^la COLUMBIA LIBRARIES OFFSITE HEALTH SCIENCES STANDARD HX64101908 QP421 .M35 The diurnal course RECAP . Q PU I A1J5, intljpCttpoflfttifork College of ^tipsiiciansi anb burgeons! ILiiirarp Digitized by tine Internet Arciiive in 2010 witii funding from Open Knowledge Commons (for the Medical Heritage Library project) http://www.archive.org/details/diurnalcourseofeOOmars '^ h redone o, Lee Colui:: 'i U'.i'-crsity THE DIURNAL COURSE OF EFFICIENCY BY HOWARD D. MARSH, Ph.D. ARCHIVES OF PHILOSOPHY, PSYCHOLOGY AND SCIENTIFIC METHODS BDITBD BT J. McKEEN CATTELL and FREDERICK J. E. WOODBRIDGE No. 7, JtTLY, 1906 Columbia University Contributions to Philosophy and Psycbology, Vol. XIV.No. 3. NEW YORK THE SCIENCE PRESS ARCHIVES OF PHILOSOPHY PSYCHOLOGY AND SCIENTIFIC METHODS Editorial communications should be addresaed to Professor J. McKeen Catteii, Garrison, N. Y,, or to Professor Fredekick J. E. Woodbridgk, Columbia University, New York City. Subscriptions and advertisements should be sent to The Science Press, Sub- Station 84, New York City. The subscription price is five dollars a volume contain- ing between six and seven hundred pages. The numbers are as follows : 1. Measurements of Twins : Edward L. Thorndike. 50 cents. 2. Avenarius and the Standpoint of Pure Experience : Wendell T. Bush. 75 cents. 3. The Psychology of Association : Felix Arnold. 50 cents. 4. The Psychology of Reading: Walter F. Dearborn. Sl.OO. 5. The Measurement of Variable Quantities : Franz Boas. 50 cents. 6. Linguistic Lapses : F. L. Wells. $1.00. 7. The Diurnal Course of EfBiciency : H. D. Marsh. 90 cents. 8. The Time of Perception as a Measure of Differences in Sensations ; Vivian A. C. Henmon. 60 cents. 9. The Psychology of Mentally Deficient Children : Kaomi Norsworthy. In press. THE JOURNAL OF PHILOSOPHY PSYCHOLOGY AND SCIENTIFIC 3IETHODS The contents of recent numbers include : The Knowledge Experience and its Relationships : John Dewey. Kant's Doctrine of the Basis of Mathematics : Josjah Royce. The Psychical Complex called an Interest : Lucinda Pearl Boggs. The Issue between Idealism and Immediate Empiricism : Charles M. Bakewell. Mental Elements of Dreams : Will S. Monroe. Feeling and Conception : Kate Gordon. Cognitive Experience and its Object : B. H. Bode. Association and Atomism : Felix Arnold. A Philosophical Confession : Harald Hoffding. A Syntactician among the Psychologists : Basil L. Gildbrsleete. The Real and the Pseudo Psychology of Religion : Irving King. The Reeducation of an Aphasic ; Shepherd Ivory Franz. An Apparent Contradiction in the Modem Theory of Judgment : W. B. PiLLSBURTf. An Empirical Definition of Consciousness : Wendell T. Bush. How Two Minds can know One Thing : William James. Some Outstanding Problems for Philosophy : Cassius J. Keyser. Immediate Empiricism : John Dewey. $3 a Year (26 numbers) 15 cents a Copy. LIBRARY OF PHILOSOPHY PSYCHOLOGY AND SCIENTIFIC METHODS Theory of Mental and Social Measurements : Edward L. Thorndike. 81.50. Science and Hypothesis : Henri Poincar6. Translated by George Bruce Haustkd, with an Introduction by Josiah Royce. $1.50. THE SCIENCE PRESS, Sub-Station 84, New York City. THE DIURNAL COURSE OF EFFICIENCY BY HOWARD D. MARSH, Ph.D. ARCHIVES OF PHILOSOPHY, PSYCHOLOGY AND SCIENTIFIC IVIETHODS EDITED BT J. McKEEX CATTELL, and FBEDEBICK J. E. WOODBBIDGE No. 7, July, 1906 Columbia University Contributions to Philosophy and Psychology, Tol. XIT, No. 3. NEW YORK THE SCIENCE PRESS 'C Press of The new era printing Company lancaster, pa. TABLE OF CONTENTS PAGK Part A. Introduction 1 Part B. Diurnal Course of Psychophysiological Effi- ciency 4 I. Vital Activities 4 1. Secretion, Urination, Evaporation 4 2. Circulation, Respiration, Temperature 4 3. Deaths relative to Time of Day 9 4. Brief Summary 13 II. Sensory Activities 13 1. Sight, Color-sense, Hearing 13 2. Skin-sensitivity, Pressure-sense 14 3. Brief Summary 14 III. Motor Activities 14 1. Subhuman Data 14 (1) Plant Life 14 (2) Animal Life 15 2. Human Data 15 (1) General Characteristics of Writer's Methods and Experiments 15 (2) Actual Tests and Results— Writer's and Others' 17 a. Speed and Accuracy of Movement 17 6. Normal Muscular Power 24 (a) Small and Large Muscular Groups 24 (&) Comparative Functioning of Such Groups 31 (c) Bilateral Symmetry 31 c. General Motor Control 32 d. Complex Motor Activity— Athletic Perform- ances and Manual Labor 32 3. Brief summary 40 iii iv CONIENTS PAGE Part C. Diurnal Course of Mental Efficiency 42 I. The Simpler Mental Activities 42 1. Attention 42 2. Discrimination— Letters and Forms 44 3. Association— Reaction-time, Naming Colors, Word Associations 46 4. Memory (mechanical)- Visual, Auditory, Visual- auditory 50 II. The More Complex Mental Activities 53 1. Arithmetic— Addition, Mental and Written Multi- plication 53 2. Translating French ; Scoring Death Eecords 55 3. School Examinations and School Marks 57 4. Students' and Authors' Preferred Hours for Work 59 III. Brief Summary 69 Part D. Chief Causal Factors in the Diurnal Efficiency Curve 71 I. Night-day Rhythm — Sleep and Activity 71 II. Habits, Meals, etc 74 III. Fatigue and Feelings of Fatigue 77 1. Muscular Fatigue 78 2. Mental Fatigue 78 (1) Measured by Motor Inefficiency 78 (2) Measured by Sensory Inefficiency 79 (3) Measured by Mental Inefficiency 82 IV. Inertness and Nervousness 88 V. Sex and Age 90 VI. Brief Summary 91 Part E. Conclusion 93 Bibliography 97 THE DIURNAL COURSE OF EFFICIENCY • A. INTRODUCTION Recurrence of phenomena is the source of all scientific knowl- edge. Rhythmic recurrence in nature is well known; for physics, it is- perhaps the most general and fundamental characteristic, under- lying the theories of the mechanics of motion, sound, heat, light and electricity. The student of astronomy finds the movements of heavenly bodies marked by like repetition. It appears in biology in a less regular way in the growth changes of vegetable and animal bodies and in a more general way in the successive generation of species ; while evolution as a whole is rather completely expressed in alternation of integration and disintegration. Social life, in all its historic aspects, gives evidence of epochal development, and some sociologists have gone so far as to make repetition (as imitation, etc.) the warp of the whole social fabric.^ Nearer the point of the present research are particular seasonal changes in individual life — yearly, quarterly, monthly, weekly, daily — of whose existence we have hints in various anthropological, physiological and pathological data, more or less closely connected with the sexual, religious and emotional life.- This paper deals only with the question whether there are nor- mally recurring variations of ability at different periods of the day. The discussion of this question is not to be found in general works on psj^chology, however modern, but in sundry monographs and journal articles appearing in the last fifteen years. This means that the more conservative psychologists are not yet ready to accept such recurrence as a fact. But the investigators themselves seem to consider it quite confirmed, despite the disagreements of their results when intercompared. Part of this assurance may be due to ignorance of what others have found. The good side of this is that experimentation under such conditions has less tendency to bias, to which workers in indi- vidual psychology are perhaps more exposed than are those in other parts of the experimental field. We want not only the subjective ^ For complete theory, see G. Tarde's Laics of Imitation and other social writings. ^Of course there is no doubt about the monthly rhythm in females. For other suggested ones, see H. Ellis, The Psych, of Sex, 2. 1902. 2 DIURNAL COURSE OF EFFICIENCY satisfaction, but also the objective recognition, of having obtained results with a definite pointing— which 'show something'. But that this should be exhibited from the negative side seems often as objec- tionable to the searcher as discovery of no 'general tendency'. The writer aims to present all the important work hitherto pub- lished, having direct bearing on the problem. On account of the scope of the latter, no collection of this material at any one point is attempted, but it is introduced where most pertinent. The greatest lack in all this work lies in two things— inextensiveness of the experimental series and omission of quantitative expression of the reliability of the results. It is easy to account for the first of these deficiencies when one remembers the nature of the problem, affecting as it does the whole duration of the day. This makes it hard indeed to secure persons to act as subjects. If the tests are short, they must be repeated on a great number of days, in which case the various conditions must be so much the longer controlled; if they are individually more ex- tended and adequate, they demand most of the subjects' time for the days on which they are taken, and in this case accidental disturb- ances are more costly. Therefore the investigator is inclined to reduce either the number of days, the number of periods per day, the number of trials per period or the number of subjects tested. This being true, the second lack mentioned is easily explained. The neglect to calculate the averages obtained — a very serious omis- sion when differences are small and results not always harmonious — arises partly from the feeling that the material at hand is too meager to justify the calculus of probabilities. An additional difficulty is that in a series of tests day after day the curve of diurnal variation is complicated with the curve of progressive improvement due to practise. This makes it especially hard to calculate the reliability of the averages for the different hours of the day. And even the mere labor of the tests and computations is a deterrent ; for it is clear that if 500 trials are thought necessary to establish a reliable measure of any function, ten times that many and ten times the cal- culations will be required to establish it for ten periods of the day. Other minor deterrents might be enumerated that operate in certain cases. One other occasion for criticism of the works referred to is found in the fact that the tests for the different parts of the day have not always been made on the same day. Such procedure implies the belief that the absolute ability one exhibits on one day is the same as he would manifest at the same hour the next day or any subse- quent one : it takes no account of accidental influences, practise INTRODUCTION 3 effects and other inherent factors of change. To keep clear of these errors, the tests for the different hours of the day should be made on the same days. My own work is not free from the first of these criticisms, but avoids the remaining two. In addition to presenting the work of others, the author, by a great variety of extended tests upon himself, has been able to estab- lish definitely the course of diurnal changes for one normal subject. Six male graduate students, one female graduate, and a group of sixteen female undergraduates were likewise employed for both motor and mental tests, and the measurement of the hourly product of 22 female factory operatives contributed additional data. The tabulation of human deaths for time of day; the results of school examinations and recitation marks for different diurnal periods; the best times of day for athletic performance and for intellectual composition ; in short, as wide a class of pertinent matter as could be secured has been applied to the solution of the question of diurnal changes of efficiency. In the experimental portion of the present work, much use has been made of tests variously employed by others. The aim has been to get such variety as the practical limitations