s ( ! ,atlantic diff] f ! ripples on ! surface ( a pond4 ,no matt] h[ ! di/urbance is fir/ set up1 :e!r by ! sun1 or by a dynamo1 or by a w>m flat-iron1 9 e case ! e!r conveys no?+ at all b ! en]gy ( wave-motion1 & :5 ! waves1 enc\nt]+ "s mat]ial ob/acle :i* does n reflect !m1 become qu5*$1 _! en]gy takes ano!r =m1 & "s k9d ( "w is done1 or heat is g5]at$ 9 ! ob/acle4 ,! :ole1 or at lea/ ! great] "p1 ( ! en]gy giv5 up by ! waves is 9 mo/ cases trans=m$ into heat1 b "u special circum/ances1 z1 = in/ance1 :5 ! waves fall ^u a gre5 leaf or a liv+ eye1 a few ( !m may p]=m "w ( an electrical or *emical nature4 ,! process ( ! transmission ( en]gy f "o body to ano!r by propagation ?r\gh an int]v5+ m$ium has long be5 spok5 ( z radiation1 & 9 rec5t ye>s ! same t]m has be5 l>gely employ$ to d5ote ! en]gy itself :ile 9 ! /age ( transmission4 ,radiation 9 ! latt] s5se--mean+ e!r wave-en]gy--includes :at is (t5 improp]ly call$ light4 ,light1 p say1 takes ab\t eight m9utes 9 travell+ f ! sun to ! e>?4 ,b :ile x is on its j\rney x is n light 9 ! true s5se ( ! ^w2 nei!r does any?+ ( ! nature ( light ev] />t f ! sun4 ,light has no m exi/5ce 9 nature outside a liv+ body ?an ! flav\r ( onions has2 bo? are m]ely s5sations4 ,if a boy ?r[s a /one :i* hits y 9 ! face1 y feel a pa92 b y d n say t x 0 a pa9 :i* left ! boy's h& & travell$ ?r\gh space f him to y4 ,! /one1 in/ead ( caus+ pa9 9 a s5ti5t be+1 might h brok5 a w9d[1 or knock$ d[n an apple4 ,j s1 ! same radiation :i*1 :5 x *ances to enc\nt] an eye1 produces a c]ta9 s5sation1 w produce a *emical decomposition if x falls ^u a cabbage1 an electrical effect 9 a sel5ium cell1 or a heat+ effect 9 almo/ any?+4 ,:y1 !n1 %\ld x be specially id5tifi$ ) ! s5sation8 ,radiation also includes1 & is ne>ly synonym\s )1 :at is (t5 miscall$ radiant heat4 ,aft] :at has be5 already indicat$1 ,i ne$ h>dly say t !re is no su* ?+ z radiant heat4 ,! tru? is t ! sun or o!r hot body g5]ates wave-en]gy 9 ! e!r at ! exp5se ( "s ( its own heat1 & any di/ant sub/ance :i* absorbs a portion ( ?is en]gy g5]ally 7b n necess>ily7 acquires an equival5t quantity ( heat4 ,! result may be exactly ! same z if heat left ! hot body & travell$ across space to ! sub/ance2 b ! process is diff]5t4 ,x is l s5d+ a sov]eign to a fri5d by a po/al ord]4 ,y "p ) a sov]eign & he receives "o1 b ! piece ( pap] :i* goes ?r\gh ! po/ is n a sov]eign4 ,x is /rictly correct to say t ! sun loses heat by radiation1 j z y lose a sov]eign by inve/+ x 9 ! pur*ase ( a po/al ord]4 ,b t is n ! same ?+ z say+ t ! sun radiates heat4 ,! t]m radiation has ! advantage ( avoid+ any sugge/ion ( ! fallacy t !re is "s ess5tial diff]5ce 9 ! nature ( ! e!r-waves :i* may happ5 to t]m9ate _! respective c>e]s 9 ! production ( light or heat or *emical action or some?+ else2 b x is1 un=tunately1 impossible 9 ! pres5t condition ( ?+s to use x z freely z "o c\ld wi% )\t p$antry1 & we mu/ /ill (t5 speak ( light or ( heat :5 radiation w\ld express our mean+ ) great] accuracy4 ,light1 !n--to use ! t]m unblu%+ly 9 its objectionable b well und]/ood s5se--has ! prop]ty ( /imulat+ c]ta9 n]ves :i* exi/ 9 _m liv+ be+s1 ) ! result t1 9 "s unkn[n & probably unkn[able mann]1 a special s5sation is call$ into play--! s5sation ( lum9osity4 ,& 9 ord] t ! creature may be able n only to p]ceive light b also to see ?+s1 t is1 to appreciate ! =ms ( ext]nal objects1 x is g5]ally provid$ ) an optical app>atus by means ( :i* ! incid5t light is suitably di/ribut$ ov] a l>ge numb] ( indep5d5t s5sitive elem5ts4 ,9 man & ! high] animals ! optical app>atus1 or eye1 consi/s ( a /iff globul> %ell1 hav+ 9 front an op5+ provid$ ) a sy/em ( l5ses1 &1 at ! back ( ! int]ior1 a delicate p]ceptive membrane1 ^u :i* ! transmitt$ light is receiv$4 ,s mu* ( ! light emitt$ or reflect$ f an ext]nal object z passes ?r\gh ! l5ses1 is di/ribut$ by !m 9 su* a mann] z to =m :at is call$ an image ^u ! membrane1 e elem5t>y po9t ( ! image receiv+ ! light :i* issues f a correspond+ po9t ( ! object1 & no o!r4 ,! contrivance evid5tly be>s a close resemblance to a photographic cam]a1 ! s5sitive plate or film1 ^u :i* ! picture is project$1 be+ analog\s to ! p]ceptive membrane4 ,i am n go+ to attempt a detail$ description ( ! human eye4 ,x w be suffici5t to po9t out briefly "s ( its pr9cipal features z indicat$ 9 ! annex$ diagrammatic section1 ,fig4 #b4 ,illu/ration3 ,fig4 #b4--;,diagram ( ! ,eye4 ,! op5+ 9 front ( ! globe is cov]$ by a slightly protub]ant transp>5t m$ium ,c1 :i* is %ap$ l a small wat*-glass1 & on acc\nt ( its horn-l /ructure has be5 nam$ ! cornea4 ,! space betwe5 ! cornea ,c & ! body m>k$ ,l is fill$ ) a wat]y liquid ,a1 kn[n z ! aque\s hum\r3 ?is liquid ) its curv$ surfaces con/itutes a m5iscus l5s1 convex on ! out] side & concave on ! inn]4 ,!n comes ! biconvex cry/all9e l5s ,l1 an ela/ic gelat9\s-look+ solid1 :i* is easily di/ort$ by pressure4 ,! convexity ( ?is l5s c be v>i$ by ! action ( a surr\nd+ muscle ,m ,m1 & 9 ?is way ! focus is adju/$ = objects at diff]5t di/ances f ! eye4 ,:5 ! muscle is relax$ & ! l5s 9 its natural condition1 ! curvature ( its surfaces is su* t a %>p image is =m$ ( objects di/ant ab\t =ty feet & upw>ds4 ,:5 by an ef=t ( w1 ! muscle is contract$1 ! l5s becomes m convex1 & di/9ct pictures c ?us be focuss$ ( ?+s :i* are only a few in*es away4 ,?is process ( adju/m5t by muscul> ef=t is te*nically kn[n z accommodation4 ,! rema9d] ( ! globe is fill$ ) ! s-call$ vitre\s body ,v1 :i* d]ives its "n f its fanci$ resemblance to liquid glass3 x might p]haps be m prop]ly lik5$ to a ?9 col\rless jelly4 ,! vitre\s body plays a "p 9 ! refraction ( ! light4 ,! p]ceptive membrane1 or ret9a ,r ,r1 :i* l9es r m ?an half ! int]ior ( ! eye-ball1 is an exce$+ly complex /ructure4 ,?\gh its av]age ?ickness is less ?an #a_/ajj in* x is kn[n to consi/ ( n9e di/9ct lay]s1 mo/ ( :i* are m>vels ( m9ute intricacy4 ,( ^! lay]s ,i %all notice only two1 ! s-call$ bacill>y lay]1 :i* is 9 imm$iate contact ) ! inn] coat+ ( ! eye-ball1 & ! fibr\s lay]1 or lay] ( optic n]ve fibres1 :i* is only sep>at$ f ! vitre\s body by a ?9 protective film4 ,! bacill>y lay] 7f bacillum1 a w&7 consi/s ( a va/ assemblage ( little elongat$ bodies call$ rods & cones1 :i* are plac$ side by side & set p]p5dicul>ly to ! surfaces ( ! ret9a1 or 9 o!r words1 radially to ! eye-ball4 ,let u try to make ! arrangem5t cle> by an illu/ration4 ,imag9e a small portion ( ! inn] surface ( ! eye-ball1 "o-t5? ( an in* squ>e1 to be magnifi$ #bjjj diamet]s 7f\r million times71 & let ! enl>g$ area be repres5t$ by ! floor ( a room #ag feet squ>e4 ,procure a quantity ( c$> p5cils1 & set !m on ! floor 9 an upright position & v close to "o ano!r4 ,x w be f\nd t ! numb] ( p5cils requir$ to fill ! space w be ab\t half-a-million4 ,to make ! analogy m complete1 let "s ( ! p5cils be %>p5$ to a long tap]+ po9t at _! l[] ends1 ! great] numb] rema9+ uncut1 j z receiv$ f ! manufactur]s4 ,neglect+ details :i* are immat]ial = our pres5t purpose1 we may reg>d ! uncut p5cils z repres5t+ ^u an enorm\sly magnifi$ scale ! rods ( ! ret9a1 & ! po9t$ ones ! cones4 ,! flat upp] ends ( ! p5cils may be pa9t$ 9 diff]5t uni=m col\rs1 & arrang$ s z to =m a l>ge picture 9 mosaic1 & if ?is is look$ at f su* a di/ance t its image on ! ret9a is a t5? ( an in* squ>e 7:i* w be ! case :5 ! picture is ab\t =ty y>ds away7 all possibility ( di/+ui%+ ! sep>ate elem5ts :i* compose x w be lo/1 & ! picture w seem to be a p]fectly cont9u\s "o4 ,al?\gh ! light :i* ent]s ! eye _c rea* ! rods & cones until x has trav]s$ all ! o!r lay]s ( ! ret9a1 yet ^! int]v5+ lay]s1 be+ transp>5t1 (f] little ob/ruction to its passage1 & x c h>dly be d\bt$ t ! rods & cones are ! special organs ^u :i* light ex]ts its action1 ! picture focuss$ ^u _! ends be+ 9 tru? an exce$+ly f9e mosaic4 ,f e sep>ate elem5t ( ! mosaic--f e s+le rod & cone--!re proce$s a sl5d] transp>5t filam5t3 all ^! make _! way ?r\gh ! int]m$iate lay]s ( ! ret9a1 )\t1 z is believ$1 any break ( functional cont9uity1 & em]ge ne> its int]nal surface2 "h !y b5d ov] at "r angles1 & ! ?\s&s ( filam5ts =m a tangle :i* l9es ! inside ( ! eye l a f9e network1 & con/itute ! lay] ( optic n]ve-fibres already ref]r$ to4 ,! filam5ts1 or n]ve-fibres1 d n h[ev] t]m9ate )9 ! eye2 !y all pass ?r\gh ! hole m>k$ ,n 9 ! figure1 & !nce1 9 ! =m ( a _m-/r&$ cable1 con/itut+ ! optic n]ve1 !y are l$ to ! bra91 to :i* ea* individual fibre is sep>ately atta*$4 ,if1 !re=e1 :at ,i h said is true--&1 ?\gh x has n1 ,i believe1 be5 all rigor\sly prov$1 yet ! evid5ce 9 its support is exce$+ly cog5t--x foll[s t e "o ( ! multitude ( rods & cones has its own indep5d5t l9e ( communication ) ! bra94 ,! m9d1 :i* is my/]i\sly connect$ ) ! bra91 is ?us af=d$ ! means ( localis+ all ! po9ts ( lum9\s excitation relatively to "o ano!r1 & furni%$ ) data = e/imat+ ! =m ( ! object f :i* ! light proce$s4 ,!re are two small regions ( ! ret9a :i* are ( special int]e/4 ,"o ( !m lies j ov] ! op5+ ,n :]e ! optic n]ve ent]s4 ,"h x is evid5t t !re c be no rods & cones1 _! place be+ :olly occupi$ by /r&s ( n]ve-fibre4 ,n[ x is rem>kable t ?is spot is totally ins5sitive to light4 ,! o!r int]e/+ portion is situat$ opposite ! middle ( ! front op5+1 & is m>k$ by a small yell[ pat*1 9 ! c5tre ( :i* is a depression or pit1 :i* is %[n 9 an exagg]at$ =m at ,f1 & is call$ ! fovea4 ,x has be5 asc]ta9$ t ! depression is due p>tly to ! abs5ce ( ! lay] ( n]ve-fibres1 :i* are "h b5t aside out ( _! natural c\rse1 & p>tly to a local r$uction 9 ! ?ickness ( "s ( ! int]m$iate ret9al lay]s4 ,?is spot1 be+ at ! c5tre ( ! field ( vision1 occupies a position ( great importance1 & ! evid5t purpose ( ! sup]ficial depression is to all[ ! light to rea* ! und]ly+ bacill>y lay] ) z little ob/ruction z possible4 ,x is notewor?y t ! bacill>y lay] b5ea? ! yell[ spot is compos$ entirely ( cones1 ! rods1 :i* else:]e are 9 excess1 be+ altoge!r want+4 ,! only o!r accessory ( ! visual app>atus to :i* ,i %all ref] is ! iris 7,i ,i1 ,fig4 #b71 a col\r$ disk hav+ a c5tral p]=ation4 ,?is c be se5 ?r\gh ! cornea & is consequ5tly a v famili> object4 ,! iris s]ves ! same purpose z ! /op1 or diaphragm1 ( a photographic l5s1 its function be+ to limit & regulate ! quantity ( light :i* is admitt$ into ! eye4 ,! size ( ! c5tral op5+1 or pupil1 v>ies automatically ) ! int5sity ( ! illum9ation3 9 a /rong light ! op5+ becomes small2 9 a feeble light or 9 d>kness x is enl>g$4 ,! pupil also contracts :5 ! eye is focuss$ ^u a ne> object & dilates :5 ! vision is direct$ to a di/ance4 ,?is brief sket* may s]ve to give "s slight idea ( ! complexity & delicacy ( ! visual app>atus4 ,only a few ( its m sali5t features h be5 t\*$ ^u2 :5 our scrut9y is c>ri$ into details ! complexity becomes bewild]+4 ,ev5 su* simple-look+ ?+s z ! cornea & ! vitre\s body turn out on close exam9ation to be mo/ elaborately con/itut$4 ,mu*1 no d\bt1 rema9s to be discov]$1 & ( :at has already be5 inve/igat$ mu* is at pres5t only p>tially und]/ood4 ,& yet1 ?\gh x is true t man is fe>fully & wond]fully made1 x is equally true t he is f> f p]fect2 & :ile !re is no /ructure 9 ! :ole human anatomy :i* exhibits s abundant a pr(usion ( m>vels z ! eye1 !re is p]haps none :i* is m>k$ ) imp]fections s /rik+4 ,_m ( its defects are ! m /rik+ because !y are s obvi\s1 be+ su* z w\ld nev] be tol]at$ 9 optical in/rum5ts ( human manufacture4 ,9 any fairly good cam]a or telescope or microscope we %\ld expect to f9d t ! l5ses 7 symmetrically figur$1 free f /ri & prop]ly c5tr$2 also t !y 7 a*romatic & effici5tly correct$ = sph]ical ab]ration4 ,9 ! eye n "o ( ^! elem5t>y requirem5ts is fulfill$4 ,! ext]nal surface ( ! l5s =m$ by ! aque\s hum\r & ! cornea is n a surface ( revolution1 su* z w\ld be fa%ion$ by a turn+ la! or a l5s-gr9d+ ma*9e2 its curvature is great] 9 a v]tical ?an 9 a horizontal direction1 & ! di/9ctness ( ! focuss$ image is consequ5tly impair$4 ,aga91 ! cry/all9e l5s is con/ruct$ ( a numb] ( sep>ate portions :i* are imp]fectly jo9$ toge!r4 ,/ri occur along ! junctions1 & ! light :i* trav]ses !m1 in/ead ( be+ uni=mly refract$1 is scatt]$ irregul>ly4 ,moreov] ! sy/em ( l5ses is n c5tr$ ^u a common axis2 nei!r is x a*romatic1 :ile ! means employ$ = correct+ sph]ical ab]ration are inadequate4 ,! pur*as] ( an optical in/rum5t :i* turn$ out to h su* faults z ^! w\ld c]ta9ly1 z ! late ,pr(essor ,helmholtz rem>k$1 be ju/ifi$ 9 return+ x to ! mak] & blam+ him sev]ely = 8 c>elessness4 ,i w\ld n1 ( c\rse1 h x believ$ t sci5tific m5 are conceit$ en\gh to imag9e !mselves capable ( design+ a bett] eye ?an is to be f\nd 9 nature4 ,t w\ld be an absurdity4 ,!y are q ready to admit t !re may exi/ suffici5tly good reasons = ! und\bt$ blemi%es :i* h be5 indicat$1 z well z = o!rs :i* w be ref]r$ to lat]4 ,x is inde$ well kn[n t ! g5]al effici5cy ( a ma*9e z a :ole may (t5 be be/ secur$ by ! sacrifice ( ideal p]fection 9 "s ( its p>ts4 ,) all its anomalies ! eye fulfils its prop] function v p]fectly1 & is reg>d$ by ^? :o h /udi$ x mo/ closely ) feel+s ( wond] & humble admiration4#c ,,*apt] ,,ii4 ,,col\r ,,& ,,its ,,p]ception4 ,x 0 expla9$ 9 ! la/ *apt] t we see ?+s ?r\gh ! ag5cy ( ! light--emitt$ or reflect$--:i* proce$s f !m to ! eye1 & is suitably di/ribut$ ov] ! ret9a by ! action ( a sy/em ( l5ses4 ,n[ ! image ?us =m$ is n g5]ally p]ceiv$ z a simple mono*romatic "o1 d>k] 9 "s p>ts1 light] 9 o!rs1 l a black & :ite engrav+4 ,x is1 9 mo/ cases at lea/1 *>act]is$ by a v>iety ( col\rs1 ! light :i* comes f diff]5t objects1 or f diff]5t p>ts ( ! same object1 hav+ ! p[] ( excit+ diff]5t col\r s5sations4 ,light :i* has ! prop]ty ( excit+ ! s5sation ( any col\r is commonly spok5 ( z col\r$ light4 ,! light reflect$ by a soldi]'s coat1 = example1 may be call$ r$ light1 because :5 x falls ^u ! eye x gives rise to a s5sation ( r$ness4 ,b x mu/ be und]/ood t ?is mode ( expression is only a conv5i5t abbreviation1 = !re c1 ( c\rse1 be no objective col\r 9 ! light or radiation itself4 ,:]e91 !n1 does col\r$ light diff] f :ite8 ,:y d ?+s appe> to be v>i\sly col\r$ :5 illum9at$ by light :i* is col\rless8 ,& h[ d col\r$ lights affect ! visual organs s z to evoke appropriate s5sations8 ,^! are que/ions--! fir/ two ( a physical "*1 ! la/ p>tly physiological & p>tly psy*ological--:i* x is n[ propos$ to discuss4 ,! matt] has already be5 t\*$ ^u1 ?\gh v slightly1 9 connection ) ! spectrum4 ,let u aga9 turn to ! spectrum & consid] x a little m fully4 ,x is easily se5 t ! lum9\s b& conta9s six pr9cipal hues or tones ( col\r--r$1 orange1 yell[1 gre51 blue1 & violet4 7,see ,fig4 #a1 page #ab47 ,^! h[ev] m]ge into "o ano!r s gradually t x is impossible to say exactly :]e any "o col\r beg9s & ends4 ,look1 = in/ance1 at ! some:at n>r[ b v conspicu\s /ripe ( yell[4 ,t[>ds ! "r ( ?is /ripe ! col\r gradually becomes gre5i%-yell[2 a little fur!r on x is yell[i%-gre51 & at l5g?1 by ins5sible gradations1 a full1 pure gre5 is rea*$4 ,! six mo/ prom95t hues ( ! spectrum are1 9 fact1 supplem5t$ by an imm5se multitude ( subord9ate ones1 ! total numb] :i* ! eye c recognise z di/9ct be+ n less ?an a ?