iilil iMil I' ' ititM(uitiiti»&ii >|J' Em&mmtt^:S^fsx(M^lo^^i ^"Psjimental psychology and pedagogy; tor 3 1924 013 083 005 Cornell University Library The original of tiiis book is in tine Cornell University Library. There are no known copyright restrictions in the United States on the use of the text. http://www.archive.org/details/cu31 92401 3083005 EXPERIMENTAL PSYCHOLOGY AND PEDAGOGY Experimental Psychology and Pedagogy For Teachers, Normal Colleges, and Universities BY R. SCHULZE TRANSLATED BY RUDOLF PINTNER, M.A. (Edin.), Ph.D. (Leipzig) PKOI-KSSOR OK I'SV 23.'i i:38 Contents Xlll III JMkTHODS of llKC()(iNITIOX 1. Single Coml) illation ....... ((() Tests by means ol' .Stimuli \\'itli cmitinuous Changes (?') Te.sts l)y means of Stimuli witlioiit eontiniKius Clianges 2. (J()ml.)ination of a Series .... ((() The Melhoil of llerogniliou . (/>) The Method oF identical Itows IV. ^Methods ov Repiiuduction .... 1. The Scoring Method .... 2. Method of the Memory Span . 3. Method of retained Elements . 4. The Prompting Method .... 5. The Learning Method .... 6. Tlie Saving Method .... 7. Method of Reconstruction PAGE ^40 240 240 244 24.'5 24.5 24.5 240 248 249 250 2 5 2 252 254 254 CHAPTER IX APPKRC1<:PT1VE Ct)MBIN ATK )NS I. TfIK DNDKKLYIXO PrIXCII'LE« of 'I'HE MiO'lHODS IIF IXVKSTIOAT ING THE C'OMBINATIONS OF THE AlTEROEl'TlVE PliOCESK 11. Tachistoscopic Experiments .... III. Statistics of the ideational Pr(.(Cess 1. Free Reprodnction (a) Normal and Ahnormal ideational Processes (i) The Methods of Investigation 2. Constrained liepvoduction .... I\'. St,\TISTICS op PiEPlloDUOTlON TiMES V. IiIeTHOD OF TllIE i\lEASORI';MEXT IN' Rf I'lo lOFCTloN ExPEKIJIENT, 1. The Oraphic .Method 2. The Registration Method 30 ESS 258 259 2G0 2(i0 201 264 2(j(; 260 208 268 C; II AFTER X SPEECH I. Analysis of the Sounds of Speech 278 II. Analysis of the Melody of Spekch 280 III. Statistics of the Forms and Combinations of Words . . 282 IV. Speech as a Means of Expuession 284 xiv Experimental Psychology and Pedagogy CHAPTER XI PHYSICAL WORK I. The Ergograph 1. Ergographs with Weiglits 2. Ergographs with Spiings 11. This Measuremeni' of physical Ability 1. Ergograph Curves .... 2. The niaximuni Ahility . III. Rhythm and Work .... IV. The Sy'Mmetry op Movement PAGE 290 290 297 299 299 301 30(j 313 CHAPTER XII MENTAL WORK I. Methods of Investigation 1. Indirect Method.s . 2. Direct Methods II. The Interpretation of Work Curves 1. Mental Ability .... 2. Ideal Practice and Fatigue Curves . 3. Real Practice and Fatigue Curves . 4. Other Components of Work Curves 316 316 317 320 320 321 324 327 CHAPTER XIII PSYCHICAL CORRELATIONS I. The Calculation of Correlations .... 1. The Correlation Formula ..... 2. A Supplement to the Correlation Formula 3. Correction of the Supplenuuited Correlation Formula II. Correlations in Psychology 1. Krueger and Spearman's Results .... 2. Ohrn's Results III. Correlations in Peda(io(iv 330 331 338 340 341 341 344 346 Coniei/is xv APPENDIX I f AGE A New Chroiioscope .......... 3.50 APPENDIX If List of Ajiparatus for Colleges or NuMiial Scliiiols 3.5(j Index ........... 3.59 TABLE FOR CONVERSION OF THE DECIMAL SCALE TO ENGLISH MEASURES 1 Millimetre = -039.37 inch. 1 Centimetre = -.393708 inch. 1 Metre = 3-2808 feet. 1 Kilometre = 1093-633 yard,s. 1 Gramme = 15-432.5 grain.s troy. 1 Kilogramme = 2-2055 lbs. avoir. ILLUSTRATIONS ruts FIG. 1. Barnes' spirometer 2. Flint's stethoineter tn measure the ilianieter of the c 3. Peterson's head measure ...... 4. Bertillon's compass lor heail measui'i 5. Kreinlein's eephalometer (i. Collins' (lyuamniiicter .... 7. (Oilman's (lynauiomi'ter fnr si(ueeziiif; 8. Ullniaii's ilynamometer for s(pieezini^- am 9. Skulls oF one-year-olil rliihl, teii-year-ohl 10. Curves of the ahsolute annual inen-ase in hci^lil children ........ 11. Increase in weight and height of Berlin eliildreu . lest dliug . ihl, aiK adult f .lena od ^'ht-year-old girls md joint nf 2]' I'v ids of the l'ln'iix,i,ill,,,N ,sV,/ '"" V rf,ll,il, Its 12. Distriliution curve of eiglitx' nieasui'emeids . 13. Error enr\-es of two nhstu'vers 14. Distribution curve of the lieiglit of 2i;i2 c Louis according to (lauss's law . 15. Distributinn curve of the lengths of tin- stalks from a tiidd near Leipzig , 16. A.symmetrical distribution curve . 17. Number of i-adiate llowers oi 17,0n0 1i h.'iiniittlii III ii'W according to Lndwig 18. Variation cui've of a plant 19. "\'ariation ciii've of the filaments of lli iiilluenci; of chemicals '20. Distribution of the ability to copy out 21. Distriliution of 1794 jiiilnres ....... 22. Distribution 01770 pictui'es ....... 23. Distriliution curve with t«"ii summits for addition for si.v iioui 24. I )istriljution curve of the height of Freiberg elemenl;uy sell children .......... 25. Height of Freiberg element;ii'y schoohchildren I 20. Growth in height of 1000 cliiidren from age (i- 27. Distribution curve of tlie aliility of a school before and after systematical practice . 28. Distriliution of 1334 thirteen-year-ohl girls ol S 29. Colour mixer .......... 30. Colour nuxer on stand ........ 31. Investigation of the stimulus threshold for colours xvij ^ rnm age 12-lri b». . -class in additioi t. Lcmis in grades rAGE 10 10 11 11 12 12 12 12 13 14 1.5 18 29 34 38 3S 39 41 42 43 44 46 51 51 xviii Expcriiucufnl PsycJwlogy and Pedagogy FIG. 32. f'olour ilisc far investigating the stinmlu.s threshold 33. ('olour discs for investigating the differenee sensitivity . 34. Colour discs for investigating stimuli that ajijiear ei^uivalent 35. Coloiu' disc for investigating differences that appear equivalent 36. Verke^' arrangement fur determining the power of discrimination of colours in nnce ........ 37. Von b'rey's stimulus hair ....... 38. Von Frey's hair-a;sthesiometer ...... 39. Investigation of pressure .sensitivity ..... 40. Investigation of the cold spots of the skin .... 41. Cold and warm spots o)i the .same part of the arm . 42. Von Frey's thermometer-rod ....... 43. Kiueniatometer ......... 44. A patient lacking cutaneous sensitivity and joint and muscle sensitivity in the right ar)n ...... 4.5. Voung children exercising the large joints .... 46. Difference sensitivity in lifting Aveiglits . .... 47. Zotli's arcoumcter ......... 48. Tuning-forks to determine difference sensitivity 49. Spearman's festhesionieter ....... 50. Ebbinghaus' icsthesiometer ....... 51. Determination of the tartual spatial threshold 52. Children modelling ........ 53. A blind girl reading ........ 54. Apparatus for testing the visual spatial threshold . 55. The Miiller-Lyer illusion 56. Apparatus fur testing the accuracy of our perceptions of depth 57. Heading les.son in an idementary class in the "Werner-Siemen; Realgyiunasium, Beilin ....... 58. Jletronome with electrical contacts ..... 59. Tapping in rhythm at ]ileasure ...... (10. Tapping arcording to a [iii'scribed rhythm .... 01. Fnd of curve in Fig. 60 ........ (j2-03. Di'awings from memoij by a Hamlnirg child . 64. Drawing from memory by an in.sane patient .... 65. Drawing from memory by an insane jtatient .... 66. Drawing from meumiy liy an insane patient .... 67. Development of the sense of form of a sister and lirother from thi age of 5 to (i (18. Arrangement for experinn-nt to investigate the pulse and lireath ing while looking at pictures ...... 69. Kymogiaph and I'ecoixling .apparatus for experiment shown in Fig. 68 70. "The Young ifan of Xaiu " ....... 71. Breathing and pulse curves while bulking at the picture " Tlie Voun"' Man of Nain " PAGE 53 53 55 56 66 80 81 82 87 87 89 90 91 92 93 95 96 97 98 99 101 101 101 104 105 106 107 108 111 112 116 117 11 lustrations XIX Fir:. 72. 73. 74. 7(). 77. 78. 79. SO. 81. 82. 83. 84. 85. 80. 87. 88. 89. 9(J. 91. 92. 93. 94. 95. 96. 07. 98. 99. 100. 101. 102. 103. 104. 105. 106 108. 109. 110. 111. 112. 113. l)irtiue " Aliriili.iiii am lokiii;^ al I 111' pici iirr timulation Brralhiiig (;ur\e wliilt' looking a( tlic iiirUu-e "The \'iiiiii (iF Naiu "........, "Abraliaiii ami Lnt," . iiivalliiiig and pulse uiii\-f.s wliili: loDking al the ]iicUire "Ah and Lot "......... Bi'eatliiiig rurve while looking at tl "The Stiirni on (lie Sea of Galilee' Bi-eathing and pulse curves while looking al Storm on the Sea of Galilee" Investigation of the jiulse and hreathing durin sense of taste ........ Pulse changes while tasting aloe ..... Pulse changes while tasting sugai Kymograph and apparatus for I'eeordiug the ]iulse Marey tambour Smoking the drum . Fixing and drying the eurves ..... Glass plate with nullinietre scale for measuiing the cur\e Gardiograp)h for investigating the heating of tlie heart . C'ardiograph ....... Plethysmograjih ...... Plethy.smograph ...... Demonstration of the hreathing rur\e . Pawlolf's method of measuring the How nf sali\'a Lombard's recorder ..... Ordinarx' recording magnet .... Normal eur^-e ....... Normal eur^'e ....... Normal curve ....... Normal curve ....... Holding back the bi'eath and its influence mi lie.- pul,~e Normal curves ....... Taste e.\]iei'iments ....... Attention ........ Reckoning ........ Pulse cur\'e before tie 200-nietre race Pulse cur\e immediately after the race . Breathing curves before and after a cycle run h and 107. Breathing curves befoi-e and aflei' beginnei- ....... "The Valley," by jlein " Schiller," fiy Bauer ..... " Autumn," by ( liilieb ..... A sweet taste— sugar ..... A sour taste — lemon ..... A l.iittei- taste— aloe about 20 ki I cycle run ihaii bill ''i'h. f the 1 bs 1 I II 119 liill 121 122 123 126 127 12« 129 130 130 13! 132 132 134 136 1 3-S 139 1-1! 141 142 1 43 143 144 I4.'i 14(; 147 147 148 14S 14^ l.'.i.i l.'il 1.13 1;-..-, 1.-..-, 15.". XX Rxperinieiifal Psycliology and Pedagogy 114-127. Wlmt fi-L-lmssao tlie cliiUh-en show ^ .... 127-131. Wliiilfeelin.us do tliecliil.lreii slinw? .... K!2-].'?5. Pautniuiniii' (.■.v])re.ssinu iiioveiiieuls wliilc lo(,ikiii[,' ;il ]iii:turcs VM\. "Tl)..' . Kymograph .......... 14(;. S])iing kyniogra]ih ......... 147. Uidinary recorder ......... 14!^. Klectro-niagnetic tuniiig-l'orlv ....... 14IJ. Testing a tuning-l'ork with tile Jaijuet. clirononieter I.')!). Tuning-fork oscillations and twi.i marks oF llie .huiuel ehrono meter ........... l.')!. A simple leaction experiment ...... 152. Contact key 153. A simple reaction experiment ...... 154. Electro- magnetic sound hammer ...... 155. Acoustical reaction hy means of the sound hammer 15(3. Hipp chronoscope ......... 157. Wnndt's control hammer ....... 158. Ebbinghan.s' gravity apparatus ...... 159. Five-finger reaction keys ....... 160. Diagram of frequency curves in reaction exj eriment.s, according to Wundt 1()1. Distribution curves of nuiscnlar, natural, and sensorial reactions 1(J2. Distribution curves of the reactions of an observer with decidec mnscular reaction tendencies ...... 163. Distribution curves of the reactions of an observer with decidei' sensorial reaction tendencies ...... 164. Progress of jjractice in changing fi'uin decided muscular to sen- sorial reaction ......... 165. The ti-aining of the will liy games ...... 16f). Optical attention ......... 167. Kighteen-month.s-old child looking at a tl>ing swallow . 168. The same child, a ivw moments later ..... 160. Blind children listening 170. A deaf and dumb boy reading from the lips of his teacher . 171-172. Defective children . . . . ■ . 173. The mimicry of attention of eiglrt-year-uld childi-en wdien think- ing of something easy ........ 174. The strongly mai-ked mimici-y of eight-year-old children in a stati of concentrated attention ....... 175. ( 'ontraction of the mtisciiliix Ji'mital'is on command . 176. Sommer's appai'atns for analysing movements in [m, dimensions PAGE 156 157 160 16] 161 161 162 165 166 167 16M 16S 169 170 171 171 174 174 176 17H 179 l-SO 182 183 184 185 185 186 1!)0 192 193 193 194 195 196 197 198 10!t 200 Illiisf rations XXI FIG. 177. 178. 17!J. 1 SI I. LSI. 1S2 183. 1S4 1 S6. 187. 188- 193- 19."!. 196- 198. 199. 200. 201- 204. 205. 206. 207. 20S. 209. 210- ' 212. 213- 215- 217. 218- 221- 223. 224. 225- 227. 228. 229. 23(1. 231. Si>iiinii.'r's ;i[ip;iratus foi' aualysiiit; llie miivriiienl s of \\\it miisclc: uC tin- fnlvllrlld Sdiuiiier's ti'idiiueiisiiiiL'il analyser ...... S(iii[i)u'i-'s triiliiiien.sional analyser ...... I 'uivu of Summer's tiidimeiisional analy.ser .... Demoiistr.ilioii taclii.^tuscoiH-, acrmiling to ^\'umlt . Olitiial stimuli for iletiTuiining tlio ,sco]ie (jf cnnsciousiirss Assimilating activity when lonkiiig- , it a jiicUiie anil 185. Assimilating powL-r of an iilea ..... Pictures for assimilation experiments, accnrilmg to van tier Torrei " Pea.sants in Corntiekl," by Thoma 192. Drawings l.iy two pupils, showing visual ami constructive type ■194. Drawings of a vase by two ]iupils ..... Drawing of the co]"iy shown in Fig. 188 ..... -197. Drawing from memory a diagram as if seen ni a mirror . C'oufusion nf side and bird's-eye view in a cliild's drawing 'I'lie same confusion of side and liird's-eye \iew in a drawing liy Dakota Indian ......... The same confusion as in Figs. 198-9 . . . . 203. Persevering ideas as distractions of perception Productions of the visual memory of young children Drawings of a blueliell ........ Ranschburg's memory apparatus ...... Disi' for memor\' experiments ...... Ranschburg's memory apparatus arranged for an experiment Wii'th's meiiioi'}' ajijiaratus ....... ■211. Wirth's memor\" a]i|iaiatus with long p,-i]ier strips . Midler's meiiior\' apparat us ....... 214. Xew memory apparatus for pc(lagogical experiments 210. The u]iiier and lower contacts of the new memory appiavatus .......... Testing the nuuiiory ........ 219. Visual objc-ets for testing tie- power of noticing of tl insane, ;iccnrding to rjernstein ....... 'I'ablc for testing the power of uotiiiug of the insane, according to Bernstein .......... 2. The iulluenee of grouping of rows of twelve syllables on the memory ......... The memory of ,girls ........ The memory of boys ........ 22(). The first recollections at dilfercnt ages of hoys and girls \'i.-ual object to test judgment of identity .... Drawing l^y an insane jierson . ...... Drawing by an insane person in a state of command automatisn Drawing li\" an epileptic boy ....... AHier's I'ptieal stimuli apparatus ...... 201 202 202 21)3 21 15 21 10 211 212 217 218 223 224 225 225 226 226 227 227 233 234 235 236 236 237 237 238 239 24t) 241 243 244 251 255 255 256 26i;i 262 263 263 209 xxii ILxperiiiicntal Psychology and Pedagogy Flu. 232. 233. 234. 235. 230. 237. 238. 239. 240. 241. 242. 243. 244. 245. 240. 247. 248. 249. 250. 251. 252. 253. 254. 255. 256- 256. Minuciiiiaiiii'.s caid cliaiiger ....... Diagram of Heiiipel',s voicc;-key ...... Hcmpers voice-key ......... Arrangement of cxpuriment for mea.suriiig ruproiluction timer with the grapliic nielhod ........ 1. Rixiujidtiuii Times. Watei' — ^\atel■, '8 sec. ......... •Jveii — oven, '9 .sec. ......... Evening — evening, 1 sec. ........ 2. Free lieproduetion. Key — cuphoard, 2'1 sees. ........ Table— chair, 2-5 Rose — garden, 4'6 sees 3. (Joiistraiiied lieprmhietion . Tree — branch, 3'3 sees. ......... Horse — liead, 4'0 sees. ......... Garden — llowei', 4'5 sees. ........ 4. Conxtraiiied lleprutliictioii. Roof — house, 4'2 .sees. ........ Leg — body, 4'() sees. ........ Back — chair, 7-3 .sees. ........ memory memory apparatus Perception time. "Ap])le" — "Yes." Measurement of reproduction times with Kanselibur apparatus, chro]ioscope, and voice-key Measurement of reproduction tinies with chronoscope, and voice-key Ivrueger and Wirth's larynx sound recorder . Writing capsule of the larynx sound recorder Marbe's apparatus for the melody of speech . Demonstration apparatus for schools .... The body as a means of expression in the games of childre -259. Recitation without ge.sture ..... Lovely was the evening, Silver clouds tle-H' by (.)\er all the .sprint; landscape JuijDiis in tlie xkij ..... 257. On t,he side of yon steep hill Were gra\es, where dead men lie, And on the wall the Cross stood there In silent giief on hiijh .... PAGE 269 270 271 27 2 273 273 273 273 273 273 273 273 273 273 273 274 276 278 279 280 281 285 286 286 Illusfrafions XX111 258. 259. 200. 261. 262. 263. 264. 265, 206. 267. 208. 269. 270. 271. 272. 273. 274. 275. 270. 277. 278. 279. 2H( ). 281. 282. 283. 284. 28,'-|. 280. 287. 28,s. 289. 290. 291. 292. Tlie po.sl-lioy Iieeded not, hut crdi-Lcd His ifliAp with all his will, And let his horn rinf; inen'ily Over dale, and hill .... On we driivu with janglinj,' nnisr Past fudd and wood and rill, l>nt long there sounded in ni\- ears Thi' tmic. from yrni, dee'p hill Poem as in Fig. 250 ...... Poem as in Fig. 257 ...... Poem as in Fig. 258 ...... Poem as in Fig. 259 Mosso's ergogiviph ....... Arm-rest for eigograph ...... Piecording arrangement of an ergograph with an endless nieasni ing tape ......... Meumann's ergograph ....... Ergograph for investigating the working power of the bieep; Dnbois' ergograph ........ Ergogramm on millimetre paper Demonstration of an ergo.graph curve on the kymograph Lehniann's ergograph .... Prof. Heniy's ergugraph .... Normal ergogra])li enrve Practice curve . . . . • Fatigue curve ...... Fatigue oirve with .short intej'vals of rest Muscle jiower before h-arning by heart . Muscle power after learning by heart Lifting 4.', kg. till exhaustion, then continuing willi \}, Infinite curve f th Continuous work on the ergogra])h with gradual deci weight according to Treves' method . Ergograph curve of a hysterical patient Ergograph curve of a ]iatient suffering from chorea Normal curve ........ Curve of a sham patient ...... Normal and sham curves ..... The first ergograph curve of .a niue-year-ohl child witl scribed rhythm ....... Single pulls on the ergograph without prescribed rhyth Single pulls on the ergograph with prescribed rhythn Kraepelin's writing apparatus .... Pre.ssure curves while writing the figures, 1, 2, 3 by normal men and women lOUt 286 286 287 288 288 289 291 292 292 293 294 295 295 290 297 298 300 300 300 300 301 301 3112 31 )-J. 302 304 304 31 15 305 305 307 307 307 309 310 XXIV Experinieutal Psychology ami Pedagogy tunes time, Noiiiia Alcnh FIG. 293. Pressure curves wliile writing tlir tigures ], 2, 3 liy an insau patient ....... 294. Pressure curves wliilc writing the figure 8 tlin-e experiment ...... 295. Pressure curves wliile writing llic tigure 8 three experiment ...... 296. i (with dot) written four times 297. i (without dot) written three times 298. i (wiih dot) written three times 299. i (with dot) written once .... 300. Involuntary symmetry of movement when mndelh 301. Co-ordination of symmetrical grou])s of muscles 302. Practising the left arm after drawing with both ar 303. Kraepel ill's electrical pencil .... 304. Practice and fatigue curves .... 305. Ideal practice and fatigue curves , 306. Mathematically constructed and real work curves 307. Practice curves in receiving and sending telegrams 308. Physical fatigue ...... 309. Effect of massage on the work of a muscle 310. Components of the work curve, according to Krae 311. Correlation between ilitferent mental abilities 312. A new chronoscope ..... 313. A new chronosco]je ..... 314. Testing the new chronoscope .... 311 311 311 312 312 312 312 313 314 314 31 S 320 322 323 324 326 327 328 345 351 352 353 EXFP]RIMENTAL PSYCHOLOGY AND PEDAGOGY TNTKODUC'JION I. The Principles of Experimental E,esearch Descartes' famous saying, " Dul^ito de omnilnis," is not only useful for the building up of a system of plidosopliy, but also for the foundation of a new science. For in both cases will he fare best who acce^^ts no general statement without sufficient proof. And so it often happens tliat such a person finds the most important questions and the most difficult problems in things that " common sense " takes as a matter of course. Psychology has been trying for more tlian a generation, and pedagogy for more than ten years, to liuild up anew their departments of knowledge. This being so, both sciences can make good use of Descartes' famous saying. Let us see wliere the use of this principle in psychology and pedagogy will lead us. Let us suppose we ask somel^ody the very trivial question whether the taste of sugar is pleasant or un- pleasant. This cjuestion is purely psychological, because we do not seek any information al:)out an outside object or about an objective process. We seek information about a process that takes place in our consciousness, and we do not raise the question as to the possibility of an objective process (say, a certain process in our nervous A 2 Expcrijiicnfal Psychology and Pedagogy system) being connected with the subjective process in our consciousness. Tlie answer, " Sugar tastes pleasant," will Ije given in the great majority of cases as if it were self -understood, and it will be difficult to persuade anyone that to raise a doubt here is anything more than a laughable eccentricity. But what if a real doid,>t did arise ? The doubter would, of course, test the taste of sugar again, and then ujjhold his jjrevious judgment with greater certainty oii the ground of his latest experience. How does this greater certainty arise ? Partly because he has had direct ex- perience of the fact, but chiefly because he has paid special attention to it. Two conditions were carried out in this little experiment. Firstly , the individual in cpiestion knew the moment when the process to be observed would l^egin , and therefore could fix his attention on the expected occurrence. What would happen if we were to fall upon the person, so to speak, from behind with an unexpected stimulus, can be seen in the well-known experiment of mixing marzipan with potatoes. The sweet taste in this case can be absolutely repulsive. It is not to be wondered at, then, if we demand in a scientific investigation a signal for the attention — "Ready" or "Now" — before every experiment, and also a definitely measured off length of time (one to two seconds) for the attention to get fixed. Secondly, the attentive observation of the phenomenon must not l)e disturbed by anything. No one could ex]>ect anyone suft'cring from toothache to take part in our exjoeriment on tlie taste of sugar. So also, no one would think of helping the observation of the process in questioii, by suddenly firing off a pistol near the subject of the experiment. On the contrary, everv- thing that might distract the attention would be avoided. Introdiictioii 3 We liave thus arrived at two important rules of metliod ; — 1. The observer must be prepared for the occurrence of tlie event. 2. The observer must follow the process of conscious- ness with close attention. But Descartes' famous saying will not allow us to rest satisfied with a single experiment. The only possibility of freeing ourselves from lasting doubt lies in several repetitions of the experiment. Where from the nature of the case such a repetition seems impossil^le, as, for example, in dreams, we can never arrive at scientifically useful results. This is the reason for the aversion of our science to the investigation of so-called abnormal states of consciousness. We therefore set \\\) a third rule : — 3. Every observation must, for tlie sake of certainty in our results, be able to be repeated several times under exactly the same conditions. It is oidy when we have carried out this principle that our metliod of procedure jjecomes what in the empirical sciences is called a scientific method of investi- gation. And in most cases it leads to a real extension of knowledge. And so it is in regard to our simple question. We discover first of all that we have forniidated our question wrongly. For it appears that in tasting sugar a great number of sensations work together — sensations of taste, of touch, and of temperature — and that we ariive at quite different judgments as to our feeling, according as these sensations are mixed. A warm solution of sugar tastes otherwise than a cold one, for examiile. And we can never come to any result, if we do not tlioroughly analvse the occurrence into its difi'erent elements and 4 Experimental Psychology and Pedagogy investigate each particular element separately. There- fore our question must be first of all only : " Is a sweet taste pleasant or unj)leasant ? " Sensations of touch and temperature must be as far as possible prevented. There would remain as another cpiite different experiment, the investigation of the combined influence of these different elements upon our feeling. (Synthesis.) Our fourth rule will then run as follows : — 4. The investigation must start with the elements of consciousness, and afterwards proceed to the investigation of the complex processes. I will give here a number of statements that I obtained in experiments on the influence of sweet and sour taste on tlie feelings.^ Subject A., Solution of sugar: — "A feeling of pleasure has not arisen. Perhaps I was dis- appointed at the weakness of the taste." "^ On another occasion, with the same solution of sugar, the judgment " pleasant " was given. Because of the uncertainty of the judgment, T then took a stuff with a stronger sweet taste, i.e. saccharine, and first gave a weak solution. The judgment was : " The stimulus is too weak ; no positive feeling of pleasure, rather the ojj^wsitc ; sweetly oily." I then gave a stronger solution. Answer : " Bitter- sweet, unpleasant." With a medium solution, the judgment was : " At first unpleasant, then it slowly changes to a weak feeling of pleasure." And so we arrive at the method of giving a graduated series of sensations of sweetness, from a very weak to a ' The clianges of the jiulse and breathing that took place during the experiment were als(j registei'ed. (Of. pp. 123-124.) ^ The same subject in the next experiment, in ■\vlricli a solution of vinegar was given to taste, remarked : " This ta.stes refreshing, and not unpleasant. More pleasant than in the former experiment." The tempera- ture of botli solutions was exactly the same. Introduction 5 very saturated solution. We obtain tlien a graduated series of judgments corresponding to tlie series of stimuli. Comparing these two series, we find a certain regularity in tlie relations between stimulus and sensation, and between sensation and feeling — a sort of psychological law, which for .Subject A. would run somewhat as follows : — Weak solutions of sugar taste indifferent or unpleasant, very strong solutions unpleasant. Solutions of a middle strength taste pleasant. (The strength of the solutions would be given according to the percentage of sugar contained.) On the basis of these considerations, we arrive at a fifth and last rule : — 5. We must in our experiments be able to alter the conditions we are investigating, under which a process of consciousness takes place, according to a definite plan, so that we may be able to deduce certain regularities from a comparison of the graduated series of stimuli with their corresponding effects in consciousness. If we casta glance over the five rules we have formulated, we find that they form a method of investigation that goes by the name of " experimenting." ^ This method of investigation has long been accepted as the best in all natural sciences. Both descriptive and explanatory natural sciences have arrived at their most important results by the aid of experiment. By applying Descartes' rule to our simple psychological problem, we have been able to deduce all the conditions necessary for an experiment. We have arrived c[uite naturally at the experimental method. The method has grown as a matter of course out of our problem. We therefore conclude that this method is the natural method for psychological investigation. 