of application would allow. It is not thought that together they give a measure of 'gen- eral intelligence' or 'general motor power'. These themselves have rather a precarious existence since analytical psychology has taken an experimental turn. Our activities are recognized as much more complex than they seem, and tests of them, on the basis of simplicity, may not be productive of the most valuable knowledge. Especially is this true of adults, to whom my own tests were chiefly confined. Still it is true that legitimate and interesting results have been reached in the past and that accruing experience must make future eft'orts more fruitful. As the total outcome of his research, the author is not prepared to 'say the last word' about diurnal rhythms, but hopes that something worth while has been attained. Acknowledgments are due to Professor Cattell for the original proposal of the problem and for other suggestions; to Professors Thorndike, Woodwcrth and JMejdan, of Columbia University, for assistance of various kinds; to those who acted as subjects at much personal inconvenience ; to certain officers and employees of the manufacturing firms of Dennison and Sons, J. English and Sons and the Trow Directory Company, of New York City, for courtesies extended during observational work in their respective factories ; to Dr. Guilfoy, registrar of the City Health Department; and to other individuals who contributed in different ways. B. THE COURSE OF PSYCHOPHYSIOLOGICAL EFFICIENCY I. Vital Activities Any strict line of demarcation between so-called physical and mental life does not exist in reality and is maintained primarily for convenience in dealing with the complex phenomena of life. For the data to be presented in this section the term psychophysiological seems best, since from the physical side only the functional aspect is dealt with. What will be called 'vital' activities are on the mental side directly concerned with emotion; 'sensory', with cognition; and 'motor,' with volition, as the words are commonly accepted. The idea is not to support a three-faculty psychology nor to stand sponsor for a rigid meaning of the terms employed. But a more systematic presentation of the subject can be given by adherence to the order named. Considering the length of time that medicine has been studied, or even that physiology has been a separate science, one would ex- pect to find the matter of physiological rhythms fully discussed. However, the text-books here are as barren as those in psychology with respect to the general problem. The first and only systematic collection of results was published by Vierordt in 1888\ The actual work was done by various medical men between 1840 and 1880, and much of it must be considered unreliable, particularly where the conditions of experimentation or observation are not stated. Only such figures as seem most trustworthy will be quoted. 1. Secretion, Urination, Evaporation Vierordt 's results here have no value in themselves without fuller information as to number of subjects and their habits of life, their age, sex, health, etc., together with the length and method of experi- mentation. The following are less deficient in these respects. 2. Circulation, Respiration, Temperature 1. Circulation. — The most commonly accepted view with regard to pulse rate is that the maximum occurs in early morning, followed by a steady decline till after midnight, broken only by the temporary "■ Daten und Tahellen. 1888. COURSE OF PSYCHOPHYSIOLOGICAL EFFICIENCY 5 influence of meals and accidental excitements. This seems only partly true. The correctness of the first statement is greater the nearer to arising the record is taken, as will be indicated below. The following table gives most of what material could be gathered on this matter, together with some of my own. Table I. Pulse Rate and Time of Day. 7 A.M. 8 9 10 11 12 1 P.M. 2 3 4 5 6 7 8 9 10 11 12 1 * 74 71 70 69 81 84 82 77 76 75 75 2, 68 * 71 68 68 *72 2: 73 * 83 76 68 *69 2, 90 * 105 97 82 *97 3 68 * 81 87 73 79 67 82 83 73 73 70 *85 4, 56 * 72 69 * 65 4, 72 * 86 83 81 76 73 79 77 75 72 78 *84 82 77 73 70 69 68 4, 71 * 78 74 70 77 72 70 72 * 78 76 71 4, 84 * 94 98 ■ 84 78 77 - 82 76 *86 72 1 — Yierordt. Self, number of days not stated. 2 — Binet. 2i self, 7 days; 2. male adult, 2 days; 23 female adult, 2 days. 3 — Storey. Self, average of to 10 determinations. 4 — Marsh. 4i male adult, days; 4. male adult, 15 days (summer); 43 same, 12 days (winter) ; 4^ female adult, 7 days. * — Indicates occurrence of meals. Not all the series approach completeness. Small dependence can be placed on 1^ ; Binet" says of his own that ' ' they are three single curves chosen from a great number because of the distinctness of the tracings", not saying that they are even representative ones; 4i is a short series but with small probable error. The first four and 4^ show the maximum rate at about 1-2 p.m. (2 also at 9 p.m.), the othei-s at 8-9 a.m. The changes to which we are subjected on awakening act as excitants and largely cause the apparently high early morning rate. Change of position alone has great influence. Ten mornings, in my own case, the rate two minutes after arising averaged 81, as compared with 72 twenty minutes later ; and again in the second series the figures were respectively 81 and 71. Subject 4^ did the same thing a like number of days, taking an additional record in bed several minutes before arising. Her aver- ages were 79, 87 and 84. These facts suggest what small reliance can be placed on bare pulse figures in tables. AYhat seems to be true here, aside from the doubly induced increase due to rising and eating, is first a morning rate less high and less well sustained than the afternoon rate ; second, a higher female than male rate, with ^Yet, as late as 1900, the Am. Text-book of Physiol, quotes these figures as bases for conclusions. This shows how meager is the information. " ' Le Changements de Forme du Pouls Capillaire,' L'Annee Psych., 3. 1897. 6 DIURNAL COURSE OF EFFICIENCY tendency to an earlier maximum^ ; and third, a general rate higher than the 70 so long accepted as about the average. As to the first point, the order is sustained by a recent study of blood and pulse pressures made by Erlanger and Hooker^ with one adult subject. Five extensive tables show the results obtained for five days, at different hours from 8 a.m. till 10:30 p.m., but their reproduction here is forbidden by their great length and uncertain value. However, the authors draw this conclusion: "We therefore can distinguish a gradual increase of pulse-pressure [directly pro- portional to pulse-rate according to them — p. 294] throughout the day upon which is built up the wave-like increase that follows upon the ingestion of meals" (p. 343). "There seems to be no relation between the amount of food ingested and the height and duration of the post-prandial [fundamental] rise" (p. 344). In addition to this direct evidence there are other facts in favor of the view that the pulse is normally lower in the morning than later in the day. IMuscular and brain activity heighten the pulse rate on account of the increase in waste products sent to the heart. If it be true that a greater muscular power accompanies increased rate of circulation, as shown by Zablondousky^ and Maggiora^ with massaged muscles, and by Kronecker^ with injection of blood into fatigued muscle, then, according to the traditional pulse theory, the greatest strength might be expected in the early morning. As a matter of fact, it is probably least then, as will be shown later. Periodicity in pulse intensity and blood pressure, dicrotism and the rapidity of vascular reflex have been investigated, but the results are unimportant. 2. Respiration. — One expects to find a positive correlation be- tw^een pulse rate and respiration and temperature, as they have a common dependence on the amount of metabolism in the body, the latter in turn being largely proportional to the excitement or activity at any period. And it is found to be so in fact. Vierordt's maximum for himself was at 2 p.m. for rapidity of respiration and volume of air and of COo expired, and minimum at ^ The high night rate of 23 is due to the evening meal and no comparable morning figure is shown, while the case of 44 shows a considerable falling off towards night. Guy and Knox, per Vierordt, found a higher rate for children Avith an early morning maximum. Miss N. Norsworthy, of Teachers College, in a recent research, the results of which are as yet unpublished, found for girls under 15 years a rate of 9.5.5, and above 15, of 93 — 295 cases; for boys under 13, 95, and above 13, 85 — 253 cases. 2 ' An Experimental Study of Blood-pressure and Pulse-pressure in jNIan,' Johns Hopkins Hosp. Reps., 12: 145-378. 1904. ^ J. Joteyko, ' Revue Generale sur la Fatigue IMusculaire,' L'Annee Psych., 5. 1898. COUBSE OF P.^YCHOPHYSIOLOGICAL EFFICIENCY 7 7 P.M. — the last hour of his series. Binet's maximum occurred at 7-9 P.M. and minimum at 8 a.m., his series being from 8 a.m. to 8:40 p.m. Tigerstedt^ (after Jiirgensen — see Fig. 1) finds the maximum volume of COo at 5-8 p.m. and minimum at 4-8 a.m. Fig. 1. Normal .Variations in Body Temperature (after Tigerstedt). Johannson^ tinds that the greatest quantity of CO2 is given off by an active subject from 8-12 a.m. and the least from 4-6 a.m.; this was practically true also during perfect rest. His experiments were on himself, in six-hour courses of activity and, again, of in- activity (in bed), requiring different days to cover the whole twenty- four hours. In accord with this evidence are certain data obtained from spirometric tests of Filipinos at the St. Louis Exposition,^ where 72 Avere tested for lung capacity in the morning and 41 in the after- noon. The average for the morning was 170.0 cu. in., with a prob- able error of 1.76 ; and for the afternoon, 177.2, with P.E. 2.60. In my own case, eight trials at each period showed averages as follows : 7-8 a.m., 189; 9-10 a.m., 192; 11-12 a.m., 191; 1-2 p.m., 188; 3- 4 P.M., 192; 5-6 p.m., 192; 7-8 p.m., 193; 9-10 p.m., 191. This fol- lows closely the curve of my motor ability as found by many tests. 3. Temperature. — Reference to Table II. and Fig. 1 makes it evident that the temperature reaches the highest point about 5- 8 P.M. and the lowest about 7 a.m. Johannson finds the same true for diurnal activity or rest conditions. Attention is called to the close similarity of the curves in Fig. 1, and to the close agreement of the results cited. ^ Lehrbiich der Physiologie. 1902. ^ ' Ueber die Tagesschwankungen des Stoffwechsels imd der Korpertemper- atur, Skand. Archiv. fur Physiol., 8. 1898. ^ Fuller reference to this work will be made later. DIURNAL COURSE OF EFFICIENCY Table II. Body Tkmpekature and Time of Day. 6 A.M. 7 8 9 10 11 12 1 P.M. 2 3 4 5 1, 37.9 * *38.4 38.2 h' 37.5 * *37.9 37.8 2, 36.4 36.7 * 37.0 37.2 37.1 *37.3 37.5 2, 36.6 36.8 * 37.1 37.1 37.1 *37.2 37.3 3, 33.3 32.8 32.9 32.5 * 32.5 33.6 34.2 * 35.5 34.5 33.5 33.9 33.2 3, 36.4 36.9 37.2 37.2 *37.3 37.4 37.4 * 37.5 37.4 37.4 37.4 37.4 6 P.M. 7 8 9 10 11 12 1A.M. 2 3 4 5 1, 38.4 *38.5 1, 38.0 * 38.2 2i 37.4 *37.2 36.9 2I 37.1 *37.1 36.7 3, 34.2 35.6 36.0 35.9 35.8 35.7 35.7 35.5 35.2 34.7 35.0 35.0 3^ 37.5 37.5 37.4 37.2 37.0 37.0 36.8 36.8 36.7 36.6 36.6 36.4 1 — Binet. Ii hand; L rectum; self, number of days not stated. 2 — Vierordt. 2i average of Jiirgensen, Liebermeister, Barensprung — rectum — days not stated ; 2; average of Gierse, Kallmann, Liclitenfels, Friihlich — mouth — days not stated. 3 — Roemer. 