\s&4 ,all ! col\rs t we see 9 nature1 ) ! exception ( ! purples 7ab\t :i* ,i %all say m pres5tly71 are "h repres5t$1 & e s+le v>iety ( tone 9 ! prismatic scale corresponds ) "o1 & only "o1 def9ite wave-l5g? ( light4 ,! s\rce ( all ^! col\rs is1 z we "k1 a beam ( :ite or col\rless light1 ! con/itu5ts ( :i* h be5 sort$ out & arrang$ s t !y fall side by side ^u ! scre5 9 ! ord] ( _! sev]al wave-l5g?s4 ,if1 !n1 ^! col\r$ con/itu5ts 7 all mix$ toge!r aga91 x w\ld be reasonable to expect t pure :ite light w\ld be reproduc$4 ,! exp]im5t has be5 p]=m$ 9 a great _m diff]5t ways1 sev]al ( :i* 7 devis$ by ,newton himself1 & ! result admits ( no d\bt :atev]4 ,! me?od :i* ,i int5d to describe is n q s simple z "s o!rs1 b x has great advantages 9 ! way ( conv5i5t manipulation1 & af=ds ! means ( demon/rat+ a numb] ( int]e/+ col\r effects 9 an easily intelligible mann]4 ,by ! simple op]ation ( mov+ aside a l5s out ( ! track ( ! light1 we c ga!r up & ?or\ghly mix toge!r all ! v>i\sly col\r$ rays ( ! spectrum & cause !m to =m ^u ! scre5 a bright circul> pat*1 :i*1 ?\gh due to a mixture ( a ?\s& diff]5t hues1 is absolutely :ite4 ,:5 ! l5s is replac$1 :i* is done 9 an in/ant1 ! mixture is aga9 analys$ into its compon5t p>ts1 & ! spectrum reappe>s4 ,! arrangem5t ( ! app>atus1 :i* is ess5tially ! same z t devis$ by ,capta9 ,abney1 & call$ by him ! col\r-pat* app>atus1 is %[n 9 ! annex$ diagram 7,fig4 #c74 ,illu/ration3 ,fig4 #c4--;,abney's ,col\r-pat* ,app>atus4 ,! light ( an electric lamp ,a plac$ inside ! lant]n is conc5trat$ by ! cond5s+ l5ses ,b ^u a n>r[ adju/able slit ,c4 ,! framework ( ?is slit is atta*$ to "o end ( a telescope tube1 :i* c>ries at ! o!r end an a*romatic l5s ,d ( ab\t #aj in*es focus4 ,! rays hav+ be5 r5d]$ p>allel by ,d are refract$ by ! prism ,e2 !y !n pass ?r\gh a circul> op5+ 9 ! brass plate ,f to ! l5s ,g1 ! focal l5g? ( :i* is #g in*es1 & =m a little bright spectrum ^u a :ite c>d held 9 a groov$ support at ,h4 ,! c>d be+ remov$1 we place at ,k a l5s hav+ a diamet] ( #e-a_/b in*es & a focal l5g? ( #ah in*es or m1 & adju/ x s t a %>ply def9$ image ( ! hole 9 ! brass plate ,f is =m$ ^u ! di/ant :ite scre5 ,l4 ,if all ! l5ses are correctly plac$1 ?is image1 ?\gh =m$ entirely by ! rays :i* con/itut$ ! little spectrum at ,h1 w be p]fectly free f col\r ev5 ar\nd ! edge4 ,if we wi% to project ^u ! scre5 ,l an enl>g$ image ( ! little spectrum1 we h only to use ano!r suitable l5s ,i 9 conjunction ) ,k3 ! diamet] ( t us$ by myself is #b-c_/d in*es1 & its focal l5g? #f-a_/b in*es4 ,:5 we h once f\nd by trial ! position 9 :i* ?is supplem5t>y l5s gives ! cle>e/ image#d x is easy to arrange a contrivance = remov+ & replac+ x correctly )\t ne$ ( any fur!r adju/m5t4 ,?is app>atus %[s !n t ord9>y :ite light may be reg>d$ z a mixture ( all ! v>i\sly col\r$ lights :i* occur 9 ! spectrum1 ! s5sation produc$ :5 x falls ^u ! eye be+ consequ5tly a comp\nd "o4 ,f ^! & simil> exp]im5ts ! sci5tific neophyte is n unlikely to draw an errone\s conclusion4 ,:ite light1 he is apt to ?9k1 is always due to ! comb9$ action ( rays ( e possible wave-l5g?1 :ile col\r$ light consi/s ( rays ( "o def9ite wave-l5g? only4 ,nei!r ( ^! inf]5ces w\ld be correct4 ,x is n true t :ite light necess>ily conta9s rays ( all possible wave-l5g?s3 ! s5sation ( :it5ess may1 z w be %[n by & bye1 be produc$ q z effectively by ! comb9ation ( only two or ?ree diff]5t wave-l5g?s4 ,nor is x true t su* col\rs z we see 9 nature are always due to light ( a s+le wave-l5g?2 light ( ?is k9d is inde$ r>ely met ) outside laboratories & lecture rooms4 ,f> m commonly col\r$ light consi/s ( mix$ rays1 & l ord9>y :ite light1 x may1 & g5]ally does1 conta9 all ! col\rs ( ! spectrum1 b 9 diff]5t proportions4 ,?is la/ ass]tion is easily prov$4 ,by means ( a slip ( c>d we may int]cept a portion ( ! little spectrum =m$ at ,h 7,fig4 #c74 ,! d>k %ad[ ( ! c>d 9 ! enl>g$ spectrum on ! scre5 is %[n 9 ,fig4 #d4 ,x w be notic$ t ! %ad[ cuts (f a "p only ( ! r$1 orange1 & yell[ light1 all[+ ! rema9d] to pass ?r\gh ! projection l5ses4 ,!re are /ill rays ( e possible wave-l5g? f extreme r$ to extreme violet1 b ! proportion ( ^? t[>ds ! r$ end is less ?an x 0 be=e ! c>d 0 int]pos$4 ,illu/ration3 ,fig4 #d4--,p>tially int]cept$ ,spectrum4 ,if n[ we remove ! l5s ,i 7,fig4 #c7 & s mix ! col\rs ( ?is mutilat$ spectrum1 ! bright r\nd pat* :]e ! mix$ rays fall ^u ! scre5 w no long] appe> :ite b gre5i%-blue4 ,if we transf] ! c>d to ! o!r end ( ! little spectrum1 s z to cause a p>tial eclipse ( ! violet1 blue1 & gre5 rays1 ! col\r ( ! pat* w be *ang$ to orange4 ,if we remove ! c>d altoge!r1 ! pat* w once m become :ite4 ,x foll[s a =tiori t :5 any portion ( ! little spectrum is eclips$ totally1 in/ead ( only p>tially1 ! light f ! rema9d] w appe>1 :5 comb9$1 to be col\r$4 ,v beautiful *anges ( hue are exhibit$ by ! bright pat* :5 a n>r[ opaque /rip1 su* z ! small blade ( a pocket knife1 is sl[ly mov$ along ! little spectrum at ,h1 eclips+ diff]5t portions ( x 9 succession4 ,! pat* fir/ becomes gre51 !n by imp]ceptible gradations x *anges successively to blue1 purple1 sc>let1 orange1 yell[1 & f9ally1 :5 ! knife has complet$ its c\rse1 all col\r disappe>s & ! pat* is aga9 :ite4 ,we may improve ^u ?is crude exp]im5t1 &1 aft] ,capta9 ,abney's plan1 prep>e a numb] ( small c>dbo>d /5cils1 ) op5+s correspond+ to any select$ p>ts ( ! little spectrum4 ,:5 a c>d s prep>$ is plac$ at ,h 7,fig4 #c7 ! bright pat* ^u ! scre5 is =m$ by ! comb9ation ( ! select$ rays1 all ! o!rs be+ qu5*$4 ,we %all f9d t "u ^! conditions ! bright pat* is g5]ally1 b n always1 col\r$4 ,illu/ration3 ,fig4 #e4--,/5cil ,c>ds4 ,! fir/ diagram 9 ,fig4 #e repres5ts a black5$ c>d1 :i* all[s only ! r$ & a little ( ! orange to pass ?r\gh4 ,:5 ?is is ins]t$ 9 ! groov$ hold] at ,h1 ! bright pat* imm$iately turns r$4 ,! second diagram %[s ano!r1 :i* transmits ! middle portion ( ! spectrum1 b blocks ! r$ & ! violet at its two ends3 ) ?is c>d ! col\r ( ! pat* becomes gre54 ,! ?ird c>d has op5+s = ! violet & ! r$ rays3 ?is turns ! pat* a beautiful purple1 a hue :i*1 z already m5tion$1 is n produc$ by light ( any s+le wave-l5g?4 ,! purples are mixtures ( r$ & violet or ( r$ & blue4 ,n[ ,i h 9 my possession ?ree pieces ( glass 7or1 to be /rictly accurate1 two pieces ( glass & "o glass-m\nt$ gelat9e film7 :i*1 :5 plac$ transv]sely 9 ! beam ( light1 ei!r at ,h 7,fig4 #c7 or any:]e else1 behave exactly l ^! ?ree c>dbo>d /5cils4 ,! fir/ glass cuts (f all ! spectrum except ! r$ & "p ( ! orange1 j z ! fir/ /5cil does1 ?\gh ! l9e ( dem>cation is n q s %>p4 ,?is is 9 fact a piece ( r$ glass1 or 9 o!r words ! light t x transmits produces ! s5sation ( r$4 ,! second glass1 l ! second /5cil1 all[s ! :ole ( ! spectral rays to pass freely except ! r$ & ! violet1 :i* disappe> z if !y 7 ob/ruct$ by an opaque body4 ,?is is a gre5 glass4 ,& ! ?ird 7:i* is really a film ( gelat9e7 cuts out ! middle ( ! spectrum b transmits ! r$ & violet ends4 ,! col\r ( ! gelat9e is purple4#e ,! glasses & ! gelat9e 9 "q act l ! c>dbo>d /5cils 9 completely cutt+ (f "s ( ! spectral rays & transmitt+ o!rs1 & !y owe _! app>5t col\rs to ! comb9$ influ5ce :i* ! transmitt$ rays ex]t ^u ! eye4 ,_m o!r col\r$ glasses m]ely weak5 "s ( ! rays1 )\t entirely qu5*+ any4 ,a piece ( pale yell[ glass1 = example1 :5 plac$ 9 ! pa? ( ! beam ( light f :i* ! spectrum on ! scre5 is =m$1 simply dim9i%es ! brightness ( ! blue region & does n :olly qu5* any ( ! rays2 & aga91 a common k9d ( violet-col\r$ glass enfeebles1 b does n q oblit]ate1 ! middle portion ( ! spectrum4 ,f su* obs]vations z ^! we inf] t ! glasses d]ive _! respective col\rs f ! light :i* falls ^u !m4 ,! fir/ glass w\ld n appe> r$ if se5 9 a light :i* conta9$ no r$ rays4 ,?is is easily prov$ by an exp]im5t ) ! col\r-pat* app>atus4 ,! spectrum be+ once m comb9$ into a bright :ite pat* 7:i* turns r$ if ! glass is = a mom5t int]pos$71 let all ! r$ rays & "p ( ! orange be cut (f ) a suitable /5cil4 ,! re-comb9$ light is no long] :ite b gre5i%-blue1 z is evid5c$ by ! col\r ( ! pat*2 & no?+ t is illum9at$ by ?is light c possibly appe> r$4 ,! piece ( r$ glass1 if plac$ 9 ! beam1 w n[ ca/ a p]fectly black %ad[1 & a squ>e ( bright r$ pap] held 9 ! middle ( ! pat* w look z black z ink4 ,x w be %[n lat] h[ we may obta9 light :i*1 al?\gh x appe>s to ! eye to diff] 9 no respect f ord9>y :ite daylight1 yet conta9s no r$ compon5t1 & is consequ5tly z p[]less z ?is gre5i%-blue light to reveal any r$ col\r 9 ! objects :i* x illum9ates4 ,if we sub/itute a /5cil :i* admits only r$ rays1 we %all obta9 a beam ( light 9 :i* no col\r b r$ c be se54 ,gre5 & blue glasses :5 expos$ to ?is light w ca/ black %ad[s1 :ile pieces ( gre5 & blue pap] w become ei!r black or d>k grey4 ,we see !n t ! col\rs ( transp>5t objects1 l ! glasses us$ 9 ^! exp]im5ts1 are br\ght out by a process ( filtration4 ,c]ta9 ( ! col\r$ +r$i5ts ( :ite light are filt]$ out & qu5*$ inside ! glass1 & x is to ! rema9+ +r$i5ts :i* pass ?r\gh unimp$$ t ! obs]v$ col\r is due4 ,! en]gy ( ! absorb$ rays is n lo/ ( c\rse1 = en]gy1 l matt]1 is inde/ructible4 ,x is trans=m$ into heat4 ,a col\r$ glass held 9 a /rong beam ( light w 9 a %ort "t become s5sibly w>m] ?an "o t is cle> & col\rless4 ,9 /udy+ col\r effects z produc$ by col\r$ glasses1 we h at ! same "t be5 le>n+ h[ ! great majority ( natural objects--n only ^? :i* are transp>5t b also ^? call$ opaque--become possess$ ( _! col\rs4 ,= ! tru? is t few ?+s are p]fectly opaque4 ,:5 :ite light falls ^u a col\r$ body1 x g5]ally p5etrates to a small dep? bel[ ! surface1 & 9 s do+ loses by absorption "s ( its col\r$ compon5ts1 j z x does 9 pass+ ?r\gh ! pieces ( glass4 ,b be=e x has gone v f>--g5]ally mu* less ?an a ?\s&? "p ( an in*--x has enc\nt]$ a numb] ( little reflect+ surfaces due to optical irregul>ities1 :i* turn ! light back aga9 & compel x to pass a second "t ?r\gh ! same ?ickness ( ! sub/ance3 x ?us becomes /ill m effectively sift$1 & on em]g+ is imbu$ ) a col\r due to su* ( ! compon5ts z h n be5 qu5*$ 9 ! c\rse ( _! d\ble j\rney ?r\gh a sup]ficial lay] ( ! sub/ance4 ,any col\r$ rays reflect$ by an object mu/ necess>ily be conta9$ 9 ! light by :i* ! object is se54 ,! foll[+ is a curi\s exp]im5t illu/rat+ ?is4 ,a l>ge bright spectrum is project$ ^u a scre5 & 9 ! gre5 or blue portion ( x is held a wall po/]4 ,! lett]s & figures ^u ! pap] are se5 to /& out boldly z if pr9t$ ) ! blacke/ ink4 ,b if ! po/] is mov$ into ! r$ "p ( ! spectrum1 ! pr9t+ at once disappe>s z if by magic1 & ! pap] appe>s p]fectly blank4 ,! explanation is t ! lett]s are pr9t$ 9 r$ ink--!y c reflect no light b r$4 ,gre5 or blue light fall+ ^u !m is absorb$ & qu5*$1 & ! lett]s consequ5tly appe> black4 ,on ! o!r h& :5 ! po/] is illum9at$ by ! r$ rays ( ! spectrum1 ! lett]s reflect j z mu* light z ! pap] itself1 & are !re=e indi/+ui%able f x4 ,any?+ :i*1 :5 illum9at$ by a s\rce ( :ite light1 reflects all its v>i\s compon5ts equally & )\t absorb+ a l>g] proportion ( "s ?an ( o!rs1 appe>s :ite or grey4 ,betwe5 :ite & grey !re is no ess5tial diff]5ce except 9 lum9osity1 or brightness1 t is to say1 9 ! quantity ( light reflect$ to ! eye1 or--to g a /ep fur!r back--9 ! amplitude ( ! e!r waves4 ,"u diff]5t conditions ( illum9ation any sub/ance :i* reflects all ! rays ( ! spectrum equally may appe> ei!r :ite or grey1 or ev5 black4 ,a sn[ball c easily be made to look black] ?an pit*1 & a block ( pit* :it] ?an sn[4 ,x mu/ h /ruck _m ( ^? :o h ?\ght ab\t ! matt] at all z a mo/ rem>kable co9cid5ce t sunlight %\ld be :ite4 ,:ite light1 z we h se51 consi/s ( a mixture ( v>i\sly-col\r$ rays 9 v diff]5t & app>5tly arbitr>y proportions1 & if ^! proportions 7 a little *ang$ ! light w\ld no long] be q col\rless4 ,no ord9>y artificial light is s exactly :ite z t ( ! sun4 ,! light ( c&les1 gas1 oil1 & electric gl[-lamps is yell[2 t ( ! electric arc 7:5 unaffect$ by atmosph]ic absorption7 is blue1 & t ( ! inc&esc5t gas burn] gre54 ,x is exce$+ly conv5i5t t ! light :i* s]ves u = ! great] "p ( our wak+ lives %\ld happ5 to be j s con/itut$ t x is col\rless4 ,b on a little fur!r reflection x w1 ,i ?9k1 appe> t ?is is n ! "r way to look at ! matt]4 ,x is precisely because ! hue call$ :ite is ! "o :i* is associat$ ) ! light ( our sun t we reg>d :it5ess z synonym\s ) abs5ce ( col\r4 ,we take sunlight z our /&>d ( neutrality1 & any?+ t reflects x )\t alt]+ ! proportions ( its con/itu5ts we consid] z be+ col\rless4 ,!re c be little d\bt t if ! sun 7 purple in/ead ( :ite1 our s5tim5ts z reg>ds ^! two hues w\ld be int]*ang$2 we %\ld talk q naturally ( a pure purple1 entirely free f any trace ( col\r1 or p]haps describe a lady's co/ume z be+ ( a gaudy :ite4 ,ev5 z ?+s are1 ! /&>d ( neutrality is n q a h>d & fa/ "o4 ,we h a t5d5cy to reg>d any artificial light :i* we may happ5 to be us+1 z m free f col\r ?an x w\ld turn out to be if comp>$ directly ) sunlight4 ,if 9 ! middle ( ! day we g sudd5ly into a gas-lit room1 we _c fail to obs]ve h[ int5sely yell[ ! illum9ation at fir/ appe>s2 9 a few m9utes1 h[ev]1 ! col\r loses its obtrusiv5ess & we cease to take mu* notice ( x4 ,! effect may be p>tly a physiological "o1 dep5d+ ^u unequal fatigue ( ! v>i\s p]ceptive n]ves ( ! ret9a2 b ,i believe t x is to a l>ge ext5t due to m5tal judgm5t4 ,! /&>d ( :it5ess1 or col\r-z]o1 c app>5tly be *ang$ )9 c]ta9 limits 9 a v %ort "t1 &1 z we %all see lat]1 ?is is only "o ( _m in/ances 9 :i* our organs ( vision seem to be incapable ( recognis+ a con/ant /&>d ( ref]5ce4 ,& n[ let u consid] h[ x comes ab\t t ea* elem5t>y portion ( ! ret9a--at lea/ 9 its c5tral region--has ! p[] ( di/+ui%+ s _m hundr$s ( diff]5t hues4 ,x is incr$ible t e little area ( microscopic dim5sions %\ld be furni%$ ) su* a multitude ( indep5d5t organs z w\ld be necess>y if ea* ( ! _m col\rs met ) 9 nature requir$ a sep>ate organ = its p]ception2 & x is n necess>y to suppose any?+ ( ! k9d4 ,exp]im5t %[s t all ! v>i\s hues ( ! spectrum1 z well z all 7includ+ :ite7 t c be =m$ f _! mixture1 may be d]iv$ f no m ?an ?ree di/9ct col\rs4 ,!re are1 9 fact1 an indef9ite numb] ( triads ( col\rs :i*1 9 suitable comb9ations1 are capable ( produc+ ! s5sation ( e tone1 t9t1 & %ade ( col\r :i* ! eye ( man has ev] beheld4 ,old-fa%ion$ books1 su* z an e>ly edition ( ,ganot's ,physics1 tell u t ! ?ree prim>y col\rs are r$1 yell[1 & blue1 & t all o!rs are produc$ by mixtures ( ^!4 ,?is 0 ! basis ( ,sir ,david ,brew/]'s !ory1 :i* atta9$ a v wide popul>ity1 & ev5 at ! pres5t "t is held z an article ( fai? among ! great majority ( intellig5t p]sons :o h n paid any special att5tion to sci5ce4 ,b x is n true4 ,a fatal objection to x is ! well-asc]ta9$ fact t no comb9ation ( r$1 yell[1 & blue1 or ( any two ( !m1 su* z blue & yell[1 = example1 w produce gre54 ,yet e pa9t] kn[s t if he mixes blue & yell[ pigm5ts toge!r he gets gre54 ,t is "o ( ! fir/ ?