1 Wuiult, Psij,:hoh)ijisi.-hc Utiidim, Cd. III. 1907. 6 Experimental Psychology and Pedagogy It may not be possible in every psychological experi- ment to obey all the five rules we have formulated. There are perfect and imperfect experiments. If in a particular case only three of the rules can be obeyed, then we must endeavour to carry out these all the more conscientiously. II. The Divisions of Experimental Psychology AND Pedagogy In the above-mentioned experiments on taste, certain striking individual differences came to light. For example. Subject B. gave this judgment about a certain solution of sugar : " This kind of sweetness is unpleasant." Whereas Subject C. remarked about the same solution : " Decidedly pleasant." Here we stand on the threshold of the psy- chology of individual differences, or individual psychology. It further came to light that the greater number of the male subjects of our experiment felt the sweet taste to be scarcely pleasant or absolutely unpleasant. We have, however, reason to believe that for all young children a sweet taste is pleasant. And this brings us to the problem of psychical development, of ontogenesis, of the develop- ment of the individual. The most important part of ontogenetic psychology is, of course, child psychology. And so through our simple experiment we have arrived at the following branches of experimental psychology : — 1. General psychology. 2 Individual psychology. 3. Ontogenetic psychology ; in particular, child j^sy- chology. One might also add phylogenesis, or the development of communities — a family, a people, a race, or mankind. The most important part of phylogenetic psychology is racial psychology. Experimental investio;ation is here. Ijifrodiiction 7 of course, almost entirely impossible. But then the experimental method is not an aljsolute necessity. For the processes of consciousness do not, as in the above- mentioned diA'isions of psychology, form the ol)jects of investigation, but rather dchnite mental products, such as language, custom, religion, and these are relatively constant objects in co]nparison to the ever-flowing ])ro- cesses of coirsciousness.^ A method of simjile oliservation is deemed sufficient for those natural sciences that deal with constant objects, say, for example, in the investiga- tion and d('Scri[)tion of a mineral, whereas in investigating processes — ajj. a. cliemical change — tlie ex]>erimental method is absolutely necessary. So in ]-acial ])sychology the method of observation is sufficient, and is indeed the only possil)le nrethod. Moreover, racial psychology is only then of interest to pedagogy, wdien we can obtain definite resrdts as to the laws of ontogenesis by a com- parison of the mental products of the child with those of different races — c.f/. a comparison of the drawings of children with those of ])rimitive races. Pedagogy, to gi\'e a. very general definition, deals with the investigation of those metliods which must be made use of in order to influence in a definite nranner the de- velopnrent of a human l)eing. This definition of pedagogy is different from that of nornuxtive pedagogy, wliicli first sets up special norms or rules, and tlien seeks the means by which these ends are to l)e attained. The nornrs are taken from the normati\'e sciences — e.(j. ethics. On tlie other hand, our idea of ])edagogy is that it must take into consideration all the possibilities of development, and then investigate by what ways and means these latent possi- bilities of development may be helped. Pedagogy has therefore first of all to test the ends set up by the nor- 1 (■/■ Wiuidt, Oiit.lini:^ 11/ Psijrholiiiiii, .luiM's tr;uisl;itiou, '-ii'd edition, IK -20. ' 8 Experimental Psychology and Pedagogy mative sciences, and find out liow far tiey can be attained ; and furtlier, it must investigate whether the attainment of the ends in question can be influenced by educational measures, and if so, at 'what stage of development these measures must be begun, and which stages are most favourable for further development. If, for example, the normative sciences demand a development towards morality and religion, pedagogy has first of all to answer the following questions : — 1. What are the psychological foundations of morality and religion ? The answer to this question must be sought for in general psychology or in racial psychology. 2. What individual differences are there ? (Moral insanity, &c.) Individual psychology must answer this. 3. How do morality and religion develop in the in- dividual child ? At what stage of development does an understanding for moral and religious questions arise ? At what stage is the interest in such questions greatest ? Child psychology must answer these questions. Only after settling these preliminary questions does the real pedagogical investigation begin. Since this always has to deal with stages of development which are meant to be influenced by definite pedagogical measures, the experi- mental method should be made use of whenever possible. The reason why this has seldom been the case up till now, is that this method had not been perfected. First of all, experimental pedagogy has to settle the cjuestion by experiment, whether the mental and physical abilities can be influenced (by educational measures), and if so, to what degree. Then it must test the iir dividual existing methods of education in respect to their useful- ness in helping to attain a special end and in helping to attain the general end of education. Introduction in. Anthropob'Ietrical Measurements Pedagogy lias in general to deal with the develo]:>nient of the whole human being, and in ])articular with the de- velopment of his mental faculties. Therefore, it has a direct interest in the development of the liody, in as far as the training of the child's bodily faculties is one of its tasks, and an indirect interest, in as far as physical and mental development influence each other. This may show itself in a parallelism between the two, or in the fact that periods of increased pliysical development are accompanied 1)y decreased mental de\'elo]nnent, and vice versa. A further f(uestion arises, as to whether physical development is hindered by overmuch instruction — e.g. by admitting children to school at a very early age, or by overloading them with mental work. From all these considerations, is anthropometry, the science of measuring the human body, of importance to pedagogy. The most important anthropometrical measurements for pedagogy are : — 1. Height. With bare feet, heels close together and knees imbent. 2. Weight. This can best be taken during the swim- ming lesson. The swimming costume should be of a definite weight for all. 3. The capacity of the lungs. The exact ainount of air exhaled can be measured by the spirometer (Fig. 1). The air enters an inverted air-bell, floating on water, and thereby moves the scale np, so that the number of cubic centimetres can be easily read off. The use of this apparatus with children presents certain diffi- lo Experimental Psychology and Pedagogy culties, so that measurements (with an ordinary tape) of the circumference of the chest, when the lungs are full and when they are empty, may suffice. During these measurements, the attention of the child should be diverted from his breathing, so that this may take place in a natural manner. The average of the two values I'll!. 1. — Barnes' spiroiuetcr. FlC. 2. — Flint's stethonietev to measure the (lianieter of the chest. thus obtained serves as a measure of the circumference of the cliest.^ The diameter of the chest can be measured by the stethometer (Fig. 2). 1 (,)iiirsfeld, "Ziir physisclieii unci geistigeii Eiitwickluiij,' ck'S Kiudcs" (Zrilsclirift fiir Scliiihirxiiiiilheitsfifliye, 1905), maintains, on the liasis uf liis investiealions, lliai tlie circiniifefence of the cliest has no direct inlluence (111 till' vital caiiacity ol' tlie lungs. Iiiti'odiicfioji 1 1 4. The .size of the head — (rf) The circumference, by means of a steel measuring tape. (6) The length and breadth, by means of a rod (Fig. 3) or a compass (Fig. 4). I-'IG. 3. — Petcr.son's head measure. l''lG. 4. — Bertillou's eetiipass for iieail lueasureineiil s. (c) The height, by means of the anthroj^ometer. (Fig. 5 shows a measurement of the head with Kronlein's cephalometer.) To establish cases of under-nourishment and of weak- ness due to illness, a measurement of the hand jjressure is important. 5. The pressure of the hand can be best tested with the dynamometer. This consists of a strong spring, which is pressed 12 Experimental Psychology and Pedagogy together by the hand. This causes, by means of a small tooth-wheel, two pointers to move along a scale. When the pressure ceases, one of the pointers remains stationary at the maximum point, and then the greatest pressure attained can be read off in kilo- grammes (Fig. 6). Figs. 7 and 8 show dynamo- meters for pressure and for pulling. Fig. 5. — Kriinlcin's cephalomuter. Fig. to the a,L;e of 11. Fi'oiu tlien on to the age of 15 they gvow (luicker than boys. 14 Experiinciital Psyclwlogy and Pedagogy These relations differ according to sex, race, climate, and especially in regard to the environment in which the child grows up. Poor children develop slower than rich ones. In general, the fact is established that the period of A AO ■■' \ ; •', : ; : A. ,> 6.0 ,; f\> ^ ; / — X- ; / xv. \ ; ; / \- ' ■/ Y 50 / \ ' .-■ ' ' / / °\ -. H / ' \ t,0 S\ /7 1 \ \ s '•^ ,* v»--5 \ / ^ V. '.'■'/ \\ V ••• / \ J.U \, \ > N \, zo c s 1 -,— ;, 1.0 ■^ .- lA ) '\ ^s KiG. 10. — UurvL'S of t.be absolute annual increase in lieight of Jena scbool-cliililrcn. (From Koeli-Hesse, Zeitsrjn-ift fiir iTcsundJifilsjifli'i/r . Voss.) maturity is the period of most rapid growth. Another maximum appears between the ages of 8 and 11. Between these two periods lies a period of slow growth. The pedagogical importance of these facts is self-evident. This rhythmical progress can be shown in all anthropo- 1 iii rodnction 15 metrical mcasiircments, although, of course, the maxima by no means always fall in the same period. It is well Uci-hi, \ l>,-,.,,;,nilor,v an.l +++ + + Ucighl, | KlumnilaiT Wri-hl. ^ Hi-h S.-.liuuJs. WiUK'lit f ''^^■l I^- l''li;. 11. — Inurrasc in weight ami heiglit ul Loilin childri'ii. (Im- ItictZ, .\ir\ur fill- Anihrni.olniji,-, Bd. 2i.l. VicWCg & Snilll.) known that the growth in height conies to an end in the period of youth, while great changes may afterwards take place in regard to weight. 1 6 Experimental PsycJiology and Pedagogy And so it is with mental development. Therefore it is our task to fix tlie general mental condition at each grade of instruction, and also the mental capacity of the child for each particular subject at each particular grade. Besides these large fluctuations in physical development, smaller fluctuations within the period of a year have been observed. It is, of course, extremely difficult to arrive at uni- versally valid results, because each child again shows in- dividual peculiarities in its growth. So we cannot be surprised that the science of anthropometry lias up till now been able to set up very few universal laws. For example, it cannot yet definitely answer the question as to whether the first school-year hinders the physical de- velopment of the child or not. Most important for all such investigations is a familiarity with the methods of mathematics. For only then can we make our figures alive, only then can we make them speak. CHAPTER 1 THE MATHEMATICAL TliEA'r:\lEXT OF RESULTS IN CRILI) PSYCH(JL()(:iY AND TEDACKJGY I. MEASUREMENTS IN PHYSIOS 1. The Law of Error As soon as measurements of any kind arc undertaken, errors are bound to appear. Speaking strictly, no measure- ment is absohitely free from error. In measuring the length of a certain rod, for example, I got the first time lOO'l mm., the second time 99"9 mm., and so on. There- fore, if I wish to make a definite statement as to the length of this rod, I must take a mmiber of measure- ments. In measuring it eighty times, I obtained lengths that fluctuated irregularly from 99-6 to 100-4 mm. The different tenths of a millimetre between 99-6 and 100-4 mm. did not all appear the same number of times, but as follows : — I.fllglllS. 'M-C, iimi. !J9-7 „ OifS „ 9!)-9 „ 100 100-1 „ 100-2 „ 10(1-3 „ 100-4 „ No. nt Ti If I measure off the tenths of a millimetre on a horizontal line (abscissas), and the number of times on a vertical hue E 1 8 Experiincntal Psychology and Pedagogy (ordiuates), I can draw a curve. (Fig. 12.) We see at once a certain regularity in the distribution of the errors, although during the measurements they appeared to arise quite irregularly. The " error " curve is absolutely symmetrical. The riglit half is exactly the same as the left. The curve rises at first very slowly, and then more cpiickly. In the middle, at the top, it is comparatively flat. Now, this kind of curve appears in all kinds of measurements. It is the same if I measure the length or the breadth or any other quality of an object. From this regular distribution of errors, we arrive at a certain 99,6 99,7 593 99,9 lOO lOO.i 100,2. i00,3 lOO.t Fig. 12. — Distribution curve ol eighty measurements oi: a. rod. law, which is called tlie Law of Error,^ and to which all physical measurements are subject. The working out and proof of this law belongs to the Theory of Probability. Without goiug deeper into this theory, we can nevertheless obtain an insight into the nature of the Law of Error by a few elementary considera- tions. First of all, it is easy to understand that our measin-emeuts of the length of the object are too long just as often as they are too short. For there is no reason why the one measurement should have an advantage over the other. Hence the synnnetry of the curve is explained. It is also quite obvious that the larger errors (99'6 mm. ' Tlie "error" curve was iir.st estalili.slieil and proved liy Gauss, and the law is therefore often called Gaus.s's Law. Matheiimtical Treatment of Results 19 and 100'4 mm.) apjiear very seldom, and that the fre- quency of the error increases according as the error becomes smaller. And that explains the general shape of the curve. As to the height and breadth of the curve, wc can say nothing a friori. But these two characteristics are not of fundamental ijuportance ; they change merely according to the number of measurements taken, the accuracy of the measuring instrument, &c. Of funda- mental importance is the form of the curve. 2. The Arithmetical Mean Since in our measurements the accidental errors are symmetrically grouped, the arithmetical mean of all the single measurements will give us the real length of tlie measured object. Let n equal the number of measurements carried out (in our case 80), and let a^, a.„ a,^, . . . a,„ stand for the separate measurements — 99"6 (once), 99'7 (three times), &c. ; then to get the arithmetical mean (A), I must take the sum {~^) of all these sej^arate measurements, and divide this by the total number of measurements {n). A= - n Substituting my figures — ■ 8000 A =; ^,, — 100 mm. This would, as a rule, satisfy the interests of ph3^sics, because it is only concerned with getting at the length of the object in c[uestion. 3. The Distribution of the Errors But we might ask a further c[uestion : What was the magnitude of the errors that appeared in our measure- ments ? To answer this, we could take the greatest and 20 Experimental Psychology and Pedagogy the least number (100'4 and 99'6 mm.) or their difference from the arithmetical mean, and say, " The maximal deviation (above and below) is in both cases 0"4 mm." Such a statement, however, would scarcely be precise. For it can easily be imagined that by some unfortunate occurrence we might get a figure that deviates very con- siderably from the mean (say, by 2 mm.). If this one figure were given to describe the accuracy of the measure- ment, it would give quite a false view. Therefore it is always best to calculate the average error (E,„). Let us call the differences of each measure- ment from the arithmetical mean A^, A^, ... A,, (A = difference). A^ and A,„ for example, in our case are each equal to 0'4 mm., and so on. I then take the sum of all these differences (2A), and divide it by the total number of measurements {n). The average error is therefore — ^^ n Working this out, I get — 92 E,„ = 7T7,~' 1'15 tenths of a millimetre. This value is a measure for the accuracy of our obser- vations or for the distributioir of the errors. An objection to this calculation might l^e raised. A large error (say, of 2 mm.) that happens by chance would have too great an influence on our calculation. The above 92 108 example would then be changed from -^n. to -^^ = 1-35, a considerably larger value. Tlierefore in very exact in- vestigations the probable error (PE), and not the average error, is taken as a measure for the accuracy. All the errors (in our case 80) are arranged in order of magnitude, and the middle one is chosen (in our case No. 40), counting Mathematical Treatment of Results 2 1 either from the smallest or the largest.^ There are, then, counting from this middle error, as many large errors as smaller ones. Therefore this error gets the name of the probable error. Then the probability of errors of smaller or of greater magnitude occurring is equally great. The probable error will always be a little smaller than the average error, because according to this method of cal- culation the extreme errors of great or small magnitude do not play so important a part. It has been established that the proltable error always amounts to \ (or, more precisely, "8453) of the average error. PE= iSiO;!- In our example, therefore — PE= ±'S4:-"')''> >: 1-15 = ±-972 tenth of a millimetre. If I always carry out my measurements with the same instrument and with the same conscientiousness, the measure of accuracy (the average or the probable error) will remain the same, whether I measure a length of 10, 20, or 100 cm. If, therefore, the accuracy of an instnmient is once and for all settled, the calculation of the measure of accuracy for a special object is of little importance for the physical scientist. Each observer has, however, his own measure of accuracy, according to the acuteness of his senses, his con- scientiousness, etc. In Fig. 13 we sec the error curve for two observers. The observer B. (the dotted line) is the better of the two. The maximum and minimum of deviation are smaller, and the errors group themselves more 1 If the smallest error is 1 and the largest 20, the middle need not lie at 10. For if there are a number (jf errors of the value 1 and also df the value 2 or 3, the i)roliable error may be 3. By tins method of reckoning, one large error of tlie value 20 or 100 makes no dilt'erence. Of fundamental import- ance is the space where most of the errors crowd together. 22 Experiinental Psychology and Pedagogy about the middle. The value 100 mm. appears thirty-three 30 25 20 15 10 99,6 99,7 99,8 99,9 100 100,1 100,2 100,3 100,4- Fig. 13. — Error curves of two observers. times, instead of only twenty-two times as with observer A. B.'s average and probable error are smaller than A.'s. y' -r / 1 / '.^— \ > 7^^-^ ^>. J ^. • \ N s ^^ ^^ -^ - 4. The Probable Error of the Arithmetical Mean We have up till now taken for granted that the real size of the measured object is the same as the arithmetical mean of the separate measurements. This is true only in the case of a very great number of measurements, or, strictly speaking, of an infinite number. Only in this case does the arithmetical mean really correspond to the real size of the object, and we can then say that the error is equal to nothing. If, on the other hand, I had only taken one measurement, I would not be able to make any statement about the j^robable error. Therefore the more measurements I take, the greater will be their accuracy, and the smaller will be the probable error of the arith- metical mean. One might then imagine that by a hundred times as many measurements the error would be a hundred times as small, but it has been proved that it is onlv ten times as small (^100). If I therefore divide the pre- viously calculated probable errors of all the separate observations by ^100 (or generally, the root of the Mathematical Treatment of Results 23 number of observations, Jn), I shall obtain the probal^le error of my arithmetical mean. In our exam])le — ■972 PE,„= --^ =-10i) tenth of a, millinietre. ,/SU This value is of great importance to the natural scientist, because it is really a measure of the accuracy of the state- ment that the measured object is 100 mm. long. The formulfc that we have so far given are accurate enough, when more than ten measurements are taken. Other fornuihie have been found to be better for a smaller number of measurements. We give below the simple formula:^, for reckoning the error that we have just dis- cussed, and ap])end thereto the more accurate fornudse (Nos. la, 2«, ;3«). The shiejile Foniiitlw. 1. A^'erage error — "-'A ',)2 2. Probable error of the se])arate observations — PE - + -84.53 •-- PE = i -St.'iS :< M 5 - ± -SI? n 3. Prol)a]dc error of the arithmetical mean — o PE == - PE,„= . ='ll More accurate Forniukc. la. Average error — ^^ F /^-P53 24 Experimental Psychology and Pedagogy 2a. Probable error of tlie separate observations — PE= ±-0745a/ — T PE= ±-0745 x l'ri3= 1-03 Q ■Ml. Proljaljle error of the aritlimetical mean — PE 1-03 PE„, = — _z PE , = -^ = -12 Formula lo. requires all errors (A) to be raised to the square root, to be added together (2A-), and to be divided by n — 1. The square root of the value thus obtained must then be taken. Formulse 2a and 3« follow quite obviously. Note the fact that q\q\\ with eighty measurements the accuracy of the simple formulae is very great. II. MEASUREMENTS IN BIOLOGY 1. The Subjection of Biological Quantities to Natural Laws If a great storm occurs during the herring-season, and if the shoals happen to be near land, then hundreds aiid thousands of herring are cast upon the shore and die. The anger of the waves seems to make no distinction among the hundreds of thousands of these creatures that it condemns to death. No one who has witnessed such an occurrence can help feeling pity for the destruction of so many living beings, whether he has a direct interest in it as fisherman or tradesman or whether he be merely a lover of nature. And only very few know that such natural phenomena arc subject to definite natural laws, that even here so-called chance acts according to a special law. A mathematician took the trouble to obtain measure- ments of many thousands of the dead aninrals — of the length of the whole body, of the head, of the fins, &c. He found a great number of very small and very large animals. The medium size of fish was scarcely found at all ; and the MathcJiiatical Trcaf/iic//f of Rcsn/fs 25 more the animals approached this medium size, the fewer there were. The same was found to l)e the case A\-ith regai'd to the length of the various parts of the bodv. Animals Avitli very long or very short tail-fins were more common than the medium kind. When the same observer took the same measurements of the fish caught at sea, it appeared that the great majority of these belonged to the niedimn size, and that, with the same regularity as before, the munl)ers of small and large animals this time de- creased in proportion to their size. This r(^gularity was found to correspond exactly to what we have seen in measurements in physics — in short, to Gauss's Law of Error. We can. then, in this case also, in accordance with our law, fix a mean value and an average or probalde error of deviation from the mean. The mean value represents the common type of fish, and the mean deviation ^weH us the extent of variation from this type. We see to some extent the tendency of nature always to reproduce these animals according to a definite size and to definite pro- portions. And we assume with certainty, that the working of the law of heredity, by which the oft'spring in general is exactly similar to the parents, is a sufficient reason for the appearance of this mean value. The extent of varia- tion represents the extent of the general conditions under which the animal lives. Further, the average strength of the storms settles the limits within which the average animal can exist ; the animal that is too large or too small perishes. Within these limits fixed by heredity and by average conditions of existence, exists another division according to the accidental conditions of existence of each separate animal (the dift'erent kinds of food it may obtain, &c.) ; and in consecpience of these accidental con- ditions, the form of the curve must be the same as that of Gauss's error curve, which also arises from purely acci- dental factors in measuring any physical phenomenon. 26 Experiincutal Psychology and Pedagogy This average error, which lias little importance in physics, is in biology of the greatest importance. It gives, as the so-called extent of variation, the limits within which those individuals grouped around the mean value may be called normal. These laws are valid for all organisms, and also for human beings. The measurements here are, of course, more difficult than those in physics. Thoma,^ for example, found that very large errors appear in simple measure- ments of the height of a human being. In taking several successive measurements of a child, errors of 4 or 5 mm. appeared in spite of the greatest caution. However, if we measure a very great number of individuals, the law of error is clearly seen. Fig. 14 shows two curves : the continuous line shows the Gauss curve calculated according to the mathematical laws ; the dotted line the distribution according to height of 2192 American school-children (eight-year-old girls). For example, 2 girls were 137 cm. high, 8 were 135 cm., &c. ; 300 girls attained the average height of 120 cm., 3 were 103 cm. high, and only 2 were 100 cm. Note how accurately the dotted line corresponds to the ideal curve. The scientists who first of all recognised the fact that life and death and all conditions of human existence are subject to definite laws that can be represented in figures, were filled with awe at the regular working of nature. Sussmilch., the author of Die gottlicJie Ordnung in den V erdnderungen des menschliclien GeschlecMs , writes ; " The all-wise creator and ruler of this world brings forth out of nothing the numberless army of human beings. The infinite being allows us for a certain time to bow down in reverence before His presence, until, our time completed, we must leave the scene of this world. Our entrance, ' R. Thoma, Uiilirsinjiiuiijeii iiber die Grossc und das Uewkht dcs mensch- llrheii KiirjiKis, Leipzig, 1882. Matlicinatical Trcatnienf of Results 27 our passing before the eyes of the Lord in this world, and our exit, all take place according to a most wonderful law." 300 2.S0 2IJ0 240 220 200 ISO \m liO 120 100 so 60 10 20 ■■■ ■■■ JIB! ■!