3^ hand; 3. rectum; subjects and days not stated. * — Indicates the occurrence of meals. Some very interesting experiments by Galbraith and Simpson^ on the temperature wave in the monkey may be briefly described here. The axillary temperature of six subjects was taken every two hours, day and night, under different conditions, as follows: (a) Twelve days— ordinary conditions— fed at 9 a.m. and at 6 P.M. — period of activity, 9 a.m. to 9 p.m. ; of rest, 9 p.m. to 9 a.m. Result: rise during activity, maximum at 5 p.m.; fall during rest, minimum at 6 a.m. This occurred every day. (&) Six days— conditions exactly reversed (artificial light at night) — fed at 6 a.m. and at 9 p.m. Result: complete reversal of the diurnal wave in 24 hours; maximum at 2 a.m. and minimum at 5 P.M. (c) Six days— active from 3 p.m. to 3 a.m., etc. Result: similar, though slower, modification of the wave ; maximum at 9 p.m., mini- mum at noon. (d) Six days— in darkness during entire time, etc. Result: wave became irregular, then gradually disappeared. (c) Six days— in light whole time, etc. Result: no wave shown, but an irregular curve; animals irritable. To verify these results, experiments on birds were next under- taken.- The owl reached the maximum at night and the minimum ^ ' Conditions Influencing the Diurnal Wave in Temperature of Monkey,' Jour, of Physiol., 30, 2: 20; Proc. Physiol. 80c. 1903. ^ ' Temperature Variations in Nocturnal and Other Birds,' op. cit. COURSE OF PSYCHOPHYSIOLOGICAL EFFICIENCY 9 in daytime. The thrush, on the other hand, reached these points at noon and midnight, the difference amounting to over 5 degrees centi- grade. Gulls, pigeons and three varieties of ducks showed similar variations ; the smaller the bird the greater was the diurnal variation. The recurrent rhythm of day and night and the consequent habitual alternation of activity and rest are seen to be the all- important factors here, as they are also, probably, for men. Unfor- tunately, no analogous human experiments have yet been made, 3, Deaths Relative to Time of Day Kirkpatrick,^ discussing fatigue, makes the remark (p. 327) that since more deatlis occur about 4 a.m. than any other hour, vitality is then probably lowest. He cites no authority for this statement, nor has the writer been able to locate any in print, though it appears to be widely accepted as a fact, even by physicians who have been personally questioned. To determine the facts 36,000 records of Table III. Deaths Relative to Time of Day. Description of Group. Zymotic cases All causes Males Females Age— M. & F, 1- 5 6-25 26-45 46-65 66-95 Season, M. & F. Nov. -Feb Mar.-June Jul y-Oct Description of Group. Zymotic cases.... All causes Males Females Age— M. & F. 1- 5 6-25 26-45 46-65 66-95 Season, M. & F. Nov.-Feb Mar.-June Julv-Oct Total No. 15,616 23,439 11,991 11.448 6 A.M. 7 10 11 12 1P.M. 2 660 659 579 631 628 668 665 598 695 687 697 693 10031007 953 959 970 981 782 955 980100910061054 493 525 491 506 501 501 409 484 471 505 500 562 510 482 462 453 469 480 373 471 509 504 501 492 227 211 209 171 178 195 152 176 202 203 180 203 113 137 115 130 122 125 96 117 123 105 122 119 268 272 245 261 272 264 229 268 275 296 286 273 241 229 237 248 238 242 185 239 239 253 247 276 154 158 147 149 160 155 120 155 141 152 166 183 369 359 348 360 380 330 270 324 312 371 372 361 358 381 373 339 325 372 269 336 375 357 351 378 276 267 232 260 265 279 243 295 293 281 278 315 Total No. 6 P.M. 10 11 12 1A.M. 709 623 603 634 688 723 412 669 624 651 660 663 994 948 984 947 9421107 8471026 994 981 9861029 530 493 520 492 471 540 441 514 490 532 485 535 464 455 464 455 471 567 406 512 504 449 501 494 201 181 181 200 162 219 138 200 211 190 190 212 119 94 135 127 127 135 98 122 130 107 114 130 256 275 267 263 263 281 248 285 250 301 268 282 272 245 249 208 238 300 201 286 258 245 244 253 146 153 152 149 152 172 162 133 145 137 172 152 325 324 375 336 327 434 323 410 379 348 361 383 354 351 319 333 331 377 276 318 332 332 359 369 315 273 290 278 284 296 248 298 283 301 266 277 ^Fundamentals of Child Study. 1903. 10 DIURNAL COURSE OF EFFICIENCY deaths occurring in New York City in 1901 were worked over by the writer. Those cases were excluded in which death was caused by suicide or accident, those of children under one year in age, and those where the hour was omitted from the doctor's certificate. Males and females were recorded separately and in the following age groups: 1-5, 6-25, 26^5, 46-65, 66-95. Three seasons of the year were also noted: November-February inclusive, March-June, July-October. For each of the months 3,000 cases were used, but after making the above exclusions only 23,439 remained. In addi- tion to these. Table III. reports 15,516 deaths from zymotic diseases, which had been tabulated for each hour of the day in the official records of the city Health Department for the years 1876-79. No others were found for subsequent years, up to date. In all but the first group 6 a.m. means from 5 :30 to 6 :30 ; what it means in the first group was not discovered. A.M Noon PM. M/D-NIDHT AM. iPDc ir-, 1 R a ip /r /2 I ? •J i y~fi 7 ? ? 'P < ^-^ ! 'F'I r ^—^'^oa 400 GOO 400 600 Fig. 2. ^>'' Females -Broken L/ne Figures at the sides indicate number of cases. Deaths Relative to Time of Day — see Table III. This table and these curves of the chief groups are by no means easy of interpretation as they stand. This is mainly due to one peculiarity, affecting every group, which must first of all be ac- counted for if possible. This is the notably apparent decrease in deaths at the noon and midnight periods. Its occurrence at these particular points is 'prima facie evidence that a real fact is not dis- closed, and its occurrence in all groups is convincing proof of some external cause. It can probably be explained if Kant's philosophic theorizing be allowed — that 'man is by nature lazy'. This seems a COURSE OF PSYCHOPHYSIOLOGICAL EFFICIENCY H far cry to the present matter, but a few words will make it plain. The attending physician must certify, as to the hour of a patient's death, that so-and-so died 'at — m.' This works happily enough till the periods in question are involved, when immediately arises the necessity of writing out in full the words 'noon' and 'midnight' for the sake of their clear distinction.^ Rather than take this trouble, the indifferent doctor will do one of two things — either leave the space blank or make the record for the closely preceding or suc- ceeding hours. In the first event a like loss is occasioned in the two periods, since the record does not then differ from those left blank for other reasons, all of which records had to be excluded. The cases thus affected would account for only a small part of the ap- parent decrease — possibly one tenth. In the other event, the ad- jacent periods should show decided increase. That is precisely what they do at midnight, sufficiently to smooth the curves from 10 p.m. till 2 A.M. inclusive, as indicated by the dotted lines in Fig. 2. It is not so evident at noon, and there may be a real decrease at that hour — a possibility somewhat emphasized by the 1 o'clock drop in the first group. The first two curves are similar throughout, save at noon. "When corrected as suggested they are easier to compare. The highest rate in each occurs, not at 4 a.m., but from 2 to 6 p.m. ; and the minimum, less well marked, immediately after— from 7 to 12 p.m. An exam- ination of the figures of the sex group shows much the same order, the females reaching a slightly earlier maximum than the males, while 3 A.M. and 5 p.m. are the only points of notable opposition. The female rate is more regular than the male. In the age group, 26-45 most resembles the female curve and 46-65 the male. In the season group, the July-October curve most nearly follows that for all deaths. Just what these results mean is puzzling. The differences are frequently much above those easily accounted for by chance. While the number of cases seems large, it is only about one fourth large enough. The sub-groups for age and times of year are far from ^ This would not be the case if many people, including doctors, did not think that M stands for midnight as well as for meridian, while some think it stands for nothing else. The first set of figures is seen, relative to the second, to show a much less noon and a much greater midnight drop. This is best accounted for by supposing that the compiler included at the former period all cases marked ' 12 M.,' whereas part doubtless belonged to the latter period. The same is less true of the second set because some of these cases were distributed to the later period. Attention is called to the fact that these points are disregarded in the ' smoothings ' shown in Fig. 2, but should be remembered in connection with the further discussion of the curves. 12 DIURNAL COURSE OF EFFICIENCY being extensive enough to throw adequate light on the causal factors involved. Even after combining the male and female figures, each period of these groups is poorly represented ; but such a combination is scarcely allowable. Confining attention to the first three groups, the venture of a definite opinion might be hazarded. At first thought one would be tempted to infer from the occurrence of the maximum at the warmest part of the day, that the heat was the important factor. But inspection of the seasonal figures shows a much lower death rate in July-October than in winter, while the curve for the summer months closely resembles the general curve. Heat alone offers no solution. The following theory is more in harmony with the facts : the vast majority of these people died of disease and not of old age. The proportion of sudden to gradual deaths was, on the whole, small. The 'average patient' (to generalize for adults) on awakening is fed, and for an hour his death rate is lowered, food influence being often eked out by a nap. His waking troubles begin thereafter and cause a rise in the rate till at noon he is again fed and reduced to a minimum of activity for an hour or two. Then the troubles and frettings are resumed more vigorously, the body having now become thoroughly 'warmed up', and the actions responsive to this increased state of stimulation sap the energy more rapidly, till the final sur- render takes place about 5 or 6 p.m. For the survivors, there is at this time another meal and consequent lowering of the curve, which is again emphasized by longer sleep, during which there is a gradual, slight rise in rate till the morning awakening. This idea is supported by the time of occurrence of the highest and lowest rates in the age groups. A close scrutiny, with the suggested smoothing always in mind, reveals that the maximum for ages 1-5 is very early in the morning — before 9 ; that for 6-25, from 7 to 11 A.M. ; that for 26-45, about 3 to 4 p.m. ; and for the following years, about 5 to 6 p.m. The curve of diurnal activity probably varies rather similarly with age. A large increase in the number of cases would be of marked value, as it is evident that these groups are not only too wide, but are also rendered too heterogeneous by the sex element. The minimum locations are likewise confirmatory, as they uniformly occur at the customary points of least activity — meal-times and night. At all events, the 4 a.m. idea is effectively exploded. The cases are numerous enough to establish this point reliably. The females of the second group show a drop at 3 a.m., which is offset by a male rise and hence a straight line appears in the curve of the whole group for this part of the night. As a matter of fact, the time from COURSE OF PSYCHOPHYSIOLOGICAL EFFICIENCY 13 1 to 5 A.M. is but a trifle higher in death rate than the median for the whole day. The body temperature and expiration of COo, as shown by Fig. 1, would lead one offhand to expect the greatest num- ber of deaths about 4 or 5 a.m. The fact that they do not come at this point of apparently lowest vitality, throws us back to the con- clusion already drawn. It appears paradoxical to call the point of greatest strength — a fact to be later demonstrated — the most likely death point, yet it is really not unreasonable. It is well known that patients usually grow worse towards evening. The successive mo- ments of pain or chafing throughout the day act in a cumulative way physiopsychologically, causing increased irritation, which ex- cites to the expenditure of more energy than the system can afford to lose. 4. Brief Summary It may be said that, in a general way and ignoring the immediate effects of meals as far as possible, the heart and lung functions and body temperature tend to vary alike, beginning with a very early minimum — about 5 a.m.