+s t a *ild le>ns :5 he is all[$ to play ) a box ( wat]-col\rs1 & no d\bt ,brew/] 0 misl$ by ! fact4 ,! tru? is1 t ! col\rs ( all1 or almo/ all1 kn[n blue & yell[ pigm5ts happ5 to be composite4 ,an ord9>y blue pa9t reflects n only blue light1 b a l>ge quantity ( gre5 z well2 :ile an ord9>y yell[ pa9t reflects a l>ge quantity ( gre5 light 9 addition to yell[4 ,:5 su* pa9ts are mix$ toge!r1 ! blue & yell[ hues neutralise "o ano!r1 & only ! gre51 :i* is common to bo?1 rema9s4 ,! spectrum app>atus w make ?is cle>]4 ,hold a piece ( bright blue glass be=e ! slit2 ! light pass+ ?r\gh ! glass w be analys$ by ! prism1 & y w see t x really conta9s almo/ z mu* gre5 z blue4 ,if a yell[ glass is sub/itut$1 n only w yell[ light be transmitt$1 b1 z be=e1 a consid]able quantity ( gre54 ,if n[ bo? glasses be plac$ toge!r be=e ! slit1 :at w happ58 ,! yell[ glass w /op ! blue light transmitt$ by ! blue glass1 ! blue glass w /op ! yell[ light transmitt$ by ! yell[ glass1 & only ! gre5 light :i* bo? glasses h ! p[] ( transmitt+ w pass ?r\gh unimp$$1 =m+ a b& ( pure gre5 col\r ^u ! scre54 ,! comb9ation ( simple blue & yell[ lights ( suitable relative lum9osities results 9 ! =mation ( :ite or neutral light4 ,if ! blue is a little 9 excess1 ! comb9$ light w be ( a blui% t9t2 if ! yell[ is 9 excess1 ! comb9ation w h a yell[i% t9t4 ,x w nev] conta9 any trace ( gre54 ,! comb9ation ( simple spectral blue & yell[ is easily effect$ by ! col\r-pat* app>atus1 & ! result w be f\nd to be> out :at has be5 said4 ,s9ce1 !n1 no mixture ( r$1 yell[1 & blue1 or ( any two ( !m1 w produce gre51 we _c reg>d ^! col\rs z be+1 9 ,brew/]'s s5se ( ! t]m1 prim>y ones4 ,b x is q possible to f9d a gr\p ( ?ree diff]5t hues--& inde$ _m su* gr\ps--:i* :5 made to act ^u ! eye simultane\sly & 9 ! "r proportions c give rise to ! s5sation ( any col\r :atev]4 ,n[ ?is exp]im5tal fact is obvi\sly sugge/ive ( a possible conv]se1 namely1 t almo/ e col\r s5sation may 9 reality be a comp\nd "o1 ! resultant ( n m ?an ?ree simple s5sations4 ,assum+ ?is to be s1 x is evid5t t if ea* elem5t>y area ( ! ret9a 7 provid$ ) only ?ree suitable col\r organs1 no?+ m w\ld be requisite = ! p]ception ( an indef9ite numb] ( di/9ct col\rs4 ,su* a hypo!sis 0 fir/ propos$ by ,?omas ,"y at ! beg9n+ ( ! pres5t c5tury2 b x came be=e its "t & met ) no att5tion until fifty ye>s lat]1 :5 x 0 une>!d by ! di/+ui%$ physici/ & physiologi/1 ,helmholtz1 :o accord$ to x 8 p[]ful support & modifi$ x 9 "o or two important details4 ,illu/ration3 ,fig4 #f4--;,helmholtz's ,curves ( ,col\r ,p]ception4 ,accord+ to ! ,"y-,helmholtz !ory1 z x is n[ call$1 !re are ?ree diff]5t k9ds ( n]ve-fibres di/ribut$ ov] ! ret9a4 ,! fir/1 :5 sep>ately /imulat$1 produce ! s5sation ( r$1 ! second t ( gre51 & ! ?ird t ( violet4 ,light hav+ ! same wave-l5g? z ! extreme r$ rays ( ! spectrum /imulates ! r$ n]ve-fibres only2 t hav+ ! same wave-l5g? z ! extreme violet rays /imulates ! violet n]ve-fibres only4 ,light ( all int]m$iate wave-l5g?s1 correspond+ to ! orange1 yell[1 gre51 & blue ( ! spectrum1 /imulates all ?ree sets ( n]ve-fibres at once1 b 9 diff]5t degrees4 ,! proportionate /imulation ( ! r$1 gre51 & violet n]ves ?r\gh\t ! spectrum is indicat$ 9 ,fig4 #f1 :i* is d]iv$ f ! r\gh sket* fir/ giv5 by ,helmholtz4 ,! yell[ rays ( ! spectrum1 x w be se51 excite ! r$ & gre5 n]ves /rongly1 & ! violet feebly2 gre5 light excites ! gre5 n]ves /rongly1 & ! r$ & violet mod]ately2 :ile blue light excites ! gre5 & violet n]ves /rongly1 & ! r$ feebly4 ,illu/ration3 ,fig4 #g4--,knig's ,curves4 ,fig4 #g %[s ano!r set ( curves giv5 m rec5tly by ,dr4 ,knig z ! result ( _m ?\s&s ( exp]im5ts made1 n only ^u p]sons ^: vision 0 normal1 b also ^u "s :o 7 col\r-bl9d4 ,knig f\nd t ! equations he obta9$ 7 be/ satisfi$ by assum+ z ! normal fundam5tal s5sations a purpli% r$ 7n to be f\nd 9 ! spectrum71 a gre5 l t ( wave-l5g? #ejej1 & a blue l t ( wave-l5g? #dgjj approximately1 ! two latt]1 h[ev]1 be+ pur] or m saturat$ ?an any actual spectrum col\r4 ,b ,knig's curves are n consi/5t ) e class ( vision :i* he exam9$1 & ! "q z to :at are ! true fundam5tal col\r-s5sations1 if su* really exi/ at all1 _c yet be reg>d$ z f9ally settl$4#f ,! ,"y-,helmholtz !ory ( col\r-vision1 :e!r or n x is de/9$ 9 ! future to be sup]s$$ by "s o!r1 has at all ev5ts prov$ an invaluable guide 9 exp]im5tal "w1 & !re are v few col\r ph5om5a ( :i* x is n compet5t to (f] a satisfactory explanation4 ,x has at pres5t only "o s]i\s rival--! !ory ( ,h]+1 :i*1 al?\gh x seems to be curi\sly attractive to _m physiologi/s1 c h>dly be said to pres5t less s]i\s difficulties ?an t :i* x seeks to displace4 ,nei!r ( ^! compet+ !ories has yet _h its fundam5tal assumptions confirm$ by any direct evid5ce1 & ! advantage mu/ re/ ) ! "o :i* be/ accords ) ! facts ( col\r vision4 ,9 my judgm5t ! old] ( ! two is to be greatly pref]r$ z a useful work+ hypo!sis4 ,c]ta9 curiosities ( vision ) :i* ,i propose to deal 9 a future *apt] dep5d ^u ! prop]ties ( :at are kn[n z complem5t>y col\rs4 ,two col\rs are said to be complem5t>y to ea* o!r :5 _! comb9ation 9 prop] proportions results 9 ! =mation ( :ite4 ,illu/ration3 ,fig4 #h4--,/5cil ,c>d = ,complem5t>y ,col\rs4 ,if we produce a comp\nd hue by mix+ toge!r ! col\rs ( any portion ( ! spectrum1 & a second comp\nd hue by mix+ ! rema9d] ( ! spectrum1 x mu/ be evid5t t ^! two hues are necess>ily complem5t>y1 = :5 !y are unit$ !y conta9 toge!r all ! elem5ts ( ! entire spectrum1 & !re=e appe> z :ite4 ,?is may be illu/rat$ ) ! aid ( ! col\r-pat* app>atus4 ,place at ,h 7,fig4 #c7 a c>dbo>d /5cil ( ! =m %[n 9 ,fig4 #h1 & focus ^u x a little spectrum1 ! pr9cipal hues ( :i* are indicat$ by ! lett]s ,r ,o ,y ,g ,b ,v 7r$1 orange1 yell[1 gre51 blue1 violet74 ,! two oblong ap]tures 9 ! c>d %\ld be ( exactly ! same height1 & ! c>d s plac$ t "o ap]ture may admit rays ext5d+ f ! r$ end ( ! spectrum to ab\t ! middle ( ! gre51 :ile ! o!r admits rays f ! rema9d] ( ! spectrum4 ,if n[ ! l[] ap]ture be cov]$1 only ! r$1 orange1 yell[1 & "p ( ! gre5 rays w pass ?r\gh ! /5cil1 & ^! be+ comb9$ by ! l5s ,k 7,fig4 #c7 w =m ^u ! scre5 a bright pat*1 ! col\r ( :i* w be yell[4 ,if ! upp] ap]ture be cov]$1 & ! re/ ( ! gre51 toge!r ) ! blue & violet rays1 all[$ to pass ?r\gh ! o!r1 ! col\r ( ! pat* w become blue2 & if bo? ap]tures be uncov]$ at ! same "t1 rays f ! :ole l5g? ( ! spectrum w pass ?r\gh ! /5cil1 & ! pat* w1 ( c\rse1 turn :ite4 ,! yell[ & ! blue :i* 7 comp\nd$ f ! two portions ( ! spectrum are1 !re=e1 9 accordance ) ! def9ition1 complem5t>y col\rs4 ,9 a simil> mann] by divid+ ! spectrum into any two portions :atev]--z1 = example1 by ! complicat$ /5cil %[n 9 ,fig4 #i--we c obta9 an indef9ite numb] ( pairs ( complem5t>y col\rs4 ,illu/ration3 ,fig4 #i4--,/5cil ,c>d = ,complem5t>y ,col\rs4 ,b x is by no means indisp5sable t bo? or ei!r ( a pair ( complem5t>y col\rs %\ld be comp\nd4 ,to prove ?is1 two /rips ( c>d ) n>r[ v]tical op5+s ,a & ,b are prep>$ z %[n 9 ,fig4 #aj4 ,! c>ds are plac$ "o above ! o!r & c be slipp$ 9 a horizontal direction1 s t ! n>r[ op5+s c be br\ght into any desir$ "p ( ! spectrum :i* is indicat$ 9 outl9e by ! dott$ oblong4 ,illu/ration3 ,fig4 #aj4--,slide = mix+ any two ,spectral ,col\rs4 ,br+ ! op5+ ,a ( ! upp] c>d into ! yell[ ( ! spectrum & ! op5+ ,b ( ! l[] c>d into ! blue4 ,! bright pat* =m$ ^u ! scre5 w !n be illum9at$ by simple blue & yell[ rays2 yet x w be :ite--n gre51 z x w\ld be if ,brew/]'s !ory 7 correct4 ,if ^u ! fir/ trial ! :ite %\ld n be absolutely pure1 x c easily be made s by p>tially cov]+ ei!r ,a or ,b--! fir/ if ! :ite is yell[i%1 ! second if x is blui%4 ,simple spectral blue & yell[ are !re=e no less truly complem5t>y col\rs ?an are ! comp\nd hues =m$ :5 ! spectrum is divid$ into two p>ts4 ,x is noticeable1 h[ev]1 t ! :ite light result+ f ! comb9ation ( blue & yell[1 ?\gh x _c be di/+ui%$ by ! eye f ord9>y :ite light1 is yet possess$ ( v diff]5t prop]ties4 ,mo/ col\r$ objects :5 illum9at$ by x h _! hues greatly alt]$2 a piece ( ribbon1 = example1 :i* 9 common light is bright r$1 w appe> :5 held 9 ! blue-yell[ light to be ( a d>k slate col\r1 almo/ black4 ,if ! op5+ ,a is plac$ 9 any "p :atev] ( ! spectrum except ! gre51 x w always be possible1 by mov+ ,b backw>ds or =w>ds1 to f9d "s o!r "p :]e ! col\r is complem5t>y to t at ,a4 ,to gre5 !re is no simple complem5t>y2 a purple is requir$1 :i* is n f\nd 9 ! spectrum1 b may be =m$ by comb9+ small portions ( spectral blue & r$4 ,= /udy+ mixtures ( ?ree simple col\rs1 a ?ird slide may be add$ to ! two %[n 9 ,fig4 #aj4 ,! foll[+ little table gives ! pr9cipal pairs ( complem5t>y col\rs4 ,,table ,,( ,,complem5t>y ,,col\rs4 ,r$ ,gre5i%-blue ,orange ,sky-blue ,yell[ ,blue ,gre5i%-yell[ ,violet ,gre5 ,purple ,,*apt] ,,iii4 ,,"s ,,optical ,,defects ,,( ,,! ,,eye4 ,m ?an "o ref]5ce has be5 made to ! fact t ! s5se ( sight1 ev5 9 its be/ normal condition1 is *>act]is$ by c]ta9 defects & anomalies4 ,"s ( ^! arise directly f causes inh]5t 9 ! design or /ructure ( ! eye itself1 & may be broadly classifi$ z physical2 o!rs are ( psy*ological orig91 & result f ! errone\s int]pretations plac$ by ! m9d ^u ! ph5om5a pres5t$ to x ?r\gh ! m$ium ( ! optic n]ve & ! bra94 ,among ! num]\s physical defects ( ! eye none is m rem>kable ?an ! abs5ce ( means = prop]ly correct+ *romatic ab]ration4 ,?is defect is rem>kable because x appe>s--at lea/ to ^? :o are )\t actual exp]i5ce 9 ! manufacture ( eyes--to be "o :i* might v easily h be5 avoid$4 ,s f> z a m]e !ori/ c judge1 an a*romatic arrangem5t ( l5ses w\ld h be5 j z simple & j z *eap 7if ,i may use ! t]m7 z ! arrangem5t ) :i* we f9d ourselves provid$4 ,x is true t we manage to g ?r\gh life v well ) our uncorrect$ l5ses1 & inde$ x is h>dly possible by ord9>y obs]vation to detect any evid5ce ( ! imp]fection4 ,yet its exi/5ce 9 a gl>+ degree is und\bt$1 & c be readily demon/rat$ by a great v>iety ( me?ods4 ,! conclusion is inevitable t ) a*romatic eyes our vision w\ld be improv$1 b :e!r !re may n possibly exi/ reasons :y su* an improvem5t c\ld only be a*iev$ at a disproportionately high co/ is a "q :i* _c at pres5t be answ]$4 ,)\t go+ into matt]s :i* are dealt ) 9 e elem5t>y text book ( optics or g5]al physics1 x may be desirable to expla9 %ortly :at is meant by ! t]ms *romatic ab]ration1 & a*romatism4 ,illu/ration3 ,fig4 #aa4--,refraction ( mono*romatic ,light by a l5s4 ,let ,l ,l1 ,fig4 #aa1 repres5t 9 section a circul> convex l5s1 & ,p a lum9\s po9t1 :i* is mo/ conv5i5tly suppos$ to be situat$ on ! axis ( ! l5s4 ,imag9e ,p to be surr\nd$ 9 ! fir/ in/ance by a glass %ade :i* transmits only mono*romatic r$ light4 ,s mu* ( ! light f ,p z falls ^u ! l5s w be refract$ to a po9t at ! conjugate focus ,f1 & aft] pass+ ?is po9t w div]ge aga92 ! refract$ light rays w1 9 fact1 =m a d\ble cone1 ( :i* ,f is ! apex4 ,if a :ite scre5 be held at ,f1 !re w be focuss$ ^u x a small cle>ly-def9$ image ( ! lum9\s po9t4 ,if1 h[ev]1 ! scre5 be mov$ ne>] to or fur!r f ! l5s1 x w cut ! cone ( light1 & ! image w !n no long] appe> z a po9t1 b z a circul> r$ disk1 :i* w be l>g] ! great] ! di/ance ( ! scre5 f ,f4 ,su* a disk is kn[n z a diffusion circle4 ,suppose n[ t we sub/itute = ! r$ glass1 surr\nd+ ! s\rce ( light1 a purple "o capable ( transmitt+ n only r$ rays b violet z well4 ,! l5s w cause bo? ! r$ & ! violet rays :i* pass ?r\gh x to conv]ge2 b s9ce ! violet rays are m refrangible--m easily refract$ or b5t aside out ( _! /raight c\rse--?an ! r$1 !re w n[ be two d\ble cones1 z %[n 9 ,fig4 #ab1 :]e ! cont\rs ( ! r$ cones are repres5t$ by solid l9es & ^? ( ! violet by dots4 ,illu/ration3 ,fig4 #ab4--,refraction ( di*romatic ,light4 ,! focus ( ! r$ rays w z be=e be at ,f1 b t ( ! violet w be ne>] to ! l5s1 z at ,h1 & ?is be+ s1 x is evid5t t a well def9$ image ( ! purple s\rce ( light _c possibly be =m$ ^u a scre5 plac$ any:]e beh9d ! l5s4 ,held 9 ! position indicat$ by ! l9e ,c ,c1 :]e x passes ?r\gh ! focus ( ! r$ rays1 ! scre5 cuts "o ( ! cones ( violet light1 & ! image at ,f w appe> to be surr\nd$ by a violet halo4 ,held at ,a ,a1 ! scre5 evid5tly receives an image ) a r$ halo r\nd x4 ,only at ,b ,b1 9 ! plane :]e ! surfaces ( ! r$ & violet cones cut "o ano!r1 w x be possible to obta9 an image )\t a col\r$ bord]2 b "h good def9ition is unatta9able1 = nei!r ! r$ nor ! violet rays are 9 focus1 & ! lum9\s po9t is repres5t$ by a purple disk or diffusion circle ( s5sible diamet]4 ,if rays ( e possible refrangibility are all[$ to fall ^u ! l5s1 z is ! case :5 ! s\rce ( light is n %ield$ by any col\r$ glass1 !re w be =m$ an indef9ite numb] ( pairs ( cones1 ! apices ( :i* w lie along ! /raight l9e jo9+ ,h & ,f4 ,x is cle> t all ^! cones _c possibly int]sect 9 a s+le plane1 & consequ5tly no position c be f\nd :]e ! edge ( ! project$ image is p]fectly free f col\r1 ?\gh at a c]ta9 di/ance f ! l5s1 :]e ! brighte/ con/itu5ts ( ! light--namely1 ! yell[ & gre5--are approximately focuss$1 ! col\r$ bord] is lea/ conspicu\s1 & is ( a purple t9t1 due to ! mixture ( ! r$ & violet rays4 ,= ^! reasons a s+le glass l5s _c1 except ) homog5e\s light1 be made to give a p]fectly di/9ct image ( a lum9\s po9t1 nor ( an illum9at$ object1 ! surface ( :i* may be reg>d$ z an assemblage ( po9ts4 ,su* a l5s1 !re=e1 is nev] employ$ :5 good def9ition is requir$4 ,! confusion result+ f ! unequal refrangibility ( ! diff]5tly col\r$ rays is said to be due to ! *romatic ab]ration ( ! l5s4 ,9 connection ) ?is matt]1 ! hi/ory ( physical optics conta9s an int]e/+ little episode4 ,x occurr$ to ,sir ,isaac ,newton t al?\gh a s+le l5s c\ld nev] be free f *romatic ab]ration1 yet x might be possible to arrange a s-call$ a*romatic comb9ation ( l5ses 9 su* a mann] z to ov]come ! defect & br+ all ! rays issu+ f a po9t1 :atev] _! refrangibility1 to "o focus4 ,exp]im5ts :i* he und]took = ! purpose ( te/+ ! matt] l$ him to =m ! conclusion t su* a result c\ld nev] be atta9$1 ! am\nt ( col\r disp]sion 9 all sub/ances be+1 z he /at$1 always exactly proportional to t ( refraction4 ,= ?is reason he confid5tly ann\nc$ t x 0 useless to attempt ! con/ruction ( a really good refract+ telescope1 & s great 0 ! au?ority atta*+ to 8 "n t = _m ye>s all ef=ts 9 t direction 7 ab&on$4 ,nev]!less f "t to "t c]ta9 philosoph]s v5tur$ to surmise t ,newton might p]haps h be5 mi/ak51 & ! curi\s ?