■■ ■■■■ ■P mm nnnnBi ■■■■■■m 11 ■■■■■I ■■■■■I IB .JH ■■■11 ■■■1 ■■■■■I ■■■■■I ■■■■■I ■■■■■I ■■■■■■ ■■■■■I !■■■■ !■■■■ KBHi ■■■■■ Ifll ■■■■■■■n^^i ■■HHiliU ini ■■■■■■■■■■n^i ■■■■■■■■■■■ni ■■■■■■■■■■■■■■■■■■Kr ^ammwmmmwmmmwmmwm I I 102 lOU .110 114 lis 122 126 130 134 13S Fig. 14. — Di.stribut-ion curve of the height of 21!)2 eiglit-yc nr-okl girls of St. Louis according to Gauss's law (continuous line) antl according to the real observations (dotted line). (From Townsend Porter, Zeitselirift fiir Ethnoloyir, 1S93. Asher fiir AnihropoJoijic, Bd. oO. Vieweg & Sohn.) Mafhciiiatical Trcafmciif of Results 29 On closer investigation, we find tliat sucli an asym- metrical cnrve always arises when we try to fix a ]iormal type in biology. [Gf. Fig. 17.) E^'cn tlie curve in Fig. M sliows on closer examination a slight asymmetry. 40 85 30 25 20 15 10 5 — - """ -40 \ -- \ 1 \ \ ^ y 1 \ \ 2 e Kl 3) / \ " ^ / 3|1 \ r^ 9 10 11 12 13 U 15 10 17 18 19 20 21 22 23 24 25 2B 27 28 29 30 31 32 33 El 35 3(; 37 38 Flti. 17.— Niuulji-r ot railiatc lluwris oE 17,0110 heads of tlic Chrijxn ull.i uunn UuriiutJu nut Hi acu'.'rding to LuiIwjl^ (From Raiikr and Ui-oiiior, .1 rrA/r /«,■ . I /i/A /•.,/»,/.»/;, , 1;,1. TiO. ^'iewol,' &. Solin.) This can be seen from the hiet that the greater ])art of the left half of the dotted line lies within the contiiuious line, whereas the right half of the dotted line falls outside of it. The reason for the asymmetry of liiological distriljution is difficult to prove theoretically, but it can be easily made clear by practical examples. AVe can easily imagine that a very slight shortening of the normal length of the tail-fins of the herrings might make these animals in- capable of fighting successfully against the elements ; 30 Experimental Psychology and Pedagogy whereas a much greater extent of variation in the breadth of the fin may be possible before this would do harm to the animals owing to its unmanageableness — i.e. the muscles of the fish might not be able to manage these large fins in a heavy sea.^ We must accept, then, this asymmetrical distribution for biological cjuantities, and it follows that our calculation will be different from that for physical measurements. As regards the chief value, we see at once that the arith- metical mean will not represent the normal tyjae, but rather that value which appears most often, the average density, so-called because at this special point the number of observations are most dense. In Fig. 15, for example, the average density is 39 cm., because this value appears with the greatest frec[uency, namely, thirty-two times. The arithmetical mean, on the other hand, is only 36 cm., a value that obviously does not represent the normal type- The so-called " central value," or median, is also often used as a measure in asymmetrical curves. To find this, all the values are arranged in order according to size, and the middle or central one is chosen. In the above example, the lengths of all the 217 stalks would be arranged in order, and the 109th one would be chosen. This one has a length of 37 cm. The central value always lies between the arithmetical mean and the average density, and ' KiDin a purely iiia.tlieiiiatical sljiiidpoint, the asyiniiietry of biolutjieal quaiitilics i/an lie thus explained ; — A syiiiiiietiacal arraiiL;eiiient can only lie expected "vvhere there is an infinite ninuher of independent causes cif errors. In such a case, it is taken for granted that the positive errors will balance tlie negative ones, because there is no reason for an opjiosite view. Where however, special laws are at work, as in the biological la\\s of growth an asymmetrical curve must always be expected. The height of a tree cannot increase indefinitely beyond the normal height, because it wuuld not then be able to withstand the storms ; and similarly the other half of the curve wdll siidv slowly down to the height of the smallest dwarf-trees. MafJicniatical Tjxatiucnt of Results 31 therefore approaches the hitter more nearly than the arithmetical mean does. We give these three values in our example : — Tlie ai'itliiiictical mean — ?>fi cm. The centMil \aliie or median = 37 „ The ave)"i^c density or mode -= W.) „ If only a few measurements have 1)een taken, the central vahie is sufficient, and it is easier to arrive at than the arithmetical mean. With a great mmifjcr of measure- ments, the density value appears in the curve without any further calculation, because here the greatest number of observations crowd together. This value should, then, have the preference. After fixing the density value, we must reckon out the average error for each side of the curve, and out of this we get the extent of variation, which is, of course, asymmetrical, reckoning from the average density. In our example the density value is 39. We then take the sum of all the deviations that lie above (^A"), and divide it by the total number of deviations above the density value («"). We therefore get for our average error of this upper half of the curve (El), according to the simple formula on page 23 — „ -A" K, , = „ ■ We now carry out the same calculation with the lower half of the curve, dividing the sum of the deviations below (2 A'') by the total number of these deviations (ii''). This gives as a mean deviation below the density value — Instead of saying " average error," it would in this case be better to say " mean variation " (,„V). Accordingly 32 Expcriinciital Psychology and Pedagogy the following values would characterise a biological dis- tribution : — 1. The density value, representing the normal type of the object investigated. 2. The mean variation, above and below this normal.^ Lastly, the probable error of the density value can also be calculated, either for the upper or for the lower half of the curve. According to the formula on page 23, and calling the probable error of the density value PEd, we get two formulas : '" — ,,V" or The importance of these formulae in the measurements of school-children will be shown later. 3. Measurements oe a Collective Object In order to arrive at exact measurements in physics, we carry out a number of measurements of the same object, and then take the arithmetical mean. In ])iolooy^ where we deal with a whole class or species of objects, of ' It is nljvii.us lliat on tlie siilu oT (lie c'm-\e, ^\■llelv tlie ,„A' is fjrrater Llie greater miiiilier of oliser\-a(.ioiis must lie, and also lliat llie iiuiiiliei' of oliservatioiLs nuist lie jjiopoilional to the size of the mean variation— ,„V'' ii" : V" = Mi.= :iA" ' The in'ohalile enxii- ■\\ill not he the sann' in hoth eases. This aiiscs fi'om the fact that a greater luimher of olisei'\-a(ioiis oecnr in the one case than in the other. Mathciuatical Treatment of Results 33 which the separate representatives are measured, we are trying to arrive at the normal type. This whole class or species Fechner calls a collective object, because we are here concerned with a collection of objects. To arrive at real values, it is desiral)le to measure all tlie indi^'idua]s of the class in question — e.q. all human beings. If this is impossible, I must limit my class. I can, for exam]:)le, measure all adult Germans (c/'. the measurement of re- cruits) or all twelve-year-old girls in Berlin, &c. Boys and girls shonld naturally not be measured together, Ijecause I know in advance that the bodily pro]Dortions of the two are different. If in any district there are very marked racial differences, I must then measure each race separately. Very special attention must be gi^'en to the cases where, instead of one type, two types appear, because of the great change wrought on natural conditions by artificial influences. It has been established that children of well-to-do parents differ greatly in their measurements from those of poor parents, and also that the rapidity of bodily development is c_[uite different. These differences appear to be \s, &c. See VVj^. IS.) 36 Experimental Psychology and Pedagogy first half under normal, the second half mider abnormal, conditions. We have here, so to speak, two different types or races. The calcidation of a middle value in all cases of curves of two or more maxima is useless. We must here investigate whether there are not two races present, or whether the one set of the individuals investigated has not grown up under conditions different from those of the other set. III. MEASUREMENTS IN PSYCHOLOGY Whoever undertakes to describe psychological processes in accurate figures, will soon discover what great diffi- culties present themselves. Suppose we are successful in measuring some way or other a sensation of brightness,^ vet it is absolutely impossible for us to measure the same sensation again and again. For every sensation, in fact every process of consciousness, is, when it has once j^assed, irretrievably lost, and will never return in exactly the same form. Since for exact measurements I rec[uire a, great number of individual measurements, I can only get over the difficulty by measuring a number of similar sensations : for example, say a number of sensations which arise in a normal state of consciousness due to the stimulus of a definite degree of brightness. We measure, so to say, a number of separate individuals — i.e. the separate sensa- tions — and then we determine the normal type of this whole class or species of sensations. It- is obvious, there- fore, that psychological quantities can only be treated mathematically as collective objects. The methods we employ here are therefore similar to those of biology. • The possibility of such lueasureniunts will be shown in llie following cha|iters. Mathciiiaticiil 'rrcafinciit of Results 37 We iiiiLst expect also in psychology asymnietrical dis- tribution, and we must calculate the density value and the extent of variation above and below. Let us take for granted that a certain individual can be ])ut into certain states of emotion in most cases by the application of even a very small stimulus (say by sensa- tions of temperature). If we took a certain number of measurements, the density value of our figures would mean a great sensitiveness of the subject in question. Again, if the mean variation of one such subject was large, and of another small, this would show an unreliable and reliable nature respectively (reliable in the sense that we can rely upon a person to react in a definite manner to a definite stimulus, or act in a definite manner ac- cording to given motives). Fig. 20 shows the pro- ductive power of an adult in copying out digits for a period of several hours. The time taken to copy out each set of 25 digits was registered, with the following results : — 4 times lie took 12 seconds. 'JO 10 / ^ \ / \ / \ / \ ^ ' \ 12 14 ii; ilS sees. Fig. 20. — Dislribution of the ability to copy out single digits. 7 13 .32 14 47 !•"' 44 If! 18 „ n 8 1*^ 1 time 19 Note the similarity of this curve to Gauss's error cur^'e ; note also the slight tendency to asymmetry. If psychological observations on one and the same individual are to be treated mathematically as a collective 250 201) luo 50 1 /\ /_ \ I \^~- / \ / v^ 120 125 145 cm. Fiu. 21. — Distribution of 1794 pictures (laudscjipes) accordiiiij; to the height of the picture. (Froui PechnerV KoUcldnnri^^^/rhir.) 38 Experimental Psychology and Pedagogy object, mucli more so must observations on groups of individuals be similarly treated. In the products of human culture, in the language, religion, and art of a people, certain psychological processes are to be found, as it were, fossilised. These, as we have mentioned above, form the subject- matter of racial psychology, and they can also be subjected to measurement. Fechner had the measurements of the height and breadth of thousands of pictures taken. Fig. 21 shows the distribution of landscapes according to their height. The height of greatest frequency was 35 cm. (more than 250 times). The distribution is extra- ordinarily regular and clearly asymmetrical. The same applies to the genre pictures (Fig. 22). This example shows that the most complicated psy- chological pro- cesses take place in accordance to certain laws. Here we are dealing with very involved sesthetical values. In Fig. 23 we have obviously a curve of two maxima, similar to that in Fig. 19. Without knowing any thin i4ilirc.) Matliciiiatical Trcafiiicut of Results 39 simple collective object, but that there must be two classes or species present. And this is really the case. I tested my capacity for addition during an uninterrupted period of six hours. The real purpose of the experiment was to measure the progress of fatigue in mental work. The distribution curve would ])resumably lia>'e been a regular asymmetrical one witli a maximum at about 160, except for the fact that certain irregularities hindered the progress of fatigue. After the fourth hour, the yA\\\ in the hand that was cojjyijig down the figures became so acute, that I instinctively let my hand slip down and struck it against JO / \ / \ / i / \ 10 / -^ ^ \ \ / \ ^ / \ .,^ ^^ 120 130 rHJ 150 li;i) 170 180 seconds. Fig. 2M. — Distributiun curve witli two .summits for addition i'oi' six lioLirs. (250 additions in 115, 120 ... 190 seconds.) my knee. Owing to this massage, some of the fatigue- substance in the hand was no doubt got rid of, and my production rose to 250 additions in 150 seconds, whereas before 160-170 seconds were needed for the same number of additions. Owing to this reason, 150 was the point of greatest frecjuency ; it appeared twenty-five times, and thus formed a second maximum. This example proves further that it is not allowable to group together work done in different stages of fatigue. Only the figures obtained during the same stage of fatigue (or of practice, &c.) can be classed together to give a middle value. Therefore a psychological investigation must as far as possible always be conducted at the same hour of the day ; all irregularities (such as indulgence in 40 Experimental PsycJwlogy and Pedagogy alcohol or tobacco, &c.) must be as far as possible ex- cluded ; in short, everything possible must be done to obtain processes of consciousness of the same quality or class. It is, of course, also obvious here, that only a great number of experiments will lead to sound conclusions. IV. MEASUREMENTS IN CHILD PSYCHOLOGY AND PEDAGOGY Child psychology and pedagogy, as well as the anthro- pology of the child, deal with the development of the human being. The anthropology of the child and child psychology investigate the natural development of the human body and mind, whereas pedagogy takes into account the changes that occur owing to a systematic influencing of the natural development of the whole being. From what has been said above, it is obvious that the objects of investigation of the three sciences mentioned must be dealt with mathematically as collective objects. One further point requires special emphasis. To exj^lain this point, let us take an example out of anthropology. 1. The Change in Asymmetry through Natural Growth In Fig. 24 in the measurements of the height of the 12- to 13-year-old Freiburg school-girls, we have an extraordinarily even and almost absolutely sym- metrical distribution. The asymmetry of biological dis- tribution is sometimes very small, and there are exceptional cases of absolute symmetiy. In the thirteenth year a period of increased growth sets in, as is normally the case. In the second curve Mathematical Treatment of Results 41 (age 13-14) the asynimetiy appears, but is very indefiuite ; in the third (■ur^'e (age 14-15) it is very appreciable. This can be easily explained Only if all the children increased by exactly the same amount each year, would the sym- metrical cur\'e of the 12- to 13-year cliildren move towards the right along the abscissa without changing its form. But, of course, certain groups of cliildren — e.fy. those who up till now were backward in development — grow at a n hf y ■c 7qe \ - n f c 1 R L S. / /9\- ■'/s. \ c hfl o'/r / 1 \ . r-f / \ 'n ) ^ / / o\ ?e \ \ L / / •/i\ 11 \ Y> n ) v . \ \ ^ A -J^e \\ V '■ rr. — /po J /z 7 /J. \ io / \ \ .-> \ ^ _ , 10 1. r f« /*"/ /. VI / ?(i / s .1 - — a_ He^g/ t f. n (^ C/J t,r rie 're s . Fig. 24. — Distiiliution curve of the heitrht of Freibcii; elementary sohool-chiklieir. (From Geissler and Uhlilzscli, ZdU':hi-\ft rl,s Varhs. ,'nt 1. The Possibility of exact Measurement in Child Psychology and Pedagogy Each teacher gives marks — very likely for the attentive- ness or industry of the pujDil. He thereby arranges the value of a j^sychical accomplishment according to a certain measure, and in doing this he serves as a proof for the naive notion that the possibility of measuring ^'sycliical processes is a thing to be taken for granted. Such measurements have two great drawbacks : Firstly, they are extremely inaccurate ; and secondly, the measurements or nrarks assigned by different teachers have no common basis for comparison ; — they are purely individual, not general. For this reason, the tremendous amount of work in assigning marks which many thousands of teachers are doomed to cany out at the end of e^'ery school year — all this work will never advance scientific pedagogy l)y one step. If, however, we could get figures which were accurate, and wliich we could compare with each other, pedagoo-y could then begin to solve problems, which at present cannot be attempted. If, for example, we could find a method by which each teacher in the same manner could determine accurately in figures the memory-power- of his pupils, we could measure and compare the memory- ]30wcr at all ages of growtli. If the same measurements The Measurement of Seiisntiou 49 were carried out every 50 or 100 years, we could tell whether in following generations the memory was better or worse, just as we can at the present day determine whether our athletes are stronger or speedier tlian the champions of the old Greek athletic contests. Yes, we could then prove the value of any educational reform, say the introduction of special classes, in regard to the de- velopment of memory, if before and after the introduction a sufficient number of measurements were taken. We must therefore discuss this extremely important question of the possibility of exact measurement of psychological values.'' In physics we make a distinction between direct and indirect measurement. It is, for example, a direct measure- ment, when we determine the length of a rod by laying a yard-stick by the side of it and reading ofl: the length in question. We are here measuring lengths with lengths. It is otherwise when we determine the temperature. We possess no direct heat " yard." Here we must measure the heat by the expansion which takes j^lace in a mercury column — i.e. quite a different quantity. For this reason, we cannot say that to-day it is " once as warm," or " half as warm " as yesterday ; we must say " it is warmer or colder." Now, it is further quite obvious that we cannot directly measure psychological values — for exanqtle, sensa- tions. That would presujjpose that we possessed a graduated sensation scale, or sensation " yard," which we could keep by us and at any time measure any sensation we chose. An indirect measurement is, however, possible. I can measure the sensations by means of the stimuli which produce them. I can say, for example, about a ' Cf. Ci. F. Lippp, i'/e jK-iycliui'lieii JSlKsamelliodai ; G. E. Miiller, Die Gesicldspunkte mid die Tatsachen drr jixiirhidn/iisflu-ii Method ik ; (). Ivulpe, Grimdriss der Piychologiey Wundt, Untiid::iige dei- pliydoIuyiaeliLn d'niirlinloi/ie, Band I. D 50 Rxperiinciital Psychology and Pedagogy sensation of light, " tliis is a sensation wliicli arises when I see one candle burning," and " this is a sensation wliicli arises in me wlien I see 1000 candles burning." But I have really gained nothing by this. If I compare the two sensations, the most that I can say is, that the one is greater than the other. I do not know whether it is 1000 times or only twice or perhaps 10,000 times as great as the other. We cannot speak here of a real measurement. Again, a comparison between different individuals is absolutely impossible by such a measurement, and in pedagogy such a comparison is exactly what we want to achieve. We only know that in observer A. as well as in observer B. the sensation caused by 1000 candles is greater than that caused by one candle ; but we can say nothing definite as to whether the sensation caused by one candle in A. is the same as or greater than that in B. We shall soon see, however, that there are cases where a com- parison between individuals is possible. 2. The Range of Application of the Methods OF Psychical Measurement (a) The Determination of Stimuli Thresholds or Limens. In special cases it is possible to overcome the difficulties described in the previous paragraph. I might ask. What is the smallest stimulus — e.g. the smallest amount of light — that is just sufficient to cause a sensation in observer A. or B. ? This stimulus is called the stimulus threshold. I can also determine the terminal stimulus — i.e. I can ask. With how many candles have A. or B. the strongest sensation of light, so that the addition of more burning candles would cause no change or intensification in their sensa- tion ? The investigation of terminal stimuli is for practical reasons (especially with children) not to be reconnnended, llic Mcasitrciucuf of Sciisaiioii 5 1 because, owing to the effects of very strong light or very high tones, damage may he done to the sense organs. The stinnikis threshold can, however, be determined for adults and childi'cn, and it leads to figures for comparison. If one child has a light sensation from a \-ery small degree of brightness, I can say that it possesses a greater sensi- ti\'ity for light than another child who requires a greater degree of brightness before a light sensation arises. The colour-mixer (Figs. 29 and 30) ^ is used to deter- KlG. 2i"l. — Cnliair mixrr. Fl(.'. an. — Colour luixcv on .stand. mine the stimulus thresholds for the intensity (brightness) or for the quality (colour-tone or saturation) of a light sensation. The apparatus, the colour-mixer, can be set in motion by a small electric-motor, driven by an ac- cumulator, or by three or four dry batteries. For individual experiments, it stands securely on the table (Fig. 30). For demonstration purposes, the other apparatus (Fig. 29) is preferable, as it may be carried about in the hand during rotation. Fig. 31 shows a child sitting before the ap- paratus during an experiment on the stinriUus threshold. ' Pirlnre.s taken from tlie catalogue of Zinimennaii, Leijizig. 52 Experiincutal Psychology and Pedagogy On the colour-mixer there is a large grey disc, the front view of which is given in Fig. 32. It has a radial sht that extends to the centre of the grey disc, so that a smaller coloured disc (say red) can be inserted. If this '•■/ 'rT \.L| Ig 1 1 / ■ '^" ""'' , -^-'WT- ;^ KiL -" ^«fl ^^■—-' [A if •« ■''■ X W' •' * 1 ym V '"^ '^-^Mi^',4: ■ ^ Fig. 31. — Iiivi'sli.natiou of llir .sUiii\ilii,s (bru.^liold Utv colours. smaller disc only projects by a few degrees, the whole disc, even the inner circle, a.p])ears during rotation quite grey. Now, when I have determined how many degrees of red must be added in order that a trace of red may be just noticed in the inner circle, I have fixed the stimulus The Measurement of Sensation 53 tliresliold for the sensation of red in question. (Fig. 31 RED FK-J. :!2. — Colciur disc fm- investigating tlii' stimulus liiiTslioM. sliows also how the degrees are measured off on the back of the disc of the colour-mixer.) (Ij) Tlie DeterminafAon of Difference Thresholds. Fig. 33 shows two colour-discs. Tlie first is divided into 180° Idue and 180° grey. During rotation a certain Fig. 33. — Colour discs for investigating tlie difference sensitivity. 54 Expcriiucutal Psychology and Pedagogy blue will result. The second disc shows the addition of more blue. By these means I can investigate hoAV many more degrees of blue must be added in order to give a second blue colour that is just noticcttbly stronger than the first. This value is a measure for the degree of sensi- tivity, or the difference sensitivity. The determination of the difference sensitivity is much more important than that of the stimulus threshold. If I wished to investigate whether a child is musical or not. I could determine the stimulus threshold by finding out how many vibrations (sixteen or more) per second a tone must make, before it is recognised as a tone. But it would Ije of much greater importance to determine how many viljra- tions must be added to a tone about the middle of the scale, in order that a second tone be distinguished from the first. Musical talent will in most cases show itself clearly in the difi'erence sensitivity. (c) The Determination of Stimuli that appear Equivalent. In the methods of investigation just described, we have been changing the stimiilus, until a definite change in the sensation took place. In the first case (stimulus thres- hold) until a sensation appeared ; in the second (difference threshold) until we arrived at a sensation that was felt to be different from the original sensation. I could now ask the opposite question : Under what conditions do two dift'erent stimuli give rise to two sensa- tions, which in one particular at least a2)pear the same ? They can never appear absolutely the same. For even if we are dealing with two lines, which, according to my judgment, are equally long, yet the two sensations are in so far different as the one line appears abo\'e or below right or left of the other, &c. Therefore, I can only say 7 he Mcasnrcuieiit of Sensation 55 that two sensations are the same in o?ic particular. In otlier words : two different stimuli can call forth sensa- tions, which in one particular seem equal or ecpiivalent. For example (Fig. 34), I can compare a, green disc («) with a grey one {h) The grey is obtained "by a certain mixture of w^hite and black. I now change the green by increasing the degrees of white, until \\w. green and grey discs appear equally bright. I thus obtain in „ GREEN ^^^^^^^ a 6 Fig. 34.