— and increasing to a late afternoon maxi- mum — about 5 P.M. There is ground for believing that women reach an earlier maximum than men, at a higher rate, and children a still earlier maximum at a still higher rate. Similar facts seem to hold good in the case of animals, where it is found that the periods of rest and activity determine the curve entirely. While the same maximal period holds for human deaths, the minimal one is from 7 to 11 P.M., females again coming to an earlier maximum than males, and young children to a very early one. Contrary to what one would anticipate, the maximal death rate in each of these groups fails very near to the time where we are led in normal life to look for the greatest physical efficiency, as evidenced by muscular power. This is most likely due to the fact that, in the main, at these periods the most effective stimuli are in operation, externally and internally, and cause an over-expenditure of energy from which it is impossible for many patients to ralty. While a sort of physical weakness prob- ably exists very early in the morning, it is not true, as held by pop- ular opinion, that we are 'nearest death' at that particular portion of the day. II. Sensory Activities 1. Sight, Color-sense and Hearing In some psychological experiments on various peoples at the St. Louis World's Fair, in which the author assisted those in charge of this section of the exposition— Dr. R. S. Woodworth and Mr. F. 2 14 DIURNAL COURSE OF EFFICIENCY G. Bruner, of Columbia University— the sight and hearing of over 100 Filipinos were tested between 9 :30 a.m. and 12 m., or between 2 P.M. and 4:30 p.m. For acuity of vision the so-called E-test was used, in which lines of E 's of successively smaller sizes and pointing in different directions are arranged on a chart like the ordinary letter charts for optical tests, and the greatest distance, at which the subject can recognize the positions of the E's in each line, is determined. The results give the average ratio of the distances so determined to the standard distances. There were 70 cases in the morning, giving an average of 1.90, and 38 in the afternoon, aver- aging 1.94. That is, the difference is little better than one of chance. The same subjects, in matching a series of shades of colors with a series of tints, averaged respectively 11.8 and 11.5 errors, thus favoring the morning, but only to the chance degree. In acuity of hearing, as tested by apparatus devised by Mr, Bruner for the purpose, 37 men in the morning showed an average ability slightly inferior to that of 23 men in the afternoon. The numbers are too small to make this difference significant. 2. Skin-seiisitivity and Pressure-sense AVhat work there is here bearing on diurnal variation is best discussed under the headings of attention (p. 43 and Table XIV.) and fatigue (p. 79 and Tables XXX., XXXI. and XXXII.). 3. Brief Summary No diurnal changes in sensory efficiency are suggested by the scanty data at hand, save a possible low skin-sensitivity in early morning. III. Motor Activities 1. Subhuman Data 1. Plaiit Life.— The following statements are of some interest for comparison with human data. Saclis says, "If a plant, which has been exposed to the alternations of day and night, be kept in dark- ness for a considerable time, the periodicity (due to light) may con- tinue for a time, according to Pf effer, as a persistent effect. ' '^ Pierce bases the following statements on work by Sachs: "The daily periodicity of light and darkness is almost coincident with the daily periodicity in growth rates. . . . The rate of growth in length, of plants furnished with all the food they need, will reach its maxi- mum about sunrise and its minimum about sunset. "- "^ American Text-hook of Botany, p. 883. 1882. Tlunt Physiology, p. 211. 1903. COURSE OP PSYCHOPHYSIOLOGICAL EFFICIENCY 15 2. Animal Life.— Hodge and Aikins/ experimenting with one of the Protozoa, found the ciliary activity of this animal, by which food is injected and detritus ejected, to be apparently without periods of rest corresponding" to those of higher animals. This constancy seems unaffected by reproduction, barometric pressure, light, heat, or sound. In their diagram of the activities of the animal, the most pronounced vesicular and stalk contractions and reproductive phases appear to be recurrent in the early morning. These are of the same order as man's motor and reproductive activities, while the uninter- mittent character of the ciliary activity is what might be anticipated when it is remembered that this undifferentiated functioning in these unicellular animals is that which, in developed form, finds ex- pression in our unremitting circulatory and respiratory perform- ances. With advance up the animal scale the number of rhythmic activities slowly increases. 2. Human Data 1. General Cliaracteristics of the Writer's Methods and Experi- ments. — To determine with scientific exactness the diurnal course of any activity would require a minute consideration of the subject's daily physical and mental condition, the character of his daily work and habits — as to regularity and irregularity, quantity and quality — and of all external and accidental influences to which he might be subjected at any period of the day. It would be impossible to ascer- tain accurately all these matters, but practically what we want to know is what one does accomplish under just such heterogeneous circumstances— the ordinary conditions of life. Allowances were made for several serious interruptions, and where illness occurred the records were not used. In order to insure the greatest variety of material, the data were gathered not merely from strictly laboratory tests, but also from those less capable of exact measurement in standard terms; and, in addition, the questionnaire method was attempted and data from general and school work utilized. The tests were applied extensively to groups of subjects for one or two days, and intensively to indi- vidual subjects for several weeks. Some information as to the sub- jects, periods and tests used and as to results and their statistical treatment must be given. Subjects. The males were seven in number, all graduate students at Columbia University, ranging in age from 26 to 34, and will be referred to hereafter as I., II.-VII. All were conversant with and practised in laboratory work; hence each acted as both operator and "Daily Life of a Protozoan,' Am. Jour, of Psych., 6: 524-533. 1895. 16 DIURNAL COURSE OF EFFICIENCY subject in his own case, after full verbal and written instructions on every test of his series. The letter 'A' will be used to represent a female graduate student, 23 years old, engaged in intellectual work 10 to 13 hours daily; 'B' is a group of young women of Teachers College, ranging in age from 20 to 35. Both A and B were naive subjects, and their tests were given by myself. Periods. The tests w^ere to be taken daily before breakfast, lunch, dinner and bedtime. The actual times at which they were taken necessarily varied somewhat for different persons, this being one of the great practical difficulties of the problem. To get reliable results one must start with reliable subjects, but they are just the ones who can least afford to distort their affairs daily for so long a time. The periods in the tables are two hours broad to allow classification of the subjects together. Tests. Each subject went through the tests for about twelve days; the exact numbers will be indicated in the following pages, to which reference must also be made for detailed descriptions of the various tests. About six sorts of tests, motor and mental, were given to each subject, except that subject I. (the author) took a much greater number. But the results of some of the tests admit of double measurement — for speed and for accuracy — and appear so in the tables. The 'number' given in the second column of some tables has reference to the trials at each period and not to the total number in the whole series. Results and Statistical Treatment. The results are from whole days' records, as already suggested.^ All results are in terms of the average and its probable error, the formula used for calculating the latter being ± 0.6745 (ct/V'«), where o- is the mean square deviation and n the number of cases. The mathematical chances are even that the average shown will not vary from the real (theoretical) average by more than the limits of the P.E. It should perhaps be remarked that it was necessary, though less desirable, to have the higher fig- ures represent greater degrees of inefficiency instead of efficiency; the larger the numbers, therefore, the less the efficiency. Corrections were frequently made for practise effect. There would be no occa- sion for this if only averages were sought, but it is absolutely re- quired when a figure for their reliability is wanted. The method used for eliminating the practise effect was to make an empirically graduated scale of allowance for the successive days, which was based, as to the whole amount, on the absolute dift'erences between the initial and final records and which took advantage, as to graduation, ' See p. 3. COURSE OF PSYCHOPHYSIOLOGICAL EFFICIENCY 17 of the established fact that the effect of practise is at first very pro- nounced and then decreases more slowly. There are two assump- tions in the method : first, that practise affects all the periods of the day equally; second, that any possible correction thus made and applied to all the periods alike is legitimate, however low it reduces the probable errors, since these will always remain higher than if a perfectly adjusted and full correction could be made. The first assumption may or may not be true, but it seems to me that no excep- tion can be taken to the second. Any correction of a constant error is almost sure to be nearer the truth than any uncorrected figure, as it is then transferred to the rank of chance where elimination of its influence b}^ others is much more likely to occur. When a 'total' average in the tables refers to only a few individuals, it should not withdraw attention from the separate records themselves, where the main truth is then most probably located. AVhat has been said above refers also to the mental tests, which were taken by each subject at the same times as the motor ones. The mass of details to be handled makes the briefest treatment necessary at every point, with elimination of repetitions as far as possible. 