+ is t !y all bas$ _! scepticism ^u :at !y consid]$ ! self-evid5t fact ( ! a*romatism ( ! eye4 ,! sy/em ( l5ses 9 ! eye1 !y argu$1 be+ unque/ionably a*romatic1 :y %\ld n an equally effective comb9ation be con/ruct$ artificially8 ,at l5g?1 m ?an eighty ye>s aft] ,newton _h made & publi%$ 8 fundam5tal exp]im5ts1 x occurr$ to a work+ optician1 ,john ,dollond1 t x might be wor? :ile to repeat !m1 & ^u do+ s he at once f\nd t ,newton 0 wrong 9 8 facts1 ! results z record$ by him be+ 9 direct opposition to ! tru?4 ,) prop] respect = ! memory ( a great man x is usual to speak ( ,newton's obs]vation z a ha/y "o1 b if 9 ^! days a junior sci5ce /ud5t 7 to be guilty ( a simil> lapse1 8 conduct w\ld n impossibly be /igmatis$ z grossly c>eless4 ,hav+ e/abli%$ ,newton's error1 ,dollond f\nd little difficulty 9 con/ruct+ a*romatic l5ses ( v satisfactory quality2 telescopes ( 8 manufacture long enjoy$ ! highe/ reputation1 & ! be/ optical in/rum5ts ( ! pres5t day are ! direct (fspr+ ( 8 inv5tion4 ,^? :o ent]ta9$ ! op9ion t ,newton's conclusion 0 errone\s 7 !re=e 9 ! "r1 b x is rem>kable t ! reason ^u :i* t op9ion re/$ 0 altoge!r invalid1 =1 z ,i h said1 ! l5ses ( ! eye are by no means a*romatic4 ,( ! _m ways 9 :i* ?is c be demon/rat$1 ! foll[+ is "o ( ! mo/ impressive4 ,let a long & n>r[ spectrum ( ! electric light be project$ ^u a :ite scre51 ! prisms & l5ses be+ c>efully arrang$ 9 su* a mann] z to ensure t ! upp] & l[] edges ( ! spectrum are cle>ly def9$ & /rictly p>allel4 ,to an obs]v] /&+ close to ! scre51 ! spectrum w pres5t ! appe>ance ( a bright p>ti-col\r$ rectangle4 ,b view$ f a di/ance ( a few feet ! spectrum w n seem to be rectangul>1 its upp] & l[] edges no long] appe>+ to be p>allel1 b to div]ge1 fan-l1 t[>ds ! blue & violet1 z %[n 9 ,fig4 #ac4 ,?is is because ! violet & "s ( ! blue rays proce$+ f an object at a little di/ance _c by any ef=t be focuss$ ^u ! ret9a4 ,!y are too mu* refract$1 & ! me*anism by :i* ! eye is adju/$ is incompet5t to dim9i% ! convexity ( ! l5ses suffici5tly to enable !m to project a cle> image4 ,e po9t is exp&$ into a lum9\s circle1 :i* is ! l>g] ! m refrangible ! rays1 & x is ! ext5sion ( ^! diffusion circles beyond ! prop] b\nd>ies ( ! image t gives ! appe>ance ( increas$ bread?4 ,x is a simple matt] to c\nt]act ! effects ( undue convexity by means ( a concave l5s4 ,if a normal-ey$ p]son1 to :om ! violet end ( ! spectrum :5 se5 f a di/ance appe>s blurr$ & wid5$1 w look at x ?r\gh suitable glasses adapt$ = %ort sight1 he w at once see x cle>ly def9$ & ( its prop] wid?4 ,illu/ration3 ,fig4 #ac4--,n>r[ ,spectrum z se5 f a di/ance4 ,let a rectangul> pat* ( :ite light hav+ ab\t ! same dim5sions z ! rectangul> spectrum be n[ ?r[n ^u ! scre54 ,! light reflect$ f ! pat* w conta91 z be=e1 all ! v>i\s spectral col\rs1 b !y w be mix$ or sup]pos$1 in/ead ( be+ spread out side by side4 ,! pat* w s5d =?1 among o!rs1 c yell[ & gre5 rays1 :i* ! eye easily focus2 x w also s5d out violet rays1 :i*1 z we h %[n1 _c be focuss$ by ! unassi/$ eye4 ,ow+ to ! exi/5ce ( diffusion circles !re mu/ necess>ily be =m$ ^u ! ret9a a violet image l>g] ?an ! approximately sup]pos$ images due to rays ( bright] col\rs4 ,view$ f a di/ance !re=e ! :ite pat* might be expect$ to exhibit a violet bord]4 ,yet x may be confid5tly ass]t$ t ! obs]v] w n be consci\s ( see+ any su* bord]1 = ?\gh "o actually exi/s1 x is possess$ ( su* comp>atively feeble lum9osity t x is lo/ 9 ! gl>e produc$ by ! bright] rays4 ,x is1 h[ev]1 possible to cut (f ^! bright] rays by int]pos+ betwe5 ! projection lant]n & ! scre5 a comb9ation ( glasses :i* has be5 f\nd by trial ) a spectroscope to transmit only d>k blue & violet light4 ,! rectangle w !n be ( a blue-violet col\r1 & :5 look$ at closely1 w /ill be q cle> & %>ply def9$1 b view$ f a little di/ance x w appe> blurr$ & ( an exagg]at$ size4 ,ano!r & p]haps ev5 bett] way ( demon/rat+ ?is la/ effect is to enclose ! s\rce ( light 7:i* %\ld be a p[]ful "o1 su* z an arc lamp or limelight7 inside a box hav+ a gr\nd-glass w9d[ 9 "o side4 ,:5 ! w9d[ is cov]$ by ! col\r$ glasses its outl9e _c be cle>ly di/+ui%$ unless ! obs]v] is ne>1 b if he uses suitable concave spectacles1 he w be able to see x q di/9ctly1 ev5 f a consid]able di/ance4 ,x is well kn[n t ideas ( di/ance are associat$ ) c]ta9 col\rs4 ,a room gives "o ! impression ( be+ l>g] :5 x is pap]$ or pa9t$ a blue-violet col\r ?an :5 its col\r+ is r$4 ,9 ! =m] case ! walls seem to retire f ! spectator1 9 ! latt] to approa* him4 ,s too a r$ spot ^u a violet gr\nd appe>s to be di/9ctly rais$ above ! surface1 :ile a violet spot ^u a r$ gr\nd appe>s to be depress$4 ,^! ph5om5a are fully expla9$ by ! imp]fect a*romatism ( ! eye4 ,:5 we look at a r$ object1 we h to adju/ ! cry/all9e l5s by means ( ! cili>y muscle 9 exactly ! same way z :5 we look at a ne> object2 9 bo? cases x is necess>y to increase ! convexity ( ! l5s1 & s dim9i% its focal l5g?1 9 ord] to obta9 a cle> image ^u ! ret9a4 ,& aga91 :5 we wi% to see a blue or violet ?+ di/9ctly1 ! cili>y muscle mu/ be relax$ & ! convexity ( ! l5s z f> z possible dim9i%$1 j z if ! gaze 7 direct$ to ! horizon4 ,we are accu/om$ to e/imate ! di/ances ( ?+s l>gely by ! muscul> ef=t requir$ to focus _! images1 & ?us x happ5s t ! col\r r$ comes to be associat$ 9 our m9ds ) ne>ness1 & violet ) remot5ess4 ,^! psy*ological effects are p]fectly well m>k$ ev5 ) ! impure col\rs met ) 9 ord9>y life1 b !y are naturally m evid5t :5 ! col\rs obs]v$ are pure1 l ^? ( ! spectrum4 ,a beautiful example is t pres5t$ by ! pair ( %ort bright spectra =m$ ^u ! scre5 :5 a d\ble slit is us$ %ap$ l ! lett] ,v4 ,! gorge\sly col\r$ ,v seems to /& out 9 /rong relief l a pair ( incl9$ bo>ds1 ! ne>] edges be+ r$1 ! f>!r ones violet4 7,see ,fig4 #ad47 ,illu/ration3 ,fig4 #ad4--,spectrum =m$ ) ,v-%ap$ ,slit4 ,9 _m o!r ways1 & ) little or no app>atus1 any "o may easily conv9ce himself t ! diff]5t con/itu5ts ( :ite light are n equally refract$ by ! l5ses ( ! eye4 ,look1 = in/ance1 at ! inc&esc5t filam5t ( an electric lamp ?r\gh a piece#g ( common d>k blue cobalt glass1 :i* has ! prop]ty ( ob/ruct+ ! col\r$ rays correspond+ to ! middle ( ! spectrum1 :ile transmitt+ ! r$ & ! blue4 ,se5 f a di/ance ( only a few in*es1 ! filam5t appe>s to be pale blue ) a bright r$ bord]1 ! blue rays be+ p]fectly focuss$1 :ile ! r$ =m diffusion circles4 ,move "s six or eight feet away & look aga92 ! col\rs w n[ be rev]s$1 ! filam5t appe>+ r$ & ! bord] blue-violet4 ,f a /ill great] di/ance--ab\t fifte5 or tw5ty feet--! :ole lamp-bulb w seem to be fill$ ) a blue-violet gl[1 due to l>ge diffusion circles1 :ile ! r$ image ( ! filam5t may be ev5 m cle>ly def9$ ?an be=e4 ,no d\bt x is p>tly ow+ to ! non-a*romatism ( ! eye t di/ant arc lights always appe> to h a yell[i% hue1 ev5 :5 ! air is q cle>2 a consid]able proportion ( _! blue & violet compon5ts mu/ necess>ily be lo/ by ext5sive diffusion4#h ,aga91 look at a sunlit l&scape or a pr9t$ wall po/] ?r\gh a comb9ation ( col\r$ glasses :i* w transmit only ! violet end ( ! spectrum4 ,y w f9d y\rself = ! "t t]ribly %ort-sight$1 ev]y?+ appe>+ blurr$ & indi/9ct4 ,b if y resort to ! usual corrective = myopia1 & put on a pair ( concave spectacles1 y\r normal vision w be re/or$2 trees & h\ses w be se5 z cle>ly z ! feebl5ess ( ! light transmitt$ by ! col\r$ glasses w p]mit1 & ! lett]s ( ! po/] w become easily legible4 ,n[1 ( c\rse1 ! int]position ( col\r$ glasses does n actually give rise to ^! blurr$ images2 x m]ely enables "o to detect _! exi/5ce4 ,"u ord9>y conditions !y always accompany ! cle>] images produc$ by ! m lum9\s rays1 & _! pres5ce _c fail to ex]t a detrim5tal effect ^u ! g5]al def9ition4 ,su* blurs mu/ at lea/ t5d to fog ! d>k] portions ( ! focuss$ picture1 & ?\gh we are n di/9ctly consci\s ( _! exi/5ce1 x is c]ta9 t if !y 7 annull$ ! acut5ess ( our vision w\ld be improv$4 ,! diffusion circles produc$ by ! r$ rays1 :5 ! eye is accommodat$ 7z x commonly is7 = ! yell[ & gre51 are less conspicu\s ?an ^? due to ! mo/ refrangible rays4 ,yet ,i f9d x impossible to focus a r$ object1 su* z ! filam5t ( an electric lamp scre5$ by a prop]ly select$ deep r$ glass1 :5 plac$ at ! ord9>y di/ance ( di/9ct vision--"s n9e or t5 in*es f ! eye--)\t ! aid ( a convex l5s4 ,9 ?is case "o is n too %ort-sight$ b too long-sight$ to see ! object di/9ctly2 9 o!r words1 ! l5ses ( ! eye _c refract ! r$ rays suffici5tly to produce well-def9$ images ^u ! ret9a1 & ! refraction has to be increas$ by artificial means4 ,?\gh1 z ,i h said1 x is difficult1 or ev5 impossible to detect any trace ( a col\r$ bord] :5 look+ at a bright object = :i* ! eye is accommodat$1 x is q easy to br+ su* bord]s into prom95ce if ! object is at a di/ance a little too great or too small = di/9ct vision4 ,a v rem>kable device = ! purpose is "o due to von ,bezold4 ,?is may be illu/rat$ by us+ a non-a*romatic glass l5s1 su* z a common magnify+ glass1 to project a transp>5cy or lant]n-slide ^u :i* is pa9t$ a t>get-l design1 consi/+ ( a s]ies ( circul> black b&s surr\nd+ a circul> black spot4#i 7,see ,fig4 #ae47 ,illu/ration3 ,fig4 #ae4--;,bezold's ,diagram4 ,suppose ! glass l5s to repres5t ! l5ses ( a gigantic eye 79 a def9ite condition ( accommodation7 & ! scre5 ! ret9a4 ,! imag9>y eye is look+ at ! design on ! lant]n-slide1 & :5 ?is is at ! di/ance ( mo/ di/9ct vision a fairly well def9$ image ( ! t>get is =m$ ^u ! ret9al scre54 ,n[ gradually move ! lant]n slide t[>ds ! l5s 7or ! l5s t[>ds ! slide71 ?us br++ x too ne> = di/9ct vision4 ,?is has ! effect ( enl>g+ ! diffusion circles =m$ by ! less refrangible rays correspond+ to ! r$ end ( ! spectrum1 & at ! same "t ( dim9i%+ ^? =m$ by ! m refrangible rays correspond+ to ! violet end4 ,! fir/ result is t ! circul> d>k b&s become r$di% br[n1 & ! spaces betwe5 !m blui%4 ,z ! di/ance betwe5 ! l5s & ! slide is /ill fur!r dim9i%$1 ! t9ts become m v>i$ & brilliant1 until at la/ !re appe>s a beautiful s]ies ( col\r$ r+s ar\nd a bright r$ c5tral spot4 ,^! effects are n produc$ :5 ! l5s employ$ is an a*romatic "o2 ) su* a l5s ! diffusion circles are all enl>g$ or dim9i%$ toge!r1 & a to-&-fro movem5t ( ! lant]n slide 7or ( ! l5s7 m]ely affects ! def9ition ( ! image )\t caus+ any p]ceptible disp]sion ( col\r4 ,n[ x is notewor?y t ! *romatic ph5om5a exhibit$ ) ! uncorrect$ glass l5s are q well %[n by ! l5ses ( ! eye4 ,x is only necess>y to hold ! lant]n-slide be=e a bright backgr\nd & gradually br+ x s close to ! eye t ! design _c be se5 di/9ctly4 ,! black b&s w !n appe> to turn br[n1 ! :ite ones blue1 & ! c5tral spot bright r$4 ,! pr9t$ diagram 7,fig4 #ae7 w itself %[ ! col\rs if x is held at a di/ance ( f\r to five in*es f "o eye 9 a good light4 ,"o m exp]im5t may be ref]r$ to4 ,look ) "o eye at a well-light$ page ( pr9t1 & ) a /rip ( br[n pap]1 held q ne> ! eye1 cov] ab\t half ! pupil4 ,! black lett]s w n[ appe> to be bord]$ ) col\r--blue t[>ds ! app>5t edge ( ! br[n pap]1 orange on ! opposite side4 ,if ! lett]s are :ite on a black gr\nd1 z sometimes happ5s 9 ! case ( adv]tisem5ts1 ! col\rs w be int]*ang$4 ,! cause ( ! col\r$ bord]s w be readily und]/ood f an inspection ( ! diagram ,fig4 #ab2 b x mu/ be rememb]$ t ! images on ! ret9a are inv]t$4 ,?us x is prov$ beyond all "q t ! l5ses ( ! eye d n =m an a*romatic comb9ation4 ,ano!r peculi>ity by :i* ! eye is affect$1 & :i* does n occur 9 optical in/rum5ts1 is t kn[n z a/igmatism4 ,! surface ( ! cornea1 :i*1 ) ! aque\s hum\r1 =ms ! out] l5s1 is n (t5 p]fectly sph]ical2 g5]ally x is %ap$ some?+ l ! b[l ( a spoon1 ! curvature be+ great] v]tically ?an horizontally4 ,rays issu+ f a lum9\s po9t d n1 aft] refraction by su* a l5s1 cross at a s+le focus1 b along two %ort /raight l9es1 ! "o horizontal ! o!r v]tical1 :i* are at diff]5t di/ances f ! l5s2 ?us a di/9ct image ( a small po9t _c any:]e be produc$4 ,illu/ration3 ,fig4 #af4--;,effect ( ,a/igmatism4 ,a v curi\s result foll[s f ?is de=mity4 ,if two /raight l9es are drawn at "r angles to ea* o!r1 z 9 ,fig4 #af1 x is impossible to see bo? ( !m q cle>ly at ! same "t4 ,:5 ! pap] is held at a c]ta9 %ort di/ance f ! eye--ab\t eight or n9e in*es--! horizontal l9e appe>s black & well def9$1 :ile ! o!r is r grey & indi/9ct2 at a great] di/ance ! upright l9e seems to be ! black]4 ,! effect is v well %[n by ! diagram1 ,fig4 #ag4 ,to mo/ p]sons ! l9es occupy+ ! middle portion w appe> ei!r mu* black] or mu* light] ?an ^? at ! two ends1 ?\gh 9 fact !y are exactly alike4 ,:5 ?is =m ( a/igmatism is excessive1 x may be correct$ by ! use ( spectacles fitt$ ) cyl9drical l5ses4 ,illu/ration3 ,fig4 #ag4--;,effect ( ,a/igmatism4 ,b !re is a diff]5t k9d ( a/igmatism--irregul> a/igmatism x is call$--to :i* e "o is m or less a victim1 & :i* _c be reliev$ by any artificial appliances4 ,=tunately x does n (t5 cause mu* practical inconv5i5ce4 ,irregul> a/igmatism is commonly demon/rat$ 9 ! foll[+ mann]4 ,) ! po9t ( a f9e ne$le1 prick a v small hole 9 a %eet ( t9foil4 ,hold up ! t9foil to ! light & look at ! hole ) "o eye1 ! o!r be+ clos$4 ,ev5 at ! di/ance ( mo/ di/9ct vision--t5 in*es or !reab\ts1--!re w probably be a ragg$ appe>ance ab\t ! hole1 z if x 7 n p]fectly r\nd4 ,b if y br+ ! t9foil an in* or two ne>] to ! eye1 ! hole w n seem to be ev5 approximately circul>2 x w assume ! =m ( a little /> ) five or m di/9ct rays4 ,! configuration ( ! /> is n g5]ally ! same = ! "r eye z = ! left2 ! rays may diff] 9 numb] & 9 relative magnitude1 & may be incl9$ at diff]5t angles to ! v]tical4 ,fig4 #ah %[s ! />s z !y appe> to my two eyes1 :5 ! illum9ation is r /rong4 ,illu/ration3 ,fig4 #ah4--,/>-l ,images ( lum9\s ,po9t4 ,if sev]al holes are prick$ 9 ! t9foil1 ea* w ( c\rse orig9ate a sep>ate />1 & all ! />s z se5 by ! same eye w appe> to be figur$ ^u ! same model1 ?