— Coliiur discs for investigating stimuli tlint appear equivalent. (Colour and brightness.) regard to one particular, viz. brightness, equivalence or sameness. (d) The Determination of Differenees that appear Equivalent- A similar case is seen in Fig. 35. By rotation of the whole disc, we get in the middle a black circle, outside a white strip, and between the two a grey one. I can now set myself the problem to change the grey strip in the middle untiJ, according to my judgment, it represents a 56 Experimental Psychology and Pedagogy brightness of grey that lies exactly in the middle between the white and black. According to my sensa- tion, then, the difference between the grey, I have 1'"IG. 35. — Colour disc for investigating diffurcnces tiat appear eqviivale)it. chosen, and the black is exactly equivalent to the difference between the same grey and the white. This is an example of determining differences that appear equivalent. In child psychology and pedagogy we would do best to limit our investigations for the sake of simplicity to the determination of stimulus and difference thres- holds. 3. The Thkee Methods op PsyoHiOAL Measurement (a) Tlie Method of Continuous Variation. To determine more accurately the stimulus threshold for colour-tones, we must make use of a colour-mixer, on which the red slit can be increased or decreased by TJie Mcasjircniciif of Sensation 57 means of a screw during the rotation of the wliole disc.i The first question to answer is : How sliall tliis increase or decrease of the red slit tal^c place ? Tlie sini2)le,st nietliod is to leaA'e this to the observer. He is permitted to turn tlie screw tliis way and tluit, until he arri^■es at the place where he can say that the middle circle is just noticeably red. This is called tli(> method of continuous ^'aria.tion, because the stimidus is varied continuously this way or that. The disadvantages of this method are, of course, obvious. We shall obtain quite different results according as the oljser^'i'r changes cjuickly this way or that, or according as he makes the changes slowly and cahnly. And worst of all we cannot tell, out of the figures obtained, how the obser^s'er has set about his task. One nray always start \\A\\\ a great deal of red, another with no red, and so on. The figures can tell us nothing of this. So this niethod of continuous variation is the most unmethodical of the psychical methods of measurement. It cannot be used in investigations in child psychology. (b) The Metlwd of Liinilsr [-ii.neiithjiitiuii of the i^tiiniiliis Tlina^liohl accordimj fu lite ]\fetlile. Now we ascend until both stimuli appear the same (just un noticeable difference of DL,), and then still further until the second appears greater (just )ioticeable difference of DL"). After several repetitions of the whole process, I reckon oirt first the upper difference threshold by taking the arithmetical mean, the mean variation, and the probable error of all the just noticeable and just unuoticeable differences of the upper threshold (exactly as in fixino; the stimulus threshold). I then repeat the same calculation for tlie just noticeable and just unuoticeable differences of the lower threshold. Of these two threshold values (DL" and DL,), I take the mean and so obtain the mean difference threshold (DL,„), or for shortness the difference threshold (DL). (c) The- Method of Constani Chani/es. This method gives the most accurate results. Here, just as in the method of limits, we work with a number of fixed gi'aduated changes. Only here these gradations are not ])resented in their natural seriueucc, but are first thoroughly mixed u]i togx'ther. Tlie observer is called on to give one of the three judgments — greater, smaller, or ef(ual. The way in which we should deal witli the results ol)tained by tliis method, has led to long and tedious discussions. Great diff'erence of opinion has arisen as to how we should classify the " ccpurl "' judgments. As it is undesirable to repeat all this discussion in a 1;iook on practical child psychology and pedagogy, we shall not recommend the rrse of this method. It is still in an ruisettled condition. Experiments according to this 62 Expa'inicutal Psychology and Pedagogy method also make a great call upon the staying-power of the observer. We therefore recommend the method of limits as the best for experiments in child psychology and pedagogy. By using the three methods of psychical measurement in the four departments of sensation — measurement (see pp. 50-55), we obtain twelve different cases. But since we recommend in child psychology only the method of limits, and this only to be applied to investigate the stimulus and difference thresholds, there remain for us the two following cases : investigation of the stimulus threshold according to the method of limits, and of the difference threshold according to the method of limits. We have given a detailed description of both. 4. The Meaninci of the Figures Obtained In our experiment on the 9-year-old child, we ob- tained the following figures : — Sensitivity for the saturation of a green colour . . 16-75^ Mean variation ........ ■■56° Probable error of the diHeronce tlire.shold . . . '17° (a) Sensitivity. What do these figures mean ? As they stand there alone they mean nothing. The figure 17 (for sensitivity) tells me simply that I must mix 17° of green with 343° of grey in order to get a green which is just noticeable. But this figure increases at once in importance, as soon as I have other figures for comparison, which were obtained under exactly the saiue conditions. I give once again the separate results of the ex]ieri- ment with the child, and for comparison the results of the same experiment with a woman of 60. As the wife of a TJic Mcasiii'cnicut of Sensation 63 drawing-teaclier and tlie mother of an arti8t, she would certainly have had exjjerience in colours. FriJa L. (9 years). Mr.g. N. ( Ascending. 2.'-)° 26° 28" 00 years). Asconding. 17° 1 0° Ui" 18" Descending. Descending. 1st 1 2ii(l 3nl 4tli -xpt'iiiiieiit .... 17" 16° 17° 17° 29" 30" 29° 29° We can see here clearly that the 9-year-old child is superior to the woman of 60, not only in sensitivity (17° versus 28°), but also in regard to the steadiness of her judgments. At first I feared that the child was only guessing, i.e. that she always said " green " after a certain number of single experiments had been made. Therefore I undertook a new set of experiments with the child in which four colours were used. I employed the so-called " method without knowledge." The child did not know which of the colours would appear. The colour-mixer was covered, set in rotation, then uncovered, and only then did the colour gradually appear. Here, of course, the ascending method alone could be used. I experi- mented under exactly the same conditions on the child, on Mrs. N., on Mrs. L., the mother of the child (40 years old), and on Mr. K. (drawing-teacher, about 40 years old). I obtained the following results : — Purple Blue . Ked . (jreeu Prida L. (il). IMrs. L. (4(1) Hfr s. N. («0). l\tr. K. (4(1). 10° 9° 18" .5° 10" 11° 17" 5° / / 1.3° 8° 10° 10° 18" 8° Note first of all, that the stimulus threshold of all 64 Expei'iuicntal PsycJwlogy and Pedagogy observers by this method without previous knowledge is not, as one would expect, larger ; it is smaller. Perhaps this arises from the fact that by the other method the constant appearance of the same colour gradually creates a slight insensitiveness for that colour, and also gives rise to a certain uncertainty in judgment. A comparison of the sensitivity of the child with that of Mrs. N. shows that the relation between the two has remained the same ; before it was 17 : 28 ; now it is 9 : 16. Also in this experiment the child shows an extraordinary evenness in its judgments. (The value 7° in red would seem to arise from some con- stant error, because the sensitivity for red for all observers, except Mr. K., shows very low values. As a matter of fact, after a careful test, it was found that the red did not correspond in brightness to the green employed.) This experiment with the child shows with what good results the exactest methods of psychology can be used. Wundt has repeatedly written against the use of experi- mental methods in child psychology, and many educa- tionalists have misunderstood his words. He obviously means them to ajiply to the psychology of childhood, i.e. to the first few years of the child's life.^ He has clearly stated his belief in the use of experimeutal methods with school-children. Now if we do decide to experiment with school-children, then we ought to take care to use the exactest metliods possible. In investigating the colour-sensitivity of school-children the " naming " method has been used, i.e. the child lurd to ]iame slips of colour of different colour-tone ; or the " covering " method, i.e. the child liad to co^'er coloured slips with slips of the same colour. Little can be achieved with tliese rough-and- ready methods. If one is reminded that the colour- sensitivity of the Japanese mouse has l.>een fixed at i',,, it would almost suggest the idea that Japanese mice are ' Wlimll, Ollllilics nfJ'siirll^ln^JIJ, p. 33(i. Lrijizig, 1907. The Measiivement of Sensation 65 more trustworthy and more interesting observers than our school-children. Fig. 36 shows how in animal psychology such investigations are made. It must be remembered that our experiments deal with the determination of the sensitivity for colour- saturation and not for colour-tones. To determine this latter, one colour must be changed gradually in the direc- tion of the next colour on the spectrum, until a change has lieen observed. By this means we arrive at the difference sensitivity for the colour in question. In our experiment the great similarity between the figures for the mother and child is very striking. Again our experiment raises the c[uestion as to whether the colour-sensitivity of children, as is generally supposed, is really very much less than that of adults. Of course we dare not generalise from our particular case. If it were not so, that the colour-sensitivity of children is less, then we would have to suppose that it is in the naming of colours that children are weaker. Pedagogy would then have to reckon with this fact, and woidd have to cultivate the colour-sense of children more than it has up till now, e^Tn with the youngest children. Comparisons l^etween adults and children have ])re- sumably very often led to false conclusions, l)ecause the figures obtained from school-children and those obtained in laboratories from students and professors have been without any allowance compared. Now the school- children generally belong to the middle or lower classes, and the professors and students to the upper class. To obtain valid comparisons we ought to take individuals out of the same class. It would be very useful to experi- ment also with the parents of the children in cpiestion. This would be of great interest in regard to children of exceptional talent, since it might lead to important results for the theory of heredity. Many a country teacher, 66 Experimental Psychology and Pedagogy b'lO. liG. — Yurkcs' arniiij,'i'iiieiit. fur rk'tuniiiniiig the power of diseriiiiination of colours in mice. One coiu|i£ir(.iuent in illuminated red, the other green. (From Claparede, Vvisvhau, 1908, Nr. 2«.) The Measurement of Sensation 67 undev whowe hands two or three generations of the same family pass, migiit obtain very important scientific results, if he conducted year after year such simple experiments on the children and parents of the same families. To arrive at striking results rpiickly is not so easy for ex- perimental psychology. More important is conscientious work done year after year, that slowly but surely helps to settle the problems we are faced witli. If investigations on the colour-sensitivity of cliildr(m had been made,^ we might, for example, be a1jle to say that the colour-sensitivity among boys in the sixth class was ec|ual to that of girls in the eighth class. Further, if the colour and form sense had tlius been investigated, we would obtain great helj) in reforming our methods of teaching drawing. Many observers maintain that there is such a difference between boys and girls, and yet, in most schools, boys and girls are taught in exactly the same way. Many will say that it is absurd laljour to carry out 100 or more experiments merely to find out one single child's sensitivity for one single colour. But we must rememlier that the methods of exact science are always slow and tedious. Think of the scientists, who spent years of worlc investigating the qualities of the cathode rays, without at first arri\'ing at results worthy of note. Then all of a sudden came the wonderful discovery of Rontgen rays and of radium with all its marvellous cpialities. If it is considei'ed worth while coimting the number of filaments in 17,000 dandelion flowers ; if a physiologist " is not frightened of carrying out 60,000 experiments on ' LI. W. Jones, " Untersiicliuiigeu iibcr die E.eizscli^\elle I'lir Farlieiisatti- guiig liei KiU'lerii," V'criifridlirliiiiiij, n i.lrs Instituts fiir cyp. J'ailuyyil^ imd Psychologic, ]M. H- Leipzig-, UUl. These experiineuls ajipeared after the Gennau eilition was jmljlished. — Tcaiixlaloi'. - C'anierei', Zeilxchiift fiic /tlnlmjic, 1881, XVIl., " Versuche uhei' den Uauiiisiiiii del- Haut liei Kiiideiii." 68 Experimental Psychology and Pedagogy his two cliildren in order to fix the fineness of the space tlireshold of the sense of toucli, then we ouglit not in experimental pedagogy to shrink from experiments that demand a great amount of time. (b) The Mean Variation. Tlie mean variation or the distribution of the figures is clearly a measure for the certainty and trustworthiness of the observer, and for the evenness of attention during the experiments. It can in some cases be cj^uite inde- pendent of the degree of sensitivity. An observer with a very indifierent threshold may yet give his judgments with extraordinary evenness. The mean variation, there- fore, gives us information about a more general and a more important characteristic of the observer than the stimulus threshold, which we set out to investigate. It can often become of more importance than the latter. We should, therefore, never omit to calculate and mention the mean variation. Our nine-year-old girl showed a very remarkable constancy in her judgments. The mean ^'ariation only amounted to "56, while for Mrs. N., as we can see at a glance, it is much greater. The certainty of the child's judgment was also shown in the fact that she only once (in the method without previous knowledge) gave a wrong judgment. She said at 9° green — " Blue or green." At 10° green came the judgment, " Green." The colour was really a green that tended towards blue. So this judgment was almost correct. The rest of the judgments came in this way : the girl always saw grey so long as the colour-tones were weak, but as soon as they were strong enoTigh, the richt judgment always came at once. On the other hand, the motlier of the child gave many The Measurement of Sensation 69 wrong judgments as long as the colour-tones were weak. The real appearance of the threshold was generally charac- terised by some sucli statement, " This time I am quite certain that it is right." It would be of great importance to investigate sys- tematically tliis problem of the trustworthiness of school- children. We might find out how it increases or decreases according to age, if there are differences Ijctween girls and boys, or if between the different classes or races. The mean variation in experiments dealing with the determination of the stimulus and difference thresliolds would be a great help to tlic investigation we have sug- gested. 5. Peecautions for the Investigation of Stimulus AND Difference Thresholds Whoever wishes to take up investigations of the kind we lla^x> described, must take care to follow strictly quite a numl)er of rules,^ if he wishes his results to be of any lasting value. We shall mention, Iniefly, a few of these regulations which bear upon the experiment we have just described — (ffl) With children the method without previous know- ledge should be used. If, however, it is thought necessary to use the method with previous knowledge (in order to obtain a descending series), then care must be taken not alway^s to begin with the same number, for example, sometimes with 25°, sometimes with 23°, and so on. The child must not get the idea that after a certain number of experiments the judgment always changes. He will, perhaps, begin to count, instead of judging each time independently. 1 Cf. G. E. MuUer, Die Gesichtfimnkte und die Tntmclun der psijehnphysisdien Mdhudih. yo Experiiiicutal Psychology and Pedagogy (b) All results and metliods must he so given, that another experimenter may repeat the experiment under exactly the same conditions. The following details are necessary : How far froni the apparatus the child sat (in our experiment, 2 metres) ; the time of day (2 o'clock) ; the light (i.e. artificial light or diffuse daylight or sunshine, &c.). The equality of the colours used must be exactly determined according to their position on the spectrum. For the isolated experimenter this is not always pos- sible. He will, therefore, do well to obtain coloured paper, which has already l^een tested in some scientific laboratory. (c) All the conditions of the experiment, except the varialDle one (in our case the colour-saturation), must be kept constant during the whole period of the experiments. For example, the grey must correspond to all the other colours in regard to brightness. This can 1)e attained by mixing black and white (see Fig. 34) until a grey, that corresponds to the other colours, results. This grey would then be chosen and adhered to for all the experiments. Further, preliminary experiments must always precede the real scries, in order to allow the ol)ser^'er to become accustomed to tlie apparatus^ and to avoid the great change in the figures due to practice. These experiments must never be included in the results. The experiments must not last too long (for children half-an-hour at a time is sufficient), and the single experiment or test must also not last too long. If we know from the preliminary experiments that the threshold lies at about 10'^, we must not l)egin with 1° (this would be useless and make the test too long) but with 4° or 5°. We have discussed intentionally the determination of thresholds at great length. Child psychology, just like general psychology^ must begin by investigating the most The Measurement of Sensation 7 1 elementary processes. The field for the employment of these psychical methods of measurement is an extra- ordinarily large one. Stimulus and difference thresholds can be determined for all sensations, and especially for the most important sensations, those of touch, hearing, and sight. We can determine thresholds for the intensity (the brightness of a sight sensation, the strength of a tone, &c.), and for the quality of the sensation (the sensitivity for differences in colours and tones, &c.). In the same Avay we can investigate spatial and temporal values. But here we enter into the field of ideas. ]|. AN ANALYSIS OF ONE SENSATION The latest reforms in education have demanded more use of the hand, a greater attention to the sense of touch. If such a problem is to be settled on the basis of psycho- logical investigation, it will be necessary to analyse the sense in c[uestion with the experimental means at hand. We must experiment with adults and children of various ages in regard to the different equalities of this sensation. An investigation of the stimidus and difference thresholds might, perhaps, show us what sensations at what ages could be developed by practice. A series of synthetical experiments might then try to describe the co-operation of the different single sensations in complicated work. Such a description would help to settle pedagogical questions of method. Then the results of such experi- ments in child psychology might lead on to experimental investigation of purely pedagogical problems. Let us follow this out in regard to sensations of touch. Our description will also be an example of the analysis of a special sense by experimental means. 72 Experimental Psychology and Pedagogy 1. The External Touch Sensations The simplest observation sliows us that in touching an object (say in modelling) we have to do with a complex of sensations. It brings to light at once the two most important components of this complex, the external and the internal touch sensations. (a) Contact and Pressure. The easiest of the external touch sensations to be isolated are those of contact and pressure. Experimental investigation is here very i|£.riMHERM»NN;LEiPzia.imM simple. To determme the stimulus threshold for ex- ample, we can place very small weights on the skin. FlG. 37.— von Frey's stimulus hair. gi^^t this WOuld UOt be the simplest case, as this would stimulate a whole surface, larger or smaller, according to the dimensions of our weights. We must, therefore, begin by stimulating a point (punctiform stimuli). For this purpose we should use the so-called stimulus hair of von Frey (Fig. 37). Or we can make such an apparatus ourselves by fastening a human hair to a piece of wood with wax. If I place this hair on an object it soon bends and exerts a constant pressure, regardless of how much it is bent. This pressure I can determine by pressing the hair upon one of the scales of a balance, and then by putting the appropriate weights (very small, of course) on the other scale. If the hair is too weak, in proportion to a certain desired weight, I simply cut a little off. By this means I can get a series of hairs (say 40 to 60) in a regular graduated scale for an exact experiment. •002 "V. oil the foreliead, and •005 gv. on the finger-tips. Ac- cordmg to an absohite iiiea,snre this approximates to j^jQ^y (^rg. From this we see that the sensi- tivity of the pressure sense is much less tliaii that of liearing (stimuhis threshold up to 1 T/ie Mcasurcnicut of Sensation 73 For preUminary experiments and for demonstrations von Frey's hair-sestliesiometer is good (Fig. 38). Here one long hair is fastened in a tube. This is enclosed in a metal case, and by screwing it round I can lengthen or shorten the amount of hair that projects, and thus regulate the pressure. In Fig. 38 at tlie left is an extra case to protect the hair when not in use. If I experiment with the finest stimidus hair, and touch the dorsal surface of the hand, I find that there are only certain points where a sensa- tion of pressure arises. These are called pressure spots. At the other points touched, I feel nothing. The stiniulus threshold for pressure on the most sensitive parts of the skin amounts to \ Fig. 08. — von l'"re_y's liair- ;v3stliusiometei'. 10,000,000 erg) and of sight (up to io-o,oJo,ODo erg). On the other hand^ we must remember that the surface of touch, because of its great extension, possesses a large mmiber of separate individual qualities. The relative values of the stimulus threshold, the difference 74 Experimental Psychology and Pedagogy in sensitivity of the different parts of the body, are of more interest than these absolute values. The following state- ment, that has been maintained by some authorities, ought to be thoroughly investigated, namely, that the lai-ger cutaneous surfaces {e.g. the upper arm) are more sensitive in children than in adults. The difference threshold for pressure can be best \''"^" " ~ '"" \"" ■( KM ^M f/: 1 mJ H 21^ JBLyStjM 1 B^***^^fc*.^' V 1 P'^ ^HBy ^H w iHI^S 1^1 fi^^^Kfl|^^^^^^HB ^^9 m ^^^H^^^B ^3 .1^ ■ "%■'' ■■^^^SH 1 B| 1 ^SB^M 1 ^^1 1^1 Fk;. ^111. — Iii\"estigati(iu uf pret^sure sensitivity. (Photogra|jheil with Wundt's periuissiun in his institute.) investigated by means of the Idnesimetcr of Stratton (Fig. 39). A point resting on the hand is loaded with a normal weight ; by means of pressing a lever an immediate increase in weight is brought about. This prevents any falling or jerking Avhich would certainly occur if the weights were each placed on the scales by the hand. The Mcasiirc7)iciif of Sensation 75 (b) Vain. If I ii.sc ail a'stlu'siometcr with a very strono; liair (e.f/. a liorse liair) I get veiy different sensations at different places. The investigation is again most easily carried out on the back of the hand. At most places I shall feel only pressure, but at some places a most distinct sensation of pain will arise, which could be described as a prick or a. hot jjiercing sensation. By this method we discover the pain spots on the skin. (c) Cold and Heat. Just as we Avent over the skin, touching it with the stiimdus hair, so we can go OA'cr it with a metal rod, which has a blunt ])oint at each end. With this we also get mostly sensations of pressure. But suddenly we shall come to a point where we have the sensation as if a small piece of ice was touching the skin. We have found a cold spot (Fig. 40). We can mark this with ink, and we find that a cold sensation always arises at this special point. On the wrist a great numlwr of these cold spots are to be found. The slight difference in temperature between the metal and tlie natural heat of the part of the body imder investigation is sufficient to give rise to a sensation of cold. It is best to have a cork holder over that part of the metal rod which is touched by the experi- menter, so that it may not be affected by the warmth of the experimenter's hand. If I wish to find warm spots, I warm the metal rod sliglitly. Warm spots are not. so conmion and do not coincide with the cold spots. Fig. 41 show^s the distribution of warm and cold spots on tlie same part of the arm. If I heat the rod only slightly above the temperature of the skin, sensations of cold arise at the cold spots. 76 Experimental Psychology and Pedagogy Tliese are the so-called paradoxical sensations of cold. At the warm spots, with the same rod and the same tem- FlG. 40. — Investigation of tlie cold spots of the sliin. perature, I get distinct sensations of warmth. For an accurate investigation of these conditions von Frey'a The Measui'emcnt of Sensation 77 heat thermometer should be used (Fig. 42). Two rul^ber tubes are attached to the two metal tubes seen in the figure, and water of a constant temperature is allowed to flow through. The exact temperature at any given time can be read off on the thermometer attached. The point, S, that touches the skin has, therefore, a constant measurable temperature. Kecent investigations have shown that sensations of Flo. 41. — Cold and warm spots on the same part of the arm. (From Titchenev's Expcriuicntal Pxi/cholofiji.) heat (in contradistinction to warmth) represent a distinct species of sensation, which probably arises from a stimulation of warm and cold spots at the same time. We see, therefore, that even the externnl touch sensations, which arise say by grasp- ing an object, are most complicated. Physiologically the surface of the skin is like a mosaic of sensitive points, which may be considered as the nerve-endings of different fibres, each class of which transmits a special class of sensations. Fronr different combinations of these special classes arise those sense-perceptions ^ sV of roughness, hardness, wetness, &c., which fig. 12.