2. Actual Tests and Besults— writer's and others'. a. Speed and Accuracy of Movement. — As a rather simple form of movement the subjects were required each period to strike with a pencil point, as rapidly as possible, 200 small squares (1 cm.) arranged in 10 columns of 20 each. The time of performance gives the measure of speed, while the number of squares missed roughly indicates the (in) accuracy. None of the subjects except the writer knew that the accuracy was to be considered. Table IV. shows the results and, for convenience of comparison, those of a test treated in the imme- diately following sub-section. A small correction Avas made for practise. The greatest defi- ciency appears at morning and night periods, a dift'erence that is significant,^ as seen by the P.E.'s of each average and the uniformity of the individual results. But the inaccuracy maximum and mini- mum seem just the reverse, and this may account for the preceding differences entirely or in part. The group of young women, B, in taking this test, were allowed 45 seconds to strike 200 squares or as many of them as possible. The results show somewhat the same inverse relation of speed and accu- racy and the same order of rate efficiency as the individual cases. The improvement here might, however, be attributed to practise. ^The exact significance of a difference can be found by various formulae: see Thorndike's Mental and Social Measurements, p. 145. But it can be seen well enough, often, by inspection of the P.E.'s. 18 DIURNAL COURSE OF EFFICIENCY < I H OOO CD CD l>. CO 00 1— ( CO Ph OOO o o s — 1 p^ OOO o o « rH t- ■»*< 1—1 o iO 00 l> CO-* lO «D 05 i-H I^ CD lO lO CD O CD O IC lO rH t^ C O OCD O OI- CD CO lO '» m n COURSE OF PSYCHOPHYSIOLOGICAL EFFICIENCY 19 Another sort of accuracy test in my own case gave about the same result. The so-called hold-and-let-go method was employed in a twelve-day series in which a light (4 gr.) wooden disk was snapped, on a smooth board, toward an aperture slightly larger, 60 cm. distant. This was done ten times per period each day and the inaccuracy measured by the average distance at which the disk lodged from the hole. These averages for the four periods, which have large P.E.'s, v/ere obtained. Right 9.3 8.3 7.6 8.7 Left 12.1 11.4 11.4 12.5 The middle periods of the day still remain better than the morn- ing and evening. A more complex movement was tested by having several subjects write the numeral words 'one' to 'twelve' five times each period and the time taken. Group B wrote as many words as possible in 90 seconds. The results, in the above table, rather confirm the order of deficiency in speed, as found for the simpler movements. At the Department of Psychometry and Anthropometry of the St. Louis World's Fair many people were subjected to various tests, as above mentioned, and their records were kindly put at my disposal. Only a part of them could be used for the problem of periodicity. Two movement tests were used which need some description (see right end of Table V.). The 'accuracy' test was devised by Dr. Woodworth and consisted of an equilateral triangular piece of hard wood into each apex of which was inserted a brass circular plate containing a central hole 6 mm. in diameter and 25 mm. deep. These holes were 12 cm. apart and an electrical arrangement was made whereby a bell was rung every time a metal 'poker', a trifle less in diameter than the holes, was pushed to the bottom of any one of them. The method was to hit the holes as rapidly as possible in succession, count being kept by the operator of the record made per minute. The accuracy was therefore measured in terms of the rate at which the accurate move- ment could be repeated. The apparatus for measuring the 'rate of tapping' was also of an electrical nature. A brass plate, about 10 cm. across, was con- nected with a mechanical counter^ ; the circuit was closed and the counter moved forward one point every time contact was made, by tapping, between the plate and a short copper rod, held by the sub- ject. The time for 100 taps was taken by a stop-watch. ' Kindly supplied by the C. H. Stoelting Co., Chicago, 111. 20 DIURNAL COURSE OF EFFICIENCY lo -rj^' o ifi o o o W5 lO «o ^ «d o' ^ o 05(N->*05t--ooo'tot^ooeor-ii— iTtiic >-( eO Ol -^ U5 iM vt-^C0iC«OIM00rf 'O ^ -*j 0) ■*-* C4-H o a Ki s -tj p^ (1) s a> T3 .« +3 COO •'I -^ Xi CO =1 d) C « t3 &-^ s ^ '^ l-H c3 -kJ ^ aj « ;;5 ^ >. 3 c t) n tc ^ P> r3 bf) a> S fl pQ s ^ =3 ^ rrj -p ^ o -^ ^ ^ M tt> «= COURSE OF PSYCnOPHYSIOLOGWAL EFFICIEXCY 21 A glance at the table of results reveals five groups of males and three of females, a separation made necessary by the special condi- tions obtaining in each. The Filipino 'band-men' are separated from the 'scouts' and 'constabulary' because, being musicians, they were a highly selected group, and in fact, as the results show, they were much quicker and more accurate, but much less strong, than the ordinary soldiers composing the other groups/ The first group of white males and females comprises individuals taken through a long series of tests and measurements, and their records will be found much lower than those of the succeeding groups, which were taken under highly competitive conditions and only in the activities shown. (4), of the males, and (2), of the females, as well as many of the Filipino records were taken by myself. Since the morning and afternoon divisions of these groups were composed of different persons, the practise effect, so troublesome in most of my other series, is avoided here. But offsetting that advan- tage is the great difference between individuals, which raises the variability of each group and diminishes the reliability of the aver- ages. The P.E.'s are indeed so large that it did not seem worth while to insert them in the table. While no special significance can be assigned to the difference between the morning and afternoon results of any single group, the general tendency of the whole table seems plain. The number of groups in which the afternoon records surpass the morning is greater than can reasonably be assigned to chance. This tendency to increased motor efficiency in the after- noon is slightly less marked in the female groups than in the male. As a whole, therefore, these results lean the same way as those in the more intensive series. Use of the same apparatus by myself gave the results exhibited in Table VI. There were tv^^o series of tapping tests. The first was for twelve days, in each of which the time was taken for 200 taps at ten different periods. In a later one for five days, the time of 200 taps was taken six times at each of seven periods, every 200 alter- nating with an accuracy test of the time required to ring the bell 100 times. The thing most worthy of note in this table, aside from the gen- eral agreement with the tendencies found above, is the exceptional occurrence and decisiveness of the maximum tapping rate at 9- 10 P.M.— less distinctly manifested in the case of the right hand than of the left. The maximum of accuracy clearly comes earlier in the 1 Of the soldiers, all those shorter than five feet or taller than five feet six inches are excluded, because it so happened that a larger proportion of short men were tested in the morning and of tall men in the afternoon. 22 DIURNAL COURSE OF EFFICIENCY 00 ©* CO to 050 CJ >-( Cs 1 Cl C5 c> c> C> ■=> o (2> ~; ci >-i CO (M CO g o o ^+1 00 t^ 1-1 C5 iC o So! C5 s^ 00 CD lO to o (^ C) *-i Ci cl C) C) coco o « I-- CO o o (M C<1 SI 00 t^ CO CO 00 Mf; oq '^ ^ to to 1-1 iC h CJ >N c> ^-H CJ ■^ CO (M o 2 CO 5^1 o o © 00 l^ o cc 00 t^ CO Oi ■* •* Co o 00 1—1 ^.Ph ^ 8 O 00 -«1 1-1 to to 1^ c> 00 5^ ^ CO CJ C> C5 c> CO IM o £ CO.- lO O o o e^ CO-* CO t^ Co Co Co CO "* CD CO ci ■^ c> cj ci to CO c^ <=> 2 9. S o o o CI C5 CO t^ 1— 1 1—1 QC 00 r-l ■* lO (M O -H CO (N iM o r-H CO 55 o o CO CO 11 Kt3 'bc'&b o o 3 c a; bt) es 0) > < > <5 93 ^ 5 5 Q "fen ^ ^ "? tiD jj ^ "^ '"^ 03 >". 7J QJ o >i « o I [ TJ cu S o '2 c 03 »-> -tJ ^3 a 43 cS a> o ■^ cS T tt' r») -u *'■' ' ' W c — o fM g *j £ 0) "■^ ^ 4-* g o Cj >^ n; u > ^ ^ n tJD &( 0) S 2 « o 03 — *> III tC " C S c ^ H c ci c o Ph a; 03 I-; C3 -u cS O o o C 93 — >. O -jf s (D T- •- -■ C0VR8E OF PSYCHOPHYSIOLOGICAL EFFICIENCY 23 day than the maximum of speed ; a shorter series of accuracy tests, with the same subject, tended to place the maximum earlier than 4 :30 P.M. ; there is no doubt that it falls somewhere in the middle portion of the day. It is obvious that the rating in accuracy, secured by use of this test, has some dependence on the rapidity of the movements involved. To see if any change of order would be efifected, if the factor of speed were eliminated, the following trial was made. 100 dots, 1 cm. apart, were arranged in ten (printed) rows, both vertical and horizontal. By a free-hand, easy and uniform movement, with a fountain pen, these rows were successively traced over both hori- zontally and vertically, 100 in each direction being done at a period. The results are in terms of the average number of dots missed every ten lines. Table VII. No. of; 7:30 A.M. 9:30 Rows. Misses. P. E. Mis. P;E. 100 0.85 .23 1001 2.10.47 11:30 Mis. P.E. 1:30 P.M. Mis. P.E. 0.80 .19 0.82 .19 1.11 .15 0.98 .2£\ 1.10 1.71 .34 1.92 .43 2.82 .54 1-83 .35\ 1.99 .U 3:30 5:30 Mis. P.E. Mis. P.E. 7:30 9:30 1 11:30 Mis. P.ET'Mis. P.E. Mis. P.E. 1.21 .33\1.2i .S3 l.ZS .31 2.72.56 3.03 .50 2.94.5^ 1. Horizontal rows. 2. Vertical rovvs. This test served to confirm the other to a considerable degree, doubtless due to the chief common elements— motor control and keen- ness, with quickness of sight. The main divergence is at the second period, which the results make the best of the day. It is seen that the diurnal curve is the same for both horizontal and vertical rows, and this helps to confirm periodic differences which the large P.E.'s made doubtful, though these are mainly due to uncorrected practise effect and shortness of the series. The curve of these figures, if plotted, closely resembles that of the writer's diurnal efficiency as subjectively estimated. The diurnal couree of speed of tapping seems from these results to differ markedly from the course of accuracy of movement. Both are alike in showing a Iom^ state of efficiency in the first hours after rising, but after that their curves are nearly inverse to each other. Pure accuracy appears to reach a maximum in the later morning hours, while tapping, in my own case at least, is at its best in the evening, when accuracy is about at its worst. Tapping also gives a different diurnal curve from other tests of speed, such as striking squares or writing numerals. In explanation of this difference, it is suggested that rapidity of tapping, as it requires a minimum of control but a maximum of neural excitement, may be expressive largely of 'nervousness.' If a person is most nervous in the evening — and this agrees with my own introspection — he would accordingly 24 DIURNAL COURSE OF EFFICIENCY be quickest in tapping at that time, but not most accurate in motor control. h. Normal Muscular Poiver. (a) Small and Large Muscular Groups. Since many of the conclusions as to muscular power and fatigue are based on work with small sets of muscles, it is a question of importance whether their efficiency at any time is a trustworthy symptom of the general muscular power. This will be discussed incidentally. First, the Cattell 1903 type of spring ergometer was used by subject I. to register 50 contractions of the thumb and forefinger of each hand, at every period, for 15 days ; subject A made 40 con- tractions at each period for 14 days. However, the first 4 days, in both cases, were omitted for practise. The figures in the table show the average number of kilos for 25 and 40 contractions, for the respective subjects, and the P.E.'s. Table VIII. Subject. No. of Con- tractions. 8—9 A.M. 3:30 to 4:30 P.M. 10:30 to 11:30 P.M. Eight. P.E. Left. P.E. Right. P.E. Left. P.E. Eight. P.E. Left. P.E. I. 1st 25.. 2d 25... Average... 300 300 300 143 1.0 107 0.9 125 i.O 115 0.8 92 0.9 103 0.9 149 1.0 111 0.8 130 0.9 115 0.6 96 0.7 105 0.7 142 0.9 109 1.1 125 1.0 109 0.8 88 0.9 98 0.9 A-@40.. 