\gh "s may be l>g] or bright] ?an o!rs4 ,illu/ration3 ,fig4 #ai4--,sutures ( cry/all9e ,l5s4 ,!re c be no d\bt t ! /ellate =m obs]v$ 9 ^! exp]im5ts1 z well z t ( ! />s ( heav5 !mselves 7:i* ) p]fect vision w\ld be se5 simply z lum9\s po9ts71 is a consequ5ce ( ! s+ul> /ructure ( ! cry/all9e l5s ( ! eye4 ,?is does n consi/ ( "o uni=m homog5e\s mass l a glass l5s1 b ( a numb] ( sep>ate portions piec$ toge!r radially1 z indicat$ diagrammatically 9 ,fig4 #ai4 ,9 ! eye ( a newly-born *ild !re are ?ree su* portions1 & ! radial junctions on "o side ( ! l5s are n opposite to ^? on ! o!r1 b are int]m$iate4 ,9 ! figure ! junctions at ! front ( ! l5s are repres5t$ by cont9u\s l9es & ^? at ! back by dots4 ,! numb] ( sutures f\nd 9 ! adult l5s is g5]ally great] ?an six4 ,b :ile x is c]ta9 t ^! radial sutures are 9 "s way closely connect$ ) ! lum9\s rays :i* appe> to proce$ f a bright po9t1 x mu/ be confess$ t no adequate explanation has yet be5 giv5 ( ! precise mann] 9 :i* ! ph5om5on is br\ght ab\t4 ,oph?almologi/s seem to h be5 cont5t$ ) vague /atem5ts ab\t irregul> refraction1 b :at k9d ( irregul>ity w\ld suffici5tly acc\nt = all ! facts ( obs]vation has nev]1 s f> z ,i "k1 be5 exactly det]m9$4 ,! problem c h>dly be v difficult ( solution1 & w\ld1 no d\bt1 readily yield to ! jo9t ef=ts ( a physici/ & a physiologi/4 ,! ph5om5a ( irregul> a/igmatism z exhibit$ by a normal eye are exce$+ly curi\s1 & p]haps ,i may be all[$ to ref] briefly to "o or two exp]im5ts :i* ,i h myself made on ! subject4#aj ,illu/ration3 ,fig4 #bj4--,multiple ,images ( a lum9\s ,po9t4 ,light f an enclos$ electric lamp ( tw5ty-five c&le p[] 0 admitt$ ?r\gh a circul> ap]ture ab\t #a_/ab-;in* 7#bmm47 9 diamet] p]=at$ 9 a brass plate2 a %eet ( gr\nd glass & ano!r ( ruby-r$ glass 7 plac$ beh9d ! ap]ture4 ,:5 ! little disk ( mono*romatic light ?us =m$ 0 look$ at ?r\gh a concave l5s ( elev5 in*es focal l5g? f a suitable di/ance--ne>ly two feet 9 my own case--x appe>$ z sev5 bright r\nd spots ^u a less lum9\s gr\nd4 ,! appe>ance is repres5t$ 9 a some:at idealis$ =m 9 ,fig4 #bj2 b ! spots 7 n q s di/9ct nor s regul>ly dispos$ z !re %[n1 nei!r 0 _! configuration exactly ! same = ! "r eye z = ! left4 ,on gradually increas+ ! di/ance ea* circumf]5tial spot became at fir/ elongat$ radially & aft]w>ds split up into two circul> ones2 at ! same "t new spots 7 develop$ ^u ! lum9\s gr\nd1 ! approximate symmetry ( ! figure be+ /ill reta9$4 ,fig4 #ba repres5ts a c]ta9 /age 9 ?is process ( expansion4 ,! appe>ance 0 happily lik5$ by an obs]v] :o repeat$ ! exp]im5t to t ( a l>ge unripe blackb]ry4 ,z ! di/ance 0 /ill fur!r increas$1 ! spots cont9u$ to multiply1 ultimately becom+ v num]\s2 _! arrangem5t h[ev] soon became mu* less regul>1 & ! def9ition ( mo/ ( !m less di/9ct4 ,at ab\t tw5ty feet !re 0 se5 a lum9\s pat*1 r\ghly circul> 9 outl9e1 & cov]$ ) irregul> speckles2 sup]pos$ ^u ?is 7 /r+s ( bright1 p>tially ov]lapp+ spots1 correspond+ app>5tly to ! sutures ( ! cry/all9e l5s4 ,illu/ration3 ,fig4 #ba4--;,increas$ numb] ( ,images4 ,:5 ! hole 0 look$ at f a mod]ate di/ance ?r\gh a n>r[ slit 7ab\t #a_/cj in* wide7 int]pos$ betwe5 ! eye & ! l5s1 !re 0 se5 only a s+le r[ ( circul> spots1 :i* 7 arrang$ s9u\sly1 z %[n 9 ,fig4 #bb4 ,a slight movem5t ( ! slit 9 ! direction p]p5dicul> to its l5g? produc$ a wave-l motion ( ! circles1 sugge/ive1 z po9t$ out by ! excell5t obs]v] be=e ref]r$ to ( ! wriggl+ ( a cat]pill>4 ,illu/ration3 ,fig4 #bb4--,multiple ,images se5 ?r\gh a ,slit4 ,by suffici5tly increas+ ! di/ance betwe5 ! s\rce ( light & ! eye1 z _m z tw5ty-f\r or tw5ty-five bright spots might be made to appe> 9 ! r[1 b !y c\ld n be c\nt$ ) any great c]ta9ty4 ,at a /ill long] di/ance or ) a l5s ( %ort] focus 7convex or concave7 !y became less di/9ct1 & f9ally seem$ to be resolv$ into a multitude ( small blurr$ images--probably sev]al hundr$s--:i* 7 sep>at$ f "o ano!r by hazy d>k l9es4 ,illu/ration3 ,fig4 #bc4--;,images ( an electric lamp ,filam5t4 ,i ?\ght t ! obs]vations might be r5d]$ easi] if ! s\rce ( light _h a m di/9ctive & conspicu\s =m ?an t ( a simple circle4 ,"s exp]im5ts 7 !re=e made ) semi-circul> & triangul> holes1 & ^! 7 9 "s respects pref]able2 b f> bett] results 7 aft]w>ds obta9$ by us+ z a s\rce ( light ! horse-%oe %ap$ filam5t ( an electric lamp1 scre5$ by a col\r$ glass4 ,:5 su* a lamp 0 look$ at ?r\gh a l5s1 concave or convex1 ( ab\t six in*es focus1 f a di/ance ( a few feet1 ! r\ghly oval pat* ( lum9osity =m$ ^u ! ret9a1 in/ead ( be+ a m]e ill-def9$ blur1 su* z w\ld be produc$ if ! transp>5t m$ia ( ! eye 7 compos$ ( homog5e\s sub/ances l glass or wat]1 appe>$ to be made up ( a cr[d ( sep>ate images ( ! filam5t1 "s be+ bright] ?an o!rs1 z is %[n 9 ! diagram ,fig4 #bc4 ,illu/ration3 ,fig4 #bd;,a4--,images ) horizontal ,slit4 ,illu/ration3 ,fig4 #bd;,b4--,images ) v]tical ,slit4 ,if a spectroscope slit 0 int]pos$ betwe5 ! eye & ! l5s1 & its wid? suitably adju/$1 only a s+le r[ ( filam5ts 0 obs]v$1 ! appe>ances ) ! slit 9 horizontal1 v]tical1 & int]m$iate positions be+ z repres5t$ 9 ,fig4 #bd1 ,a1 ,b1 ,c4 ,z be=e1 x 0 f\nd possible by gradually retir+ f ! lamp to br+ ! numb] ( images up to ab\t tw5ty-five1 b att5tive exam9ation %[$ t mo/ ( ^! really consi/$ ( clu/]s1 ea* compos$ ( p]haps fifte5 or tw5ty confus$ images ( ! filam5t4 ,a /rong] l5s /ill fur!r sep>at$ ! con/itu5ts ( ! clu/]s1 exhibit+ a total numb] ( indi/9ctly se5 images :i* 0 e/imat$ to am\nt to ne>ly five hundr$4 ,assum+ ! diamet] ( ! pupil ( ! eye to be "o-fif? ( an in*1 ^! obs]vations seem to indicate z a cause ( ! ph5om5on "s fairly regul> anatomical /ructure1 situat$ 9 or ne> ! cry/all9e l5s & compos$ ( elem5ts measur+ ab\t #a_/bjjj in* 9 l5g? or bread?4 ,:e!r ! /ructure :i* gives rise to ^! multiple images is to be f\nd 9 ! fibres ( ! cry/all9e l5s itself1 or 9 ! membranes :i* cov] x1 is a "q ^u :i* ,i w n v5ture an op9ion4 ,illu/ration3 ,fig4 #bd;,c4--,images ) oblique ,slit4 ,x is inde$ wond]ful t an organ affect$ by peculi>ities ( :i* ^? t h be5 ref]r$ to are m]ely specim5s1 %\ld give su* well-def9$ pictures z x does :5 accommodat$ = ! objects look$ at4 ,,*apt] ,,iv4 ,,"s ,,optical ,,illusions4 ,optical illusions g5]ally result f ! m9d's faulty int]pretation ( ph5om5a pres5t$ to x ?r\gh ! m$ium ( ! visual organs4 ,!y are ( _m diff]5t k9ds1 b a l>ge class1 :i* at fir/ sight may seem to h little or no?+ 9 common1 arise1 ,i believe1 f a s+le cause1 namely1 ! inability ( ! m9d to =m & adh]e to a def9ite scale or /&>d ( measurem5t4 ,9 specify+ quantities & qualities by physical me?ods1 ! /&>ds ( ref]5ce t we employ are inv>iable4 ,we may1 = example1 measure a l5g? by ref]5ce to a rule1 an int]val ( "t by a clock1 a mass or weight by comp>ison ) /&>dis$ lumps ( metal1 & 9 all su* cases--provid$ t our in/rum5ts are good ones & skilfully us$--we h e confid5ce 9 ! con/ancy & uni=mity ( our results4 ,b two l5g?s1 :i* :5 te/$ ) ! same foot rule are f\nd to be exactly equal1 are n necess>ily equal 9 ! e/imate =m$ ( !m by ! m9d4 ,look1 = in/ance1 at ! two l9es 9 ,fig4 #be4 ,accord+ to ! foot rule ea* ( !m is j "o in* 9 l5g?1 b ! m9d unhesitat+ly pron\nces ! upright "o to be consid]ably long] ?an ! o!r2 ! /&>d :i* x applies is n1 l a physical "o1 id5tical 9 ! two cases4 ,_m o!r examples might be cit$ illu/rative ( ! g5]al unc]ta9ty ( m5tal e/imates4 ,illu/ration3 ,fig4 #be4--;,illusion ( ,l5g?4 ,! v>iation ( ! vague m5tal /&>d :i* we unconsci\sly employ seems to be gov]n$ by a law ( v wide if n univ]sal application4 ,?\gh ?is law is 9 itself simple & intelligible en\gh1 x _c easily be =mulat$ 9 t]ms ( adequate g5]ality4 ,! be/ result ( my ef=ts is ! foll[+ unwieldy /atem5t3--,! m5tal /&>d :i* is appli$ 9 ! e/imation ( a quality or a condition t5ds to assimilate itself1 z reg>ds ! quality or condition 9 "q1 to ! object or o!r entity "u comp>ison ( :i* ! same 7quality or condition7 is an attribute4 ,9 pla9] b less precise language1 !re is a disposition to m9imise extremes ( :atev] k9d2 to und]e/imate any deviation f a mean or av]age /ate ( ?+s1 & consequ5tly to v>y our conception ( ! mean or /&>d condition 9 su* a mann] t ! deviation f x :i* is pres5t$ to our notice 9 any p>ticul> in/ance may seem to be small r ?an l>ge4 ,?us1 :5 we look at a ?+ :i* impresses u z be+ long or tall1 ! m5tal /&>d ( l5g? is at once increas$4 ,x is z if1 9 mak+ a physical measurem5t1 our foot rule 7 automatically to add "s in*es to its l5g?1 :ile /ill suppos$ to repres5t a /&>d foot3 cle>ly any?+ measur$ by means ( ! augm5t$ rule w\ld seem to conta9 a few] numb] ( feet1 &1 !re=e1 to be %ort] ?an if ! rule _h n und]gone a *ange4 ,x is n an uncommon ?+ = p visit+ ,switz]l& = ! fir/ "t to express disappo9tm5t at ! app>5tly small height ( ! m\nta9s4 ,a m\nta9 ( #aj1jjj feet c]ta9ly does n seem to be tw5ty times z l(ty z a hill ( #ejj4 ,! fact is t a diff]5t scale ( measurem5t is appli$ 9 ! two cases2 ?\gh ! obs]v] is unaw>e ( x1 ! m\nta9 is e/imat$ 9 t]ms ( a l>g] unit ?an ! hill4 ,illu/ration3 ,fig4 #bf4--;,illusion ( ,l5g?4 ,if we m5tally comp>e two adjac5t ?+s ( unequal l5g?1 su* z ! two /raight l9es 9 ,fig4 #bf1 !re is a t5d5cy to reg>d ! %ort] "o z long] ?an x w\ld appe> if se5 alone1 & ! long] "o z %ort]4 ,! l[] ( ! two l9es 9 ! figure is j twice z long z ! o!r1 b x does n look s2 ea* is reg>d$ z diff]+ less ?an x really does f an imag9>y l9e ( int]m$iate l5g?4 ,illu/ration3 ,fig4 #bg4--;,illusion ( ,l5g?4 ,two div]g5tly oblique l9es atta*$ to ! ends ( a /raight l9e z at ,a1 ,fig4 #bg1 sugge/ to ! m9d ! idea ( l5g?s great] ?an t ( ! /raight l9e itself2 ! latt]1 be+ ?\ght ( z comp>atively small1 is !re=e e/imat$ 9 t]ms ( a small] unit ?an w\ld be employ$ if ! atta*m5ts 7 abs5t1 & consequ5tly appe>s long]4 ,if1 on ! o!r h&1 ! atta*m5ts are made conv]g5t1 z at ,b1 %ort] l5g?s are sugge/$2 ! l5g? ( ! giv5 l9e is reg>d$ z exce$+ an av]age or mean2 ! /&>d appli$ 9 e/imat+ x is accord+ly increas$1 & ! l9e is made to seem unduly %ort4 ,9 spite ( appe>ances to ! contr>y1 ! two l9es ,a & ,b are actually ( ! same l5g?4 ,by duplicat+ ! atta*$ l9es1 z %[n 9 ,fig4 #bh1 _! mislead+ effect becomes int5sifi$4 ,"h we h a well-kn[n illusion ( :i* sev]al explanations h be5 propos$4 ,! fallacy is1 ,i ?9k1 suffici5tly acc\nt$ = by v>iation ( ! m5tal /&>d1 9 accordance ) ! law to :i* ,i h call$ att5tion4 ,illu/ration3 ,fig4 #bh4--;,illusion ( ,l5g?4 ,a numb] ( o!r p>adoxical effects may be ref]r$ to ! op]ation ( ! same law4 ,fig4 #bi %[s a curi\s specim54 ,at ea* end ( ! diagram is a %ort upright l9e2 exactly 9 ! middle is ano!r2 betwe5 ! middle & ! left h& end are ins]t$ sev]al m l9es1 ! space to ! "r ( ! middle be+ left blank4 ,any "o look+ casually at ! diagram w\ld be incl9$ to suppose t x 0 n equally divid$ by :at purports to be ! middle l9e1 ! left h& portion appe>+ s5sibly long] ?an ! o!r4 ,illu/ration3 ,fig4 #bi4--;,illusion ( ,di/ance4 ,x is n difficult to indicate ! s\rce ( ! illusion4 ,:5 we look at ! left h& portion we att5d to ! small subdivisions1 & ! m5tal unit becomes correspond+ly small2 :ile 9 ! e/imation ( ! portion :i* is n subdivid$ a l>g] unit is appli$4 ,z "o m example ,i may ref] to a famili> trap = ! unw>y4 ,ask a p]son to m>k ^u ! wall ( a room ! height above ! floor :i* he ?9ks w correspond to t ( a g5tleman's tall hat4 ,unless he has be5 beguil$ on a =m] occasion1 he w c]ta9ly place ! m>k sev]al in*es too high4 ,obvi\sly ! height ( a hat is unconsci\sly e/imat$ 9 t]ms ( a small] /&>d ?an t ( a room4 ,! illusion pres5t$ by ! horizontal & v]tical l9es 9 ,fig4 #be 7p4 #acb7 dep5ds1 ?\gh a little less directly1 ^u a simil> cause4 ,we habitually apply a l>g] /&>d 9 ! e/imation ( horizontal ?an ( v]tical di/ances1 because ! horizontal magnitudes to :i* we are accu/om$ are ^u ! :ole v mu* great] ?an ! v]tical ones4 ,! heights ( h\ses1 t[]s1 spires1 trees1 or ev5 m\nta9s are insignificant 9 comp>ison ) ! horizontal ext5sion ( ! e>?'s surface1 & ( _m ?+s ^u x1 to :i* our notice is con/antly direct$4 ,= ?is reason1 we h come to associate horizontality ) great] ext5sion & v]ticality ) less1 &1 9 con=mity ) our law1 a giv5 di/ance appe>s long] :5 reckon$ v]tically ?an :5 reckon$ horizontally4 ,h5ce ! illusion 9 ,fig4 #be4 ,b x is n only 9 reg>d to l5g?s & di/ances t ! law 9 "q holds good2 9 mo/1 if n all cases 9 :i* a psy*o-optical e/imate is possible1 ! m5tal /&>d is un/able & t5ds to assimilate itself1 z reg>ds ! quality or condition to be e/imat$1 to ! entity 9 :i* ! same is manife/$4 ,?is is true1 = example1 9 judg+ ( an angle ( incl9ation or slope2 ( a motion 9 space2 ( lum9\s int5sity1 or ( ! purity ( a col\r4 ,e cycli/ kn[s h[ difficult x is to =m a correct judgm5t ( ! /eepness ( a hill by m]ely look+ at x4 ,n only may a slope seem to be great] or less ?an x really is1 b "u c]ta9 circum/ances a dead level sometimes appe>s z an upw>d or d[nw>d incl9ation1 :ile a g5tle asc5t may ev5 be mi/ak5 = a desc5t1 & vice v]sa4 ,we usually specify a slope by its incl9ation to a level plane :i* is p>allel to ! plane ( ! horizon1 or at "r angles to ! direction ( gravity4 ,at any giv5 spot ! level is1 physically consid]$1 def9ite & unalt]able4 ,9 =m+ a m5tal judgm5t ( an incl9ation1 we employ z our /&>d ( ref]5ce an imag9>y plane :i* is int5d$ to be id5tical ) ! physical level4 ,b our m5tal plane is n absolutely /able2 :5 we ref] a slope to x1 we unconsci\sly give ! m5tal plane a slight tilt1 t5d+ to make x p>allel ) ! slope4 ,h5ce ! incl9ation ( a simple slope1 :5 mislead+ complications are abs5t1 is always und]e/imat$4 ,illu/ration3 ,fig4 #cj4--;,illusion ( ,incl9ation4 ,?is may be illu/rat$ by ! diagram ,fig4 #cj4 ,if ,a ,b repres5ts a truly horizontal l9e1 ! slope ( ! oblique l9e ,c ,d is correctly specifi$ by ! angle ,c ,o ,a4 ,b if we h no in/rum5t at h& to fix ! level = u1 we %all infallibly imag9e x to be 9 "s su* position z t indicat$ 79 an exagg]at$ degree7 by ! dott$ l9e ,e ,f1 :ile ! true level ,a ,b w appe> to slope oppositely to ,c ,d4 ,?is class ( illusion is rem>kably well demon/rat$ by ,zlln]'s l9es1 ,fig4 #ca2 ! two ?ick l9es :i* appe> to div]ge f left to "r1 are 9 tru? /rictly p>allel4 ,illu/ration3 ,fig4 #ca4--,zlln]'s ,l9es4 ,i ne$ n discuss 9 fur!r detail ! v>i\s illusions to :i* a cycli/ is subject$ :5 slopes ( diff]5t incl9ations succe$ "o ano!r3 !y all foll[ simply f ! same g5]al pr9ciple4 ,a ?+ is said to be 9 motion :5 x is *ang+ its position relatively to ! e>?1 :i* = all practical purposes may be reg>d$ z motionless4 ,! /ate1 z reg>ds motion1 ( ! e>? & any?+ rigidly atta*$ to x1 !re=e con/itutes ! physical z]o or /&>d to :i* ! motion ( ev]y?+ t]re/rial is ref]r$4 ,b ! correspond+ m5tal /&>d1 especially :5 x _c easily be *eck$ by comp>ison ) "s /ation>y object1 is liable to deviate f ! physical "o2 x t5ds 9 fact to move 9 ! same direction z ! mov+ body :i* is "u obs]vation1 & ! app>5t spe$ ( ! body is consequ5tly r less ?an x %\ld be4 ,! influ5ce ex]t$ ^u ! judgm5t sometimes ev5 p]si/s = an appreciable p]iod aft] ! excit+ cause has ceas$ to be op]ative1 z :5 ! mov+ body is lo/ sight ( or has sudd5ly come to re/2 9 su* cases fix$ objects1 be+ comp>$ ) ! delusive m5tal /&>d1 appe> = a few seconds to be mov+ 9 ! opposite direction4 ,i h devis$ a lant]n slide 7,fig4 #cb7 by ! aid ( :i* ?is ph5om5on may be r5d]$ v evid5t4 ,9 a squ>e plate ( metal is cut a v]tical slot1 :i* is %ad$ 9 ! figure2 beh9d ! plate is an opaque disk1 :i*1 by means ( suitable me*anism1 c be made to rotate ab\t its c5tre4 ,! disk has a spiral op5+ cut 9 x ( ! same wid? z ! slot1 z indicat$ by ! dott$ l9e4 ,! slide is plac$ 9 an optical lant]n1 & ! light pass+ ?r\gh ! ap]ture =m$ :]e ! slot is cross$ by ! spiral op5+1 produces a small bright pat* ^u a :ite scre5 hung at a suitable di/ance f ! lant]n4 ,illu/ration3 ,fig4 #cb4--,slide = %[+ ,illusions ( ,motion4 ,:5 ! disk is turn$ 9 ! direction indicat$ by ! arr[1 ! bright pat* moves upw>ds & ultimately disappe>s2 b at ! mom5t ( its disappe>ance a fre% pat* />ts f bel[1 :i* also moves 9 ! upw>d direction2 ?us !re is =m$ ^u ! scre5 a cont9u\s succession ( asc5d+ bright pat*es4 ,aft] ^! h be5 obs]v$ = ab\t a qu>t] ( a m9ute1 ! disk is sudd5ly /opp$1 & ! p]si/5ce ( ! fallaci\s m5tal /&>d is at once demon/rat$4 ,= ! bright pat* does n appe> to be at re/1 z x actually is1 b to creep /eadily d[nw>ds1 cont9u+ to d s m & m sl[ly = p]haps z long z t5 seconds4 ,! upw>d motion ( ! bright pat*es _h l$ ! obs]v] to assume a sl[] upw>d motion z ! z]o1 or /&>d ( no motion1 & ref]5ce ( ! really /ation>y pat* to ?is physically false /&>d induces ! illusion t ! pat* is desc5d+4 ,?is exp]im5t is mo/ successful :5 ! bright pat*es are project$ ^u ! middle ( a l>ge scre54 ,! disk %\ld turn ab\t ?ree times 9 a second1 & ! room %\ld be feebly illum9at$1 b n q d>k4 ,illu/ration3 ,fig4 #cc4--;,illusions ( ,motion4 ,a v rem>kable illusion :i* no d\bt dep5ds ^u ! same pr9ciple z ! la/1 ?