— vonFrcy's ° . tliurmomi'ter-rod. we know from experience, and which, after accurate analysis, ought to be able to be de- 78 Experimental Psychology and Pedagogy rived from tlie primitive elements. For example that the sensation of wetness arises fmidamentally from the circmiistance of a smooth, cold object moving over a part of the body, can be shown by passing the smooth part of the metal rod we used for cold spots over the back of the hand. The observer, who, of course, must keep his eyes shut during the experiment, generally tries to wipe away " the water." 2. The Internal Touch Sensations Here the conditions are much more complicated. If I investigate the pressure sensitivity of the skin by placing weights on the hand at rest, I find that the difference sensitivity is about ?,, i.e. I must add \ kg. to 1 kg. in order to notice a difference. (With the kinesimeter Stratton lowered it to -^^r.) If, on the other hand, I raise the weights in comparing them (Fig. 46), the difference sensitivity becomes much greater (yV)^ ^^^^^ '^^ know from self-observation that here not only pressure sensations but other sensations take part, mainly sensations of position and effort, the fineness of which must Ije sepa- rately investigated. (a) Positu on. Sensations of position can be best isolated. They seem mostly to take place in the joints. They can be inves- tigated by means of the kinematometer (Fig. 43). One member of the body (say the arm) is so fastened so that only one joint can move. The length of the movement can be measured off in degrees on the apparatus. The experimenter moves the arm into a certain position The Mcasitreniciit of Sensation 79 (passive movement), and the observer lias to say how far the new position is from the old one. The external touch sensations arc not essential to ideas of position. Patients, whose skin is absolutely insensitive, yet possess^ with closed eyes, an accurate idea of the position of the KlG. 43. — Kiiieiuatometei'. (From Stoi'riog, Pliihis. Siudkn, XII. Eiiy-elmaiai.) members of their body, as long as their " joint " sensations are preserved. If these are also wanting, they no longer possess sensations of position. The patient in Fig. 44 was without cutaneous sensitivity and joint sensitivity in the right arm. With open eyes he could hold both hands in the same position, but if his eyes were shut, the position of the right hand changed without him being aware of it. 8o Experiniental Psychology and Pedagogy (b) Effort. In all active movements, e.g. if I raise a weight or make a voluntary movement in the kinematometer, sensations of effort are added to sensations of position. They seem chiefly to have their origin in the muscles and Fig. 44. — A patient lacking cntancous sensitivity and joint and muscle sensi- tivity in the right arm. With his eyes open he can raise liotli hands to the same position ; as soon as he shuts his eyes the jiosition of the right hand changes unconsciously and involiinlarily. (From Striimpell, Zritsehrift fur Ncrrcnli ill^uiiih, VMVl. ^'ogel.) are difficult to analyse because they are always bound up with sensations of position. They can best be investigated by lifting weights. (c) Movement. Sensation of movement is in the main a, complex chiefly of sensations of effort and position. The difference The Measurement of Sensation 8i sensitivity of sensations of movement can be investigated by means of the Ivinematometer. I move the arm of the observer from the original position along to a certain point and then back again (passive movement), and now Kio. 45. — Young children exercising the large joints. (From T;ul(l, Nnic Vi'tgt :ur kiinstlcrischen Er:.iihun;j cl,r Jujji nil. Voigtliinder.) I tell him to move it back again to the same point (rx'tive movement) It seems as if the difference sensitivity of children for movements of the larger joints is comparatively greater than that of adults. If this is verified it means that we should begin to develop the whole hand and arm of 82 Experimental PsycJiology and Pedagogy small children before going on to the development of the finer joints like those of the fingers. This would support the latest methods of teaching young children to draw, by which the arm is first of all brought Ki«. 4t). — Dill'erenci' sensitivity in liCtiiie; wuii^hls. Webei's Law. into play. Fig. 45 shows American children at the blackboard. The organic sensations (e.r/. hunger), which are generally included among touch sensations, can be only mentioned here, because it is almost impossible to experiment with them. The Mcasurcinciit of Sensation 83 III. WEIiER'S LAW I tested the difference sensitivity of the above-men- tioned nine-year-old girl for lifting weights. T put a light dish with a 100 gr. weight into her right hand. Her eyes were l)lindfolded. She was told to lift the dish once up to about the height of her eyes and then to lay her hand down again on the table. A soft covering should be on the table. I then put another similar disli with a 101 gr. weight into her hand, and told hei- to lift it as before. No dift'crence was noticed. Then followed the normal weight again (100 gr.), and t}i(>n for comparison 102 gr. and so on. At 120 gr., i.e. at 20 gr. difference, a difference was noticed.^ I then went through the experiment with a 50 gr. weight. Already at 61 gr. came the judgment " heavier." Therefore with a weight half as lumvy as the former, only about half as much iie(>ded to be added in order tliat a difference should be j^erceived. We see therefore that the addition to the stimulus, which is needed to call forth a just noticealile difference in the sensation, must stand in the same proportion to the normal stimulus. I nmst each time add \ of the normal weight to make a difference perceptible. We have thus arri^-ed at the law, which gets its name froiu the discoverer, Ernst Heinrich Welxu'. In its simplest fashion we can formulate it thus : — The additional stimulus that is necessary in oixler to proceed from one given sensation to another just notice- ably greater, is always for that particular sense a constant fraction of the given stimulus.- ' The same ex])uriiiient. witli the molliei' slioweil that she was suiierini- to hei' cliikl ill lifting weights. At 110 gv. came the judgment "lieavier." Compare on the other hand the colour-sensitivity, page 03. ° This is the simplest expression for the facts under discussion. Geuei'- ally W'eher's law is so formulated : — The stimulus must increase in 84 Experimental Psychology and Pedagogy This fraction is \ for the child in question in regard to the lifting of weights. If I begin with this cliild at 20 gr., I must add 4 gr., if at 200 gr. I niust add 40 gr. and so on, always on the supposition that Weber's law holds good. What does Weber's law mean ? We must call to mind that it is really incorrect to speak of measuring sensations. We have not measured sensations but the relations of two sensations to each other. ^ In all our measurements we had to compare two sensations with each other.^ And it is the accuracy of our comparison that we have measured and not really the size of the sensations. If this is true then Weber's law must possess universal validity, wherever a comparison of the intensity of two sensations takes place. This seems actually to be the case. It has proved itself valid above all in investigations on pressure and force sensations (the pressure and lifting of weights), and on the intensity of light and sound sensations. There are exceptions to AVeber's law, just as there are to every other law. We shall soon learn one of the most important of these exceptions. geometrical progression in order that the sensation may increase in aritli- metical progression ; or: — Sensation is proportional to the logarithm of the stimulus. Fechner came to this foi-mulation by help of the (.liii'erential calculus. These two forniulatio]is however are only valid on the hypothesis that sensations (or sensatiou-ililTerences) can be measured according to their absolute value, a supposition that we cannot accept. We leconmiend there- fore for child psycholog)' and pedagogy the simple fornirdation given in the text, which does not need any such hypothesis. 1 Strictly speaking our chapter ought not to be called " The j\Ieasurement of Sensation," but rather "The Measurement of the Power of comparing Sensations." ^ This also holds good in the determination of the stimulus threshold. I can only perceive the existence of the very weakest, just noticeable sensation in comparison with minimal sensations that always fill our consciousness. No absolutely (juiet place exists. There is no absolute darkness. If I shut my eyes in the darkest room, I still have quite a numljer of very weak, subjective sensations of light. The McasnrcmeJit of Sensation 85 Ebbingliaus es in liis G'timdzUgen der PsijcJtoIoffie the taljle printed below for tke difference sensitivity at different degrees of briglitness. Light Intensity. 0-i I 2 5 10 1:0 no 10(1 200 Difforcuco Light ISunsitiv 1 1 1 (i ity. Intensity. r,( )0 1,00(J "! ■S 1 li 1 2,000 5,000 10,000 1' 20,001.1 "1 "i' I u i -1 .'p fiO.OOO 1 00,01 )0 200,01 )0 — Difroronco Sensitivity. We see from this ta})le that to a liglit of tlie intensity 1, we must add a sixth to perceive the difference ; of tlie in- tensity 2 an eiglitli, and so on. Tlie brightness of the in- tensity 1 was about equal to the light that one would get. " if one were to let the light of an ordinary good stearine- candle fall on to a -^^ery white piece of rough paper from a distance of f metre, and then to look at this light through a small hole one square mm. large. The intensity 2000 equals the lighting of the same paper from the same dis- tance by a strong uncovered electrical arc-lamji of 2000 candle-power." ^ AVe see at once from the table that Weber's law is only valid for the middle degrees of bright- ness, from about 2000 to 10,000 candle-power. Only here is the addition of brightness that is necessary to call forth a new sensation constant, namely about ^^^^. On the other hand, with very weak or very intense stimuli the sensitivity is much less, i.e. not so fine. The explanation of this difference is obvious. In the middle degrees of brightness, to which our eyes are accus- tomed, from the sunlight-flooded landscape down to a Elibingliaus, Grundzugc, Bd. I. p. 523. 86 Experimental Psychology and Pedagog y dull dayliglit, the relations of the separate intensities of light do not alter, although the general brightness may be at times different. Therefore the whole picture remains pretty much the same for our sense-perception in bright or in dull daylight. Only when it becomes considerably blighter or darker, do these relations alter, and then our whole perception of it alters as well. It ha^^pens by very weak light that Ave cannot judge relations Ijetween things at all. What can pedagogy expect from investigating Weber's law on children ? ^ We may say that the more jDrecisely Weber's law appears in a child, the more is his faculty of com^^arison developed, i.e. the faculty of comparing separate sensations with another in regard to their intensity. Secondly we may note how large the " middle region " is, within which Weber's law holds good. There must lie the stinndi, which are adequate for the sensation of children. It is possiljle, for exam])le, that among children nmch stronger intensities of sensa- tions of hearing belong to the normal sensations than among adults. We could test this, of course, by seeing whether Weber's law holds good for them A\'ithin the same region. Then again in regard to the sense of colour, if the child is more sensitive to sim])le pure colours, then Weber's law in regard to the intensity of the mixed colours will not be as 2:»recise as for adults. Those regions, of course, should be first of all investi- gated for the purpose of comparison, A\diere Weber's law has proved itself valid for adults, e.y. for the intensity of sound sensations. For such experiments Zoth's accoumeter is good (Fig. 47). An electro-magnet holds a small steel ball ' No iinpoi'tant investigations are kncjwii to the anllior. TJie Measurement of Sensation 87 which, when the current is broken, falls on a smooth steel plate. The determination of the objective intensity of Fk;. 47. — Zoth's accoumeter. Fig. 4S. — TuniiiEr-forks to dL-termine diti'erence sensitivity. the sound is fixed by taking the product of the weight of the ball and the height of fall (which can be read off a 88 Experimental Psyelwlogy and Pedagogy scale) as the measure. Tlie method of investigation is the same as the determination of any difference thresliold.^ ' By using the small ball in Zotk's apparatus we can investigate the stimulus threshold. It is often used to determine the sharpness of hearing in anthropometrical measurements, whieh often include a test of the senses. The test is made either with the apparatus itself or by using a watch. The ticking of the watch is compared with the apparatus by fixing from what height the smallest l)all must fall in order to give a sound which is e({ual to the tick of the watch. With this watch the tests can be made. If different watches are of different intensities, these differences can also be fixed, and the watch with the greater intensity Avill be used at a coriespondiiigly greater distance. We add here the simple method liy which the difference threshold for tones can be determined. Two similar tuning-forks are used. One of them has a sliding weight to make the tone deeper (Fig. 48). According to the method of limits the normal fork is first struck and then the second, which is tuned much lower. The difference is then gradually lessened, until equality appears and so on, as in every experiment according to the method of limits. CHAPTER HI PJatCKl'TIONS AND IDEAS J. SPATIAL IM'JllUEPTntN 1. Spat[al Pei;,('eption by Touch Perceptions, like all otlier psychical complexes, can he divided up into their elements, sensations. Tliey ar(> not however a mere sum of these elements. In their com- position there arises by means of the arrangement of sensations something new, in the first place their spatial and temporal characteristics. Fk;. 49. — Siieiivman's rcMthesiomuter. Our spatial perceptions, as is well known, arise chiefly by means of the senses of touch and sight. To test our spatial touch perceptions we must go back to the simplest case. If we touch a certain part of the skin with a stimulus hair (Fig. 37) we not only get a pure sensation of touch, but we localise it on a certain part of our body, by arranging it in space with the help of indistinct associations of spatial sight ideas. The accuracy of this arrangement I can measure, if I let two stimuli work upon the skin quite near each other. I can do this with the help of 90 Experimental Psychology and Pedagogy an Eesthesiometer (Fig. 49). The apparatus is made of aluminium so. that it may be placed upon the skin lightly.^ If I wish to measure the spatial threshold of the touch sense with this apparatus, I begin with the single point and then place the two points very close together on the skin, and so on " until the child notices that there are two points (Fig. 51). The spatial threshold of the sense of touch is very different on different parts of the skin. On the finger-tips it is 1 to 2 mm. ; on the upper arm it is 6 to 7 cm. With children it is a little smaller. They are therefore more sensitive than adults. The cause of this can easily be explained. The surface of the skin increases considerably by growth, and very few new nerve-endings arise. There- fore the adult has far fewer pressure spots on a square centimetre of skin than the child has. The spatial power of the sense of touch always works, as we have before mentioned, with the help of associations of visual space ideas, and the closer these two elements, the spatial touch and sight ideas, are melted together, the more is gained for the formation of spatial perceptions. Just here lies the great value of modelling and similar Fig. .'ii'. — Elibinghaus ffisthoyiumoter. " Any oriliniiry pair of coiiipaHses may be taken if iLsed carefully. The distance between the two points must be carefully measured each time and both points must touch the skin simultaneously. It is well to blunt the points slightly. ('/'. Ebliinghaus' ^Esthesiometer (Fig .50). ^ According to the method of limits. Perceptions and Idem 91 92 Experimental Psychology and Pedagogy activities, not in using the sense of touch alone. Touching an object on all sides, running a pencil over the outline of a picture, modelling — all these activities help to develop the spatial sense best of all when the eyes give their assistance. Fig. 52 shows how the spatial touch and Fig. 52. — Children modelling. Drawing-room in an elementary school in Leipsic. (From WeLsscnborn, Neve Bahncn, 190G.) sight perceptions melt into one by means of the attention when the child is modelling. Those blind from birth cannot of course have associa- tions from visual perception, they must therefore make use of different means to help themselves to proceed from the local signs of the organs of touch to spatial perception. In fact they use associations from internal sensations of touch, as is clearly shown in the reading of the blind. 'o PcrcepfioJis and Ideas 9 They first of all run over with the fingers of one hand the letters of the text, which are formed of raised points, and they let the other hand follow closely after feeling each Fig. W.'t. — A blind ijirl readintc. point. The image of the movement of the first hand is associated with the simple touch sensations of the second. They thus achieve some sort of spatial sense-perception of the letters and are so able to read. The girl shown in Fig. 53 reads correctly and fluently like a normal child. Such a capacity is of course only possible owing to the 94 Expe7'iinental Psychology and Pedagogy fact that among the blind the spatial threshold for touch sensations and the difEerence sensitivity for sensations of movement become very fine through practice. Normal children could also achieve a similar perfection with suffi- cient practice. Since with the latter the visual spatial perceptions make up the chief part in the formation of spatial perception, the development of the internal touch sensations is neglected. The unfortunate ones, who are shut out by fate from the world of visual perception, show us clearly how lavish nature has been to man in the possi- bilities of mental development, so lavish that even the blind are a-ble to attain a spatial image similar to that of the normal human being. 2. Spatial Perception by Sight In testing our perception of space from sight let us start, not with the stimulus threshold, but with the difference threshold. For this purpose we draw two lines near to each other, but not too near. Then we increase one of them a little at a time until a difference is noticed. For very accurate investigations the apparatus for testing the spatial threshold is used (Fig. 54). For marking off the distances two points are used. They are marked on two glass sheets that lie one on tlie othei", and these sheets of glass can be screwed up or down by means of a handle, and so the points come nearer or move further apart. As with the sense of touch, the ])crception may take place when the eyes are at rest or in motion. I can fixate an object or follow a line with my eyes. If I wish to investigate, whether the spatial ]ie]'ception of the child is more exact when the eyes are at rest or in motion, I can let one form (say a triangle) be compared with another in two different ways. Either a special Perceptio)is and Ideas 95 point must be fixated or the eyes must follow round tlie outline of the figure. We can then easily determine by which method of observation the judgments are more accurate. I may also use for my investigations the well- known optical illusions, which like the Miiller-Lyer illusion Ktg. ~A. — Aiiparatiis for testiiiE; tlii.' visual spatial tliresholil. (Fig. 55), arise because eye-movements take place. In Fig. 55 the eye underestimates the up^^er line because of the short oblicpie lines that lead the eye inward, and similarly overestimates the lower line. These illusions appear in children just as well as iir acbdts. It proves that in children the eye-movements take the same part in the development of visual spatial perception. 96 Experimental Psychology and Pedagogy Accurate investigations ^ of the diiJerence threshold for spatial sight perceptions of children show that six-year- old children can judge as accurately as fourteen-year-old ones, and that the six-year-old ones are on an average better than adults. It would be of great use if these results were further verified. The investigation can be carried out without any apparatus. Giering gives all details as regards method. If Giering's results hold good, and we can scarcely doubt them, then we must demand in our educational system that more attention be paid to the development of the spatial perception of the child in the first years of school by means of drawing, modelling, &c. The girls were inferior to the boys in the difference <■ > < KiG. 55. — The MiiUer-Lyer ilUision. sensitivity, because the six-year-old boys gave much better judgments than the fourteen-year-old girls. This result should also be verified. If it proves true, we must discuss the problem, as to whether it would not l^e better to pay more attention to the extraordinary colour-sensi- tivity of young girls (sec p. 62) instead of to their spatial perceptions. Besides the estimation of the lengths of lines, we can also test the comparison of surfaces of different magnitude with each other, the estimation of the distance of an ol^ject and the perception of depth. It is of special importance to consider how the different elements in the child's per- ception of depth, those obtained from touch and those ' UicTing, H., Ihia A injciiiiiasx lii'i Srh iilliinleni , Zrit.-n-lir. fiir Paiit-lmliKii,-, ild. 31J. 1005. Perceptions and Ideas 97 from sight, arc related to each other. No such ex^^eri- ments are known to the author. Fig. 56 shows an apparatus by wliiclr we can test as to whether we can estimate with one eye tlie distance of a rod or of a tliiu string, which can be brought nearer or moved away from the eye. In conclusion let us repeat again that in spatial percejotion the best results are always obtained when per- ceptions of sight and of toueli are combined together. Printed letters, for example, will be much better remem- Flli. 5G. — Apparatus fur testing the accuracy o[ our perci'iitioris i.f depith. (Giciing, ZcUsvltrlfl fiir I'siich. v. Physio/, iter f i >,' Fig. 65. — Drawing from memory by an insane patient (adolescent insanity in an advanced stage). (From M'llir, Zrilschr'ift fiir rmr/cioaiidtc Psijcholorjlv, 1908. Barth.) Perceptions and Ideas 107 ing as his cJiildren come from a liilly or a flat country, a city or a country-place, from rich or poor parents. Fig. G6. — Drawicij from memory by an insane iiatient (adolescent, insanity). (From Mohr, Zcitschrift fiir angcioandtc Psycholoiiic, 1908. Earth.) io8 Experimental Psychology and Pedagogy Lotte W. : l.'(A?e 5J) Funnel. 2. (Asje 5J) Bat. 3. (Ag-e 6) Windmill. 4. (Age 6J) Daisy. .5, (Age GI-) Mother bathing cliild. 7 and 8. (Age G) Lamp and hat Lotte W.: 1. (Age .5|) Frog. 3. (Age 6|-) Snail. Helmut W. : 2. : .\.ge 3_^) Pig.' 4. (Age 4.V) Horse. ' 5. (Age fii) Sheep. 0. (Age (U) Goat. Helmut W. : 1. (Age CJ) Sleeping dog. 2. (Age (;-J) Policeman. :->, 4, and 5. (Age 6J) Stag family. 6. (Age (IJ) Hoise witli harness. Fig. G7. — Development of the sense of form ot a sister and brother from the ago o£ 5 to G. (I'lom Weissenboru, Newc Balmvn, 1905-G.) Perceptions and Ideas 109 It would Ije entirely wrong, if lie were to make these difl'erences the starting-point and foundation for a difference in schools. Say, for example, the differences between rich and poor children as an argument for the advantages of our English elementary and preparatory schools. For the sum of ideas that the six-year-old child brings to school depends ujoon the experiences that a cliild has been able to gather in favoui'able and unfavourable conditions and not upon his capacity or talents. A different kind of school, if its existence is necessary, can only have a scientific excuse, if it is based upon differences of intellect and talent. Here investigations of the difference sensitivity, the memory, &c., are alone conclusive. CHAPTER IV FEELINGS I. THE METHOD OF OUTWARD EXPRESSION However difierent tlie definitions of tlie nature of feeling may be among psycliologists, they very nearly all agree in one point, and that is that the feelings are those processes of consciousness that lie nearest to our self.'' For ex- ample, Wundt says, " Feeling is given us in experience, as a subjective reaction of consciousness on an outward impression." We see at once from this that there will be special difficulties in investigating the feelings. If introspection (self-observation), especially for children, is in any case difficult, how much more difficult must it be if the pro- cesses to be observed stand in the closest relation to the observing self, as is the case with feelings. There is, however, another difficulty. Sensations are comparatively constant phenomena. If I observe the green of a meadow for a certain time, the green sensation by no means remains the same, it begins to change from the very first moment of observation. This can be easily proved by observing the colour only with one eye and then from time to time by opening the other eye in order 1 The simple feelings are along with the sensations the onh' elements of consciousness. They should therefore really have been treated before our chapter on perceptions and ideas. We must remind the reader again that our book is arranged rather to follow the methods that may be em]iloyed. Had we followed the psychological sequence we should have liad to separate the simple and compound feelings, and we deem it best to deal with these in one and the same chapter. Feelings III 1 1 2 Experimental Psychology and Pedagogy to compare the change that has been going on. We see from this how a colour loses in saturation for an eye upon which a colour stimulus is working continuously. And yet this is only a difference in intensity. On the other hand everyone knows that with feelings, in accordance Fig. G9. — Kymograph and recording ap[iaratas for experiment ah(twu iu Fig. G8. with their subjective nature, we have to do with a very changeable affective 2:)rocess, whereby one feeling may in some circumstances change into its exact opposite. The most beautiful melody may become unbearable if we are forced to hear it a thousand times in succession. A Feelings 1 1 3 sensation, however, is in its essence tlie same even if repeated a thousand times. There is still a further difficulty. In experimenting with sensations, we used the simj^le impression method.