400 126 1.9 115 2.2 137 1.7 125 1.9 124 2.3 119 2.2 With subject I. the record was kept for each half of the 50 con- tractions. The table shows some fluctuations, but on the whole both subjects show most strength in the middle period, as will be found to be generally true for strength.^ Lombard,- working with the Mosso ergograph as modified by himself, found the maximum power of the flexor muscles of the second finger to be reached at 10 a.m. and 10 p.m. (the higher maxi- mum) daily, and the minimum at 4 a.m. and 4 p.m. (the lower mini- mum). He assigns the diurnal rhythm of atmospheric pressure as the cause of these changes in motor power. In the same article it is ^ A cruder form of test was used by I., involving the hold-and-let-go method already mentioned. A lead ball, weighing 1)0 cm., was snapped, or ' plumped ' marble fashion, as far as possible on a soft board, marked off in centimeters. It registered its own distance each time by spots, which were erased after every 10 snaps (the number taken with both hands at 7 a.m., 12 m., 5 and 10 P.M, for 12 days). The results, in average cm. distances, were as follows: for the right hand, 54, 56, 57, 54; and for the left, 50, 55, 56, 53. Only the greatest- least differences here are of consequence, though the order seems without doubt a real one with me. ^ ' Some Influences Affecting the Power of Voluntary Muscular Contrac- tions,' Jour, of Physiol., 13. 1892. COURSE OF PSYCHOPHYSIOLOGICAL EFFICIENCY 25 found that, among various influences increasing the power, food and sleep are very important. With this true his morning maximum should be much in excess of the night one, but, as a matter of fact, it is less. It is also hard to see how the slight barometric change occurring at Worcester, where the experiments were made, could be so decidedly influential. Lombard^ also found the diurnal variations in -height of knee-jerk to be directly related to the barometric pres- sure and inversely to the temperature, but mainly dependent on hunger and fatigue, which are highly depressive (p. 67). Here the order was a morning maximum and a minimum at night, the decline being irregular; in general the knee-jerk was larger after each meal (p. 68), Patrizi,- using the Mosso machine in experiments on. himself and another adult male in early morning, early afternoon, evening and midnight, discovered the 2:30 p.m. period to be the best; the evening was better than the morning, and the midnight result was about equal to that of the morning — an order but slightly different from that of my own results. He also quotes Buch as having found (with the dynamometer) a low muscular power in the morning; greater after luncheon, and greatest after dinner. Harley,^ also making use of the Mosso ergograph, in experiments on himself, found the diurnal variations under ordinary conditions to be: maximum about 3 p.m. and minimum at 9 a.m. (his experi- mental day was only from 9 a.m. to 8 p.m.), as shown by Table IX., in kilogram-meters. Table IX. Variations in Normal Muscular Po'^a'er — Ergograph (after Harley) . Hour 9 10 11 12 12345678 Work 6.2 6.4 8.7 7.0 9.5 8.2 9.9 7.8 8.9 8.7 7.2 8.8 The agreement of this result with what has preceded is obvious. Christopher'* employed the ergograph for tests 90 seconds long every hour from 8 :30 a.m. to 3 :30 p.m., one day, on 1,127 schoolboys and girls. More intensive tests w^ere made on four children of each sex. He concludes that '' Variations of Knee-jerk, etc.,' Am. Jour, of Psych., 1: 5-71. 1887. -J. Joteyko, 'Rev. Generale sur la Fatigue Musculaire,' L'Annee Psych., 5: 1-54. 1898. ^ ' Effect of Sugar and Smoking on Muscular Work,' Jour, of Physiol., 17. 1894. * ' Report on Child-study Investigation,' An. Kept, of Bd. of Ed. of Chicago. 1898-1899. 26 DIURNAL COURSE OF EFFICIENCY 1. The extremes of endurance and fatigtie are greater in the morning than in the afternoon. 2. Strength is not as great in the afternoon, but is better sus- tained than in the morning. 3. In children the intellectual capacity varies as the physical condition. Smedley^ repeated the experiments on six boys and six girls of another school and practically confirmed the other work, as reference to Fig. 3 will indicate. Hou.t 0^ tKe liau. — A../M. T,m. i 3ct T .5 it» c <» .6 220 r Solid lint = o-yerflu-^e Xt-coti of (in t>«iy'iU ej OlUoti StKool. 3relt«R t. i " « •• •• f •. .. •. •'' , These results suggest that fatigue may be a more influential factor with children than with adults and may modify the diurnal curve accordingly. This whole matter will be treated at length later. Oseretzkowsky and Kraepelin,- using Mosso's type of ergograph, found that the average height of contraction (measuring the amount of performance) was greater at 2 p.m.; but the number of contrac- tions (measuring the endurance) at 10 a.m. These experiments, being made only at the hours named, contribute only confirmatory evidence to what has already been produced. Storey^ reports work, with a modification of this type of machine, on students of Stanford University, number not given. The first ^ ' Report on Child-Study Investigation,' An. Rept. of Bd. of Ed. of Chicago. 1899-1900. ^ ' Ueber die Beeinflussung der Muskelleistung durch verschiedene Arbeits- bedingungen,' Psych. Arbeit en, 3: 643 seq. 1901. ^ Some Daily Variations in Height, Weight and Strength,' Am. Phys. Ed. Rev., 6. 1901. COURSE OF PSYCHOPHYSIOLOGICAL EFFICIENCY 27 series was from 8 a.m. to 6 p.m. and shows the maximum at 4 p.m. and minimum at noon— but slightly below the 8 a.m. figure. In a second series — details lacking — from 7 a.m. till 11 p.m., the highest point occurred at 9 p.m. and lowest at 7 a.m. Subsequently/ for himself, he found that "there is a normal decrease in ability to do muscular work between 2 and 5 p.m." (p. 193). This must be a misprint, for at another place he says "the morning max. comes at or near 10 a.m and the afternoon max. at or near 4 p.m.," and this is borne out by his tables, which are, however, not always clear. Very recently- he has made an extended study of the matter which, on the whole, confirms his earlier results and conclusions. Half the article describes the rather complex apparatus used— his own modification of the Lombard ergograph, employing weight or spring resistance, as desired. His method of exhibiting results is open to criticism, for only the roughest comparison of groups is possible, and that not of the best sort. How much was gained or lost is usually the most essential thing to know. The results show the number of times that gains or losses in power were found for the successive hours, when each was compared with the preceding. The unequal numbers of trials at the different periods show that the false assumption Avas entertained— alluded to before^ as made by many authors— that incomplete records of some of the days repre- sented give reliable data for such relational procedure as the prob- lem involves. Storey himself, a well-trained subject, comes first in the table. A language student, 21 years old, taking no regular exercise, gave the results in the second division of the table. A law student and athlete, 21 years old, is reported in the third section, while three adult mechanics gave the results of the last section. Table X. Compabison OF Hourly Ergographic Records- —After Storey. No. of Subjects. 6-7 7-8 8-9 9-10 10-11 11-12 12-1 1 1-2 2-3 3-4 4-5 5-6 6-7 7-8 8-9 9-10 1 Gaining. 6 24 16 12 4 5 11 28 14 2 3 5 3 Losing. 1 5 4 20 1 3i 1 6 16 17 1 3 5 1 Gaining. Losing. 6 2 6 1 9 ! 5 3 5 5 2 8 1 4 1 Gaining. 11 8 5; 12 10 Losing. 1 6 10 3 4 3 % of gains. 161 52 21 j 3 % of losses. 1 33' 6 15 48 ^ ' Daily Variation in the Power of Voluntary Muscu'ar Contraction,' Am. Phys. Ed. Rev., 7. 1902. ^ Studies in Voluntary Muscular Contraction, Stanford Press, pp. 60. 1904. ^ See above, p. 2. 28 DIURNAL COURSE OF EFFICIENCY AVith the Upham dynamometer Storey then tested, in all, several hundred Stanford students by one grip before and one after an hour's gymnasium practise in classes at 10-11 a.m. (1,653 records for each hand) ; 11-12 a.m. (1,772) ; 3-4 p.m. (1,366) ; 4-5 p.m. (1.990) ; 5-6 P.M. (2,726). His main object was to ascertain the influence of gymnasium work on motor power, but, considering this work a constant factor, the above periods can be compared. The results are thus stated : ' ' there is evidence of a morning period of muscular ability between 10 and 11 a.m.; a subsequent diminution between 11 and 12; a tendency to rise between 3 and 4 p.m.; about the same condition between 4 and 5 ; and finally there is evidence of a tendency to lose power between 5 and 6 p.m." (p. 49). This, he truly saj^s, verifies all the preceding work. He fails to call attention to the early morning degree of inefficiency, though its presence is indicated by the high gains at the 8-9 and 9-10 periods. These results being in such a form as to give no definite idea of the real quantitative difference of efficiency at the different hours can not be closely compared with the work of other authors, but the divergence between his results and mine lies almost wholly in the extent of the noon and night drops. As to the general form of the diurnal curve, they are alike. The present writer also tested the strength of the grip, first using Collin's oval dynamometer. This instrument was also used in a set of leg-back tests, to be treated very soon. For convenience both results are incorporated in Table XI. These figures point rather consistently one way, being similar to Storey's save in the particulars mentioned above. My own individual order, of an early minimum and late after- noon maximum, is placed beyond all doubt by three other series. The first was for 30 days in autumn, under great regularity of external and subjective conditions. Five grips at each of 7 periods were taken. The second was also for 30 days, morning and night, 2 grips at each. The third was for 12 days, 11 periods and 2 grips at each. The last two were taken with the Narragansett dynamom- eter, registering higher than the Collin 's instrument, which was used in the first series. The great amount of material secured enabled me, however, to transmute the results of the one into the unit of the others with but slight absolute error, that would not at all affect the relative standing of the periods. To eliminate practise effect, the first ten days' records of the first two series are omitted; no allowance was made in the third, which immediately succeeded the second in point of time but stands second in the table. COURSE OF PSYCHOPHYSIOLOGICAL EFFICIENCY 29 Table XII. Musculab Power — Hand Dynamometer. Subject I. No. of 7 : 00 A.M. 8 : 30 A.M. 10:00 A.M. 11:30 A.M. 1-1 : 30 P.M. 3:00 P.M. Grips. Kilos P.E. Kilos P.E. Kilos P.E. Kilos P.E. Kilos P.E. Kilos P.E. f 100 60.7 0.17 63.5 0.16 64.9 0.18 Right... j 24 40 60.8 0.27 60.2 0.23 6.30 0.26 65.2 0.23 66.2 0.22 66.1 0.32 66.2 0.28 r 100 56.6 0.18 59.3 0.18 60.7 0.18 Left 24 40 52.2 0.36 54.1 0.19 54.1 0.30 56.0 0.33 58.2 0.28 58.3 0.37 58.1 0.32 Subject I. No. of 4-4 : 30 P.M. 6:00 P.M. 7-7:30 P.M. 9:00 P.M. 10-10:30P.M. 11-12:00P.M. Grips. Kilos P.E. Kilos P.E. Kilos P.E. Kilos P.E. Kilos P.E. Kilos P.E. f 100 66.2 0.19 63.7 0.18 61.7 0.17 60.7 0.21 Right... j 24 40 67.3 0.25 69.0 0.30 67.1 0.22 66.0 0.23 64.2 0.25 63.9 0.33 r 100 61.7 0.20 59.5 0.17 58.0 0.17 56.6 0.22 Left j 24 40 59.2 0.24 60.1 0.22 58.0 0.35 56.3 0.31 54.1 0.31 57.7 0.19 Attention is here called to Table V., showing various groups of males and females tested as to forearm strength. It will be noted that the men did almost uniformly better in the afternoon than in the morning, while the females were about evenly divided. How these results are to be construed has already been mentioned and needs no fuller comment at this point. The larger muscular groups are not so readily tested. At an international meeting of the University Physical Directors at Prince- ton, in December of 1903, an attempt at agreeing on the best gym- nasium methods of testing general muscular efficiency was unsuc- cessful, and it was decided to temporarily continue the old ones with certain modifications. This suggests the difficulty of getting ade- quate tests. Floor Dip. This was meant to test the arm muscles mainly, and consisted in extending the (rigid) body parallel with the floor, on toes and hands, and then alternately raising and lowering it by the arms as rapidly and as often as possible, efficiency being measured by the number of times it was thus completely raised. The tests v/ere on myself for twelve days, but the first three w^ere omitted and a small correction made on the rest for practise. The results are decisive for the periods shown. 