\gh its =m is entirely diff]5t1 is t to :i* ! diagram ,fig4 #cc relates4 ,s f> z ,i am aw>e1 x has n be=e be5 notic$4 ,two int]sect+ /raight l9es1 ! "o upright & ! o!r slop+1 z %[n 9 ! figure1 are drawn ^u a c>d4 ,! c>d is to be held v]tically be=e ! eyes at ! di/ance ( mo/ di/9ct vision1 & wav$ up & d[n ?r\gh a di/ance ( a few in*es4 ,! oblique l9e w !n appe> to oscillate transv]sely1 z if x 7 n rigidly atta*$ to ! c>d4 ,?is is ! result ( und]e/imat+ ! spe$ at :i* ! c>d is mov$4 ,r ?an recognise ! true /ate ( ?+s1 ! m9d pref]s to accept ! sugge/ion t ! upw>d or d[nw>d movem5t ( ! po9t ( int]section is 9 "p due to oppositely direct$ horizontal movem5ts ( ! l9es !mselves ^u ! surface ( ! c>d4 ,:5 ! c>d is desc5d+ ! v]tical l9e is suppos$ to slide a little to ! "r & ! oblique l9e to ! left1 :i* w\ld h ! effect ( l[]+ _! po9t ( int]section indep5d5tly ( ! d[nw>d movem5t ( ! c>d itself4 ,:5 ! c>d asc5ds1 ^! horizontal movem5ts are suppos$ to be rev]s$1 & ! po9t ( int]section consequ5tly rais$4 ,! assumption is exactly analog\s to t made :5 an angle ( slope is unwitt+ly m9imis$4 ,ano!r example ( ! in/ability ( a m5tal /&>d occurs 9 ! e/imation ( lum9osity4 ,! lum9osity ( a bright object1 if reckon$ 9 t]ms ( ! same unit z t appli$ 9 judg+ ( a less bright "o1 w\ld appe> to be great] ?an x actually does appe>1 & ?is q indep5d5tly ( any effects ( fatigue4 ,illu/ration3 ,fig4 #cd4--;,illusion ( ,lum9osity4 ,! fact is well illu/rat$ by a famili> exp]im5t4 ,fig4 #cd is photograph$ f a transp>5cy made by sup]pos+ sev]al diff]5t l5g?s ( gelat9e film s z to =m a s]ies ( /eps4 ,at ! "r-h& end ( ! image ! light has pass$ ?r\gh only "o lay] ( ! film2 9 ! next division x has trav]s$ two lay]s1 9 ! next1 ?ree1 & 9 ! la/1 f\r4 ,! lum9osity ( ea* ( ! f\r squ>es into :i* ! oblong is divid$ is1 9 a physical s5se1 q uni=m1 b ! m5tal /&>d ( lum9osity v>ies = diff]5t p>ts ( ! image1 increas+ or decreas+1 z ! case may be1 n p] saltum1 b smoo?ly & cont9u\sly1 ) ! result t ea* squ>e looks bright] t[>ds ! left ?an t[>ds ! "r4 ,! appe>ance1 :i* is (t5 lik5$ to t pres5t$ by a fragm5t ( a flut$ column1 is equally well %[n :5 ! diagram is illum9at$ in/antane\sly by an electric sp>k1 & _c1 !re=e1 be acc\nt$ = by ret9al fatigue4 ,if ! squ>es are sep>at$ f "o ano!r by di/9ct l9es ( dem>cation1 h[ev] f9e1 ! /&>d ( lum9osity becomes uni=m = ea* squ>e1 & ! illusion vani%es4 ,?is fact suffici5tly disposes ( ! hypo!sis :i* has be5 advanc$ to ! effect t ! ph5om5on is due to physiological causes4 ,i n[ propose to discuss a curi\s consequ5ce ( ! fluctuation ( unaid$ judgm5t z reg>ds ! purity ( a col\r4 ,:5 any col\r occupies a pr$om9ant place 9 ! field ( vision1 we are apt to consid] x z be+ less pure1 or pal]1 ?an we %\ld if x 7 less conspicu\s1 our /&>d ( :it5ess t5d+ to approximate itself to ! col\r 9 "q4 ,= ! sake ( cle>ness let u fir/ conf9e our att5tion to a def9ite col\r--say r$4 ,an absolutely pure r$ is "o t is entirely free f any admixture ( :ite2 9 proportion z x conta9s m & m :ite1 ! m impure1 or 9 o!r words1 ! m pale does x become1 until at la/ all trace ( p]ceptible r$ness is lo/ & ! col\r is indi/+ui%able f :ite4 ,illu/ration3 ,fig4 #ce4--;,illusion ( ,col\r4 ,a conv5i5t way ( pictur+ ! scale ( purity is %[n 9 ,fig #ce4 ,! %ad$ oblong may be suppos$ to repres5t a pa9t$ /rip ( c>dbo>d or pap]4 ,at ! extreme "r h& end ! col\r is suppos$ to be absolutely pure r$2 t[>ds ! left ! r$ gradually becomes pal] or m dilute1 & at ! middle ( ! diagram x has m]g$ into p]fect :it5ess4 ,! figures #j to #ajj f left to "r d5ote ! p]c5tage ( free r$ conta9$ 9 ! mixture at diff]5t p>ts ( ! scale2 ! lum9osity is suppos$ to be uni=m ?r\gh\t4 ,n[ ! :ite light ) :i* ! r$ is dilut$ may be reg>d$ z consi/+ ( two p>ts1 "o ( :i* is ( exactly ! same hue z ! pure r$ itself1 & ! o!r an equival5t proportion ( ! complem5t>y col\r1 :i* 9 ! pres5t case w be gre5i%-blue4 ,! fact !re=e really is t1 z we pass along ! scale f #ajj to #j1 ! total quantity ( r$ 9 ! mixture is n r$uc$ to no?+1 b only to "o half1 :ile at ! same "t gre5i%-blue is add$ 9 proportions increas+ f n\ght at ! extreme "r to #ej p] c5t4 ( ! :ole at ! middle ( ! c>d4 ,! ord9ates ( ! quadrilat]al figure ,e ,d ,b ,f %[ ! proportion ( r$1 & ^? ( ! triangle ,e ,f ,b ! proportion ( gre5i%-blue1 at diff]5t p>ts ( ! scale4 ,reg>d+ ! portion ( ! /rip :i* lies above ! po9t m>k$ #j1 z repres5t+ ! z]o ( col\r--t is1 :it5ess or greyness1 :i* is ess5tially ! same z :it5ess--let u cont9ue ! diagram 9 ! negative direction1 gradually r$uc+ ! quantity ( r$ until x falls f #ej p] c5t4 ( ! :ole at ,f to no?+ at ,a1 & at ! same "t increas+ t ( ! gre5i%-blue f #ej p] c5t4 at ,f to #ajj p] c5t4 at ,a4 ,! resultant hue 9 ! portion ( ! c>d betwe5 ,f & ,a w be gre5i%-blue1 :i* beg9s to be p]ceptible z a v pale t9t j to ! left ( ,f1 & increases 9 purity z ,a is approa*$1 at :i* po9t ! col\r w be entirely free f any admixture ) :ite4 ,we h 9 ! scale ?us pres5t$ to our imag9ation a pair ( col\rs1 ea* occupy+ "o-half ( ! scale1 & gradually dim9i%+ 9 purity t[>ds ! middle l9e2 "h only1 j at ! /age :]e "o col\r m]ges into ! o!r1 is !re no col\r at all1 & ?is region repres5ts ! fix$ physical z]o or /&>d f :i* is reckon$ ! purity ( a col\r correspond+ to any o!r portion ( ! scale4 ,! complet$ scale1 x w be obs]v$1 ?\gh orig9ally int5d$ only = ! case ( r$1 turns out to be equally s]viceable = gre5i%-blue3 if we consid] gre5i%-blue z positive1 !n ! r$1 be+ on ! o!r side ( z]o1 mu/ be reg>d$ z negative4 ,any o!r possible pairs ( complem5t>y col\rs may be simil>ly treat$4 ,?is device enables u at once to und]/& ! consequ5ce ( m5tally displac+ ! z]o1 :ile physically ! scale rema9s un*ang$4 ,:5 r$ is ! prevail+ col\r 9 ! field ( vision1 we are incl9$ to consid] x unduly pale2 9 o!r words we imag9e x to be ne>] ! z]o ( ! scale ?an is actually ! case1 & s are l$ to %ift our /&>d ( :it5ess f ! middle slightly t[>ds ! r$ end ( ! scale4 ,! new position assign$ to :ite1 be+ a little to ! "r ( ! po9t m>k$ #j 9 ,fig4 #ce1 is "o :]e1 "u cu/om>y circum/ances1 ! col\r w\ld be call$ pale r$4 ,at ! same "t1 an object :i* is normally :ite1 & is exactly mat*$ at ! middle ( ! scale1 w\ld be a little to ! left ( ! imag9>y z]o1 & w\ld consequ5tly appe> to be ( a gre5i%-blue t9t4 ,?is app>5t trans=mation ( :ite or grey into a decid$ col\r is mo/ /rik+ :5 ! induc+ col\r is consid]ably dilut$ ) :ite or is ( feeble lum9osity4 ,a small fragm5t ( neutral grey pap]1 plac$ ^u a mu* l>g] piece ( a bright r$ hue1 g5]ally appe>s at ! fir/ glance#aa to be gre5i%-blue1 b if ! light is at all /rong1 only slightly s4 ,if1 h[ev]1 a %eet ( :ite tissue pap] is laid ov] ! :ole1 ! gre5i%-blue t9t imm$iately becomes />tl+ly di/9ct1 & may ev5 appe> m decid$ ?an ! r$ itself z se5 ?r\gh ! tissue4 ,! same piece ( grey pap]1 :5 plac$ ^u a gre5 gr\nd1 appe>s rose-col\r$1 & ^u a blue gr\nd1 yell[1 ! effect be+ always greatly increas$ by ! dilu5t action ( sup]pos$ tissue pap]4 ,!re seem to be sev]al reasons1 p>tly physical & p>tly psy*ological1 :y ^! contra/ col\rs1 z !y are call$1 are m pron\nc$ :5 ! col\r t calls !m into exi/5ce ei!r has a some:at pale t9t or is feebly illum9at$4 ,probably ! mo/ important is ( a purely physical "*4 ,! refract+ m$ia ( ! eye are mu* less p]fectly transp>5t ?an a good glass l5s is2 !y are s5sibly turbid or opalesc5t1 & 9 consequ5ce ( ?is defect "s ( ! light :i* falls ^u !m is irregul>ly scatt]$ ov] ! ret9a4 ,if we look at a bright r$ object ) a small :ite pat* ^u x1 ! image ( ! pat* z =m$ ^u ! ret9a is n1 physically speak+1 p]fectly :ite1 b slightly col\r$ by diffus$ r$ light2 ow+ h[ev] to ! psy*ological influ5ce to :i* our att5tion has be5 direct$1 ! fa9t r$ coloration is n consci\sly p]ceiv$2 ! same m5tal displacem5t ( ! z]o :i*1 :5 ! excit+ col\r 0 feeble1 l$ u to reg>d :ite 7or grey7 z blui%-gre51 n[ causes :at is actually pale r$ to appe> :ite4 ,!re is no ne$ :atev] to assume t ! contra/ col\rs ) :i* we h be5 deal+ are ( physiological orig9 & due to an inductive action excit$ 9 portions ( ! ret9a adjac5t to ^? ^u :i* col\r$ light falls4 ,on ! contr>y1 x w\ld be a matt] = surprise if ! case 9 "q pres5t$ an exception to ! compreh5sive law :i* gov]ns ! fluctuation ( ! m5tal judgm5t4 ,( ! op]ation ( ?is law ,i h quot$ sev]al v div]se in/ances1 & ! numb] might easily h be5 increas$4 ,nor is x only 9 relation to optical ph5om5a t ! law holds good2 9 its mo/ g5]al =m1 supplem5t$ x may be 9 "s in/ances by obvi\s coroll>ies1 x is applicable to almo/ e case 9 :i* physical attributes ( :atev] k9d are ! subject ( unassi/$ m5tal judgm5t4 ,,*apt] ,v4 ,,curiosities ,,( ,,vision4 ,! function ( ! eye1 reg>d$ z an optical in/rum5t1 is limit$ to ! =mation ( lum9\s images ^u ! ret9a4 ,f a purely physical po9t ( view x is a simple en\gh piece ( app>atus1 &1 z 0 =cibly po9t$ out by ,helmholtz1 x is subject to a numb] ( defects :i* c be demon/rat$ by ! simple/ te/s1 & :i*1 if !y occurr$ 9 a %op-b\ght in/rum5t1 w\ld be consid]$ intol]able4 ,:at takes place 9 ! ret9a itself "u lum9\s excitation1 & h[ ! s5sation ( sight is produc$1 are que/ions :i* belong to ! sci5ces ( physiology & psy*ology2 & 9 ! physiological & psy*ological dep>tm5ts ( ! visual ma*9]y we meet ) an additional ho/ ( objectionable peculi>ities f :i* any humanly-con/ruct$ app>atus is by ! nature ( ! case free4 ,yet 9 spite ( all ^! drawbacks our eyes d u excell5t s]vice1 & provid$ t !y are free f actual mal=mation & h n suff]$ f injury or disease1 we d n (t5 f9d fault ) !m4 ,?is1 h[ev]1 is n because !y are z good z !y might be1 b because ) incessant practice we h acquir$ a v high degree ( skill 9 _! use4 ,if any?+ is m rem>kable ?an ! ease & c]ta9ty ) :i* we h le>nt to int]pret ocul> indications1 :5 !y are 9 "s sort ( con=mity ) ext]nal objects1 x is ! p]t9acity ) :i* we refuse to be misl$ :5 our eyes are do+ _! be/ to deceive u4 ,9 our e>lie/ ye>s we began to f9d out t we mu/ n believe all we saw2 exp]i5ce gradually taught u t on c]ta9 po9ts & "u c]ta9 circum/ances ! indications ( our organs ( vision 7 uni=mly mean+less or fallaci\s1 & we soon discov]$ t x w\ld save u tr\ble & add to ! com=t ( life if we cultivat$ a habit ( completely ignor+ all su* visual s5sations z 7 ( no practical value4 ,9 ?is mo/ ( u h be5 rem>kably successful2 s mu* s1 t if1 f motives ( curiosity1 or = ! sake ( sci5tific exp]im5t1 we wi% to direct our att5tion to ! s5sations 9 "q1 & to see ?+s z !y actually appe>1 we c only d s ) ! greate/ difficulty2 sometimes1 inde$1 n at all1 unless ) ! assi/ance ( "s specially contriv$ artifice4 ,9 ! pres5t *apt] x is propos$ to discuss a few ( ! less famili> vag>ies ( ! visual organs1 & to %[ h[ !y may be demon/rat$4 ,"s ( ! exp]im5ts may1 x is to be fe>$1 be f\nd r difficult2 success w dep5d ma9ly ^u ! exp]im5tali/'s ability to lay aside habit & prejudice1 & give close att5tion to 8 visual s5sations2 b x is h>dly to be expect$ t an unskill$ p]son w at ! fir/ attempt obs]ve all ! ph5om5a :i* w be ref]r$ to4 ,among ! mo/ annoy+ ( ! ecc5tricities :i* *>act]ise ! s5se ( vision is t kn[n z ! p]si/5ce ( impressions4 ,! s5sation ( sight :i* is produc$ by an illum9at$ object does n cease at ! mom5t :5 ! excit+ cause is remov$ or *ang$ 9 position2 x cont9ues = a p]iod :i* is g5]ally said to be ab\t a t5? ( a second1 b may sometimes be mu* m or less4 ,x is = ?is reason t we _c see ! details ( any?+ :i* is 9 rapid motion1 b only an indi/9ct blur1 result+ f ! confusion ( successive impressions4 ,if a c>dbo>d disk1 :i* is pa9t$ 9 conspicu\s black & :ite sectors is caus$ to rotate at a suffici5tly high spe$1 ! divisions are completely lo/ sight (1 & ! :ole surface appe>s to be ( a uni=mly grey hue4 ,b if ! rapidly rotat+ disk is illum9at$ by a prop]ly tim$ s]ies ( electric fla%es1 x looks z if x 7 at re/1 & 9 spite ( ! int]mitt5t nature ( ! light1 ! black & :ite sectors c be se5 q cont9u\sly1 ?\gh z a matt] ( fact ! int]vals ( d>kness are v mu* long] ?an ^? ( illum9ation4 ,p]si/5t impressions ( ?is k9d are (t5 spok5 ( z positive aft]-images4 ,!re is a v rem>kable ph5om5on accompany+ ! =mation ( positive aft]-images1 especially ^? foll[+ brief illum9ation1 :i* seems1 until comp>atively rec5t times1 to h entirely escap$ ! notice ( ! mo/ acute obs]v]s4 ,x 0 fir/ obs]v$ accid5tally by ,pr(essor ,c4 ,a4 ,"y1 :5 he 0 exp]im5t+ ) a l>ge electrical ma*9e :i* _h be5 newly acquir$ = 8 laboratory4 ,he notic$ t :5 a p[]ful ,leyd5 j> dis*>ge took place 9 a d>k5$ room1 any conspicu\s object 0 se5 twice at lea/1 ) an int]val ( a trifle less ?an a qu>t] ( a second1 ! fir/ "t vividly1 ! second "t fa9tly4 ,(t5 x 0 se5 a ?ird "t1 & sometimes1 b only ) great difficulty1 ev5 a f\r? "t4 ,he gave to ?is ph5om5on ! "n ( recurr5t vision2 x may p]haps be m appropriately d5om9at$ ! ,"y effect4 ,by means ( ! p[]ful ma*9e pres5t$ to ! ,royal ,in/itution by ,mr4 ,wim%ur/1 us$ 9 conjunction ) a batt]y ( ,leyd5 j>s1 ! ,"y effect has be5 successfully %[n to a l>ge assembly4 ,b x is q easy to demon/rate x on a small scale ) any influ5ce ma*9e :i* w give a sp>k ab\t an in* long4 ,"o ( ! t]m9als ( ! ma*9e %\ld be connect$ by a wire ) ! inn] coat+ ( a half-p9t ,leyd5 j>1 ! o!r ) ! out] coat+1 & ! dis*>g+ balls %\ld be set a qu>t] ( an in* ap>t4 ,! obs]v]'s eyes mu/ be %ield$ f ! direct light ( ! sp>k by any conv5i5t scre51 su* z a l>ge book set on end4 ,! be/ object = ! exp]im5t is a %eet ( :ite pap]1 plac$ 9 an upright position a few in*es away f ! t]m9als ( ! ma*9e & expos$ to ! full light ( ! dis*>ge4 ,! room be+ d>k5$1 let ! ma*9e be work$ sl[ly1 :ile ! eyes are turn$ t[>ds ! :ite pap]4 ,?is w be se5 = a mom5t :5 ! sp>k passes1 &1 aft] a d>k int]val ( ab\t "o-fif? ( a second1 x w make ano!r brief appe>ance4 ,aft] a fur!r %ort int]val ( d>kness1 a second recurr5t image w (t5 be se54 ,x may be rem>k$ t ! effect is mo/ /rik+ :5 ! eyes are n direct$ exactly ^u ! :ite pap]1 b above or on "o side ( x2 ! prop] di/ance ( ! pap] f ! sp>k-gap %\ld be f\nd by trial4 ,"u fav\rable conditions ,i h obs]v$ z _m z six or sev5 reappe>ances ( an object :i* 0 illum9at$ by a s+le dis*>ge4 ,^! foll[$ "o ano!r at ! usual rate--ab\t five 9 a second--& produc$ a tw9kl+ or quiv]+ effect1 closely resembl+ t att5d+ a fla% ( lightn+ :i* is n directly se54 ,!re c inde$ be little d\bt t ! prov]bial quiv] ( ! lightn+-fla% is 9 _m cases m]ely an effect ( recurr5t vision1 ?