^ We set a stimulus, an impression, to work and observed the changes in our consciousness. These pure impression methods can only be indirectly used in experimenting on the feelings. Suppose that a person feels a certain sweet substance unpleasant. I dare not draw the conclusion that there is some anomaly in feeling in this case. It may be that because of some physiological changes, the observer has not the sensation of sweet at all. Naturally therefore the usual feeling of pleasantness cannot arise. In this casfe it would not be an anomaly of feeling but of sensation. We must therefore, in using the impression method, always ask two things, firstly whether the ex- pected sensation has arisen, and secondly, what feeling arose in connection with it. 1. The Nature of the Expression Method In face of these difficulties it is very good that we possess another method for investigating the feelings, the so-called expression method. The expression method agrees with the impression method in so far as we set a stimulus to work, the effect of which is controlled by introspection (this time of course of the feelings). Alongside of this, however, the experi- menter observes certain bodily changes, the so-called expression movements of the observer. The study of these movements is the essential part of the expression method. • All methods of expei-imental psychology are in reality impression methods. But it is usual to call those, ^yhich only make use of the impres- sion of a stimulus and the following introspection, impression methods. H 1 1 4 Experimental Psychology and Pedagogy Everybody knows that the expression of the face changes under the influence of strong feelings ; that the heart beats differently, that the rhythm of breathing is different, when we are moved by pleasure or pain. We all know that sorrow or great joy may cause tears to flow. The flow of the secretion of certain glands is also influenced by our feelings. The expression methods try accurately to determine (cj[uantitatively) all these bodily symptoms, and thereby arrive at definite conclusions. The following bodily changes can be measured : — 1. The real expression movements. (a) The movements of mimicry, facial changes of expression. (6) Pantomimical movements, the changes in the movements of the members of the body and of the whole body. 2. The so-called expression movements. (a) Changes of the pulse. (6) Changes of the breathing. (c) Changes in the secretion of glands. 2. The Use of the Expression Method Let us suppose that we have established by a number of experiments on children, that a feeling of pleasure makes the heart beat slower and strono;er than usual. We now experiment on a new child, and we find the pulse quite normal, in accordance with an indifferent state of feeling. Suddenly we notice that the pulse is beating more strongly and more slowly. It would be quite false to conclude that the child has now a feeling of pleasure. Eor a thousand other things may cause the pulse to slow down and beat more strongly. Even in the case wliere we Feelings 1 1 5 set a stimulus to work (say sugar), from wliicli we exj^ect a feeling of pleasure, tlie ifact that the pulse alters in a certain way does not in itself justify us in drawing any conclusion as to the feelings of the observer. We need always the statement of the observer as to his feelings. We involuntarily ask, " What is the use then of this expression method % " Wundt's answer to this question is, " The appearance of these symptoms is not a proof but an index of the existence of a certain affective state." Wliat this means and what results the expression method has arrived at may best be seen from a history of the psychology of feeling. One of Wundt's pupils was the first to investigate the changes of the pulse under the feelings of pleasure and pain, and found certain re- lations between these feelings and the changes of the pulse. There appeared, however, by certain feelings of pleasure and pain other changes of the pulse, which could not be arranged under the general law established, and which were, therefore, difficult to explain. Wimdt re- garded this as a sign, that perhaps the old theory of feeling, which accepted only two simple cpialities (pleasure and pain), might be wrong, and that with keener intro- spection we might discover other simple feelings. And his introsjDection led him actually to the assertion that other pairs of feeling exist, namely arousing and quieting feelings, and feelings of strain and relaxation. This was the way in which Wundt's three-dimensional theory of feeling was introduced into psychology based on the results of the expression method.^ ' There still exists no unanimity as to which changes of the pulse correspond to the particular feelings. If for this reason alone many psychologists do not accept the three-dimensional theory, then thej' do not understand the meaning of the expression method. It can only help observation, it can never arrive at decisive psychological lesults. The real decision of such a question must rest on introspection. 1 1 6 Experimental Psychology and Pedagogy In pedagogy we can use the expression method in a similar fashion. Let us choose a very difficult case. Three new pictures are given to us to judge — Fig. 70, " The Young Man of Nain ; " Fig. 73, " Abraham and Lot ; " Fig. 76, " The Storm on the Sea of Galilee " ' — pictures that un- doubtedly strongly affect adults, and each picture in a particular way. Will these pictures also cause strong and different feel- FiG. 70.—" The Young Man of Nain.'^ (Voigtlander, Leipzig.) ings in school-children ? Of course we could ask the chil- dren, as many psychologists would propose to do in such a case. Anybody who has to deal with children would know what the result would be. Nothing is more difficult than to get a statement of their feelings from them, and therefore nothing is more absurd than to put such ques- tions, without testing the truth of their answers by some 1 Litliographic drawings of Haueisen, published by R. Voigtliiuder Leipzig. Feelings 117 1 1 8 Experimental Psychology and Pedagogy objective criterion. For such an investigation the ex- pression method is exactly suited. I recorded the breatliing and pulse of a fourteen-year-old boy and showed him at the same time a picture. Fig. 68 shows the arrangement of the experiment. The boy is sitting behind a cardboard screen which has an opening at the end facing him. When this is opened or closed an electric contact is pressed. This closes the circuit Fic 73. — "Al)raham and Lot." (Voigtliinder, Leipzig.) and a small electro-magnet with a writing-point makes a mark on the drum of a kymograph. Fig. 69 shows the apparatus, above the tambour for recording the breathing, below that for the pulse and in the middle the small electro-magnet. Ijet us now examine the changes of pulse and breathing when the picture appeared (Fig. 71), " The Young Man of Nain " (Fig. 70). The first mark on the middle line denotes]that the picture is visible to the observer. The pulse curve (the lower one) at the same moment rises Feelings I20 Experimental Psychology and Pedagogy with one leap very high ; when the opening is shut and the picture no longer visible (the second mark on the middle line), the pulse sinks almost as rapidly. At each opening and shutting there seems to be an involuntary jerky movement of the hand. We see this in the other curve of the picture, " Storm on the Sea of Galilee " (Curve 77), but not in " Abraham and Lot " (Curve 74). The pulse in Fig. 71, apart from this great displace- FlG. 7(i. — "The Sturm on IbeSea of Galilee. (Voigtliinder, Loii)zig.) ment, does not seem to show any essential changes. At least I would not dare to draw any further conclusions from a curve of such irregularity. On the other hand the change in breathing is quite remarkable (upper curve). Up to the opening regular lareathing, then superficial, irregular breaths, and then, as soon as the picture vanishes, two deep breaths. The deep breaths at the end appear characteristic. We see them also in the curve of a twelve-year-old lioy (Fig. 72). The effect of the picture, '" Abraham and Lot," is Feelings 1 2 1 quite different (Fig. 74). Breathing and pidse continue quietly and evenly as before. There is no trace of a great displacement of the pulse, no trace of a deeper breathing, either with the fourteen-year-old or the twelve- year-old boy (Fig. 75). " The Storm on the Sea of Galilee " gave irregular breathing and unsteadiness of the pulse. Whether dee]:i breathing followed, I could unfortunately not establish, as the available space on the drum came suddenly to an end. It is worthy of note, however, that the pulse is Fto. 77. — Breathing and pul.se curvc-s while loolcing at the picture "The Storm on the Sea of Galilee." (Heinz H., age 14.) stronger during the insj^ection of the picture than before. A strong pulse is, according to Wundt, a sign of excite- ment. Now I am ready to believe the boy, if he tells me that " The Storm on the Sea of Galilee " was the picture that excited him most of all. Without the objective proof with the pulse, I would have attached very little value to his statement. On the basis of the curves obtained, and taking into account the statements of the boy, I would now venture to make the statement that the three pictures excited feeUngs in the boy, strong feelings of quite differing 122 Expcrhncntal Psychology and Pedagogy Fig. 78. — IiiveHtigation of the pulse and hreathing iluring stimulation of tl)c sense of taste. Feelings 1 23 characters. To analyse these feeUngs more accuratelv would be the problem of a more thorough investigation. It would be interesting to investigate whether tlie pic- tures in our schools would similarly aflect the pulse and breathing. The investigation of children with the help of tlie ex- pression method promises not only for pedagogy but for psychology important results, since it is to be expected that children will react more naturally and more viva- ciously to certain simple feelings than adults will. It is difficult with adults to obtain exjjression symptoms of ^jleasure after stimulation with a sweet taste, for the Fjg. 7!I. — Pulse changes wliile tastiiia; ainc. (Heinz H., ai,'e li.) simple reason that a sweet taste with many people does not cause a feeling of pleasure. I obtained, in the experi- ment mentioned at the beginning of the book, negative results from three out of five observers. These were the answers after stimulation with sugar : — A. " Fairly pleasant, then too SAveet ; " B. " A real feeling of pleasure has not arisen ; " C. " This kind of sweetness is really unpleasant to me." With children, especially with girls, better results are to be expected. Fig. 78 shows the arrangement for such taste experiments. The boy re- ceives a few drops of a liquid on his tongue. At the same time pulse and breathing curves are taken. The bitter substance, aloe, causes a drastic change in the pulse curve (Fig. 79). The moment the stimulus 124 Experimental Psychology and Pedagogy works, which is recorded by the mark on the upper line, the pulse becomes smaller and quicker, just as Wundt's theory requires for unpleasant feelings. The effect of sugar (Fig. 80) is less marked. Only in the latter part of the experiment does the pulse increase. We note no slowing down of the pulse. Here it would be advisable to make experiments with girls. The expression method not only serves as an index for feelings that are present, but may also in certain cases help to verify our introspection. Let me give an example. Many defective children show a preference for obnoxious smells. They maintain that these smells are pleasant. NV^J^A^ mm^ mm^ Fig. so. —Pulse chano-es while tasting sugar. (Heinz H., age 14.) It may be possible that in such a case a defect in the sense of smell is present, and that the specific quality of such a smell does not really come to sensation. If, however, the expression method shows us that the characteristic symptoms of pleasure are really present in such cases, then the statement of the child is verified. We then know that there is some radical defect in the child's feelings, and he should be sent to some home for special treatment. n. THE INVESTIGATION OF THE SYMPTOMS OF OUTWARD EXPRESSION 1. The Investigation of the Pulse We must now describe in detail the apparatus, which is used in psychology for investigating the pulse. The Feelings 125 medical practitioner makes use of many methods, in- spection of the pulse by observing those places where the pulse can be seen beating, palpation, i.e. feeling the pulse, especially the so-called radial pulse at the wrist, ausculta- tion, i.e. listening to the pulse beat, especially of the heart. All these methods are too rough and ready for psychology, because it has to take note of much finer difi'erences than the medical man has. It is therefore necessary to have a written record of the pulse, so that we may accurately investigate all its characteristics. We need, therefore, a graphic method. (a) Pressure. An apjDaratus that gives a written record of the ])ulse is called a spliygmograph. If we cross one leg over the other we notice that the upper leg moves up and down, following the rhythm of the pulse. A great pressure is exerted on the artery in the hollow of the knee. When a new wave of blood comes, it meets this impediment, forces its way through, and thus forces the leg slightly upwards. If I were to let a piece of smoked paper move along just touching my toe, the foot would record the movements of the pulse on this paper. We would get more accurate results, if we were to attach (say in the region of the heart) an easily movable lever (made out of a reed or piece of straw) with another lever (the writing-point) so lengthened that the smallest movement of the heart would be registered on a larger scale. It will be much more convenient if we can separate the apparatus that receives the pulse beats, which must of course rest on the jjulse, from the apparatus that writes them down. Such a sphygmograph is shown in Fig. 81. A support is fixed on the wrist with rubber bands. In the 126 Experimental Psychology and Pedagogy middle of this support is a flat liollow capsule of metal, across which on the rinder side a fine piece of rubber is drawn. On this rubber a thin sheet of tin is fastened by means of sealing-wax, and in the middle of this there is a small wooden button, the pelotte (P) . I move the capsule 1 Kt, 81. — Kjmouiaiih .uiil d,pjj.iiatus fui reuouTing the iiulf-e. down until the pelotte just touches the pulse. Every beat of the pulse pushes the pelotte and the thin piece of tin upwards into the capsule, the rubber, of course, offering scarcely any resistance. When the pulse sinks, the pelotte sinks as well. The air in the capsule will be compressed at each rise of the jselotte. These air waves are conducted along a rubber tube into the writing appa- Feelings 127 ratiLS (T), the so-called Marcy tambour, which is just such another capsule, biit this time the rubber is stretched across the upper side. In the middle there is a small vertical point, upon which a steel needle rests. This needle moves with every movement of the pulse. To enlarge the movements, another lever is fastened at the end of the needle. This is a steel needle bent in an upward direction. At the end of this there is a writing- With :i ilouble lever. Fin. S2. — JIarcy tambour. point (A). Compare also Fig. 82. The movements of the pulse are thus enlarged fifty to eighty times. A well-recorded pulse is about 1 cm. high. To the left of the hand in Fig. 81 we see the so-called kymograph, the wave-writer. It consists of a clockwork, that sets a big drum in motion. Before the experiment the drum is detached and covered with a glossy piece of paper. It is well to damp the paper on the inside so that it may he smoothlv on the drum. The paper is then smoked by holding the drum over a petroleum-lainp and b^y turning 128 Experimental Psychology and Pedagogy it slowly (Pig. 83). Then the drum is put into place, the screw (V) turned until the writing-jjoint just touches the Fig. 83. — Smoking; the dram. paper, and the clockwork is set in motion. The writing- point now records the pulse curve. When the paper is covered with records, we take it carefully off the drum, cutting it at the place where it is fastened together. We Pec/ nip's 129 dip it into a solution of shellac in alcohol, such as is used to fix carbon drawings, and let it dry (Fig. 84). Thus we Fl(!. 84. — Fixing and dryiiiii; tin: curves. can preserve our curves for later study. To measure the curves we make use of a glass measure divided np into millimetres (Fig. 85). The form of the receiving apparatus will, of course, I 130 Experiineiital Psychology and Pedagogy vary according to the part of the body we are investigating. To investigate the carotid, the large artery in the neck, we Fig. 85. — Glass plate with millimetre scale for measuring the curves. Fig. 86. — Cardiograph for investigating the beating of the heart. (From LangendorfE, Physiologisclie Qrapliik). make use of the carotid capsule (Fig. 81, C), which is a Feeliuirs 131 hollow funnel-shaped capsule, on the same piinci])le as the other, but without a pelotte in the middle. We press this against the carotid, keeping our elbow on the table in order to hold the apparatus quietly. The carotid capsule is convenient for purposes of demonstration, but is not to be recommendecl for exact investigations, as it is very difficult to hold the apparatus Cjuietly enough for these purposes. Fig. 86 shows the arrangement for recording the palpitation of the heart. Fig. 87 shows the cardiograph Fig. 87.— Cardiograph. that is used. We note a similar metal capsule with pelotte as in the sphygmograph. The support has three feet that can be screwed up or down, so that it may be placed in proper position on the body. (b) Volume. With the apparatus described above the movements of the pulse were measured in this manner, and the changing pressure at a certain point of an artery was recorded. We might take another kind of measurement. A measure- ment of the amount of in-flowing and out-flowing blood. 132 Experimental Psychology and Pedagogy A whole member of the body (say the arm) is put into a rubber sleeve and then into some special kind of jar. When this is filled with water, the rubber presses tightly c I'"iG. 88. — Plethysmograph. Fig. 89. — Plethysraogrniib. on to the arm. The glass tube is then half-filled with water (Fig. 88). Now according as the blood flows in or out of the arm, the volume of the arm is slightly increased or diminished. The water in the glass tube rises or falls Fcelinp's i^'x oo accordingly. This apparatus is called a plethysmograpli (a volume-writer). We can connect by means of a rubber tube with a writing-point, just as with the sphygmograph, and so get a record for the volume pulse (Fig. 89). For many reasons, which need not be mentioned here, it is preferable in experimenting with children to use the sphygmograph. 2. The . Investigation of the Breathing Far easier and simpler is the investigation of the breathing. The first experiments with the pulse often turn out very unsatisfactory. It often requires half-an- hour before one can get a useful pulse record. In in- vestigating the breathing the technicalities are very simple. The pneumograph is a simple flat rubber capsule with a tube at one end. This is tied on to the breast by means of a bandage, high up or lower down according as we wish to investigate the thoracic or abdominal breath- ing. It is best to test both at once. When the chest ex]3ands in breathing, the rubber ball is compressed. The pneumograph is connected with a writing-point, which is exactly the same as for a sphygmograph, except that here only one needle is necessary, since the move- ments of the pneumograph are naturally more extensive than those of the syhygmograph. For demonstration purposes an extra long straw may be attached (40 cm.), as in Fig. 90. We see here also how strong the movements are during mental work. We find similar movements during intense attention, say listening to the weak tick- ing of a watch. Especially in investigating the attention, breathing records ought to lead to important residts. In exact experiments the writing-point should be of the ordinary length. It is necessary also^ in investigating either the pulse or the breathing, to be quite familiar with 134 Experimental Psychology and Pedagogy ■ ;•• . ■; ■■ *.^,...,:;,-,_.. ,, ^^^^== — t : ■ "^ - '".'if^fiS ■''mmmsm^ S; -£.1 fjtJ^BHaa^^ w' ■■■■YfTJlHIff * "^ r .J^/j^^raBWWJ^ ?*l^ 'P?^^ i- .^ r^^i SEiv/ik^HHH '':■?.' i',:'l^ ^fe.;it-i, «. W^ >t-^Sfarii|^^P jtjW ',^i^, ' ^'4"% M^*sJ^?«*W^" .;;^> • -^ jrjBrr^iJK! P*:«: i #. ■'♦ ' ^ iW^rr ■ "■,»■■ ■ - - f'U,;..A«t' ^/1W 1 \ V K9l! ^^ ^ J ^^H^i \ ■H ^H "^^ - "'"vJ^^H ■- .. .' S'f^^RSH &r 1. &' aH^^^^H IHI i ■' 1 r 1 iSfil n ^^H^JHB; I ^ ' ■ ■' i ^^KKBI^Kk- !'■ J_---^ ■^ '■ — .,_^_^ . i ^^^^^^^H^Ht •a ^ ■■HM : ~^-^^_^ ■ 1 ■B^^V'' '^a 4. J ^I^bS^^^^^ ;' ^_^-^ -:, * " ^^^^^bh 1. [■■ r'- h V. X H 1 1 ■ ^ 1 H^B ^^^1 ■ P^^ffi^H ^i-'^^^^^B' '''' -^^^^^H J'celiitgs 1 35 the physiological conditions of the pulse and the breath- ing/ In putting the sphygmograph or the pneumograph into place, there always arises a strong pressure in the capsule which must be got rid of, or else the writing-point will remain pointing upward. To effect this there is a valve in the rubber tube connecting the receiving with the recording apparatus. This valve can be seen in Fig. 90 at about the same height as the watch. This valve is kept open until the apparatus is in place, so that there may be no pressure in the receiving or recording apparatus. It sometimes happens during the experiments, owing to some movement of the observer, that an extra positive or negative pressure arises, causing the writing- point to rise too high or sink too low. This is rectified by opening the valve. Both pulse and breathing exjaeriments have their ad- vantages and disadvantages. Breathing records are easy to obtain, but the movements of breathing are to a certain extent dependent upon the will. If the child notices what is wanted, he may consciously or unconsciously spoil our results. I have generally tried to avoid this difficulty by employing the method without previous knowledge. For example, I say to the child, after fastening on the pneumo- graph, " I am going to try two things. First I will get a record of your breathing and then we shall examine a picture together." In this way I changed from one to the other, of course taking a breathing curve when the child was engrossed in the picture. It is bad not to say anything about the apparatus to the child, for if you do not he will be distracted by trying to think what it is for. Such small deceptions often cause a lot of trouble, but they are absolutely necessary. The movements of the pulse are not dependent upon ' See von Fiey, Die Untersuchiing des PhZscs. Berlin, 18IJ2. 136 Experimental Psychology and Pedagogy Fig. 91. — Pawloir's method ot oiea.'^uriiic: the flow of saliva of the dog. Top, left: determination of the weight of the flow. Top, riglit: graphic record of the flow ; A, at the word " Dinner " ; B, at the sight of the food ; C, at the sight of a led placard. (From Cla[:arede, riiischmi, XII. HJOS). Feelings 1 37 the will. But pulse records are technically much more difficult, and again the pulse is influenced by the breathing (whereas changes in the pulse only influence the breathing very sliglitly). It is therefore useful, in taking jDulse records, to take records of the breatliing at the same time. It is also desirable to take pulse records at different parts of the body at the same time. The changes in secretion of the glands, especially of the salivary glands, have been used as expression symptoms most of all in animal psychology. If a hungry dog sees his food the saliva begins to flow. The amount of saliva produced can be measured (Fig. 91, top left) ; or by means of two capsules (as with the sphygmograph) each dro23 that falls is marked on a kymograph (Fig. 91). Fig. B shows the curve that results when the dog sees his food. The thick black line denotes the amount of saliva that dropped into the glass. The zig-zag line below is made by the apparatus marking the seconds. A dog may be accustomed always to get his food when he hears the word " dinner." If this is so, we find that an increased secretion of saliva takes place when the dog simply hears the word without seeing the food (Fig. A). A dotr may also be accustomed to get his food after seeing a large red piece of paper. If it has become accustomed to this, the saliva begins to flow as soon as it sees the red paper (Fig. C). Such experiments have been carried out by the Russian scientist, Pawloft', and his pupils. 3. Pulse and Breathing Curves An analysis of the pulse and breathing curves can deal with three things — the length, the height, and the form of tlie curve. Let us consider these characteristics for a pulse curve. The length of the pulse denotes obviously the rapidity 138 Experimental PsycJwIogy and Pedagogy of the palpitation of the heart. If I wish to measure it accurately I must have a time record underneath. For this purpose Jaquet's chronometer is suitable (Fig. 149). This consists of a very accurate clockwork, which is in connection with a lever with a writing- point, which gives an upward jerk every second, fifth of a second, &c., according as we adjust it. If I wish to make a more detailed study of the form of the curve, I make the kymograph revolve quicker, and fix the chrono- meter for a fifth of a second. Instead of Jaquet's chronometer I can also use the metronome used in our time experiments (Fig. 58). The negative wire and one or both of the positive wires are con- nected to an electro-magnet with a writing-point, the so-called re- cording-magnet. Fig. 92 shows Lombard's recording apparatus, and Fig. 93 an ordinary cheap record- ing magnet suitable for school pur- poses. A study of the length of the pulse is to be greatly recommended. It is the charac- teristic which presents fewest difficulties in explanation. The height of the pulse may arise from the strength of the beat of the heart. Or, secondly, it may be caused by an enlargement of the artery at the place tested. The walls of the arteries may become tighter or looser according to the condition of the blood or the activity of special nerve-centres. Now if the artery is not very tight the wave of blood will cause a stronger upward movement, and our record will show a higher mark. And, thirdly, there may occur in certain parts of the body (the alD- dominal cavity or the brain) a narrowing of the blood- FlG. 92. Lombard's recorder. Feelings 1 39 vessels from some cause or other. Less ]jlood can there- fore flow to these parts, and the blood-pressure at the places we are testing becomes stronger without the heart beating any stronger, and without any change arising at the places in question. And, fourthly, the height of the pulse changes considerably according as T fasten the sphygmograph tightly or loosely, according to the part of the artery that I touch. ^ The smallest change in the position of the arm often causes very considerable changes in the height of the pulse, as we shall see later in an example. We see, therefore, that we must draw oar Fig. 93. — Ordinary recording magnet. conclusions fronr the height of the pulse with the greatest caution. The form of the pulse naturally differs according to the part of the body I test. Pulses that lie approximately the same distance from the heart are in general of the same form. Each single pulse generally rises pretty abruptly, and then sinks slowly down (Fig. 94 et seq.). During the fall there generally appears a second (sometimes a third or even fourth) small rise, the so-called dicrotism. The normal pulse is dicrotic, a pulse with two beats, i.e. with two summits, the chief rise and the dicrotism during the fall. A probable explanation of dicrotism is the following. 