7 A.M., 17.2 (P.E., .15) ; 12 M., 21.3 (.34) ; 5 P.M., 22.6 (.28) ; 10 P.M., 19.3 (.15). Floor Squat: a test for leg muscles in w^hich the subject alter- nately squatted and raised himself 100 times as rapidly as possible. An extended effort to measure the deficiency present by the shorten- ing of the time the subject could hold his breath after the exercise, 3 30 DIURNAL COURSE OF EFFICIENCY P4 X eo t^ 05 (M -* .-(■«*< IM CO CO lO lO CO 1 1 (M O ^ CO Oi OS 1— I Tt< t^ lO iC to (M (M lO 05 1-H M< C<) iM ■^ (M -^ eo 1 i-H O (M iC CO o O t^ CO 1— 1 rti lO ■<^ lO lO iC lO eo 00 O i-H (N t^ t-- eo CO o (M iM OJ o 00 O r-Jo a> o 00 o eo t^ 1-J i-Io >-i t-H Oi rHrHO O rH t^ i-H r-i O OS (M t^ O" rH O 05 1— I CO r-i ffq >-i 00 o 00 1-<"C^ r-i 1> CO i-H 'St::> hH t-HK' :i n •d fl 03 o 1 5" 3 05 ^X 'X COURSE OF PSYCHOPHYSIOLOGICAL EFFICIENCY 31 or by, the increase in pulse rate, proved fruitless. Then forty-pound dumb-bells were held in the hands and quick fatigue secured by squatting: as before, in a short series. The results were too meager for quantitative expression but quite similar in tendency to those of the floor dip for arm muscles. Hand-foot Dynamometer: designed to test combined arm, back and leg strength. It is made up of the Collins oval dynamometer, having a kilo scale for use when the instrument is pulled endwise and equipped with the necessary handle and foot attachments for pulling. A short length to the whole, 14 inches, insured full use of the leg muscles as the subject sat on the floor and pulled. The operations were : right hand and foot, left hand and foot, and finally both hands and feet, three times each period. The results are shown in Table XI., while Table V. gives some results for two Filipino groups, with whom Tiemann 's pulling apparatus was used, also regis- tering in kilos. There is no radical discrepancy between these and preceding results. Though the P.E. is sometimes large, it is clear from the whole course of the figures that the middle of the day, in- cluding the afternoon, is a time of greater muscular strength than either morning or evening. The curve of strength efficiency seems, therefore, well established for the following course, and this probability will be strongly rein- forced below by further data : a beginning minimum in early morn- ing, a fairly rapid rise till 11, a level or slight decline till 1 p.m. (±: 1 hour), an increase to the maximum at 5 (ih 1 hour), thence a fall till bedtime. Explanation of this is attempted in Part D. (6) Comparative Fmictioning of Muscle Groups. As to whether the work of very small groups can be taken as indicative of the gen- eral muscular condition, an assumption long and often made with little attempt at proof, it appears from the above tables that the assumption is in reality well founded. In order to avert confusion, attention was not called to this point in passing, nor to the one suc- ceeding, and their verification must, therefore, be retrospective on the part of the reader. (c) Bilateral Symmetry. Regarding this matter, a review of Tables V., VI., VIII., XI. and XII. will disclose the fact that, although fluctuations occur, yet as a whole the figures show the same general course of efficiency for both sides of the body. The other evidence obtained agrees on this point. Lombard^ with the ergo- graph noted that the strength may be greater on one side when least on the other, but that 'the major variations occur simultane- '^^ Fatiguing Voluntary Work,' Jour, of Physiol., 14. 1893. 32 DIURNAL COURSE OF EFFICIENCY ously on the two sides' (p. 114), and Storey, in a work already cited, with many subjects found the same. c. General Motor Control. Experiments with the ataxiagraph to determine relative steadiness in standing for different times of day, and with the automatograph to determine changes in involun- tary movement, were planned but could not be carried out for lack of time, though the author is convinced by rough, tentative tests in balancing, etc., that the field is fruitful for this problem and would throw light on the other results. d. Complex Motor Activity. Of course there is no jump be- tween the following and Avhat has preceded. Attempt was made to get data on activities of a more comprehensive character, first by means of a questionnaire designed particularly for physical directors and athletes, and containing questions as to what their experience would lead them to say on the matter of recurrent periods of varying ability, and their objective grounds for the opinion. The few replies that have been received are suggestive. Mr. J. G. Lathrop, Physical Director^ of Schools at Southboro, Mass., says: "I have never considered it possible for me or any one under me to do as good performances in the morning as in the afternoon and that, so far as I know, is the experience of others. This, I consider, applies to any form of athletic work." Dr. G. L. Meylan, Medical Director of Columbia Gymnasium, after long experience as physical director, corroborates these words of Mr. Lathrop. Mr. Davis, his assistant, holds the same decided opinion, and both describe various performances in substantiation thereof. Miss Louisa Smith, Director of the Bryn Mawr Gymnasium, re- plied thus : ' ' We have athletics only in the afternoon and no compari- sons can be made. But I have noticed this : when we used to have both afternoon and evening classes, the students of the afternoon classes did the better work in educational gymnastics. In our physical examinations, too, I have noticed that those strength tests that are taken in the early part of the day average better 'than those taken between 9 and 10 in the evening." Here we find the same as set forth above for females. Mr. C. H. Robinson, Harvard athlete, 21 years old, submits this : "Last winter [1903] I competed in the shot-put at the Boston Athletic Association games and the best distance I could get was 40 feet. These games came Saturday night. On the Monday fol- lowing I returned to my practise in the afternoon and easily reached ^ Now track-team coach at Harvard University. COURSE OF PSYCHOPHYSIOLOGICAL EFFICIENCY 33 44 feet. I have found this same fact true when I worked in the morning." He thinks the time of the maximum is determined by that of practise — which is only partly true. Mr. T. G. Meier, another Harvard athlete, 23 years old, says that he "has never noticed any particular difference between morn- ing, afternoon and night scores"; but mentally the morning seems best, with periods of lesser alertness after lunch and dinner, and then an efficient one from 8 p.m. to 12 m. This case must be noted as an apparent exception. Under the present heading will also be discussed certain observa- tions of manual labor in factories, made by the author at various times through several months. The main results are given in Table XIII., which reveals also the chief conditions. But some words must be added in fuller description of each group. The first group^ is the most noteworthy, by reason of the regu- larity of the work, the basis of pay, and the number of hours per day. Its members were all experienced stitchers except the last, in whose case a correction was necessary for practise effect. They worked at top speed, presumably, as long as the material was supplied by the 'gathering machines', which was about 45 or 50 minutes out of each hour. That is, all were obliged to stop these few minutes while the supply 'piled up' for the succeeding hour's run. At 12, 5 and 9 p.m. they worked only half an hour. Other than this the individual stops were insignificant. One was sick four days, another transferred to other work for a day, etc., so that although observations were con- tinued eleven days no one is reported for the full time. The first day's work was disregarded entirely, as it was found to be much affected by the initial experience of being observed. The same was true with the second group. Ordinarily, 120 magazines at a time were given to each worker, while, on my part, the time required to do this number, or any multiple of it, was taken just as often as it could be caught exactly, and subsequently reduced to the form shown in the table. In the second group the first two girls were experienced and piece-workers; the rest were inexperienced and worked partly by piece and partly by time, hence it was useless to attach P.E.'s to their averages. When paid by the thousand, they did from 30 to 60 per cent, more than when paid by the day. Through the kind- ness of the foreman* the work was given to them in even hundreds; otherwise it would have been impossible to get correct records of ^ The work was the binding of a 900,000 edition of Everybody's Magazine by the Trow Directory Co., New York City. * These observations were made at Dennison and Sons, New York City. 34 DIURNAL COURSE OF EFFICIENCY more than two girls at a time. In fact, several days were devoted to the first two before the above arrangement was instituted, and tab was kept of the individual holders made. After that the time per 100, for each of the eight, was taken as frequently as it could be caught, together with the corresponding time of day. More labor than for the first group was required to reduce these data to a uni- form basis. The ten-minute basis was chosen to agree with the first group, where it had been selected to avoid the use of four-place numbers for the hour. During most of the time the work was dis- tributed to the workers by the author, which insured better control of the conditions. The third group covers cases where reliable records for only one or two days were secured.' In the pursuit of this phase of the general problem much valuable time was wasted in seeking oppor- tunity for suitable observational work,'' but the trouble did not end there. Some people naturally object to being watched so closely and especially w^hen on time-work. This would not be in evidence till after the first day. In several instances records were thrown away for wilful 'soldiering'. To make these results from various sorts of work comparable it Avas necessary to treat them as indicated in the table. That is, -f- 11 for the first subject means that she was, between 8 and 9 a.m., 11 per cent, above her average efficiency for the whole time observed. Table XIII. Female Industrial Labor. I. Magazine Wire-stitching. — Average Number Done per Each 10 Minutes , of the Hour. — Piece-rate Basis of Pay. 8-9 A.M. 9-10 10 -11 11 -12 12-1 (12:30) Subj. Age. Days. Nights. No. P. E. No. P. E. No. P. E. No. P. E. No. P. E. 1. 27 9 4 148 1.38 156 1.09 158 0.47 155 1.39 157 1.06 2. 24 9 3 144 1.42 156 1.54 155 2.71 155 2.32 152 1.79 3. 20 10 5 146 1.81 153 1.43 153 0.88 150 1.26 158 1.23 4. 32 10 3 138 1.03 149 0.79 150 1.04 156 1.11 156 1.51 5. 20 9 3 128 2.62 140 2.33 142 2.02 141 2.32 149 1.43 6. 26 10 4 109 0.85 110 0.71 108 1.14 109 0.82 110 1.05 7. 23 5 3 110 1.24 115 1.59 119 2.48 111 2.28 118 2.34 8. 