\gh sometimes1 ( c\rse1 z has be5 %[n by photographs1 ! dis*>ge is really multiple4 ,"s ye>s ago ,i call$ att5tion to a v diff]5t me?od ( exhibit+ a recurr5t image4 ,! app>atus us$ = ! purpose consi/s ( a vacuum tube m\nt$ 9 ! usual way ^u a horizontal axis capable ( rotation4 ,:5 ! tube is illum9at$ by a rapid succession ( dis*>ges f an induction coil1 & is made to rotate v sl[ly by clockwork 7turn+ once 9 e two or ?ree seconds71 a v curi\s ph5om5on may be notic$4 ,at a di/ance ( a few degrees beh9d ! tube & sep>at$ f x by an int]val ( p]fect d>kness1 comes a gho/4 ,?is gho/ is 9 =m an exact reproduction ( ! tube2 x is v cle>ly def9$1 & ?\gh its app>5t lum9osity is some:at feeble1 x c 9 mo/ cases be se5 )\t difficulty4 ,! v>i$ col\rs ( ! orig9al are1 h[ev]1 abs5t1 ! :ole ( ! phantom tube be+ ( a uni=m blui% or violet t9t4 ,if ! rotation is sudd5ly /opp$ ! gho/ /ill moves /eadily on until x rea*es ! lum9\s tube1 ) :i* x coalesces & s disappe>s4 7,see ,fig4 #cf1 :]e ! recurr5t image is repres5t$ by dott$ l9es47 ,illu/ration3 ,fig4 #cf4--,recurr5t ,vision demon/rat$ ) a ,vacuum ,tube4 ,m rec5tly a fre% s]ies ( exp]im5ts 7 und]tak5 9 connection ) ! ,"y effect & c]ta9 alli$ matt]s1 ! results be+ embodi$ 9 a communication to ! ,royal ,society 7,proc4 ,roy4 ,soc41 #ahid1 vol4 #ef1 p4 #acb74 ,among o!r ?+s an attempt 0 made to asc]ta9 h[ f> a recurr5t image 0 affect$ by ! col\r ( ! excit+ light4 ,) ?is object two me?ods ( exp]im5t+ 7 employ$4 ,9 ! fir/1 col\r$ light 0 obta9$ by pass+ :ite light ?r\gh col\r$ glasses2 9 ! second & m p]fect s]ies ( exp]im5ts1 ! pure col\r$ light ( ! spectrum 0 us$4 ,among o!r results x 0 f\nd t1 ct]is p>ibus1 ! recurr5t image 0 mu* /rong] ) gre5 light ?an ) any o!r1 & t :5 ! excitation 0 produc$ by pure r$ light1 h[ev] int5se1 !re 0 no recurr5t image at all4 ,illu/ration3 ,fig4 #cg4--,recurr5t ,vision ) ,rotat+ ,disk4 ,= a repetition ( my fir/ exp]im5t a me*anical lant]n slide is requir$ conta9+ a metal disk ab\t ?ree in*es 9 diamet] :i* c be caus$ to rotate sl[ly & /eadily ab\t its c5tre4 ,ne> ! edge ( ! disk is a small circul> ap]ture4 ,! slide is plac$ 9 a limelight lant]n1 & a bright image ( ! hole is focuss$ ^u a di/ant scre51 all o!r light be+ c>efully %ut (f4 ,:5 ! disk is turn$ sl[ly1 ! spot ( light ^u ! scre5 goes r\nd & r\nd1 & x is g5]ally possible to see at once t ! bright prim>y spot appe>s to be foll[$ at a %ort di/ance by a mu* feebl] spot ( a violet col\r1 :i* is ! recurr5t image ( ! fir/4 7,see ,fig4 #cg47 ,x is ess5tial to keep ! direction ( ! eyes p]fectly /eady1 :i* is n a v easy ?+ to d )\t practice4 ,if a gre5 glass is plac$ be=e ! l5s1 ! gho/ w be at its be/1 & %\ld be se5 q cle>ly & easily1 provid$ t no attempt is made to foll[ x ) ! eyes4 ,) an orange glass ! gho/ becomes less di/9ctly visible1 & its col\r g5]ally appe>s to be gre5i%-blue1 in/ead ( violet z be=e4 ,:5 a r$ glass is sub/itut$1 ! gho/ completely disappe>s4 ,if ! spe$ ( rotation is suffici5tly high1 ! r$ spot is consid]ably elongat$ dur+ its revolution1 & its col\r ceases to be uni=m1 ! tail assum+ a light blui%-p9k t9t4 ,b h[ev] great ! spe$1 no complete sep>ation ( ! spot into r$ & p9k portions c be effect$1 & no recurr5t image is ev] f\nd4 ,! spectrum me?od ( obs]vation c only be c>ri$ out on a small scale1 & is n suit$ = exhibition to an audi5ce4 ,x1 h[ev]1 af=ds ! be/ means ( asc]ta9+ h[ f> ! app>5t col\r ( ! recurr5t image dep5ds ^u t ( ! prim>y1 a matt] ( "s !oretical int]e/4 ,illu/ration3 ,fig4 #ch4--,recurr5t ,vision ) ,spectrum4 ,! arrangem5t adopt$ is %[n 9 ! annex$ diagram 7,fig4 #ch74 ,l is a lant]n conta9+ an oxyhydrog5 light or an electric arc lamp1 ,s is an adju/able slit1 ,m a projection l5s1 ,p a bisulphide ( c>bon prism1 ,d a metal plate 9 ! middle ( :i* is a circul> ap]ture #b millimetres 7#a_/ab in*7 9 diamet]4 ,a bright spectrum1 #f or #g c5timetres 9 l5g? 7ab\t #c in*es71 is project$ ^u ?is metal plate1 & a small select$ portion ( x passes ?r\gh ! r\nd hole2 !nce ! col\r$ light goes ?r\gh ! l5s ,n to ! little mirror ,q1 :i* reflects x ^u ! :ite scre5 ,r4 ,by prop]ly adju/+ ! position ( ! l5s ,n a %>p mono*romatic image ( ! r\nd hole 9 ! plate ,d is focuss$ ^u ! scre5 ,r4 ,to ! back ( ! mirror ,q is atta*$ a horizontal arm :i* is n q p]p5dicul> to ! mirror1 its incl9ation be+ capable ( adju/m5t4 ,! arm is turn$ sl[ly by clock-"w1 ?us caus+ ! col\r$ spot on ! scre5 to revolve 9 a circul> orbit ab\t #cj c5timetres 7#a foot7 9 diamet]1 its recurr5t image foll[+ at a %ort di/ance beh9d x4 ,:5 ! mirror turns once 9 #a-a_/b seconds1 ?is image appe>s ab\t #ej beh9d ! col\r$ spot1 ! correspond+ "t-int]val be+ ab\t "o-fif? ( a second4 ,us+ ?is app>atus1 x 0 f\nd t :ite light 0 foll[$ by a violet recurr5t image2 aft] blue & gre51 :5 ! image 0 brighte/1 its col\r 0 also violet2 aft] yell[ & orange x appe>$ blue or gre5i% blue4 ,on ! o!r h&1 :5 a complete spectrum 0 caus$ to revolve ^u ! scre51 ! :ole ( its recurr5t image f end to end appe>$ violet2 !re 0 no suspicion ( blue or gre5i%-blue at ! less refrangible end4 ,= ?is & o!r reasons giv5 9 ! pap] x 0 conclud$ t ! true col\r 0 9 all cases really violet1 ! blue & gre5i%-blue app>5tly se5 9 conjunction ) ! mu* bright] yell[ & orange ( ! prim>y be+ m]ely an illusory effect ( contra/4 ,x seems likely1 !n1 t ! ph5om5on :i* has be5 spok5 ( z recurr5t vision1 is due pr9cipally1 if n entirely1 to an action ( ! violet n]ve-fibres4 ,recurr5t vision is1 no d\bt1 g5]ally mo/ conspicu\s aft] a v brief p]iod ( ret9al illum9ation1 su* z 0 employ$ 9 ! exp]im5ts :i* we h be5 discuss+2 ?is is evid5tly due to ! fact t ! effect is mo/ easily p]ceiv$ :5 ! s5sibility ( ! ret9a has n be5 impair$ by fatigue4 ,b by a little ef=t x may be detect$ ev5 aft] v prolong$ illum9ation1 & a practis$ obs]v] c h>dly avoid notic+ a %ort fla% ( blui% light :i* manife/s itself ab\t a qu>t] ( a second aft] ! lights 9 a room h be5 sudd5ly ext+ui%$2 ! ph5om5on =ces itself ^u my att5tion almo/ e night :5 ,i turn (f ! electric lights4 ,x ne$ h>dly be po9t$ out t x repres5ts only a transi5t phase ( ! well kn[n positive aft]-image1 & x _h ev5 be5 obs]v$ 9 a vague & unc]ta9 sort ( way long be=e ! date ( ,pr(essor ,"y's exp]im5t4 ,helmholtz1 = example1 m5tions ! case ( a positive aft]-image :i* seem$ to disappe> & !n to bright5 up aga91 b he goes on to expla9--errone\sly1 z x turns out--t ! seem+ disappe>ance 0 illusory4 ,m4 ,*>p5ti]1 ( ,nancy1 ^: "w 9 physiological optics is well kn[n1 0 ! fir/ to notice & record a rem>kable ph5om5on :i*1 9 "s =m or o!r1 mu/ pres5t itself _m times daily to e p]son :o is n bl9d1 b :i* until ab\t sev5 ye>s ago _h be5 absolutely & univ]sally ignor$4 ,! law :i* is associat$ ) ,*>p5ti]'s "n is ?is3--,:5 d>kness is succe$$ by light1 ! /imulus :i* ! ret9a at fir/ receives1 & :i* causes ! s5sation ( lum9osity1 is foll[$ by a brief p]iod ( ins5sibility1 result+ 9 ! s5sation ( mom5t>y d>kness4 ,x appe>s t ! d>k p]iod beg9s ab\t "o sixtie? ( a second aft] ! light has fir/ be5 admitt$ to ! eye1 & la/s = ab\t an equal "t4 ,! :ole alt]nation f light to d>kness & back aga9 to light is p]=m$ s rapidly1 t except "u c]ta9 conditions1 :i*1 h[ev]1 occur frequ5tly en\gh1 x _c be detect$4 ,illu/ration3 ,fig4 #ci4--;,*>p5ti]'s ,d>k ,b&4 ,! app>atus :i* ,*>p5ti] employ$ = demon/rat+ & measur+ ! duration ( ?is effect is v simple4 ,x consi/s ( a black5$ disk ) a :ite sector1 m\nt$ ^u an axis4 ,:5 ! disk is illum9at$ by sunlight & turn$ r sl[ly1 ! direction ( ! gaze be+ fix$ ^u ! c5tre1 !re appe>s ^u ! :ite sector1 close beh9d its lead+ edge1 a n>r[ b q conspicu\s d>k b&4 7,see ,fig4 #ci47 ,! portion ( ! ret9a :i* at any mom5t is app>5tly occupi$ by ! d>k b&1 is t ^u :i* ! light reflect$ by ! lead+ edge ( ! :ite sector imp+$ "o sixtie? ( a second previ\sly4 ,b no special app>atus is requir$ to %[ ! d>k reaction4 ,9 ,fig4 #dj an attempt has be5 made to illu/rate :at any "o may see if he simply moves 8 h& betwe5 8 eyes & ! sky or any /rongly illum9at$ :ite surface4 ,! h& appe>s to be foll[$ by a d>k outl9e sep>at$ f x by a bright int]val4 ,! same k9d ( ?+ happ5s1 9 a m or less m>k$ degree1 :5ev] a d>k object moves across a bright backgr\nd1 or a bright object across a d>k backgr\nd4 ,illu/ration3 ,fig4 #dj4--;,*>p5ti]'s ,effect %[n ) ! ,h&4 ,9 ord] to see ! effect di/9ctly by ,*>p5ti]'s orig9al me?od1 ! illum9ation mu/ be /rong4 ,if1 h[ov]1 ! arrangem5t is slightly v>i$1 s t transmitt$ in/ead ( reflect$ light is made use (1 comp>atively feeble illum9ation is suffici5t4 ,a v effective way is to turn a small metal disk1 hav+ an op5 sector ( ab\t #fj1 9 front ( a %eet ( gr\nd or opal glass beh9d :i* is a lamp4 ,by an arrangem5t ( ?is k9d ^u a l>g] scale1 ! effect may easily be r5d]$ visible to an audi5ce4 ,! eyes %\ld n be all[$ to foll[ ! disk 9 its rotation1 b %\ld be direct$ /eadily ^u ! c5tre4 ,! acute & educat$ vision ( ,*>p5ti] enabl$ him1 ev5 :5 work+ ) 8 black & :ite disk1 to detect ! exi/5ce1 "u fav\rable conditions1 ( a second1 & sometimes a ?ird1 b& ( greatly dim9i%$ int5sity1 ?\gh he rem>ks t ! obs]vation is a v difficult "o4 ,:at is probably ! same effect c1 h[ev]1 z po9t$ out 9 my pap] ( #ahid1 be %[n q easily 9 a diff]5t mann]4 ,if a disk ) a n>r[ radial slit1 ab\t half a millimetre 7#a_/ej in*7 wide1 is caus$ to rotate at ! rate ( ab\t "o turn p] second 9 front ( a bright backgr\nd1 su* z a %eet ( gr\nd glass ) a lamp beh9d x1 ! mov+ slit assumes ! appe>ance ( a fan-%ap$ lum9\s pat*1 ! brightness ( :i* dim9i%es ) ! di/ance f ! lead+ edge4 ,& if ! eyes are /eadily fix$ ^u ! c5tre ( ! disk1 x w be notic$ t ?is bright image is /reak$ ) a numb] ( d>k radial b&s1 sugge/ive ( ! ribs or /icks ( a fan4 ,ne> ! circumf]5ce z _m z f\r or five su* d>k /reaks c be di/+ui%$ )\t difficulty2 t[>ds ! c5tre !y are less conspicu\s1 ow+ to ! ov]lapp+ ( ! successive images ( ! slit4 ,! effect is r\ghly indicat$ 9 ,fig4 #da4 ,illu/ration3 ,fig4 #da4--,multiple ,d>k ,b&s4 ,! d>k reaction kn[n z ! ,*>p5ti] effect occurs at ! beg9n+ ( a p]iod ( illum9ation4 ,!re is also a d>k reaction ( v %ort duration at ! end ( a p]iod ( illum9ation4 ,x %\ld be expla9$ t1 ow+ to :at is call$ ! prop] light ( ! ret9a1 ord9>y d>kness does n appe> absolutely black3 ev5 9 a d>k room on a d>k night ) ! eyes c>efully cov]$1 !re is always "s s5sation ( lum9osity :i* w\ld be suffici5t to %[ up a really black image if "o c\ld be produc$4 ,n[ ! d>kness :i* is exp]i5c$ aft] ! ext9ction ( a light is = a small fraction ( a second m int5se ?an common d>kness4 ,! fir/ m5tion ( ?is d>k reaction p]haps occurs 9 an article contribut$ to ,nature 9 #ahhe1 9 :i* x 0 /at$ t :5 ! curr5t 0 cut (f f an illum9at$ vacuum tube ! lum9\s image 0 almo/ in/antly replac$ by a correspond+ image :i* seem$ to be int5sely black ^u a less d>k backgr\nd1 & :i* 0 e/imat$ to la/ f a-qu>t] to a-half second4 ,abnormal d>kness1 x 0 add$1 foll[s z a reaction aft] lum9osity4 ,illu/ration3 ,fig4 #db4--,tempor>y ,ins5sitiv5ess ( ! ,eye4 ,9 ! ,royal ,society pap] be=e ref]r$ to ! po9t is fur!r discuss$1 & a me?od is describ$ by :i* ! /age ( reaction may be easily exhibit$ & its duration approximately measur$4 ,if a transluc5t disk1 made ( /\t draw+-pap] & hav+ an op5 sector1 is caus$ to rotate sl[ly 9 front ( a lum9\s backgr\nd1 a n>r[ radial d>k b&1 l a /reak ( black pa9t1 appe>s ^u ! pap] v ne> ! edge :i* foll[s ! op5 sector4 ,f ! space cov]$ by ?is b& :5 ! disk 0 rotat+ at a kn[n spe$1 ! duration ( ! d>k reaction 0 calculat$ to be ab\t "o-fiftie? ( a second2 my orig9al e/imate 0 !re=e an excessive "o4 ,! exp]im5t is illu/rat$ 9 ,fig4 #db4 ,"o m int]e/+ po9t %\ld be notic$ 9 ! tra9 ( visual ph5om5a :i* att5d a p]iod ( illum9ation4 ,! s5sation ( lum9osity :i* is excit$ :5 light fir/ /rikes ! eye is = ab\t a sixtie? ( a second mu* m int5se ?an x subsequ5tly becomes4 ,?is is %[n by ! fact1 :i* is obvi\s en\gh :5 once att5tion has be5 direct$ to x1 t ! bright b&1 :i* 9 ! ,*>p5ti] disk int]v5es betwe5 ! d>k b& & ! lead+ edge ( ! :ite sector1 appe>s to be mu* m /rongly illum9at$ ?an any o!r portion ( ! sector4 ,! complete ord] ( visual ph5om5a obs]v$ :5 ! ret9a is expos$ to ! action ( light = a limit$ "t may !re=e be summ$ up z foll[s3-- 7#a7 ,imm$iately ^u ! impact ( ! light !re is exp]i5c$ a s5sation ( lum9osity1 ! int5sity ( :i* increases = ab\t "o-sixtie? ( a second3 m rapidly t[>ds ! end ( t p]iod ?an at fir/4 7#b7 ,!n ensues a sudd5 re-action1 la/+ also = ab\t "o-sixtie? ( a second1 9 virtue ( :i* ! ret9a becomes p>tially ins5sible to r5ew$ or cont9u$ lum9\s impressions4 ,^! two effects may be repeat$ 9 a dim9i%$ degree1 z (t5 z ?ree or f\r times4 7#c7 ,! /age ( fluctuation is succe$$ by a s5sation ( /eady lum9osity1 ! int5sity ( :i* is1 h[ev]1 consid]ably bel[ ! mean ( t exp]i5c$ dur+ ! fir/ "o-sixtie? ( a second4 7#d7 ,aft] ! ext]nal light has be5 %ut (f1 a s5sation ( dim9i%+ lum9osity cont9ues = a %ort "t1 & is succe$$ by a brief int]val ( d>kness4 7#e7 ,!n foll[s a sudd5 & cle>ly-def9$ s5sation ( :at may be call$ abnormal d>kness--d>k] ?an common d>kness--:i* la/s = ab\t "o-sixtie? ( a second1 & is foll[$ by ano!r int]val ( ord9>y d>kness4 7#f7 ,f9ally1 9 ab\t a fif? ( a second aft] ! ext9ction ( ! ext]nal light1 !re occurs ano!r transi5t impression ( lum9osity1 g5]ally violet col\r$1 aft] :i* ! uni=mity ( ! d>kness rema9s undi/urb$4 ,fig4 #dc1 :i* is copi$ f my pap]1 gives a r\gh diagrammatic repres5tation ( ! above describ$ *a9 ( s5sations4 ,no acc\nt is "h tak5 ( ! comp>atively feeble aft]-images :i* succe$ ! recurr5t image1 & may la/ = sev]al seconds4 ,i propose n[ to say a few words ab\t a curi\s ph5om5on ( vision :i* a %ort "t ago excit$ consid]able int]e/4 ,illu/ration3 ,fig4 #dc4--,visual ,s5sations att5d+ a p]iod ( ,illum9ation4 ,illu/ration3 ,fig4 #dd4--;,b5ham's ,top4 ,9 ! ye> #ahie ,mr4 ,c4 ,e4 ,b5ham br\ght out a pretty little toy :i* he call$ ! ,artificial ,spectrum ,top4 ,x consi/s ( a c>dbo>d disk1 "o half ( :i* is pa9t$ black1 :ile on ! o!r half are drawn f\r successive gr\ps ( curv$ black l9es at diff]5t di/ances f ! c5tre1 z %[n 9 ,fig4 #dd4 ,:5 ! disk rotates r sl[ly1 ea* gr\p ( black l9es g5]ally appe>s to assume a diff]5t col\r1 ! nature ( :i* dep5ds ^u ! spe$ ( ! rotation & ! int5sity & quality ( ! light4 ,"u ! be/ conditions ! inn] & out] gr\ps ( l9es become bright r$ & d>k blue2 at ! same "t ! int]m$iate gr\ps also appe> t9t$1 b ! hues :i* !y assume are r unc]ta9 & difficult to specify4 ,by f> ! mo/ /rik+ ( ! col\rs exhibit$ by ! top is ! r$1 & next to t ! blue2 ?is latt] is1 h[ev]1 sometimes describ$ z blui%-gre54 ,"s exp]im5ts c>ri$ out by myself 9 #ahif 7,proc4 ,roy4 ,soc41 vol4 #fj1 p4 #cgj7 seem to indicate pretty cle>ly ! cause ( ! rem>kable bright r$ col\r1 & also t ( ! blue4 ,! m feeble t9ts ( ! two int]m$iate gr\ps ( l9es p]haps result f simil> causes 9 a modifi$ =m1 b ^! h n yet be5 inve/igat$4 ,9 ! r$ col\r we h ano!r /rik+ example ( an exce$+ly common ph5om5on :i* is habitually disreg>d$2 inde$ ,i c f9d no record ( its ev] hav+ be5 notic$ at all4 ,! fact is t :5ev] a bright image is sudd5ly =m$ ^u ! ret9a aft] a p]iod ( comp>ative d>kness1 ?is image appe>s = a %ort "t to be surr\nd$ by a n>r[ col\r$ bord]1 ! col\r1 "u ord9>y conditions ( illum9ation1 be+ r$4 ,if ! light is v /rong1 ! transi5t bord] is gre5i%-blue1 b ?is col\r1 z w be expla9$ lat]1 turn$ out to be m]ely an aft]-effect ( r$4 ,sometimes1 :5 ! object is 9 motion1 bo? r$ & blue are se5 toge!r4 ,! obs]vations 7 fir/ made 9 ! foll[+ mann]4 ,a black5$ z9c plate1 9 :i* is a small r\nd hole cov]$ ) a piece ( ?9 writ+-pap]1 is fix$ ov] a l>g] op5+ 9 a wood5 bo>d2 ?us we are furni%$ ) a %>ply-def9$ transluc5t disk1 :i* is surr\nd$ by a p]fectly opaque sub/ance4 ,an arrangem5t is provid$ = cov]+ ! transluc5t disk ) a %utt]1 :i* c be op5$ v rapidly by releas+ a /rong spr+4 ,if ?is app>atus is held betwe5 ! eyes & a lamp1 & ! transluc5t disk is sudd5ly disclos$ by work+ ! %utt]1 ! disk appe>s = a %ort "t to be surr\nd$ by a n>r[ r$ bord]4 ,! wid? ( ! bord] is p]haps a millimetre 7#a_/be in*71 & ! appe>ance la/s = some?+ l a t5? ( a second4 ,mo/ p are at fir/ q unable to recognise ?is effect1 ! difficulty be+1 n to see x1 b to "k t "o sees x4 ,^? :o h be5 accu/om$ to visual obs]vations g5]ally p]ceive x )\t any difficulty :5 !y "k :at to look =1 & no d\bt x w\ld be v evid5t to a baby :i* _h n advanc$ v f> 9 ! education ( its eyes4 ,! obs]vation is made r less difficult by a fur!r device4 ,if ! disk is divid$ into two p>ts by an opaque /rip across ! middle1 x is cle> t ea* half disk w h its r$ bord]1 & if ! /rip is made suffici5tly n>r[1 ! r$ bord]s along its edges w meet or p]haps ov]lap1 & ! :ole /rip w1 = a mom5t aft] ! %utt] is op5$1 appe> r$4 ,a disk 0 ?us prep>$ by gumm+ across ! pap] a v n>r[ /rip ( t9foil4 ,! effect produc$ :5 su* a disk is sudd5ly expos$ is indicat$ 9 ,fig4 #de1 ! r$ col\r be+ repres5t$ by %ad+4 ,illu/ration3 ,fig4 #de4--;,demon/ration ( ,r$ ,bord]s4 ,a simpl] app>atus is1 h[ev]1 q suffici5t = %[+ ! ph5om5on1#ab & ) practice "o c ev5 acquire ! p[] ( see+ x )\t any artificial aid at all4 ,i h _m times notic$ fla%es ( r$ ^u ! black lett]s ( a book t ,i 0 read+1 or ^u ! edges ( ! page3 bright metallic1 or poli%$ objects (t5 %[ x :5 !y pass across ! field ( vision 9 consequ5ce ( a movem5t ( ! eyes1 & x 0 an accid5tal obs]vation ( ?is k9d :i* sugge/$ ! foll[+ easy way ( exhibit+ ! effect exp]im5tally4 ,an inc&esc5t electric lamp 0 fix$ beh9d a r\nd hole 9 a %eet ( metal :i* 0 atta*$ to a bo>d4 ,! hole 0 cov]$ ) two or ?ree ?icknesses ( writ+ pap]1 mak+ a bright disk ( ne>ly uni=m lum9osity4 ,:5 ?is arrangem5t 0 mov$ r quickly ei!r backw>ds & =w>ds or r\nd & r\nd 9 a small circle1 ! edge ( ! /reak ( light ?us =m$ appe>$ to be bord]$ ) r$4 ,if ?is exp]im5t is p]=m$ ) a /rong light beh9d ! pap]1 ! /reak becomes bord]$ ) gre5i%-blue in/ead ( r$4 ,) an int]m$iate degree ( illum9ation1 bo? blue & r$ may be se5 toge!r4 ,mo/ ( ! effects t h s f> be5 describ$ 7 produc$ by transmitt$ light1 b reflect$ light w %[ !m equally well4 ,if y place a pr9t$ book 9 front ( y ne> a good lamp & int]pose a d>k scre5 be=e y\r eyes1 !n1 :5 ! scre5 is sudd5ly )drawn1 ! pr9t$ lett]s w = a mom5t appe> r$1 quickly *ang+ to black4 ,"s practice is requir$ be=e ?is obs]vation c be made satisfactorily1 b by a simple device x is possible to oblit]ate ! image ( ! lett]s be=e ! r$ness has _h "t to disappe>2 ! col\r !n becomes q easily p]ceptible4 ,hold two scre5s toge!r side by side1 a black "o & a :ite "o1 9 su* a mann] t an op5 space is left betwe5 !m4 7,see ,fig4 #df47 ,9 ! fir/ place let ! black scre5 cov] ! pr9t+2 !n quickly move ! scre5s sideways s t ! pr9t$ lett]s may be = a mom5t expos$ to view ?r\gh ! gap1 /opp+ ! movem5t z soon z ! page is cov]$ by ! :ite scre54 ,dur+ ! brief glimpse t w be _h ( ! black lett]s :ile ! gap is pass+ ov] !m1 !y w1 if ! illum9ation is suitable1 appe> to be bright r$4 ,illu/ration3 ,fig4 #df4--;,black & ,:ite ,scre5s4 ,illu/ration3 ,fig4 #dg4--;,disk = ,r$ ,bord]s4 ,we may g a /ep fur!r4 ,cut out a disk ( :ite c>dbo>d1 divide x into two equal p>ts by a /raight l9e ?r\gh ! c5tre1 & pa9t "o half black4#ac ,at ! junction ( ! black & :ite portions cut out a gap1 :i* may conv5i5tly be ( ! =m ( a sector ( #de4 7,see ,fig4 #dg47 ,/ick a long p9 ?r\gh ! c5tre & hold ! arrangem5t by ! po9t$ end ( ! p9 a few in*es above a pr9t$ page ne> a good light4 ,make ! disk sp9 at ! rate ( ab\t five or six turns a second by /rik+ ! edge ) ! f+]4 ,z be=e1 ! lett]s :5 se5 ?r\gh ! gap w appe> r$1 & p]si/5ce w r5d] ! repeat$ impressions almo/ cont9u\s s long z ! rotation is kept up2 any "o see+ ! pr9t+ = ! fir/ "t ?r\gh ! rotat+ disk w\ld believe t x 0 done ) r$ ink4 ,c>e mu/ be tak5 t ! disk does n ca/ a %ad[ ^u ! page1 & t ! int5sity ( ! illum9ation is prop]ly adju/$4 ,i h devis$ sev]al r m elaborate contrivances = mak+ ! disks rotate at a uni=m spe$2 "o ( ^! is %[n 9 ,fig4 #ej4 ,9 none ( ^! exp]im5ts does an ext5d$ black surface ev] appe> r$1 b only black dots or l9es4 ,& ! l9es mu/ n be too ?ick2 if _! ?ickness is mu* m ?an a millimetre 7#a_/be in*71 ! l9es1 z se5 by an obs]v] f ! usual di/ance = read+1 d n become r$ ?r\gh\t1 b only along _! edges4 ,! r$ appe>ance does n 9 fact orig9ate 9 ! black l9es !mselves3 ^! s]ve m]ely z a backgr\nd = %[+ up ! r$ bord] :i* fr+es ext]nally ! :ite portions ( ! pap]1 & ! wid? ( ?is bord] does n exce$ ab\t "o-fif? ( a degree4 ,b by employ+ a suffici5tly l>ge disk & select+ designs or lett]s compos$ ( l9es ( suitable ?ickness1 ! col\r effect has be5 %[n to a l>ge audi5ce4 ,:5 ! disk is turn$ 9 ! opposite direction1 s t ! gap is prec$$ by :ite & foll[$ by black1 ! l9es ( ! design appe> at fir/ sight to become d>k blue in/ead ( r$4 ,att5tive obs]vation1 h[ev]1 %[s t ! app>5tly blue t9t is n =m$ ^u ! l9es !mselves1 z ! r$ t9t 01 b ^u ! :ite gr\nd j outside !m4 ,?is introduces to our notice ano!r bord] ph5om5on1 :i* seems to pres5t itself :5 a d>k pat* is sudd5ly =m$ on a bright gr\nd1 = t is ess5tially :at takes place :5 ! disk is turn$ ! rev]se way4 ,i made "s attempts to obta9 m direct evid5ce t su* a d>k pat* appe>$ = a mom5t to h a blue bord]1 & aft] "s tr\ble succe$$ 9 do+ s4 ,a circul> ap]ture 0 cut 9 a wood5 bo>d & cov]$ ) :ite pap]2 a lamp 0 plac$ beh9d ! bo>d1 & ?us a bright disk 0 obta9$1 z 9 ! =m] exp]im5t4 ,an arrangem5t 0 prep>$ by means ( :i* "o half ( ?is bright disk c\ld be sudd5ly cov]$ by a metal %utt]1 & x 0 f\nd t :5 ?is 0 done a n>r[ blue b& appe>$ on ! bright gr\nd j beyond & adjo9+ ! edge ( ! %utt] :5 x _h come to re/4 ,! blue b& la/$ = ab\t a t5? ( a second1 & x seem$ to disappe> by retreat+ into ! black edge ( ! %utt]4 ,! ph5om5on is illu/rat$ 9 ,fig4 #dh1 :]e ! %ad$ b& indicates ! blue bord]4 ,illu/ration3 ,fig4 #dh4--;,demon/ration ( ,blue ,bord]4 ,we h !n to acc\nt1 if possible1 = ! two facts t1 9 ! =mation ( ^! transi5t col\r-bord]s1 ! r$ s5sation occurs 9 a portion ( ! ret9a :i* has n itself be5 expos$ to ! direct action ( light1 :ile ! blue occurs 9 a portion :i* is /eadily illum9at$1 bo? col\r s5sations be+ ref]r$ to localities adjac5t to ^? 9 :i* a *ange ( illum9ation has sudd5ly tak5 place4 ,accept+ ! ,"y-,helmholtz !ory ( col\r vision1 ! effects mu/1 ,i ?9k1 be attribut$ to a sympa!tic affection ( ! r$ n]ve fibres4 ,:5 ! v>i\s n]ve fibres occupy+ a limit$ portion ( ! ret9a are sudd5ly /imulat$ by :ite light 7or by any k9d ( light :i* conta9s a r$ con/itu5t7 ! imm$iately surr\nd+ r$ n]ve fibres are = a %ort p]iod excit$ sympa!tically1 :ile ! violet & gre5 fibres are n s excit$1 or 9 a mu* less degree4 ,& aga91 :5 light is sudd5ly cut (f f a pat* 9 a bright field1 !re occurs a sympa!tic ins5sitive reaction 9 ! r$ fibres j outside ! d>k5$ pat*1 9 virtue ( :i* !y cease = a mom5t to respond to ! lum9\s /imulus2 ! gre5 & violet fibres1 by cont9u+ to respond un9t]rupt$ly1 give rise to ! s5sation ( a blue bord]4 ,x is p]haps desirable to ref] briefly to ano!r propos$ explanation ( ! ph5om5on1 :i* occurr$ to myself at an e>ly /age ( ! inve/igation1 & has s9ce be5 sugge/$ by _m diff]5t p]sons4 ,! explanation 9 "q is ( a purely physical "*1 & dep5ds ^u ! non-a*romatism ( ! eye4 ,illu/ration3 ,fig4 #di4--;,disk = exp]im5ts on ! orig9 ( ,col\r-bord]s4 ,)\t go+ into details1 x w suffice to quote a s+le exp]im5t :i* is ( itself fatal to any su* !ory4 ,prep>e a disk l t %[n 9 ,fig4 #di1 & sp9 x above a page ( pr9t+4 ,! lett]s b5ea? ! zone :i* is p>tly black & p>tly :ite w1 "u ! usual conditions1 turn r$1 b ^? b5ea? ! rema9d] ( ! disk w reta9 _! blackness4 ,! dem>cation is q def9ite1 & a s+le pr9t$ ^w may be made to appe> r$ 9 ! middle & black at its two ends4 ,n[ x is1 ( c\rse1 impossible t ! l5ses ( ! eye %\ld be p]fectly accommodat$ = ! lett]s :i* appe> black1 & at ! same "t s]i\sly out ( focus = ! o!rs4 ,?is explanation1 !re=e1 simple & obvi\s z x may seem1 is altoge!r unt5able4 ,:e!r or n ! hypo!sis :i* ,i h sugge/$ is correct 9 all its details1 x is1 ,i ?9k1 suffici5tly obvi\s t ! r$ & blue col\rs ( ,b5ham's top are due to exactly ! same causes z ! col\rs obs]v$ 9 my own exp]im5ts1 = ! ess5tial conditions are ! same 9 bo? cases4 ,! la/ curiosity :i* ,i w notice is connect$ ) ! fact already m5tion$1 t :5 ! illum9ation is /rong1 ! transi5t bord]-col\rs are ne>ly rev]s$1 gre5i%-blue appe>+ 9 place ( r$1 & brick-r$ 9 place ( blue4 ,i 0 = a long "t q unable to imag9e any reasonably probable explanation ( ?is circum/ance1 b a clue 0 f9ally obta9$ f consid]ation ( ! fact t gre5i%-blue is ! complem5t>y col\r to r$1 & 9 a subsequ5t memoir 7,proc4 ,roy4 ,soc41 vol4 #fa1 p4 #bfi7 "s exp]im5ts 7 describ$ :i* %[1 z ,i believe conclusively1 t ! gre5i%-blue bord]s se5 9 a /rong light are simply negative aft]-images ( ! usual r$ "o4 ,^! negative aft]-images are ( ! famili> k9d t are obs]v$ aft] "o has gaz$ = "s "t at a bright col\r$ object4 ,if a r$ waf] ly+ ^u a %eet ( :ite or grey pap] is look$ at /eadily = ab\t half a m9ute1 & ! gaze is !n sudd5ly transf]r$ to "s o!r "p ( ! pap]1 a gre5i%-blue gho/ ( ! waf] w be se54 ,! portion ( ! ret9a ^u :i* ! r$ image at fir/ falls becomes fatigu$ & p>tially ins5sible to r$ light2 x is !re=e unable to appreciate ! r$ compon5t ( ! :ite light aft]w>ds reflect$ to x by ! pap]1 & ! s5sation ( ! complem5t>y col\r consequ5tly pr$om9ates2 h5ce ! gre5i%-blue gho/1 :i* is call$ ! negative aft]-image ( ! waf]4 ,! new exp]im5ts %[ t1 if a c]ta9 condition is fulfill$1 ! usual prolong$ />e becomes unnecess>y1 a mom5t>y glance suffic+ to produce a /rong b fugitive aft]-image4 ,! condition is t ! "p ( ! ret9a :]e ! image is to be =m$1 %all h be5 d>k5$ imm$iately be=e excitation by ! bright object4 ,! ret9al n]ves1 :5 9 d>kness1 rapidly acquire a /ate ( s5sitiv5ess f> exce$+ ! normal av]age 9 ! light1 b quickly dim9i%+ aga9 "u ! influ5ce ( illum9ation4 ,?is peculi> s5sitiv5ess may1 inde$1 be bo? ga9$ & lo/ 9 a small fraction ( a second1 & is !re=e v fav\rable = ! rapid g5]ation ( negative aft]-images4 ,once m mak+ use ( ! black & :ite scre5s depict$ 9 ,fig4 #df1 let ! black scre5 fir/ cov] ! pap] ^u :i* ! waf] is ly+2 ?is w d>k5 a portion ( ! ret9a1 & r5d] x s5sitive4 ,!n let ! scre5s be quickly mov$ sideways1 s t ! waf]1 aft] hav+ be5 se5 = a mom5t ?r\gh ! op5+1 may be imm$iately cov]$ by ! :ite scre54 ,a bright b evanesc5t gre5i%-blue gho/ w succe$ ! r$ impression4 ,b ! mo/ curi\s ?+ is t if ! illum9ation is /rong1 & ! scre5s are mov$ at ! prop] spe$1 no trace ( r$ w be se5 at all2 x w appe> exactly z if ! actual col\r ( ! waf] se5 ?r\gh ! gap 7 gre5i%-blue4 ,i am in=m$ t analog\s ph5om5a h be5 obs]v$ 9 o!r bran*es ( physiology2 a well-def9$ reaction sometimes occurs :5 no direct evid5ce c be detect$ ( ! exi/5ce ( ! excitation to :i* ! reaction mu/ be due4 ,z 9 ! =m] exp]im5ts1 ! effect may be %[n cont9u\sly by means ( a rotat+ disk ) an op5 sector4 ,! annex$ diagram 7,fig4 #ej7 indicates a conv5i5t app>atus = ! purpose4 ,! disk is made ( ?9 metal1 & prop]ly balanc$2 ! d>k portion ( ! surface is cov]$ ) black velvet1 & ! light portion ) unglaz$ grey or buff pap]4 ,x %\ld turn "s six or eight times 9 a second1 :ile its front is well illum9at$ ei!r by bright diffus$ daylight or by a p[]ful lamp4 ,a r$ c>d plac$ beh9d ! rotat+ disk is made to appe> gre51 a gre5 c>d p9k1 & a blue "o yell[1 :ile a black pat* pa9t$ ^u a :ite gr\nd appe>s light] ?an ! gr\nd itself4 ,i h prep>$ "s designs :i* demon/rate ! ph5om5on 9 a v /rik+ mann]4 ,"o ( ^! is a picture ( a lady ) indigo-blue hair1 an em]ald-gre5 face1 & a sc>let g[n1 :o is repres5t$ z admir+ a violet sunfl[] ) purple leaves4 ,se5 ?r\gh ! disk1 ! lady's tresses appe> flax51 h] complexion a delicate p9k1 & h] dress a light peacock-blue2 ! petals ( ! sunfl[] also become yell[1 & its foliage gre54 ,o!r designs %[ equally rem>kable trans=mations ( col\r4 ,illu/ration3 ,fig4 #ej4--;,disk = trans=m+ ,col\rs4 ,i h m5tion$ only a few among _m curi\s ph5om5a :i* h pres5t$ !mselves 9 ! c\rse ( ^! inve/igations4 ,x is n improbable t a c>eful /udy ( ! subjective effects produc$ by int]mitt5t illum9ation w\ld lead to results t5d+ to cle> up sev]al d\btful po9ts 9 ! !ory ( col\r vision4 ,william ,byles ,sons1 ,pr9t]s1 #abi1 ,fleet ,/reet1 ,london1 & ,brad=d4 ,,footnotes3 #a ,x %\ld be cle>ly und]/ood t ! l5g? ( ea* wave ( a s]ies is measur$ by ! di/ance betwe5 ! cre/s ( two successive waves4 ,! l5g? ( wat]-waves :i* break ^u a sea %ore is n ! l5g? along ! cre/ ( a s+le wave measur$ 9 a direction p>allel to ! %ore1 z ! un9itiat$ are apt to suppose4 ,! true wave-l5g?1 or di/ance f cre/ to cre/ ( successive waves1 c be well obs]v$ f ! top ( a cliff4 #b ,9 practice1 wave-l5g?s are express$ 9 t5-million?s ( a millimetre4 ,! wave-l5g?s ( ! l9es ,a & ,h ( ! sol> spectrum1 :i* approximately co9cide ) ! limits ( visibility1 are #geid & #cifh t5-million?s ( a millimetre4 #c ,possibly ! human eye is at pres5t 9 process ( trans=mation f an inf]ior type to a diff]5t & m p]fect "o4 #d ,x is sometimes necess>y to place ! l5s ,i on ! o!r side ( ,k4 #e ,x is easy to f9d specim5s ( r$ & gre5 glass suitable = ?is exp]im5t4 ,! prop] k9d ( purple is n s commonly met )4 #f ,"s rec5t exp]im5ts on artificial col\r-bl9dness 7,proc4 ,roy4 ,soc41 ,feb41 #ahih7 h l$ ,mr4 ,bur* to ! conclusion t !re are really f\r fundam5tal col\r-s5sations--a r$1 a gre51 a blue1 & a violet4 ,8 results are1 h[ev]1 ?\ght to be capable ( a diff]5t int]pretation4 #g ,or ?r\gh sev]al pieces sup]pos$4 #h ,a violet-col\r$ haze may sometimes be actually se5 ar\nd ! opal globes ( ! electric lamps 9 ! /reets4 #i ,a focus electric lamp 0 us$ 9 ! lant]n4 #aj ,proc4 ,roy4 ,soc41 ,jan41 #ahii4 #aa ,aft] a few seconds' obs]vation ! gre5i%-blue col\r (t5 becomes mu* m int5se1 b ?is is an effect ( fatigue1 ) :i* we are n at pres5t conc]n$4 #ab ,see ,nature1 vol4 #ee1 p4 #cfg 7,feb4 #ah?1 #ahig74 #ac ,or1 = be/ results1 use a balanc$ metal disk cov]$ ) black velvet & :ite pap]4