1 The tonograpU is an apparatus for measuring the pulse, absolutely free from any dejections. It consists of a small sharp tube that is introduced directly into the arteiT. This is only used in experiments on animals. 140 Expcriincutal Psychology and Pedagogy As soon as the blood streams into the aorta, this is in the neighbourhood of the heart enlarged on account of the sudden pressure, and this enlargement in the form of a blood wave is transplanted rapidly along the arteries. When it comes to the capillary vessels it cannot continue any further. These narrow vessels present an absolute impediment to the rapidly moving wave, which is thrown back, just as the waves of a lake are thrown back by the shore, even although there may be several small outlets at that part of the shore. The water flows slowly through these outlets (just as the blood does through the capillary vessels), but the more rapid movement of the waves is reflected backwards. This reflected wave appears as dicrotism, once, twice, or three times. If therefore in our pulse the dicrotism is a long way from the chief summit, then the rapidity of the wave is not so great. This may arise from one of two reasons. Either the beat of the heart is weak (the stronger the beat, the cj^uicker the wave), or the walls of the arteries are loose. The tighter the walls are, the quicker the wave rushes along. If the walls were absolutely stiff, say of glass, then the wave would reach the capillary vessels at once. AVe see, there- fore, that the changes in the form of a pulse curve do not lead us to absolutely unambiguous conclusions.^ ^ The same difficulties arise, if we wish to use the lateness of the pulse (how long it takes to come after the heart-beat) as an expression symptom. The piilse wave comes later the further away from the lieavt the artery is. Two different pulses can be recorded, say the radial pulse and the carotid pulse, and the dift'erence in time fixed. From this we can reckon the rapidity of the pulse-wave, if we know the distance from the heart of the places tested. To avoid the necessarily complicated calculations of such experiments, I propose the following method : Lead the tuljes of Ijotli sphj'gmographs to the same tambour, which will now record both pulses. This will of course give a very complicated picture. If, however, a third tube open at the end is introduced into the tambour, each pulse will be recorded only by a veiy small rise. The distance can then be directly measured. (The open tube must be so long that the air waves only slowly flatten out, slowly enough to leave them sufficient strength to raise the writing-point sli;4litly.) Feelings 1 4 1 .Psychological analysis must therefore lay most importance upon the changes in the length of the pulse. In Figs. 94-102 we see at the top the breathing curve, under this the time record (in seconds with the Jacjuet Fl(_;. 'J-i. — Noriual cuive. chronometer), and at the bottom the pulse curve. These curves were taken during the experiment inentioned before, to test the effect of sweet and bitter substances. Fig. 94 shows a normal curve — normal according to Fig. 95. — Normal curve. Traube-Hering waves. the statement of the observer and according to what objectively can be seen (regular breathing and a regular pulse). The pulse recorder, owing to the strength of the pulse, " tossed " a little, and therefore the summits of the curve are a little too high. In Fig. 95 we see how the pulse-wave as a whole 142 Experimental Psychology and Pedagogy moves lip and down, and we also notice that it follows the rhythm of the movements of the breathing. These Traube-Hering waves, so-called from the discoverers, pro- bably arise from the condition of the blood (percentage of oxygen), by which certain nerve-centres are excited (the so-called vaso-motor centres), which cause a rhythmical change in the tension of the walls of the arteries. These so-called secondary waves (the separate pulse waves are the primary waves) have, therefore, no psychological meaning. They can easily be detected by their rhythmical j^rogress, especially if we have a simultaneous curve of the breathing. Fig. 96. — Normal curve. Mayer's waves. In Fig. 96 we see still more extensive oscillations of the whole curve, and they last longer. They are the so-called Mayer waves, and their cause is not yet fully explained. If one notices such oscillations in a pulse curve, it is best to stop the experiment, since an explanation of such a curve is too difficult. The curve in Fig. 97 begins with a high pulse which suddenly sinks down (because of a small movement of the observer), and the pulse rapidly decreases owing to this outside circumstance. In the middle of the curve it has sunk almost to zero. I then open the valve (twice in succession) and the writing-point jumps up and the pulse becomes almost as high as at the be Feelings 14 o Fig. 98 shows tlie different clianges in the pulse that occur in different observers, if the breath is held. The J'lG. 97. — Normaljcurve. Dependence of the height.ofi thefpulse.on the position of the writing-point. ^yVTVAAA ^lu^^^wvu^M-^J>^JU^^^^M^^M^f ,,>jw^w^^A^^'^"'^*-'^^-f^J^^^ ^^Hu^W.^^u^.j^^Ma.*'-'"-'^^'-"'-'-'" "^-^ ^f#AMmwWiimvWfiWwwtpjwjwwwj^y^^ Fig. 98. — Holding back the breath and its influence on the pulse. lowest curve shows scarcely any change. The second shows considerable changes in form, size^ and length of pulse. The upper curve in Fig. 99 was the first that I took 144 Expcriinental Psychology and Pedagogy of the observer in question. It is a so-called normal curve, where no stimulus is at work. The curve shows in pulse and breathing very obvious irregularities. When I asked the observer to give me his experiences during this " normal " curve, I obtained the following answer : " It seemed as if I were at the dentist, who was standing behind me and getting ready to stop my teeth." The observer had before been present as an onlooker at experiments. Fig. 99. — Normal curves. where " fright " stimuli (say a sudden loud noise, smash- ing of glass, &c.) had been used. After I had assured the observer that I would give absolutely no stimulus, I obtained the second curve, which is quite normal. In experimenting with children much greater care must be taken that they become familiar with the condi- tions of the experiment. The first curve in Fig. 100 is a normal curve. The statement of the observer was, " I was dozing peacefully. I thought, how pleasant it is that it is so quiet here, and how quietly and subdued the clock ticks. In the middle Feelings H5 of the experiment I noticed a slight tugging at the heart." The quiet course of the pulse and breathing corresponds exactly to the introspection. The little disturbance at the heart can be clearly seen on the pulse curve. Curve 2 is normal up to the middle. Then the observer received as stinrulus, a solution of sugar. The breathing curve is immediately affected by the movements of swallowing. The observer's experience is, " From the Fig. 100. — Taste expei'imeutH. 1. Normal curve. 2. Solution of sugar. 3. Vinegar. very beginning my state of consciousness was different from that of the previous experiment, more restless, not so unrestrained. This increased at the signal, 'Now.'^ A feeling of strain arose. A real feeling of pleasure did not arise. Perhaps I was disappointed at the slight effect of the stimulus. My state of consciousness then became gradually quieter." In Curve 3 a solution of vinegar was given, again about the middle. Observer's statement, " Before tlie 1 ,S\ich a signal was always given before applj'ing the stimulus. K 146 Experimental Psychology and Pedagogy stimulus I felt a slight tendency to sneeze, which gave rise to a feeling of something comic. At the moment of the signal, ' Now,' this feeling disappeared entirely. The taste itself was rather refreshing and not unpleasant. Perhaps more pleasant than in the previous experiment. There my expectation seemed disappointed, here my expectation seemed relieved." Here we have a case where a sensation of sweetness does not cause a feeling of pleasure, where there is, even with vinegar, only a slight change in feeling to be noted. Corresponding to this we note no essential change in the VA/VA-^V^^ VV^WVAA m^^mmmmmmvmmmmmmmmm.Mm "^^^mmmmmmmmmmm^mmumms^'^^^^'^' Fia. 101. — Attention. Counting the ticks of a watch. pulse. On the other hand these experiments show how exact introspection often leads to very important side- results. The " restless " state of consciousness during the whole of the second experiment is clearly seen in the extraordinary rapidity of the j)ulse-movements. The change to a quiet state is shown in the slowing down of the pulse towards the end of the curve. We also see the effect of the comic feeling on the pulse in the third curve — two very irregular forms. The breathing, however, con- tinues quite regularly. The sudden disappearance of the feeling is shown in the oscillation of the curve as a whole (Mayer's wave), which takes place before the stimulus is applied. In the two curves in Fig. 101 the time-recorder shows Feelings 1 47 two marks. During this period the observer had to follow the weak ticking of a watch. We see how this state of attention shows itself, especially in the breathing curve. Fig. 102 shows the pulse during mental work. At the Fig. 102. — Reckoning. word " Attention \" (!) we notice a lengthening in the breathing curve and a slight Mayer's wave in the pulse. This wave becomes more marked during the difficult task 28 X 12 ; and still more marked Avhen I shouted " Wrong " {fahch), although the multiplication was Fig. 103. — Pulse curve before a 200-ruetre race. correct. The observer stated that " Wrong " caused a very unpleasant feeling, and he began intensely to work out the multiplication again, considering at the same time whether I had not misunderstood his result. That bodily exertion influences pulse and breathing is well known. Figs. 103 to 106 show this clearly. It is in- 148 Experimental Psychology and Pedagogy teresting to note that the largest changes appear in people out of practice (compare Figs. 106 and 107). It would Fig. 104. — Pulse'curve immediately, after the race. (From Schmidt, I'nser Korper. Voigtlimder.) Fig 105. — Breathing curves before and after a cycle run of about 20 km. (From Schmidt, as in Fig. 104.) Figs. lOG and 107. — Breathing curves before and after a cycle run of a lieginner. Fig. 106. Before the start. Fig. 107. I. Loss of breatli immediately after the run. II. 2 mins. 10 sees, later. (From Tissie. L'liyyiene du vdorApedisk.) be interesting to know in what state the pulse and breathing of our children is after the gymnastic lesson, and what differences the different systems of gymnastics cause. Feelings 1 49 It would be very important to test pupils (es2)ocially of higher schools) before and after an examination. III. THE INVESTIGATION OF THE MOVEMENTS OF OUTWARD EXPRESSION In investigating simple feelings it is best to make use of the symptoms of expression as we have shown. In investigating the emotions and moods we do well to in- clude, for purposes of comparison, the changes in real movements of expression. Let us show by means of an example the methods we can make use of here. A few years ago the following question was vigorously discussed : " Are children capable of understanding and enjoying a work of art ? " Let us try to answer this question. 1. Outward Expression by Speech Speech, as is well known, arose out of movements of expression (interjections), and it can be called a system of expression movements, if we take the idea " expression movement '"' in its widest sense. Let us therefore use this means of expression in our investigation of the feelings. Let us show the child a picture and then ask him or her to make some statement about it. After showing the picture in Fig. 108, I obtained the following accounts : — Girl A. We see a beautiful meadow through which a violet river runs. On the left side of the river there stands a row of birch-trees. They have white barks. Behind is a forest. It will soon be smnmer, for the birches have green leaves, and the meadow looks already quite green. B. On the side birch-trees stand, with white barks. 150 Experimental Psychology and Pedagogy The beginning, where the leaves begin, looks red, the other green. C. This is the foot of a hill, which goes high up. We can see this from the wood behind. D. The river must be running quickly, because it tears a lot of the bank away. E. It flows slowly, for its way is easy. Therefore there are such curves. Fig lOS— "The "^alky," by Hein. (Teubner.) P. If one looks at it from the side, one can see the river properly flowing. G. The birches are placed nearly along the bank. H. The meadows lie lower than I or the dam upon which the birches are standing. I. The air seems to be clear. K. The dark-green wood is thick. Not a bit of skv peeps through. We see at once that this method is not suitable. As I hidings 1 5 1 soon as speech sways the course of associations, the feehng- element recedes. The children try to explain the objective content of the picture. Thoughts and judgments as to content appear. Of the efi'ect on the feelings we learn nothing.! We might learn more if we could manage to call forth interjections or exclamations. I managed this in regard Fig. lOU. — "Schiller," b_v Bauer. (Teubncr.) to the picture of Schiller (Fig. 109) by saying, " Look at this head very carefully. The mouth will soon open and begin to speak, just one short word." One of the pupils ' Perhaps we might get sliglitly better results if we permitted tlie children to speak in their own dialect or patois. Dialect expresses a man's inner feelings more accurately than the literary language. It gets nearer to him. This can be seen by asking the children to write down a sentence with a word that the teacher gives them. We chancre now from the literary language to the child's dialect and note the cliauges in the sentences exj. (Cat) "The cat is a mammal." (Pussy) "Oh how the poor pussy miows ! Are you hungry, pussy ? " k ' 152 Expcriineiital Psychology and Pedagogy immediately got up and exclaimed, " Ye miscreants, get ye gone ! Wliat do ye want with me % " ^ 2. Outward Expression by Drawing The drawings of children, their expressions while drawing, might also be considered as a means of ex- pression. We might tell them, after examining a land- scape, to draw what they like, or to draw a figure that would suit the landscape. Children of the age with which I was experimenting (twelve-year-old girls) have not the courage to express themselves freely in their drawings I therefore chose the following indirect method. I said to the children, " When I look at a landscape for a long time, something wonderful sometimes hapj^ens to me. A figure suddenly appears in it. Look at this picture for a long time, so long, until perhaps a figure will appear. Then describe to me as accurately as possible how it looks and what it is doing." The picture was '"' The Valley," by Hein (Fig. 108). The result was as follows :-- A. A little girl is sitting on the bank of the stream with a fishing-rod in her hand. B. A little girl with a red dress and a red hood is sitting in the meadow and holds her head between her hands. C. A mother is sitting with her girl and boy on the bank. The little boy is lying on his back, holding his hand in the water. The girl is also lying on the grass. The mother is sitting. D. Through the avenue of birches a woman dressed in black is walking. She forms a great contrast to the white barks of the trees. 1 The girl used in the original German such phrases as might easily appear, say in Schiller's " Robbers." Slie obviously felt that iSchiller would nse such expressions to the group of children crowding round his portrait. — Translufor's Note. Feelings 1 53 E. On the left bank a girl is lying asleep. She has a green dress on. It matches the nieadoAV. F. On the bank of the stream near to the black bush a girl is standing looking into the water. G. A boy is lying on the bank, his elbows on the ground and his head resting on his hands. With the picture "Autumn," by Ortlieb (Fig. 110), I obtained the following results : — A. On the hill a woman is standing, dressed in black. Fig. 110. — "Autumn," I'Y Ortlieb. (Tcubncr. ) B. I imagine a traveller dressed in black. He is standing at the fence at one side, leaning on the fence and looking at the road that disappears in the distance. C. A woman dressed in black is leaning on the fence and looking at the dark fir-trees. D. A woman dressed in black is standing on the road. Both in the colour and the pose of the figures de- scribed, we see clearly that the mood of the child corre- sponds to the mood suggested in the picture.^ 1 For a full account of tliese experiments, see lUlderhdrachtuNfien, Arbciten aus der Ahteiluiuj fiir Kuitstpfletjc des Le'qniijer Lchrervereins. B. (). Teubner, Leipzig, 1906. 154 Expcrivieiital Psycliology and Pedagogy This method has proved itself better than the other, even although it appears complicated. 3. Outward Expression by Mimicry I have often watched the mimicry of children {i.e. the changes in the muscles of the face) , when I have shown them a picture for the first time.^ If we wish to analyse these movements it is necessary, for the purposes of later study, to fix them in some way. The special apparatus, which is used for recording the changes in the facial muscles, is here of no use. We shall describe it, therefore, when we deal with the mimicry of attention. I chose therefore another method, the photographic one, which up till now had only been used in investiga- tions on the insane. I showed the children a picture and photographed them at the same moment. By this method we certainly do not fix the whole course of the feeling but only a section of it. Nevertheless such photographs give us sufficient material for comparison. Figs. 114-125 show the same child in front of all the pictures that were shown — a scale of afiective processes reaching from unrestrained merriment down to the deepest earnestness. It is worthy of note that this girl, who was not one of the best in the class, reacted most sensitively to the more difficult moods. She was a year older than her school-comrades. ' How capable of expression children's faces are, is seen in Figs. 111- 11.3. The children are tasting sugar, lemon, and aloe respectively. Note the turning up of the eyes of all three children when tasting a sweet substance, the same as with an infant sucking milk. I had hoped to find again the " sweet," "sour," and "bitter" expressions in looking at pictures, and so receive some help in an analysis of the feelings. This only happened with "sweet." At the "sweetest" of the twelve pictures ("A Wealth of Flowers," by Eiese) the same expression of the mouth appeared. Perhaps I gave too strong solutions of sour and bitter. For a fuller description of these experiments, see R. Schulze, Die Mi'inik der Kinder heim kiinstlerischen Geniesscn. Voigtlander, Leipzig, Feelings 155 Tt/m » 1 ii^«»2^Bi.., 1 1 &^^'. 1^ n V''"''l Fig. 111. — A sweet taste — suorar. Fig. 112. — A sour taste — lemon Fig. 113. — A bitter taste — aloe. Fig. 126. Figs. 114-127.— What feelings do the children show 7 An.swer (Observer A): "The fact that the children's faces make me laugh, makes it pretty clear. They show unrestrained laughter and joy. The subject of the picture lies within the children's circle of ideas. It must represent things well known to them. The older girl in the middle restrains herself a little. She seems to say to herself, ' This is too childish for you.' Perhaps it is one of Caspari's pictures." When the twelve pictures were shown, Caspari's "The Insatiables" was at once chosen as the one corresponding to the photograph. Fig. 118. Fjg. 119. Fig. 124. Fig 123. Fig. 1:^8. Fig. 128-131. — Wliat feelings do the children show ? Answer (Observer A): "It is the direct opposite o( the last picture (" The Insatiables "). A little of the previous aoiusement is still re- flected in their faces, but the children restrain it." What do you mean by the direct opposite ? " Perhaps it has something to do with a biblical story. Perhaps angels are on the picture. Cer- tainly something religious." — What picture might it be ? " The subject is not a childish one like 'Christmas Eve.' Per- haps it is 'The Resurrection.' The seriousness shown has some- thing conventional about it." When the pictures were shown, the right picture, "The Crucified Christ," was chosen. Observer B : "The great change in the feelings shows that some- thing exceedingly serious has full}' obliterated the previous impression. Some dying or ill pierson or something very sad." When shown, " The Crucified Christ" was chosen. Observer C: "A picture, very likely a religious one, full of feeling. Quite a contrast to the previous picture. The positions of the hands and bodies arc interesting, especially of thej'girl at the right. All are as chosen. Fig. 129.- ^The Crucified Christ." Her eyes are turned upwards. All are affected" in the same degree." When shown, " The Crucified Christ " wa 158 Expcrinieiital Psychology and Pedagogy Fig. 126 shows a larger number of pupils looking at the picture shown in Fig. 127. Fig. 128 shows the same children looking at the picture " The Crucified Christ " (Fig. 129).^ These were the first pictures which I showed the children. The difi:erence in the expression of the children cannot be mistaken. Now we must find out how Fig. 130. Fig. 131. accurately the expression corresponds to the feeling-tone of the picture. To test this I showed the twelve photographs of the children to four persons, a lady, a scholar, an artist, and a teacher, and told them to describe the feelings of the children and to give some idea as to the picture that would correspond to such feelings. Under the Figs. 126 and 128 some of these judgments are printed. It is astonishing with what degree of certainty observer A, for example, describes the picture, and with Fig. 126 guesses even the painter, without knowing that one of Caspari's pictures was among the twelve. Then I showed the observers the twelve pictures and asked them to find the photograph of the children that corresponded to each picture. Observer A accomplished 1 Figs. 130 and 131 are taken from Duchenne's book — Mn\(nisme de la physiognomie hmnaine. Duclienne stimulated the muscles of the face with an electric current to show what muscles took part in certain feelings. Figs. 130 and 131 reproduced an expression of religious feeling. The second girl in Fig. 128 shows exactly the same expression as the head in Fig. 130, I'^ecliiigs 1 59 this without one mistake, the other observers made only a few sHght mistakes. Our first experiment is tlie more important of the two. Out of an analysis of the observers' judgments we must find what feelings can be gathered out of the faces of the children. 4. Outward Expression by Pantomime It had not been my intention in these experiments to investigate the pantomimic movements, i.e. the move- ments of the whole body and its members. By a lucky chance the photographs also included the hands of the children. I was very much astonished to find here, as well, quite a number of signs as to tlie feelings of the children. Strong excitement shows itself clearly in the tense ujDi'ight position of the body, in the clenching of the fists or in rubbing the right thumb strongly. A quieting feeling shows itself in a sinking together of the body, a slight sjjreading out of the hands or touching softly with the finger-tips. We see in Fig. 132 a picture of joyfid excitement — the clenched fist, the body held straight up, the mouth slightly open and drawn tight (Fig. 136 shows the picture that corresponds to the photograph.) The limp body, the groping fingers and the laughing mouth in Fig. ' 133 show cheerful quietness. (An Alpine valley in lovely colouring was the corresponding picture.) Fig. 134 shows serious quietness ; the body is sunk together, the hands are lying loosely together, the mouth is normal and the brows are drawn together (Fig. 137 was the corresponding picture). Fig. 135 shows serious excitement. The body is held i6o Experimental PsycJwlogy and Pedagogy straight up, the thumb is rubbed strongly, and the brows are drawn tightly together (Fig. 138 was the corresponding picture) . Figs. 134 and 135 are of special importance. Both times landscapes were shown, each of a serious nature. The exciting element in the picture of a thunderstorm, and the quieting element in that of the moonlight — both are clearly shown in the expression of the child's face. i<'iGS. 132-135. — Pantomimic expression movements while looking at pictures. (From Ncue Bahncn. "\'oigtlander.) That hands can also speak, we see from examining Figs. 139 to 143. They tell us about the inner feelings of their possessor. Figs. 139 and 143 show a comfortable groping about with the finger-tips, a pleasant enjoyment. (The Alpine valley was the picture corresponding to Fig. 139, and a shepherd with his flock to Fig. 143.) How different are the hands in Figs. 141 and 142 ! What energy is shown in the clenched fist or in the strong rubbing of the right thumb ! In Fig. 141 the hands of the third scholar cannot be seen. She has raised them and crossed them over her breast. These are pictures of Feelings i6i great excitement, serious in Fig. 141 (" The Dawn," Hang's well-known picture out of the period of the Napoleonic wars) and joyful in Fig. 142 ("The Goblin," Fig. 136). Fia. 13(J.— "The Goblin." (A'liigtLmder.) FiO. 137 — " Wlien the Moon rises,'' by Graf Freiburg. (VoigtUinder.) Fic. l:-!s. — "Poplars in a Storm,'' by Kaujpmann. (A'oigtlander.) 