19 7 3 111 1.67 114 1.94 118 2.02 117 2.41 115 2.79 Av. 24 8.6 3.5 129 1.59 137 1.43 138 1.59 137 1.74 139 1.65 1 For these records I am mainly indebted to J. English and Sons and Dennison and Sons, New York City. 2 What obstacles one encounters on this score is suggested in the following: 5 each of apparently the best representatives of 20 lines of manufacture were selected and a carefully written letter sent to each, stating aims and needs, together with stamped addressed return envelope. Of these 100 letters, 6 were returned unopened, 7 unfavorably answered, 9 favorably answered, and 78 ignored entirely. Of the 9, 5 factories were distant from New York, 3 were unfit for the work, and 2 were used. COURSE OF PSYCHOPHYSIOLOGICAL EFFICIENCY 35 1-2 P.M. 1 2 -3 3-4 1-5 5-6 (5:30) Subj. Age. Days. Nights. No. P. E. No. P. E. No. P. E. No. P. E. No. P. E. 1. 27 9 4 156 0.90 158 1.42 152 0.81 153 1.64 154 1.71 2. 24 9 3 157 2.62 162 1.23 158 2.90 162 2.91 154 1.79 3. 20 10 5 156 1.02 \ 156 1.53 151^ 1.56 153 1.51 156 1.11 4. 32 10 3 155 1.19 151 1.01 147 1.07 152 1.00 151 0.97 5. 20 9 3 143 2.40 144 2.21 143 1.87 143 2.43 147 2.60 6. 26 10 4 113 0.97 113 0.85 113 0.62 108 0.78 106 1.07 7. 23 5 3 126 1.24 125 2.27 121 2.49 114 3.23 112 3.42 8. 19 7 3 111 1.57 112 1.66 112 1.82 112 2.28 105 1.77 Av. 24 8.6 3.5 140 1.49 140 1.52 137 1.64 137 1.97 136 1.80 Age. 27 24 20 32 20 26 23 19 24 9 9 10 10 9 10 5 7 8.6 Nights, 4 3 5 3 3 4 3 3 3.5 6-7 No. P. E. 157 163 156 160 148 117 121 110 1.12 2.86 1.58 3.82 3.48 2.06 2.51 1.78 141 2.40 No. P. E. 158 160 156 155 148 116 115 118 1.01 0.62 1.72 2.41 2.29 2.00 2.39 1.39 141 1.73 8-9 No. P. E. 163 158 157 152 142 116 115 114 0.93 0.00 1.31 1.12 2.61 1.60 2.35 2.83 140 1.60 9-10 (9:30) No. P. E. 164 159 157 162 140 121 111 113 1.28 62 1.63 3.75 3.56 1.92 2.43 2.71 141 2.24 II. Making Paper Coin-cases. — Average Number Done per Each 10 Minutes of the Hour. Subj. Age. Days. Pay 8-9 A.M. 9-10 10-11 11 -12 12-1 (12:30) No. P. E. No. P. E. No. P. E. No. P. E. No. P. E. 1. 18 10 Piece 94 1.66 97 1.21 94 1.29 92 1.42 89 1.90 2. 16 9 74 2.10 79 1.14 76 0.81 72 0.56 71 1.99 3. 16 6 47 48 46 46 47 4. 17 5 ns CQ 61 61 62 58 61 5. 16 4 iV. 49 50 48 43 39 6. 16 3 -i a> O « •^ Oh w '-' tf M « 2 ZO 5.8 1 8.0 1 10.0 10.8 1 7.3 1 9.8 — 1 3.5 1 2.7 Griesbach, therefore, demands (1) no school work in the after- noon, (2) later beginning in morning — on account of residual fatigue from the previous day, (3) abolishment of examinations, (4) less home work and less learning by heart. Special attention is called to the high afternoon sensitivity compared with that of the early morning — true both of the work-day and the rest-day following. His method has been much tried and criticized, but only a few of the leading articles can be referred to. MacDougall,^ after adverse criticism of other methods, says of Griesbach 's that it is 'decidedly satisfactory'; that by it can be found the best hours and days for study and the length of study periods proper for different ages. He produces no experimental proof of his own. Leuba^ attacks the method and results. He tested three adult male students for fourteen days, at 9 a.m., 11 a.m., 1 p.m., 5 p.m. and 9 P.M., on cheek and thumb, 6,000 judgments in all. He found ^ Am. Jour, of Physiol, 19. 1898. ^'Validity of the Griesbach Method of Determining Fatigue,' Psych. Rev., 6. 1899. CAUSAL FACTORS I\ DIURNAL EFFICIENCY CURVE 81 divergent results as betAveen cheek and thumb and as between one and two point determinations on the thumb alone, and his results were also diametrically opposed to those of Griesbach, Vannod and Wagner. They showed, on the whole, a marked increase in sensi- tivity up to 5 P.M. and thence a steady decrease. The results for this group and for the one described below are graphically repre- sented in his publication, but have been transmuted into figures by the present writer, as indicated in Table XXXI. Work and rest days are given separately. The second series was on six adult female students, 2,000 judgments, at 8 :15 a.m., 10 :15 a.m., 11 :15 p.m. and 1 :15 p.m. It gave similar results, diverse for forehead and cheek as to the single individuals, but showing considerable con- stancy in the total curves. The females also start with a relatively low sensitivity and increase as steadily as the males. Table XXXI. ESTHESIOilETER TeSTS : THRESHOLD IN MM. — ReTISION AFTEB Leuba. Male 9:00 A.M. 11:00 A.M. 1:00 P.M. 5:00 P.M. 9:00 P.M. Subjects. Work. Rest. W. R. W. R. W. R. W. R. 1 14.5 14.0 15.1 13.3 14.5 13.6 14.0 14.1 14.4 14.6 2 15.8 15.4 15.2 15.5 15.7 16.2 14.9 16.1 14.8 16.6 3 8.0 8.4 7.5 7.2 7.4 7.1 7.1 6.7 8.1 7.6 Average. 12.8 12.6 12.6 12.0 12.5 12.3 12.0 12.3 12.4 12.9 Female 8 : 15 A.M. 10:15 A.M. 11:15 A.M. 1:15 P.M. Subjects. Forehead Cheek. F. C. F. c. F. C. 1 16 22.0 15 22.0 14 20.5 14 20.5 2 11 13.0 9 11.0 11 15.0 10 12.0 3 7 11.0 6 11.5 4 8.0 3 8.8 4 12 21.9 17 21.0 14 22.0 13 22.1 o 8 10.0 6 8.8 7 8.0 8 8.0 6 8 11.0 5 11.0 5 11.1 5 9.0 Average. 10.5 14.8 9.7 14.2 9.2 14.1 8.8 13.4 Germann^ used both one and two points, noting the number of errors at the threshold. The experiments were made on an adult female student studying eight or nine hours daily; she was tested at about 8 :30 a.m. and 9 :30 p.m. ; 2,450 judgments were made. Table XXXI. gives the results of fourteen days, in each of which a pair of tests were taken. Table XXXII. Esthesiometeb Records — After Germann. Errors. A.M. = P.M. A.M. > P.M. A.M. < P.M. Total. 2-point. 1-point. 4 1 2 8 9 9 2 4 3 ^ ' On the Invalidity of the Esthesiometric Method as a Measurement of Mental Fatigue,' Psych. Rev., 6. 1899. 82 DIURNAL COURSE OF EFFICIENCY This means that the percentage of erroneous judgments in the morning was equal to that in the evening on four days ; greater on eight days ; less on two days ; etc. Plenee he concludes that "in at least one normal case the percentage of errors in cutaneous tactile discrimination bears no constant nor even relative correspondence to the mental fatigue experienced by the subject." This conclusion is hasty, as it possibly confuses fatigue and feelings of fatigue and fails to take account of the state of inertia which has been emphasized as occurring early in the morning. It is suggested, therefore, that the morning insensitivity observed by these various experimenters is not at all a symptom of technical, nor of 'residual,' fatigue, nor yet of subjective weariness, but rather of inert physical condition unconnected directly with any one of them. It is also a mistake to assume that insensitivity can arise only as a result of fatigue, and just as much a mistake to suppose that much fatigue is necessarily present in adults in the afternoon or after several hours' effort. INIany other authors also have demon- strated the unsafeness of this method for measuring mental fatigue.^ 3. 3Ieasured by Mental Inefficiency .— A closer approach to a purely mental test of fatigue is found in the reaction-time or asso- ciation-time method first used by Cattell." In a course of experi- ments Avitli different sorts of stimuli, reactions (1,950 in all) were continued for a whole day, from 7 :30 a.m. to 8 :30 p.m., with short intervals for eating. The results are thus summed up : " The first result to be noted is the very slight effects of fatigue ; in no case is the time lengthened more than a couple of hundredths of a second and the mean variation is but little increased. "We reach the un- expected conclusion that the processes which are most automatic — naming colors and simple reaction-time to sound— are the most affected by fatigue." This method was also employed by Roemer and by Ellis and Shipe.' The latter determined the reading time for words of four letters, in morning and at noon, and state their results in the words quoted on page 79, above. In regard to a second series of various tests, they say, "Of 24 perfect records reaction-time was shorter 14 times at 5:30 p.m.; mean variation less 13 times; total figures added, more 17 times; more added correctly, 15 times; more cubes written, 19 times; more nonsense syllables learned, 15 times. One of the best records came just after a long, hard examination (in the » As Cattell, Ebbinghaus, Kraepelin ('99), Bolton, Meiimann, Kraepelin ('03). » ' The Time taken up by Cerebral Operations,' Mind, 12: 44. 1886. ^ Op. cit. CAUSAL FACTORS IN DIURNAL EFFICIENCY CURVE 83 afternoon) for which the student, after working and worrying all the day before had continued work till 2 a.m. the night before." Later, seven subjects were tested (1) during a w'eek before March examinations, (2) during the examination Aveek and (3) during the one following. The results practically agreed with those of previous series, and the authors report 'the same lack of agreement between tests and the same failure to indicate fatigue in the afternoon.' If they 'failed to indicate fatigue' they must have agreed somewhat, it w^ould seem. Then three subjects were used (four, three and two days, re- spectively) at 9-10 A.M. and 5-6 p.m., and results were obtained 'equally inconsistent with each other and with the undoubted facts.' These last citations show the assumption in the writers' minds that fatigue is the normal afternoon condition in the case of all adult students. It appears from their reports that the afternoon reactions and results in general were superior to those of the morning, even though the morning tests were taken rather late. Their series of experi- ments were short but, as far as they go, confirm the views expressed in the previous portions of this paper. One of the earliest tests to be used was that of 'dictation exer- cises,' introduced by Sikorsky in 1879— the first attempt at objective measurement of mental fatigue. It was later used by Oehrn ('89), Hopfner ( '94), Friedrich ( '96), and preferred to all others by Henri and Binet ('98). According to its advocates, it is easily given to groups, easily cheeked up, easy to understand, has no practise effect, and the different characters of the errors show the different types of fatigue (as for arithmetic, gymnastics, etc.). Only those of Sikorsky and Friedrich off'er direct evidence on the problem in hand. Sikorsky, according to Binet, ^ gave 500 dictations to six classes, the members ranging in age from nine to sixteen years, at 9 a.m. and 3 P.M., before and after the school session. The number of errors made by each class is given in the following table.^ Binet classified these into phonetic (omissions and substitutions of letters), graphic (in written form, etc.), psychological (omissions or substitutions of words), indeterminate (all others). These are also shown in the table in per cents.^ ^La Fatigue Intellectuelle, 288. 1898. ^ The figures are evidently percentages of something : Binet does not say what. ^Neither of the revised columns totals 100 per cent.: Binet does not say why. 84 DIURNAL COURSE OF EFFICIENCY Table XXXIII. Fatigue — ^Measured by Dictation Exercises. Binet, after Sikorsky. Binet, revision after Sikorsky. No; of Class. 9 A.M. 3 P.M. Per Cent. Diflferenee. Kind of Error. 9 A.M. 3 P.M. Class 1 " 2 " 3 " 4 " 5 " 6 123.5 121.5 72.4 66.5 61.4 45.7 156.7 145.3 102.8 94.2 81.0 80.0 + 33.2 + 23.8 + 30.4 + 27.7 + 19.6 + 34.3 Phonetic. Graphic. Psychic. Indeterminate. 62.6 % 8.9 % 4.5 % 6.0 % 77.3 % 11.3% 8.9 % 11.9 % This table interprets itself. The differences in efficiency are seen to be very large, uniformly so, and can leave no doubt as to the cor- rect inference. But still greater differences were found by Friedrich with a class of 51, averaging ten years of age. Table XXXIV. Fatigue — Per Errors in Dictations — Binet, after Friedrich. t." . c . tTto u , 3 u tT . u . 4) s a s a s a 3 a s a O W-3 3 a 3 i^' S K-l o_o o o tS-2 o o WT3 w-S W o to ^ ^2 n V _ =3 _ « 'O QJ 1^ "S °* ti 0/ t3 V t2 1^ CO S II II Go m <* ^^ -^o <« <