1 62 Experimental Psychology and Pedagogy Fig. 140 sliows the exact opposite. The hands lie loosely together. A perfect calmness. (The picture was " When the Moon rises/' Fig. 137.) Fig. 13!i. Fig. 140. Fig. 141 Fig. 142. Fig. 143. %M W \f Fig. 1.39—143. — Speaking hands. (From Neuf' Biihiii'ii. A'oigtUlD'ler. ) Whoever now puts the question, as to whether chil- dren really understand pictures, will surely be able to answer it. CHAPTER y TI[E WILL I. THE TIME ERllOR IN ASTRONOMICAL OBSERVATIONS 1. Astronomical Methods of Time Measurement The director of Clreenwicli Observatory, the astronomer Maskelyiie, noted in tlie publications of the Observatory in 1795 that he had been forced to dismiss his assistant, Kinnebrook, because he had accustomed himself to a false method in observation. Kinnebrook always ob- served the stars half a second or a wliole second later than the director of the Observatory. The honour of the un- fortunate assistant was saved many years later by the German astronomer Bessel, who established, by com- parative observations, that the records of any two ob- servers never exactly corresponded. Eacli observer differs from any other in a certain manner, so that all his observations differ from those of a second by a certain amoimt. Tlicse phenomena Bessel classed under the name of " the personal equation," without being able to give any expkination. The method of the astronomers was the so-called " eye and ear " method. The observer directs his tele- scope to a certain part of the heavens where the transit of a star is expected, and observes its entrance into the field of vision of the telescope and its passing a thread that is stretched across the middle of the field of vision. At the same time he listens to a clock that strikes the seconds and he then determines at what stroke the transit takes 164 Experimental Psychology and Pedagogy place, or at what distance from the thread a certain stroke of the clock was to be heard. To get rid of " the personal equation " another method was introduced, the " eye and hand " method. The observation through the telescope was the same, but the striking of the clock was abolished. Instead of this the observer pressed down a tapping-key and let go at the moment the transit occurred. The letting-go of the tapping- key set a recording-magnet by means of electricity in action, and the writing-point of this magnet at the same moment made a mark on a small moving strip of paper. On the same paper seconds were marked by another magnet which was connected with an accurate clock. The precise time when the tapping-key had been re- leased could be now read off. At this moment the observer had seen the star passing the thread. But even with this method, which is used in our observatories to-day, the " personal equation " still re- mained, and it was left for experimental psychology to explain the cause of this phenomenon. 2. The Transit of a Star In order to test the processes that occur in astronomical observation, let us try to observe the transit of a star by means of some simple apjjaratus. In Fig. 144 we see the kymograph, which we have already used in recording pulse curves. It can be strongly recommended because it is both cheap and convenient (it can be placed horizontally or vertically). We see that on the axle, which comes out of the clockwork, a small wheel, the friction-wheel, is attached. When this turns, the round plate resting upon it, and therefore the drum, is set in motion. Now according as we place this friction- wheel near or far from the axle of the drum, the drum llic IV ill 165 turns more slowly or more quickly. We also see clearly in the figur'e the two contact-arrangements Ijelow on the axle. Each has a small vertical point. One of them is just touching a piece of brass, and so closes an electric Fig. 141. — Kymograpli with Clockwork. (From Pd-ohrx Catalviji'c) circuit, which goes from the little screw in the front, through the spring that is pressing against the axle of the drum at the bottom, through the contact-rod, the steel- point, the piece of brass, and a screw at the other side (not seen in the picture) from where it could be led further. As soon as the kymograph turns further round, the steel- 1 66 Experimental Psychology and Pedagogy point moves ofi the piece of brass and the circuit is again broken.^ The speeds necessary for astronomical observations and for other psychological time measurements cannot be attained with such a simple instrument. More com- FlCJ. 145. — Kymograph. plicated and therefore very expensive apparatus is re- quired. I have therefore made a small change on the simple kymograph, by means of which a very quick and regular movement of the drum is achieved. The appa- ratus is so simple to manage that even the unskilled may obtain good results. Fig. 146 shows the improved ' The kyiiioc;ra.ph .shown in Fig. 145 i.s one tliat is well .snited for recording pulse and breathing curves (Chaps. IV., II.) The IVill 167 apparatus.^ The spiral spring, S, is tightened by being phaced in position against the bar of steel, W. If the lever F is pressed down the drum will make one revolu- tion, and then be held again in the catch which the lever F has opened. If the lever is continuously held down the drum will continue to revolve. The roller R can be FlO. 14-t). — spring kyinncrraph. turned round a ratchet-wheel, Sp, to tighten the spring. By this means the velocity of the drum may be regulated. To test the velocity, we cover the drum with smoked paper (see p. 128), set it in motion, and let a long steel spring, to which a horse-hair has been attached, mark its oscillations thereon (Figs. 151 and 153 show such oscillations) . ' The jiicture shows an apparatus without clockwork, friction-wheel, and friction-plate. These of course may still remain on the apparatus, so that both slow and cjuick speeds can be olitained. 1 68 Experimental Psychology and Pedagogy These time tests are best taken before and after the experiments, so that the observer may not be disturbed in his observations by the noise of the sjDring. FlO. 1-17. — Ordiuary recorder. (J tec). A little electrical bell may also be used for recording time. Take away the bell and lengthen the clapper with ]''il;. 148. — Electro-magneiio tuning-fork. a long writing-point (Fig. 155 shows at the top such a writing-point and the curve drawn by it) . For accurate time measurements in scientific work electro-magnetic tuning-forks are used. Fig. 148 shows The JViIl 169 lyo Experimental Psychology and Pedagogy such a tuning-fork with the electro-magnet between the two arms. The current flows from the battery to the screw at the right, through the fork itself, through the bent platinum wire to the screw at the left, from there through the electro-magnet to the screw in the middle. From here we connect with a recording-magnet, which writes down the oscillations on a kymograph. When the circuit is closed the two arms of the fork are bent inwards. This prevents the point from touching the screw and the circuit is broken, the arms swing outwards and so on, just on the principle of an electric bell. The tuning-fork may be fastened on a stand (Fig. 149), and so mark its own oscillations by means of a writing- FiG. 150. — Tuning-fork oscillations and two marks of the Jaquet cbronometer. ]Doint. It is most convenient to use a tuning-fork with 100 oscillations per second. The tuning-fork itself may be tested by taking a curve of a Jaquet chronometer at the same time. Fig. 149 shows the arrangement of the apparatus and Fig. 150 the curves obtained from such a test. The chronometer was fixed for marking fifths of a second, and the curve shows that the tuning-fork recorded exactly twenty oscillations during this period. Similarly we can test the spring or the clapper of the electric bell with the metronome shown in Fig. 58. But let us return to the transit of a star. We mark on the drum a vertical line from top to bottom and this stands for the star. (In Fig. 151 we cannot see this line, as it is at the other side of the drum.) The thread of the telescope is represented by the metal stand to the left of the apparatus. It would be better, of course, if we were The Will 171 to stretch a white thread vertically in front of the kymo- graph. The moment the " star " passes the stand, the child must let go the tapping-key (Fig. 152). This breaks the I'lG. 151. — A simple reaction experiment. Grapliic method. Optical stimulus. circuit, which goes from a battery to the recording-magnet, then to the tapping-key, and then back to the battery. Before starting the experiment we place the kymograph so that the "star," the white line, stands exactly behind the stand. We then draw another vertical line on the other side of the drum, exactly at the place where the writing-point of the recording-magnet MjiiiiiiiiiP^^^^^^ ''ijiiiiiiii ■iiiiiiiiiiiil'- Fig. 152. — Contact key. 172 Experimental Psychology and Pedagogy touches the drum. (This Hne can be seen in Fig. 151). Now the experiment can begin. The child presses down the key and looks attentively at the stand. The ex])erimenter stretches the spiral spring, and while releasing the catch says, " Ready." The drum turns, the writing-point of the recording-magnet is pulled downwards by the current and draws a horizontal line in this position. Now when the first vertical line passes the stand, the other vertical line must be exactly at the writing-point of the recorder. The child lets go the key " at once," and the writing-point jerks upward. We see however, from Fig. 151, that the letting-go of the key does not follow " at once," but about one oscillation of the recorder (in this case \ second) later. The child, in observing the transit of a star, has made a mistake of \ second. If we test different observers according to this method, we find that each child has his peculiar way of observing, each one has his " personal equation." Among others we find a great number of individuals who let go the tapping- key before the " star " has gone past. They see the star, so to say, before it is really there. How can we explain this % If we consider this closely, we see that we have here a very complicated process to deal with. It concerns the co-ordination of two groups of muscles, the muscles of the finger and of the eyes. Some observers do not fixate the stand, but follow the approaching line with their eyes (or at least with their attention in indirect vision), and at the same time keep the muscles of their fingers in readiness, in order to achieve as nearly as possible a union of the two movements (the aj^proach of the eyes to the stand and the lifting of the finger). According as they pay more attention to the one or to the other group of muscles, the finger is lifted before or after the objective meeting of stand and line. The JFi/I 173 A psychological analysis would only be possible in this case, if the conditions coidd be simplified. 11. REACTION EXPEET?i[ENT8 WITH THE GRAPHIC MET J [Ob 1. Reaction to an Optical Stimulus Psycliologically considered our experiment will be much simpler, if I hide the left side of the child's field of vision by means of cardboard. Now it cannot follow the approaching line any longer, neither by means of eye- movement, nor with its attention in indirect vision. It can only fixate the stand, and when the line ap- pears, carry out the finger-movement as tj^uickly as possible. What we are measuring in this experiment, a so-called reaction experiment, is an act of Abolition of the simplest kind. Each process of volition, looked at from the outside, can be limited by two moments — the appearance of a stimulus, and at the other end the movement of some muscle. So it is when a baby grasps at an apple. The appearance of the image of an apple is the stimulus, and with the grasping at the apple the process of volition is completed. In our experiment the stimulus was an optical one. We can, however, with a very small change carry out the same experiment with an acoustical stimulus. 2. Reaction to an Acoustical Stimltlus Fig. 153 shows the arrangement of an experiment with an acoustical stimulus. This time only one vertical line is drawn on the drum, and the writing-point is so placed that it just touches the line when the steel-point at the 174 Experimental Psychology and Pedagogy bottom of the kymograph is passing over and touching the piece of brass. In doing this the steel-point causes a Fig. Xvt'A. — A simijle reaction experiment. Grapliic metliod. Acoustical stimulus. sound. The observer has to react on this sound. He should be sitting with his back to the apparatus and his Fig. 151. — Electro-magnetic sound hammer. eyes shut. The experiment proceeds exactly as the previous one we have described. A louder sound must be used for demonstration before The JVill 175 a large audience. For this purpose the somid-hammer sho\vn in Fig. 154 is used. If I send an electric current through the magnet (by connecting the two screws right of E with a battery), the hammer, H, is drawn down and strikes the anvil. If I arrange another circuit by means of the two screws, one at the right and the other at tlie left, then this will only be closed when the hammer touches the anvil. The following is our arrangement for the experiment (Fig. 155) : — First circuit : from the white battery standing on the table, to the recorder at the top of the stand. Tliis recorder was made out of an electrical bell. (Time record.) Second circuit : from the contact screw of the kymo- graph to the electro-magnet of the sound-hammer. (The dry battery stands left in the picture.) This circuit is closed the moment the steel pin touches the brass contact when the kymograph is in motion. Third circuit : from the left battery in tlie front to the sound-hammer, through the hammer, anvil, second recorder, and back to the battery. This circuit is closed the moment the hammer touches the anvil. This moment is marked by the recorder. Fourth circuit : from the right battery in the front to the tapping-key, to the third recorder and back to the battery. This recorder marks the moment when the child lets go the tapping-key. For this experiment I used two kymographs, one with clockwork and one with the spring arrangement, and arranged them on the table as seen in the picture. I thus obtained a long slip of paper to take my records on. In the experiment seen on the picture there are about eight oscillations between the striking of the hammer and the letting go of the tapping-lcey. Eacli oscillation measured 60 thousandths of a second. We generally 176 Experiinental Psychology mid Pedagogy The JJ^ill 177 write 6O0-, o- (sigma) being equal to a tliousandtli of a second. The reaction, therefore, lasted 480a- or about \ second, a very long time for an acoustical reaction. The child was not yet practised in these experiments. The normal reaction time for sounds for adults is IOO-I2O0-, for optical stimuli 180-250t. With our arrangement we coidd also employ a touch stimulus, by connecting an induction-coil with the brass contact of the kymograph, whereby the observer gets a slight electrical shock. The reaction times for touch stimuli are generally still shorter than for sounds. But these differences are of little interest to us, since they most of all depend upon physiological conditions. The stimulus works quicker in the " mechanical " senses (touch and sound), than in the " chemical " ones (sight, smell, and taste). The reaction times for smell and taste are generally very much longer than those for sight. III. REACTION EXPERIMENTS WITH THE REGISTRATION METHOD Instead of recording the time directly (the graphic method), we can also obtain time measurements by means of a special clock. From the position of the indicators we read off the length of the reaction process (the re- gistration method) . Hipp's chronoscope (Fig. 156) is generally used. It is an electric clock, which is driven by means of a weight, and which can be set in motion or stopped by pulling two cords. As soon as one of the cords is pulled the wheels are set in motion (see diagram on Fig. 250). At the same time the metal spring F oscillates up and down, and it is so tuned as to make 1000 oscillations per second. The quickest wheel of the clock S moves the distance M 178 Experimental Psychology and Pedagogy of one tooth further, every time the spring jumps up- wards. This wheel therefore jumps precisely one tooth further every thousandth of a second. The little indi- cator of the upper clock-face is connected to this wheel. Each movement of this indicator denotes, therefore, a thousandth of a second. The indicator of the lower clock-face shows tenths of a second. If at the beginning of a reaction ex- periment the lower indicator Ts^S^Jk pointed to 23 and the upper to 84, and at the end of the experiment the lower to 25 and the upper to 56, we have there- fore moved from 2384 thou- sandths to 2556 thousandths. The reaction time will there- fore be 2556 - 2384 = 172 thou- sandths of a second or 172cr. Now the indicator Z., and a therefore the indicator Z^ do 3 not revolve when the clockwork alone is set in motion, because i the axle xx of the indicator Z„ KiQ. 156.— Hipp chronoscope. goes tlirough all the wheels as an independent axle, and, along with the cross-bar h, fixed on D the axle, is held tight by the spring, which keeps the lanchor m away from the magnet, and therefore, by means of the lever, presses xx to the left, so that the cross-bar li engages the teeth of the crown wheel K,, which is stationary. Therefore the indicator is generally stationary even when the clock- work is in motion. If, however, I send an electric current through the electro-magnet F.„ the anchor tn is pulled downwards, and thereby the screw y moves towards the The IVni 179 right. The spring g jerks the axle xx, the cross-bar li, and the indicator Z., towards the right, so that the cross- bar li engages tlie teeth, of the crown wlieel Ivj (as the first diagram on Fig. 250 shows). Now since the crown wlieel K|, which is in connection with the wheel S (by means of the wheel R„), is in continuous motion, so the indicators are now set in motion by the clockwork. As Fl(i. IfiT. — Wiindl.'s control haiuiiiiT. soon as I shut off the current, the anchor m is pidled up- wards by the spring F and the axle xx with the cross-bar li, and the indicator Z., is again pressed towards the left, and the indicator stops even although the clockwork continues. The indicator only moves with the clock- work as long as a current goes through the electro- magnet. For a reaction experiment we must so arrange it that the stimulus (say the striking of the sound-hammer) i8o Experinieutal Psychology and Pedagogy closes the circuit, and the reaction (say the letting go of the key) breaks the circuit again. There is however the following difficulty. The mea- surement can only be accurate, if the j^uUing of the anchor on to the electro-magnet follows as quickly as the pulling of it away by the spring. Both forces must therefore be accurately balanced against each other. The chronoscope must therefore be accurately tested, by means of Wundt's control hammer (Fig. 157) or Ebbing- I'lG. IGS. — Ebljingliaus' gravity apparatus. haus' gravity apparatus (Fig. 158). Besides this the strength of the current must be measured during the experiments and kept constant by means of a resistance. The arrangements we must make for experiments with such apparatus become, as can be imagined, fairly com- plicated, and therefore they cannot be recommended for pedagogical experiments. For this reason I have described the graphic method in much greater detail, although it is unfortunately seldom made use of in reac- The IV ill i8i tion experiments. The only inconvenience is the reckoning ont of tlie curves. The advantages of the metliod are great simplicity and accuracy. The registration method could only be recommended if we could simplify the chronoscope.^ iV. THE MCTIIOJ) (_)F INSERTION What we liave measured in our reaction experiments is the time of a simple volitional ])rocess. In this time are included, according to Wimdt, the following seven partial processes : — 1. The stimulation of the organs of sense. "| Physiological 2. Tiansmission to the central nervous system, j processes. 3. Entrance of stimulus into the field of conscious--. Psychological pi'ocesses. ness. 4. Entrance of stimulus into the fixation-point of consciousness. 5. The start of the volitional process. 6. Transmission from the central nervous system 1 . , . , , ,, , I Physiological to the muscle. f n r„i , ■ 1 , ■ ,■ . 1 1 processes. 7. The stimulation or the muscle. .' It can scarcely be hoped tJiat we shall ever be able to measure these different processes separately, and so arrive at pure psychological times. This may however be possible by another method. After I have determined by many experiments the simple reaction time of a child, I may complicate the process, I can insert a recognition or choice of stimuli, &c. We shall of coiirse obtain longer reaction times. If we subtract from these figures the simple reaction time, we shall get times which may be called " recognition times," " choice times," &c. Discriminative reactions can be very easily carried out ' See Appendix I. 1 82 Experiineiital Psychology and Pedagogy with our arrangement. There are two contacts with springs on our kymograph. We can therefore choose two springs which give difEerent tones, and we can tell the observer only to react to a certain one of them. In our experiments we change about from one tone to the other irregularly. Similarly we can carry out selective reactions, by demanding the observer to react with the right hand on the high tone, and with the left on the deep tone. Of course there must be two keys and two recorders. Lastly, several stimuli may be used, say one for each finger, and Fig. 159. — Five-fintcer reaction keys. for this purpose two five-finger reaction keys are used (Fig. 159). This method has been called the " insertion " method, because we insert more complicated processes. We have now learnt the three most important psycho- logical methods : — 1. The pure impression method (for sensation and perception) . 2. The expression method (for feelings). 3. The insertion method (for volitional processes). V. MU80ULAE, SENSORIAL, AND NATURAL REACTION Let us return once again to the simple volitional process, say to the acoustical reaction represented on Fig. 153. Even in this simplest case the " personal equation " remains, i.e. there are individuals whose The JVill li o volitional processes take on an average a longer time than those of other individuals. For example, the reaction time to an optical stimulus may be for one individual I8O0-, and for another 250(t. An accurate psychological investigation shows that essential differences in the process of volition go hand in hand with these differences in time. The one observer turns his attention almost exclusively to the stimulus (sensorial or complete reaction), the other almost exclusively to the movement to be carried out (muscular or shortened reaction). Besides these we have the so-called natural or central reaction, where the ob- server divides his attention as equally as possible between Fl((. ICO. — Diagram of frequency curves in reaction experiments, according to Wundt. (From Wiindt, OiitUncs of Psycholo(jii, p. 225. Engelmann.) the stimulus and the movement. A perfectly equal division of attention is however never possible. In every single experiment either the stimulus or the movement receives an essentially larger share of the attention. If I therefore carry out 500-1000 experiments and arrange them according to their length of time on a frequency curve, I will not get a simple curve with one summit, but such an one as is shown in Fig. 160. The two summits show that sometimes more attention was directed to the stimulus (this might give as central value 250■■V .'*■-"' ", - t 1 I'l 1m> ''' I'iiiir'ifc ^^B ■■^ "' "^f^«i. ^^^■•^-^ I'iG. IfiG. — Optical attention. Head and bands raised. How great the importance of the process of attention is in the development of the child can be seen from its physical concomitants, which in young children engage the muscles of almost the whole body. Besides the out- wardly visible symptoms there are changes in the breath- ing and circulation. (We have already described the Coi/sciousNcss and Aftciitioii methods of investigating these.) No one can be in doubt about the fact, that the children in Figs. 166 to 168 are in a state of attention. In Fig. 167 we see how the attention first of all shows itself in the adjust- ment of the sense organ in question. The finer changes, the accommodation of the lens of the eye to the dis- tance, the adaptation of the eye to the brightness Fjg. 1(17. — Eigliteen-iiuiiitli,s-old child looking at a Hying swallow. Optical attention. Fid. UiS. — 'I'll!' same rhilil, a few njonieiits later. 'i'lie tiptical attention is nuioli nioi-o intense. (From Sante de Sanctis, Db Mimik dis /h ub us. ilarliold.) of the object, all these, of course, cannot be seen in the photograph. In Fig. 168 the attention increases, and we see how the motor excitement takes pos- session of the whole body in such a manner that all the muscles (here especially those on the left side from where the stimulus comes) have the tendency to turn the body, and especially the sense organ, towards the stimulus. N 194 Experimental Psychology and Pedagogy The difference between acoustical and optical attention is obvious by comparing Figs. 169 and 170. Fig. IC'J. — Blind children listening. Acoustical attention. Defective children (Figs. 171 and 172) show an abnor- mal (exaggerated) expression of attention or an absolute want of any symptoms of expression. Coj/sciousiiess and Attention 195 At a later age tlie mimicry of attention is gene- FiG. 170. — A (leaf and dumb boy reading from the lips of his teacher. Optical attention. rally limited to changes in the muscles of the forehead. 196 Experiinental Psychology and Pedagogy During the first years of school, however, the expression of attention is generally fairly well marked ^ (Figs. 173 and 174), and it would therefore be a very thankful task to investigate by means of the photographic method these ^^hysical symptoms of expression in school-children of every age. In the upper classes an investigation of the muscles of the forehead is to be preferred. Here the method described by Professor Sante de Sanctis is to be recommended : '^ KlOS. 171 and 172. — Defective chilili en. Left, ina state of inrtifterenoe. Rigbt, optical attention ; hei'e the mimici'v seems to denote pain, as if the boy wislied to protect his eyes against some too powerful light. (From Sante de Sanctis, as in Fig. I In the eight-year-old chihjren (Fig. 173) the miiiiici;y of allentioii, which i.s most of all concentrated in the muscles of the forehead, spreads out to the most outlying groups of muscles even when the thinking is simple. When the task is more difficult, as in aritlimetic, it is as if a wliirlwind rushed over the small faces and shook every hranch and twig (h'ig. 174). - Sante de Sanctis; AJiiiilk ilrx Deiil'iiix. jNlarhold, Halle, IDOti. Consciousness and Attention 197 J r^»- * f \T 1 « JMi i''-'' fl ■^^^^Kj fei? / >x ■PvJ ^1^^ s> ■ " ■ J n 1 ^^^B ^^IQrW 1 •<«4 1 ^ 1 j ^ ^, -.; ^ ^ !■ 1 1^ '■■'-it: •< i :-:;:; !P|f€ 't V' \ \ \} r«i \ 1 - '' ^HudP^^F't (^Bf' ^ri^^^^^BA ^^^■P^bl#^ '^ T %\ \ 1 ^Kifl^M \ \ , ^■^ > ( « ^ ^ o o -a 2 198 Experimental PsycJiology and Pedagogy Consciousness and Attention 199 graph permanent." In laying on tlie paper the child must, of course, be in the state of attention we wish to investigate. Fig. 175 shows a photograph of the muscles of the forehead obtained by this method. A similar method was invented by Professor Sommer. He takes a piece of smoked paper and presses it quickly Bf:*'«Sk'-?'^<&S^!^ ^JBJJHHBBP'K . ' ' yjl ^J^i^^^^':'^'. - 'i^ ^. -,i^^I4 . J' . ' - . . * -- TT Vu\. 175. — Contraution of the inii: