BOUGHT WITH THE INCOME 
 FROM THE 
 
 SAGE ENDOWMENT FUND 
 
 THE GIFT OF 
 
 Sietirg W. Base 
 
 1891 
 
 ^...^.:^.:^.o..7..3: ' ^/e:;//..()... 
 
 6896-1 
 
Cornell University Library 
 QP 431.G81 
 
 Physiology of the special senses / 
 
 3 1924 024 828 810 
 
The original of this book is in 
 the Cornell University Library. 
 
 There are no known copyright restrictions in 
 the United States on the use of the text. 
 
 http://www.archive.org/cletails/cu31924024828810 
 
PHYSIOLOGY OF THE SPECIAL 
 
 SENSES 
 
PHYSIOLOGY OF THE 
 SPECIAL SENSES 
 
 BY 
 
 M. GREENWOOD, Junr. 
 
 M.R.C.S., L.R.C.P., F.S.S. 
 
 STATISTICIAN TO THE LISTEE INSTITUTE OF PREVENTIVE MEDICINE ; DIRECTOE 
 
 OP THE LONDON HOSPITAL STATISTICAL DEPAHTMENT ; LATE SENIOR 
 
 DEMONSTRATOR OF PHYSIOLOGY IN THE LONDON HOSPITAL 
 
 MEDICAL COLLEGE ; AND EXAMINER IN PHYSIOLOGY 
 
 TO THE UNIVERSITY OF ST. ANDREWS 
 
 LONDON 
 EDWARD ARNOLD 
 
 1910 
 
 [All rights reserved] 
 
' PEEFACB 
 
 I HOPE that this volume may be of service to two classes of 
 readers. Students of Psychology may desire to obtain more 
 information regarding the physiological side of the senses 
 than is usually found in works professedly dealing with 
 psychology. I think such students will find this book of 
 use if they read it in conjunction with Professor Myers' 
 admirable treatise on Experimental Psychology.^ 
 
 Another class of readers which I have had in mind are 
 those who are either taking up physiology as a branch of 
 liberal education, or with a view of presenting themselves 
 for certain higher professional examinations. In either 
 case a somewhat more detailed knowledge of physiology 
 is required than can be obtained from the general text- 
 books of the subject, while time hardly permits of much 
 use being made of original sources of information. The 
 necessity in such cases arises for books intermediate between 
 the text-book and the original memoir. In most branches 
 of physiology this want is abundantly supplied, but I am 
 not acquainted with any such aids to study in the case 
 of the senses. This book may help to fill the gap. 
 
 The object I have had in view has compelled me to 
 
 1 A Text-hooJc of Experimental Psychology, by Dr. C- S. Myers, 8s. 6d, net 
 (London : Edward Arnold), 
 
vi PREFACE 
 
 restrict the work within definite limits. I have abstained 
 from describing the anatomy and histology of the sense 
 organs, since information on these matters is to be found 
 in any general text-book, and I assume the reader to be 
 acquainted with the rudiments of physiological optics and 
 acoustics. 
 
 Bibliographical references have been confined for the 
 most part to easily accessible works. If the reader is 
 tempted to follow up the clues given, the ends of this 
 book will have been completely attained. 
 
 M. GREENWOOD, Junk. 
 LoiTGHTON, January 1910. 
 
CONTENTS 
 
 CHAP. PA8E 
 
 I. Introduction — The Laws of Muller, Weber, and 
 
 Techner 1 
 
 II. General Physiology of Cutaneous Sensation . 8 
 
 III. Pain 18 
 
 IV. Protopathic and Epicritic Sensibility . . 24 
 V. Taste and Smell 35 
 
 VI. The Sense of Position and Movement. . 48 
 
 VII. The Sense of Position and Movement {Concluded) 58 
 
 VIII. Hearing — Historical Sketch . . .64 
 
 IX. Physiology of the Bar ... 70 
 
 X. The Comparative Physiology of Vision . 86 
 
 XI. Retinal Processes, Electrical, Phototropic, and 
 
 Chemical Responses .... .96 
 
 XII. Visual Adaptation — Peripheral Vision, Total 
 
 Colour-blindness ... . . 101 
 
 XIII. Recurrent Vision-theories of Adaptation 112 
 
 XIV. Trichromatic Vision . . . . . .124 
 
 XV. Dichromatic Vision . . .139 
 
 XVI. After-images oe Successive Induction . . 155 
 
 XVII. Historical Theories of Vision . . 166 
 
 XVIII. The Young- Helmholtz Theory of Colour Vision . 177 
 
 XIX. Hering's Theory of Visual Sensations . .189 
 
 XX. Simultaneous Contrast ... . 203 
 
 XXI. The Physiology of "Space" 215 
 
 INDEX 237 
 
 vii 
 
PHYSIOLOGY OF THE SPECIAL SENSES 
 
 CHAPTEE I 
 
 INTRODTJCTION— THE "LAWS" OF MULLER, 
 WEBEB, AND FECHNER 
 
 That province of Physiology to which has been assigned 
 the investigation of our special sense mechanisms is of wide 
 extent. We have to deal with the representation in con- 
 sciousness of effects produced upon the bodily structures 
 by different physical agents under conditions of the most 
 varied nature. In some cases, the results are manifest; in 
 others, their existence may be inferred with more or less 
 plausibility; in yet others, although the conscious reaction, 
 the sensation, is distinct enough, the physiological change 
 which is associated with it eludes our imperfect means of 
 investigation. 
 
 From very early times sense physiology and psychology 
 have proved themselves of absorbing interest alike to philo- 
 sophers and men of science, and a history of the subject if 
 adequately treated would make an interesting record of the 
 progress of scientific thought. In this book I can do no more 
 than indicate very imperfectly how the workers of long ago. 
 and of to-day have endeavoured to mould their theories and 
 observations into that organised body of knowledge which 
 is a science. 
 
 The main problem, then, is this : Given a series of physical 
 processes of assumed constancy, and a number of conscious 
 states also assumed to have a real existence — from the stand- 
 point of science as distinct from metaphysics — we have to 
 trace out the intervening physiological processes. 
 
 It is scarcely necessary to remark that Sense Physiology 
 has nothing to say regarding a causal link between physio- 
 
2 PHYSIOLOGY OF THE SPECIAL SENSES 
 
 logical changes and sensations. We are merely concerned 
 with a time sequence, and make no attempt to bridge the 
 gulf between the worlds of physiology and psychology. 
 
 The scope of our inquiry being so wide, it is natural to 
 ask, first, whether there are any general formulae, or laws, 
 describing some relations between external stimuli and con- 
 scious reactions valid for all the sense organs. Three such 
 formulae have been enunciated, the " laws " of Miiller, Weber, 
 and Fechner, and we shall begin our task with an investi- 
 gation of them. 
 
 The first in point, both of time and importance, is un- 
 doubtedly Miiller's " law," which may be enunciated in the 
 following terms : — 
 
 1. Different stimuli acting upon the same sense mecha- 
 nism are followed by the same kind of sensation. 
 
 2. The same stimulus acting upon different sense mecha- 
 nisms calls up diff'erent sensations. 
 
 This formula is often called the " Law of Specific Sense 
 Energy," but the terminology is bad, since the word energy 
 has acquired a meaning differing much from that in vogue 
 when the law was first published. 
 
 In discussing this law, two phrases are employed which 
 need definition. An "Adequate Stimulus" is that form 
 which usually acts upon the sense organ under considera- 
 tion ; for instance, light is an adequate stimulus to the eye. 
 An "Inadequate Stimulus" is one differing in kind from 
 those generally effective ; thus a blow on the eyeball is an 
 " inadequate '' stimvilus for the organ of vision. 
 
 If the use of common words in uncommon senses be bad, 
 then this terminology cannot be defended ; be this as it may, 
 in employing it one conforms to the established custom. 
 
 It is clear that the, anatomically, more complex sense 
 organs are shielded from the possible action of inadequate 
 stimuli; for example, the organ of hearing is not readily 
 exposed to the incidence of light waves, and the situation 
 of the eye guards against its mechanical stimulation. It is 
 not, therefore, easy to test Miiller's " law " in these instances, 
 although many attempts have been made. 
 
 Nagel remarks, with justice, that the supposed proofs in 
 
INTRODUCTION 3 
 
 the case of sight are not convincing ; the flash of light seen 
 by an unanaesthetised patient on section of the optic nerve 
 might well be due to mechanical stimulation of the retina, 
 and a similar objection can be urged against other like 
 experiments. The most elegant illustration of the law 
 is afforded by the results of stimulating the central end 
 of the chorda tympani, chemical, mechanical and thermal 
 excitation alike calling up a sensation of taste. The reader 
 will of course notice that this is not a proof, but an illustra- 
 tion ; inadequate stimulation of the supposed end organs of 
 taste themselves has not led to very definite results. 
 
 Although we have no rigorous proof of the correctness 
 of MtlUer's " law," the formula is of real scientific value. At. 
 we shall see, many early thinkers were hampered in their 
 speculations by an axiom dating back to remote ages which 
 asserts that '' causes " and " effects " are necessarily similar 
 in a narrow sense. Mtiller was, perhaps, the first physio- 
 logist definitely to abandon this ancient and sterile dogma, 
 his " law " defining accurately the modern point of view 
 in discussion of this type.^ 
 
 The next formula, Weber's " law," carries into the province 
 of experimental psychology, and another definition of terms 
 is required. If corresponding to a stimulus of given intensity 
 A we experience a sensation of " magnitude " B {vide infra), 
 then if the stimulus be increased or diminished very slightly 
 no difference in the sensation will be experienced. When, 
 however, the increase or decrease exceeds a certain limit, the 
 intensity of the sensation changes. Again, if a stimulus be 
 increased in intensity from nothing, only after a certain 
 point has been reached will any sensation be experienced. 
 The absolute magnitude of the stimulus in the second case, 
 and the difference of the magnitudes in the first case, measure 
 the threshold or liminal values, and may be alluded to as 
 threshold or liminal stimuli. 
 
 For information as to the manner in which such experi- 
 ments are carried out, together with details as to possible 
 sources of error, the reader must consult special works ; space 
 
 1 See Helmholtz, Ilandb. d. Pkys, Optik. 2(c. Aujl., p.-. 248, etc. 
 
4 PHYSIOLOGY OF THE SPECIAL SENSES 
 
 only permits of a short note as to the leading principles 
 involved. 
 
 Most methods of experiment fall under one of the three 
 following groups : — 
 
 (1) The Method of Smallest Perceptible Differences. 
 This simply consists of passing from one intensity of 
 
 stimulus, by small increments, to the just noticeable next 
 " higher " sensation. Most of Weber's own work was carried 
 out in this way. 
 
 (2) The Method of Right and Wrong Answers. 
 
 Two stimuli, A and B, are chosen of nearly equal magni- 
 tudes, e.g. a pair of weights. The subject is then asked 
 which is heavier, his answer recorded, and the experiment 
 repeated many times. 
 
 If the number of correct answers be only about equal to 
 the number of incorrect ones, it is concluded that the dif- 
 ference between the two stimuli is inappreciable. The 
 experiments are then repeated with weights differing by a 
 greater amount, values being finally reached for which the 
 number of correct judgments definitely exceeds — when 
 tested by an adequate statistical process — the number of 
 incorrect ones; it is inferred that at this stage a definite 
 subjective appreciation of the difference has been attained. 
 
 (3) The Method of the Mean Error. 
 
 Using the former illustration, a subject is asked to choose 
 from a large number of weights all those exactly equal to 
 a standard. All the weights he chooses will not be exactly 
 equal; the differences from the standard are recorded, ^ and 
 their sum, divided by the total number of trials, is thought to 
 give the average error the subject may be expected to make. 
 This magnitude will be a lower limit for the threshold value. 
 
 It may be said at onpe that results obtained by the 
 employment of any of these methods require the use of 
 most refined statistical reasoning before any importance 
 whatever can be attached to them, and such an analysis has 
 not always been forthcoming. 
 
 As a result of investigations of this kind, carried out by 
 Weber, Fechner, and others, the following conclusions were 
 ^ Without regard to sign. 
 
INTRODUCTION 5 
 
 reached: — The just noticeable increase of a stimukis bears a 
 constant ratio to the original stimulus, or, two stimuli in 
 order to be discriminated must be in a constant ratio, the 
 latter being independent of the absolute magnitudes of the 
 stimuli. The actual value of this ratio, although constant 
 for any one sense mechanism, varies from organ to organ. 
 This is Weber's well-known " law." 
 
 Accepting Weber's "law" as true, an attempt Avas made 
 by Fechner to express sensations in terms of quantitative 
 units. Having adopted certain assumptions of which the most 
 important perhaps is, that all just noticeable differences of sen- 
 sation contain an equal number of sensation-units, it was easy 
 to deduce from Weber's " law''^ that the sensation varies as the 
 logarithm of the stimulus. This corollary is Fechner 's " law." 
 
 The validity of Weber's "law" and Fechner's corollary 
 have been hotly contested. The problem is complex, and 
 little is gained by either confident abuse ^ or praise. The 
 following remarks, which agree essentially with v. Kries' 
 opinions,^ indicate some of the more important difficulties 
 in the way of accepting Fechner's conclusions. 
 
 Experimental methods of measuring threshold stimulus 
 values are open to criticism from both the theoretical and 
 practical standpoints. The theoretical objections are inti- 
 mately associated with certain deductions from the theory of 
 correlation, and cannot be examined here. A practical diffi- 
 culty is the tacit assumption that the perceptive mechanism 
 of the person who is the subject of our experiments is a 
 constant factor throughout a series of observations. This is 
 not, however, true. What is a just noticeable stimulus at 
 
 1 The simplest way of putting it is : If E be the measure of a sensation 
 and dE a liminal increment, R the measure of a stimulus and dR a small 
 increment, then 
 
 K — = dE, -where K is a constant (Weber's law), 
 R 
 
 hence / K =- = E, 
 
 hf- 
 
 or K loge R + Ki = E (Fechner's law). 
 
 ' Such, for example, as will be found in Prof. W. James's Principles of 
 Psychology, vol. i. pp. 548-9. 
 
 8 See particularly "Zur Psychologie der Sinne," Nagel's Handb. d. Physiol. 
 d. Menschen, vol. iii. pp. 16, etc. 
 
6 PHYSIOLOGY OF THE SPECIAL SENSES 
 
 one time may be inappreciable at another, owing to a change, 
 physiological or psychological, in this factor. A higher 
 threshold value is often observed when we proceed from a. 
 weaker to a stronger stimulus than in following the reverse 
 order. It follows that the experimental methods employed 
 to establish Weber's " law " may yield, and do yield, inconstant 
 results, and a system of measurements which gives different 
 readings when operating through the same range is of little 
 service. 
 
 It has also been shown that even under favourable 
 conditions the formula is definitely invalid for very weak 
 or very strong stimuli. 
 
 We seem, therefore, forced to conclude that the experi- 
 mental basis of Weber's " law " is not firm enough to permit 
 our attaching much importance to the formula itself. 
 
 These objections are directed to the foryn Weber's "law'" 
 takes ; that some relation of this kind exists is probable, 
 for all judgments involve comparison ; but it may well be 
 doubted whether the present state of knowledge permits any 
 more definite assertion. 
 
 If the conclusions just stated be sound, Fechner's corollary 
 evidently fails, but, over and above this, special objections to 
 it can be assigned. For instance, Fechner's " law " assumes 
 that sensations are quantitatively measurable, and this is, to 
 say the least, not self-evidently true. Measurements only 
 have a meaning when the unit is clearly and uniquely 
 defined, or, under very special circumstances only, when its. 
 existence can be inferred. Changes of temperature could be 
 measured in various ways, and to say that the temperature 
 of A is 1 degree higher than that of B conveys a definite 
 idea to the mind of him who realises that the scale of units 
 is fixed and has a constant physical significance. 
 
 In the case of sensations, no such scale has been given.^ 
 To say that one body has twice the mass of another is. 
 intelligible, because the unit of mass can be defined ; * 
 
 ' See in this connection Lloyd Morgan's Croonian Lecture, Proc. Roy, 
 Soc, vol. Ixviii. p. 459. 
 
 ' Even in such a case the definition is not particularly easy ; the reader 
 should consult La Science et Vlfypothise, by H. Poincare, chap, viii. 
 
INTRODUCTION 7 
 
 to say that one sensation is twice as intense as another means 
 nothing until the unit is fixed. It seems doubtful, then, 
 whether we are in a position to measure sensations at 
 all, and until we are, the formulation of a quantitative law 
 is hardly justifiable. 
 
 Some of these points may come up for consideration once 
 more in connection with the various organs, but it is well 
 to note from the outset that the numerous statements which 
 have been made with regard to the sense organs in general 
 are liable to very real objections. We are not yet in sight of 
 that simple, all-embracing formula to supply which is the 
 aim of science.! 
 
 BOOKS AND PAPERS RECOMMENDED FOR 
 FURTHER STUDY 
 
 (Bibliographical references under this heading are intended to help 
 students desiring to pursue the subject further. They are invariably 
 confined to works readily accessible in a moderately good library.) 
 
 MtJLLicR's Law 
 
 TV. Nagel, Die Lehre von den Spezifischen Sinnesenergien, Handb. d. 
 Physiol, d. Menschen, herausgegeben von W. Nagel, vol. iii. pp. 1-15. 
 
 L. Asher, Das Gesetz der Spezifischen Sinnesenergie und seine 
 Beziehung zur Entwickslungslehre, Zeitsch. f. Psychologie und Physi- 
 ologic der Sinnesorgane, 1906, vol. xli. p. 157 (this journal will be 
 denoted in future by the letters Z.P.P.S.O.). 
 
 Weber's and Fechner's Laws 
 
 /. V. Kries, op. cit. (a thoughtful and instructive criticism). 
 
 G. F. Lipps, Grundriss der Psychophysik, Leipzig, Sammlung Goschen, 
 1903. An elementary account from the psycho-physical standpoint. 
 
 G. E. Miiller, Die Gesichtspunkte und die Tatsachen der psycho- 
 physischen Methodik, Ergebnisse d. Physiologie, zweiter Jahrgang, 
 1903, second part, pp. 267-517. (A very full, and somewhat difficult 
 account, with complete bibliography of modern psycho-physical 
 literature.) 
 
 Professor James's remarks can be read with pleasure {op. cit.), 
 although his criticism does not appear adequate. 
 
 ' It is not of course implied that to assume a, quantitative relation 
 between stimulation processes and sensation processes is necessarily illegitimate. 
 In fact, as will appear later, such an assumption often leads to extremely 
 useful working hypotheses ; the student ought, however, to realise what 
 caution the proceeding requires. 
 
CHAPTER II 
 
 GENERAL PHYSIOLOGY OF CUTANEOUS SENSATION 
 
 In studying any brancli of knowledge it is an obvious 
 advantage to begin with its simplest and most readily in- 
 telligible manifestations, and to approach only by gradual 
 stages its more complex parts. In most cases this plan 
 can be readily followed, but in sense physiology peculiar 
 difficulties are encountered. To judge from the number of 
 pages allotted in a general text- book to the sense organs 
 of the skin, on the one hand, and to the eye, on the other, 
 it might be supposed that the physiology of the former 
 is much easier than that of the latter. Such is by no 
 means the case; in fact, the converse proposition is quite 
 arguable. At the same time, the study of cutaneous sen- 
 sation is in many respects peculiarly instructive, and it will 
 therefore be convenient to undertake it at once. 
 
 At the outset we shall see that the tendency of modern 
 workers has been to conceive of the cutaneous surface as a 
 universe of points, each reacting in a specific manner to the 
 application of stimuli. We shall study the results which 
 have led up to this, and then examine important modifica- 
 tions of the conception which have been rendered necessary 
 by more recent discoveries. 
 
 Pre- Aristotelian writings on cutaneous sensations are too 
 vague to be of service to the modern reader. The scientific 
 history of the subject really commences with Aristotle.^ 
 
 Aristotle recognised distinctly that the sense of " touch " 
 is separable into many sense modalities, although his limited 
 means of investigation did not permit him to render the 
 
 ' For information as to Greek Sense Physiology the reader is strongly 
 advised to consult Professor J. I. Beare's admirable work, Greek Theories of 
 Elementary Cognition from Alcmeeon to Aristotle. Clarendon Press, 1906. 
 
PHYSIOLOGY OF CUTANEOUS SENSATION 9 
 
 distinctions precise. He also anticipated the view that the 
 organs are affected by a change of state. He regarded the 
 sense of touch as fundamental, and the only one necessary 
 to existence. 
 
 In both these conclusions Aristotle has been followed 
 by the majority of subsequent writers. Without attempting 
 to detail the course of investigations Avhich have led up 
 to the modern epoch, I shall describe in as few words as 
 possible the state of knowledge gained up to a few years ago. 
 
 At least four distinct qualities of sensation can be dis- 
 tinguished when stimuli are applied to the skin — those of 
 touch or pressure, warmth, cold, and pain. Some writers 
 have advocated an even finer analysis, regarding, for instance, 
 the sensation of itching as specific, but this is hardly war- 
 ranted by experiment. It is quite clear that a series of 
 points can be mapped out over the skin, stimulation of 
 which produces the sensations of heat, cold, pain, and touch ; 
 the question of pain, however, will occupy us subsequently. 
 The principles of experimental methods for this sort of work 
 are simple enough, although the actual instruments vary 
 enormously in detail. For experimenting on tactile per- 
 ception, V. Frey, who has done pioneer work in cutaneous 
 sense physiology, devised a series of valuable instruments.^ 
 They consist of a number of wooden rods, to the ends of 
 which hairs of differing thickness are fixed at right angles. 
 If the hair be applied perpendicularly to a point on the skin 
 and pressure exerted, then with the application of a certain 
 pressure the hair will be visibly bent. The amount of 
 pressure required to bend visibly any given hair can be 
 experimentally determined, and with a large series of hairs 
 of different thickness we can use pressures measured and 
 graded with some accuracy. 
 
 Toulouse and Vaschide have justly remarked^ that dif- 
 ferent readings, will be given with the same hair unless the 
 pressure is arrested at the instant of bending, a point not 
 
 ' Bloch described a similar but less perfect instrument in 1891 (Arch. 
 dei'%s.,Avrill891). 
 
 2 Toulouse, Vaschide, and Pieron, Technique de Psychologic Experi- 
 mentale, Paris, 1904, p. 64. 
 
10 PHYSIOLOGY OF THE SPECIAL SENSES 
 
 easy to determine with absolute precision. I also think 
 that the degree of moisture in the air makes some difference. 
 Notwithstanding these and certain other objections, the 
 method seems the most accurate at present in use. To 
 study temperature sensations, a hollow pencil-shaped rod 
 through which water at a definite temperature can be 
 circulated is the best, although French workers claim to 
 have obtained good results by using minute drops of water 
 at various temperatures as stimuli. 
 
 The topography of sensation points was first completely 
 worked out by Magnus Blix, and his conclusions have been 
 endorsed by the best modern observers (Sommer). 
 
 These conclusions are : — 
 
 (1) Pressure points are closely related to the distribution 
 of hairs, each hair having a pressure point near its site of 
 eruption (v. Frey), the point corresponding on the surface to 
 the situation of a hair follicle. There may, however, be some 
 pressure points which do not correspond to hairs. 
 
 (2) Cold reacting points are more numerous than warm 
 points, Sommer finding 12-13 cold points sq. cm. against 1-2 
 warm points, but the distribution is by no means uniform. 
 
 Respecting the anatomical structures which correspond 
 to the sense points, hardly anything is certainly known. For 
 instance, it is not easy to tell whether any given nerve fibre 
 conveys centripetal or centrifugal impulses, the ordinary 
 methods of determining the matter being generally im- 
 possible of application. The relation of pressure points to 
 hairs renders it probable that the nerve fibre entering the 
 root sheath and forming ring-like arborisations round the 
 upper part of the follicle are especially associated with the 
 pressure sense. 
 
 It is also possible that the Meissner corpuscles are pressure 
 end organs, and may even, in the hairless parts of the skin, 
 replace the hairs.^ 
 
 It has been stated that Krause's end bulbs and Ruffini's 
 organs are associated respectively with sensations of cold 
 and warmth, but experimental evidence is of the slenderest. 
 
 ' v. Frey, Beitrage s. Sinnetphys. d. Haut, Konig. Sachs. Gesellschnft d. 
 Wisscnsoh., Marz 1895. 
 
PHYSIOLOGY OF CUTANEOUS SENSATION 11 
 
 Sensations of pressure can normally be evoked from 
 the whole cutaneous surface, the mucous membrane of 
 the mouth, the tongue, teeth, and nares. Kespecting the 
 cornea and glans penis much difference of opinion has 
 existed, but the reason probably may be sought in a failure 
 to distinguish the exact forms of stimuli employed, as will be 
 clear subsequently. 
 
 Strtlmpell has recently^ asserted that sensations of 
 pressure can be set up in internal parts, tendons, fasciae, 
 muscles, and periosteum. Thunberg ^ has suggested that in 
 this case tactile impressions are confused with those of pain. 
 
 The adequate {vide supra) stimulus for the sense of pres- 
 sure is of course mechanical, but since any stimulus applied 
 to the surface does not operate directly upon the end organs 
 — if such there be — the question arises as to what physical 
 conditions in the immediate vicinity of the possibly recipient 
 structures give rise to sensations of pressure. 
 
 This problem was solved by v. Frey and Kiesow ^ in 1899. 
 When a point, e.g. the end of a hair, is pushed against the 
 skin, the pressure is. greatest at the part immediately in 
 contact with the point, and diminished in the deeper and 
 lateral regions ; that is to say, there is a fall of pressure from 
 without inwards. Conversely, if a small plate be affixed 
 to the skin and traction made upon it, the pressure is least 
 externally, and increases from without inwards. In the first 
 case we have a negative and in the second a positive slope of 
 pressure from without inwards. V. Frey and Kiesow found 
 that the sensations experienced in the two cases appeared 
 to be identical, provided the magnitude of the pull in one 
 case were equal to that of the push in the other; that 
 is to say, the end organ is excited by a change in the ex- 
 isting pressure relations of end organ and superficies, 
 whether the change be positive or negative, increase or 
 decrease. 
 
 Although the variation in pressure in the neighbourhood 
 
 1 JDeuUche med. Wochen., 1904, pp. 1411 and 1460. 
 ' Nagel's Ilandb. d. Phys. d. Menschen, vol. iii. p. 657. 
 » Z. P.P. S.O., vol. XX. pp. 126, etc., especially p. 153. Those desirous of 
 expeilmenting on the skin should note the cautions given in this paper. 
 
12 PHYSIOLOGY OF THE SPECIAL SENSES 
 
 of an end organ is an essential condition for the latter's 
 excitation, other factors modify the result. 
 
 For instance, if the stimulus be a falling weight, its 
 velocity is a factor. The greater the velocity, within limits, 
 the less the weight required to produce a just noticeable 
 sensation. 
 
 The sensation does not, however, vary as the kinetic 
 energy of the exciting mass. Two weights allowed to drop 
 from heights inversely proportional to their masses would 
 possess equal amounts of kinetic energy on reaching the 
 skin, but the smaller weight produces the more intense 
 sensation. 
 
 The extent of surface to which the stimulus is applied 
 is also an important matter. It might be thought that the 
 sensations excited in two areas would be identical if the 
 pressure per unit of surface were the same in both, cases. 
 This is not so if, for instance, an area of -5 sq. mm. be com- 
 pared with one of "25 sq. mm. ; a greater pressure per unit 
 is required in the second than in the first case to produce 
 a just noticeable sensation. The optimum surface is approxi- 
 mately '5 sq. mm. ; as we increase or decrease the extent of 
 surface stimulated from these dimensions, the liminal value 
 increases (v. Frey and Kiesow, op. cit.). A partial explana- 
 tion is found in the above-mentioned fact, that pressure 
 sensations are excited by a slope of pressure from the point 
 of application to the end organs. The greater the surface 
 pressed upon, the less steep the slope from within outwards 
 owing to lateral thrusts. This does not account for the fact 
 that if the surface be diminished below a certain value the 
 pressure must be increased, and the proposed solutions of 
 the problem seem hardly adequate. 
 
 The above-mentioned experiments were made on hairless 
 portions of the skin ; the conclusions, however, are, with some 
 modification, valid for hair-clothed surfaces. 
 
 The tactile apparatus of the hairy skin is more excitable 
 and more readily fatigued than that of the smooth areas. 
 Stimuli not distinguishable on a smooth can be detected if 
 applied to a hairy surface. This lowering of the magnitude 
 of a threshold stimulus in the hairy regions perhaps depends 
 
PHYSIOLOGY OF CUTANEOUS SENSATION 13 
 
 on two factors,^ the diminution of the skin area stimulated, 
 since the pressure is only effective through the base of the 
 hair follicle, and the lever-like action of the hair itself. 
 
 The comparison and discrimination of two pressure 
 stimuli have been the objects of many investigations. The 
 quantitative results are discordant, but the following con- 
 clusion appears to be well founded. 
 
 It is easier to compare weights applied successively than 
 when they act simultaneously ; i.e. a more accurate com- 
 parison can be made between the memory of a sensation and 
 an actual sensation than between simultaneous sensations 
 (Weber). 
 
 Again, the rapidity with which stimuli operate influences 
 one's powers of discrimination. Generally, the slower the 
 rate of change, the higher the liminal stimulus value 
 (Dohrn). 
 
 It is also easier to perceive that a given weight differs 
 from another than to say whether it is heavier or lighter. 
 When such discrimination is possible, it is" easier to detect 
 an increase in weight than a decrease. If successive weights 
 be applied with a rapidity such that, in equal times, the same 
 fractions of the original weights are added, the ratios of 
 increase of stimulus to original stimulus give results in fair 
 accordance with Weber's law. It has, however, been shown 
 (Seashore, Hall, and Motora) that the results are variable 
 with the experimental conditions, and no definite proof 
 of the law as applied to tactile impressions is forth- 
 coming. 
 
 Experiments performed with the object of determining 
 what frequency of pressure stimuli gives a fused sensation 
 have yielded most discordant results,^ and need not be dis- 
 cussed. The length of time the sensation persists after 
 removal of the stimulating body depends on the skin area ; 
 in some places the sensation vanishes at once, in others, 
 e.g. the forehead, the after-effect may with moderately strong 
 stimuli persist some little time. 
 
 , . ^ . J IT 1 m,j..--. J. .7. ...1 li n m 4t'J.i.<J-.t:Xli, 'f.-uT- 
 
 1 Aubertand Kammler, M/ldyvra AlUi., vol. IJr. u«l. 
 
 ' Sergi, Z.P P.S.O., iii. 179 ; v. Kiies and Auerbach, Du Bois Reym. Arch., 
 1889 ; V. Vintsohgau and DUrig, PJl. Arch., Ixix. 377. 
 
14 PHYSIOLOGY OF THE SPECIAL SENSES 
 
 These, then, are the ''classical" results of investigations 
 on the sense of touch, and we can pass to those dealing with 
 the temperature senses. 
 
 Sensations of temperature can be elicited from the whole 
 cutaneous surface— the skin of the external ear, parts of the 
 nasal cavity, inclusive of the anterior nares, and the mucous 
 lining of the beginning and end of the digestive tract. There 
 are differences of opinion as to whether such sensibility 
 exists in the cornea, conjunctiva, and glans penis ; ^ we shall 
 see in a later chapter how it is that confusion has existed 
 on this point. There has been much dispute as to what 
 constitutes the adequate stimulus of the temperature end 
 organs, two theories, those of Weber and Herin. having been 
 favoured at different times. ^^ 
 
 Weber supposed that a temperature sensation was de- 
 pendent upon an alteration in the existing end organ tem- 
 perature; thus a rise of temperature in the heat end organ 
 produces a sensation of warmth, and a fall in the temperature 
 of the cold end organ one of cold. In either case such a 
 change could be brought about in four ways. Thus a sensa- 
 tion of warmth might be produced by : (1) A diminution 
 in heat loss with a constant influx of heat ; (2) a constant 
 heat loss and increased, influx of heat; (3) increased heat 
 loss combined with a proportionately greater influx of heat ; 
 (4) diminished heat loss and heat influx, the former 
 diminution being the more considerable. 
 
 It has been shown that whichever of these conditions be 
 fulfilled, the sensation is the same. 
 
 It is further known that, within certain limits, given a 
 constant external temperature sensations originally present 
 disappear, a constant rate of heat influx and efllux being 
 presumably established. 
 
 These facts agree with Weber's hypothesis, but certain 
 experiments seemed difficult to reconcile with it. If a metal 
 disc at 2° C. be applied to the forehead for 30 seconds, the 
 sensation of cold persists for some 20 seconds after removal 
 of the stimulus, that is, while the heat loss is diminished and 
 
 ' See Thunberg, op. cit. , p. G70. 
 
PHYSIOLOGY OF CUTANEOUS SENSATION 15 
 
 the temperature of the end organs actually rising. Hering 
 accordingly proposed a different theory. 
 
 When a certain skin area gives us no temperature sen- 
 sations, the end organs are to be regarded as being at the 
 physiological zero of temperature. Any alteration in the 
 thermal condition will give rise to a sensation, the intensity of 
 which will depend on the difference from this physiological 
 point of reference. The actual situation of the physiological 
 zero point is presumed to vary within tolerably wide limits, 
 not being the same for all skin surfaces, nor for the same skin 
 surface at different times. 
 
 On entering a warm room a sensation of warmth is 
 elicited, because the physiological thermometer is set for a 
 lower point ; but if the stay be prolonged, the physiological 
 zero is readjusted to a higher level, and the temperature 
 sensations disappear. 
 
 The difference between the theories is that Weber 
 supposes sensations to arise during variation, Hering when 
 the end organ temperature is constant. 
 
 Holm seems to have demonstrated that, if precautions be 
 taken to prevent the spread of stimulation beyond the point 
 originally involved, the sensation only persists for a short 
 time after withdrawal of the stimulus, the time being so 
 short as to cqrresipond to the probable time during which the 
 end organ tempetature varies. The same investigator studied 
 the experiment alluded to before, in which, after application 
 of a cold piece of metal to the forehead, the sensation of cold 
 persisted for some time after removal of the stimulus. Holm 
 noticed that a distinct interval separates the appearance 
 of the cold after-sensation from that at first experienced, and 
 suggested that we are here dealing with a case of paradoxical 
 sensation, that the cold end organs are stimulated by the 
 warm blood which arrives in increasing amount owing to 
 vasrodilatation, and that they respond specifically with a 
 sensation of cold. The existence of such paradoxical sen- 
 sations has long been known (Striimpell, Alrutz, Thunberg, 
 and others). 
 
 Holm also found that after-sensations of cold do not 
 follow the application of moderately intense stimuli, but only 
 
16 PHYSIOLOGY OF THE SPECIAL SENSES 
 
 that of strong ones. It may therefore be said that the 
 experiment is too complex to be regarded as decisively 
 negativing the theory of Weber, and that it is still question- 
 able which of the hypotheses is to be preferred. 
 
 With respect to the factors which influence the threshold 
 value of temperature stimuli, the following may be enu- 
 merated as the most important : — 
 
 (1) Situation of the point stimulated; warm and cold 
 points are by no means uniformly distributed over the body 
 surface. For instance, Goldscheider ^ found that cold points 
 were far more numerous over the extremities than warm 
 points. This author's methods have been adversely criticised, 
 and there is no doubt that the topography recorded by 
 different investigators is highly variable. I believe the 
 inconstancy is partly due to real variation in the dis- 
 tribution of end organs, and partly to inexact methods of 
 study. 
 
 (2) The actual size of the region stimulated. Thus 
 Weber asserted that water at 29'5° Reaumur feels warmer 
 than water at 32° R. if the whole hand be dipped into the 
 former and only one finger into the latter. At the same 
 time it is to be remembered that the considerations enume- 
 rated under (1) may account for Weber's result. It is 
 difficult to plan an experiment in which these sources of 
 error are avoided. 
 
 (3) The physical condition of the stimulating body, its 
 thermal capacity, conductivity, and more or less perfect 
 application to the skin, are important, and play their part 
 in the divergent results noted under (1). 
 
 Finally, it may be said that the numerous researches, 
 dating back to the middle of the last century, in which 
 the discrimination of temperature difi'erences has been 
 studied, have not yielded constant enough results to demand 
 special attention in an elementary work. Attempts to 
 establish the validity of Weber's law in the case of tempera- 
 ture sensations have not been successful. 
 
 ' Arch.f, Psi/chiatric und Nervenlcr., xviii. 659. 
 
PHYSIOLOGY OF CUTANEOUS SENSATION 17 
 
 BOOKS RECOMMENDED FOR FURTHER STUDY 
 
 An excellent account o£ modern work on the subject will be found in 
 Professor Thunberg's article (Nagel's Handb. d. Phys., Bd. iii. p. 647). 
 Of the original papers, those by v. Frey should be read first. 
 
 A valuable discussion of the experimental methods of dealing 
 with the subject is contained in chapter i. of " The Grouping of Afferent 
 Impulses within the Spinal Cord," by H. Head and T. Thompson 
 (Braiuj for 1906, part cxvi.). 
 
CHAPTEE III 
 
 PAIN 
 
 Although what is commonly termed pain is not exclusively 
 associated witli cutaneous stimulation, it will be found that 
 the matter is best considered at this point. The discussion 
 must be brief, since most of the problems which arise 
 belong rather to the psychology than the physiology of the 
 sense organs. 
 
 Many definitions of pain have been proposed by various 
 writers, of which the following examples will suffice : — 
 
 "Pain is a change of sensibility repugnant to him who 
 experiences it " (Mantegazza). 
 
 "Pain is any disagreeable sensation (sensation penible) 
 perceived by the nervous centres, but varied in its modalities, 
 effects, and causes (Eloy). 
 
 " A sensation such that one does not desire to experience 
 it afresh " (C. Richet). 
 
 While it must be allowed that it is not an easy task to 
 construct a satisfactory definition of pain, it appears that 
 the authors cited confuse what should, if possible, be kept 
 distinct. But little reflection is necessary to convince the 
 reader that a sensation may be unpleasant, may be repugnant 
 in the highest degree, without falling under the category 
 which includes, for instance, the sensation experienced when a 
 needle is thrust into the skin. 
 
 The extraction of a tooth without gas is a disagreeable 
 experience, and so is the death of a friend, but it is surely an 
 abuse of language to employ the word pain indifferently for 
 the sensory phenomena associated with such disparate causes. 
 
 We have here an example of the " fallacy of words," dis- 
 cussed by Bentham and other logicians. German writers 
 have attempted to remedy this evil by using the terms 
 
 18 
 
PAIN 19 
 
 "Unlust" and "Schmerz." Any excessive stimulation of an 
 afferent system is associated with "Unlust," and the word 
 can also be made to cover what popular English writers convey 
 by the phrase mental or moral pain. It is, however, true 
 that many authors use " Schmerz " almost as loosely as " pain " 
 or " douleur." Indeed, however one may strive for accuracy, 
 the difficulties seem almost insuperable. Although there is 
 fairly good evidence that the sensation associated with stimu- 
 lation of the skin by a sharp instrument, such as a needle, is 
 sui generis and dependent on the existence of specific organs, 
 yet no absolute distinction can be drawn between this and 
 the " pain " of visceral disease. 
 
 I think therefore, although it is a frank confession of 
 weakness, that it will be convenient to use the word pain 
 without a qualifying adjective somewhat in the sense of 
 " Unlust," while I shall speak of the sensation which appears to 
 have a specific basis as cutaneous pain. I am not at all clear 
 bow far visceral pain is or can be separated from " Unlust." 
 
 Funcke appears to have been among the first physiologists 
 actually to advance the opinion that cutaneous pain is dis- 
 tinguisbable from other afferent impressions arising from the 
 skin. He suggested that the separation might occur first in 
 the spinal cord, or that the impulses are insulated throughout 
 their whole course from periphery to sensorium. The re- 
 searches of V. Frey made it possible to render these ideas 
 more precise. It was found that by the employment of suit- 
 able mechanical stimuli clearly defined points not generally 
 coincident with the touch s-pots could be mapped out which 
 yielded maximal sensations of pain. In the ordinary way it 
 is not possible to induce a sensation of cutaneous pain entirely 
 free from one of touch, but this can be done if only very 
 sharp needles or bristles are used and the epidermis is 
 moistened. 
 
 From such experiments, and from the further observation 
 that the points would also respond to chemical stimuli by 
 sensations of cutaneous pain, it was possible to conclude that 
 specific organs for cutaneous pain exist. The " pain points " 
 are characterised by a very long latent period when subjected 
 to weak stimulation, and a special inactivity of response to 
 
20 PHYSIOLOGY OF THE SPECIAL SENSES 
 
 rapidly alternating or oscillating stimuli. They do not 
 coincide ■with pressure points, than which they are at least 
 four times more numerous. 
 
 The belief that cutaneous pain was induced by direct 
 stimulation of nerve fibres, is contradicted by the fact that 
 it can be elicited by the employment of stimuli too weak to 
 act as ordinary nerve excitants. Further, weak mechanical 
 stimuli, just above the liminal value, provoke a delayed sen- 
 sation of pain, the delay being greater than can be accounted 
 by the time required for the conduction of an impulse to the 
 spinal cord, and greater than the interval which separates 
 stimulus and response when the former is applied directly to 
 a nerve fibre. 
 
 Although the interpretation of these experiments is more 
 dubious than some think, they are not inconsistent with the 
 hypothesis of specific organs for cutaneous pain. V. Frey 
 suggests that the free nerve endings of the skin are the 
 organs in question, and the following confirmatory evidence 
 is adduced : — 
 
 (1) Diminution of the surface area stimulated does not 
 diminish the effectiveness of the stimulus. 
 
 (2) The liminal value of an electrical stimulus is lower 
 than that of any other form. 
 
 (3) Corrosives produce, first of all, a sensation of pain. 
 These facts suggest that the pain end organs are more 
 
 superficial than those of touch or temperature. The only 
 nervous elements definitely superficial to the end bulbs 
 supposed to be associated with touch are the free intra- 
 epithelial nerve endings, hence they should be the. required 
 organs. 
 
 It has also been asserted that, in the cornea, the only 
 receptive elements are the intra-epithelial nerve endings, and 
 that from the cornea painful sensations alone can be elicited. 
 V. Frey suggests that the actual excitant might be fluid 
 squeezed from the cells into the inter-cellular spaces. Such 
 fluid not being sufficiently concentrated owing to the im- 
 permeability of the cell wall for certain salts, its low con- 
 centration may render it a chemical stimulus for the nerve 
 endings. Apart from other objections which can be raised. 
 
PAIN 21 
 
 it should be noticed that Donaldson finds sensations other 
 than those of cutaneous pain can be derived from the 
 cornea. 
 
 I have already remarked on the confusion which has 
 existed respecting the meaning of the word pain. Even 
 with respect to cutaneous pain, as above defined, there is 
 room for further specialisation. Thunberg, for instance, 
 seems to hold the view that aching and stabbing or pricks 
 ing (cutaneous) pains depend on totally different mechanisms. 
 He points out that on pinching up a large fold on skin, 
 pressure, however slight, elicits a sensation of aching pain 
 alone, while the application of superficial stimuli under 
 normal circumstances excites only pricking sensations. 
 Thunberg has not, however, shown satisfactorily that the 
 difference of sensation may not be related to the very dif- 
 ferent stimuli employed and their conditions of application. 
 
 A few notes on visceral pain v^ fitly be introduced at 
 this stage. It was formerly believed that the viscera are 
 insensitive under normal circumstances, but become capable 
 of originating painful sensations when diseased or injured. 
 
 The work of Lennander, RamstrOm, and others has proved 
 that this conception is erroneous. 
 
 Lennander demonstrated that all painful sensations set 
 up in the pelvic viscera can be assumed to arise in areas in- 
 nervated by the lumbosacral and intercostal nerves, espe- 
 cially in the parietal layer of the peritoneum. For instance, 
 increased peristalsis set up by an inflammatory condition of 
 the gut acts as a mechanical stimulus, exerting traction on 
 the peritoneum. According to Ramstrom, traction alone is 
 an effective stimulus for the parietal peritoneum. Lennander 
 believes that all organs innervated by the sympathetic or 
 by the vagus below the origin of the recurrent laryngeal are 
 insensitive ; he has found that pain is not excited by stimula- 
 tion of the anterior wall of the vagina, the uterus, ovaries, 
 and Fallopian tubes. 
 
 Next, as to referred pain. The fact that the pain asso- 
 ciated with visceral disease is often localised in a cutaneous 
 area rather than in the affected region, was noted many years 
 ago by Lange. Henry Head was, however, the first to investi- 
 
22 PHYSIOLOGY OF THE SPECIAL SENSES 
 
 gate the subject systematically and render the results of care- 
 ful observation and experiment available in practice. 
 
 It appears from Head's researches that certain stimuli 
 applied to the viscera are capable of eliciting pain, but pain 
 not localised in the viscera but referred to some cutaneous 
 surface. 
 
 The area of reference was constant and, apart from actual 
 pain, was often hypereesthetic. The constancy of these pain- 
 ful and hyperEBsthetic zones render them of much value 
 for diagnostic purposes. Physiologically, the results can be 
 best described in terms of the following hypothesis : It may 
 be laid down as a fact of experiment and general observation 
 that if a painful stimulus be applied to, e.g., a point of a 
 normal foot, the resulting sensation is localised near to the 
 point of stimulation ; if, however, the sensibility of the skin 
 is impaired in that region, the sensation will be localised 
 in the nearest (physiologically speaking) unaffected point. 
 That is to say, if two spots, A and B, are physiologically 
 associated, for instance, by nervous inter-connections, and if A 
 be an area of lower sensibility than B — it may be normally, 
 or pathologically, owing to disease of the afferent paths — 
 then a stimulus applied to A is localised in B. A viscus is 
 the A of our illustration, and B some corresponding or con- 
 nected skin area. Following the train of ideas a little 
 further, if A send up aff'erent impulses to some common 
 " centre " C, an alteration of the state of equilibrium in C 
 may result, with the consequence that if stimuli be now 
 applied to B, an abnormal conscious response may be 
 obtained owing to the modification of the recipient " centre." 
 
 The nature of the connecting link between our A and B 
 has also been studied. It was first suggested that posterior 
 root ganglia receiving fibres from viscus and corresponding 
 skin area formed the link. Subsequent work suggested that 
 the correlated parts could be grouped more satisfactorily by 
 a schema derived from our knowledge of the primitive 
 segmentation of the cord. 
 
 Kecent work has, on the whole, rather tended to support 
 the original hypothesis, but the matter is not yet definitely 
 settled 
 
PAIN 23 
 
 I have now described some of the main facts known as 
 to cutaneous pain. In the next chapter we shall consider 
 some important modifications of our beliefs due to recent 
 studies. 
 
 BOOKS, ETC., RECOMMENDED FOR FURTHER STUDY 
 
 Pain — 
 
 Thunberg, op. cit. 
 
 Psycho-phyaiologie de la Douleur, par /. loteyho et M. Stefanowsha, 
 
 Paris, Alcan, 1909, pp. 249. (Contains a fairly good bibliography. 
 
 Must be read critically.) 
 Richet. Art. Douleur, Diet, de Physiologie. (The same remarks 
 
 apply as in the previous case.) 
 
 Referred Pain, etc.— 
 
 Consult the numerous papers by Head and his collaborators which 
 have appeared in Brain from 1888 onwards. 
 
CHAPTER IV 
 
 PROTOPATHIO AND EPICRITIC SENSIBILITY 
 
 The reader who lias examined the results epitomised in the 
 last two chapters will have noticed that, mainly in con- 
 sequence of the researches initiated or inspired by Max v. 
 Frey, a fairly coherent description of the sensory mechanisms 
 of the skin has been reached. In effect, it has become 
 customary to regard the cutaneous surface as a vast congeries 
 of points each responding to stimuli by a specific afferent im- 
 pulse, the heat spots responding by a sensation of heat, the 
 touch spots by one of touch, and so on. The whole con- 
 ception is conveniently resumed by the phrase " Punctate 
 Sensibility." 
 
 Although this belief certainly reflects the opinions of a 
 distinct majority of workers, and adequately presumes a large 
 number of careful observations, objectors and difficulties have 
 not been wanting. Oppenheimer, for instance, has sharply 
 criticised v. Frey's interpretations of his experiments, and the 
 state of sensation after injury to nerves or section of them 
 has been difficult to reconcile with the doctrine of punctate 
 sensibility. 
 
 It was plain that a careful study of the sensory changes 
 . associated with division of cutaneous nerves must throw 
 needed light on the nature of the mechanisms concerned. 
 It was equally apparent that the necessary conditions of such 
 study would never be fulfilled by the casual subjects of 
 accidental injury, since the training and experience necessary 
 for really efficient work on cutaneous sensibility are, perhaps, 
 greater than for any other branch of sensory research. Dr. 
 Henry Head of the London Hospital was the first man of 
 science to realise practically the importance of these con- 
 siderations. To the enthusiasm and ability of this worker, 
 
 21 
 
PROTOPATHIC AND EPICRITIC SENSIBILITY 25 
 
 together with his collaborators, Rivers, Sherren, Ham, and 
 Thompson, we owe the initiation of a new epoch in sense- 
 physiology. Other workers have not been slow to avail 
 themselves of Head's pioneer work, e.g. Trotter and Davies, 
 who have recently published an interesting contribution to 
 the subject. 
 
 Signs are not wanting that the method of direct experi- 
 ment upon man, naturally under conditions which preclude 
 as far as humanly possible the risk of permanent injury, will 
 enable us to complete and co-ordinate the results of ordinary 
 experimental physiology. 
 
 When one remembers the essential difficulty of psycho- 
 physiological work, it is not surprising to find that the results 
 of the pioneer researches are not in all respects concordant. 
 It is also probable that subsequent progress will necessitate 
 essential modifications in the theoretical schemata proposed 
 by the observers. I shall describe those portions of the work 
 which seem to be best established, and finally indicate the 
 theory of the matter which seems at the moment to be the 
 most adequate working hypothesis. It is to be remarked 
 that any one really interested in the matter should examine 
 the original papers. 
 
 When the median nerve is divided, sensation is entirely 
 lost over a considerable part of both index and middle fingers ; 
 over the palm, within the area said by anatomists to be 
 supplied by this nerve, sensation is usually diminished but 
 not completely lost. Similarly, division of the ulnar pro- 
 duces complete anaesthesia of the little finger and of a 
 variable portion of the ulnar aspect of the palm. Partial 
 loss of sensation occurs over a larger aspect of the palm 
 and of the ulnar surface of the ring finger. 
 
 On examining what is meant by saying that sensibility is 
 diminished in a part of the ulnar territory, we find that what 
 really happens is that certain sensations are completely lost 
 and others retained. If in a patient who has divided his 
 ulnar nerve the ulnar half of the palm be stimulated with 
 cotton wool no sensation will be produced, while the lightest 
 touch will be appreciated directly the line corresponding to 
 the axis of the index finger is passed. Compass points 2 
 
26 PHYSIOLOGY OF THE SPECIAL SENSES 
 
 centimetres apart are not distinguished, nor temperatures 
 between 22° and 40° Centigrade. 
 
 " When the hand has settled down after the shock of the 
 injury that has divided one or more nerves to the palm, it 
 will be found that although the area we have spoken of is 
 totally insensitive to certain higher forms of stimulation, a 
 stimulus producing pain over normal parts, e.g. a prick of a 
 pin, causes a more unpleasant effect than over normal parts. 
 If the nerve has been united, sensation begins to return after 
 a variable interval. The first sign of recovery is a gradual 
 diminution in the extent of the area insensitive to pain and 
 to all forms of heat and cold." 
 
 Eventually no part remains irresponsive to all stimuli, and 
 at a time when the boundary of light touch is still distinct, 
 responsiveness to pin prick and thermal stimuli has much 
 improved. 
 
 At this stage, painful stimuli produce effects which radiate 
 widely and are particularly impleasant. Ice and water at 50° 
 C. can be detected as cold and hot, but no intermediate 
 degrees. 
 
 These observations were made on hospital patients. We 
 now come to the experiment on Head himself. 
 
 The radial and external cutaneous nerves were divided 
 near the elbow with aseptic precautions, and the ends sutured 
 together. Immediate examination of the surface innervated 
 by the divided nerves showed the following condition : — 
 
 (a) Very moderate pressure was appreciated and well 
 localised, but touches with cotton wool or deformations of the 
 skin produced by drawing hairs outwards were without effect. 
 
 (6) Compass points even 8 centimetres apart could not be 
 distinguished. 
 
 (c) Heat, pain, and cold were all lost. 
 
 {d) Between the margin of the analgesic area and that 
 insensitive to cotton wool lay a zone where the prick of the 
 pin was abnormally painful. 
 
 (e) None of the cold spots marked out before the opera- 
 tion reacted to ordinary stimuli. 
 
 It will be seen that the responsiveness still existing within 
 the analgesic area was merely to forms of stimuli which pro- 
 
PROTOPATHIC AND EPICRITIC SENSIBILITY 27 
 
 duced inward deformation of the skin, and may be supposed 
 to have acted on the end organs below the cutaneous level ; 
 that is to say, there was, as we should expect, a conservation 
 of deep sensibility. The importance of the observation is, 
 that it illustrates that much which is attributed to cutaneous 
 sense organs is really due to a quite different afferent system. 
 Trotter and Davies, who have made somewhat similar obser- 
 vations, desire to restrict the term "Sense of Pressure" to 
 the afferent correlate of deforming stimuli and to keep the 
 word touch for strictly cutaneous sensibility. Although a 
 distinction of this kind is desirable for the sake of a clear 
 terminology, I am not certain that the separation of the 
 effects by an introspective analysis is so definite as these 
 authors hold. 
 
 Within the analgesic zone, no change was observed for 
 seven weeks ; at the end of that time sensibility to pin prick 
 returned, and was complete in about 200 days. At this stage 
 it was found that ice and water at 50^ C. could be sensed as 
 respectively cold and hot, but no " intermediate " sensations 
 of warmth or coolness were detected. The areas from which 
 these sensations were obtained coincided with the hot and 
 cold spots. The cold spots reacted to stimuli from 0° to 24°, 
 the hot ones to temperatures from 38° to 45°. A curious 
 observation is the following : A cold spot of unusual activity 
 was stimulated by means of a copper cylinder 1 mm. in 
 diameter cooled to the temperature of melting ice. This 
 produced a sensation of cold. Water at 20° C. was then 
 placed in a test tube with a flat bottom of 1 cm. diameter 
 applied to the skin in such a way that it stimulated a con- 
 stellation of spots, including the one originally stimulated. 
 The resultant sensation of cold was more vivid than that 
 produced by stimulating a single spot with a considerably 
 lower temperature. 
 
 Pain, as recognised at this period, exhibited marked pecu- 
 liarities. No exact punctate distribution could be ascertained. 
 The minimal stimulus requisite to produce a response was 
 higher than for normal parts, but the response when pro- 
 duced was peculiarly unpleasant, radiating and tending to 
 be localised in remote parts. 
 
28 PHYSIOLOGY OF THE SPECIAL SENSES 
 
 These phenomena of radiation and remote reference were 
 specially characteristic of this stage of recovery. Deep sen- 
 sations — in the sense used above — were well localised, but 
 as recovery proceeded this power was lost or obscured. 
 
 "To evade the localisation due to deep sensibility we 
 employed minute drops of ether or of ethyl chloride, instead 
 of the ice-cold tube ; the characteristic radiating and referred 
 sensations of cold then appeared unhampered by the conse- 
 quences of pressure. Such a stimulus applied to the wrist 
 might cause the whole affected area on the back of the hand, 
 including the greater part of the thumb, to become icy cold, 
 and stimulation of a group of spots on the forearm was 
 followed by an intense coldness over the whole dorsal surface 
 of the thumb." 
 
 We have, therefore, at this stage a fairly well defined 
 group of sensations, viz. sensibility to temperature stimuli 
 restricted to two categories, the hot and the cold, and elicited 
 by stimuli of widely different intensities ; sensibility to pain, 
 abnormally difficult to bring over the infra-conscious threshold 
 associated with marked " Unlust," radiating and often referred. 
 
 This kind of sensibility is termed by Head Protopathic. 
 
 After a variable time the skin again becomes sensitive to 
 light touch and moderate grades of temperature stimulation. 
 Compass points are discriminated, and the radiation char- 
 acteristic of the former stage disappears. These later- 
 returning sensibilities are called Epiceitic. 
 
 It is evident that the desirability of forming these two 
 classes will depend on whether we can answer the following 
 questions in the affirmative : — 
 
 (1) Does any part of the body normally exhibit a re- 
 sponsiveness of the protopathic type only? 
 
 (2) Can protopathic and epicritic areas be delimited with 
 accuracy either (a) in point of tiine or (b) spatially within 
 the region deprived of cutaneous afferent paths by experi- 
 mental division of nerves ? 
 
 Head and Rivers have, in my opinion, made out a very 
 strong case in favour of their conclusion, that the glans penis 
 is endowed with protopathic and deep sensibility alone. 
 
 "The glans is entirely insensitive to stimulation with 
 
PROTOPATHIC AND EPICRITIC SENSIBILITY 29 
 
 cotton wool and with the tactile hairs. ... As soon as hairs 
 of greater bending-strain, the so-called ' pain-hairs,' are used, 
 the glans is found to be sensitive to from 70 to 90 grm./mm^. 
 This produces a characteristic diffuse boring or stinging pain 
 much more unpleasant than over the skin of the penis or 
 foreskin; v. Frey specially remarks that the pain is of a 
 different character from that over the normal skin." 
 
 Although no sensations of heat seemed attainable, perhaps 
 owing to the absence of heat spots, water from 0° to 21° C. 
 produced an intense sensation of cold. In the other subject 
 heat spots were present round the meatus, and water above 
 40° was always called warm. No response was obtained with 
 temperatures between 26° and 37° C. It is obvious that these 
 results agree well with the characters of protopathic sensi- 
 bility as above described. 
 
 Although much interesting evidence is adduced, it is 
 not clear that the answer to our second question is quite 
 so decisive. 
 
 A small triangular area on Head's wrist is described in 
 the following terms : — 
 
 Deep sensibility was undoubtedly present. There was 
 also definite responsiveness to cutaneous stimulation, either 
 in the form of moderately fine v. Frey hairs or cotton wool, 
 and the point of application of such stimuli was well 
 localised. Cutaneous pain was entirely absent. " Strong 
 interrupted currents unbearably painful over the normal 
 skin produced the characteristic whirring sensation devoid 
 of any element of pain. The observations on thermal 
 sensibility within this area are not quite so definite. It 
 seems to be clear that at first no stimuli below 22° produced 
 a response, that is, at a time when touch was accurately 
 localised. In the same way no heat spots were present, but 
 temperature stimuli acting upon comparatively large surfaces 
 (diameters of 1 or more centimetres) were said to call up a 
 sensation of warmth if between 42° and 48°. Tubes at 50° 
 or higher temperatures elicited a sensation of contact, but no 
 thermal response." Head and Rivers observe that one of 
 their chief experimental difiiculties was " the tendency which 
 Head showed to call cold stimuli ' warm ' within the limits of 
 
30 PHYSIOLOGY OF THE SPECIAL SENSES 
 
 the triangle. Whenever a thermal sensation was produced 
 at all, it was one of warmth ; some of the most satisfactory 
 warm sensations were evoked by an ice-cold tube, and yet, 
 at this time, temperatures of 50° C. and above were not 
 appreciated." 
 
 Eventually responsiveness to painful stimuli was restored 
 and a thermal sensibility of the ordinary punctate form. 
 
 In describing the state of afi'airs within the " triangle " I 
 have availed myself of the complete account published in 
 Head and Rivers' latest (November 1908) memoir. This paper 
 was not, I think, published when Trotter and Davies wrote 
 their criticism of the work. Some of their remarks, therefore, 
 are at best applicable to the statement of the case as described 
 in the preliminary communication of Head, Rivers, and 
 Sherren. In view of the fact that stimuli between 42° and 
 48° called up a sensation of warmth, together with the 
 other phenomena noted, I am of opinion that the sugges- 
 tion of Trotter and Davies that the thermal sensibility 
 was hallucinatory, or — as generally termed — paradox warmth, 
 occurring within an area of ordinary thermo-ansesthesia, 
 cannot be accepted. Indeed, although as a pure theorist I 
 feel it to be presumptuous to criticise the opinions advanced 
 by such careful and experienced observers as Trotter and 
 Davies, their views on the mechanism of thermal sensibility 
 are not, I think, entirely consistent. Thus they write {op. 
 cit, p. 147) : " Of two grades of the sensation of warmth it 
 cannot be said that there is anything in the one which is not 
 in the other, the difference is one purely of intensity; on 
 the other hand, between the sensation of heat and the most 
 marked warmth there is a distinct difference of quality — heat 
 possesses a certain abruptness and, so to say, brightness 
 which warmth lacks." 
 
 On the following page of their memoir, however, I find 
 this passage {op. cit, p. 148) : " When a temperature spot 
 has its sensitiveness reduced, the sensations roused by the 
 appropriate temperatures are reduced in intensity, so that, 
 for example, on a heat spot a temperature which gave before 
 a sensation of warmth now gives indifferent warmth or 
 indifference, while on a cold spot a temperature which gave 
 
PROTOPATHIC AND EPICRITIC SENSIBILITY 31 
 
 cool now gives indifferent cool or indifference. The same of 
 course applies to the maximal sensation in each case — warmth 
 is felt instead of heat, cool instead of cold." On the whole, 
 the reader will perhaps conclude that Head and Rivers have 
 established a prima facie case in support of the view that, 
 the second fundamental question must be answered in 
 the affirmative, so far as temporal separation is con- 
 cerned. 
 
 As to whether this is also true for the second part of the 
 question seems much more doubtful. It does not appear 
 in view of Trotter and Davies' work, that in the partly 
 anaesthetic areas due to nerve section a sharp border be- 
 tween regions of protopathic plus deep sensibility and deep 
 sensibility alone can be made out. The question, however, 
 turns so largely on the respective values of the techniques 
 used by Head and Rivers on the one hand and Trotter 
 and Davies on the other, that I do not think those who have 
 not actually experimented themselves can usefully formulate 
 opinions. 
 
 Head and Rivers express their general conclusions in the 
 following way : — 
 
 (1) The skin is supplied by two anatomical distinct 
 systems, the epicritic and the protopathic, which regenerate 
 at different periods afcer section of an afferent nerve, and may 
 be one or other, alone present in certain regions. The glans 
 penis is an example of a part supplied by the protopathic 
 system only, while the triangular area on Head's arm was, 
 for some time, innervated through the epicritic system alone. 
 
 (2) Protopathic sensibility depends on the existence of 
 specific end organs, the areas between such organs being, 
 in the absence of epicritic innervation, insensitive to purely 
 cutaneous stimulation. 
 
 (3) In a part supplied by protopathic end organs alone 
 any sensation evoked radiates widely and tends to be referred 
 to distant parts. Radiation and reference are abolished 
 when the part becomes sensitive to cutaneous tactile stimuli 
 and intermediate thermal stimuli. 
 
 (4) Thermal adaptation is a function of the epicritic 
 system. 
 
32 PHYSIOLOGY OF THE SPECIAL SENSES 
 
 (5) Touch spots are not analogous with heat, cold, or pain 
 spots. 
 
 (6) All protopathic organs have a high threshold; the 
 heat spots do not react to temperatures below 37°, nor 
 the cold spots to temperatures above 26°. The pain spots 
 on the back of the head are not sensitive to pressures below 
 70 grm./mm^. 
 
 Epicritic organs, on the other hand, have a low threshold, 
 responding to temperatures between 26° and 38°, and to 
 tactile stimuli (on the back of the hand) of 12 grm./mm^ 
 
 It will have been noticed that a cardinal distinction 
 between epicritic and protopathic sensibility is the uniformly 
 lower threshold value of the former. It is also to be re- 
 marked that when the afferent impulses have reached the 
 spinal cord no separation into epicritic and protopathic 
 groups can be made out. In certain forms of haemorrhage 
 into the cord, motion is lost in one limb and all forms of 
 sensation in the other. The careful researches of Theodore 
 Thompson seem to have established the following conclusions 
 with respect to intra-spinal conduction : — 
 
 The ascending paths are three in number, one carrying up 
 painful and thermal impulses of all forms, a second con- 
 veying tactile impulses, and a third those connected with the 
 sense of position. The distinction between protopathic and 
 epicritic systems is therefore definitely sub-spinal. 
 
 We may perhaps sum up the discoveries of Head and 
 Rivers with the help of the following theoretical schema : — 
 
 In the peripheral nerves there exist two conducting 
 mechanisms functionally and, it may be, anatomically 
 separate. One system is relatively impermeable to in- 
 coming impulses, and represents a more primitive type of 
 nervous organisation; the other, more delicately organised, 
 less resistant to impulses, representing a more complex type 
 of nervous system. This second system, in virtue of its more 
 delicate organisation, will be regenerated after nerve injury 
 in most cases more slowly ; it will fail to transmit the results 
 of strong stimuli, being as it were temporarily disorganised 
 by them, and Avill be less widely distributed than the former 
 system. The first, or, to use a rude electrical analogy, high 
 
PROTOPATHIC AND EPICRITIC SENSIBILITY 33 
 
 tension system is Head and Rivers' protopathic system, the 
 second, or low tension, their epicritic system. The sole or 
 main difference between the two classes is one of finer or 
 coarser organisation in the afferent paths up to the cord. 
 In the cord itself, resistance to all forms of impulse becomes 
 the same. 
 
 These brilliant discoveries enable us to form some rational 
 conception of the manner in which, through the conflict of 
 sensory impulses, peripheral specialisation has come about. 
 At one time we may suppose high threshold stimuU alone 
 produced a response ; fine sensory discrimination was not an 
 important factor in the struggle for existence owing to the 
 imperfect adaptation of the whole organism to its environ- 
 ment. At this stage all sensibility was protopathic. As 
 adjustment became less imperfect, selection for finer powers 
 of sensory discrimination began to operate, and we reach the 
 germs of an epicritic system. 
 
 This train of thought can be applied to the process of 
 natural selection, not merely between organisms taken as 
 a whole, but between diff"erent parts of the same organism, 
 the inter-cellular struggle for existence, to which Le Dantec 
 has called attention. 
 
 Certain parts of the superficies are from their situation, 
 and the necessity for their performance of specialised 
 functions, not likely to have to deal with the finer processes 
 of sensory discrimination. An admirable illustration is 
 afforded by the external genitalia. It is clear enough that 
 coitus does not require a sensory mechanism of the 
 specialised grade associated with the epicritic system, and 
 we have seen that the glans penis is furnished with proto- 
 pathic and deep sensibility alone. 
 
 A further extension of the ideas sketched in the preced- 
 ing paragraphs may be left to the reader, who will, I think, 
 speedily realise how fruitful and important in many directions 
 are the researches discussed in this chapter. 
 
34 PHYSIOLOGY OF THE SPECIAL SENSES 
 
 PAPERS RECOMMENDED FOR FURTHER STUDY 
 
 Head, Rivers, and Sherren, The Afferent Nervous System from a 
 New Aspect, Brain, 1905, xxviii. 99. 
 
 Head and Thompson, The Grouping of Afferent Impulses within 
 the Spinal Cord, Brain, 1906, xxix. 538. 
 
 Head and Rivers, A Human Experiment in Nerve Division, 
 Brain, 1908, xxxi. 324. 
 
 Trotter and Bavies, Experimental Studies in the Innervation o£ the 
 Skin, Journal of Physiology, 1909, xxxviii. 134. 
 
 Thompson, On Certain Changes in Sensation which Occur in Cases of 
 Cross Lesions within the Spinal Cord. M.D. Thesis, London, 1906. 
 Jarrold and Sons. 
 
CHAPTER V 
 
 TASTE AND SMELL 
 
 The senses of taste and smell have many points of resem- 
 blance, one being the scantness of our knowledge respecting 
 the physiology of either. 
 
 The researches of the Greeks respecting the psycho- 
 physiology of taste have not been fruitful, although in the 
 Timieus, Plato uses language which can be understood 
 to imply that the physical basis of gustatory stimulation 
 is chemical. 
 
 At present we must confess ignorance as to the real 
 mechanism of the stimulation processes, and admit that the 
 end organs themselves, and the paths by which impulses 
 arising in the end organs are transmitted to the brain, have 
 not been determined with great exactitude. 
 
 The ostensible end organs, the taste buds comprising 
 " gustatory " and supporting cells, are somewhat widely dis- 
 tributed. They occur on the upper surface of the tongue, 
 the soft palate, the posterior surface of the epiglottis, and, 
 to some extent, within the larynx. 
 
 Whether the gustatory cells are really end organs in the 
 sense in which this term may be applied to the retinal cones 
 and rods, or whether they merely act as props upon which 
 the nerve endings twine, is a matter of doubt. 
 
 Within the sensory province of taste we can recognise but 
 four sensation-qualities — the sweet, the salt, the acid, and 
 the bitter. 
 
 Investigation can be made with the help of several 
 methods. The simplest, that used by Oehrwall and many 
 others, is to apply a fine camel's-hair brush dipped in test 
 solutions to individual papillae, the latter being illuminated 
 by light reflected from a small concave mirror. 
 
 •' ° 35 
 
36 PHYSIOLOGY OF THE SPECIAL SENSES 
 
 Toulouse and Vaschide employ a large number of test 
 solutions of different strengths ; a measured volume, the 
 same in each experiment, is placed in the subject's mouth, 
 and he must immediately state what he feels. Numerous 
 blank experiments — the same volume of distilled water — 
 are interpolated, and care is taken that all the solutions shall 
 be at body temperature, a very important point. 
 
 Sternberg has introduced an ingenious method by the aid 
 of which stimuli can be minutely localised. This observer 
 finds that the vapour of ether when blown upon the tongue 
 produces a sensation of extreme bitterness, and that, in a 
 similar way, chloroform vapour is very sweet. A glance at 
 the diagram of his instrument (Fig. 1) will give the reader 
 an idea as to its practical use. 
 
 Fig. 1. — A and Z rubber tubes leading from two-way tap to glass-bulbs 
 C and D, which contain Ether and Chloroform respectively. By 
 means of the rubber-bulb E a stream of air can be directed along 
 E,A,C,l, or E, Z, D, 2, at will. 
 
 Of the three methods, that of Toulouse and Vaschide 
 is probably the most convenient for clinical work, and that 
 of Sternberg for localising experiments. 
 
 It is quite clear that the tongue, which is the organ 
 of taste, at least in a popular sense, is not equally sensitive 
 in its various parts for all forms of stimuli, 
 
 According to popular belief, bitter substances are best 
 tasted with the back of the tongue, and sweet ones with the 
 tip. Schreiber found that responsiveness to bitter stimuli 
 was restricted to the back and extreme borders of the tongue ; 
 the insensitive areas for salt and sweet comprise the antero- 
 medial portion of the dorsum linguae ; while the whole upper 
 
TASTE AND SMELL 37 
 
 isurface responds to acid except a very small portion rather 
 nearer the tip than the root. Within this latter region, of 
 course, no sensation can be elicited. 
 
 The actual measurements of the insensitive area are 
 variable, and it does not seem to exist in children. 
 
 With regard to the individual papillae, Oehrwall's original 
 work has not been superseded. He asserted that the filiform 
 papillae are quite insensitive. Oehrwall tested 125 papillae 
 (filiform, fungiform, and circumvallate) ; 27 did not react to 
 acid, bitter, or sweet ; 60 reacted to all three ; of the other 28. 
 some reacted to acid and sweet but not to bitter, others to 
 one form of stimulus only. The fact that Oehrwall found 
 that an electrode applied to, e.g., an acid-reacting papilla 
 called up a sensation of acid cannot be considered to prove 
 Mtiller's law for this sensory field. There is reason to think that 
 electrolytic decomposition of the saliva is responsible for the 
 result obtained. 
 
 There is no good evidence that sensations of taste can be 
 excited by mechanical stimulation (v, Kiesow) of the tongue. 
 
 In connection with taste, phenomena analogous to visual 
 contrast, both simultaneous and successive, have been 
 observed. After tasting weak sulphuric acid distilled water 
 seems sweetish, and a O'l per cent, solution of sodium chloride 
 increases the stimulus value of a weak glucose solution. 
 
 Simultaneous application of sugar and salt to opposite 
 borders of the tongue increases the effectiveness of both 
 stimuli. Some drugs affect particular sensations, a well- 
 known example being the case of GymneTiia sylvestre. If 
 the leaves of this plant be chewed, or if a sodium salt of 
 gymnecic acid be painted on the tongue, a very strong solu- 
 tion of sugar does not taste sweet nor quinine bitter. Salt 
 and acid stimulation are not however affected. These effects, 
 which pass off in a few minutes, are very striking. 
 
 Considerable attention has been paid to determining 
 whether chemical affinities exist between substances which 
 ■elicit the same sensation of taste. Wilhelm Stirnberg has 
 been especially active in this direction, but it cannot be said 
 that any very definite results have been obtained. 
 
 The path by which gustatory impulses reach the brain 
 
38 PHYSIOLOGY OF THE SPECIAL SENSES 
 
 has also been much studied, and the results are curiously 
 conflicting. 
 
 Taking the front of the tongue, it is certain that the 
 impulses first run in the lingual nerve and then in the 
 chorda tympani. Thus Lussana divided the lingual on one 
 side after its junction with the chorda and on the other side 
 ahove the junction. In the latter case, taste was unaffected. 
 Many cases of division of the chorda, or middle ear disease 
 affecting it, have been published, in which taste was abolished 
 over the front of the tongue. 
 
 The effect of stimulating the central end of the chorda 
 tympani has already been described (Chapter I.). 
 
 As we get higher up, our difficulties increase. Some 
 hold that the impulses travel by the pars intermedia of 
 the seventh nerve, others that they re-enter the fifth by way 
 of the Geniculate ganglion, great superficial petrosal, and 
 Vidian nerve. 
 
 The evidence on both sides seems to be definite, but 
 is really not so. For example, the fact that complete ex- 
 tirpation of the Gasserian ganglion, in the Hartley-Krause 
 operation for trigeminal neuralgia, has been unaccompanied 
 by loss of taste, would seem to settle the question. But 
 Vaschide has pointed out that what is removed in this 
 operation is not always the Gasserian ganglion. He cites an 
 amusing illustration.^ An eminent French surgeon ex- 
 hibited at the meeting of a learned society the Gasserian 
 ganglion excised by him for trigeminal neuralgia. He 
 was ill-advised enough to allow a pathologist who was 
 present to cut sections of the " ganglion," which was found 
 to contain neither nerve cells nor fibres. 
 
 I do not therefore think it necessary to insert the various 
 " proofs." 
 
 The course of the impulses from the back of the tongue 
 is also uncertain. That they first travel in the glosso- 
 pharyngeal is hardly doubtful, but it is possible that they 
 then reach the fifth by way of Jacobson's nerve and the 
 small superficial petrosal. 
 
 The balance of opinion seems, however, to incline in favour 
 
 ' Vaschide, art. " Gotlt," Riohet's Diet, de Physiol., vol. vii. p. 617. 
 
TASTE AND SMELL 39 
 
 of the glossopliaryngeal, of which nerve that of Jacobson 
 is, developmen tally, a branch. 
 
 It should be mentioned that some fibres from the 
 superior laryngeal branch of the vagus are distributed over 
 the root of the tongue. Their relation to taste is un- 
 known. 
 
 We have no knowledge as to the cortical representation 
 of taste, hardly even a conjecture worth repeating. There is 
 a tendency to pick out the anterior sylvian convolution as 
 the cortical " centre," but the evidence is trifling. 
 
 The relation of taste to smell will be discussed when 
 we have examined the latter sense. 
 
 Smell 
 
 Although both Aristotle and Plato were the authors of 
 ingenious speculations on the physical basis of smell, it 
 cannot be aflirmed that their work has had much influence 
 on the progress of physiological knowledge respecting this 
 sense. In general terms, it may be said that while the 
 number of exact observations on smell which have now been 
 accumulated is great, their interpretation is obscure. 
 
 The area of nasal mucosa associated with the perceptive 
 mechanism of the sense is smaller than used to be thought, 
 comprising a rectangular strip situated over the upper part 
 of the superior turbinated bone, together with a corresponding 
 patch upon the septum. 
 
 That the " olfactory cells " of the histologist are the end 
 organs is hardly doubtful, since the proximal extremities are 
 non-moduUated nerve fibres of the first cranial (olfactory) 
 pair. 
 
 Three methods of examination may be described. 
 
 Zwaardemaker's Olfcictometer 
 
 This consists of two concentric cylinders, the inner ter- 
 minating in a nose piece. The outer cylinder is lined with 
 some odorous substance, e.g. indiarubber or " hoof cement." 
 When the inner cylinder is pushed in level with the outer 
 
40 PHYSIOLOGY OF THE SPECIAL SENSES 
 
 one, air inhaled through the instrument does not pass over 
 the layer of indiarubber. 
 
 The further the inner cylinder is drawn out, the greater 
 the surface of indiarubber exposed to the in-drawn air. 
 
 (I 
 
 Fig. 2, 
 
 The extent to which it is necessary to withdraw the inner 
 cylinder before the odorous substance can be recognised is 
 a measure of responsiveness to that particular stimulus. 
 
 Method of Toulouse and Vaschide 
 
 A series of solutions containing camphor is prepared. 
 The solutions vary in strength from 1 in 1000 to 1 in 
 1,000,000. These are placed in flasks, each holding 10-15 c.c, 
 and having an opening of 17 mm. diameter. A similar 
 flask containing distilled water serves as a control. 
 
 Grazzi's Method 
 
 A square of blotting-paper, 5 cm^., is moistened with 10 
 drops of 20 per cent, benzoic acid in alcoholic solution. The 
 alcohol is allowed to evaporate, and the blotting-paper is 
 preserved. A series of cards are made which exactly cover 
 the square of blotting-paper, and in the centre of each 
 a hole is made. The diameter of the aperture varies from 
 0-5 to 5 cm. A card is placed over the blotting-paper, 
 and a little hollow cylinder, 30 by 5 cm., is applied to the 
 card, including the hole. If the patient cannot detect the 
 odour of benzoic acid, another card is used with a larger 
 hole in it, until the threshold is reached. 
 
 Of these different plans, that of Zwaardemaker has 
 yielded the best laboratory results, while the method of 
 Toulouse and Vaschide is more convenient for the testing 
 roughly of a large number of people. Grazzi's method is 
 
TASTE AND SMELL 41 
 
 not, so far as I know, largely used, and I cannot give 
 any details as to its practical applicability. The routine 
 investigation of the olfactory sense demands a certain 
 amount of patience and care on the part of the operator. 
 The methods generally adopted in the wards of hospitals 
 by clinical clerks and others are unlikely to yield any results 
 of value. The same remarks apply to the investigation of taste. 
 
 It would seem probable that in order to stimulate the 
 olfactory end organs, substances must be either gaseous or 
 sufficiently finely divided to be suspended in a gaseous 
 medium. 
 
 It is generally held that liquids are odourless. This 
 was first rendered probable by Weber, and more recently 
 by Haycraft. 
 
 "The person to be experimented on lies upon his back 
 on a bench with his head hanging downwards over the end 
 so that the olfactory region of the nose is lowest. Two 
 glass funnels are connected by tubing with a T junction, 
 and from the third limb of the T another tube passes into 
 one nostril; warm normal saline is allowed to flow from 
 one funnel until both nasal cavities are filled, and the air- 
 free solution falls upon the ground. This is then displaced 
 by warm-scented normal saline." ^ In this way Haycraft 
 could detect no odour in a 5-10 per cent, solution of eau 
 de Cologne. 
 
 Experiments of this kind have not escaped criticism. 
 
 Aronsohn has objected that strong solutions of sub- 
 stances, such as eau de Cologne, produce inflammation of 
 the nasal mucosa, and that the negative results depend on 
 the latter circumstance. This author has brought forward 
 some evidence to show that moderately strong solutions 
 of odoriferous bodies can be smelled, although it is not 
 certain that in his experiments all air was driven out of 
 the nostrils. It is, however, a fact in favour of Aronsohn's 
 view, that solutions of many salts which do not stimulate 
 the sense of smell under ordinary circumstances appear 
 to do so when the nostrils are filled with a solution. 
 
 ' Haycraft, Sekafer's Text-hook of Physiology, vol. ii. p. 1256. 
 
42 PHYSIOLOGY OF THE SPECIAL SENSES 
 
 There is reason to think that the stimulation of the true 
 olfactory end organs is best effected by moving particles, 
 and that in this way we can distinguish between " common " 
 and real olfactory sensation in the nose. 
 
 If one breathes deeply for a few minutes it is possible 
 to hold the breath comfortably for 80 to 120 seconds at least. 
 If a bottle of ammonia is unstoppered under the nose while 
 the breath is held, one experiences a tingling sensation, but 
 no distinct smell of ammonia. If after waiting a minute 
 one pinches the nose to imprison some of the vapour and 
 goes into another room containing no ammonia, at the first 
 inspiration through the nostril the smell of ammonia can 
 be detected. 
 
 Although the distinction between olfactory and "common" 
 sensation cannot be made absolute, the general conclusion 
 which may be drawn from this experiment appears to be 
 sound, and is supported by other pieces of evidence. Some 
 amusing experiments can be carried out with snuff. The 
 pungency of snuff is said to depend on the presence of free 
 ammonia and free nicotine, while in the process of manu- 
 facture various perfumes or " sauces " are added. 
 
 Of the varieties commonly sold, "Princes Mixture," 
 " Morton's Mixture," and " Kendal Brown " are scented, while 
 " Rappee " and Wilson's " S.P." are plain. If a pinch of snuft' 
 containing but little perfume be taken, it will be found that 
 the odour of the snuff is only detectable during the actual 
 inspiratory sniff', while the pungent sensation produced in 
 the nose persists when the breath is held, and may even not 
 appear before the "smell" has passed off. It may be remarked 
 that the pungency of the snuffs sold appears to vary with 
 the dryness and state of division. Thus " S.P.," which is 
 dry and finely divided, is extremely pungent, while " Brown 
 Rappee," which is coarse and somewhat moist, is far less so. 
 
 "The path taken by the inspired air stream through the 
 nostril is of some interest, and has been studied in various 
 ways. E. Paulsen divided the head of a cadaver in the 
 mesial plane and placed squares of red litmus paper over the 
 mucous membrane of the nasal cavity. The two halves were 
 then placed in apposition, and a stream of air saturated with 
 
TASTE AND SMELL 43 
 
 ammonia vapour was drawn through by a pump attached to 
 the trachea. In this way the path was mapped out by the 
 blue patches of litmus. 
 
 As a result of this and similar experiments it would seem 
 that no air is directly drawn over the olfactory mucous 
 membrane, but sweeps round below it. A partial displace- 
 ment must, however, be produced, and the olfactory "sniff" 
 probably sets up vortices. As Nagel points out, olfactory 
 stimulation through the posterior nares is more important, 
 biologically, to man than that set up by way of the anterior 
 nares. 
 
 A classification of the different specific sensations of smell 
 is impracticable, as has been tacitly recognised by mankind 
 at large. In no language do we find a distinctive vocabulary 
 for the objects of this sense. Similarly, the attempt to re- 
 duce the various forms of olfactory stimuli to terms of a few 
 components, despite the indefatigable researches of Zwaarde- 
 maker, has not led to any very satisfactory conclusions. 
 
 In the course of his work Zwaardemaker made many 
 interesting discoveries as to the antagonism between certain 
 odours. Thus iodoform may be antagonised by balsam of Peru, 
 and the odour of musk by that of bitter almonds. Such an 
 antagonism might be physical or chemical, but certain results 
 could hardly be explained in this way. For instance, Zwaar- 
 demaker, using a double olfactometer, i.e. two instruments of 
 the type already described, performed the following experi- 
 ment : — 
 
 The vapour of a 2 per cent, solution of acetic acid was 
 led into one nostril and that of a 1 per cent, solution of 
 ammonia into the other. If ammonia alone could be detected, 
 the amount of the acetic acid vapour passing to the other 
 nostril was increased, and conversely. Finally a point was 
 reached at which no odour was perceptible at all, although 
 either acting by itself produced a sensation. The following 
 are the chief antagonistic pairs from this point of view : — 
 
 Musk and Oil of Bitter Almond?. 
 Caoutchouc and Paraffin. 
 Ammonia and Acetic Acid. 
 Volatile Oils and Iodoform. 
 
44 PHYSIOLOGY OF THE SPECIAL SENSES 
 
 This is, perhaps, the only well authenticated example of 
 two stimuli, each separately effective, when simultaneously 
 applied failing to produce a response with which we are 
 acquainted in sense physiology. 
 
 As would be expected, many of the analogies to contrast 
 and fatigue already noticed in the case of gustatory stimula- 
 tion can be made out for smell. Thus after smelling 
 cumarin, a mixture of cumarin and vanilla only smells of 
 vanilla. After stimulation with tincture of iodine, alcohol 
 and copaiba balsam appear odourless ; some ethereal oils are 
 weakened, others not affected. Prolonged stimulation with 
 ammonium sulphide removes the responsiveness to sul- 
 phuretted hydrogen and bromine, but not to ethereal oils. 
 
 Normally, the delicacy of the sense of smell is variable. 
 Toulouse and Vaschide distinguish between the general and 
 specific threshold values for smell. The general threshold 
 value is the strength of solution which the subject recognises 
 not to be distilled water, but cannot tell exactly what it is. 
 The specific threshold value is the strength of the solution 
 which can be detected and recognised as a specific body. 
 
 Some of their main results were as follows (the fraction 
 represents the strength of the camphor solution used) : — 
 
 General Threshold. Specific Threshold. 
 
 33 males . . . 9/100,000 4/10,000 (mean values). 
 
 37 females . . 1/100,000 5/10,000 „ 
 
 Vaschide and Toulouse also tested 163 children from 
 Villejuif, dividing them into three groups : 3-5 years, 6 years, 
 12 years. The general threshold diminished steadily to the 
 age of 6 years and then increased. The specific threshold 
 diminished steadily. Ten dilutions, from 1/1,000,000 to 
 1/100,000, were submitted to ten children in series. All 
 the children recognised 5/1,000,000. 
 
 The possible sources of fallacy in work of this kind are 
 many, so that it is not surprising that different observers 
 obtain contradictory results, especially as they have not 
 employed any adequate statistical method of reducing their 
 observations. 
 
 There is no satisfactory evidence of any sexual difference 
 
TASTE AND SMELL 45 
 
 in olfactory acuteness. There is some reason to think that 
 smokers are less sensitive than non-smokers. It is not true 
 that the blind have an abnormally acute sense of smell 
 (Griesbach). 
 
 Minor abnormalities in the sense are common, and may 
 be due to disease — e.g. catarrh — or are congenital. Cloquet 
 reported the case of a person unable to smell vanilla ; Nagel 
 refers to an English chemist who could not detect any odour 
 in hydrocyanic acid ; while many persons, it is said, cannot 
 smell mignonette. Temporary anosmia (loss of smell) can be 
 induced by the local application of cocaine or morphia. On 
 the other hand, hyper-sensitivity, or hyperosmia, is not rare 
 in certain neurotic conditions, and can be induced by the 
 administration of certain drugs, of which strychnine is the 
 best known. 
 
 The inter-relationships of taste and smell are numerous, 
 and most of what is generally called taste must be referred to 
 olfactory stimulation. This needs no proof to any one who 
 has suffered from a cold in the head. It is not so generally 
 known that stimuli which only operate on one of the two 
 mechanisms may be antagonistic. Thus tinct, aurantii, 
 which has no real " taste," neutralises the bitter taste of 
 quinine, and effervescing drinks diminish the nauseous flavour 
 (olfactory stimulation) of castor oil. 
 
 The path of impulses to the brain would seem to be 
 involved in no uncertainty, but there is a little evidence that 
 olfactory impulses 'may reach the brain by channels other 
 than the olfactory nerves. 
 
 Magendie destroyed the olfactory nerves in a dog right 
 down to the lamina cribrosa. The animal recovered from 
 the operation, and Magendie presented it with several paper 
 parcels of the same size, some containing cheese and others 
 wood. The dog selected and unwrapped the packets of 
 cheese without hesitation. 
 
 Claude Bernard performed an autopsy on a womanof twentj-^ 
 nine in whom the olfactory bulbs were absent and the lamina 
 cribrosa imperforate. He traced out her friends, and found 
 that her sense of smell had been perfectly normal. She had 
 complained of people having smoked in her room and of a 
 
46 PHYSIOLOGY OF THE SPECIAL SENSES 
 
 neighbouring sewer. She had also been fond of smelling 
 flowers. Somewhat similar cases have been reported by 
 Lebeo and Heschl. While it is foolish and unscientific to 
 distrust evidence merely because it is old, we should bear in 
 mind that these cases were published a good many years ago, 
 and that, at the least, such a condition as they reveal is 
 excessively rare. 
 
 Finally, a few words must be devoted to the psychology 
 of smell. The relation of smell to memory is, as Professor 
 James remarks, one of the favourite topics of popular psycho- 
 logy. Oliver Wendell Holmes happily summarised one's 
 ordinary experience in the following words : " Memory, 
 imagination, old sentiments and associations are more readily 
 reached through the sense of smell than by almost any other 
 channel." Technically, this amounts to saying that smell is 
 an affective sense ; i.e. stimulation of its mechanism results 
 rather in a general alteration of feeling-tone of the general 
 atmosphere, as it were, of consciousness than in any distinctly 
 localised or delimited impression. An illustration of this 
 is that we do not project our olfactory sensations to any 
 great extent, if at all. In consequence, olfactory impressions 
 tend to be associated with a sum-total of feeling-tone and 
 with other impressions which also affect the conscious 
 atmosphere as a whole. It is accordingly easy to understand 
 why, in many animals and some men, olfactory stimulation 
 is associated with sexual excitement. 
 
 Bunge, among others, has urged that the sense 6i smell 
 is protective in character, and points out that noxious gases, 
 e.g. sulphuretted hydrogen, are malodorous. The objections 
 to this are of comparatively little weight, in my opinion. 
 It is perfectly true that some highly poisonous gases, such 
 as carbon monoxide, have no smell, but it seems to me that 
 few or none of them are of frequent occurrence in a state 
 of nature. It is a tempting speculation — but no more than 
 a speculation — that all gases originally excited the end organs 
 of smell; that natural selection favoured those animals 
 whose end organs reacted most strongly to toxic gases; 
 and that panmixia eliminated gradually responsiveness 
 to normal atmospheric constituents, or that responsive- 
 
TASTE AND SMELL 47 
 
 ness to the latter was correlated with some harmful 
 variation. 
 
 The cortical representation of smell is doubtful; some 
 results suggest the hippocampal region, but the connections 
 of the olfactory tract are numerous and complicated, while 
 experimental work on lower animals is beset with serious 
 difficulties. 
 
 The delicacy of the receptive mechanism in such animals 
 as the dog must be exquisite, since the blunted sense of man 
 can detect some substances {e.g. mercaptan) in a dilution 
 of one in fifty thousand millions. ! 
 
 WORKS RECOMMENDED FOR FURTHER STUDY 
 
 Taste— 
 
 L. Marchand, Le Gout, Paris (Biblio. de Psychol. Exper.), 1903, 
 
 pp. 328. A full account with literature to date of publication. 
 W. Nagel, art. Geschmackssinn, Nagel's Handb., Bd. iii. p. 621. A 
 
 very readable sketch. 
 N. Vaschide, art. Gout, Richet's Diet, de Phys., vol. vii. p. 571. 
 
 The most complete account in existence. A very full and fairly 
 
 critical bibliography. 
 W. Sternberg, Geschmack und Geruch, Berlin (Springer), 1906, p. 149. 
 
 Contains some interesting remarks on technique. 
 
 Smell — 
 
 R. Zwaardemaker, Physiologie des Geruchs, Leipzig, 1895. The 
 
 standard treatise. 
 H. Zwaardemaker, art. Geschmack, Ergebnisse d. Phys. Jahrb., 1903, 
 
 2nd Abth., pp. 699, etc. 
 W. Nagel, art. Gerchssinn, Handb. d. Phys., vol. iii. p. 589. Thoroughly 
 
 good. 
 (The best accounts in the English language are by Professor J. B. 
 Hayoraf t in the second volume of Schafer's Text-book of Physiology. ) 
 
CHAPTER VI 
 
 THE SENSE OF , POSITION AND MOVEMENT 
 
 In the next two chapters I shall discuss a division of our 
 subject which is difficult to arrange in orderly fashion, but of 
 extreme interest. 
 
 The ordinary experience of life teaches us that in health 
 we possess the faculty of poising our bodies in space with 
 great precision ; we are also capable of adjusting the several 
 parts, in particular the limbs, to the performance of duties 
 requiring a minute accuracy of localisation. 
 
 These powers of general and local equilibration show 
 themselves to a marked extent at a very early period of 
 life ; indeed great as are the effects of special training and 
 experience, late acquirements are small in comparison with 
 the powers developed as a normal event in the process of 
 growth. The difference between the most sJiilful of dancers 
 and the ordinary child o£ six is not so vast as that separating 
 the normal child and one of the same age who, either through 
 disease or accident, cannot walk. 
 
 Extending our survey to other members of the animal 
 kingdom, we see in all but the most primitive phylla equally 
 remarkable powers of equilibration. The problem is to 
 determine what mechanism or mechanisms are essential for 
 the performance of these functions. 
 
 It is at once obvious that we can make an elementary dis- 
 tinction between the efferent and afferent processes involved. 
 Evidently both general and local equilibration demand an 
 efficient muscular system and an efficient efferent nervous 
 system through which impulses can reach the muscles or 
 corresponding executive organs. This side of the problem 
 does not require any discussion. When we turn to the 
 afferent and receptive side of the matter, however, difficulties 
 begin to accumulate. In the first place, is it necessary to 
 
THE SENSE OF POSITION" AND MOVEMENT 49 
 
 suppose that there are any special paths or receptive struc- 
 tures dedicated to the sense of equilibration, whether local 
 or general, or can we account for the facts by way of the 
 actual or stored (memorised) impressions which result from 
 the stimulation of other sense organs, those of touch, smell, 
 hearing, and vision? Now it is quite probable that many 
 facts akin to those just mentioned can be accounted for 
 quite simply. An example is the sense of direction in 
 pigeons, and — although this is less clear — the homing flight 
 of bees can perhaps be regarded as another illustration. 
 Exnej's observations on pigeons are so interesting that I 
 shall describe them. Pigeons were placed in a basket covered 
 with a black cloth, and the basket was shaken or made to 
 turn on its axis during the whole outward journey. They 
 were released 37'7 kilometres from Vienna, separated from 
 that city by mountains and in a spot which their owner 
 believed they had never before visited. A second batch 
 were treated in the same way, except that the basket con- 
 taining them was not rotated. The pigeons were released 
 one after the other, not one before its predecessor has dis- 
 appeared. All reached home safely. In a second experi- 
 ment galvanism was tried. Four pigeons were used, two 
 old and two young ones, one old and one young pigeon 
 being galvanised and rotated ; the distance was 54 kilometres. 
 The old galvanised pigeon reached home in 3 hours 32 
 minutes, the old normal pigeon two days later. Neither 
 of the young pigeons got home. These experiments evidently 
 suggest that visual impressions are, as Forel holds, adequate 
 to account for the facts. Exner also tried to find out whether 
 the pigeons acquire any information on their outward joprney. 
 He anaesthetised two pigeons during the whole of a journey 
 of 43 kilometres, and then left them four days at Oberhol- 
 labrun separated by hills from Vienna; they were then re- 
 leased together with a control pigeon. Of the three, one of 
 the anaesthetised pigeons reached home in 4 hours 20 minutes, 
 the other two birds being lost. In another experiment, over 
 a less distance, two old and two young birds were used, one 
 of each age being anaesthetised. Both the old birds got 
 home, the anaesthetised one first ; the young ones were lost. 
 
 D 
 
50 PHYSIOLOGY OF THE SPECIAL SENSES 
 
 Forel in describing these experiments remarks : " I am 
 astonished that Exner does not draw from such clear results 
 the simple conclusion that it is the experience and the 
 knowledge of places by vision that orients pigeons." ^ 
 
 The quotation illustrates the somewhat trenchant fashion 
 in which Forel settles rather complicated matters, but it 
 does appear that the results are consistent with his view, 
 although, as we shall see later, that is good reason for 
 thinking that the pigeon has quite special equilibrating 
 organs. It may be remarked that the term orientation is 
 hardly used in connection with this class of experiment in 
 the same sense as that employed by physiologists. There 
 is a distinction, not however always easy to draw, between 
 orientation and power of direction. 
 
 Forel summarises his views in the following terms: 
 " The faculty of orientation outside the body itself of 
 the individual rests neither in a special mysterious force 
 nor in an unknown sense, static, geotropic, or other, nor in a 
 sixth sense, nor in the semicircular canals. It is the result 
 of the experience of known senses, combined or not, especially 
 of sight and smell, according to the case and the species. In 
 aerial orientation it is vision which most predominates. 
 The case of the carrier-pigeon constitutes the most sur- 
 prising case of aerial orientation. Since it is explained 
 by vision, it is -idle to search in the domain of other 
 mysterious causes. In terrestrial orientation the sense of 
 smell often plays a predominant part, but gives place to 
 sight in many animals, among which are man, monkeys, 
 arboreal reptiles, certain insects, etc. In the orientation of 
 subterranean and cave-dwelling animals, smell and touch 
 reign as masters. In spiders it is touch which is the 
 principal orienting sense." ^ 
 
 There seems to me little doubt that Forel's assertions are 
 too sweeping. As Bonnier has noticed, we can trace in 
 the animal kingdom a series of organs consisting when 
 reduced to the simplest terms of a solid particle suspended 
 
 ^ The Senses of Insects, by A. Forel, trans, by M. Yearsley. London, 
 Methuen, n.d., p. 214. 
 " Forel, op. cit., p. 242. 
 
THE SENSE OF POSITION AND MOVEMENT 51 
 
 in a liquid. The simplest type, found in certain jelly-fish 
 (^^ig- 3), is a tentacle which has lost its suppleness 
 and mobility and acquired a considerable freedom of 
 inertia. We then come to a modification consisting of 
 calcareous particles (otolithic organ, Fig. 4) suspended more 
 or less freely. Eventually we reach a degree of specialisation 
 at which the otolith is lodged in a separate cavity con- 
 taining a fluid distinct from that which bathes the animal 
 as a whole (Fig. 5). Numerous experiments, some of which 
 will be described, tend to prove that the homologues of these 
 organs in the higher mammals and in birds are really 
 connected with the faculty of equilibration ; it is tempting 
 
 Fig. 3, 
 
 Fig. 5. 
 
 to postulate a similar function in the case of animals, ex- 
 periments upon which are difficult both to plan and 
 interpret. 
 
 In the case of man, while it may be freely admitted that 
 the other senses, especially that of sight, generally modify 
 and can, to some extent, replace the special sensory organs of 
 equilibration, the importance of the latter are unquestionable. 
 Many writers distinguish carefully between the equilibrating 
 sense and the perception of muscular movement. Such a 
 distinction looks well enough on paper, but hardly corre- 
 sponds to anything found by experiment. I shall therefore 
 not attempt any strict classification, but describe the best 
 ascertained facts, subsequently touching on the more im- 
 portant theories which have been proposed to cover them. 
 
 Let us first consider sensations of position referred to a 
 vertical axis. For these investigations the subject is fastened 
 in the dorsal decubitus, on a plank capable of rotation round 
 
52 PHYSIOLOGY OF THE SPECIAL SENSES 
 
 Fia. 6. — Diagram of Plank Bed capable 
 of Kotation Round a Horizontal 
 Axis. 
 
 a horizontal axis (Fig. 6). The subject's eyes must be closed 
 and care taken to minimise the information derivable from 
 cutaneous sensations due to sliding or changed pressure on the 
 soles of the feet as the plank is rotated. The subject's know- 
 ledge of his position is tested 
 in a simple way. After he 
 has been rotated a certain 
 amount, he is made to hold 
 a light rod in what he thinks 
 to be a position at right angles 
 to the true horizontal — i.e. 
 vertically. 
 
 Delage and others have 
 found that rotation through 
 an angle of 60 degrees is cor- 
 rectlyestimated. Lessdegrees 
 of rotation are under- and 
 greater amounts over-estimated. Bending the head forwards 
 increases the error up to 60° and diminishes it between 60° 
 and 90°. Analogous results are obtained when rotation is 
 made round a sagittal axis. As before, the position of the 
 head is important. Thus Sachs and Meller found that if 
 the head be fixed vertically and the body inclined, the in- 
 clination is over-estimated ; if the body be vertical and the 
 head inclined, the opposite result is obtained. 
 
 Next as to the relation between the sense of position and 
 optical orientation. If a straight line be ruled on a vertical 
 screen, its inclination to the true vertical can be estimated 
 with considerable accuracy. Roughly, an inclination of 1 
 degree in a line 40 centimetres long seen from a distance of 
 2 metres can be detected. If the head be rotated, or the 
 whole body, on a sagittal axis, a really vertical line appears 
 to be bent in a direction opposite to that of the body move- 
 ment (Sachs and Meller, Aubert). The presence in the field of 
 vision of objects known to be vertical causes the disappearance 
 of this and kindred illusions. If, on the other hand, the eyes 
 be directed to a horizontal plane, e.g. the ceiling, when lying on 
 the back, judgment of direction is far less efficient. Generally, 
 the assumed principal horizontal axis of reference, with 
 
THE SENSE OF POSITION AND MOVEMENT 53 
 
 respect to which inclinations are judged, coincides with the 
 long axis of the body, but such concepts are far less accurate 
 than those of position in a vertical plane. Evidently these 
 experiments suggest that the equilibrating organ is influenced 
 by gravity. 
 
 When the body is moved without altering its objective rela- 
 tion to the vertical, various illusions are produced. It is 
 well known that from an express train moving round a curve 
 houses bordering the line appear bent towards the carriages. 
 The interpretation of this is, however, not free from difficulty, 
 and less equivocal observations have been made. In rota- 
 tion round a vertical axis, in an experimental modification 
 of a "merry-go-round," the true vertical is judged to be 
 inclined, and a plummet which has taken up a position due 
 to the resultant of the two forces, the vertical component of 
 gravity and the angular horizontal rotatory force, is judged 
 to be really vertical ; for instance, a pendulum will take up 
 a position depending on the magnitude of the resultant of 
 these two forces and appear to be vertical (Purkinje). The 
 bearing of these results and their agreement with those 
 detailed in the last paragraph is clear. It is also noteworthy 
 that the illusion is not found in many deaf mutes (Kreidl). 
 These are the main data regarding the sense of position as 
 a whole. We next turn to the sense of position of separate 
 parts of the body. 
 
 It- would seem that normal persons form more accurate 
 judgments as to the position of parts of the body which can 
 be seen than of those, judgments regarding which are, uncon- 
 trollable by visual impressions. Thus, while our judgment 
 as to the degree of flexion or extension of the fingers is good, 
 we are often at fault respecting the toes, and still oftener in 
 the case of the tongue. This is not, however, entirely a 
 matter of visual memory, since the blind form quite accurate 
 conclusions as to the attitude of their fingers. It is also not 
 a matter of cutaneous sensibility, since cutaneous aneesthesia 
 does not abolish the power. Deep anaesthesia, as in a 
 patient thoroughly investigated by Strtlmpell, abolishes the 
 sense of local position almost entirely. 
 
 Sensations of movement of the whole body have been 
 
54 PHYSIOLOGY OF THE SPECIAL SENSES 
 
 examined in detail. For the experiments to be pure, it is 
 of course necessary that the subject should remain quite 
 passive. Roughly speaking, we can classify the experimental 
 results into those dealing with rectilinear and curvilinear 
 motion. In the former case, uniform motion produces no 
 sensation, only negative or positive acceleration ; in the 
 latter class an illusion of motion in the opposite direction 
 may be produced. In ordinary life the presence in con- 
 sciousness of the effects of other sensory stimuli hinders the 
 development of distinct illusions, but even during uniform 
 motion a change in the relative positions of head and trunk 
 gives rise to a sensation of movement. It is also found that 
 the subject's axis must be inclined in a direction determined 
 by compounding the force producing rotation with the 
 vertical component of gravity in order that he may appear 
 to be vertical. If rotated in the true vertical position, the 
 subject will think his body is inclined. These illusions are 
 of course particular cases of those discussed in the previous 
 section. 
 
 Another example is the observation of Yves Delage, that in 
 swinging the inclination of the plane in which the swing is 
 moving is only correctly appreciated when the head is vertical ; 
 if the latter be inclined to the right, the plane appears tilted 
 to the left. The results of experiments on the movement of 
 individual parts are precise enough to be formulated under 
 four headings : — 
 
 1. The sense of movement is not necessarily greatest in 
 those parts the position of which is most accurately judged. 
 The movements of the toes are as well appreciated as those 
 of the fingers — I speak of course of passive movements. 
 
 2. The sensibility partly depends on a succession of tactile 
 excitations, partly on afferent impulses conducted by the 
 centripetal fibres of muscles, tendons and joints. 
 
 3. Practice is markedly important. Loeb, Ostermann, and 
 others have shown that in right-handed persons if the right 
 and left hands be passively moved through the same dis- 
 tance, the subject invariably fancies the left hand has exe- 
 cuted a greater movement. An opposite result is obtained 
 in left-handed people. 
 
THE SENSE OF POSITION AND MOVEMENT 55 
 
 4. Cutaneous anaesthesia does not abolish the appreciation 
 of active movement, but the passage of a faradic current 
 through the joint at which movement occurs wholly or to 
 a large extent disturbs the accuracy of judgment (Gold- 
 scheider). 
 
 Akin to but not identical with the sense of local move- 
 ment is that of resistance. That the sensation produced 
 by pushing against a stiff indiarubber ball is not merely a 
 mixture of tactile and " movement " impulses is shown by the 
 facts that such sensations are produced when no movement 
 takes place, and also when the skin is anaesthetic. Gold- 
 scheider's paradoxical experiment is another proof. If one 
 lowers a weight attached to a string until it touches the 
 ground, the sensation does not disappear immediately after 
 contact with the floor, but one has the illusion of the string 
 being replaced by a solid rod. This is perhaps due to the 
 maintenance for a fraction of time of tension in muscle, 
 tendon, and string. 
 
 The old idea that sensations of movement are " innerva- 
 tion sensations," a sort of automatic register of outflowing 
 " motor energy," cannot be maintained. The evidence both 
 of experiments and clinical observations demonstrates that an 
 intact centripetal nexus is a conditio sine qua non of the sense. 
 At least two sets of end organs can reasonably be regarded 
 as the chief sensory elements in the process, viz. the muscle 
 and tendon end organs. Perhaps the joint end organs 
 are somewhat more important, because (1) movement is 
 almost always localised in the joint itself, and (2) faradisation 
 of the joint much dulls the sense. It cannot, however, be 
 said that either consideration is decisive. That the sensation 
 of resistance is an affair of tension seems probable from 
 the Goldscheider paradox ; perhaps the tension stimulates 
 mechanically the end organs associated with the centripetal 
 fibres connected with tendon bundles. I do not think any 
 very conclusive evidence exists as to these points. 
 
 We now come to the practically interesting questions of 
 vertigo and the so-called rotation reflexes. Vertigo is a term 
 used to describe two different phenomena, viz. the sensation 
 produced by looking down from a great height, and an illusion 
 
56 PHYSIOLOGY OF THE SPECIAL SENSES 
 
 of movement either as occurring in the subject himself or 
 in surrounding objects. Vertigo in the latter sense alone 
 concerns us here. 
 
 Although the definition anticipates the results of ex- 
 periments which I have not yet described, we can best 
 regard vertigo as that sensory condition which is induced 
 when the information derived by consciousness from the 
 labyrinthine organs is in conflict with that obtained through 
 other sensory channels. All or any of the following ante- 
 cedents may produce vertigo : — 
 
 (1) Rotation of the body. 
 
 (2) Movement in a curved path, especially rapid 
 oscillation. 
 
 (3) Passage of a galvanic current through the head. 
 
 (4) Local cerebral injury, especially abscess, 
 
 (5) Cerebral anaemia, as in ordinary syncope. 
 
 (6) Certain poisons, particularly alcohol and tobacco. 
 
 (7) Any interference with binocular vision. 
 
 (8) Conflict of visual and other sense impressions. 
 
 (9) Disease of the internal ear. 
 This list could easily be extended. 
 
 With reference to rotatory vertigo, it is clear that the 
 direction of apparent movement depends upon the posi- 
 tion of the head during the previous rotation. The reader 
 can easily satisfy himself of the truth of this statement 
 by turning round sharply a few times with the head in 
 diff"erent positions. If during the rotation the head be held 
 in the ordinary erect posture, "things go round" in a 
 horizontal plane; if the head be dorsi-flexed and then 
 brought back into the vertical at the end of rotation, 
 objects turn in a vertical plane. 
 
 Rotatory vertigo is not present in many deaf mutes. 
 James found that of 519 deaf mutes tested, 186 ex- 
 perienced no rotatory vertigo, while this was present in 100 
 per cent, of 200 normal persons similarly examined. It 
 is unnecessary to examine the other forms of vertigo in 
 detail; they all depend, as a little reflection will convince 
 the reader, on a conflict between the afferent impulses arising 
 in our various intelligence departments. 
 
THE SENSE OF POSITION AND MOVEMENT 57 
 
 Among the correlates of a change in the bodily posture, 
 rotatory movements of the eye-ball have attracted special 
 notice. The table contains some of the observations made 
 recently by W. Nagel. 
 
 NAGEL'S EXPERIMENTS 
 
 Angular Ro- ^ 
 tation of 
 Head J 
 
 Compensa-"] 
 tory Rota- 1 
 tion of I 
 Eye-ball J 
 
 lount of"l 
 Jompen- }- 
 lation J 
 
 Amount of"| 
 C( 
 sation 
 
 10° 
 
 20° 
 
 30° 
 
 40° 
 
 50° 
 
 60° 
 
 70° 
 
 80° 
 
 90° 
 
 100° 
 
 1-3° 
 
 3-8° 
 
 5-2° 
 
 5-4° 
 
 6-3° 
 
 6-7° 
 
 6-8° 
 
 8-0° 
 
 8a° 
 
 8-6° 
 
 1 
 7-7 
 
 1 
 5-2 
 
 1 
 
 5-8 
 
 1 
 7-4 
 
 1 
 
 7-9 
 
 1 
 9 
 
 1 
 10-3 
 
 1 
 10 
 
 1 
 
 11-1 
 
 1 
 11-8 
 
 It would appear from the experiments of Delage that the 
 amount of compensatory rotation is not the same for equal 
 degrees of rotation to right and left. A similar process 
 occurs when the head is rotated round a transverse axis. 
 These eye movements are closelyassociatedwith the horizontal 
 nystagmus which is observed when the body is rotated. 
 
 I have now described the chief general experiments dealing 
 with the sense of position and movement; the more specialised 
 researches which are concerned with theories of equilibration 
 will be examined in the next chapter. 
 
 BOOKS RECOMMENDED FOR FURTHER STUDY 
 
 A. Forel, The Senses of Insects, translated by M. Yearsley. London, 
 Methuen. 
 
 P. Bonnier, L'Audition. Paris, 1901, Doin. 
 
 W. Nagel, Die Lage-Bewegungs- und Wiederstandsempfindungen, 
 Nagel's Handb. der Physiol., Bd. iii. p. Y34. Braunschweig, 1905. (This 
 memoir gives sufficient indications to enable the reader to consult the 
 enormous literature of the subject.) 
 
CHAPTER VII 
 
 THE SENSE OF POSITION AND MOVEMENT {Concluded) 
 
 Floueens, in 1828, first announced that injury to the semi- 
 circular canals is associated with disturbances of equilibration. 
 Numerous observers have confirmed and extended his 
 results, so that we possess a wealth of details respecting the 
 changes which follow experimental interference with any 
 part of the labyrinth, although theoretical interpretation is 
 not in all cases beyond dispute. I do not propose to re- 
 capitulate more than a few of the chief researches; the 
 works cited at the end of this chapter will help the student 
 to a wider knowledge of this interesting field of observation. 
 It may be said that no branch of experimental physiology 
 has been more fruitful in producing new and ingenious 
 methods of investigation, and that research in this direction 
 is only profitable in the hands of those gifted with special 
 aptitude for delicate manipulations. 
 
 Ewald found that extirpation of the labyrinth in pigeons 
 produced muscular weakness, some uncertainty in move- 
 ment — but not necessarily ataxia — and abolished the power 
 of flight. One-sided removal caused frequent rotation of the 
 head towards the operated side and irregularity in progres-. 
 sion, but the power of i3ight remained. There was also 
 unilateral muscular weakness. In frogs, muscular weakness 
 is associated with bilateral ablation of the labyrinth (Schrader). 
 After unilateral destruction, the muscular weakness is not 
 marked, but there is rotation of the head towards the operated 
 side and a tendency to curvilinear springing. In all animals 
 investigated, compensatory eye movements cease to occur 
 after destruction of the labyrinth. The exact relation of the 
 labyrinth to general muscular tone is not, for obvious reasons, 
 well understood. 
 
 The results of experiments on individual canals are quite 
 
 58 
 
THE SENSE OF POSITION AND MOVEMENT 59 
 
 definite. Unilateral destruction of a single canal produces 
 little or no effect, but when the corresponding canals of both 
 sides are destroyed — e.g. in doves — pendular movements of 
 the head in the plane of the affected canals generally occur. 
 If one canal be stimulated in the avipullary region, the head 
 moves in the opposite direction (Ewald, Breuer, Lee), but 
 this does not occur if the ampulla be anaesthetised with 
 cocaine before stimulation. Lee carried out exhaustive ex- 
 periments with the dog-fish. He ascertained that if the fish 
 were passively rotated round different axes, compensating 
 movements of the eye-ball and fins took place. If the fish 
 were rotated in the plane of a semicircular canal, exactly the 
 same movements of the eye-ball followed as were produced 
 by stimulating the canal in question. The same physiologist 
 divided the animal's auditory nerve on one side and noticed 
 rolling movement round the longitudinal axis towards that 
 side. If both nerves were cut, no power of static equilibrium 
 was retained, and the fish would lie indifferently on its back, 
 belly, or side, and swim in these positions. He stimulated the 
 ampulla of the anterior semicircular canal and noticed an 
 upward rotation of the eye on the same side and a down- 
 ward rotation on the opposite side. If the nerve were divided, 
 the opposite result was obtained; if the nerves of both 
 anterior ampullse were divided, the fish dived downwards. 
 On section of the nerves of both posterior ampullse the fish 
 swam upwards, sometimes putting its head out of water. 
 These results hardly leave room for doubt as to the parti- 
 cipation of the canals and labyrinth in the equilibration 
 process. 
 
 Clinical evidence, on the whole, tends to support the con- 
 clusions drawn from experimental work, but, as seems gene- 
 rally true for sense physiology, the observations are not 
 definite enough to have great importance attached to them. 
 Most of the cases are connected with the symptom complex 
 known as Meniere's diseases and the pathology of deaf mutism. 
 In the former condition, leading characteristics of which are 
 vertigo and defective equilibration, disease of the labyrinth 
 and canals has often been found. In deaf mutism, some 
 good work has been done. Myguid found in 118 autopsies 
 
60 PHYSIOLOGY OF THE SPECIAL SENSES 
 
 on deaf mutes that 80 exhibited pathological changes in some 
 part of the labyrinth ; of these, 40 per cent, were cochlear 
 and vestibular anomalies, and in 56 per cent, the semicircular 
 canals were affected. It is a suggestive fact that disorders 
 of equilibration are not invariably noticed in deaf mutes. 
 Kreidl found that the normal horizontal nystagmus observed 
 during rotation round a vertical axis was absent in 50 out of 
 109 cases of deaf mutism examined. The results of James 
 were mentioned in the last chapter. 
 
 It only remains to sum up the attempts which have been 
 made to resume in a general theory the facts of experiment 
 and observation. Goltz originally suggested the hypothesis, 
 that the endolymph excited afferent nerve endings by pressing 
 more strongly on different parts of the labyrinth in accordance 
 with the situation of the head. Breuer, however, pointed out 
 that this explanation could only hold if the membraneous 
 labyrinth were suspended in a compressible medium, such as 
 air, and that it actuall}'' floats in the incompressible perilymph. 
 The theory which resumes a majority of the facts is that due 
 to Breuer, Mach, and Crum Brown. 
 
 Mach schematised the essential conditions of the problem 
 in the following way : " Suppose in a body B there is a 
 cavity on the walls of which there are nerve endings, and that 
 this cavity contains a solid or liquid A. By its weight A will 
 exercise a greater pressure on one part of the walls of the cavity 
 than on the others, and it will thus determine the position 
 of the body B relative to the vertical. With each acceleration 
 of B, A will press in the opposite direction, and this contra+ 
 pressure will be added to the acceleration due to weight, so 
 that the direction of the pressure as well as its intensity will 
 change in the cavity. In like manner, with each angular 
 acceleration communicated to B, A will oppose a rotation in 
 the opposite direction. Thus B will obtain knowledge both 
 of its position and of its progressive acceleration in a straight 
 line, and, in the case of rotation, angular acceleration will 
 also be indicated. The vestibule and semicircular canals in 
 Mach's scheme constitute B, the vestibule possibly having to 
 do with the sense of acceleration of movement in a straight 
 line, while the semicircular canals serve for angular accelera- 
 
THE SENSE OF POSITION AND MOVEMENT 61 
 
 tion. Each excitation produced in this manner . . . will give 
 rise to a sense of rotation." '■ 
 
 Or, the matter can be looked at in a slightly different way. 
 Suppose the three canals full of liquid, a rotation of the head 
 in one direction will cause currents of fluid in the opposite 
 direction owing to the liquid's inertia, and the amount of 
 flow in each canal will depend on the plane of rotation. The 
 crista of each ampulla is furnished with cells possessing hairs 
 or cilia, and these cells are in close relation with afferent nerve 
 fibres. It is not difficult to imagine that a back current in 
 the endolymph will excite the hair cells, and through them 
 the afferent fibres. In' some such fashion might arise a 
 sensation of movement in the plane of the canal and in a 
 direction opposite to that of the actual movement. If the 
 movement continue, friction will soon arrest the eddy ; but 
 if motion be suddenly stopped, the liquid will tend, for the 
 previous reason, to continue flowing in the direction of former 
 movement, and another sensation will be induced. The several 
 hypotheses of Mach, Breuer, and Crum Brown differ in minor 
 points, but are all of the above type. 
 
 The actual dimensions of the semicircular canals are such 
 that no actual streaming of the liquid can be supposed to 
 occur. This does not in any way invalidate the reasoning 
 on which the Mach-Breuer theory is based, since there must 
 evidently at each positive and negative acceleration be a 
 change in the existing fluid pressure relations. In general 
 terms, one may conclude that the Mach-Breuer theory de- 
 scribes satisfactorily the chief experimental data. Thus we 
 should expect to be able to perceive accelerations, but not be 
 conscious of uniform movement. Further, if any bending of 
 the ampuUary hairs in one direction call up a sensation of 
 movement in the opposite direction — this is, of course, an 
 assumption — we can understand the genesis of movement 
 in the opposite direction which is associated with negative 
 acceleration. 
 
 Of workers who decline to accept the hydrodynamic 
 theory of Mach, Breuer, and Crum Brown, the most industrious 
 
 ^ J. G. McKendrick and A. A. Gray, Schdfer's Text-hook of Physiology, 
 Edinburgh, 1900, vol. ii. pp. 1200-1. 
 
62 PHYSIOLOGY OF THE SPECIAL SENSES 
 
 is Cyon. According to Cyon, the planes of the semicircular 
 canals form a system of physiological co-ordinates to which 
 we refer all our notions of space. Illusions of position which 
 occur when the head is inclined are due to the altered position 
 of these planes of reference. Cyon has written at much length 
 on these questions, and his methods of controversy, which 
 not infrequently take the form of offensive personal attacks 
 on those who have the temerity to differ from him, are not 
 calculated to secure unprejudiced consideration of his views. 
 Nagel writes : " The way in which Cyon supposes the system 
 of canals to act as a source of the conception of space is 
 not, as I must frankly admit, intelligible to me; for this 
 reason I am compelled to abandon a closer investigation of 
 his theory — if one can use that term." ^ I confess to being in 
 the same state of mental confusion as Professor Nagel. 
 
 The sense of position as distinct from movement is perhaps 
 referable to the end organ of the utricle. The sensory 
 epithelium of the utricle differs from that of the semicircular 
 canals in having a mass of small solid particles (otoliths) ap- 
 plied to the hairs. It is clear that with varying degrees of 
 rotation of the head the extent to which the hair cells are 
 pulled on by the otoliths will be altered; it is also evident 
 that so long as the new position is maintained, the stimulation 
 will endure. This mechanism therefore, unlike that of the 
 semicircular canals, provides for a long-continued excitation, 
 and would be a basis for the origin of general sensations of 
 position. The theory, although by no means free from ob- 
 jection, is perhaps the best at our command. It is also pos- 
 sible that the utricular end organ gives rise to sensations at 
 the beginning and end of motion in a straight line, but here 
 we are on very unsafe ground, and the question must still 
 be regarded as quite an open one. 
 
 ' Nagel's Ilandb. d. Physiol., vol. iii. p. 804. 
 
THE SENSE OF POSITION" AND MOVEMENT 63 
 
 BOOKS AND PAPERS RECOMMENDED FOR 
 FURTHER STUDY 
 
 Flourens, Mem. Acad. d. Sciences de Tlnstit. de France, 1828, ix. 5. 
 
 Lee, Journ. Physiology, xvii. 192. 
 
 Mach, Grundlinien der Lehre von den Bewegungsempfindungen. 
 
 Leipzig, 1875. 
 Crum Brown, Journ. Physiology, viii. 327. 
 Breuer, Med. Jahrb., p. 72. Wien, 1874-5. 
 *Nagel, Handb. d. Physiol., Bd. iii. p. 734. 
 *McKendrieh and Gray, Schiifer's Text-book of Physiology, vol. iii. 
 
 1194. 
 * Sherrington, ibidem, p. 1002. 
 
CHAPTER YIII 
 
 HEARING— HISTORICAL SKETCH 
 
 In previous chapters of this book I have described the 
 sense organs from the point of view of a modern physiologist. 
 References to the thought and work of a remote antiquity 
 have been scanty, and the reader may well have been en- 
 couraged in the common and erroneous idea that the modern 
 student can safely disregard that which was believed in 
 olden times. My excuse is that information respecting the 
 sense organs so far considered requires for its arrangement in 
 proper historical perspective a wealth of space and mastery 
 of antiquarian detail far exceeding my means. In the case of 
 hearing and vision, this excuse will not serve. Professor J. I. 
 Beare has, in his masterly treatise,^ conveyed so intelligible 
 an idea of the psycho-physiology of hearing and sight as in- 
 terpreted by the Greek philosophers, that an ignorance of 
 their language can no longer be pleaded in defence of 
 neglecting their matter. I shall in this chapter summarise 
 certain conclusions at which we may arrive in regard to 
 this historical branch of psycho-physiology. The time is 
 approaching when the student of natural science will be as 
 ashamed to confess ignorance of the history of his subject 
 as any educated man is to admit that he knows nothing of 
 the history of his country. 
 
 The earliest writer as to whom we have explicit informa- 
 tion is Alcmseon of Cretona. His view was that the air 
 (vacuum) of the intra-tympanic region was the intermediary 
 of hearing. This air catches up and reverberates an external 
 sound, the reverberations being conveyed to the "point 
 of sense," or interpreting structure, which is the brain. 
 Alcmseon had got to the stage of recognising something 
 
 ' Greek Theories of Elementary Cognition from Alcmceon to Aristotle, by 
 John I. Beare, Oxford, 1906, pp. 354. 
 
HEARING— HISTORICAL SKETCH 65 
 
 internal, an air-filled cavity, as a physiologically necessary 
 organ. Empedocles went further. According to him, just 
 as the organ of vision contains a lantern, the organ of 
 hearing contains a bell or gong which is rung by the outer 
 sound. Presumably this ringing is conveyed in some fashion 
 to the "point of sense," and the perception of sound is 
 awakened. Exactly what this gong-like body corresponds to 
 is not certain; scholars differ in their translations of the 
 word used, but the best opinion seems to be that something 
 inside the ear is meant. 
 
 The obviously weak point of the theory was clearly 
 noticed by Theophrastus, who observes : " Empedocles ex- 
 plains hearing by stating that it is due to intra-aural sounds. 
 But it is strange of him to suppose that he has made it self- 
 evident how we hear by merely stating this theory of a 
 sound, as of a gong, within the ear. For suppose that we 
 hear the outer sounds by means of this gong ; by what do we 
 hear the gong itself when it rings ? For this — the very point 
 of the whole inquiry — is neglected by him." ^ 
 
 It is a far cry to the days of Theophrastus, but succeeding 
 ages have not produced a theory of hearing which will pass 
 through his criticism unscathed. 
 
 The theory of Democritus has little physiological im- 
 portance. He regarded hearing as a mode of contact be- 
 tween the atoms of sound and the soul atoms of the body. 
 
 Anaxagoras, again, did not advance the matter to any 
 extent. The following sentences epitomise his theory: 
 "Anaxagoras held that ^wvr] (vocal sound) is produced by 
 the breath (or air in motion) which collides against the fixed, 
 solid air, and, by a recoil from the shock, is borne onwards to 
 the organs of hearing, just as what is called an " echo " is 
 produced." 
 
 The views of Plato are somewhat difficult to summarise. 
 There is no evidence that he was acquainted with the 
 physiological anatomy of the ear, nor is it certain that 
 Archer-Hind is correct in attributing to him a theory of 
 the physical basis of pitch which approximates to that 
 
 ' Beare, op. cU., p. 97. ^ Ibid., p. 104. 
 
 E 
 
66 PHYSIOLOGY OF THE SPECIAL SENSES 
 
 now taught. '■' He defines vocal sound as (on its physical 
 side) air in motion, impelled from the seat of intelligence, 
 through the mouth, and (as physiological stimulus of hearing) 
 a shock caused by the air, through the ears, to the brain and 
 blood, propagated to the soul." ^ 
 
 Plato, like his predecessors, seems to have believed that 
 air was actually projected from the sounding body to the ear. 
 It is, in Beare's opinion, doubtful whether the word KivqdeKi 
 can fairly be translated to mean " vibrations," since the con- 
 ception of air as an elastic vibrant medium, in the modern 
 sense, cannot safely be attributed to any of the Greek writers. 
 
 Many scholars, for instance Trendelenberg, maintain that 
 a portion of Aristotle's writings on sound has been lost. 
 Even from his undoubted works, however, we can con- 
 struct a valuable account of the sense of hearing. In 
 considering the physiological psychology of the senses as 
 expounded by Aristotle, it is necessary, to bear in mind 
 certain leading principles. For Aristotle, the subject-matter 
 of any sensory mechanism can always be regarded from two 
 points of view, the actual and potential. In making this 
 distinction, Aristoile often comes near to the standpoint 
 of the modern physiologist who accepts Mliller's law as, at 
 least, a true principle. It will be seen that this dualism 
 is predicated both of the physical " cause " or stimulus and 
 of the physiological or psycho-physiological recipient. Be- 
 ginning with sound, the "object" of hearing, a distinction 
 is made between sound in general, including noises, and 
 articulate sound which, although strictly applied to vocal 
 sound, can be extended to any musical sounds, however 
 produced. In sound, taken generally, we have its actual 
 and potential aspects ; some things, e.g. wool, cannot be made 
 to produce sound, and are both actually and potentially sound- 
 less. Others, e.g. brass, are potentially sonant even when 
 not actually giving rise to sound. "As it is possible for a 
 person possessing the faculty of hearing not to hear actually 
 at some given moment, so a thing may have the property of 
 sounding without actually doing this. When, however, that 
 
 ' Beare, op. oil., p. 107. 
 
HEARING— HISTORICAL SKETCH 67 
 
 ■which can hear realises its potentiality, and also when that 
 which can sound does sound, then the realised faculty of 
 hearing and the realised sound both concur; so that the 
 former may properly be named ' actual hearing,' and the 
 latter ' actual sounding.' " ^ The actualisation of potential 
 sound depends on the existence of a medium, so that the 
 Avhole process requires the co-operation of three things — an 
 object, a receiver, and a medium. In the case of land animals 
 the medium is air, and a local movement of it is essential. 
 If there be a general or total movement, the necessary con- 
 dition is not fulfilled, and no sound is produced. A boat 
 on a river which moves because that in which it is fixed 
 moves, produces no sound. The organ of hearing, the 
 receiver, consists of air physically homogeneous with the 
 external air, its peculiarity being that it is confined in a 
 chamber so as to have a proper motion of its own. " Thus 
 it has a peculiar resonance like a horn ; and this, while it 
 lasts, is a sign that the auditory faculty is unimpaired. . . . 
 We can hear to some extent under water, because the water 
 does not enter the air-chamber of the ear. If it did so, 
 hearing would be at an end. Hearing ceases to be possible 
 if the tympanic membrane is injured, just as blindness ensues 
 if the membrane covering the eye is injured. As the water- 
 holding eye is joined with the watery brain, so the air-holding 
 ear is connected with the air-holding hinder part of the 
 cranium." ^ Sound, physically considered, is a movement 
 of something, and, unlike light, travels through the air; 
 this is proved by the fact that we see a blow struck some 
 time before we hear the sound.^ Articulate sounds are due 
 to the conformation of the air in motion, and are less well 
 heard at a great distance owing to a blurring of this 
 conformation. Pitch differences are only potential in the 
 sound as a physical event, but are actualised in the sound 
 heard. " The sharp is that which moves the sense much 
 
 1 Beare, op. cit., p. 112. " Ibid., p. 115. 
 
 ^ " Sed tonitrum fit uti post auribus accipiamus, 
 
 Fulgere quam cernant oculi, quia semper ad auris 
 Tardius adveniunt quam visum quse moveaut res." 
 
 • — Lucretius, vi. 1G4. 
 
68 PHYSIOLOGY OF THE SPECIAL SENSES 
 
 in a little time, the grave that which moves it little in 
 much time." There is, however, a physical substratum for 
 this, in so far that grave sounds correspond to slow and 
 sharp sounds to rapid motion. It may be observed that 
 Aristotle probably understood by motion actual translation 
 of masses of air ; we must not read into his statement our 
 modern notions of pitch. Turning to the physiological 
 anatomy of the ear, we have but vague indications as to 
 Aristotle's knowledge of the matter. " One (viz. the inner) 
 part of the ear is nameless, the other is called the 'lobe.' 
 The whole consists of cartilage and flesh. Inwardly its for- 
 mation is like that of spiral shells, the bone at the inner 
 extremity (into which, as last receiver, sound comes) being 
 in shape like the (outer) ear. This inner ear has no passage 
 into the brain, but it has one to the palate, and a vein extends 
 into it from the brain." ^ It would seem, from this passage, 
 that Aristotle had examined the cochlea, but regarded the 
 middle ear and cochlea as forming together one chamber. 
 
 A discussion of the biological and intellectual importance 
 of hearing relatively to the other senses does not mu«h 
 concern sense physiology. The teaching of Aristotle in 
 these respects might stand with little modification to-day. 
 He points out that although the primacy among the sense 
 belongs to vision if we regard the mere preservation of life, 
 yet, for intellectual development, the sense of hearing is 
 more important. This is because words are in their nature 
 general, and to be regarded as the signs of ideas. " The 
 impressions of sight, on the other hand, are primarily of the 
 nature of particulars, and appeal rather to the individual. 
 Those received from \6yo<i through the sense of hearing are, 
 almost from the first, of the nature of universals, and therefore 
 almost directly {i.e. so far as we understand them) stimulate 
 the faculty of intelligence." ^ Aristotle, of course, pointed 
 out that these results are secondary, that the ivninediate 
 data of hearing, like those of every other sense, necessarily 
 consist of particulars. A full discussion of these and kindred 
 matters . would be out of place in a physiological book, a 
 
 1 Beare, op. cit., p. 122. ^ Ibidem, p, 124. 
 
HEARING— HISTORICAL SKETCH 69 
 
 remark which also applies to the subject of Greek harmonics. 
 The above sketch will be sufficient to enable the reader 
 to realise the importance and value of Greek physiological 
 psychology. I must once more impress on him the neces- 
 sity of examining his subject historically as well as ex- 
 perimentally. 
 
 RECOMMENDED FOR FURTHER STUDY 
 
 An adequate guide to the literature will be found in the work 
 o£ Professor Beare, from which I have so frequently quoted. 
 
CHAPTER IX 
 
 PHYSIOLOGY OF THE EAR 
 
 Sound, whether considered from the physical, physiological, 
 or psychological side, has been the object of such close study 
 in the last two hundred years, that our knowledge of many 
 important phenomena is relatively minute. It is my inten- 
 tion, however, to deal with the problem in an even less 
 detailed manner than has been the case with the other 
 divisions of sense physiology. In my opinion, founded on 
 the experience of some years' teaching, many of the most 
 interesting researches on tonal analysis, the resonator theory 
 of the cochlea, etc., can only be made intelligible to those 
 having received a wider preliminary training in physical 
 acoustics than I may assume to be the case with the reader 
 of this book. It would, of course, be possible to supply this 
 defect by writing a physical introduction to the subject, but 
 to do so would force me to exceed reasonable limits of space 
 and time. Not only am I convinced of the necessity of 
 taking this course, but further, it happens to be one which 
 can be followed with the least inconvenience to the English 
 reader. Most of the classical writers on vision, for instance, 
 can only be consulted by those who have been wise enough 
 to acquire a knowledge of French and German. Helmholtz's 
 great treatise on the physiology of the ear, a work which, 
 altogether apart from the question as to how far the theories 
 it promulgates may be correct, must be the starting point of 
 a course of reading on the subject, has been translated into 
 English. To this book, and to the others mentioned at the 
 end of the chapter, the reader's attention is particularly 
 directed ; a perusal of this sketch cannot, and will not, give 
 him an adequate idea of the profoundly interesting dis- 
 coveries which have been the fruit of investigation. 
 
PHYSIOLOGY OF THE EAR 71 
 
 From the physical standpoint, simple tones are pendular 
 vibrations of the sounding medium, and their graphical repre- 
 sentation corresponds to that exhibiting the change in value 
 of the sine of an angle as the angle varies continuously. 
 Such tones can difier one from another in at least two respects, 
 viz. (1) the number of vibrations in a unit of time, (2) the 
 amplitude of such vibrations. For reasons which will be 
 more particularly referred to below, it is usually stated that, 
 sounds differing in the first way differ in 'pitch, and if the 
 change be of the other kind, in intensity. In order that the 
 ear may perceive a given note, the pitch and amplitude must 
 lie within certain limits. The determination of the upper 
 and lower boundaries of pitch is a matter of difficulty. Most 
 of the earlier workers were unable to exclude the possibility 
 of their results being vitiated by the presence of impure tones. 
 Another point is that while, in the middle of the scale, over- 
 tones are feebler than fundamentals, in the lower region the 
 reverse relation holds. There is some reason to think that 
 vibrations at the rate of 15 per second cause a perceptible 
 note. A somewhat analogous uncertainty holds with regard 
 to the upper limit of pitch perception. Edelmann, using an 
 improved form of the well-known Galton whistle, obtained 
 an audible note having a frequency of 50,000 vibrations per 
 second. The objection to such experiments is the fact that 
 Galton's whistle does not appear to furnish a note of constant 
 intensity. 
 
 We all know that tones in different regions of the scale 
 possess a different character, or, to adopt a familiar German 
 expression, tone colour. It has been laid down that, other 
 things being equal, the intensities of two notes of different 
 pitch are directly proportional to the kinetic energies of the 
 respective moving masses.^ Helmholtz, however, showed 
 with a syren apparatus,^ that when the same amount of 
 mechanical work was expended in the production of two notes 
 of high and low pitch respectively, the former appeared to be 
 the more intense. Now, the kinetic energy or, what comes to 
 the same thing, the mechanical work is proportional to {an)^. 
 
 ' Helmholtz, Tonempfindungen, English trans., p. 75. ' Ibid., p. 174. 
 
72 PHYSIOLOGY OF THE SPECIAL SENSES 
 
 where a is the amplitude, and n the number of vibrations in a 
 unit of time. If this is to remain constant, n must decrease 
 as a increases, hence when a is kept constant, the vahie will 
 increase with n, and when two notes have the same amplitude 
 but different pitches, the higher pitched note will be the 
 louder. An objection to this method of proceeding is that 
 loudness is subjective, and that it is very diflScult to compare 
 with respect to this character two notes of quite different 
 pitches. 
 
 The determination of the minimum amplitude of vibra- 
 tion corresponding to an audible note is a matter of practical 
 importance, as it affords a criterion of the normality, or 
 otherwise, of the subject's hearing powers, and is largely used 
 by aural surgeons. Many methods are in use, of which 
 that originally proposed by v. Conta appears to be the 
 most satisfactory. It has been shown by Wead that the 
 
 kinetic energy of a vibrating tuning-fork is equal to -^ . a', 
 
 where I is its length, d its thickness, b the breadth of the 
 prongs, a the amplitude of vibration, and E Young's 
 modulus for steel. When such a fork has been set in 
 vibration, its energy falls off in a manner dependent solely 
 on the amplitude. Thus it can be shown, by reasoning 
 based upon Laplace's equation for small oscillations, that 
 if at any instant the amplitude is a„, t seconds later the 
 
 amplitude is a„^(, where e is the Napierian base, and h is 
 a constant. 
 
 Now, writing Wead's equation in the simple form 
 V = Ya"^, tj, sees, is after the fork has been started V = F {a^-'"pf. 
 
 A certain fraction, say -, enters the auditory meatus of the 
 
 subject. Therefore, if he can just hear the fork, his thres- 
 
 F 
 hold intensity Ip=- (a.„e""''')2. 
 
 Now, supposing a deaf man can still hear the fork after 
 the shorter period t^. Then his threshold intensity Iq is 
 
 -(«„e"")^ But if both normal and deaf listened under 
 
 approximately the same physical conditions, n = n.,, hence 
 
PHYSIOLOGY OF THE EAR 73 
 
 j^ = ^-^^3^''^2 = e"^'"?-"'»', or the deaf person's auditory acuity 
 
 j-=e-^''i'p-'!'x J-. Of these quantities, h can be ascer- 
 tained by experiment for any given fork, tp and ty are 
 found in the examination of the subject and control. It 
 is apparent that to render the process at all accurate, 
 many normal persons must be examined under approxi- 
 mately the same physical conditions, and that a tuning- 
 fork is not an altogether satisfactory instrument. Marrage, 
 three or four years ago, introduced a syren for testing 
 purposes, but I do not know whether it has, in practice, 
 proved more useful than the tuning-fork method. In any 
 experiment it is of course necessary to exclude so far as 
 possible all extraneous sounds, and it is of some interest 
 to know whether all such sounds are equally disturbing. 
 The experiments of Mayer, partly confirmed by Stumpf, 
 suggest that low pitched notes produce, within certain 
 limits of intensity, more marked interference in the per- 
 ception of high notes than do the latter on the percep- 
 tion of the former. These disturbing influences naturally 
 become more important as we approach the limen. Absolute 
 silence, indeed, cannot be realised, and the absolute threshold 
 value cannot be measured. When all external noises are 
 hushed, we hear the beating of our arteries and the sound 
 of our breath, etc. In order that a note may be perceived 
 at all, not only must it possess a certain vibration frequency 
 and a certain intensity, but it must endure for a finite 
 time. This can be experimentally studied by means of an 
 apparatus so arranged that the note of, e.g., a tuning-fork 
 is allowed to reach the ear through a tube which can be 
 automatically opened and closed very quickly. Numerous 
 workers have tackled this problem, and it seems that two 
 vibrations of a tone vibrating at the rate of 72 per second are 
 sufficient to set up a corresponding tone sensation (Pfaundler, 
 1878). Abrahams and Brlihl, who employed a syren, assert 
 that while two vibrations are sufficient in the middle of the 
 scale, the necessary number increases steadily as the pitch 
 rises. This is, of course, exactly what one would expect. 
 Passing to the subject of compound notes, only a few 
 
74 PHYSIOLOGY OF THE SPECIAL SENSES 
 
 sentences can be devoted to the physics of the matter. A 
 compound note, e.g. a simple harmony, or the air move- 
 ment produced when a violin string is bowed, consists of a 
 fusion together of the effects produced by each individual 
 component. This fusion is of such a nature that the 
 individual constituents can be recovered from the mixture 
 either (1) Experimentally ; thus if three tuning-forks vibrat- 
 ing at the rates of 100, 200, and 400 per second respectively 
 blend to form a single compound movement of the air, then 
 three tuning-forks having the same vibration frequencies 
 as those of the blending instruments set up on resonance 
 boxes in the neighbourhood will each be set in motion. 
 This is called sympathetic resonance. (2) If the form of 
 the wave motion representing the compound tone is given, 
 it is always possible to analyse this compound into simple 
 constituents. This is called harmonic analysis, and its 
 development is owing to the investigations of the mathe- 
 matical physicist Fourier. 
 
 One of the chief objects of modern work has been to 
 ascertain whether the facts of tone sensation suggest or 
 are consistent with the existence of some physiological 
 mechanism which can analyse sounds, as is done by a 
 collection of tuning-forks or resonators in the presence 
 of a compound note. An important limitation must at 
 once be observed. Suppose we could show that those 
 characters of a sound which we hear are precisely the ones 
 which produce a physical eifect on resonators, we should 
 still have advanced only a little way in the physiological 
 and not at all in the psychological description of the^ 
 process of hearing. We should still have to answer, and 
 fail to answer, the question of Theophrastus, that the 
 gong may be rung by the outer sound, but what is it that 
 hears the gong ? This is of course recognised by all in- 
 vestigators ; the reason why so much labour has been devoted 
 to finding analogies between the process of sound reception 
 and sympathetic resonance is, that some such method of 
 reception would afford the simplest explanation for the 
 elaborate structures of the internal ear which, all agree, 
 are intimately connected with the process. 
 
PHYSIOLOGY OF THE EAR 75 
 
 Bearing these important but frequently forgotten facts 
 in mind, we may sum up the results of modern work in 
 the following terms: — The analysis of a compound note 
 which is effected by a set of resonators can also be carried 
 out by the human ear, although the power of doing so 
 and the attention requisite to distinguish the partials yary 
 much from subject to subject. The fundamental work 
 was done by Helmholtz in 1859. Helmholtz selected for 
 study the vowel sounds of human speech because they 
 can be produced as evenly sustained musical notes. While 
 the accuracy of these conclusions has been generally ad- 
 mitted, some discussion has arisen as to whether any tonal 
 phenomenon can be detected by the ear which does not 
 affect a set of resonators. If one thinks for a moment 
 of what happens in a sonant medium, for instance the air, 
 when a sound is produced, one notices that there is a regular 
 production of condensations and rarefactions graphically 
 represented by the crests and troughs of the "waves" 
 of a sine curve or on. the figure traced out by a piece of 
 stiff paper attached to the prong of a tuning-fork and 
 writing on a uniformly rotated drum. Supposing one takes 
 a tuning-fork vibrating at the rate of 100 per second, there 
 will be in each second one hundred such crests and the 
 same number of troughs. If now a second tuning-fork, 
 also vibrating at the rate of 100 per set, be thrown into 
 vibration a two hundredth of a second later than the first, 
 then the crest of a wave in the first set of vibrations will 
 always be accompanied by a trough in the second set and 
 vice versa, so that if the two sets of vibrations have the 
 same amplitude of vibration, the rarefaction phase of one 
 will be counteracted by the condensation phase of the 
 other, no movement of the sounding medium will be pro- 
 duced, and no sound will be heard. The two waves are 
 said to be in a state of phase opposition, and the result is 
 said to be a case of total interference. If instead of two 
 forks of the same vibration rate set in motion consecutively 
 we take two forks, one vibrating at 100 and the other at 
 101 per second, and start them together, then once, and 
 only once, in each second the trough of one wave system 
 
76 PHYSIOLOGY OF THE SPECIAL SENSES 
 
 exactly corresponds to the crest of the other set, and the 
 result is that a weakening in or cessation of the even flow 
 of the sound is produced once a second. This phenomenon 
 is called a beat. 
 
 Now with regard to the phase of a note, some physiological 
 difEeulty has arisen. It is, so far as a resonator is concerned, 
 quite immaterial whether aerial motion which excites it 
 begins by a phase of condensation or by one of rarefaction, 
 but it is not equally clear that the ear cannot distinguish 
 the two cases. In simpler language, it is doubtful whether 
 the ear can or cannot distinguish between a push or a 
 pull applied to the tympanic membrane. As a result of 
 certain experiments by Lord Kelvin, experiments which 
 are too complicated to be described here, it used to be 
 taught that the ear can distinguish differences of phase. 
 Recent opinion has not on the whole tended to accept 
 this view. Lindig, for instance, has apparently demonstrated 
 that Lord Kelvin's results cannot be applied to pure 
 harmonies, that in such no alteration of timbre is produced 
 by total reversal of phase relations. 
 
 Another phenomenon which appeared to be inexplicable 
 on the basis of sympathetic resonance was that of beat 
 tones. If the difference in frequency of two forks be con- 
 tinuously increased by filing one of them, the number of 
 beats of course increases steadily. It was said that when 
 the difference in frequency is sufficiently great, these beats 
 themselves fuse into a note which is perceptible to us but 
 does not excite a physical resonator. The researches of 
 Helmholtz,^ Schaefer, Hermann, and others seem to have 
 demonstrated that these so-called beat tones are really 
 examples of combination tones occurring under special 
 circumstances. The genesis of a combination tone is 
 theoretically simple enough. If two forks having vibration 
 rates of 200 and 400 per second respectively are set in motion, 
 in addition to the compound wave generated by their fusion 
 
 * Helmholtz's views will be found on pp. 156 et seq., op. cit. A 
 mathematical investigation is given at p. 411. Tlie older literature is 
 collated at pp. 527-535, Beoent work is discussed by Schaefer, Nagel, 
 vol. iii. pp. 552 et seq. 
 
PHYSIOLOGY OF THE EAR 77 
 
 a component is found which has a vibration rate equal to 
 their difference, namely, 200 per second, and another with a 
 frequency equal to their sum, namely, 600 per second. The 
 former is called a difference and the latter a summation tone, 
 and the whole class of such tones is termed a group of 
 combination tones. Now such notes correspond to physical 
 realities, and can, under certain conditions, excite resonators. 
 It is thought that the so-called beat tones are akin to these 
 combination tones, and that their reception by the auditory 
 apparatus does not require us to postulate the existence of 
 a mechanism different from that of a sympathetic resonator. 
 I speak, however, with much reserve on this point. The 
 subject of combination tones is a very difficult one to master. 
 I am not at all sure that I understand the true bearing of 
 the work which has been done, and am quite sure that my 
 knowledge of physical acoustics is far too elementary to allow 
 me to make the subject clear to my readers. I must therefore 
 be content with the remark that, so far as one can judge, 
 there is not at present any good reason for holdiDg that tones 
 which do not or cannot affect a physical resonator affect the 
 receptive apparatus of the inner ear. 
 
 We must now, having coiisidered the facts in a general 
 way, examine briefly the physiological anatomy of the ear. 
 Right up to the internal ear the meaning of the various 
 structures is fairly easy to understand ; at that point we 
 become a prey to theories. 
 
 That the auricle acts as a reflector, directing the waves 
 into the external meatus, is a reasonable supposition. Some 
 two centuries ago Boerhaave attempted to prove that the 
 pinna behaved like a parabolic reflector, but this was shown 
 long ago by Savart to be erroneous. A method of investi- 
 gation is to sound a compound note and observe whether its 
 quality changes in accordance with the position of the source of 
 sound in relation to the meatus. Sylvanus Thompson pointed 
 out that the tick of a watch has a slightly different quality 
 when the watch is held in the middle line behind the head 
 from what obtains when it is held in the middle line and the 
 same distance in front of the ears. The exact cause of this 
 is, however, doubtful, and we know — from the experience of 
 
78 PHYSIOLOGY OF THE SPECIAL SENSES 
 
 times when clipping the ears was a recognised method of 
 poUtical argument — that removal of the pinna is a more 
 serious detriment to one's personal appearance than to the 
 sense of hearing. The external auditory meatus may also, 
 in virtue of its curvature, possess some slight value as a 
 resonator in the case of high pitched notes, but its pro- 
 tective function is clearly more important. This protective 
 function is exercised in virtue of two characters, (1) the 
 varying curvature, (2) the secretion of cerumen. The 
 curvature renders it difficult for foreign bodies to impinge 
 directly upon the membrana tympani, and the aromatic 
 odour and bitter taste of the cerumen probably deter insects 
 from taking up their abode in an inconvenient situation. 
 The movements of the auricle are, in man, of little or no 
 importance to the sense of hearing. The membrana tympani, 
 as the reader will remember, is composed of two sets of fibres 
 circularly and radially arranged, the former making an 
 external ring. The tension of the membrane is maintained 
 by the tensor tympani, the tendon of which is inserted into 
 the handle of the malleus a little below the horizontal axis 
 of rotation of that bone, so that it must exert an almost 
 constant degree of traction on the membrane. The form of 
 the membrana is conical, but the walls of the hollow cone are 
 convex outwards, owing to the disposition of the central 
 fibres. Had the membrana been a uniformly stretched mem- 
 brane like a drum-head, the amplitude of its movements must 
 have steadily increased from within outwards, but the peculiar 
 structure avoids this, together with the weight of the 
 malleus ; in addition to this, the arrangement of the radial 
 fibres affords a special functional advantage. These fibres, 
 as we have seen, are slightly convex externally, but their 
 curvature is very slight. Helmholtz^ pointed out that the 
 curvature being very slight, small variations of it would be 
 associated with relatively large excursions of the centre of 
 the fibre in a plane at right angles to their direction. The 
 consequence is that very slight changes in air pressure in 
 
 1 Helmholtz, Tonempfindunr/en, etc., English trans., p. 135. The proof is 
 given more fully by McKendriok and Gray (Sohafcr's Text-booh of Physiology, 
 vol. ii. pp. 1151-1156). 
 
PHYSIOLOGY OF THE EAR 79 
 
 the meatus will produce relatively great effects on the trans- 
 mitting mechanism, the membrana tympani ; that is to say, 
 this special form of membrane secures the maximum degree 
 of effect to sounds of feeble intensity. The action of the 
 intrinsic muscles of the middle ear now deserves attention. 
 We saw that the tensor tympani is so inserted that it main- 
 tains a constant tension in the tympanic membrane when 
 the muscle is uncontracted or in a constant state of tone. 
 When the iimscle contracts it necessarily renders the mem- 
 brane more tense, and is generally believed to damp loud 
 notes by diminishing the excursion of the membrana. Some 
 few persons can contract the muscle voluntarily, but generally 
 the contractions are reflex, the arc being auditory nerve- 
 brain -trigeminal nerve. It is certain that the contractions 
 raise the iritra-labyrinthine pressure, but the assumed 
 damping effect is not satisfactorily proved. The action of 
 the stapedius is still more doubtful. Henle was of opinion 
 that the muscle held the plate of the stapes firmly in place, 
 and prevented it being forced into the fenestra ovalis when 
 strong pressures force in the membrana tympani. Pollitzer 
 has given a different interpretation; according to him, the 
 stapedius is to be regarded as an antagonist of the tensor 
 tympani, drawing the foot-plate of the stapes out of the 
 fenestra ovalis and, by releasing the tension of the chain of 
 ossicles, producing relaxation of the tympanic membrane. 
 It is asserted, but clinical experience is not quite concordant 
 on the point, that paralysis of the facial nerve caused by 
 pressure in the aqueduct of Fallopius is associated with 
 increased sensitiveness to loud sounds. 
 
 The three auditory ossicles act roughly as a bent lever 
 of the second order. The fulcrum is the horizontal axis of 
 rotation of the malleus, the handle of the latter is the long 
 arm, and the long process of the incus carrying the stapes is 
 the short arm. The ratio of the arms is effectively 1'5 : 1, so 
 that there is an increase in force and diminution in ampli- 
 tude of movement. This arrangement is satisfactory when 
 one remembers the far greater extent of surface presented by 
 the membrana tympani than that of the membrane covering 
 the fenestra ovalis. That most perceptible sounds are con- 
 
80 PHYSIOLOGY OF THE SPECIAL SENSES 
 
 veyed to the internal ear by this mechanism is shown by the 
 serious effects on hearing accompanying disease of the middle 
 ear, even when the internal ear is unaffected. Bezold, how- 
 ever, has made it probable that very high notes may be 
 transmitted to the internal ear even when the ossicles are 
 completely ankylosed together. Only one other structure 
 in the middle ear needs mention, the Eustachian tube. The 
 prime function of this canal is to enable the air pressures 
 within and without the tympanic cavity to be equalised, but 
 it is, as Maeh and Kessel pointed out, of importance that the 
 opening should be normally closed, since otherwise, owing to 
 the fact that all parts of the head are affected by the wave of 
 sound, the excursions of the tympanic membrane would be 
 too much damped, it being, as it were, simultaneously pushed 
 on both sides when the Eustachian tube is opened. The 
 importance of the power of opening the tube for normal hear- 
 ing is well shown by the effect on this sense of exposure to 
 compressed or rarefied air. It may be remarked that under 
 all high pressures the mere act of swallowing is not, in 
 most cases, sufficient to open the tube; it is necessary to 
 expire sharply with mouth and nostrils closed. 
 
 We now come to the structures of the internal ear, and 
 the student who is not already familiar with the very com- 
 plicated apparatus therein contained should consult one of 
 the better text-books of histology or anatomy, since space only 
 permits a brief reference to the matter here. The cochlea, 
 which is undoubtedly an essential structure for hearing, is a 
 convoluted tube wrapped round a central pillar, the modiolus, 
 and divided into two cavities, the upper scala vestibuli and 
 the lower scala tympani, by the lamina spiralis. The scala 
 tympani abuts on the middle ear by the fenestra rotunda, 
 which is covered by a membrane. The patency of this 
 foramen is important, since if its wall becomes rigid the 
 practical incompressibility of liquids makes it impossible for 
 the stapes to move. The lamina spiralis is incomplete at the 
 apex of the cochlea, where the two scalae communicate by 
 way of the helicotreme. A portion of the scala vestibuli is 
 cut off by a special membrane, Reissner's membrane, and 
 ends blindly below, communicating by a side channel with 
 
PHYSIOLOGY OF THE EAE 81 
 
 the saccule. This section of the scala vestibuli is called the 
 canalis cochlese, and contains a series of complicated struc- 
 tures, consisting of (1) peculiarly shaped cells with hair-like 
 processes, arranged in two rows; (2) tiny rigid structures 
 arranged in pairs, the rods of Corti ; (3) cells thought to be 
 sustentacular ; (4) a tectorial and a basement membrane, of 
 which the latter is composed of fibres, the length of which 
 varies continuously from apex to base of the cochlea, and 
 which is closely associated with the ends of the cochlear 
 division of the auditory nerves, the nerve endings being also 
 in very close relation with the hair cells standing on the base- 
 ment membrane. The structure of the cochlea varies in its 
 fine details in different animals, and has been much elucidated 
 by the beautiful preparations of Gray, whose work should be 
 consulted. 
 
 The exact meaning of these elaborate structures has been 
 much debated, and nothing like agreement as yet prevails ; 
 of the various modes of action suggested, the most ingenious 
 and, in spite of various defects, the most plausible is that 
 propounded by Helmholtz.^ In the opinion of Helmholtz, 
 the cochlea should be regarded as a vast congeries of re- 
 sonators, composed of separate vibrators, each tuned to 
 correspond to a single vibration period, and altogether con- 
 taining enough vibrators to respond to all notes within the 
 range of audibility, say from 15 to 50,000 vibrations per 
 second. Each of these vibrators 2 is capable of exciting a 
 nerve filament which physiologically "corresponds" to it, 
 so that a nervous impulse " corresponding to " {not identical 
 with, in any physical sense) the frequency of the vibrator is 
 transmitted to the brain. The mass of each vibrator is such 
 that its inertia is slight ; it is easily set in motion, and easily 
 brought to rest again. Damping arrangements exist in the 
 cochlea which will quickly extinguish movements of the 
 vibrating stmctures. The method of action is taken to be 
 somewhat as follows. When a simple tone falls on the ear, 
 there is a pendular movement of the base of the stapes and 
 
 1 Op. cit., pp. 142-152. 
 
 2 It is known that structures somewhat analogous with the hair cells of 
 the cochlea, found in Crustacea, vibrate sympathetically. 
 
 F 
 
82 PHYSIOLOGY OF THE SPECIAL SENSES 
 
 the structures in the internal ear are set in motion, but any 
 particular structure, the vibration period of which corre- 
 sponds to that of the exciting tone, will be specially affected. 
 A particular nerve filament is thus stimulated, a particular 
 impulse is conveyed, and there, by processes utterly unknown 
 to us, the sensation of a tone of definite pitch passes into 
 consciousness. The loudness of the sound is supposed to be 
 conditioned by the amplitude of movement of the cochlear 
 resonator. When a compound wave of pressure is com- 
 municated to the structure of the inner ear it is analysed 
 by the vibrators, the periods of which correspond to its 
 partials, and all the impulses separated by the individual 
 resonators pass to the brain as an imperfectly fused whole, 
 which can by an effort of attention be again resolved into 
 constituents, these psychic constituents " corresponding " to 
 the separate excitations. It may be admitted that the 
 cochlear structure affords, so far as the number of units 
 which might be conceived to resonate, an anatomical basis 
 for the theory, provided one takes the fibres of the basilar 
 membrane to be the vibrators. There is also some experi- 
 mental evidence consistent with the hypothesis. Exner and 
 Pollak have recently shown that periodic phase variations 
 {vide supra) affect the intensity of tonal sensations, just as 
 they affect the intensity of response in sympathetic resonators. 
 Numerous experiments have been performed on animals in 
 which specified portions of the cochlea have been destroyed 
 and deafness to certain notes has been asserted to occur. 
 Deafness to certain parts of the scale, tonal lacunse, has 
 been frequently observed in man, and, in a few cases, post- 
 mortem examination has revealed pathological changes in 
 certain portions only of the cochlea. I believe this evidence 
 is of little value, particularly the experimental portion. Not 
 only is it extremely difficult to produce limited injury of the 
 cochlea in small animals, but the examination of their 
 auditory sensations cannot be carried out with sufficient 
 exactness to lead to very definite results. Some evidence 
 from the clinical side is, perhaps, rather more worthy of 
 consideration. I allude particularly to the condition studied 
 by Jacobson, and dignified with the awe-inspiring name of 
 
PHYSIOLOGY OF THE EAR 83 
 
 Diplacusis binauralis dysharmonica, in which the patient 
 hears with one ear a greater or smaller portion of the musical 
 scale " falsely." Such a condition is easy to describe in 
 terms of Helmholtz's hypothesis, difficult to account for on 
 most of the others. Suppose on this theory that the c fibre 
 in the left ear is pathologically changed in dimensions so 
 that it vibrates to d. Then when d is sounded, the d fibre 
 of the right cochlea responds with both the d and c fibres of 
 the left cochlea, but the note associated with the c fibre is 
 normally c, hence in binaural hearing the sensation of c will 
 be erroneously associated with d, and dissonance results. 
 
 Most of the secondary phenomena of compound tones 
 agree well with Helmholtz's hypothesis. Two tones near 
 each other in pitch should excite a zone of fibres, and so 
 lead to variations of intensity of the nature of beats ; if the 
 difference of pitch is great, the excitation of the intervening 
 zone will not be marked enough to produce any effective 
 disturbance of this nature. Both these deductions appear to 
 be experimentally true (Stumpf). The question of combina- 
 tion tones and their relationship with the so-called beat 
 tones has been mentioned above. I attempted, some years 
 ago, to throw light on the theory bj' leading to each ear the 
 same note, but so arranging the source of soundlthat the note 
 reached one ear half a wave length later than the other. On 
 the resonance hypothesis, this should produce almost total 
 interference. I was in no case able to produce total inter- 
 ference, but merely to alter the localisation of the sound, 
 although in some cases variations in intensity were noticed. 
 I subsequently found that Sylvanus Thompson had previously 
 carried out similar experiments with the help of a somewhat 
 more perfect apparatus and a better arrangement. His 
 results also failed to reveal any case of total interference, 
 merely peculiar localisations of the sound. 
 
 The main objection to the theory is the extreme diffi- 
 culty, in view of the histological structure, of understanding 
 how the fibres of the basilar membrane can vibrate in the 
 way postulated by Helmholtz. Pierre Bonnier has criticised 
 the theory from this standpoint, and many of the objections 
 he raises are formidable. Numerous other theories have 
 
84 PHYSIOLOGY OF THE SPECIAL SENSES 
 
 been proposed, some of which are highly ingenious. Those 
 of Waller, Ebbinghaus, Hurst, Mayer, and Ewald, while 
 assuming that coehleal analysis occurs, lay stress on varia- 
 tions in the form of the compound waves, and deny the 
 existence of resonators in Helmholtz's sense. Ewald, in 
 particular, has shown that sound waves produce peculiar 
 patterns on oiled membranes of sizes comparable to that of 
 the cochlea, and that different compound waves produce 
 different patterns. He has published some interesting micro- 
 photographs in support of his views. Sir T. Wrightson and 
 A. Keith have drawn attention to the peculiar conformation 
 and articulation of the stapes, and are inclined to attribute to 
 it a share in the process of tonal analysis. The theory of 
 Rutherford, which transfers analysis to the cerebral structures, 
 is, in effect, a confession of ignorance, and does not present 
 any features of sufficient interest to need detailed examina- 
 tion. In my opinion, the theory of Helmholtz accounts 
 satisfactorily for a greater number of facts than any of the 
 others, but at the same time the difficulties in the way of its 
 full acceptance are so numerous, that we must not regard it 
 as an established truth. 
 
 The psychology of hearing, more particularly in relation 
 to the theory of harmony and aesthetics, does not fall within 
 my province to describe; the reader who has mastered the 
 account of the matter contained in Professor Myers' admir- 
 able text-book should consult the other works mentioned 
 below. I would once again remind him that the physiology 
 and psychology of hearing, like most subjects of real interest, 
 cannot be mastered without sustained attention, and that the 
 works in question will require on his part strenuous applica- 
 tion, which will, however, be amply repaid. 
 
 BOOKS RECOMMENDED FOR FURTHER STUDY 
 
 (The papers and memoirs dealing with the subject are too numerous to 
 be referred to. The reader had better study the books mentioned in 
 the order in which they are given.) 
 
 1. Text-bool£ of Experimental Psychology, by C. 8. Myers, London, 1909, 
 especially pp. 20-62. 
 
PHYSIOLOGY OF THE EAR 85 
 
 2. Die Lehre von den Tonempfindungen, etc., by H. v. Helmholtz, 5th 
 edition, Braunschweig, 1896 (English translation by A. J. Ellis, with 
 literature to 1885. London, Longmans). 
 
 3. Tonpsychologie, by O. Stunvpf, two vols., Leipzig, 1883 and 1890. 
 
 4 Der Gehdrssinn, by ^. Sehaefer, Nagel's Handb. d. Phys., vol. iii pp. 
 476-588. 
 
 5. The Ear, by /. G. McKendrkh and A. A. Gray, Schafer's Text-book 
 o£ Physiology, vol. ii. pp. 1149-1194. Edinburgh, 1900. 
 
 6. L' Audition, by Pierre Bonnier. Paris, 1900. 
 
 The histology of auditory structures is well described in Dahlgren and 
 Kepner's Principles o£ Animal Histology, London, 1908, pp. 215-224. 
 
CHAPTEK X 
 
 THE COMPARATIVE PHYSIOLOGY OF VISION 
 
 Although it is idle to discuss whether any one of the sense 
 organs is more or less important for the well-being of the 
 organism as a whole than any other, yet if we judged by 
 the bulk of relevant literature, pride of place would have to 
 be assigned to the eye. This statement is true not only with 
 respect to human but also in the case of general physiology, 
 and I propose to discuss in the present chapter some problems 
 of vision in animals other than man. 
 
 The powerful stimulus to the scientific study of natural 
 history which we owe to the great biologists of the early and 
 middle years of the last century, above all to Darwin, Alfred 
 Russel Wallace, and August Weismann, resulted in the 
 publication of numerous observations on the sense of animals, 
 especially those of the social hymenoptera. In this field 
 Lord Avebury, Forel, Plateau, and more recently Wassmann, 
 have been especially active. The majority of their researches 
 were directly or indirectly inspired by Darwin, and in general 
 they may be described as the sort of work which Gilbert 
 White would have performed had he lived to come under 
 the influence of the biological renaissance. In other words, 
 they are characterised by patient observation and the perfor- 
 mance of comparatively simple experiments, the results being 
 described in language as far as possible free from technicalities ; 
 comparatively little attention is paid to those minutiae of 
 structure which require for their elucidation somewhat 
 elaborate training in the refinements of modern histology. 
 It cannot be denied that the advantages of this kind of 
 treatment, viz. an appeal to a large public and an encourage- 
 ment of workers in conditions ignorantly supposed to be un- 
 favourable to scientific research, are associated with certain 
 weaknesses. Thus the reader will often notice a tendency to 
 
COMPARATIVE PHYSIOLOGY OF VISION 87 
 
 somewhat facile generalisations and to the employment of 
 a perhaps excessively anthropomorphic or anthropocentric 
 method of interpretation. As generally happens, these short- 
 comings called into existence a school of opposite tendencies, 
 a school the opinions of which are well expounded by Theodor 
 Beer^ and, more recently, J. P. Nuel.- The main thesis of 
 this school can be summarised in a few words. Since no 
 animal can attain to objective knowledge of the sensory and 
 psychological processes taking place in any other animal, the 
 study of sensations or of higher psychical processes depends 
 on the interpretation of physiological reactions to stimuli. 
 
 In the case of man himself, the faculty of speech and 
 the close agreement in structure found in individuals of the 
 same race enable us to infer the existence in one another of 
 sensations and other conscious states with some degree of 
 probability. Still, even here we reason from similarities of 
 anatomical structure and physiological reaction, so that 
 errors frequently arise. Instances will readily occur to the 
 student of medicine; thus stimulation of an afferent nerve 
 causes (with one exception) a rise of blood pressure, and is 
 normally associated with pain. Now this pain is a subjective 
 matter, while the blood pressure can be objectively recorded ; 
 but we should greatly err if we asserted that the objective 
 rise of blood pressure is an accurate indicator of the amount 
 of pain felt. Again, when ethereal vibrations of wave length 
 between 670 and 660 /i/i strike the retina, some 96 per cent, 
 of English adults experience a sensation which they all de- 
 scribe by the same word, red; the remainder, however, are 
 affected in a manner which is not only different, but which 
 the "normals" can by no effort of the imagination picture 
 to themselves. If, therefore, under the most favourable con- 
 ditions, viz. in the study of man by man, errors of interpre- 
 tation occur, it is manifest that in dealing with animals, the 
 anatomical and physiological characters of which differ from 
 those of man in a way not to be mistaken, the attempt to 
 
 1 "Ueber Primitive Sehorgane," Wiene klinische Wochenschrifi, 1901, Nos. 11, 
 12, 13. 
 
 2 "La Vision," BibiioMque Internationale de Psych. Expirim., Paris, 1904, 
 Octave Doin, pp. 3-113. 
 
88 PHYSIOLOGY OF THE SPECIAL SENSES 
 
 place a conscious reaction, a sensation, side by side of a 
 physiological response, is hazardous. We shall project out- 
 wards a more or less faithful picture of our own mental 
 processes, and tend illegitimately to postulate the objective 
 existence of that which is or may be subjective. To adapt 
 Voltaire's taunt, we shall not merely fashion God in our own 
 image, but the whole animal kingdom. For reasons such 
 as these, Beer, Nuel, and their fellows hold that there is no 
 such thing as comparative psychology, but only comparative 
 physiology. They object to use in describing the senses of 
 animals such words as " sight " and " vision," which they 
 hold to be tainted by human associations. " Already the use 
 of the words ' seeing,' ' vision,' gives rise to misunderstand- 
 ings, for they presume the existence of a very complex organ 
 and function, even a psychic representation, as in the case of 
 man. To the word 'light' itself (employed to designate a 
 form of energy) there adheres a sensorial after-taste which 
 still continues to lead readers and writers into error." ^ A 
 new terminology has accordingly been introduced which, at 
 any rate, has the merit of being unfamiliar. Although we 
 may admit that this group of physiologists has pointed out 
 some important errors in the methods of the older writers, 
 its adherents are perhaps less logical and consistent than 
 they suppose. A characteristic of the whole school, to judge 
 from Beer and Nuel's monographs, is a fine scorn of those 
 who have been unfortunate enough to work in ignorance of 
 the new terminological light, a scorn which not infrequently 
 shows itself in somewhat meticulous censures. One illustra- 
 tion will suffice. 
 
 Professor Nuel writes : " Generally, the critical spirit of 
 authors does not reach so far. They commence by admitting 
 as an axiom the existence of sensations in an animal, and look 
 on movements observed in it as incited by these sensations ; 
 that is to say, they set out from two gratuitous assumptions, 
 as we shall see, to satiety later on. Thus Lubbock, to cite 
 only one example, seeing that bees react differently to lights 
 coloured differently (for us), concludes that ' the observations 
 
 > Nuel, op. eit., p. 26. 
 
COMPARATIVE PHYSIOLOGY OF VISION 89 
 
 demonstrate clearly that bees possess the faculty of distin- 
 guishmg colours.' And, nevertheless, the observed facts 
 could be just as easily explained by a simple difference in 
 intensity of a single luminous sensation. But we shall see 
 that even this single sensation is not demonstrated in the 
 case of the bee by the experiments of the authors. The 
 same writer, seeing that ants react to ultra-violet rays, draws 
 the following conclusions : ' It is therefore probable that the 
 ultra-violet rays produce in ants the sensation of a distinct 
 colour, as different (for the ant) as is the green from the 
 red (for ourselves).' He even asks himself 'if the white 
 light of these insects is not different from our white, since it 
 contains (for the ant) one more colour.' 
 
 " At the base of this reasoning, erroneous but generally 
 admitted in comparative biology, will be found another error 
 which is just as fundamental, and consists in admitting that, 
 at least in the case of man, visual movements or motor 
 reactions to light are provoked either by the sensations of 
 light themselves, or by other psychic states (psychic repre- 
 sentation, pleasure, displeasure, will, etc.), which, according to 
 the psychologists, are derivatives of sensations." * 
 
 It seems to me, for the following reasons, that this criticism 
 is purely verbal. The fact that we havp no objective know- 
 ledge of any sensations in animals has long been a common- 
 place of educated men. Nuel has no justification in supposing 
 that the authors he criticises are not familiar with this plati- 
 tude. Being men, however, and writing for men, they use 
 the language of human beings. To say that an ant " prefers " 
 red to violet light is a short way of saying that its motor re- 
 actions under the two conditions are such that, were it endowed 
 with a consciousness similar to that of the speaker, such a 
 preference could be inferred to exist. If any one asserts that 
 the ant does as a matter of fact possess a consciousness 
 
 ^ Nuel, op. cit., pp. 11-12. Beer {op. cit., p. 34) writes, somewhat more 
 •cautiously: " We have no right and no necessity to attribute to lower animals 
 with no demonstrable associative memory sensations at all, and what is built 
 up on sensations, pleasure, Unlust, Common Sensation, etc. They may have 
 these or not ; provisionally we do not require for a functional knowledge of 
 their activities to regard them psychologically otherwise than as to us a new 
 feind of apparatus." 
 
90 PHYSIOLOGY OF THE SPECIAL SENSES 
 
 similar to that of man, and twists his experimental results to- 
 agree with such a dogma, he is no doubt acting unscientific- 
 ally, but I see no evidence of this in any of the numerous 
 passages singled out by Nuel for attack. Indeed, the war 
 might well be carried into the enemy's country. While 
 ostensibly employing a terminology which is non-committal 
 as to psychic states, Nuel, Beer, and others of their school 
 often make merry at the expense of those who, however 
 vaguely, attribute consciousness to "lower" animals. Now 
 the one animal about the conscious states of which I " know " 
 anything is myself; therefore when I have to consider any 
 living creature, the structure and reactions to stimuli of 
 which compel me to classify it with myself somewhere in the 
 animal kingdom, I must, in default of evidence which, by 
 definition, I cannot obtain, attribute to it some degree of 
 "consciousness" or "blunt Occam's razor." When we turn 
 to the inorganic world, or even to the vegetable kingdom, 
 the physiological points of difference outweigh the points of 
 resemblance, and the attribution of consciousness ceases to be 
 justifiable. The suggestion of Beer and Nuel, that the con- 
 ventional method of writing about the sense of animals is 
 akin to the pre-Galilean physics which spoke of water seeking 
 its level and nature abhorring a vacuum, is, I think, purely ad 
 captandum. I admit that anthropomorphic language has- 
 sometimes led writers to go beyond their evidence, and that 
 Beer and his colleagues have done good service in pointing 
 this out, but I cannot think that the evil is as great as they 
 assert, or that the remedy they propose is really an efficient 
 one. As a matter of fact, the very authors who are particu- 
 larly scornful of others' anthropomorphism not infrequently 
 commit the same fault themselves. For instance, Professor 
 Nuel writes : " The insect which orientates itself in flying 
 among small obstacles reveals a remarkable development 
 of its moto-reactions, but no true icon-reactions. To judge 
 by what occurs in man, the latter even appear to be impos- 
 sible during so rapid a movement." ^ 
 
 Why, on Professor Nuel's principles, s/ioicW we "judge by 
 
 ' Nuel, op. cit., p. 91. 
 
COMPARATIVE PHYSIOLOGY OF VISION 91 
 
 what occurs in man " ? Further on he writes : " Of what 
 nature is this visual memory of places in the hymenoptera ? 
 There is certainly no question of a revival of previous icono- 
 reactions, of which bees are certainly deprived, and which, 
 besides, could not be produced during so rapid a translation 
 as the flight of a hymenopteron." ^ 
 
 The man who writes such sentences as these ought, I 
 think, to be a little careful how he accuses others of anthropo- 
 centric errors. I shall therefore, in the brief notes which ■ 
 follow, make no attempt to avoid the common descriptive 
 words used by Forel, Wassmann, and other comparative 
 biologists. 
 
 Our knowledge of the anatomy and histology of the eyes 
 of Invertebrata apart from Insecta owes much to the work 
 of Hesse, ably summarised by Beer in the memoir already 
 referred to. Exner has studied the faceted eye of insects in 
 great detail, while the development of histological technique, 
 particularly the silver methods of Golgi and the work of 
 Ramon y Cajal, has enabled anatomists to describe the 
 structure of the eyes of most vertebrata with remarkable 
 precision. 
 
 In some quite simple animals, simple, that is, from the 
 point of view at which we have so far arrived, the light- 
 perceiving structures are decidedly complex; indeed the 
 complexity of the latter by no means goes hand in hand 
 with the general complexity of the organism, so that no 
 satisfactory zoological classification is possible. Generally 
 speaking, we find one or more elongated structures formed 
 from cytoplasm or excreted by it, often exhibiting a laminated 
 structure, and, in the majority of cases, associated with pig- 
 ment. Pigment, however, is not invariably present, as, for 
 instance, in the case of the " photirzellen " of several annelid 
 worms and the more complex photirzellen of the leech (Fig. 7). 
 A stage beyond the " photirzellen " is seen in the ocelli of 
 leeches (Fig. 8), while yet more specialised are the ocelli of flat 
 worms (Fig. 9). Some of the polychsetes exhibit true camera 
 eyes (Fig. 10), and in cephalopods the retinal structure is 
 highly complex. 
 
 ' Nuel, op. cit. , p. 104. 
 
92 PHYSIOLOGY OF THE SPECIAL SENSES 
 
 Fig 
 
 In arthropods we meet witli a specialised type of eye, the 
 form of which may be studied in the cockroach (Periplaneta 
 orientalis). The outer surface shows many divisions, and a 
 vertical section through the principal axis proves, that the 
 superficial divisions mark the bases of long truncated cones, 
 the ends of which rest on a semicircular basement membrane. 
 
 Each cone is a single unit (or 
 ommatidium) of the eye, con- 
 sisting of numerous parts, for an 
 account of which reference must 
 be made to works on compara- 
 tive histology. 
 
 Our knowledge of the physio- 
 logy of ocelli and "photirzellen" ^ 
 is so incomplete, that we must 
 content ourselves with saying 
 that they are concerned in some 
 way with the reception of lumi- 
 nous stimuli; with respect to 
 the compound eyes of insects, 
 we know a little more. The 
 work of Exner, Grenacher, and 
 Oscar Schmidt has demonstrated 
 that a separate image cannot be 
 formed on each retinular ele- 
 ment, and that the older theory 
 of Johannes Mtiller is more 
 plausible. According to this 
 theory, each facet will receive 
 rays from a different part of 
 the object, owing to the small 
 size and great number of the 
 facets, and as a result of the co-operation of all the facets 
 a sort of mosaic pattern will be produced on the retina, 
 not an immense number of small distinct images. Exner 
 seems to have proved that the transparent structures cannot 
 project an image, but only concentrate the rays which 
 form the visual field of each facet. Evidently, if this theory 
 1 Which may be translated by " Light-receiving Cell." 
 
 7. — Section tlirough aa Ocellus of 
 Hirudo mcdicvnalis (Beer), 
 ejj^ Epithelium. LC Light-receiving Cell. 
 
 S Sensory Bulbs. 
 SN Nerve to Bulbs. 
 
 no Optic Nerve. 
 m Muscle. 
 
COMPARATIVE PHYSIOLOGY OF VISION 93 
 
 be correct, the distinctness of vision, or, to speak more accu- 
 rately, the localisation of the source of the different rays, 
 depends on the co-ordination of the facets ; the greater the 
 number of facets, the more accurate will be the localisation 
 in eyes of the same total volume, since each facet will be 
 smaller and illuminated by a smaller portion of the object. 
 The advantage of having numerous small facets will be 
 increased if the whole eye is markedly convex, because in 
 that event not only will the visual field be increased, but 
 fewer facets will receive rays from the same external point. 
 We should further expect, if the theory of mosaic vision be 
 
 -^m 
 
 Fie. 8. 
 
 Fig. 9. 
 
 u, Light-receiving Organ of Zumbricus 
 
 castaneus (Beer). 
 
 LC Light-receiving Cell. 
 ep Ordinary Epithelial Cell. 
 
 h Light-receiving Organ of Lumbricus 
 
 ruiellus (Beer). 
 
 LC Light-receiving Cell. 
 ep Ordinary Epithelial Cell. 
 
 sound, that visual acuity, in the human sense of being able to 
 distinguish accurately the forms of objects, would be poor in 
 insects, while the perception of a movement should be good. 
 All these statements are supported by the experiments of Forel. 
 In support of the assertion that insects perceive movement 
 better than form may be cited his observations on wasps. 
 Forel noticed that a wasp hunting flies was frequently 
 deceived by a nail in the wall about the same size and shape 
 as a fly, pouncing on it more than once. When dead flies or 
 spiders of about the same size were placed on a table the wasp 
 carried them off indifferently, but took no notice of insects 
 
94 PHYSIOLOGY OF THE SPECIAL SENSES 
 
 much larger or much smaller than the dead flies mixed with 
 them. Forel also adduces, in favour of the view that the 
 more facets the more distinct the vision, the great visual 
 powers of the dragon-fly as compared with the much less 
 endowed ant. Forel has also proved hy varnishing the eyes 
 that humble-bees, wasps, and numerous dipterous insects are 
 unable when thus blinded to direct their flight satisfactorily, 
 while this power remains if both antennse and the forepart of 
 the pharynx are removed. 
 
 The numerous experiments of Lord Avebury and Forel 
 have made it probable that bees and wasps can distinguish 
 
 certain colours, and the objections 
 of Plateau to their results do not 
 seem to be well founded, but I 
 have no space to give the detailed 
 evidence. There is also reason to 
 think that some power of distin- 
 guishing the form of objects is 
 possessed by these insects. Forel 
 carried out numerous experiments 
 on a wasp, of which the following 
 may be quoted : — 
 riG.io.— SectionthroughanOoeiius "The following day my wasp re- 
 of Pianaria torva (Beer). tumed twice following to eat at the 
 i? ffght-'r^Sving Cell. cross left at the same place. I then 
 
 R EoTuke structures. took her and removed bothantennse. 
 
 She flew away, but returned half- 
 an-hour later to eat, always at -the cross, which I had left in 
 the same place. After her departure I put a similar cross, 
 at the side of the first cross, but without honey, then on the 
 other side a narrow band with honey, finally removing the 
 honey cross. The wasp returned, flew straight on to the 
 cross, alighting just in the centre (where the honey was on 
 the other cross), and searched it vainly for a considerable 
 time. Then, although deprived of antennse, she began to 
 search, doubtless recollecting that the white papers on which 
 the honey was had already often changed in place and 
 aspect. 
 
 " She quickly found it on the narrow band, not without, 
 
COMPARATIVE PHYSIOLOGY OF VISION 95 
 
 however, having passed -within a few millimetres of it two 
 or three times without noticing it, which would not have 
 happened to her if she had had her antennae. She only 
 noticed it when her mouth touched it.''^ 
 
 The great difficulty of such experiments as these is that 
 of eliminating all disturbing factors, and the reader who 
 desires to form definite conclusions will be compelled not 
 only to study the works of these observers, but to perform 
 experiments on his own account, otherwise he will be in 
 danger of neglected sources of error which are not usually 
 described in experimental records. 
 
 It is unnecessary to refer to the experiments which have 
 been made on the vision of vertebrata ; in general terms, it 
 may be said that the perception of movement can be more 
 easily shown to exist than the kind of vision associated with 
 stimulation of the fovea centralis retinaj in man. This is, 
 perhaps, true even for animals, such as the monkey, which 
 possess a retinal fovea histologically similar to that of man, 
 although in the monkey many fine movements seem to 
 prove the existence of a type of vision altogether comparable 
 with that of man. Of course any one who has played with a 
 dog does not need to be told that the friend of man is far 
 more keenly interested by moving than by stationary objects. 
 
 It will be clear that many important problems in con- 
 nection with the comparative physiology of vision still await 
 solution, and that the subject is one of peculiar interest. 
 
 BOOKS AND PAPERS RECOMMENDED FOR 
 FURTHER STUDY 
 
 The student should first read the account of the histology of visual 
 tissues in Principles of Animal Histology by V. Dahlgren and W. A. 
 Kepner, Macmillan, 1908, then the works mentioned in this chapter in 
 the following order : (1) Beer's paper, (2) Nuel's book (first part), (3) 
 Forel. Ample references to the vast literature of the subject will be 
 found in these sources. 
 
 '■ Forel, The Senses of Insects, Yearsley's translation, London, Methuen, 
 pp. 27-28. 
 
CHAPTER XI 
 
 RETINAL PROCESSES, ELECTRICAL, PHOTOTROPIC, 
 AND CHEMICAL RESPONSES 
 
 The physiology of the eye falls naturally into two main 
 divisions : in the first we regard the organ as an optical 
 instrument, investigate its structure, and determine the 
 constants of the lens system together with the latter's defects 
 and compensations. We also investigate the relations sub- 
 sisting between the optical mechanism and the afferent or 
 efferent nervous paths, together with the manner in which 
 these relations may be modified or destroyed. This branch 
 does not fall within the scope of a treatise such as the 
 present. I assume the reader to possess some knowledge of 
 the easier parts of physiological optics, and intend to deal 
 solely with the other division of our subject. This investiga- 
 tion starts with the arrival of a stimulus at the retina, ex- 
 amines the resulting changes in this structure, and attempts 
 to associate them with the materials yielded by an analysis 
 of visual sensations and stimuli. In order that we may not 
 lose sight of the wood for the trees, it is necessary to dis- 
 tinguish carefully between the probable and the possible, 
 under a penalty of losing ourselves in a maze of pure specu- 
 lation. When a beam of light falls on the retina, certain 
 marked physical changes occur; in some of these cases a 
 distinction is possible, because the reaction does not appear 
 to be the same over the whole surface of the retina, a fact 
 which apparently depends on the non-uniformity of structure 
 displayed by this membrane. This statement cannot be 
 generalised, because certain forms of response are such that 
 we can, under possible experimental conditions, only consider 
 their relation to the retina as a whole ; an example is the 
 photo-electric response. If the retina and optic nerve be 
 connected up through a galvanometer, a " current of rest '' is 
 
RETINAL PROCESSES 97 
 
 observed, its direction depending on whether the inner or 
 outer surface of the retina is used. When hght falls on the 
 retina, this current undergoes a somewhat complex variation. 
 In the isolated retina there is first a positive then a negative 
 variation ; on cutting off the light a positive variation is 
 produced. In conditions most like the normal state the 
 positive phase is better marked in the case of the frog, but 
 in birds and mammals a negative phase alone is obtained. 
 This result appears in the absence of the visual purple, 
 although the reaction is more intense if this pigment be 
 present, and it is clear that the whole effect varies in in- 
 tensity with the part of the spectrum employed. 
 
 Another general change of state is the fact that after 
 exposure to light the retina is less readily stained with acid 
 dyes, but this is not so definite nor so generally accepted as 
 the electrical response.^ Thirdly, we have the well-known 
 phototropic reaction of the pigment epithelium. In a frog's 
 eye which has been kept in the dark the pigment layer 
 is easily separated from the rods and cones ; the pigment 
 granules are in the cell bodies round the outer limbs of the 
 rods. After exposure to light, the layers can only be separated 
 with difficulty; the pigment is much more abundant between 
 the outer limbs of the rods, and also passes between the 
 inner limbs as far as the external limiting membrane. This 
 change can be induced in the frog by ten minutes' exposure 
 to light. The various parts of the spectrum are not equally 
 efficient in producing this change, the long waved lights, 
 especially red, being but feebly active. The pigment will 
 not move if the central nervous system is destroyed, and the 
 reaction may take place in one eye as a consequence of light 
 falling on the opposite retina. A third response is cone 
 shortening, which appears to be due to contraction of the 
 inner limbs. Just as in the last case, cerebral injury abolishes 
 the power of movement, and the change may be produced 
 indirectly by stimulation of the opposite eye. Here, again, 
 the various spectral lights differ in power to stimulate, but 
 certain peculiarities may be noticed. Red light produces a 
 
 1 See Dittler, Pflilij. Arch., 1907, oxx. 44. 
 
98 PHYSIOLOGY OF THE SPECIAL SENSES 
 
 greater effect than on the pigment cells, and shortening may 
 follow stimuli other than light, e.g. changes of temperature. 
 Cone shortening also seems to take place in mammalian eyes, 
 while the pigment change is only definiteljestablished for 
 the frog. ^ S«, ^.ci^^-? P^^i /f/7 . V^jZ" 
 
 Finally we come to the important question of Visual 
 Purple or Retinal Red. This is a substance of a deep purple- 
 red hue, present mainly {perhaps entirely) in the outer limbs 
 of the rods. On exposure to light the pigment is rapidly 
 bleached without passing through an intermediate yellow 
 stage, as has frequently been asserted. This bleaching is 
 limited to the part upon which light falls, and the latter is 
 alone capable of decolourising the pigment in a living eye. 
 Regeneration takes place either in darkness or red light, 
 apparently depending on the existence in or production by 
 the pigment epithelium of a substance called by Kiihne 
 Rhodophyllin, as suggested by the following experiment. 
 When a frog is curarised, oedema occvirs between retina and 
 choroid, and the former is separated from the pigment layer, 
 which tends to adhere to the choroid. If the frog be now 
 exposed to light until bleaching results, no regeneration 
 occurs on detaching the retina and placing it in a dark 
 chamber ; but if the pigment layer is simply placed in 
 contact with the retina, new formation proceeds as under 
 normal conditions. We seem, therefore, entitled to assume 
 that regeneration is independent of direct continuity, and is 
 due to chemical or physical processes. It was once believed 
 that the formative substance is derived from the coloured 
 globules in the epithelial cells, but this is probably incorrect, 
 as the process seems to occur when they are wholly or mainly 
 absent, as in the albino rabbit. Visual purple can be ex- 
 tracted from the retina with a solution of bile salts, and its 
 absorption spectrum has been examined ; the absorption does 
 not appear to be identical in purple from diiFerent animals, 
 but the significance of the variations is unknown. 
 
 If we compare the bleaching powers of the homogeneous 
 lights, we find that red and yellow are practically inoperative, 
 the parts of the spectrum falling between the Fraunhofer D 
 and E lines having, on the other hand, maximal powers. 
 
RETINAL PROCESSES 99 
 
 If we ask ourselves what value these facts have for a 
 study of visual processes, we shall conclude that some dis- 
 tinctions must be made. In the first place, experimental 
 conditions requisite for observing electrical changes render 
 the co-ordination of the latter with sensation-differences 
 hardly practicable. Haas suggested that the electromotive 
 response could be regarded as a measure of sensation-in- 
 tensity, and plotted the curve obtained when the light 
 intensities were taken along one co-ordinate axis and the 
 electromotive response along the other. If the stimuli were 
 increased in geometric and the electromotive responses 
 varied in arithmetic ratios, as they should if Fechner's law 
 holds, the curve would be a straight line. Haas' figure shows 
 that this is only very roughly true, and in any case the 
 assumption that the electrOmotiva response is a measure 
 of sensation-change appears to be quite arbitrary. Perhaps 
 all we can safely infer is that light produces some change 
 of state in the retinal structure, a change not of the same 
 magnitude, nor even, it may be, qualitatively identical Avhen 
 we employ stimuli of different wave lengths; the exact 
 relationship between physical process and psycho-physio- 
 logical response is still obscure. These remarks also apply 
 to the phototropic movement of the pigmented epithelium 
 and the shortening of the retinal cones, but on turning to 
 the facts relating to visual purple, we are tempted to pursue 
 our investigation further. Here we have a substance only, 
 or at least mainly, present in certain areas of the retina; 
 if it play a part in the physiological processes of vision, these 
 should be different in different regions. We might, of course, 
 have the final result, the visual sensation, identically the 
 same even if the physiological precursors were quite difierent ; 
 but if, in fact, we find that visual sensations yielded by stimuli 
 acting on different parts of the sensitive area are not uniform, 
 that a modifying factor applied to the whole retina affects 
 it differentially, we have some ground for attributing definite 
 importance to the visual purple. To obtain the grounds 
 necessary for forming a judgment on this point, we must 
 first ascertain what alterations in visual responsiveness are 
 produced by changing the totality of conditions to which the 
 
100 PHYSIOLOGY OF THE SPECIAL SENSES 
 
 eye is subjected; secondly, we must knoAv whether the 
 changes are entirely or chiefly confined to any particular 
 region ; and lastly, whether any structural peculiarities give 
 us a plausible explanation. 
 
 It has long been known that the nature of the response 
 of the eye to a stimulus largely depends upon whether, 
 before the experiment, the subject has rested in a dark 
 room or been exposed to light, that is to say, whether there 
 is a condition of light or dark adaptation. The detailed study 
 of these and allied phenomena will form the subject of the 
 next two chapters. 
 
 RECOMMENDED FOR FURTHER STUDY 
 
 TV. Nugel, Die Wirkungen des Liohtes auf die Netzhaut (Nagel's Hand- 
 bueh der Physiol., Bd. iii. pp. 91-108) contains an excellent account, 
 with full references to the literature. 
 
CHAPTER XII 
 
 VISUAL ADAPTATION— PERIPHEEAL VISION 
 TOTAL COLOUR-BLINDNESS 
 
 As was mentioned at the end of the last chapter, the fact 
 that different parts of the retina are unequally responsive 
 to light has long been known, and the path which led up 
 to this knowledge was a study of the facts of visual adapta- 
 tion to varying intensities of light. 
 
 Aubert was one of the first exact observers to notice that 
 the response of the eye to feeble stimulation was heightened 
 by a sojourn in darkness. He also noticed that the threshold 
 value of a stimulus in such cases varied inversely as the area 
 stimulated. Since this could only mean that when light 
 fell on the peripheral part of the retina its value as a 
 stimulus was enhanced, Aubert's observation really contained 
 the nucleus of all subsequent work. His general results were 
 speedily confirmed. Charpentier, among others, found that 
 the central part of the "dark" retina, although exhibiting 
 an increased responsiveness as compared with the same 
 region in the "light" eye,'^ was much less responsive than 
 the periphery. This increase was more marked in the case 
 of the short waved spectral colours, and there is even some 
 doubt whether adaptation makes any difierence at all to the 
 activity of red rays. It is also rather uncertain whether the 
 fovea centralis retinae is affected by resting in a dark room ; 
 the evidence is conflicting, and the matter is still too con- 
 troversial for it to be discussed here. 
 
 A particular case of adaptation which is of much interest 
 is " Purkinje's Phenomenon," an effect which can be described 
 in the following way. If one examines an ordinary spectrum, 
 the brightest part of it seems to occupy the neighbourhood 
 
 ' I sball speak of an eye which has been rested in the dark as a " dark " 
 and one previously exposed to light as a "light " eye. 
 
 101 
 
102 PHYSIOLOGY OF THE SPECIAL SENSES 
 
 of tlie yellow or orange-yellow ; if, however, the physical 
 intensity of the spectrum be diminished, for instance, by 
 moving the source of light farther away from the prism, 
 the maximum of apparent brightness shifts towards the 
 violet end. With the feeblest illumination which enables 
 one to distinguish the spectral colours at all, the brightest 
 part is at the junction of the green and blue. Ewald Hering 
 has, in my opinion, demonstrated that these changes are due 
 to adaptation, using an experimental method of characteristic 
 simplicity and elegance. Two rooms which could be in- 
 dependently darkened are separated by a light-proof partition, 
 in which two slits are cut and covered with pigmented glass. 
 The amount of light transmitted by these slits could be 
 varied independently with the aid of reflectors. So long 
 as the room occupied by the observer is kept at a constant 
 illumination, diminishing the physical intensity of the light 
 traversing the two slits, which are covered with blue and 
 red glass respectively, does not change their relative in- 
 tensities. If, however, the observation room is darkened, the 
 blue slit immediately appears brighter than the red one, 
 even while the physical intensity of the light passing through 
 them is unaltered. The effect is much enhanced by a stay 
 in darkness, and is more noticeable in indirect (peripheral) 
 than in direct (foveal or central) vision. Burch has noticed 
 a similar phenomenon, and we may fairly consider Purkinje's 
 effect as dependent not on physical intensity, but adaptation. 
 Before going further, we must ask ourselves one question, 
 What does one mean by saying that different colours are 
 equally or unequally bright ? I am not acquainted with 
 any really complete answer, and shall fall back upon a purely 
 empirical justification. Ask a dozen normal persons to look 
 at a spectrum in daylight, and one finds they all agree in 
 picking out some part of the orange-yellow, which they call 
 the brightest point of the spectrum. They mean, I take it, 
 that this part produces somehow a predominant effect in 
 consciousness. How this comes about is matter for a psycho- 
 logical discussion; it is no part of the pure physiology of 
 vision. If we adopt this relatively humble standard of 
 brightness, comparative results are attainable in numerous 
 
VISUAL ADAPTATION 103 
 
 ways. One of the best methods is that of " flicker," to which 
 Haycraft and Kivers have devoted much attention. When 
 a series of sectors are whirled round on a machine called 
 a colour mixer, the velocity necessary to produce a fused 
 sensation depends upon the brightness of the sectors ; hence 
 with different sets of sectors equal in size, the velocity of 
 rotation which just extinguishes the sensation of flicker 
 affords a measure of brightness, which may be taken, within 
 fairly wide limits, to vary inversely as the rapidity of 
 rotation. It may be well to remark here that work of this 
 kind, like most researches on the sense of sight, is easy 
 neither to perform nor to interpret. To carry out satisfactory 
 experiments on a single person requires much attention to 
 details, and even when the obstacles have been surmounted 
 we have to consider the question of individual variations. 
 The problem of variations in the reaction to sensory stimuli 
 is just as urgently in need of adequate statistical treatment 
 in the case of sense physiology as in any other biological 
 field, and even less likely to obtain it. 
 
 Hering several years ago noted the relative darkening 
 of red as one passes from central to peripheral vision ; he 
 compared a pure red, a spectral mixture of red and blue- 
 green (656 /i/i + 470 /ti/i), and daylight. The converse was 
 found to hold for spectral green (505 /ti/t), but his results 
 were not quite satisfactory, for momentary dark adaptation 
 occurred during the experiments. Tschermak, who studied 
 the whole question systematically in the " light " eye, found 
 in indirect vision a relative diminution in brightness for 
 light of wave length between 693 and 525 /i/i, no change 
 from 525 to 516 jxii, an increase from 516 to 466 /n/i. 
 Similar changes were observable in " dark " eyes. 
 
 Another method was to start with a large field of colour- 
 less light, produced by mixing together complementaries, 
 and then to diminish its size. It has been found that a 
 colourless mixture of spectral red and bluish-green becomes, 
 with such a real diminution, redder and darker; if the 
 change of size be an increase, it becomes greener and brighter. 
 In fact, both for " light " and " dark " eyes colourless matches 
 valid for the periphery do not hold for the fovea centralis. 
 
104 PHYSIOLOGY OF THE SPECIAL SENSES 
 
 and vice versa. Apart from adaptation, it is interesting to 
 see whether merely changing the intensities (physical) of 
 two mixtures renders the match invalid. On this point 
 much difference of opinion exists, which is not surprising in 
 view of the experimental difficulties which have to be en- 
 countered. One subsidiary phenomenon which was studied 
 in this connection deserves a word of description, but the 
 main question is too complicated and too controversial to 
 be worth following up.^ In studying Purkinje's phenomenon, 
 we found that if the intensity of a spectrum were diminished 
 a point was reached at which the brightest part appeared 
 to have been shifted towards the violet. What, it may be 
 asked, happens if the illumination be still further diminished ? 
 Under normal circumstances we soon reach a point at which 
 the whole spectrum appears colourless, differing however 
 in brightness in the various parts. Keduction beyond this 
 yields an intensity which is associated with no sensation 
 at all. Hence it seemed necessary to distinguish between 
 the absolute liminal intensity of a spectrum, that is, the least 
 intensity corresponding to a colourless sensation, and the 
 " specific " threshold value for which the spectral zones could 
 be seen to differ in hue. In determining these absolute and 
 specific thresholds, great divergences were found between the 
 results of different workers. The observations of Burch go 
 far to prove that the existence of a specific threshold, at any 
 rate in the case of foveal or direct vision, depends upon the 
 after-effects of previous stimulation. This physiologist first 
 performed some qualitative experiments which are most 
 suggestive. A Bunsen burner was completely covered by a 
 metal chimney, so as to prevent any escape of light while 
 not interfering with ventilation. By bringing the flame into 
 contact with the metal chimney the latter could be heated 
 gradually to a point at which it became luminous. If the 
 experiment were performed in a room with windows covered 
 by ordinary blinds — that is to say, in a room from which light 
 had not been absolutely excluded — thefirst appearance of light 
 
 ' Further information will be found in the article by the present writer 
 contributed to Further Advances in Physiology, London, 1909 (Arnold), pp. 
 354, etc. 
 
VISUAL ADAPTATION 
 
 105 
 
 Adaptation Values {Burch). 
 During the period of increasing adaptation. 
 
 Time in Dark. 
 (Mins.) 
 
 Intensity of 
 
 Minimum Visible 
 
 Bed. 
 
 Intensity of 
 
 Minimum Visible 
 
 Blue- Violet. 
 
 Eatio, V:E.. 
 
 9 
 21 
 60 
 
 16-63 
 11-14 
 
 2-4 
 
 254-76 
 
 57-61 
 
 6-39 
 
 15-32 
 5-17 
 2-24 
 
 B. After spending two hours in the dark room. 
 
 Time in Dark. 
 (Mins.) 
 
 Intensity of 
 
 Minimum Visible 
 
 Ked. 
 
 Intensity of 
 
 Minimum Visible 
 
 Blue-Violet. 
 
 Eatio, V : E. 
 
 120 
 122 
 125 
 127 
 130 
 
 1-0 
 1-20 
 2-69 
 5-04 
 60-02 
 
 1-0 
 
 1-63 
 
 6-14 
 
 12-44 
 
 225-72 
 
 1-0 
 
 1-34 
 
 1-91 
 
 2-47 
 
 4-51 
 
 c 
 
 V 
 
 QE 
 
 Fig. 11. — Buroh's apparatus for the study of " achromatic thresholds." 
 A. An electric lamp, 16 candle-power. B. Paper reflector. 
 
 CC. Wall of dark room. D. Stand with polarismg prism (E). 
 
 F. Spectroscope with double image prism (ff) over eyepiece. 
 There is a stop with two slits (allowing red and blue-violet light to pass through) in the 
 eye-piece. By rotation of the polarising prism the relative intensities of the red and blue-violec 
 lights can be varied. 
 
A. Relative Stimulus Values of different Spectral Regions (Parinaud). 
 
 Fraunhofer Lines. 
 
 Adapted Retina. 
 
 TJnadapted Retina. 
 
 B 
 
 OTU 
 
 tJtj 
 
 
 
 lio 
 
 ToD 
 
 D 
 
 T^ 
 
 tAt 
 
 E 
 
 1 
 
 tJu 
 
 F 
 
 1 
 
 tJc 
 
 G 
 
 iJo 
 
 T.^ 
 
 H 
 
 ^hs 
 
 ? 
 
 B. Increased Responsiveness of Peripheral Retina {Dark-adapted). 
 
 Stimulus : A bluish-white object, 35 degrees in diameter (v. Kries, 
 p. 171).' 
 
 Eesponsivenesa 
 
 (Arbitrary 
 
 Scale). 
 
 Temporal 
 
 Eccentricity in 
 
 Degrees. 
 
 Nasal 
 
 Eccentricity in 
 
 Degrees. 
 
 Insensitive 
 Zone. 
 
 1-0 
 
 1-07 
 
 0-85 
 
 1-92 
 
 1-78 
 
 1-22 
 
 1-06 
 
 2-28 
 
 7-12 
 
 1-70 
 
 1-38 
 
 3-08 
 
 16-02 
 
 2-3 
 
 1-92 
 
 4-22 
 
 28-48 
 
 3-0 
 
 2-58 
 
 5-58 
 
 44-50 
 
 3-75 
 
 3-33 
 
 7-08 
 
 64-08 
 
 4-04 
 
 4-04 
 
 8-08 
 
 1 Distances from the fovea centralis of any point on the retina can be 
 conveniently measured in terms of the angle subtended at the nodal point by 
 a segment of the retina out by a plane passing through the fovea, nodal point, 
 
 FiQ. 12. 
 
 and position of the object, and bounded by a straight line through the nodal 
 point from the object intersecting the retinal segment and by the principal 
 axis. E.g. if A be an object, B the nodal point, and CO a section of the 
 retina, the position (eccentricity) of the image of the point A is defined by 
 the angle a. 
 
VISUAL ADAPTATION 107 
 
 was a pearl-grey tint, the achromatic threshold which others 
 had noticed. But when the room was changed to one 
 without windows and absolutely dark, the first appearance 
 of luminosity was not grey, but dull red. On repeating this 
 experiment, after spending a few minutes in a lighted room, 
 the former grey appearance was once more obtained. Quan- 
 titative experiments were then performed with the apparatus 
 shown in the diagram on page 105. 
 
 Burch was also able to show that a form of after-images 
 (his " dazzle tints ") are very persistent, and may endure as 
 long as two and a half hours after exposing the eye to light, 
 so that his contention can almost be regarded as established 
 in the case of direct vision. That an absolute as distinct 
 from a specific threshold exists in the case of peripheral 
 vision seems, on the whole, probable in view of the careful 
 work of Armin Tschermak, who used very prolonged dark 
 adaptation ; but Burch's work is quite sufficient to show how 
 cautious one must be in interpreting results obtained by 
 these methods, and how complex they are. 
 
 Direct experiment, therefore, seems to have established 
 the following points : — 
 
 (1) The peripheral regions of the retina are relatively 
 more sensitive than the fovea to light of moderate or short 
 wave length. 
 
 (2) Adaptation to darkness is characterised by an increase 
 in responsiveness to short-waved light, and this change is 
 mainly, if not entirely, extra-foveal. 
 
 The tables illustrate these statements. The intensity 
 values are arbitrary, the measurement of eccentricity is in 
 terms of the angle subtended at the nodal point of the eye. 
 
 We now come to indirect evidence tending in the same 
 direction ; of this the facts relating to total colour-blindness 
 are the most striking. 
 
 Total colour-blindness is almost always a congenital defect, 
 and is characterised, apparently, by a complete absence of 
 colour perception in the ordinary sense; it differs toto cielo 
 from the condition of partial colour-blindness, which will be 
 described in a later chapter. A person in this state may see 
 a spectrum merely as a grey strip unequally bright in the 
 
108 PHYSIOLOGY OF THE SPECIAL SENSES 
 
 diiFerent parts which seem to us of distinct hues. To de- 
 scribe the sensations of a second human being must be 
 impossible, but perhaps we may say, for the sake of com- 
 parison, and without pretending to real exactness, that a 
 totally colour-blind man receives from a coloured print im- 
 pressions similar to those excited in ourselves by an un- 
 coloured one. A summary of the observations of Hering on 
 a typical case will give the reader a better idea of the facts 
 than any catalogue of signs and symptoms. 
 
 Hering 
 
 s Case of Total 
 
 Colour-hlindness. 
 
 
 Equally Bright Circle 
 for Colour-blind. 
 
 
 White Valency of a 
 
 Coloured Circle. 
 
 
 White 
 Valency.' 
 
 Grey Circle matching 
 the Coloured Circle for 
 
 
 
 White. 
 
 Black. 
 
 
 a Normal "Light" Eye. 
 
 
 Degrees. 
 
 Degrees. 
 
 Degrees. 
 
 Degrees. 
 
 Bluish-red . 
 
 13-0 
 
 347-0 
 
 18-8 
 
 40 
 
 Yellowish-red 
 
 5-5 
 
 354-5 
 
 11-4 
 
 46 
 
 Orange 
 
 37-0 
 
 323-0 
 
 43-4 
 
 159 
 
 Yellow 
 
 136-5 
 
 223-5 
 
 140-2 
 
 283 
 
 Arsenic-green 
 
 228'0 
 
 132-0 
 
 230-5 
 
 £05 
 
 Green . 
 
 152-0 
 
 208-0 
 
 155-5 
 
 137 
 
 Greenish-hlue 
 
 109-5 , 
 
 250-5 
 
 113-7 
 
 89 
 
 Ultramarin e-blue 
 
 88-3 
 
 271-7 
 
 92-8 
 
 34 
 
 Violet . 
 
 47-5 
 
 312-5 
 
 52-7 
 
 32 
 
 The subject was a man of twenty years, whose colour 
 vision had always been abnormal. He said that he could 
 read without difficulty, provided the light were not too 
 intense, but that his eyes were readily fatigued by bright 
 illumination. In twilight his vision was especially good, 
 particularly if the light were very feeble. On examination, 
 the following results were obtained. No objective changes 
 were detected with the ophthalmoscope, nor was any part of the 
 retina insensitive ; there was no totally blind area (or scotoma). 
 His power of distinguishing two spots unequally bright — 
 physically — was much below that of a normal person, in 
 
 ' For methods of measuring white "valency" consult Hering, pp. 567, 
 etc. For the present purpose the figures in the third column of the table may 
 be regarded as a recalculation of the amounts of white in the sectors which 
 match the coloured circles, so as to admit of comparison with the figures of 
 the fourth column. 
 
VISUAL ADAPTATION 109 
 
 bright light ; in a dark room, it was much superior. With a 
 spectrum, it was found that the area of red which produced 
 any sensation was much diminished; there was shortening 
 of the red end, and those parts which were effective seemed 
 less bright than to a normal eye. The violet end, on the 
 other hand, was not shortened, and it seemed relatively 
 brighter than to the normal " light '' eye, while the region of 
 maximal brightness was in thp neighbourhood of Fraunhofer's 
 E and C lines. Brightness matches between coloured sectors 
 and mixtures of black and white gave the results indicated 
 in the table, which contains comparative values for the nor- 
 mal "light" eye. It is to be noted that the colour-blind's 
 matches were valid, i.e. good matches, for a normal "dark" 
 eye. 
 
 Tests were also carried out polariscopically. At one end 
 of a horizontal tube, blackened on the inside, a cork plate 
 was fixed ; the plate was perforated in the middle and a 
 doubly refracting prism inserted. The other end of the tube 
 was closed by a lid, in which two equal and symmetrically 
 placed semicircular openings were made. With this con- 
 trivance the ordinary and extraordinary images obtained by 
 polarisation appeared to form a series of circles when the 
 tube was directed to a source of light, e.g. a piece of baryta 
 paper stretched over a glass plate. Between the eye and the 
 prism a small telescope and a Nicol prism were introduced, 
 together with a graduated arc. The diaphragm of the tele- 
 scope removed the lateral images, leaving only two magnified 
 white circles, the halves of which could have their brightness 
 altered in opposite directions by rotating the Nicol prism. 
 In front of one opening in the tube a coloured glass was 
 placed, and the Nicol so arranged that for the normal " dark " 
 eye or for the totally colour-blind eye, both halves appeared 
 equally bright. The next table gives the readings for two 
 observers. Some of the variations may be explained by the 
 fact that the normal-sighted person was unpractised in this 
 sort of work. 
 
 Precisely similar results were obtained in matching spectral 
 colours. 
 
 The analogy between normal vision under conditions of 
 
110 PHYSIOLOGY OF THE SPECIAL SENSES 
 Polariseopic Matches (Hering). 
 
 Yellow Glass. 
 
 Blue Glass. 
 
 Total Colour-blind. 
 
 Normal Dark-adapted 
 Eye. 
 
 Total Colour-blind. 
 
 Normal Dark-adapted 
 Eye. 
 
 D egrees of Eotatio n. 
 
 Degrees of Eotation. 
 
 Degrees of Rotation. 
 
 Degrees of Rotation. 
 
 22-3 
 
 22-9 
 
 18-2 
 
 18-35 
 
 22-6 
 
 22-6 
 
 18-0 
 
 17-95 
 
 22-3 
 
 23-0 
 
 180 
 
 18-15 
 
 21-9 
 
 22-5 
 
 18-1 
 
 18-4 
 
 22-3 
 
 22-7 
 
 17-8 
 
 18-9 
 
 21-8 
 
 231 
 
 
 
 221 
 
 21-8 
 
 
 
 dark adaptation and the vision of tlie totally colour-blind will 
 also be apparent from the next two tables. The first gives the 
 intensity values of the different parts of the spectrum for a 
 normal "dark" eye, and the second Abney's observations on 
 two other cases of total colour-blindness. The units of inten- 
 sity in these tables are not comparable, but it will be observed 
 that the maxima occur in the same part of the spectrum. 
 
 " TicilvjU " Values of a 
 
 Spectrum (Schaternikoff). 
 
 Wave length. 
 
 Intensity Value. 
 
 Wave Length. 
 
 Intensity Value. 
 
 In milllonths of a 
 
 
 In millionths of a 
 
 
 mm. 
 
 
 mm. 
 
 
 670-8 
 
 18-0 
 
 529-3 
 
 2736-0 
 
 651-8 
 
 36-5 
 
 522-3 
 
 2532-3 
 
 634-3 
 
 83 3 
 
 515-4 
 
 2219-3 
 
 618-1 
 
 216-9 
 
 508-7 
 
 1944-0 
 
 603-1 
 
 423-2 
 
 502-2 
 
 1475-8 
 
 589-3 
 
 881-7 
 
 490-0 
 
 1016-0 
 
 577-1 
 
 1424-9 
 
 478-6 
 
 6330 
 
 566-4 
 
 2110-7 
 
 468-0 
 
 364-5 
 
 556-0 
 
 2609-7 
 
 458-7 
 
 208-8 
 
 546-0 
 
 2899-0 
 
 451-1 
 
 111-2 
 
 537-2 
 
 3000-0 
 
 443-9 
 
 69-6 
 
 Since the direct evidence which we studied first suggested 
 that adaptation was an affair of the extra foveal part of the 
 retina, one might expect the fovea centralis in the totally 
 colour-blind to be relatively insensitive. As a matter of fact, 
 in seven out of eighteen cases which have been investigated, 
 an absolute or relative central or para-central scotoma was 
 
VISUAL ADAPTATION 
 
 111 
 
 observed, and in most cases very imperfect fixation or even 
 nystagmus was noticed. It may, however, be regarded as 
 certain that the existence of such an insensitive area is not 
 necessarily found in typical total colour-blindness ; as we shall 
 see later, the point has been laboured chiefly for theoretical 
 reasons. 
 
 Luminosity Values of tivo Cases of Total Colour-blindness (Ahney). 
 (No. 40 in Abney's scale is close to the B line.) 
 
 Scale of Spectrum. 
 
 K. B.'B Luminosity 
 Value. 
 
 P.'s Luminosity Value. 
 
 56 
 
 2-5 
 
 
 54 
 
 9-0 
 
 
 52 
 
 160 
 
 7-6 
 
 50 
 
 27-5 
 
 19-0 
 
 48 
 
 42-5 
 
 39 
 
 46 
 
 61-0 
 
 65-0 
 
 44 
 
 82-5 
 
 85-0 
 
 42 
 
 96-0 
 
 98-0 
 
 40 
 
 100-0 
 
 99-0 
 
 38 
 
 95-5 
 
 915 
 
 36 
 
 87-5 
 
 90-0 
 
 34 
 
 75-0 
 
 80-0 
 
 32 
 
 61-5 
 
 65-0 
 
 30 
 
 43-0 
 
 50-0 
 
 28 
 
 37-0 
 
 36-0 
 
 26 
 
 30-0 
 
 26-5 
 
 24 
 
 24-0 
 
 19-5 
 
 22 
 
 18-5 
 
 14-0 
 
 20 
 
 14-5 
 
 100 
 
 18 
 
 11-5 
 
 ... 
 
 16 
 
 9-0 
 
 5 '5 
 
 14 
 
 7-0 
 
 ... 
 
 12 
 
 5-0 
 
 
 In general terms, we may say that the analogy betAveen 
 normal vision under conditions of dark adaptation and the 
 vision of the totally colour-blind is quite close. In the next 
 chapter some additional evidence will be considered, and the 
 theory of the matter discussed. 
 
 RECOMMENDED FOR FURTHER STUDY 
 Sufficient indications to enable the reader to consult original sources 
 of information will be found in the article Visual Adaptation, by 
 M. Greenwood, Further Advances in Physiology, London, 1909 (Arnold), 
 pp. 351-377. Of the literature there referred to, the memoir by 
 Tschermak is the most exhaustive. 
 
CHAPTER XIII 
 
 RECURRENT VISION-THEORIES OF ADAPTATION 
 
 A SET of experiments of apparently little importance has 
 thrown considerable light on the question of adaptation. 
 These deal with the effects which follow the application of 
 luminous stimuli for very short intervals of time. 
 
 Such experiments can be carried out in at least two ways. 
 By a contrivance similar to the shutter of a camera the eye 
 can be stimulated for a very short time, or, the gaze being 
 fixed, a source of light may be moved across the field of 
 vision. In the latter method, a disc with a slit in it can be 
 rotated in front of a lantern. If the length of the slit is I, 
 and V is the velocity of movement, l/v measures the time 
 during which each retinal element is exposed to the light, 
 and we can make this time as short as we please. There is 
 of course no difference in principle between the two methods, 
 but the second is rather easier to employ, and the results 
 obtained with it are of special interest. 
 
 If a bright object is rotated in this manner on a screen 
 with a slit in front of a light, the whole sensory effect com- 
 prises the following phases : — 
 
 (1) A primary image ; the immediate consequence of the 
 stimulus, also its strongest effect. As compared with the 
 image due to a stationary illuminant, it is more or less elon- 
 gated into a streak of light. 
 
 (2) Immediately following upon the primary image is a 
 short, dark streak. 
 
 (3) After the dark streak comes a second period of illu- 
 mination which, if the stimulus be coloured, appears tinged 
 with the complementary. In a successful experiment this 
 effect may be so pronounced that it seems as if a second 
 object were following in the track of the first, so that it has. 
 
 112 
 
RECURRENT VISION 113 
 
 been termed the "satellite" or "ghost," and the whole 
 phenomenon is described as that of "recurrent vision." 
 
 (4) The end of the satellite is not sharply defined, and is 
 followed by another interval of darkness. 
 
 (5) The field once more brightens, but somewhat faintly, 
 and an image of the same colour as the primary one appears. 
 
 (6) Lastly, another dark interval is obtained. 
 
 It would hardly be expected that there should be com- 
 plete accord in the description of so complicated an effect 
 as witnessed by different observers. Hess and v. Kries, to 
 mention two highly trained and experienced sense physio- 
 logists, differ materially in their accounts as to what can and 
 what cannot be seen. 
 
 The work of v. Kries has received, in all essential points, 
 confirmatioijj^from the experiments of Macdougall, and I 
 shall follow Jiis description as closely as possible. 
 
 With respect to the experiment as a whole, we have three 
 phases of illumination — the primary image, the satellite 
 image, and the tertiary. In apparent brightness these are 
 ranged in the order of their appearance. When the original 
 stimulus is of low intensity, no tertiary image is obtainable ; 
 with a still less intensity, the satellite also disappears. The 
 lengths of the images can also be made to vary, and the dark 
 intervals to vanish. Fixing our attention on the primary, 
 the following characters have been found. The image is 
 sometimes striped, but apart from this is uniform when 
 viewed under conditions of light adaptation. As dark 
 adaptation proceeds the image not only increases in ex- 
 tension and brightness, but with chromatic stimuli ceases to 
 be uniform. Thus with blue light, only the anterior border 
 is deep blue, and is followed by a whitish stripe. Macdougall 
 found that the white part begins at a distance corresponding 
 in his experiments to a time interval of ^ second. With 
 other colours, except red, the same result is obtained, but less 
 distinctly. 
 
 The satellite image begins J-i second after the com- 
 mencement of the primary, and is generally complementary 
 to it. This rule must, however, be modified in the following 
 way. If the primary is pure white, the secondary is bluish ; 
 
 H 
 
114 PHYSIOLOGY OF THE SPECIAL SENSES 
 
 in fact, the secondary is always modified in tlie direction of 
 bluishness. Even -with a feeble blue primary, the secondary 
 may still have a faint bluish tinge. As regards brightness, 
 the result depends on the adaptive value of the stimulus, i.e. 
 two lights of equal stimulus values for the " dark " eye give 
 equally bright secondary images. Red, with its relatively 
 low stimulus value for the " dark " eye, only gives a secondary 
 when its physical brightness is great. 
 
 As we should expect from its character, the satellite is 
 largely dependent on the adaptation of the eye, increasing to 
 a maximum as dark adaptation proceeds, and then diminish- 
 ing, although it persists after prolonged dark adaptation 
 (Macdougall). We associate, therefore, with the satellite the 
 ordinary characters of peripheral vision, (1) relatively greater 
 efficiency of the shorter waved lights, (2) increased intensity 
 on dark adaptation ; it only remains to add failure on foveal 
 stimulation. With respect to this failure, a somewhat heated 
 discussion took place between v. Kries and Hess, the latter 
 asserting that a distinct satellite could be obtained in central 
 vision. V. Kries has, however, indicated certain possible fal- 
 lacies in Hess' technique, and his assertion that no secondary 
 can be produced at the fovea centralis has been confirmed 
 by the work of Hamaker and of Macdougall. Passing to the 
 tertiary image, the following characters may be noticed. The 
 hue is best appreciated when red is chosen as the stimulus ; 
 in such a case, the tertiary may be very distinct. With in- 
 creasing dark adaptation, the tertiary gains in brightness but 
 loses in chromatic value ; owing to the high adaptive values 
 of green and blue when these lights are used at moderate in- 
 tensities, the coloration of the tertiary image can only be 
 seen at the beginning of the experiment. There is some 
 difference of opinion as to whether this tertiary image can be 
 seen in direct vision ; since red light is the most suitable 
 stimulus for calling it up, it should be so perceived. Perhaps 
 there are two factors in the production of the tertiary — 
 a chromatic element unaffected by dark adaptation, and a 
 brightness element which is so affected. 
 
 The last group of facts to which I desire to refer in this 
 connection are those which have to do with observations 
 
KECURRENT VISION 
 
 115 
 
 upon the pupillary reaction. Schirmer asserted that the 
 pupil width and its reaction were related to the adaptive 
 condition of the eye. With complete adaptation to a given 
 grade of light, the pupil reaches after an initial widening or 
 narrowing a physiological mean position. Garten found that 
 momentary illumination produced in "light" eyes a weak 
 sudden, in the " dark " eye a slow powerful contraction. Pro- 
 ceeding further on these lines, Sachs discovered that the 
 pupillo-motor response to coloured lights followed closely 
 their adaptation values. Abelsdorff confirmed these results, 
 using the apparatus sketched. 
 
 Triplea: Zamp. 
 
 ,SUt_ in Telescope. 
 -Sqiccratbw Prismi. 
 
 - CyHndior of Blacked- Paper 56 c.th. long. 
 
 ■QmvesD L ens. Focal Length 10 c.tw. 
 
 ^^ UncuxommodateA Eye. 
 
 Fig. 13. — AbelsdorfE's apparatus for studying the pupillo-motor response. 
 
 One of the two lights serves as a standard ; the subject, 
 looking through the strong convex lens at the slit, sees a 
 bright point surrounded by a diffusion circle. If the light 
 falling on the eye be changed, the diffusion circle increases 
 or diminishes in size. By rotation of a Nicol prism the new 
 intensity can be increased or diminished, and the intensity is 
 found for which the diffusion circle produced by the standard 
 
116 PHYSIOLOGY OF THE SPECIAL SENSES 
 
 light is not increased or diminished by changing to the 
 tested colour. This method is more accurate, in Abelsdorff's 
 opinion, than one might suppose, the mean experimental 
 error being about 7 per cent. The next table gives some of 
 Abelsdorff's results : — 
 
 Pupillo-Motor Values. 
 
 Wave Length. 
 MM. 
 
 Comparison Lighta. 
 
 600 ix^. 
 
 Light Adaptation. 
 480 MM- 
 
 Darlf Adaptation. 
 480 (ifi. 
 
 640 
 620 
 600 
 580 
 560 
 540 
 520 
 500 
 
 Mm. 
 •5271 
 •8523 
 •9720 
 •9536 
 ■8303 
 •5518 
 •3333 
 •1181 
 
 Mm. 
 •3920 
 •8376 
 •9822 
 •9090 
 •8739 
 •6141 
 •2936 
 •09141 
 
 Mm. 
 •2666 
 •5670 
 •7260 
 •8065 
 •8865 
 •9200 
 •5755 
 •1612 
 
 Brightness Values. 
 
 (Readings obtained when the same subject adjusted the intensities 
 of the lights until they seemed to be equally bright.) 
 
 Wave Length. 
 
 Comparison Lights. 
 
 
 
 
 Iili.. 
 
 600 ja/i. 
 
 Light Adaptation. 
 
 DarlJ Adaptation. 
 
 
 
 480 ^fi. 
 
 480 II.H.. 
 
 
 Mm. 
 
 Mm. 
 
 Mm. 
 
 640 
 
 •5253 
 
 ■3518 
 
 •2529 
 
 620 
 
 •8204 
 
 •7230 
 
 •5515 
 
 600 
 
 •9431 
 
 •9090 
 
 •8536 
 
 580 
 
 •9431 
 
 •9090 
 
 •8535 
 
 560 
 
 •8811 
 
 •8613 
 
 •9540 
 
 • 540 
 
 •6259 
 
 •6354 
 
 •9540 
 
 520 
 
 •3700 
 
 •3189 
 
 •5750 
 
 500 
 
 •1168 
 
 •0944 
 
 •1612 
 
 The close agreement between the pupillo-motor values and 
 those of apparent brightness justifies the method, and its 
 importance lies in the fact that we can employ it in experi- 
 
EECURRENT VISION 117 
 
 ments on animals. We have no direct means of investiga- 
 ting adaptive changes in any animal except man, but we 
 can measure this pupillary response ; if we find it changing 
 in the manner described, it is suggested (not, of course, 
 demonstrated) that in such animals the visual responsiveness 
 may be similarly affected. It has been found that the 
 intensities of red and blue, which appeared equally bright to, 
 and exerted the same pupillo-motor effect upon, a human 
 " light " eye, did not produce identical changes in the pupils 
 of the dove and the owl. For the former the red, for the 
 latter the blue, was the stronger stimulus. Indeed, the 
 pupillo-motor response to blue in the owl's eye was greater 
 than in the case of a total-colour-blind (Abelsdorff). 
 
 We can now appropriately examine the theories which 
 have been propounded to describe these different classes of 
 fact. In the first place, can we justly say that there is any 
 functional difference between the spot of distinctest vision, 
 the fovea centralis retinae, and the paracentral or peripheral 
 regions of the retina in respect of colour vision ? 
 
 In view of the long series of experiments bearing upon the 
 Purkinje effect, the increase in brightness of the short-waved 
 spectral lights at the expense of the long- waved vibrations, 
 it seems clear that this effect is mainly if not entirely 
 peripheral. Even those who claim to have observed foveal 
 adaptation admit it to be much less marked than peripheral 
 changes of responsiveness. We have also seen that these 
 adaptive changes consist in a greatly increased responsive- 
 ness to light of short wave length, such appearing more 
 intense than under " light " conditions. We have to account 
 theoretically for a localised change in responsiveness with 
 respect to certain forms of stimulus. 
 
 The difference in histological structure between the fovea 
 centralis and the surrounding area caused Schultze, many 
 years ago, to suggest a functional separation, a conclusion 
 which he supported by evidence drawn from the study of 
 comparative anatomy. When Schultze wrote, however, our 
 knowledge of adaptive changes was little advanced, and his 
 conception went unheeded. , The modern development of 
 Schultze's hypothesis is due to the independent researches of 
 
118 PHYSIOLOGY OF THE SPECIAL SENSES 
 
 H. Parinaud and J. v. Kries, both of whom, together with 
 their colleagues and pupils, have published numerous 
 memoirs on the subject. The actual priority appears to 
 belong to Parinaud, whose first paper was published in 1881. 
 
 Essentially the theories of Parinaud and v. Kries are 
 exceedingly simple. Two distinct visual mechanisms exist : 
 of these, one is concerned in the elaboration of chromatic 
 and achromatic stimuli, and is represented in the retina by 
 the cones; the other mechanism deals with achromatic 
 responses only, and is represented by the rods and visual 
 purple. The former mechanism is the only one which can 
 act in bright light, and its responsiveness is little, if at all, 
 increased by resting in darkness ; the latter is brought into 
 play when the eyp has been shielded from stimulation, being 
 the sole or chief agency for twilight vision ; it is character- 
 ised by special responsiveness to ethereal vibrations of short 
 wave length. In view of the double nature of the 
 mechanisms postulated, the theory has been named the 
 " Duplicity Theory " (" Dtlplizitatstheorie "). Let us see how 
 far the hypothesis covers the experimental observations I 
 have enumerated. If the theory were true, we should expect 
 (1) spectral maximal brightness to change in favour of the 
 violet end when the physical intensity of the light is 
 diminished ; (2) this change ought not, however, to occur 
 in images formed at the fovea centralis retinse ; (3) no 
 achromatic threshold ought to be obtained for any light at 
 the fovea centralis or for red light anywhere. 
 
 Each of these deductions has been shown to receive 
 support from experiment and observation. 
 
 Another way of testing the theory is to see whether we 
 have any forms of vision in which, apparently, the basal 
 mechanism is similar to that associated by the theory with 
 twilight vision, and uncomplicated by any other type of 
 vision. The subjects of total-colour-blindness appear to 
 afford us a case in point. We found that the brightness 
 judgment of these people agree well with those of normal 
 men in a state of dark adaptation; that there is evidence 
 in such cases of diminished or absent foveal responsiveness, 
 bad fixation, inferior acuteness of vision, nystagmus and 
 
KECURRENT VISION 119 
 
 abnormally good vision in twilight. This might well be 
 regarded as a case in which the twilight mechanism alone 
 is operative. Conversely, one pathological condition is 
 consistent with the activity of the hypothetical daylight 
 mechanism existing by itself. This is the condition of 
 "night-blindness," or hemeralopia, as it is badly called, which 
 has been investigated by Parinaud, Nettleship, Messmer, and 
 others. According to Parinaud, subjects of this disease have 
 vision of the foveal type ; their colour sense is normal, but 
 the spectrum is shortened at the violet end, and responsive- 
 ness to feeble stimuli is abnormally poor. It is the latter 
 condition which incapacitates a person who exhibits the 
 peculiarity to a marked degree from working in twilight or 
 poor artificial light. The investigations of Messmer have 
 made it seem very improbable that night-blindness is a 
 simple condition. In some cases, dark adaptation is very 
 slowly induced, but after a sufficiently long time attains a 
 normal degree of intensity. In other cases, a certain degree 
 of adaptation is produced in the normal time, but the 
 amount is far less than in ordinary persons. Owing to the 
 complexity of the condition, we must admit that the light 
 it throws on our problem is not so great as we could wish. 
 Provisionally, we may perhaps say that some cases of night- 
 blindness can be interpreted by supposing that the hypo- 
 thetical daylight mechanism is alone functioning, while most 
 of the examples of total-colour-blindness agree with the 
 supposition that the twilight mechanism is chiefly at work. 
 It must be pointed out that there cannot possibly be a com- 
 plete identity between the normal peripheral mechanism 
 and the visual system of a total colour-blind. Poor as is 
 the latter's visual acuity, it is far superior to that of the 
 peripheral parts of the normal retina. 
 
 The complex results in sensation which are due to short 
 or moving stimuli have perhaps confused some readers ; let 
 us see whether our hypothesis is capable of arranging them 
 in an orderly manner. The peculiar striping of the primary 
 image has already been mentioned, how blue tails ofi' into 
 white ; this suggests the interplay of two processes. I also 
 emphasised the dependence of the secondary image on 
 
120 PHYSIOLOGY OF THE SPECIAL SENSES 
 
 adaptation, and its probable absence at the fovea centralis. 
 We can perhaps sum up the effects in terms of our hypo- 
 thesis thus : — 
 
 The cone mechanism responds by two effects, the main 
 part of the primary and the colour component in the tertiary. 
 The rod apparatus responds in a threefold manner; it 
 gives us the white tail of the primary, the whole of the 
 secondary, and contributes, although slightly, to increasing 
 the brightness of the tertiary. This description clears up 
 some difficulties, but raises others. If we are to regard the 
 twilight mechanism as solely responsible for the secondary 
 image, it is clear that we derive sensations of colour as 
 well as sensations of luminosity without hue from that 
 mechanism, since the secondary image is often coloured. 
 The mechanism cannot therefore be identical with that of 
 a totally colour-blind eye ; we must either give up the view 
 that total colour-blindness is a condition in which the rods 
 and purple react as in a normal person, or regard the 
 secondary as due to something beyond rod stimulation. 
 The latter alternative is the one v. Kries is disposed to adopt, 
 but we have then the difficulty of understanding why the 
 secondary is entirely peripheral. Are we to suppose that 
 there is a functional difference between the central aad 
 peripheral cones ? If not, why in this case can the cones 
 only respond if the rods are at the same time stimulated ? 
 
 Probably this will be found the most serious difficulty 
 in the way of a complete acceptance of the duplicity hypo- 
 thesis. It is hoped, although not expected, that further 
 experimental work will clear the matter up. 
 
 A certain amount of confirmatory evidence has been 
 brought forward from the side of comparative physiology. 
 It has long been known that the relative numbers of rods 
 and cones differ in various animals. Thus, the rods are 
 very large and almost exclusively present in the retinae of 
 nocturnal animals, such as owls, bats, and hedgehogs. In 
 many other creatures, on the other hand, including most 
 birds, cones predominate. It was, indeed, on the strength 
 of this that Schultze advanced a theory essentially similar 
 to that of Parinaud and v. Kries. Ktlhne subsequently 
 
RECURRENT VISION 121 
 
 showed that visual purple was present only in retinae con- 
 taining rods, although he was not able to extract the pigment 
 from all rod-containing eyes, an exception being the bat. 
 Trendelenburg has recently extracted visual purple from 
 the retinae of more than one species of bat. Experimentally, 
 as we have seen, Abelsdorff found that, judging by the pupillo- 
 motor response, the owl is specially sensitive to short-waved 
 light and the dove insensitive. 
 
 We all know that most nocturnal animals can see badly 
 in broad daylight, while such birds as the pigeon exhibit 
 a marked degree of night-blindness. In Parinaud's words : 
 " It is a matter of common observation that hens and 
 pigeons see very imperfectly in artificial light, and defend 
 themselves with difficulty against the hand that tries to 
 seize them ; that as soon as the sun goes down these animals 
 seek their night shelter, the old adage, ' To go to bed 
 with the hens,' meaning to go to bed early, evidently having 
 its origin in this fact." ^ 
 
 Biological investigation appears to show, therefore, a 
 <;o-existence of rods and visual purple with vision of the 
 twilight type and of cones with optimal vision in daylight. 
 Reasoning from analogy is, however, proverbially dangerous, 
 and the importance of the facts described can easily be 
 ■exaggerated. We have, and can have, no direct knowledge 
 of the actual type of vision possessed by the animals men- 
 tioned; in some, notably the bat, it is a matter of doubt 
 how far vision is the true directive sense under twilight 
 conditions. The early roosting of diurnal birds may also 
 be due to causes other than a state of night-blindness. 
 
 The whole case can be summed up in the following way : — 
 
 (1) There is a marked difference between central and 
 peripheral vision in regard to the phenomenon of darkness 
 adaptation, the former being little, if at all, affected in the 
 process. 
 
 (2) These differences may be provisionally interpreted on 
 the hypothesis that visual sensations are bound up with 
 two distinct mechanisms : (a) That of the cones, with which 
 
 ' La Vision, by H. Parinaud, Paris, 1898, p. 66. 
 
122 PHYSIOLOGY OF THE SPECIAL SENSES 
 
 chromatic responsiveness and colourless sensori-reactions in 
 bright light are associated ; (6) that of the rods, upon which 
 depend achromatic reactions under conditions of darkness 
 adaptation. 
 
 The objections to this view are neither few nor unim- 
 portant. It has not been proved that no central adaptation 
 occurs. The equations (colour matches) of totally colour- 
 blind persons and those of the normal " dark " eye agree 
 well, but not completely. 
 
 The interpretation of the secondary image of recurrent 
 vision is not complete. 
 
 The first of these objections, as well as the kindred one 
 that a central scotoma does not exist in all cases of total- 
 colour-blindness, may be parried by supposing that a trace 
 of visual purple and a few scattered rods are present in the 
 fovea. Recent measurements by Fritsch give an absolutely 
 rod-free zone of only "2 mm., corresponding to an angular 
 distance of less than a degree. We must not attach much 
 weight to failures in the demonstration of such small non- 
 adaptable areas, even supposing that they are completely 
 rod-free. 
 
 The difficulty regarding " total-colour-blindness " and 
 normal "dark" equations is not formidable. Sufficient 
 measurements have not been made to enable us to affirm 
 that the differences are significant. The most serious diffi- 
 culty as to the secondary image of recurrent vision has 
 already been discussed; it may prove the crucial point of 
 the theory. 
 
 If one adopts the above view of the role of the visual 
 purple and rods as elements in the physiological processes 
 of vision with low intensities of light, one is tempted to 
 speculate as to the nature of their activity. We have, 
 however, only negative evidence. We can say with some 
 confidence that retinal fluorescence is not an important 
 factor in the process ; beyond this we cannot at 
 present go. 
 
 In conclusion, the reader must clearly understand that 
 the cones of the peripheral retinas are not functionless, that 
 peripheral vision in daylight is quite different from the type 
 
RECURRENT VISION 123 
 
 studied in this chapter. The following table shows this 
 very distinctly : — 
 
 Na line = 100. 
 Wavelength. . 680 651 629 608 589 573 558 530 513 
 Peripheral value 
 
 daylight . . 9-6 37-5 77'5 101 100 79-6 52-2 28o 14-6 
 Peripheral value 
 
 twilight . . ? 3-4 14-0 35-5 100 256 351 321 198 
 
 KEOOMMBNDED FOR FURTHER STUDY 
 
 The reader should consult the bibliography on pp. 376-7 of " Further 
 
 Advances in Physiology." Of the papers mentioned, the most 
 
 important are — 
 Die Helldunkeladaptation des Auges und die Funktion der Stabchen 
 
 und Zapfen, by A. von Tschermah, Ergebn. d. Physiol., 1st Jahrg. 
 
 pt. h. pp. 695, etc. 
 /. von Kries, Die Gesichtsempfindungen, Nagel's Handb., vol, iii. pp. 
 
 109-282. 
 On Recurrent Vision, Macdoiigall, British Journ. of Psychol., i. 78 ; 
 
 V. Kries, Z.P.P.S.O., xxix. 81 ; Hess, ibid., xxvii. 1. 
 
CHAPTEE XIV 
 
 TRICHROMATIC VISION 
 
 Up to the present we have considered visual sensations 
 from the standpoint of pure comparison. Without analysing 
 either the sensations themselves or their physical forerunners, 
 we have been content to ascertain whether a physical agent 
 of approximate constancy is or is not followed by the same 
 phenomenon in consciousness irrespectively of the exact region 
 stimulated or the general experimental conditions (light or 
 dark adaptation). We have found that the end product 
 does vary with the part of the retina examined and the 
 latter's adaptive state. We now enter upon a wider and 
 more difficult part of our inquiry, namely, a presentation 
 and arrangement of the chief physiological facts of visual 
 sensation. The difficulty of this task is apparent when one 
 considers what is involved. We must attempt to describe 
 the interplay of a complex of factors, physical, physiological, 
 and psychological ; some of these are fairly well understood, 
 others hardly at all, so that we can frequently do little more 
 than echo the words of a great master of the subject — 
 " The confession of actual doubt is better than the delusion 
 of dogmatic certainty." ^ 
 
 In the first place, are visual sensations solely dependent 
 upon their physical forerunners ? Under like physiological 
 conditions and with equal physical stimuli are the responses 
 identical ? Reflection makes it clear that the answer is — no ; 
 what we term the colour of an object is not a simple percept, 
 but a construct. The difficulty has been admirably put by 
 Hering, whose remarks deserve close attention. He writes : 
 "The layman is convinced that external objects possess 
 definite colours, that snow is white, soot black, and gold 
 yellow. He attributes to these colours an existence inde- 
 
 1 Helmholtz, Eandh. d. Pliys. Opt., p. 379. 
 
 124 
 
TRICHROMATIC VISION 125 
 
 pendent of the eye, and characterises them as the real 
 {wirldiche) colours of the respective objects, distinguishing 
 them from the accidental {zufdllige) colours which the same 
 objects exhibit under unusual circumstances, e.g. with insuf- 
 ficient illumination, or illumination differing very markedly 
 from ordinary daylight. The red of the mountain peak in 
 the Alpine glow ; the corpse-like pallor of a face illuminated 
 by the sodium flame ; the bright spots on the floor of a room 
 into which the sunlight streams through bright window- 
 panes, are instances of such occasional colours which we 
 attribute to a corresponding peculiarity of the illumination, 
 and do not regard as characteristics of the objects in ques- 
 tion. One who has learned that snow owes its whiteness 
 and gold its yellowness to the particular rays of daylight 
 reflected by each, easily forms the opinion that the " real " 
 colour of external objects must be black, since he regards 
 black, erroneously, as being due to a complete absence of 
 light rays. Yet whenever he thinks of the snow, he always 
 pictures it as white ; and so do all of us, whether we have 
 thought much or little about the nature of colours. Thus-the 
 mineralogist, to whom snow is a heaping together of little 
 colourless transparent water crystals ; the chemist, for whom 
 these crystals are composed of countless atoms and mole- 
 cules: the physicist, looking beyond atoms and molecules 
 to forms of energy — all perforce associate with the concep- 
 tion of snow the white colour." ^ 
 
 It amounts to this then, the colour sensations usually 
 experienced when we are stimulated by certain objects are 
 supposed by us to be necessarily connected with these 
 objects. ''What the layman calls the real colour of an 
 object is a colour of the object constantly present in his 
 memory," and may be termed a " memory colour." The 
 importance of such colours is obvious; they render it im- 
 possible for us to describe visual sensations completely in 
 terms of any physical measurements, since the sensation 
 value of a stimulus is not solely dependent on that stimulus 
 considered by itself. Two stimuli of equal value — physically 
 
 ^ Hering, Grundziige der LeJire 1:0m Lichtsinn, Leipzig, 1905, p. 6. 
 
126 PHYSIOLOGY OF THE SPECIAL SENSES 
 
 speaking — will not necessarily elicit the same sensation even 
 if compared under approximately the same physiological 
 conditions. The different results which may follow the 
 application of a constant stimulus are thought to be demon- 
 strated by an experiment devised by Hering i (Fig. 14). 
 
 If the apparatus be arranged as indicated in the diagram, 
 with a suitably chosen artificial light, both papers appear 
 brown, the yellow working rays reflected from the blue paper 
 overpowering the blue rays. If the window be now shut 
 and, without altering the artificial light, the two pieces of 
 paper be removed from the photometer and examined, it 
 
 ^^ 
 
 D 
 
 Fig. li.—A. Wooden Photometer Tube. B. Cylinrier with Wooden Prism 
 having Blue and Brown (2 and 1) Papers on either side. D. Artificial 
 Light. C. Mirror reflecting daylight. 
 
 -will be seen that the "really blue" paper, although it is 
 sending to the eye precisely the same mixture of rays as 
 before, looks blue just as in daylight, only perhaps a little 
 darker, while the brown paper still looks brown. In Hering's 
 opinion, the constancy of colours in objects is one of the 
 most striking phenomena in the field of visual sensations. 
 Without it a piece of chalk on a dull day would have the 
 same colour as a piece of coal on a sunny day, and, in the 
 course of a single day, would pass through a remarkable 
 series of gradations between white and black. The conse- 
 quence would be that we should no more attribute whiteness 
 to chalk or blackness to coal than hotness or coldness to iron. 
 Interesting as are these reflections, it is permissible to 
 
 ' Hering, op. eit, p. 15. 
 
TEICHROMATIC VISION 127 
 
 doubt whether the invocation of " memory colours " is neces- 
 sary to account for the constancy of colour phenomena. 
 It is equally reasonable to suppose that chalk never on 
 the dullest day resembles coal, because relatively to coal it 
 always reflects more white light. Hering's experiment ap- 
 pears to me to involve too many factors for us to regard it 
 as demonstrating his theses. Even, however, if we accept 
 Hering's view, it by no means follows that we ought to dis- 
 card experimental methods founded upon physical measure- 
 ments. In quite popular language, the position may be 
 stated as follows : We look at two scraps of paper, and say 
 that in one case we experience a sensation of blueness and 
 in the other one of redness; we then find that vibrations 
 of one type are predominately reflected by the one and of 
 another type by the other object. Although we must not 
 say that this difference in vibration period is the only cause 
 of the sensational difference, it may be convenient to use 
 it as some criterion. This experimental convenience is some 
 justification for saying that colour tone depends on wave 
 , length, colour saturation on wave purity, and intensity on 
 wave energy. I hope to show that, by adopting this con- 
 vention, we can simplify our description of the facts without 
 seriously prejudicing their theoretical interpretation. I shall 
 therefore pursue the following order- — the empirical methods 
 and results of colour-mixing and the phenomena of after- 
 images will be described; then the theoretical deductions 
 which have been made from these results, together with 
 experiments specially designed to test these deductions ; and 
 subsequently certain peculiar and special phenomena of 
 colour vision will be considered. 
 
 The physical basis of this kind of work is to be found in 
 the conception of a simple or homogeneous light, and dates 
 from Newton's researches on prismatic analysis. By homo- 
 geneous light we are to understand ethereal vibrations, the 
 wave lengths of which fall within certain narrow limits. 
 Prisms and gratings enable us to filter such lights from 
 a mixture, and to employ them for our experiments. A 
 pure light is uniquely defined by its wave length, and any 
 such light may possess any intensity. The physical unit of 
 
128 PHYSIOLOGY OF THE SPECIAL SENSES 
 
 intensity is not easily fixed, perhaps the best attempt being 
 that of Krarup,^ who has applied the energy measurements 
 of Angstrom. In any spectrum the intensities of the in- 
 dividual lights depend on the source of illumination and 
 the method of analysis, i.e. on the extent of surface over 
 which light of a given wave length is spread. In an " in- 
 terference " spectrum diffusion is uniform ; in a prismatic 
 spectrum, on the other hand, it increases from red to violet, 
 so that the short-waved light is relatively less intense than 
 the long-waved. Spectra obtained by the two methods are 
 accordingly not directly comparable. 
 
 By colour or light mixing we understand an arrangement 
 by which two or more homogeneous lights fall upon the 
 same retinal area. That our experiments may be pure we 
 
 Fig. 15. 
 
 have to use pure light of constant intensity, which often 
 necessitates the use of complex optical apparatus. Pigments 
 cannot be used, because the light proceeding from a mixture 
 of pigment is not equal to the sum of the lights emitted 
 by the constituents, owing to selective absorption. A simple 
 method, sufficient for many purposes, is the rotation of 
 coloured sectors on a colour mixer such as is generally used 
 for demonstrations. A more accurate method is that due 
 to Helmholtz. Sunlight reflected from a heliostat enters 
 a dark room by a slit, and then passes through a prism P 
 (Fig. 15) and an achromatic lens Lj. The screen S^ is placed 
 in the focal plane of the eyepiece, and a spectrum projected 
 upon its anterior surface. Between lens and screen a 
 diaphragm D is inserted, and the screen is furnished with 
 
 ' H. Krarup, Physisch-opthaXmologisohe GreueproUeme, Leipzig, 1906. 
 
TRICHROMATIC VISION 129 
 
 two vertical slits which allow definite portions of the spectrum 
 to pass through. A second achromatic lens Lg, of somewhat 
 shorter focal length, projects an image of the diaphragm 
 upon a second screen Sj. The aperture of the diaphragm 
 must be so small that each of the pure lights is distributed 
 over the whole of it. Under these circumstances the field 
 projected on the second screen is made up of a uniform 
 mixture of the two colours; colour tone and intensity are 
 varied by adjusting accurately the position and breadth of 
 the slits. 
 
 Other more elaborate and exact methods can be used, 
 but their description would be tedious. In most cases it is 
 desirable to obtain a comparison as well as a mixed field, 
 and this necessarily increases the difficulty of the arrange- 
 ment. 
 
 Examination of a spectrum teaches us that, apart from 
 extremes of intensity, change in wave length goes with 
 change in chromatic quality or colour tone, provided the 
 alteration in colour tone exceeds a lower limit. At the 
 ends of the spectrum, however, increase in wave length 
 above or diminution below a certain value is not accom- 
 panied by a corresponding sensation difference. The reason 
 of this limitation is perhaps physiological, but its discussion 
 must be omitted.^ 
 
 Experiments carried out by one of the mixing processes 
 just mentioned have enabled us to formulate certain general 
 statements which were first clearly enunciated by Grassmann. 
 These conclusions are as follows : If, in a mixture, one 
 component be continuously varied, the appearance of the 
 mixture will likewise vary (unequal lights mixed with equal 
 lights produce unequal mixtures). If two lights look the 
 same, then if each be mixed with a third light, the resultant 
 mixtures will look equal. This can, of course, be generalised. 
 A corollary is, that proportional increase of the intensity of 
 each component in a mixture does not destroy a match. 
 
 Passing to the actual observations, we at once note that 
 the effect of mixing spectral extremes is the production of 
 
 ^ See Krarup, op. cit., pp. 15-19. 
 
130 PHYSIOLOGY OF THE SPECIAL SENSES 
 
 a colour — purple — not present in the spectrum at all. It thus 
 follows that any graphical representation of our results must 
 take the form of a closed curve, since passing from red to 
 violet we can either travel over the range of spectral colours 
 or by way of purple. If we mix lights not belonging to 
 the extreme ends of the range, our results are quite different. 
 The simplest cases are those of mixing colours of wave 
 length not less than 540 /t/t. For instance, a red (670 ytt/t) 
 mixed with a yellow (580 /a/u,) gives a pure colour of in- 
 termediate wave length. The greater the proportion of 
 the long- waved component, the nearer will the position 
 of the mixture approach the red end of the spectrum, 
 and conversely. The mixing relations for this part of the 
 spectrum are therefore quite straightforward ; but the result 
 obtained — i.e. that two simple lights when mixed give a 
 simple light the wave length of which is intermediate between 
 those of its components — is only valid for a small part of 
 the spectral range. If we mix a blue-green (510 fifj,) with 
 a blue (460 fifi) the mixture, although resembling, perhaps 
 closely, a pure intermediate — e.g. 490 fi/j, — does not match 
 it perfectly. The mixed colour is paler, or, as we say, " less 
 saturated," than the spectral one. This is still more evident 
 when we choose our colours in such a way that the wave 
 length of one is greater and that of the other less than 
 617 fifi. If one constituent is taken a little nearer the 
 red than 560 /x/j,, and the other dimiirished in each ex- 
 periment, then, with suitable proportions, the mixtures pass 
 from greenish-yellow — becoming paler and paler — until we 
 reach a combination which corresponds to a sensation of 
 whiteness. As we tend to assign a unique position to white 
 in our sensation-scale, it is customary to complete the mixing 
 laws by the following statement: Any light mixture what- 
 ever can be matched by a mixture of a definite homogeneous 
 light (or a definite purple) and white light. That our re- 
 sults may be as general as possible, it is well to note that 
 there is no necessity for according a special place to white. 
 
 " If we regard any homogeneous or compound light what- 
 ever as fixed, and then mix it with the whole series of pure 
 lights completed by purple, at the same time varying the 
 
TRICHROMATIC VISION 131 
 
 proportions in the mixture from the zero of the one to that 
 of the other, we obtain all possible varieties of stimuli ; that 
 is, any possible combination is matched by some member of 
 the series." ^ 
 
 It is, however, convenient in practice to separate white, 
 and I shall continue to follow the usual classification. 
 
 We have seen that our mixing experiments give us varia- 
 tions in colour tone and variations in whiteness — that is, two 
 variables — so that our results should be expressible graphically 
 by some plane figure. We have also noticed that we can 
 pass from red to violet, and then from violet again to red, 
 without passing through the spectral range more than once, 
 so that our graph should be a closed figure ; finally, in virtue 
 of the fact that the position of any mixed colour depends 
 directly upon the relative proportions of its components, we 
 infer that the method of obtaining the position of the centre 
 of inertia of masses is applicable. If three colours, A, B, C, 
 none of which can be mixed from the other two, be repre- 
 sented by three points in a plane, then, on assigning to them 
 values in terms of any unit, the situations and quantitative 
 values of their mixtures can be ascertained. Thus a colour 
 mixed from a, parts of A and h parts of B will lie on the line 
 A B at the point at which the centre of inertia of the two 
 masses a and h (representing the proportions of the two 
 colours in the mixture) would be situated. In order to 
 establish the correctness of this method it is necessary to 
 prove that, given the experimental laws of colour-mixing, this 
 construction is valid in all possible cases, i.e. that the situa- 
 tion of the mixed colour coincides with that of the mass 
 centre of two equivalent masses (1) when the two constituents 
 ■can be mixed from the three chosen colours, (2) when one can 
 and the other cannot so be mixed, (3) when neither can so be 
 mixed. The proof is too long for insertion in a book of this 
 size, but it requires only an elementary knowledge of mathe- 
 matics for its comprehension.^ We can show that the 
 co-ordinates of the point at which the mixture must be situated 
 according to the mixing laws are the co-ordinates of the 
 
 1 J. V. Kries, Nagel's Handl., vol. iii. p. Hi'. 
 
 2 See Helmholtz, Phyi. Optil.:, second edition, pp. 328-3cO. 
 
132 PHYSIOLOGY OF THE SPECIAL SENSES 
 
 mass centre of weights situated at the points where the 
 components lie. 
 
 It is clear that a diagram constructed on these principles 
 will Tary in its actual form in accordance with our choice of 
 units and fixed points. Thus Newton chose white as a fixed 
 point, and arranged the simple lights at equal distances from 
 it, so that the diagram was a circle ; we should obtain this 
 result, except that the part of the curve passing from violet 
 to red must be a straight line, since purple can only be 
 mixed from violet and red, and its representation must be on 
 the chord joining their representative points. If we recur 
 to our experimental results, however, we shall see that the 
 best diagram to adopt is that figured 16. From red to 
 
 Greetv 
 
 YMow 
 
 \Vi6Ut 
 
 Fig. Ifi.— Coloar Table. 
 
 yellow the curve is a straight line, since we found that 
 mixtures of colours between those limits matched pure 
 spectral lights. We then get a sharp flexure in the part of 
 the diagram beyond green, representing the low saturation of 
 mixtures from this region, and finally we have the straight 
 line through purple. 
 
 Suppose now we select three colours in the spectral series, 
 e.g. red, green, and blue, then in accordance with our previous 
 deductions all mixtures of these are represented in the 
 triangle R, G, B (Fig. 17) ; this includes a good deal of our com- 
 plete diagram, but not all of it. If we choose red, green, and 
 violet, the triangle R, C, V includes nearly the whole spectral 
 diagram ; in other words, nearly all spectral colours can be 
 matched by mixing three chosen lights in suitable proportions. 
 
TRICHROMATIC VISION 133 
 
 There is not indeed a complete coincidence, because of the 
 bend in the spectral figure between green and blue, which 
 means that mixtures of green and violet are less saturated 
 than spectral cyan blue ; we can, however, generalise a little. 
 If spectral greenish-blue be mixed with red in certain propor- 
 tions, it matches a mixture of green and violet, or 
 
 aGBl+/3R = 7Gr + 6V. 
 
 G 
 
 Fig. 17. 
 
 Hence aGBl = 7Gr + eV — /3R, or we have the unmixable 
 colour in terms of our three chosen colours. I do not, how- 
 over, think this mode of expression desirable, since it is hardly 
 capable of objective interpretation. It is to be remarked 
 that colour equations, as they are termed, of the form 
 R + Gr+V = R, seem to be justifiable ways of expressing 
 experimental facts. Addition is uniform, the same result 
 being always obtained when the same quantities are summed ; 
 it is commutative, the order of operations does not affect the 
 result; it is associative and homogeneous. These characters 
 are found in our mixing process. Green + (mixed with) red 
 + (mixed with) violet = (match) red + green + violet = (red 
 + violet) + green. If we define subtraction, in terms of 
 urithmetical quantity, as uniform, non-commutative, and 
 non-associative, similar analogies can be observed ; but this 
 is of little importance, since a justification of the use of the 
 symbol of addition will suffice for our purposes. 
 
 We can now consider some experimental points. Since, 
 practically, all chromatic stimuli may be expressed in terms 
 of three, researches are conducted in the following way : A 
 definite spectrum — e.g. the prismatic spectrum of an Auer 
 gas lamp and three lights — e.g. a red, a green, and a blue of 
 arbitrary but constant intensity are selected. Each part of the 
 
134 PHYSIOLOGY OF THE SPECIAL SENSES 
 
 spectrum is then matched by mixing the three together 
 until the mixed colour looks exactly like the homogeneous 
 one. This process of spectral gauging has been applied by 
 Koenig and Dieter ici ^ to the investigation of normal colour 
 vision, and their results are generally regarded as correct. 
 Experiments involving the manipulation and mixture of 
 three pure lights are beset wiih technical difficulties, and the 
 results are not entirely free from objection, so that I do not 
 propose to give details, especially in view of the fact that 
 we can, for many purposes, obtain almost as much information 
 by another method. All I desire to emphasise is the general 
 conclusion that the experimental facts of colour vision can be 
 graphically represented by a plane figure, and the possible 
 forms of stimuli reduced to terms of three independent vari- 
 ables — that is to say, normal colour vision is trichromatic. 
 
 It is important to notice that a choice of variables is, from 
 the theoretical standpoint, arbitrary; indeed, if a table be 
 constructed in terms of three primaries, A, B, C, a second 
 can be deduced in terms of another three, A', B', C, because 
 in view of what has gone before, it is plain that we can 
 always define A' by an equation of the form A' = aA + &C + cC, 
 and similarly for the other variables, the process merely 
 involving a change of co-ordinate axes. Clearly, if we 
 choose our primaries so near together that we cannot ex- 
 perimentally reproduce all the spectral stimulus values, 
 our table will involve negative directions, but this is of no 
 theoretical importance. 
 
 Although, as I have said, it is not necessary to give full 
 details as to spectral gauging, I may note the results obtained 
 with complementaries. The usual definition of comple- 
 mentary colours — i.e. two lights which, when mixed together, 
 give a colourless mixture — is ambiguous, depending on 
 whether we define white subjectively or by reference to a 
 mixture of known physical composition. Experimentally, 
 the results obtained with and without a comparison field of 
 known composition, i.e. by the second or the first method, 
 do not difi'er widely. The table contains the values ascer- 
 tained by Helmholtz (without a comparison), and those of 
 
 ' Z.P.P.S.O., iv. 292. 
 
TRICHROMATIC VISION" 
 
 Complementary Colours for Tioo Obitervern. 
 
 135 
 
 Observer 
 
 V. KUIES. 
 
 Obsbkver 
 
 V. Trey. 
 
 Helmholtz. 
 
 Long-waved 
 light. 
 
 Short-waved 
 Light. 
 
 Long-waved 
 Light. 
 
 656-2 
 
 Short-waved 
 Light. 
 
 Long-waved 
 Light. 
 
 Short- waved 
 light. 
 
 656-2 
 
 492-4 
 
 485-2 
 
 656-2 
 
 492-1 
 
 626 
 
 492-2 
 
 626 
 
 484-6 
 
 607-7 
 
 489-7 
 
 612-3 
 
 489-6 
 
 612-3 
 
 483-6 
 
 585-3 
 
 485-4 
 
 599-5 
 
 487-8 
 
 599-5 
 
 481-8 
 
 573-7 
 
 482-1 
 
 587-6 
 
 484-7 
 
 587-6 
 
 478-9 
 
 567-1 
 
 464-5 
 
 579-7 
 
 478-7 
 
 586-7 
 
 478-7 
 
 564-4 
 
 461-8 
 
 577-6 
 
 473-9 
 
 577-7 
 
 473-9 
 
 563-5 
 
 From 433 
 
 575-5 
 
 469-3 
 
 572-8 
 
 469-3 
 
 
 downwards 
 
 572-9 
 
 464-8 
 
 570-7 
 
 464-8 
 
 
 
 571-1 
 
 460-4 
 
 569-0 
 
 460-4 
 
 
 
 571-0 
 
 452-9 
 
 568-1 
 
 452-1 
 
 
 
 570-4 
 
 440-4 
 
 566-3 
 
 440-4 
 
 
 
 670-1 
 
 429-5 
 
 566-4 
 
 429-5 
 
 
 
 D 
 
 C 
 
 D 
 
 Frfi. 18. 
 
 
136 PHYSIOLOGY OF THE SPECIAL SENSES 
 
 V. Frey and v. Kries (with a comparison). The diagram 
 is constructed from the second series. It will be seen that, 
 at the red end, as the wave length changes, that of its com- 
 plementary changes at first very slowly, then faster, until 
 finally very great changes in wave length are associated with 
 relatively little alterations in colour tone, while at the middle 
 of the range the converse statement is true. 
 
 It will have been noticed in the diagram that the two 
 observers do not agree exactly in their values, and some of 
 these personal variations are of interest. The best-known 
 illustration is afforded by the individual differences found 
 in matching a homogeneous yellow by a mixture of red and 
 green ; a good match for one eye is too green for another, too 
 red for a third. Maxwell, who first noticed this, suggested 
 that the differences depend upon unequal quantities of yellow 
 pigment in the macula, since yellow pigment absorbs very 
 little long-waved light, only becoming markedly active for 
 the yellow-green region of the spectrum, its powers diminish- 
 ing again towards the violet. The correctness of this sugges- 
 tion has been rendered highly probable by the researches of 
 Hering, Sachs, v. Frey, and v. Kries. It has been shown that 
 (1) the variations are consistent with the hypothesis that the 
 eye contains an absorbing pigment ; (2) the macular pigment 
 behaves as a yellow pigment from this point of view (Sachs) ; 
 (3) the amount of macular pigment is variable ; (4) finally, 
 a red-green mixture which matches a homogeneous yellow 
 on direct fixation appears too green when examined para- 
 centrally, and the central equation R + BIG = V appears dis- 
 tinctly bluish-green in indirect vision. These differences 
 are therefore of a physical order, and need not detain us. 
 
 There are, however, in addition some cases which cannot 
 be brought under the last category. Lord Rayleigh dis- 
 covered that when testing the equation Homogeneous 
 Yellow = Red + Green, certain persons required more red and 
 others more green than the majority. These results have 
 been confirmed by Donders, Hering, v. Kries, Koenig, and 
 Dieterici, the type requiring an excess of green being the 
 more frequent. These forms of trichromatic vision have 
 been termed by v. Kries the anomalous greens and the 
 
TRICHROMATIC VISION 137 
 
 anomalous reds, and it is certain the peculiarities do not 
 depend simply on physical diiferences. If the changes were 
 due to differences in pigmentation, then, in matching a red- 
 green mixed with homogeneous lights from 670 ;u./i to 570 nfx, 
 these subjects would require throughout the same propor- 
 tional increase in the green component of the mixture, which 
 is not the case. Similar typical differences are observed in 
 peripheral matches. These results prove that distinct varia- 
 tions in trichromatic vision exist which are irrespective of 
 differences in macular pigmentation, but perhaps no advan- 
 tage is obtained by assigning them to separate classes. 
 Hering in examining the colour sense of a series of persons 
 found they could be divided roughly into two groups, one 
 set making the brightness match. Spectral red (660 /u/u) : 
 Spectral blue (447 (ifi) : : 1.15 : 1 ; the other for the same 
 match requiring the discs in the ratio of 7 to 1. The spectral 
 light judged to be pure green was not the same in the two 
 classes, but nearer the red end in the first group. There 
 were also typical differences in the composition of colourless 
 red and bluish-green or greenish-yellow and violet mixtures ; 
 in each case the first group required relatively large quan- 
 tities of the short-waved component. The first group were 
 said to be relatively yellow sighted, the second relatively 
 blue sighted, in allusion to their special responsiveness to these 
 colours ; v. Kries' red anomaly would be an extreme case of 
 yellow sightedness, his green anomaly a marked example of 
 blue sightedness. It is therefore not improbable that these 
 abnormal trichromatics are extreme variants of a frequency 
 system representing the whole range of visual types. The 
 matter can only be settled when the quantitative mixing- 
 ratios for a definite match have been determined on a large 
 number of persons taken at random ; we may then find that 
 the results are in accordance with some well-known frequency 
 distribution, the normal equation representing a modal value. 
 At present, the theoretical interpretation of these cases is 
 extremely difficult. Summing up the results obtained in this 
 chapter, we see : — 
 
 (1) A complete sensational analysis of vision cannot be 
 effected in terms of stimuli. 
 
138 PHYSIOLOGY OF THE SPECIAL SENSES 
 
 (2) Fixing our attention on stimuli only, mixing results 
 are functions of two variables, and graphically expressible 
 by means of a plane figure. 
 
 (3) Colour differences can be expressed in terms of three 
 chosen stimuli, the choice being theoretically restricted, but 
 in practice certain definite stimuli are conveniently chosen. 
 
 (4) Of the discrepancies observed between the matches 
 of certain individuals, some depend on peculiarities of 
 physical structure ; others are to be provisionally regarded 
 as extreme variants from a modal type of vision. 
 
 I hope to have made it clear that these statements are 
 essentially descriptive, designed to resume experimental 
 facts as simply as possible. Colour diagrams and the asser- 
 tion that normal colour vision is (roughly) trichromatic are 
 short ways of expressing experimental results, and contain 
 no hidden theoretical meaning ; their truth, or falsehood, is 
 a matter of observation. As we shall see later, the results 
 have been the starting point of various hypotheses regarding 
 the nature of visual processes, but the success, or failure, of 
 these attempts is quite another story. 
 
 In the next chapter we shall treat in the same objective 
 fashion of those examples of partial colour-blindness which 
 are physiologically interesting, and shall find that, just as 
 normal vision may be called trichromatic, these types are 
 dichromatic. 
 
 EECOMMENDED FOR FURTHER STUDY 
 
 Hermann v. Helmholtz, Handbuch der Physiologischen Optik., second 
 edition, 1896, pp. 311-341. (The classical work, one of the greatest 
 books of any age, contains full bibliography up to 1895.) 
 
 Ewald Hering, op. cit. (Most suggestive and readable.) 
 
 /. V. Kries, Nagel's Handb., vol. iii. 109-127. 
 
 Sachs On Macular Pigment, Pfliig. Arch., vol. 1. p. 574. 
 
 /. V. Kries, Beitr. z. Phys. d. Gesichtempf., Du Bois R. Arch. f. Phys., 
 1878, p. 503. 
 
 Lord Eayleigh, Experiments on Colour, Nature, 1881, p. 264. 
 
 Bonders, Du Bois R. Arch. f. Phys., 1884, p. 518. 
 
 J. V. Kries, Z.P.P.S.O., vol. xiii. p. 241, and vol. xix. p. 63. 
 
 IF. Nagel, Arch. f. Augenheilkunde, vol. xxxviii. p. 31. 
 
CHAPTEE XV 
 
 DICHROMATIC VISION 
 
 We have seen that, from the purely experimental point 
 of view, the characteristics of normal colour vision admit of 
 relatively simple arrangement, that, in fact, all the results 
 of stimulation can be expressed in terms of three different 
 stimuli; we must next consider some types of vision 
 which, if themselves abnormal, throw light on the normal 
 mechanism. 
 
 The existence of abnormal visual systems is known to 
 have been recognised for more than two centuries,^ but 
 John Dalton, the great chemist, was the first to attract 
 much attention to the subject.^ Goethe in his Farbenlehre, 
 which appeared in 1812, gave the following description : — 
 
 "We will here first advert to a very remarkable state 
 in which the vision of many persons is found to be. As it 
 presents a deviation from the ordinary mode of seeing 
 colours, it may fairly be classed under morbid impressions ; 
 but as it is consistent with itself, as it often occurs, may 
 extend to several members of a family, and probably does 
 not admit of cure, we may consider it as bordering only 
 on the nosological cases, and therefore place it first. I was 
 acquainted with two individuals not more than twenty years 
 of age who were thus affected. . . . They agreed with the 
 rest of the world in denominating white, black, and grey 
 in the usual manner. . . . They appeared to see yellow, red- 
 yellow, and yellow-red like others. . . . But now a striking 
 difference presented itself. If the carmine was pressed thinly 
 over the white saucer, they would compare the light colour 
 
 1 Turberville, Phil. Trans., 1684: Huddart, Phil. Trans., 1777; Whison, 
 Phil. Trans., 1778. 
 
 * J. Dalton, Literary and Philosophical Society of Manchester, 1794. (Cited 
 in Koenig's bibliography.) 
 
140 PHYSIOLOGY OF THE SPECIAL SENSES 
 
 thus produced to the colour of the sky, and call it blue. If 
 a rose were shown them beside it, they would in like manner 
 call it blue ; and in all the trials that were made, it appeared 
 that they could not distinguish light blue from rose colour. 
 They confounded rose colour, blue, and violet on all occa- 
 sions: these colours only appeared to them to be dis- 
 tinguished from each other by delicate shades of lighter, 
 darker, intenser, or fainter appearance. Again, they could 
 not distinguish green from dark orange, nor, more especially, 
 from a red brown. 
 
 "If any one, accidentally conversing with these indi- 
 viduals, happened to question them about surrounding ob- 
 jects, their answers occasioned the greatest perplexity, and 
 the interrogator began to fancy his own wits were out of 
 order. With some method we may, however, approach to 
 a nearer knowledge of the law of this deviation from the 
 general law. 
 
 " These persons, as may be gathered from what has been 
 stated, saw fewer colours than other people; hence arose 
 the confusion of different colours. . . ." ^ 
 
 Since the time of Goethe a great deal of attention has 
 been devoted to the subject, and its literature has attained 
 formidable proportions. 
 
 The most obvious distinction between the vision of the 
 partially colour-blind and our own is an inability to perceive 
 differences which are plain to us. As Goethe says, such 
 persons see fewer colours than we do. This difficulty is 
 of special interest, because it is something definite ; it is 
 comparatively easy to find out whether a person can dis- 
 tinguish between the effects produced on him by two stimuli 
 which certainly affect us differently, while the actual physio- 
 logical or psychological nature of the effect may remain 
 obscure. How a certain green light affects a colour-blind 
 eyoi by what sensation it is followed, we can only guess, but 
 we know that the effect is indistinguishable from that pro- 
 duced by a certain intensity of red light. We have, then, 
 to deal with a condition in which the conscious responses 
 
 ' Goethe's Theory of Colours, translated by Eastlake, pp. 45-7. London, 
 1840 (Murray). 
 
DICHROMATIC VISION 
 
 141 
 
 to varied physical stimuli are fewer than in normal persons. 
 The question is, whether the experimental results can be 
 summed up in the simple manner that was possible in the 
 case of normal vision. It will be found that, with certain 
 limitations, the results can be so described. 
 
 We learned that for most experimental purposes normal 
 colour is expressible in terms of three colours. The vision 
 
 Proportions of Standard Red and Blue in a Mixture of apparently the same 
 Tint as various Spectral Colours in four cases of Partial Colour-Blindness 
 (v. Kries). 
 
 Spectral Wave 
 
 Lengths in 
 millionths of a 
 
 Deuteranope (1). 
 
 Deuteranope (2), 
 
 Protanope (1). 
 
 Protanope (2). 
 
 
 
 
 
 
 
 
 
 millimetre. 
 
 Ked. 
 
 Blue. 
 
 Red. 
 
 Blue. 
 
 E«d. 
 
 Blue. 
 
 Bed. 
 
 Blue. 
 
 670-8 
 
 33-0 
 
 
 34-4 
 
 
 5-3 
 
 
 4-9 
 
 
 656 
 
 48-0 
 
 
 56-4 
 
 
 
 9-1 
 
 
 
 8-4 
 
 
 642-0 
 
 79-0 
 
 ... 
 
 95-0 
 
 
 
 19-0 
 
 
 
 18-0 
 
 
 628-0 
 
 107-0 
 
 ... 
 
 126-0 
 
 
 
 38-0 
 
 
 
 38-5 
 
 
 615-0 
 
 147-0 
 
 ... 
 
 138-0 
 
 
 
 63-0 
 
 
 
 63-0 
 
 ... 
 
 603-0 
 
 151-0 
 
 
 155-0 
 
 
 
 90-0 
 
 
 
 84-0 
 
 
 591-0 
 
 137-0 
 
 ... 
 
 144-0 
 
 
 
 109-0 
 
 
 
 105-0 
 
 
 581-0 
 
 124-0 
 
 • . . 
 
 129-0 
 
 
 
 111-0 
 
 
 
 113-0 
 
 
 571-0 
 
 103-0 
 
 
 108-0 
 
 
 
 120-0 
 
 
 
 126-0 
 
 
 561-0 
 
 82-0 
 
 ... 
 
 89-0 
 
 
 
 108-0 
 
 
 
 106-0 
 
 
 552-0 
 
 64-0 
 
 ... 
 
 65-0 
 
 
 
 2-0 
 
 
 
 101-0 
 
 ... 
 
 544-0 
 
 52-0 
 
 
 56-0 
 
 
 
 78-0 
 
 
 
 85-0 
 
 ... 
 
 536-0 
 
 41-0 
 
 6-3 
 
 37-4 
 
 6-0 
 
 65-0 
 
 
 
 67-5 
 
 ... 
 
 525-0 
 
 26-0 
 
 12-0 
 
 21-0 
 
 10-3 
 
 38-3 
 
 11-0 
 
 46-8 
 
 
 515-0 
 
 15-0 
 
 28-0 
 
 13-7 
 
 21-6 
 
 20-6 
 
 34-0 
 
 32-8 
 
 13-0 
 
 505-0 
 
 7-7 
 
 36-0 
 
 7-5 
 
 32-2 
 
 9-8 
 
 35-0 
 
 17-2 
 
 29-0 
 
 496-0 
 
 3-7 
 
 48-0 
 
 4-1 
 
 46-3 
 
 4-8 
 
 47-0 
 
 8-4 
 
 330 
 
 488-0 
 
 1-6 
 
 62-0 
 
 1-9 
 
 58-0 
 
 2-2 
 
 57-0 
 
 5-3 
 
 49-0 
 
 480-0 
 
 0-9 
 
 64-0 
 
 0-9 
 
 67-0 
 
 0-9 
 
 66-0 
 
 2-9 
 
 71-0 
 
 469-0 
 
 0-3 
 
 70-0 
 
 0-3 
 
 65-6 
 
 0-3 
 
 67-0 
 
 1-0 
 
 69-0 
 
 460-8 
 
 ... 
 
 67-0 
 
 ... 
 
 68-6 
 
 ... 
 
 54-0 
 
 
 66-0 
 
 of partial colour-blinds is expressible in terms of two. If we 
 choose as our fixed lights a red and a blue, these, mixed in 
 certain proportions, will match every part of the spectrum as 
 it appears to the partially colour-blind, and also unanalysed 
 daylight. Normal colour vision is trichromatic, this is di- 
 chromatic. But, just as the required proportions in the 
 mixture differed in various types of trichromatic vision, we 
 also find two types of dichromatics, distinguished by their 
 
142 PHYSIOLOGY OF THE SPECIAL SENSES 
 
 respective mixing ratios. The two classes are known by 
 various names, of which I shall employ Protanope and 
 Deuteranope as being quite non-committal from the theo- 
 retical standpoint. Experimentally one proceeds just as in 
 the examination of normal colour vision. A certain red and 
 a certain blue are chosen, and the whole spectrum systemati- 
 
 ze 
 
 
 
 
 
 
 J 
 
 \^ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 \^ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 I 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 4 
 
 
 
 
 \> 
 
 1^ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 ^ 
 
 \ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 / 
 
 y 
 
 \ 
 
 |y> 
 
 iv^ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 // 
 
 
 
 \\. 
 
 \ 
 
 \ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 I 
 
 / 
 
 
 
 \, 
 
 'v^ 
 
 y 
 
 ^ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 / 
 
 
 
 
 
 \v 
 
 
 'v^ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 . 
 
 f 
 
 
 
 
 
 N 
 
 ^ 
 
 ^ 
 
 N 
 
 
 
 
 
 
 
 
 
 
 t 
 
 
 
 1 
 
 
 
 
 
 
 
 \ 
 
 !^ 
 
 \N 
 
 
 
 
 
 
 
 
 
 
 / 
 
 
 
 / 
 
 
 
 
 
 
 
 
 \ 
 
 \ 
 
 ^^ 
 
 
 
 
 
 
 
 
 A, 
 
 y 
 
 
 i 
 
 
 
 
 
 
 
 
 
 
 ^ 
 
 \ 
 
 \^ 
 
 
 
 
 
 
 
 
 
 
 / 
 
 
 
 
 
 
 
 
 
 
 \ 
 
 ^ 
 
 N,^ 
 
 S 
 
 N.. 
 
 
 
 
 
 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 ^ 
 
 ^ 
 
 s. 
 
 ■^ 
 
 s. 
 
 
 
 B 
 
 ■^ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 V 
 
 ^ 
 
 ■~-~ 
 
 
 :i; 
 
 — 
 
 
 
 is 
 
 9 10 11 1Z 13 n IS 16 il 18 19 ZO Zl 
 
 «3;iJ"^-S"T*e^«*<0 "9 14 LA ^ flO 
 
 T^^p^JOt^yjlne^Cfa f^ ^ C3 m oft 
 
 12 3 4 5 6 7a 
 
 lO Jo 5? 
 
 Waxa Lengths cF Sptctnxjw. 
 RetL YaMjes far FartiaL ColourBlinds (vSries). 
 The two auvesA are' DetUeranopce, the' two £ Protanopes. 
 
 cally matched against various mixtures of the chosen standard 
 lights. The tables and diagrams embody the results of 
 J. V. Kries.i 
 
 An examination of the curves shows that the cases fall 
 into two groups in respect of red values. The four blue 
 value curves, on the other hand, are not very dissimilar ; in 
 fact, remembering that the physical differences of macular 
 
 1 "Ueber Farbensysteme," Z.P.P.S.O,, vol. xiii. pp. 241-324. An abbre- 
 viabed account, with the diagrams, by the same author in Nagel's Uandbuch 
 d. Physiol, d. Menschen, Braunschweig, 1905, vol. iii. pp. 153-55. Some account 
 of the experimental technic will be found in Z.P.P.S.O., vol. xii. pp. i, etc. 
 
DICHROMATIC VISION 
 
 143 
 
 absorption would be most influential in the part of the 
 spectrum where the blue curves agree worst, we may regard 
 the similarity as fairly close. Referring to the tables, we see 
 that for matching spectral colours of wave length greater 
 than 530 fifi, no blue at all is required, so that the forms 
 of the red value curves beyond this point are especially 
 instructive. In one group the curve rises fairly sharply to 
 a maximum about 571 /u./a and then falls steeply, suggesting 
 
 so 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 10 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 / 
 
 
 
 ■ 
 
 
 2:: 
 
 i? 
 
 :;^ 
 
 ?— ' 
 
 -X 
 
 eo 
 
 
 
 
 
 
 
 
 
 
 ,. 
 
 ■■-r^ 
 
 •y- 
 
 '" 
 
 
 
 
 — N 
 
 N 
 
 
 
 
 
 
 
 
 
 
 
 
 y. 
 
 7 
 
 
 
 
 
 
 
 
 \ 
 
 
 t 50 
 
 
 
 
 
 
 
 y 
 
 '/ 
 
 ■> 
 
 / 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 // 
 
 <!* 
 
 
 / 
 
 
 
 
 
 
 
 
 
 
 
 
 \ 
 
 
 
 
 
 .^ 
 
 'A 
 
 
 / 
 
 
 
 
 
 
 
 
 
 
 
 
 
 -§ « 
 
 
 
 
 / 
 
 
 
 / 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 CO 
 
 o 
 
 
 ^' 
 
 
 / 
 
 ^^ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 '^ 
 
 y 
 
 '> 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 "S 
 
 § 20 
 S 
 
 ^"^ 
 
 / 
 
 / 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 / 
 
 / 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 ■^ 
 
 / 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 i 
 
 21 
 
 za 
 
 2^1 
 
 Blue Values of Partial Colour-Blinds. 
 
 a relatively low stimulus value for the long-waved light; 
 in the other group the rise is more gradual, attaining a 
 maximum at 603 fifi, while the curve does not fall to so 
 low a point at the red end. Hence, within the region under 
 examination, we may say, with fair accuracy, that the four 
 subjects are grouped into one pair relatively more sensitive 
 to short-waved light and one pair relatively more sensitive 
 to long-waved light. To the former group we apply the 
 name Protanopes, to the latter that of Deuteranopes. But, 
 it may be objected, we have only examined four cases. It 
 
144 PHYSIOLOGY OF THE SPECIAL SENSES 
 
 is necessary to see whether the results attained are confirmed 
 when experiments are carried out on a larger scale. If the 
 conclusions just stated be valid, in matching yellow some 
 partial colour-blinds (protanopes) will require enormously 
 more red than others (deuteranopes). This is a test which 
 can be quickly applied, while the above examination is too 
 elaborate to be often repeated. V. Kries tested twenty cases, 
 using red (lithium line) and a fixed yellow (sodium line). 
 The results are contained in the table below, and it will be seen 
 that one set (protanopes) required on the average five times 
 as much red as the other group (deuteranopes) to match 
 
 rrotanopes. 
 
 Deuteranopes. 
 
 214-0 
 
 36-5' 
 
 213-0 
 
 36-3' 
 
 211-0 
 
 36-3 » 
 
 205-0 
 
 36-5' 
 
 196-0 
 
 38-4 
 
 198-0 
 
 37-3 
 
 210-0 
 
 37-0 
 
 200-0 
 
 37-0 
 
 210-0 
 
 37-8 
 
 203-0 
 
 37-0 
 
 225-0 
 
 36-9 
 
 ... 
 
 38-0 
 
 the standaid yellow. There is, accordingly, some evidence 
 that the classification is a definite one. 
 
 The relative insensitiveness to long-waved light explains 
 the observation that in protanopes, who correspond to the 
 class (badly) named red blinds, the red end of the spectrum 
 appears shortened, although the shortening is of very little 
 importance from our point of view. It is to be remarked 
 that within each class slight individual variations are found. 
 In the two protanopic eases, from 552 /i/x onwards to violet, 
 one subject constantly demanded more red in his mixture 
 than the other. 
 
 An interpretation is easy : in the observer requiring less 
 red, the homogeneous light was probably weakened by 
 macular absorption, since the diminution in red values ao-rees 
 
 ' One subject. 
 
DICHROMATIC VISION 145 
 
 with the known increments of absorption as we pass towards 
 the violet. This difference is of some importance in con- 
 nection with an interesting peculiarity of dichromatic vision. 
 Since daylight can be matched, for the dichromatic eye, by 
 a mixture of red-blue, and since all spectral colours can be 
 matched by mixing the same two standards in varying pro- 
 portions, we should expect some spectral colour to match, 
 more or less closely, unanalysed daylight, and to appear to 
 the dichromatic the same as what we call white or grey. 
 Such a point of the spectrum is called " the neutral point." 
 As for protanopes, the stimulus value of the light falls off 
 rapidly towards the red ; their " neutral point " should be 
 nearer the violet than that of the deuteranopes ; this is, in 
 fact, generally the case, but if there is a good deal of pigment 
 in the macula lutea, the mixed light undergoes selective 
 absorption, and the homogeneous match is nearer the red 
 end of the spectrum. Owing to this fact, the typical differ- 
 ence in the situation of the neutral point may not be found at 
 all. The table on p. 146 gives results obtained on a deuteranope 
 with a moderately pigmented macula, on a protanope with 
 much and a protanope with little pigment in the macula 
 region. Evidently simple determination of the neutral point 
 would not enable us to distinguish the two forms. 
 
 A study of the neutral point, however, immediately brings 
 to our notice that characteristic which has attracted the 
 attention of practical men. Spectral light between 490- 
 500 fifji. induces normally the sensation of green; for the 
 dichromatic it has the same effect as unanalysed daylight or 
 a particular mixture of red and blue. But this mixture con- 
 tains much red and little blue, so that a normal person asked 
 to name the simple colour which it most resembled would 
 say red. In other words, both the dichromatic classes con- 
 fuse red and green. Even here we observe a class difference. 
 In matching a given bluish-green the protanope, being 
 relatively insensitive to long waves, requires much red in 
 his blue-red mixture ; the deuteranope takes about the same 
 amount of blue, but much less red. Accordingly, a protanope 
 confuses a light bluish-red (physically speaking) with a green 
 that appears to us much darker, e.g. a scarlet with an olive- 
 
 K 
 
146 PHYSIOLOGY OF THE SPECIAL SENSES 
 
 green ; a deuteranope matches a much bluer-red with green, 
 which we should take to be about equally bright. Although 
 both groups confuse certain reds with certain greens, " the 
 red that appears equal to a given green differs markedly 
 both in colour tone and intensity in protanopes and deu- 
 teranopes " (v. Kries). The fact that both forms of common 
 colour-blindness are associated with a confusion of red and 
 green is of great practical importance, since these colours 
 are universally employed for signalling in connection with 
 railways and ships. In most countries persons desirous of 
 acting as engine-drivers or navigating officers are submitted 
 to examination with a view to the detection of colour-blind- 
 
 Neutral Points {v. Kriex) 
 
 Mixed Light Used. 
 
 Wave Length (in fifi) of the homogeneous Match. 
 
 Deuteranope. 
 
 1st Protanope. 
 
 2nd Protanope. 
 
 Magnesium oxide } 
 in daylight ^ 
 
 499 
 
 498 
 
 490 
 
 Mirrored cloud ') 
 weakened by ^ 
 ground glass ) 
 
 499 
 
 497 
 
 489 
 
 Do. do. weakened ) 
 by smoked glass | 
 
 495 
 
 494 
 
 486 
 
 ness. While it is not the function of a theoretical writer to 
 discuss at length a matter of this kind, a few remarks on the 
 utilitarian aspect of the matter are not, perhaps, out of place. 
 A test which can be easily applied, and is consequently dear 
 to the mind of lay officials, is that associated with the name 
 of Holmgren. The subject is given a large number of skeins 
 of wool, and is told to separate out the reds and greens. 
 This test is, in my opinion, altogether inefficient. Under the 
 conditions of the test, the marked differences in luminosity 
 of the wools often enables persons who suffer from typical 
 colour-blindness to sort out the skeins correctly, especially 
 if they have previously practised the experiment. Such 
 
DICHROMATIC VISION 147 
 
 persons, however, would be highly dangerous navigators or 
 engine-drivers, because at times when the differentiation of 
 a red from a green light is vital, e.g. in a slight haze or 
 in snowy weather, such adventitious aids as differences of 
 intensity, apart from colour tone, will be absent. Edridge 
 Oreen, to whom we owe many important observations on 
 these points, has perfected a lantern test which is in every 
 way superior to that of Holmgren, and nobody who takes 
 on himself the responsibility of testing colour vision for 
 administrative purposes can safely rely on the other and less 
 efficient method. For further details, special works must be 
 ■consulted. 
 
 Returning to the physiological characters of the two 
 common types of colour vision, we have to consider what 
 relations their visual systems bear to that of normal persons. 
 The mere fact that the systems are 'dichromatic tells us com- 
 paratively little ; we should indeed conclude that sensations 
 of colour are less numerous for dichromatics than for us, but 
 they might be quite different. What we seek is a relation 
 between stimuli. As early as 1837 Seebeck expressed the 
 opinion that two lights or mixtures of lights which appeared 
 equal to the normal eye never appear unequal to colour- 
 blind eyes, that equations valid for the normal eye are 
 always valid for the colour-blind eye. If this be true, it is 
 important, because in that event a dichromatic lacks some- 
 thing the normal person has, but possesses nothing the 
 normal man has not got ; in other words, dichromatism is 
 an error of defect. 
 
 Experimental evidence, particularly that furnished by v. 
 Kries, on the whole bears out Seebeck's view. In the first 
 place, mixtures of red (670-8 t^jj-) and yellowish-green (550 ^/i) 
 which match a homogeneous yellow when the proportions 
 are chosen to give a good match for a normal eye, give also 
 equality for either a deuteranopic or protanopic subject. 
 Similarly, any match which is found to be valid for both 
 types is described as a good match by a normal person. 
 " One employs the frequently cited equations between a 
 homogeneous yellow and a mixture of red and yellowish- 
 green (670-8 /*/* and 550 Myu). As all lights in this region are of 
 
148 PHYSIOLOGY OF THE SPECIAL SENSES 
 
 equal stimulus value for the colour-blind whatever the ratio 
 of red to yellowish-green, one can always give the homogeneous 
 yellow such an intensity that the match is good either for a 
 protanope or a deuteranope, but in general the matches of 
 the one are not valid for the other. As we should expect, a 
 strongly red mixture is for the protanope equivalent to a 
 yellow of relatively feeble intensity ; a deuteranope finds in 
 a match arranged by the protanope the mixture too bright 
 and the pure yellow too dark. The relation is reversed for 
 strongly green mixtures. With extraordinary accuracy, 
 however, we find that for the ratio of red to yellowish-green 
 that has for the trichromatic an equal colour tone with the 
 homogeneous yellow, both groups agree ; trichromatic equa- 
 tions are valid for both protanopes and deuteranopes. 
 Conversely, if we try to find an equation valid for both 
 groups, we arrive precisely at the one valid for a normal 
 person." i 
 
 Still stronger evidence is afforded by the following con- 
 siderations : — V. Kries prepared a table giving the mixing 
 ratios for a normal person, matching spectral colours between 
 670"5 and 550 /j-jj-. The following are his results : — 
 
 Spectral Keglon, 
 
 Proportions of Standard Colours. 
 
 Wave length. 
 
 670-8 n/i. 
 
 662 nu. 
 
 670-8 
 
 88-5 
 
 
 6280 
 
 251-0 
 
 10 
 
 6150 
 
 276-0 
 
 27 
 
 603-0 
 
 270-0 
 
 49 
 
 591-0 
 
 202-0 
 
 67 
 
 581-0 
 
 123-0 
 
 76 
 
 671-0 
 
 73-0 
 
 91 
 
 661-0 
 
 21-0 
 
 80 
 
 552-0 
 
 
 71 
 
 He also found that a certain deuteranope required 33 units 
 of the standard red to match the spectral 670-8, and 64 units 
 of the standard 552 to match the spectral 552 /n/x. 
 
 If deuteranopic vision is a reduction form of the normal 
 
 ' V. Kries, Nagel's Handh., iii. 160. 
 
DICHROMATIC VISION 
 
 149 
 
 system, it should be possible from these data to calculate the 
 stimulus values of intervening lights. Take as an illustra- 
 tion 591 iJ-iJ., which is matched by 202 units of 670-8 /x/i plus 
 67 units ot 552 /^/x. We first change from the arbitrary in- 
 tensity units of the normal subject by dividing by 88-5 and 
 71 respectively, and then express in the intensity scale of the 
 deuteranope by multiplying by 33 and 64 respectively. The 
 stimulus value should therefore be 
 
 202 67 
 
 gg:^ X 33 + yY X 64 = 135 (approx.) units. 
 
 This, then, is the stimulus value of 591 fj,/j, expressed in 
 the intensity values of 670-8 fi/j. and 552 /x/x for the 
 given spectrum, on the hypothesis that a match good for 
 a trichromatic eye is valid for a dichromatic; the observed 
 value was 137. In this way the following table was obtained, 
 and the agreement is very good : — 
 
 Stimulus Values for Deuteranope and Protanope. 
 
 Wave 
 Length. 
 
 Deuteranope. 
 
 Protanope. 
 
 
 
 
 
 
 Calculated. 
 
 Observed. 
 
 Calculated. 
 
 Observed. 
 
 670 
 
 33 
 
 33 
 
 4-9 
 
 4-9 
 
 628 
 
 106 
 
 107 
 
 28-8 
 
 38-5 
 
 615 
 
 126 
 
 145 
 
 54-2 
 
 63-0 
 
 603 
 
 145 
 
 151 
 
 86-0 
 
 84-0 
 
 591 
 
 135 
 
 137 
 
 108-0 
 
 105-0 
 
 581 
 
 114 
 
 124 
 
 117-0 
 
 1130 
 
 571 
 
 110 
 
 103 " 
 
 137-0 
 
 126-0 
 
 561 
 
 76 
 
 82 
 
 111-0 
 
 106-0 
 
 552 
 
 64 
 
 64 
 
 101-0 
 
 101-0 
 
 A further, but less satisfactory, verification is afforded by 
 the construction of a normal colour table from observations 
 made upon dichromatics. Bearing in mind the principles 
 upon which the normal colour " triangle " depends, it is clear 
 that in applying the method to dichromatic systems all 
 those colours which appear alike to the colour-blind eye 
 must lie on a straight line, since the line joining two points 
 in the colour " triangle " contains all points corresponding to 
 
150 PHYSIOLOGY OF THE SPECIAL SENSES 
 
 colours which can be mixed from the colours represented by 
 the first pair. It is also clear that the mixtures of confusion 
 colours with any other colours lies on a series of straight 
 lines. Hence it can be shown ^ that (1) all such lines 
 meet in a point or are parallel, (2) the point of intersection 
 corresponds to a colour which has no stimulus value for 
 the colour-blind eye. This point is usually termed the 
 " Fehlpunkt " of the system.^ 
 
 Bearing these preliminaries in mind, the method of 
 constructing a colour "triangle" for normal vision from 
 dichromatic observations is simple. 
 
 Let ABC be an equilateral triangle, then it is assumed 
 that the protanopic red value, the deuteranopic red value, 
 
 and the common blue value are 
 independent one of another, and 
 their representative points are 
 the apices of the triangle. The 
 positions of the various lights 
 are then determined in the 
 following manner : Take, for 
 instance, 603 jxji. The pro- 
 tanopic red value is 90, the 
 deuteranopic red value 155, 
 Fig. 19. the blue value 0. Divide AB in 
 
 C'so that AC':C'B = 90 : 155. 
 C marks the situation of 603 /xju in our "triangle." 
 Take again 505 /i/^. The protanopic red value is 7"5, the 
 
 deuteranopic red value 9'8, the blue value — — i — = 33-7o. 
 
 Divide AB in C" so that AC" : C"B = 7-5 : 9-8. Join C"C ; 
 divide C"C in C" so that Q"C"' : CC"'= 33-75 : 9-8 + 7-5. 
 The diagram above shows a colour "triangle" constructed 
 in this way by v. Kries. The only modifications are that 
 (1) the unit for blue values has been increased fivefold, i.e. 
 the tabled values (p. 149) have been divided by five for con- 
 venience of reproduction; (2) "the very small W (red) 
 values for lights of wave lengths less than 505 /^jit are not 
 sufiSciently exact to permit of the evaluation of the ratio 
 
 * See proof at the end of the chapter. 
 
DICHROMATIC VISION 151 
 
 (protanopic red value : deuteranopic red value) with certainty. 
 On this account, that part of the boundary line extending 
 from blue to green would be correspondingly irregular, 
 whereas observations on normal sighted persons show that it 
 must present a steady curvature, mixtures of the two lights 
 appearing a little less saturated than the spectral lights. 
 I have for this reason drawn the boundary line with a 
 constant curvature, so that the ratio gradually passes from 
 that corresponding to lights between 515 j^/^ and 525 /i/i to 
 that observed for lights from 469 /xju, to 480 [j-iJ'. These 
 two values are indicated in the drawing by the dotted lines 
 from C ; accordingly a curve drawn in exact correspondence 
 with the observations would run irregularly between these 
 lines. It should also be remarked that a construction of 
 
 ao-8 
 
 Fig. 20. 
 
 this sort is not entirely free from objection, as it combines 
 the results of several observers with undoubtedly diiferent 
 amounts of macular pigmentation." ^ 
 
 That the process is not so satisfactory as the one already 
 considered will be readily admitted. But we notice again 
 that the deduced " triangle " agrees very fairly with that 
 given in the chapter on trichromatic systems, so that it 
 affords some confirmation of the surmise that dichromatic 
 systems are in reality reduction forms of the trichromatic 
 type. 
 
 The reader is specially cautioned that the conclusions 
 hinted at in the last few pages are not to be pushed too far ; 
 for reasons, some of which will be considered in the chapters 
 on theories, all the prominent workers of to-day recog- 
 
 1 v. Kries, Nagel's Handb., iii. 162. 
 
152 PHYSIOLOGY OF THE SPECIAL SENSES 
 
 nise that partial colour-blindness cannot be completely or 
 adequately described in terms of a hypothesis postulating so 
 simple a relation as that just suggested between normal and 
 abnormal kinds of vision. The experimental results are, 
 however, by no means without interest, and I have for that 
 reason devoted some space to the matter. Summing up 
 the rather difficult topics dealt with in this chapter, we 
 may say that — 
 
 (1) The two common forms of partial colour-blindness 
 are distinguishable one from another by varying responsive- 
 ness to long-waved and medium-waved lights. Protanopes 
 are relatively insensitive to long waves, deuteranopes to 
 short waves. 
 
 (2) Each is an example of dichromatic vision, using that 
 term in a strictly experimental sense. 
 
 (3) Each is, very approximately, a reduction form of 
 normal trichromatic vision. 
 
 The forms of partial colour-blindness just described are 
 of everyday occurrence, and their recognition of obvious 
 practical importance ; another type, less common, and there- 
 fore less completely studied, is that known as Blue or Yellow- 
 Blue Blindness. This condition, unlike the last, is not always 
 congenital, and frequently one-sided. Definite pathological 
 changes in the retina have also been observed in several 
 examples of the defect, and the whole visual field may 
 not be involved. Koenig's observations make it probable 
 that this type also is dichromatic, two suitably chosen 
 simple lights matching the whole spectrum. Blue blindness 
 may be a reduction form, since all normal matches appear 
 to be valid for such an eye. The neutral point is in the 
 yellow between 566 /x/x and 570 ij-ij- ; the " Fehlpunkt " is close 
 to the violet. Our knowledge of this condition is not 
 sufficiently exact for it to be profitably discussed in a 
 text-book.^ 
 
 The experimental production of a form of colour-blind- 
 ness by Burch has so direct a bearing on the theories of 
 
 * The curious reader will find sufficient references to the literature 
 of Blue Blindness in my article on " Theories of Colour Vision " in Further 
 Advances in Physiology, edited by Leonard Hill, London, 1909 (Arnold). 
 
DICHEOMATIC VISION 153 
 
 colour Tision, that I shall consider his work in the chapter 
 devoted to the theories of the subject. We shall next turn 
 our attention to the phenomena of successive induction or 
 after-images. 
 
 APPENDIX TO CHAPTER XV. 
 
 Proof of the existence of a " Fehlpmikt." ' 
 
 Suppose the quantity r o£ a colour situated at R in the diagram 
 matches the quantity jr of a colour at Gr, 
 
 r = nr + (l — n)r. 
 
 But the quantity nr of 'R=ng of G. 
 
 Therefore if m be a proper fraction, r of R matches {l — n)r of lS, + ng 
 of G. The situation of this mixture is S 
 
 if RS_ ng ,j> 
 
 SG {l-n)r ■ ■ ■ ^ ' 
 and the quantity s of the mixed colour obtained = Mjf + { 1 - ra))-, and this 
 is independent of n. 
 
 If, now, we mix b parts of some other colour B with s of S, we obtain 
 a mixed colour whose appearance is independent of n. Let its measure 
 be t and its situation T, 
 
 TS b 
 
 J = 6 + s = 6 + mo + (l-ra)r-, and, as before, =— = p: r- . . . (la) 
 
 ""■''' jjX ng + {i -n)r 
 
 Draw BH perpendicular to RG and TL perpendicular to BH. Call 
 LH, X, BH, h, TL, y, HG a, and RG c. 
 
 |=gg = — (by similar triangles) = ^ , „„ , /i — ^„. ■ ■ ■ (1)6 ^ 
 
 V _TL_SH_ SG-tt 
 h-x BL BH h ' 
 
 But, from (1) SG = c (!-"> 
 
 ng + {\-n)r 
 
 ■ V ^ jc-aXl-ny-ang .^, 
 
 'h-x hlng + {l-?i)r] 
 
 Eliminating n between (16) and (Ic), we 
 
 have 
 
 = ybh(g - r) - x[crg + br{c-a) + abg'] 
 
 + bh[{c-a)r + ag] . . . (Id) 
 
 This is a linear relation between the S'lG- 21. 
 
 co-ordinates x and y otT with respect to an origin H. 
 
 Therefore the locus of T is a straight line. 
 
 1 Helmholtz, op. cit., pp. 363, etc. 
 
154 PHYSIOLOGY OF THE SPECIAL SENSES 
 
 Let TQ be the locus intersecting RG in Q. QH. = y when x = Q, 
 . „ _ (c-a)'- + ag 
 
 ■ • y^ F=-^' 
 
 i.e. is independent of b, the amount mixed from B. 
 
 Therefore all lines of equal mixtures from RGB intersect at the same 
 point Q, or they are parallel, in which case r=g, and i/q is infinite. 
 
 The distance of Q from R is given by 
 
 3/o-c + a=-^ = QR . . . (1/) 
 r-g 
 
 If we mix q of the colour Q with g of the colour G, so that R results 
 g=|butRG = c, 
 
 r-g q ^ ■> 
 
 Since, by hypothesis, r=g and q = r — g is not necessarily = 0, the 
 colour Q has no stimulus value tor a colour-blind eye. 
 
 RECOMMENDED FOR FURTHER STUDY 
 
 The literature of Colour Blindness is enormous, and most of the 
 papers mingle a description of experimental or clinical observations with 
 theoretical discussions. I should advise the student first to read : — 
 
 /. von Kries, Ueber Farbensysteme, Z.P.P.S.O., vol. xiii. pp. 241- 
 324. Having mastered this, he might read the account in Helmholtz's 
 Handbuch, and then the more recent controversial papers. 
 
 A complete bibliography of the literature up to 1896 is appended to 
 the second edition of Helmholtz's Handbuch. Dr. Edridge Green's work 
 on Colour Blindness and Colour Perception, in the International 
 Scientific Series (Kegan Paul), contains many ingenious speculations and 
 matter of practical interest, but should be read in a critical spirit. 
 
CHAPTER XVI 
 
 AFTER-IMAGES OE SUCCESSIVE INDUCTION 
 
 That after the withdrawal of a stimuhis some sensation 
 process persists has long been known. In 1634 Peiresc 
 described the positive and negative after-images of the 
 window, and a few years later Bonacursius maintained in 
 opposition to the learned Jesuit Athanasius Kircher that one 
 could see as well in the dark as by daylight, convincing his 
 opponent by the following experiment. Kircher was taken 
 into a room and made to look fixedly at a drawing covering 
 a window slit ; the room was then completely darkened, and 
 Kircher once more perceived the drawing on looking at a 
 piece of white paper in his hand. The Jesuit's own explana- 
 tion of the result was that light absorbed by the eye streamed 
 out in the dark and illuminated the paper. Similar experi- 
 ments were described by Mariotte in 1668, and Newton was 
 familiar with the phenomena, which he held to be of psycho- 
 logical origin. In the eighteenth century many workers 
 observed these after-images or " accidental colours," particu- 
 larly the negative or complementary images, which, as we 
 shall find, are more easily studied. Writing in 1734, Jurin 
 observed that the "contrary sensation is apt to arise in us, 
 sometimes of itself, and sometimes from such causes as at 
 another time would not produce the sensation at all, or at 
 least not to the same degree." ^ 
 
 Scherfer (1761) investigated complementary images, and in 
 copying a picture painted it with a green face shaded with 
 yellow, white hair and eyebrows, black eyeballs with white 
 pupils, and green lips, so that the after-image had the colours 
 of the original. Kobert Waring Darwin (1786) studied both 
 positive and negative after-images, explaining clearly enough 
 the experimental precautions essential for their demonstra- 
 
 1 Quoted by Burch, Proc. Roy. Soc, 1900, Ixvi. 204. 
 
 165 
 
156 PHYSIOLOGY OF THE SPECIAL SENSES 
 
 tion, e.g. the necessity of excluding any extraneous light if 
 the "direct spectrum" (positive after-image) is to be seen. 
 Goethe in his Farbenlehre, which appeared in 1810, records 
 several observations of negative after-images, of which the 
 following is an amusing example : — 
 
 "I had entered an inn towards evening, and as a well- 
 favoured girl with a brilliantly fair complexion, black hair, 
 and a scarlet bodice came into the room, I looked attentively 
 at her as she stood before me at some distance in half shadow. 
 As she presently afterwards turned away, I saw on the white 
 wall, which was now before me, a black face surrounded with 
 a bright light, while the dress of the perfectly distinct figure 
 appeared of a beautiful sea-green." ^ 
 
 During the nineteenth century the subject has been 
 closely examined by an array of workers, amongst whom 
 Brewster, Fechner, Briicke, Helmholtz, and Hering are, 
 perhaps, the most prominent. I shall give a short account 
 of the facts, avoiding, so far as possible, any theoretical pre- 
 suppositions. 
 
 No special apparatus is required for the demonstration of 
 some fundamental results. If one looks fixedly for 20 to 40 
 seconds at a moderately illuminated white spot or source of 
 light and then shuts the eyes or looks towards a sheet of 
 white paper, a black spot usually surrounded by a bright 
 halo is visible. If the object be coloured, then performing a 
 similar experiment we see a spot tinged with the comple- 
 mentary of the original impression. These are the negative 
 or complementary after-images with which all are familiar. 
 
 To obtain an after-sensation identical in character with 
 that experienced during the original stimulation, a positive 
 after-image, somewhat greater care is necessary, since eye or 
 body movements rapidly cause its disappearance or inhibit 
 its production. " After waiting a sufficient time with care- 
 fully covered eyes, one turns the eyes (still covered by the 
 hands) toward the object, taking care to remain quite still, 
 then takes the hands away rapidly, and with the same speed 
 brings them back again. This movement of the hands must 
 
 1 Goethe's Theory of Colours, Eastlake's trans., 1840, p. 22. 
 
AFTER-IMAGES OF SUCCESSIVE INDUCTION 157 
 
 be executed gently, without violent exertion or shaking of 
 the body. If the experiment has been properly carried out, 
 the positive after-image seen behind the covering hands is 
 sometimes so sharply defined and bright, that it seems as 
 though the hands were transparent and one still saw the real 
 object. One has sufficient time to note details in the after- 
 image which there was no time to study during the actual 
 exposure." ^ 
 
 The first thing to grasp is that no discontinuity appears 
 to exist between positive and negative after-images. Suppose 
 one develops an after-image by looking at a bright object, and 
 that on closing the eyes one obtains a bright image on a dark 
 ground, i.e. a positive after-image ; then if instead of closing 
 the eyes after performing such an experiment one looks at 
 a moderately illuminated field, a dark image on a bright 
 ground is seen, i.e. a negative after-image. Between these 
 termini we can obtain a continuous series of intermediates ; 
 for instance, we can choose a field of brightness such that, 
 on fixating it, no after-image is obtained at all. This amounts 
 to saying that the nature of an after-image is in the main 
 determined by two factors: (1) The quality and intensity 
 of the first or primarily exciting stimulus ; (2) the character 
 of the secondary or, as we may call it, modifying stimulus. 
 In any after-image sensation, both factors are necessarily 
 involved ; even when the retina is shielded from external 
 stimuli, a stimulation process is active, bound up with the 
 sensation called by some the light chaos, by others the 
 autocthonous retinal light. Naturally, this is not an ex- 
 haustive enumeration of the modifying factors; the con- 
 ditions noted are necessary, but not sufficient. Individual 
 peculiarities, both psychological and physiological, are in- 
 volved, but our first study must be directed to the constant 
 factors which give us a definite point of departure. Let 
 us divide the experiments into two chief groups : (1) The 
 eye is exposed to one extrinsic stimulus only; (2) two 
 extrinsic stimuli are successively employed. 
 
 In (1) we have the primary stimulus compounded with 
 
 1 Helmholtz, Handb. d. Phys. Optik., second edition, p. 504. 
 
15 8 PHYSIOLOGY OF THE SPECIAL SENSES 
 
 the intrinsic retinal factor alone, in (2) the state of affairs 
 is more complex. Before applying a stimulus, the entoptic 
 sensation or "light chaos" is indeterminate in character, 
 varying like most subjective phenomena with the observer. 
 Helmholtz speaks of " manifold light points in motion often 
 like convoluted vessels or scattered moss fibres and leaves." 
 The appearance is often like that of waves passing slowly 
 across the field of vision. If we regard fixedly a moderately 
 bright object for 40 seconds and then darken the eyes, we 
 see in the light chaos a dark patch surrounded by a bright 
 halo ; that is to say, the effect is much the same as noticed if 
 one looks at a moderately bright surface, the " light chaos " 
 behaves like such a surface. 
 
 Turning to the more important case (2), we find a com- 
 plex of results. Let us call the first extrinsic stimulus the 
 Ketuning Light, the second extrinsic stimulus applied to 
 the same area the Reacting Light. The effect produced 
 by the reacting light will depend on the change induced 
 by the retuner, and can be measured by comparing its effect 
 Avith that obtained on stimulating regions not exposed to 
 the retuning light; the stimulus applied to such an area 
 may be termed the Comparison Light. The whole process 
 is conveniently named Retuning. 
 
 I have already made it seem probable that stimulus 
 values are not the same in different parts of the retina, 
 and that light-dark adaptation is selective ; hence, that our 
 results may be pure, we must work with small directly 
 fixated fields and under conditions of constant adaptation, 
 preferably light adaptation. 
 
 The first question that arises is as to the validity of 
 the mixing laws. We can match a homogeneous yellow with 
 a red-green mixture; if we retune with a homogeneous 
 yellow, a reacting yellow does not, of course, match a com- 
 parison yellow, and similarly for a retuning and reacting 
 mixture. The question is whether the responses to the 
 two originally equal stimuli are changed to a like extent, 
 whether the simple and mixed colours still match. Appa- 
 rently the answer must be affirmative. 
 
 " On small directly fixated fields optical equations suffer 
 
AFTER-IMAGES OF SUCCESSIVE INDUCTION" 159 
 
 no alteration if the tuning of this part of the retina be altered 
 by means of any light stimulus whatever. For instance, a 
 homogeneous yellow and a centrally equivalent red-green 
 mixture appear after previous yellow illumination both paler, 
 after previous blue illumination both more saturated yellow, 
 but they still match perfectly. In the same way, the equality 
 of unanalysed daylight and a white formed by a complemen- 
 tary mixture is not destroyed if both fields, by means of 
 previous colour stimulation, are strongly tinged by the com- 
 plementary of the retuning light." ^ 
 
 This very important result has not escaped criticism; 
 we are all agreed that it cannot be obtained by stimulating 
 the peripheral region of the retina. The correctness of the 
 experiments so far as the fovea centralis is concerned is 
 testified to by Btihler, who has re-investigated the matter and 
 can be provisionally accepted. 
 
 Accepting the persistence of optical equations as an 
 experimental fact, conclusions may be drawn which enable 
 us to deduce some interesting formuliie. Thus we can 
 obtain a relation between the change in appearance of 
 different reacting lights and any specified retuning. Put 
 briefly, we may say that the change in appearance of a light 
 as a result of retuning consists in a lowered or heightened 
 responsiveness of the eye to that light. If R be the 
 stimulus measure, the sensation result will depend on a 
 product of the form a ■ R, where a is a measure of the retinal 
 excitability for R, a value dependent on the tuning ; in 
 determining the effect produced in any given case of re- 
 tuning, we have only to consider the value of a. Suppose we 
 have two retinal points A and B differently tuned, and R^ 
 falling on A matches Rg at B, S^ at A similarly matching Sg at 
 B. Hence (adopting the ordinary convention as to equality), 
 Rj^ + Sj = R2 + Sj. 
 
 Now let A and B be retuned by exposure to light of stimulus 
 value T, which alters the excitability for S^ and S^ respectively. 
 We now have at A, R^ + aS]^, and for B, Rg + aSj, i.e. R^ + S^ 
 — (1 — a.)Si, and R2 + S2-(l-(x)S2, which must match, since 
 by hypothesis S^ at A is equal to S2 at B. This is the 
 
 1 V. Ki-ies, Nagel's Handb., iii. p. 210. 
 
160 PHYSIOLOGY OF THE SPECIAL SENSES 
 
 essence of tlie so-called " Law of Coefficients," and is a short 
 way of stating the experimental facts mentioned above. 
 The law obviously fails in the case of feeble stimuli, because 
 then the autocthonous stimulus (the light chaos) becomes 
 measurable relatively to Sj, Sg, and T, and can no longer 
 be neglected, as in the above expressions. Let us next ex- 
 amine some matters of detail. 
 
 First, what happens when the retuning and reacting 
 lights are identical ? This merely requires for answer a study 
 of the sensation changes accompanying prolonged fixation of 
 any object, and we all know the answer. Brightness differ- 
 ences grow smaller and smaller, being at length effaced, 
 which suggests a relationship between the degree of re- 
 tuning and the strength of the stimulus. 
 
 Next, suppose we retune as before with white, but use 
 a chromatic reactor. We find (under conditions of light 
 adaptation) that the reactor must have the same qualitative 
 composition as the comparison, but greater intensity. 
 V. Kries in his experiments obtained a good match if the 
 amounts of coloured light in reactor and comparison were 
 approximately in the ratio of 3 or 4 to 1. With a colour 
 mixer, the blue sectors were 270° : 97° ; the red, 270° : 84° ; 
 the yellow, 270° : 97°. If the coloured sectors were made 
 equal, and the white so chosen that the discs were 
 approximately equal in brightness, the sensation excited in 
 the retuned area was always that of far too low saturation. 
 White retuning, therefore, seems to change chromatic 
 stimulus values ; but the experiments upon which this con- 
 clusion is based have been adversely criticised by Hering, 
 whose objections will be considered in a later chapter, 
 when we come to deal with the theoretical side of the 
 question. 
 
 The next case is that of retuning with a colour, and here 
 also the simplest example is when retuner and reactor are 
 the same. Without a comparison, loss in brightness and 
 saturation are apparent; with a comparison, a difference 
 in colour tone may also be perceptible. Voeste's results 
 are as follows : A yellow (560 fxfj), a green (500 jj-ij.), 
 and a blue (460 fj-n) undergo no change in hue on pro- 
 
AFTER-IMAGES OF SUCCESSIVE INDUCTION 161 
 
 longed fixation; intermediates pass away from the green 
 towards the red or the blue. Thus a light between 500 and 
 460 ixfji, will match a comparison of less wave length than 
 itself; one between 500 and 560 i^ti, a light of greater wave 
 length. 
 
 If we use a grey field as reactor it produces a sensation 
 closely allied to that normally associated with the retuner's 
 complementary, and this is not a mere blend of the sen- 
 sation obtained in the darkened eye after retuning plus 
 the whiteness normally excited by a grey field. If this 
 were so, after retuning with a given yellow the subjective 
 blue could always be compensated by mixing the same 
 amount of yellow with the reacting field so that it might 
 appear colourless. As a matter of fact, however, the more 
 white there is in the reacting field, the more yellow must 
 be mixed with it — as we shouid expect from the coeflScient 
 rule. If the reactor be grey, the complementary of the 
 retuner may be very well marked; if the reactor be the 
 complementary itself, a sensation of colour is produced far 
 exceeding in purity that associated with any region of the 
 spectrum; this result is universally admitted to be of con- 
 siderable theoretical importance, as we shall find later. 
 Finally, if some colour not complementary to the retuner 
 be used as reactor various sensations are experienced, 
 although in general the complementary of the retuner 
 predominates. After red retuning, v. Kries matched a re- 
 acting yellow (589 fjufx) with a green-yellow (556 yu/i); 
 after green retuning, it matched an orange (605 /*/i). Hess, 
 after blue retuning, matched a reacting 517 ^j" with a 
 comparison 565 /^/-t. These experiments are interesting, 
 but a systematic study of the eftects of retuning when each 
 spectral colour is used in turn as a reactor, with a com^ 
 parison field, has yet to be effected, and we cannot at present 
 lay down any general laws. 
 
 The time relations of the process of retuning are of 
 some interest. Many researches have been published, the 
 technique of some being simple, of others elaborate. One 
 plan is easy; a white object on a dark ground is fixated, 
 and one ascertains, after a definite interval, what comparison 
 
 L 
 
162 PHYSIOLOGY OF THE SPECIAL SENSES 
 
 light placed near the object matches it at the moment of 
 observation. The table contains v. Kries' results. By a 
 similar method, we can measure the rate at which the 
 effect passes off with a definite amount of retuning. Un- 
 fortunately, the diflSculty of avoiding two important sources 
 of error — adaptation and extra-foveal stimulation — is great, 
 and we can lay but little stress on the results. For instance, 
 it is not known whether the effect of retuning passes off 
 smoothly, as Fechner and Helmholtz believed, or in pulses, 
 
 Changes in 
 
 White 
 
 Value 
 
 {v.K'i 
 
 'ies). 
 
 
 
 strength erf Stimulus 
 (Arbitrary Units). 
 
 Apparent Diminution After 
 
 Seconds. - 
 
 3 
 
 6 • 
 
 10 
 
 20 
 
 40 
 
 80 
 
 160 
 
 1 
 
 •91 
 
 •81 
 
 ■66 
 
 •58 
 
 •43 
 
 •23 ' 
 
 .,5 
 
 1-9.5 . . 
 
 •8G 
 
 •74 
 
 •62 
 
 •52 
 
 •32 
 
 •18 
 
 •09 
 
 3-9 
 
 •82 
 
 •71 
 
 •62 
 
 •34 
 
 •21 
 
 •14 
 
 •08 
 
 34-7 
 
 ■74 
 
 •57 
 
 •42 
 
 •25 
 
 ■16 
 
 •08 
 
 •03 
 
 as held by Aubert. It is likely that some of Aubert's 
 evidence is vitiated by imperfect technique, but his hypo- 
 thesis cannot be disproved. 
 
 Exner attempted to measure the change in sensation from 
 the moment at which the stimulus is applied by an ingenious 
 method, which has been used in a modified form by Burch. 
 This method depends on the fact that with a given stimulus 
 the sensation increases to a maximum, then, in consequence 
 of retuning, diminishes. At a definite instant a stimulus, 
 eg. & bright semicircle in a dark field, is applied ; then after 
 a known short interval a bright field is exhibited; finally, 
 after another known interval a dark field appears. In this 
 way two neighbouring points have the same stimulus applied 
 to them, but on one the stimulus begins to act a fraction 
 of a second earlier than on the other. If the first stimulus 
 
AFTER-IMAGES OF SUCCESSIVE INDUCTION 163 
 
 effect has reached its maximum but the second has not, the 
 resultant after-image will be positive; if both have passed 
 their maxima, the after-image of the semicircle will be 
 negative, because the first excitation process will have fallen 
 to a lower level than the second. Between these extremes 
 we have a point at which no differentiation appears, the field 
 seeming to be uniform. This latter result will be obtained 
 when one stimulus has just passed, the second just fallen 
 short of its maximum, and accordingly gives a measure of 
 the time necessary to attain the first maximum; 
 
 Ingenious as are these experiments, their value as mea- 
 sures of the time relations in retuning is not, perhaps, very 
 great. 
 
 Evidently they depend on the assumption that sensational 
 intensity changes after the stimulus has been withdrawn, 
 just as it does while the stimulus is still acting. Our study 
 of recurrent vision, particularly the brightness relations of 
 the tertiary image, has taught us that this assumption is 
 of doubtful truth. We thus know but little of the excitation 
 changes quantitatively, and even from the qualitative stand- 
 point our information is not complete. The colour changes 
 undergone by the after-image of a colourless object gene- 
 rally form a white-red-green-red-blue cycle. The image of 
 a coloured object passes from positive to negative (com- 
 plementary) through a colourless or reddish-white phase. 
 It is quite certain that the duration of after-image effects 
 is much longer than has been generally thought. Burch 
 passed from an ordinary into a completely darkened room 
 and noted the sensations experienced with the following 
 results : — 
 
 " 0*5 mins.^ Confused after-images of portions of objects 
 recently looked at. These images gradually fade and 
 give place to a luminous fog, made up of what I have 
 called dazzle tints, i.e. coloured impressions of luminosity 
 without form. 
 
 " 10-15 mins. Fog no longer uniform but ' spotty,' with 
 patches of a brownish or greenish-bronze colour. 
 
 " 18-21 mins. A group of yellow or green luminous points. 
 
 > G. J. Burch, Proc. Roy. Soc, B., 1905, pp. 212-213. 
 
164 PHYSIOLOGY OF THE SPECIAL SENSES 
 
 " 23-28 mins. Green dazzle tints predominate, beginning 
 to break up, in some cases, into patches of green showing a 
 fainter blue between. . . . 
 
 " 106-120 mins. Gradually a luminous fog seems to iill 
 the surrounding space. For a minute or two it increases in 
 brightness, and then breaks up near the middle in two or 
 three places and seems to roll away on all sides. Then it 
 returns, and again breaks up, generally leaving an island 
 about the size of the yellow spot, which lasts two or three 
 seconds longer than the rest. The colour of these clouds is 
 a rich pure violet, like that of the calcium lines H and K in 
 the arc spectrum. This phenomenon may go on from ten 
 minutes to half-an-hour, the violet patches getting smaller 
 and appearing at longer intervals till they die away." 
 
 To sit patiently in a dark room studying the time relations 
 of after-images may seem crude and uninteresting to those, 
 who associate experimental research with expensive apparatus, 
 and cannot be made to realise that some of the most brilliant 
 achievements of physiology have been effected with the help 
 of little except ability. It is probable that much-needed 
 light will be thrown on the process of retuning by an exten- 
 sion of work, such as that of Burch. A series of experiments 
 must be carried out, the eye being exposed for a definite 
 time to a given light before each period of observation and 
 the resultant changes are compared. In some such fashion 
 we may be able to unravel the complicated tangle of results 
 already mentioned. 
 
 Summing up the definitely ascertained conclusions, we 
 find:— 
 
 (1) The distinction between positive and negative after- 
 images is not absolute but relative, depending on the nature 
 of the reacting stimulus. 
 
 (2) An image-producing or returning stimulus changes 
 the stimulus value of a reacting (subsequently applied) 
 light, but only in such a way that the sensation-response 
 following exposure to the reacting light is increased or 
 diminished quantitatively. Colour equations do not lose 
 their validity. 
 
 (3) The latter statement, as expressed in the " Coefficient 
 
AFTER-IMAGES OF SUCCESSIVE INDUCTION 165 
 
 Law," is true in the case of foveal vision, but not for peripheral 
 stimulation. 
 
 (4) We do not know the time relations of the retuning 
 process, nor whether the latter proceeds uniformly or in 
 pulses. 
 
 I have now reviewed briefly the chief experimental data 
 upon which a theory of visual sensations (exclusive of spatial 
 perceptions) must rest. The question of simultaneous contrast 
 has indeed been omitted, but owing to certain peculiarities 
 of this branch of the subject it is best postponed until the 
 more generally interesting theories of colour vision have been 
 examined. We shall therefore now pass on to the theoretical 
 side of our investigation, and endeavour to co-ordinate the 
 apparently heterogeneous facts in some logical manner. We 
 shall see that, wide as are the apparent differences of opinion, 
 and many as are the difficulties to be encountered, both the 
 main theories embody something of interest and importance. 
 It will add to the clearness of our discussion if a preliminary 
 chapter is devoted to some ancient theories of vision, and 
 these will be examined in the following pages. 
 
 BOOKS AND PAPERS BBCOMMENDBD FOR 
 FURTHER STUDY 
 
 Helmlwltti, Hand. d. Physiol. Optik., second edition, pp. 501, etc. 
 
 V. Kries, Nagel's Handb., vol. iii. pp. 205, etc. 
 
 Tschermak, Ueber das Verhaltnis von Gegenfarbe, Kompensationsfarbe 
 
 und Kontrastfarbe (Pfliig. Arch., 1907, cxvii. 204). 
 Goethe's Farbenlehre, Eastlake's trans. 
 
 F. Klein, Nachbilder, Uebersicht und Nomenklatur (Engelmann's Arch. f. 
 
 Physiol., 1908, Supp. Bd. p. 219). 
 
 G. J. Burch, Proc. Roy. Soc, 1900, Ixvi. 204. 
 
CHAPTEE XVII 
 
 HISTORICAL THEORIES OF VISION 
 
 In attempting to estimate the scientific value of ancient 
 theories of vision, it is necessary to bear in mind certain 
 limitations which were imposed by incomplete means of 
 investigation. Even with modern apparatus it is not easy to 
 obtain a precise knowledge of the elaborate structures con- 
 tained in the eye, hence workers unprovided with the simplest 
 microscope knew almost nothing of facts which form part of 
 the mental equipment of every educated man in our own 
 time. Roughly, we may say that all the early theories agree 
 in regarding the " pupil " of the eye and the " image " within 
 it as of primary importance. Again, the flash of light seen 
 on pressing or rapidly moving the eye was held to prove the 
 existence of an inherent or native " fire," also of great signi- 
 ficance. Thirdly, the presence of a watery substance within 
 the eye required some explanation. The problem, as it pre- 
 sented itself to the earliest writers, was to assign their proper 
 shares in the visual act to the " fire," the " image," and the 
 "water." As knowledge of psychology or physiological 
 psychology progressed, the theories became more intricate, 
 and it is possible to distinguish between theories of physio- 
 logical optics, theories of colour vision, and more general 
 inquiries as to the nature of perception in general and spatial 
 perception in particular. Work under the last heading is 
 much the most subtle, and has proved of enduring interest 
 and value for modern thinkers. The speculations as to 
 anatomy and physiological optics are, for the reasons above 
 recited, of little importance ; while the theories of colour vision 
 occupy an intermediate place, having in some respects proved 
 fruitful, in others barren. As any one who is familiar with 
 the subject will recognise, it is not in all cases easy to separate 
 
 out an author's theory of vision into the sections enumerated, 
 
 i«t 
 
HISTORICAL THEORIES OF VISION U7 
 
 and my account of the matter will not be so clear as I could 
 wish. I shall try in this chapter to outline the main theories 
 of vision, especially in regard to colour vision ; purely ana- 
 tomical or physical theories will not be examined. 
 
 One of the earliest Greek writers on this subject was 
 Alcmseon of Cretona (fl. 500 B.C.), but our knowledge of his 
 opinions is fragmentary. He appears to have thought that 
 seeing is accomplished by the passage of rays from the 
 ocular " fire " to the object, and that these returning to the 
 eye, altered in some way, are reflected in the diaphanous 
 "water." The fire is therefore the active element in vision. 
 It is not clear how Alcmason harmonised the conception of a 
 visual ray from the " fire " with that of mirroring in the "water." 
 
 Empedocles (circa 450 B.C.) evolved a more subtle theory, 
 but it is not easy to reconcile different statements attributed 
 to him. According to the first doctrine enunciated by 
 Empedocles, like perceives like. All bodies are characterised 
 by the following properties : (1) All are made up of four 
 elements — earth, air, fire, and water. (2) All are permeated 
 by minute passages or pores, and all give off emanations 
 which enter the pores. Thus, in perception, emanations 
 from the object pass into the pores of the percipient organ. 
 But that this passage may be effected it is necessary that 
 the emanations and the pores should correspond; if the 
 former be too large or too small for the latter, no percep- 
 tion can occur. Hence with the eye alone can we perceive 
 emanations of colour, because these are " symmetrical " with 
 the pores of the eye alone. This correspondence is the basis 
 of sense specificity. Further, there is a symmetrical arrange- 
 ment within the eye itself with respect to the different forms 
 of stimulation. By means of the intra-ocular fire we per- 
 ceive the emanations of fire — i.e. white ; with the water we 
 see water — i.e. black ; and so on. " With earth we see earth, 
 with water we see water, with air we see the bright air ; just 
 as with love we (perceive) love, and with hate, baleful hate." ^ 
 Empedocles is said to have regarded four colours — white, 
 black, red, and green — as primaries (Stobseus), but only 
 
 1 Beare, Greeh Theories of Elementary Cognition from Ahmceon to Aristotle, 
 p. 18. 
 
168 PHYSIOLOGY OF THE SPECIAL SENSES 
 
 examines black and white in detail. He also taught that 
 rays issued from the visual "fire," but how he associated 
 this with his general doctrine of pores and emanations is 
 not clear. It is probable, judging from some passages in 
 Lucretius' great poem,i that a somewhat similar doctrine 
 was held by Epicurus. 
 
 Democritus (? 460-357 B.C.) agreed with Empedocles in pos- 
 tulating the entrance of particles from an object into pores 
 contained in the perceiving structure, and in the dictum that 
 " like is perceived by like." But he denied that there were 
 four qualitatively distinct elements, believing that all things 
 were made up of homogeneous atoms moving in a vacuum and 
 infinitely numerous. Vision is due to the mirroring of an 
 object in the eye, the latter's character being somehow 
 determined by its moist and porous nature. This part of 
 Democritus' theory was sharply criticised by Aristotle, who 
 remarked, " It is absurd that it should not have occurred to 
 him to doubt why the eye alone sees, but nothing else in 
 which energies are apparent. That the sight is aqueous is 
 true; yet it does not happen that it sees because it is 
 aqueous, but because it is diaphanous, which is also common 
 to air." 2 
 
 Democritus seems to have been the first writer to attempt 
 a detailed theory of colours, the simple ones being white, 
 black, red, and green. White is the smooth, because any- 
 thing which is not rough, does not throw shadows, and is 
 not difficult to penetrate, is bright. Black consists of atomic 
 figures of an opposite kind, viz. those which are rough, 
 uneven, and dissimilar ; on this account they cast shadows, 
 and their pores are neither straight nor easily permeable. 
 Red is formed of the same kind of atomic figures as the hot, 
 but the figures of red are larger. A proof that red is com- 
 posed of such atoms as those forming hot is, that " we our- 
 selves are red when heated, just as other things are when 
 ignited, as long as they continue to have the character of 
 ' the igneous ' ; but ignited things are redder in proportion 
 as they are formed of large figures, such as flame, coals, or 
 
 » Lucr., De Rer. Nat., ii. 833. 
 
 ' Aristotle, De Sensu et SensUibus, Taylor's trans., vol. iii. 133. 
 
HISTORICAL THEORIES OF VISION 169 
 
 wood, whether green or dry, and also iron and other metals 
 which are subject to ignition."^ Green is formed of the 
 solid and the void, the tint varying with their position and 
 arrangement. Each colour is purer the more the figures of 
 which it is essentially composed are free from admixture 
 with others. All other colours are mixtures of these four. 
 For instance, purple is mixed from white, black, and red, 
 red being in largest and black in smallest amount. Owing 
 to white coming midway, the colour is pleasant to the sense. 
 
 It is hardly possible for us to criticise this theory with 
 understanding. Here, as so often in considering ancient 
 science, one is conscious of the gulf which separates us from 
 the world which has passed away. Much which seems 
 meaningless or nonsensical probably only does so because 
 not only has the author's work come down to us in a 
 fragmentary state, but it is impossible to see things from 
 his point of view, or even to find out what that point of view 
 was. In ancient times, Theophrastus, who no doubt under- 
 stood it better than we do, criticised Democritus' theory in 
 so far as it deals with colour. He says that Democritus made 
 a difficulty by suggesting four primaries rather than two — 
 black and white — and objects that the position of the atoms 
 rather than their shape or figure should be the cause of 
 colour. The matter is, however, so obscure, that it can 
 hardly be pursued in this place. 
 
 The views of Anaxagoras (499-428 B.C.) and of Diogenes 
 of Appolonia (fifth century B.C.) can, for similar reasons, be 
 passed over rapidly. Anaxagoras held, in opposition to his 
 contemporaries, although the opposition is more formal than 
 real, that perception is the result, not of like operating on 
 like, but of the reaction between contrary and contrary. 
 The "image" is not reflected upon a part of like colour 
 to the object, but upon a different colour. 
 
 Diogenes, who believed that an all-pervading air was the 
 ultimate agency in nature, has left no distinct theory of 
 colour vision. 
 
 We now come to one of the great names in scientific 
 history, that of Plato, whose views, although far less 
 
 ' Beare, op. cit., p. 32. 
 
170 PHYSIOLOGY OF THE SPECIAL SENSES 
 
 suggestive than those of Aristotle, are well worth con- 
 sideration. 
 
 Plato has considered the problem from two different points 
 of view. His account of the physical side of the matter, as 
 one may term it, is contained in a passage in the I'imcBus, 
 of which the following quotation gives an idea : — 
 
 " And of the organs they first contrived the eyes to give 
 light, fixing them by a cause on this wise. They contrived 
 that as much of fire as would not have the power of burning, 
 but would only give a gentle light, the light of everyday life, 
 should be formed into a body ; and the pure fire which is 
 within us and akin to this they made to flow through the 
 eyes in a single entire and smooth substance, at the same 
 time compressing the centre of the eye so as to retain all the 
 grosser element, and only to allow this to be sifted through 
 pure. When therefore the light of day surrounds the stream 
 of vision, then like falls upon like, and there is a union, and 
 one body is formed by natural affinity according to the 
 direction of the eyes, wherever the light that falls from 
 within meets that which comes from an external body. And 
 everything being affected by likeness, whatever touches or 
 is touched by the stream of vision, their motions are diffused 
 over the whole body and reach the soul, producing that per- 
 ception which we call sight." ^ 
 
 . In the genesis of colour, particles are discharged froni 
 external things and impinge upon the eye, some being larger, 
 some smaller, and some equal in magnitude to the parts of 
 the eye. All colours are compounded of four — white, black, 
 bright, and red. Bright when mixed with red and white 
 becomes golden-yellow; red blended with black and white 
 yields violet. 
 
 From statements in the TiTuceus and Republic it 
 would seem that Plato, unlike Democritus, believed in the 
 objective existence of colour, but, as Helmholtz pointed out, 
 his views seem to have varied. In the Thecetetus, colour is 
 considered from a different standpoint, as will be clear from 
 the next quotation : " We shall see that every colour, white, 
 black, and every other colour, arises out of the eye meeting 
 
 1 Plato, Timceus, Jowett's trans., ii. 375. 
 
HISTORICAL THEORIES OF VISION 171 
 
 the appropriate motion, and that what we term the substance 
 of each colour is neither the active nor the passive element, 
 but something which passes between them and is peculiar to 
 each percipient. . . , 
 
 "When the eye and the appropriate object meet together 
 and give birth to whiteness, and the sensation of whiteness 
 which could not have been given by either of them going to 
 any other object ; while the sight is flowing from the eye, 
 and whiteness from the colour-producing element, the eye 
 becomes fulfilled with sight and sees, and becomes not sight, 
 but a seeing eye ; the object which combines in forming the 
 colour is fulfilled with whiteness and becomes not whiteness, 
 but white." ^ 
 
 It may be said that Plato's views are interesting, but do 
 not embody any complete theory of colour vision. 
 
 Aristotle (384-322 B.C.) elaborated a theory which is not 
 the least important of his contributions to natural know- 
 ledge and, clothed in modern notation, still survives. 
 
 According to Aristotle, the object of sight is colour. Colour 
 is at the surface of all visible bodies, but in order to be seen 
 requires the presence of light, which is the medium of vision. 
 This basal proposition formed part of the teaching of Epi- 
 curus, to judge from a notable passage in the work of his 
 greatest exponent,^ but in developing the conception Aristotle 
 far surpassed Epicurus. 
 
 Light, says Aristotle, presupposes a diaphanous substrate, 
 which in its turn is the medium of light. Examples of this 
 " diaphanous " are air, water, and many solids. The realisa- 
 tion or actualisation of this potential quality of being dia- 
 phanous is light, its absence darkness. When the former 
 condition of actual light is established in the diaphanous 
 medium, any coloured body sets up a movement in it be- 
 tween object and eye ; this is the essential process in colour 
 perception. The diaphanous substrate upon which depends 
 the existence of light and, a fortiori, colour is not peculiar 
 to the bodies called transparent or diaphanous, but is a 
 species of universally difi'fised natural power ; it is not indeed 
 
 ^ Plato, Thecetetus, Jowett's trans., vol. iii. 538-9. 
 
 2 Lucretius, De Ra-wn Nat., ii. 730-833, especially 795-8. 
 
172 PHYSIOLOGY OP THE SPECIAL SENSES 
 
 capable of existence independently of " body," but subsists in 
 varying degrees in all things. The colour of a body either 
 forms its surface or is upon its surface, the latter opinion 
 being the more exact since the indeterminate " diaphanous " 
 of air and water exhibits colour, -which, however, owing to 
 the indeterminate boundary, is variable. This explains the 
 changing hues of sea and sky. 
 
 Bodies with a definite boundary have a fixed colour, so 
 that one might again define colour as the surface limit of the 
 diaphanous in determinately bounded body. This definition 
 is consistent with the first given, viz. that which stimulates 
 the actualised " diaphanous " (light) between the object and 
 the eye, but the latter is a definition in terms of vision and 
 the medium of vision, the former in terras of the object as 
 it exists apart from vision. 
 
 Colour is a genus comprising seven species ; it is a quality, 
 and cannot therefore exist without a substrate. The seven 
 species are white, black, golden-yellow, crimson, violet, leek- 
 green, and deep blue. The colour genus (like all other 
 genera of sensible qualities) consif^ts of species lying between 
 extremes ; outside these extremes there can be no colours, 
 between them are specific boundaries. By subdividing 
 the scale limited by the extremes, we cannot obtain an 
 infinite number of distinct colours, because a sensible quality 
 is discrete, not continuous. By dividing the substrate we do 
 not arrive at any new colour ; the halves of a white object 
 are white. It is true that by sufficiently fine division no 
 colour whatever may be perceptible, but on reuniting these 
 portions we again obtain white. The two limits are black 
 and white; when one is actually existent, the other is 
 potentially existent. The transition from white to black is 
 effected through the successive degrees, which are the species 
 of colour. The substratum, of which these are the qualities, 
 is one, and is in strictness that which is changed ; the colours 
 alternate. 
 
 Colour is not purely subjective ; it is true that it depends 
 upon the eye, but it also depends upon the object. Actual 
 colour depends upon the possibilities of these two being 
 realised together, but the coloured object existed in nature 
 
HISTORICAL THEORIES OF VISION 173 
 
 as a potential colour before the act of vision, and apart 
 from it. " It is light that at once transforms the potential 
 colour to actuality, and the potentially seeing to an actually 
 seeing eye."^ 
 
 In the colour scale (as among the elements) there is a 
 sort of opposition of positive and negative. White is the 
 positive, black, the negative. This is Aristotle's general 
 doctrine of colours ; he also treats certain of them in detail. 
 
 The presence of some fire-like element is the cause of 
 light in the diaphanous, and in its absence we have darkness. 
 In all determinately bounded bodies we may assume some- 
 thing analogous with the presence and absence of this fiery 
 element. Its absence means blackness, its presence white- 
 ness. Therefore, in determinately bounded bodies, blackness 
 is privation of whiteness. Blackness and whiteness are con- 
 traries within one sensory province, that of colour, and 
 from them all the other colours are to be explained. " The 
 transition from white to black is possible through continuous 
 degrees of privation: that from white to black is likewise 
 possible by an ascending scale in the opposite direction. 
 The various colours are species which fall between the two 
 contraries, and' are generated of certain combinations of 
 these." ^ For instance, in passing from white to black we 
 first come to crimson. As the intervening stages in the pas- 
 sage mark relative extremes, change can start from any point. 
 
 With regard to the actual mode of origin of the inter- 
 mediate colours, what is actually effected in the above- 
 mentioned process ? Aristotle's views on this point are not 
 quite consistent in his different writings, but he appears to 
 condemn the doctrine of atomic juxtaposition and that of 
 superposition, taking the view that a complete blending 
 occurs: by blending is meant a compound or mixture of 
 so intimate a nature that no individual part retains its 
 original qualities unmodified. 
 
 The colour spoken of as grey is sometimes described 
 as standing at the mid-point of the , scale between black 
 and white, sometimes as a sort of relative black. Golden- 
 yellow is also nearly akin to white. Red is produced by 
 
 ' Beare, op. cit., p. 64. " Ibid., p. 69. 
 
174 PHYSIOLOGY OF THE SPECIAL SENSES 
 
 light streaming through black, as when the sun shines 
 through a fog. Purple is akin to crimson, but differs from 
 it in having more of the dark constituent. " Sometimes 
 the light of a lamp shows not white but purple, the ray 
 that is sent from it being feeble, and being reflected from 
 a dark colour. This increasing feebleness of the ray brings 
 us from purple to leek-green and violet successively. The 
 stronger ray yields crimson against the dark ground (or 
 when mixed with dark); the next in strength gives leeJe- 
 green; the weakest, violet."^ 
 
 In this connection, Aristotle refers to positive and nega- 
 tive after-images. After looking at the sun and closing 
 the eyes we see the object first of the same colour as before ; 
 this changes to crimson, then to purple, then to black, and 
 finally vanishes. The order illustrates the genesis of colours 
 from the blending of white and black. Simultaneous con- 
 trast receives an explanation along these lines — the brightest 
 rainbow is seen in the darkest cloud, white wool has its 
 colour intensified when placed next black wool, etc. Aris- 
 totle rejected altogether the theory of emanations and pores, 
 while his conception of a vibratile movement imparted to 
 the actualised "diaphanous" may, perhaps, be regarded as 
 a partial anticipation of the modern doctrine of a lumini- 
 ferous aether. We cannot, however, push the comparison 
 far, since he maintained, in opposition to Empedocles, that 
 light does not travel. 
 
 The eye itself is like other organs to be defined in terms 
 of its function, and thus is an eye only so long as it can 
 see. The eye of a corpse is so called only in a somewhat 
 incorrect way. The eye is the organ which is stimulated 
 by colour, but this process of stimulation must in some 
 way be transmitted to the " soul." It seems to follow that 
 the diaphanous medium which acts objectively is also 
 functioning within the eye itself in order to transmit the 
 stimulation inwards. The eye consists of heterogeneous 
 parts; that part which is specially concerned in vision is 
 the part generally translated pupil, but probably, at least 
 after the time of Empedocles, meaning rather the crystalline 
 
 ' Beare, op. cit., p. 75. 
 
HISTORICAL THEORIES OF VISION 175 
 
 lens. Surrounding this internal moist part comes what 
 Aristotle terms the black (? iris), and outside of this again 
 is the white (? sclerotic). The pupil and vision are to the 
 eye what body and soul are respectively in the whole 
 living creature.^ 
 
 For perfect vision there must be a proper amount of 
 moisture in the eye. Those creatures which have too little 
 see well by night but ill by day, because owing to the 
 deficiency the eyes are over-stimulated by daylight. The 
 other group, with too much moisture, can see well by day- 
 light but badly in the night, because there is not relatively 
 to the water enough fire in the eye. The membrane which 
 covers the "pupil" must be transparent, white, thin, and 
 smooth. 
 
 Aristotle rejects the theory that the eye consists of fire, 
 a theory which, as we have noticed, rests on the obser- 
 vation of phosphenes (flashes of light seen when the eye is 
 rapidly moved or when it is pressed). If, asks Aristotle, 
 the eye is igneous, why do we not always see these phos- 
 phenes, instead of only under exceptional circumstances ? 
 Further, why does not the eye see itself always ? Besides, 
 if the visual part of the eye were really fire, we ought to 
 be able to see in the dark, and Plato's explanation of the 
 circumstance that we cannot see in the dark, viz. an ex- 
 tinction of the visual ray in the darkness, is inadmissible. 
 Such fire as is made with coals may indeed be extinguished 
 by cold and moisture, but not light. 
 
 The previous pages will be sufiicient to give the reader 
 some idea of the nature of Greek thought respecting the 
 visual processes. A few general remarks may fittingly con- 
 clude the chapter. 
 
 For the reasons mentioned at the beginning, the state- 
 ments respecting the anatomical and physical side of the 
 problem of sight are of little interest. It cannot be said 
 that even Aristotle's view of the matter is particularly 
 helpful. It will have been noticed how the dogma of like 
 producing like hampered work on this side, already suffi- 
 ciently difficult, and the value of Miiller's doctrine {vide supra, 
 ^ See Beare, op. cit., p. 80. 
 
176 PHYSIOLOGY OF THE SPECIAL SENSES 
 
 Chapter I.) receives a practical illustration. Since the external 
 medium is transparent, there must be some internal trans- 
 parency ; since " fire " is visible, there must be some internal 
 "fire" by which it is perceived, and so on. Anaxagoras 
 might appear to be an exception, but a little thought con- 
 vinces one that the exception is only formal. In adopting 
 the converse of the dogma, like produces like, he too stands 
 committed to the belief that there is some necessary con- 
 nection in kind between the processes occurring within us 
 and those conceptually existent outside of us. 
 
 The theories of vision as a problem in physiological 
 psychology, on the other hand, stand on quite a different 
 plane. As we shall see in the chapter on Hering's theory 
 of visual sensations, Aristotle's teaching was the basis of 
 Goethe's theory of colours, and the doctrine of Goethe, freed 
 from certain mathematico-physical absurdities, partly a result 
 of imperfect physical knowledge, has been transformed by 
 Hering into a hypothesis of rare ingenuity and intellectual 
 value. 
 
 To pursue the history of visual theories from tho Greek 
 period to the modern epoch would be of little more than 
 antiquarian interest. In another chapter I shall refer to 
 the work of Berkeley in connection with the theory of space ; 
 his contribution to the theory of colour vision in the strict 
 sense is not of much importance from the physiological 
 standpoint. 
 
 RECOMMENDED FOR FURTHER STUDY 
 
 There ia a considerable literature dealing with the sense physiology 
 and psychology of the ancients, but it is mainly written from the stand- 
 point of the psychologist or that of the philologist. The reader will find 
 ample material for extending his knowledge in the work of Professor 
 Beare, to which I have frequently called attention. Jowett's trans- 
 lation of Plato's Dialogues can be profitably consulted for the earlier 
 speculations. The standard work on Aristotle is Aristotles ueber die 
 Farben, von 0. Prant, Miinchen, 1849. 
 
CHAPTER XVIII 
 
 THE YOUNG-HELMHOLTZ THEOKY OF 
 COLOUE VISION 
 
 As I attempted to show in tlie chapter on normal vision, 
 Newton's researches demonstrating the (conceptually) com- 
 plex nature of white light and the physical substratum of 
 chromatic stimuli — as co-ordinated by the undulatory hypo- 
 thesis — enable us to form a coherent, but not necessarily, 
 or even probably final, account of the physical elements in- 
 volved in retinal stimulation. On the other hand, the 
 specificity of physiological response, which finds a not quite 
 accurate expression in Miiller's law, releases us from many 
 of the difficulties attendant upon the primitive theories of 
 vision, the propounders of which were hampered by the 
 dogma of physico-physiological identity. The acceptance of 
 .these facts has left wide scope for ingenuity in the con- 
 struction of theories as to the psycho-physiological mechanism 
 which links together the subjective and objective visual 
 worlds; but the liberty of choice which is thus accorded 
 to the scientific imagination has its own disadvantages. 
 Perhaps the most obvious is the impossibility in which we 
 stand of " proving " — in the more popular sense of that much 
 abused word — any theory of colour vision to be true or false. 
 So long as it was held that a conceptual system of physics 
 was a valid physiological and psychological system, that our 
 physiological categories were to be tested against our physical 
 classifications, it was comparatively easy to judge and condemn 
 a theory of vision. We now hold that the writ of the physical 
 court does not run in the kingdom of physiology, and we do 
 not know to what jurisdiction a physiological theory of 
 colour vision is amenable. This is one of the reasons, in 
 fact the chief reason, why an enormous number of rival 
 theories are in the field, and why their study has seemed to 
 
 1" M 
 
178 PHYSIOLOGY OF THE SPECIAL SENSES 
 
 many highly educated men as unprofitable as the mediseval 
 disputes of the schoolmen. To those who have the leisure, 
 and the particular form of intellectual curiosity which is 
 attracted to the study of human ingenuity divorced from 
 the immediate phenomena of nature, a complete history of 
 modern theories of vision would be instructive and valuable ; 
 but by those Avhose leisure is less ample, and whose curiosity 
 does not take this form, a different path must be followed. 
 One must consider the theories of vision which, it may be, 
 rather from the energy of their propounders than from 
 any overwhelming speculative merit in the theories them- 
 selves, have borne the most fruit in respect of experimental 
 observations and methods. Adopting this criterion, two 
 theories stand out above the rest, those of Helmholtz and 
 Hering. If, therefore, these theories only are discussed in 
 the present work, it is not because I hold them to be 
 necessarily' superior as theories to others — for instance, those 
 of Ladd-Franklin and Schenck — but because they have borne 
 so much more experimental fruit. In a famous passage, 
 R. L. Stevenson describes a conversation. " What I advance," 
 said one, " is true." " Yes," replied the other, " but not the 
 whole truth." " Sir," was the retort, " there is no such thing 
 as the whole truth." In what follows, the reader should 
 have that aphorism constantly in his mind. 
 
 In attempting to arrive at an adequate interpretation of 
 any class of experimental facts, it is permissible to advance 
 by more than one route, just as many problems in geometry 
 can be solved by analytical as by pure geometrical methods. 
 As a rule, it does not much matter which path we choose, so 
 long as we keep count of any assumptions made and avoid 
 the introduction of unnecessary steps. We shall now start 
 from the experimental laws of colour mixing and see whither 
 we arrive by following the most obvious route. We found 
 that for most experimental purposes, we could say that a 
 sensation " produced by " a colour stimulus could be matched 
 by a sensation due to a stimulus obtained by mixing not 
 more than three lights together. We saw that these three 
 lights did not coincide accurately with any spectral colours, 
 but that if we admitted negative values into our colour 
 
YOUNG-HELMHOLTZ ON COLOUR VISION 179 
 
 equations, our vision could be regarded as definitely tri- 
 chromatic even in terms of known stimuli. In order, how- 
 ever, to avoid this, let us so choose our stimuli that only 
 positive values of each are employed. This means that they 
 must be so chosen that the colour " triangle " is circumscribed 
 by the lines joining their representative points in a plane, e.g. 
 
 Red/i- iWolet' 
 
 Fia. 22. 
 
 This is our first assumption, and is little more than a 
 generalisation of experimental facts. 
 
 Thus far, we have merely asserted that a stimulus R', 
 say, can be expressed by the equation R' = a;.R+2/.G-)-2:.V, 
 where x, y, z are real positive quantities. But our only 
 measure of equality of stimulation being corresponding 
 equality of sensations, we imply (and this is our second and 
 most important assumption) that there is a definite relation- 
 ship between the physiological excitatory processes which 
 lead up to sensations and the stimulus magnitudes. 
 
 This assumption is justified if we can show that it leads 
 to the formulation of a useful working hypothesis, and if 
 we assume the simplest relationship consistent with the ex- 
 perimental facts. What, then, is the simplest relation we 
 can suppose to subsist between the stimulatory and ex- 
 citatory processes ? Clearly that just as stimuli may be 
 reduced to terms of three independent variables, excitatory 
 processes are represented by three independent variables. 
 Thus, taking our previous example, any stimulus R' = a;.R 
 + 2/.G + zN, then A =f^{x, y,z),B =f^(x, y,z),C =f^{x, y, z), and 
 conversely, x = Y^[K B, C), y = ^lK B, C), z^Y^{K B, C). 
 
180 PHYSIOLOGY OF THE SPECIAL SENSES 
 
 These latter expressions may be taken as " elements " or unit 
 excitatory processes, or any linear function of them may 
 be so taken.^ 
 
 The conception is merely that three independent physio- 
 logical processes exist, each of which is defined by a func- 
 tional equation connecting it with the three independent- 
 stimulus values which, as we have seen, measure the effect 
 of any given stimulus. 
 
 This statement contains the fundamental part of the 
 hypothesis first sketched by Thomas Young and elaborated 
 by Helmholtz. It is important to distinguish the essentials 
 of the theory from its subsidiary parts. 
 
 The effect produced by any chromatic stimulus is sup- 
 posed to depend upon changes set up in three independent 
 " substances," nothing being postulated with respect to them 
 except that the magnitude of change in each is a function, 
 i.e. depends on the proportions of three independently variable 
 stimuli in terms of which the given stimulus can be ex- 
 pressed. Conversely, any given stimulus value is a function 
 of the independent activities of three visual "substances." 
 The nature of the " substances '' from a physiological point of 
 view, also the exact relations between them and the stimuli, are 
 left undiscussed. For convenience of illustration, Young and 
 Helmholtz assumed that the activity of each substance was 
 associated with a single colour sensation, and chose red, green, 
 and violet as " elementary sensations " from the present stand- 
 point, " Substance " A, when stimulated, was supposed to give 
 rise to the sensation of red ; B, under similar conditions, to 
 that of green ; and C, to that of violet or violet-blue ; in this 
 way the well-known "valency curves" of the text-books 
 were obtained. The advantage of this method is that the 
 theory seems more definite, but the disadvantage is entailed 
 that if the illustration proves irreconcilable with facts of 
 experiment, the reader omits to notice that what is found 
 wanting is merely an illustration, not the basal theory. It is 
 especially important to remember that the colour operations 
 discussed in Chapter XIV. do not pretend to describe any 
 
 ■ For a proof of this statement, see Helmholtz, Handh. d. Physiol, Opt., second 
 edition, pp. 342-3. The section beginning on p. 341 should be closely studied. 
 
YOUNG-HELMHOLTZ ON COLOUR VISION 181 
 
 direct relation between the hypothetical "substances" and 
 stimulus magnitudes. 
 
 In precisely the same way, for the sake of illustration, 
 Helmholtz suggested the existence somewhere in the retino- 
 cerebral apparatus of three sets of fibres, each corresponding 
 to one of the hypothetical "substances." This suggestion 
 was less happy because its utility was not so great as, and 
 the chance of misunderstanding greater than, in the pre- 
 vious case. The red, green, and violet fibres have been 
 persistently misunderstood; it is time to remember that 
 they are pure abstractions, no more essential to the theory 
 just sketched than is the employment in algebra of the 
 letter x to denote an unknown quantity an essential pro- 
 cedure in that science. To avoid this fertile source of 
 misunderstanding I shall use Professor v. Kries' term, 
 Components, and speak of the theory as the " Three Com- 
 ponents Hypothesis." 
 
 Another mistake is illustrated by the following quotation 
 from a paper by Miss Calkins : i — 
 
 "From the point of view of a psychological analysis of 
 our conscious sensations of colour, the postulates of this 
 theory are not in accordance with the facts of observation ; 
 for, even granting that violet is a simple fundamental colour 
 sensation (which many observers regard as complex), it 
 can hardly be denied that yellow is just as well characterised 
 and definite a sensation as red or green. Yellow looks 
 yellow, and does not seem at all like a mixture of red 
 and green, or indeed any other colour mixture." 
 
 Objections of this type are, I think, irrelevant. The 
 theory is only an attempt to express physiological processes 
 in terms of experimental facts, i.e. of stimulus values; it 
 has nothing to do with the psychological analysis of sensa- 
 tions. Whether such an analysis can be performed is a 
 question which must be decided by psychologists, it is no 
 part of our inquiry. We are only concerned with sensations 
 in so far as they are the signs of the existence of physio- 
 logical changes. 
 
 Having reduced the hypothesis to its simplest terms, let 
 
 ' Arcli. f. Pkysiologie, Supp. Vol., 1902, pp. 244 et seq. 
 
182 PHYSIOLOGY OF THE SPECIAL SENSES 
 
 us apply it to the facts resumed in earlier chapters. That 
 it describes normal visual systems is, of course, apparent, 
 since they form its starting point. Abnormal trichromatic 
 systems are satisfactorily described if we suppose one of 
 the visual components to be defined in a peculiar way. 
 Thus, if a component A is normally defined as f[(x, y, z), 
 in these cases it is some other function, /[(as, y, z), say, of 
 the stimulus values. 
 
 Turning to dichromatic systems, we have seen that, from 
 the experimental standpoint, they are reduction forms of 
 a normal system. Theoretically, the simplest reduction 
 we could imagine would be the absence or ineffectiveness 
 of one of the three normal components. Thus if we take 
 R, G, V (for the sake of clearness) as the normal components, 
 in the absence of R all sensations (sensation being used with 
 the meaning above defined) are functions of G and V only> 
 a condition approximcding to that of protanopes. If G 
 be absent and R present, we get a form resembling 
 deuteranopia. 
 
 Under such conditions, not only would the relation 
 between dichromatics and trichromatics be easily intelligible, 
 but we could determine from observations on dichromatics 
 the components of a normal system, or, more precisely, 
 the stimuli which act exclusively upon such components. A 
 stimulus inoperative upon a dichromatic eye must act 
 exclusively upon the missing component. We have learned 
 (Chapter XV.) how to find the position of such a stimulus in the 
 colour triangle (the " Fehlpunkt "), and the two such points 
 obtained for the two systems of dichromatic vision deter- 
 mine the stimulus relations of two visual components in 
 the normal eye. 
 
 All this depends, however, on the assumption that 
 one component is absent and everything else unchanged in a 
 dichromatic eye. Modern work renders it more than doubt- 
 ful whether we may make this assumption. In 1885, when 
 the second edition of his classical treatise was being pre- 
 pared, Helmholtz wrote as follows to Lord Rayleigh: 
 "I have never doubted that our colour system depended 
 on three variables, and no more. In regard to colour- 
 
YOUNG-HELMHOLTZ ON COLOUR VISION 183 
 
 blindness, the recent observations of Donders and of my 
 assistant, Dr. A. Koenig, show that this defect cannot be 
 referred simply to the lack of one of the fundamental colours, 
 but that two of the primaries (red and green) appear to 
 acquire a more even distribution in the spectrum, so that 
 now one and now the other makes a more vigorous im- 
 pression; in other words, the resulting curve approximates 
 now more to the red and now to the green sensation. In 
 addition to this we have every shade of lessened power 
 of discrimination. Consequently different individuals re- 
 quire very different mixtures of lithium and thallium light 
 in order to make up sodium light." ^ 
 
 But if we cannot regard dichromatic vision as differing 
 from the normal merely in the absence of a component, we 
 can, in terms of the fundamental hypothesis, assert that it de- 
 pends upon a visual system made up of two variables defined 
 by such expressions as A' = Fj (x, y, z) and B' = F2 {x, y, z). 
 This way of looking at the matter, although consistent and 
 logical enough, is not free from objection. We could not 
 deduce from such expressions any definite statements re- 
 specting the components of normal systems ; the description 
 is too general to admit of detailed verification. To put 
 the matter in a nutshell, the older conception of partial 
 colour-blindness as due to the absence of one normal com- 
 ponent is simple and, if true, practical, but is not, in fact, 
 quite adequate ; the more general statement is adequate, 
 but not very helpful. Next, what has the Three Components 
 Hypothesis to say with regard to the phenomena of after- 
 images ? 
 
 A noteworthy feature of after-image experiments is, of 
 course, that stimulation with a given light increases respon- 
 siveness to its complementary. It would appear, therefore, 
 easy to imagine that activity of the three components, or 
 any one or two of them, in a certain way diminishes their 
 responsiveness in one direction, increasing it in another 
 direction. This amounts to supposing that we have a 
 condition comparable with the state of the reflex arcs, so 
 brilliantly described by Sherrington; the nervous path is 
 - 1 Eolingsberger's Life of Hdmholtz, English trans., Oxford, 1907, p. 436. 
 
184 PHYSIOLOGY OF THE SPECIAL SENSES 
 
 occupied by one form of motor discharge, and this very 
 occupancy paves the way for a discharge different, and even 
 opposite, in kind. 
 
 To so highly general a statement as this no objection 
 will be found, but if we investigate details, difficulties arise. 
 Eor instance, the apparent saturation of spectral colours is 
 greatly enhanced by previously stimulating the eye with 
 their complementaries. Helmholtz accordingly supposed that 
 all the spectral colours act on each visual component. But 
 if this be true, the simpler interpretation of colour-blindness 
 once more fails. Observations of dichromatics suggest that 
 lights having wave lengths greater than 550 /^/^ do not 
 affect the third component (the " blue " or " violet ") at all, 
 because no standard blue had to be mixed with the standard 
 red in order to effect a good match with colours in this 
 part of the spectrum. To a normal eye, however, the satu- 
 ration of spectral yellow (589 /^/i) is unquestionably en- 
 hanced by previous exposure to blue. Either spectral colours 
 do not affect all three components, in which case the theory 
 does not cover after-image efifects, or the simpler explanation 
 of partial colour-blindness must be abandoned. In view of 
 what has already been said, the reader will perhaps agree 
 that the second alternative is the more plausible, and con- 
 clude that after-images are adequately described at the cost 
 of strengthening our suspicion that dichromatic and trichro- 
 matic systems cannot be co-ordinated in any simple manner. 
 So far we have found that the Hypothesis of Three 
 Components describes with sufficient clearness the facts of 
 normal colour vision, including the phenomena of after- 
 images; that it also describes the facts of partial colour- 
 blindness, but in very general terms, the earlier direct 
 explanation being insufficient. We have next to see whether 
 any experimental evidence can be found pointing to the 
 existence of independent visual components satisfying the 
 conditions of the theory or conceivably capable of so satis- 
 fying them. 
 
 Evidence of this kind has been afforded by the experi- 
 ments of G. J. Burch. This observer exposed his eye to 
 direct sunlight in the focus of a 2-inch lens behind coloured 
 
YOUNG-HELMHOLTZ ON COLOUR VISION 185 
 
 glasses. A gelatine film stained with magenta and com- 
 bined with a medium ruby glass was found to transmit a 
 fairly pure red, three thicknesses of green glass were used 
 for green, and a tank of cupric ammonia-sulphate for violet. 
 Similar arrangements were made for the other hues, and 
 in some experiments a large spectroscope was employed. 
 Two minutes' exposure was sufficient to produce the maximal 
 effect in the case of red. After exposure to red light, the 
 following effects were noticed. Scarlet geraniums appeared 
 black, calceolarias and sunflowers green, purple flowers, such 
 as clematis, violet. Pink roses were sky-blue. Fatiguing 
 with violet light caused objects reflecting violet light to 
 appear black, purples and reds seemed crimson. Green 
 stimulation made the foliage appear reddish or bluish-grey. 
 After these exposures, "the colour by which the eye has 
 been dazzled, and to which it is now blind, tint all those 
 objects which naturally reflect none of this." This state- 
 ment is illustrated by a simple experiment of Burch's. 
 Suppose the eye to be somewhat fatigued by green, as during 
 a long summer walk in the country, if the eye be directed 
 to a small red spot on a black surface, e.g. a geranium petal 
 on the black cover of a book, and one walks quickly with it 
 into a dark shed or barn, the colour of the petal changes 
 from red through orange and yellow, becoming eventually, 
 perhaps, whitish. On coming into the light again the red 
 reappears. 
 
 These interesting experiments suggest that stimulation 
 with red, green, and violet alters responsiveness with respect 
 to these stimuli alone, and the same is the case with blue. 
 Orange stimulation, on the other hand, affects not only the 
 appearance of the orange, but that of the red and the green 
 as well. Both positive and negative effects pass off rapidly 
 in the case of artificial red-blindness (in ten minutes), more 
 slowly after violet fatigue (in two hours). 
 
 It is, I think, obvious that the state of affairs presented 
 by these experiments is highly complex. We are dealing 
 with a change in responsiveness analogous to the retuning 
 effects of Chapter XVI., but of greater magnitude and in 
 less simple form. Take the experiment quoted as to the 
 
186 PHYSIOLOGY OF THE SPECIAL SENSES 
 
 apparent hue of a geranium petal after green stimulation ; 
 this is an ordinary after-image effect, and differs in no way 
 from the results obtained by other workers ; in the case of 
 exposure to stronger green light, the effect is similar. How 
 does this, we may ask, differ from the experiment with 
 intense orange light ? Simply in the fact that " retuning " 
 with orange affects responsiveness to red and green as well, 
 so that they, like orange itself, will cause the production 
 of negative after-images. The conclusion that in this case 
 the mechanism involved in the production of orange is com- 
 pounded of a mechanism yielding a sensation of redness 
 and a mechanism responding with a sensation of greenness 
 is reason3,ble, and finds confirmation in a recent experiment 
 performed by Burch. 
 
 We know that responsiveness to green is increased, re^ 
 latively to that for red stimulation, by resting the eye in 
 darkness; hence if orange or yellow depends upon a fusion 
 of two physiological processes, one concerned with green, 
 the other with red, then, under conditions of feeble illumina- 
 tion and dark adaptation, the yellow should appear greenish, 
 because the mechanism responding by a sensation of green^ 
 ness is more active under these conditions than that asso- 
 ciated in the same way with redness. One speaks of a 
 physiological process responding by the production of a sen- 
 sation purely for brevity ; it is not psychologically accurate, 
 but the psychological reader is not likely to be misled. 
 Burch found the result to be as expected — "the sodium 
 lines appeared pale green when of the minimum visible 
 intensity." 
 
 These results do, therefore, support a contention that 
 components in the sense of our theory may possibly have 
 a physiological counterpart. I do not think, however, that 
 the view that four components — a red, a green, a violet, and 
 a blue — exist is proved by Burch's experiments, valuable 
 though these are. To prove that any light acts upon only one 
 component it would be necessary to show that after dazzling 
 with, e.g., blue, any mixed light was altered by the subtrac- 
 tion of blue, and that any light not containing blue had 
 that colour added to it. The facts that, the condition is 
 
YOUNG-HELMHOLTZ ON COLOUE VISION 187 
 
 transitory is perhaps attended with some risk, and almost 
 certainly involves psychological complications, render exact 
 observations difficult. Under the circumstances it is per- 
 haps best to say that although the experiments are perfectly 
 consistent with a component hypothesis of the type dis- 
 cussed, it would be rash, on the strength of them, to make 
 any general statement as to the nature of the components 
 from the physiological standpoint. Theoretically, it does 
 not matter whether we adopt three or four components ; 
 the algebraical form of our theory would not be changed, 
 but we should lose the practical advantages of considering 
 normal colour vision to be, experimentally, trichromatic, 
 which would be a serious objection. 
 
 Before finally summarising the case presented, two matters 
 need attention. First, as to monochromatic vision or total 
 colour-blindness. It has been asserted that such a condition 
 cannot be described in terms of the Young-Helmholtz theory. 
 As a matter of fact, the assertion is inaccurate ; symbolically 
 we could cover the facts by supposing that the functions 
 defining the variables are identical, thus : A = /^(x, y, z) 
 = B = f^[x, y, z) = G = f^{x, y, z), or graphically we can put 
 it that, the three valency curves coincide. In any case, the 
 reader will have probably seen reason to think that mono- 
 chromatic vision depends upon a mechanism entirely dis- 
 tinct from the precursors of normal foveal vision, and that 
 its treatment should be kept separate from that of the 
 phenomena with which we are here concerned. 
 
 In the second place, no reference has been made to Simul- 
 taneous Contrast. The reason is, that it seems doubtful 
 whether the phenomena of simultaneous contrast are not of 
 quite a special kind. It is true that the original hypothesis 
 of Helmholtz, which assumed that contrastive effects are 
 dependent upon factors purely psychological in nature, can 
 hardly be maintained without some modification, but I shall 
 reserve a discussion of the point for a later chapter. It must 
 be said, however, that if subsequent work should compel us 
 to assign a purely physiological basis to the facts of simul- 
 taneous contrast, it will probably be necessary to modify the 
 theory of components in such a way that it will become some- 
 
188 PHYSIOLOGY OF THE SPECIAL SENSES 
 
 what more complicated than it is at present. Leaving this 
 matter for further consideration, we can say that the com- 
 ponent hypothesis associated with the names of Young and 
 Helmholtz supposes (1) that colour sensations depend upon 
 the activity of three independent physiological substances of 
 unknown nature and situation ; (2) that the relationship be- 
 tween these components and the complex of stimuli is ex- 
 pressible quantitatively by saying that the responsiveness 
 of each component is measured by a real linear function 
 of three standard stimuli; (3) the results of stimulating 
 these components are unit sensations in a purely physio- 
 logical sense, not units of consciousness; (4) no spectral 
 light acts upon only one of the components. 
 
 The theory describes with sufficient accuracy the main 
 facts, and there is some direct experimental evidence — that 
 of Burch — which is consistent with its truth. The main 
 objection to the hypothesis in its modern form is its highly 
 general nature and want of direct applicability to the im- 
 mediate data of physiological and physical research. How 
 far this is a real objection may be a matter of discussion, 
 but it at least inclines one to examine those theories which 
 are, in the colloquial phrase, less up in the clouds. Such 
 an examination will be the object of our next chapter. 
 
 RECOMMENDED FOR FURTHER STUDY 
 
 Helmholtz's treatise should, of course, be the first work consulted. 
 The reader should then turn to v. Kries' article in Nagel's Handb. The 
 following papers of Burch describe his experiments : — 
 
 (1) Phil. Trans., B., vol. cxci. pp. 1-34. 
 
 (2) Proc. Roy. Soc, vol. Ixvi. pp. 216-219. 
 
 (3) Proc. Roy. Soc, B., 1905, p. 214. 
 
CHAPTEE XIX 
 
 HERING'S THEORY OF VISUAL SENSATIONS 
 
 In the last chapter I attempted to trace out the theoretical 
 consequences developed by Young and Helmholtz from the 
 experimental facts of colour-mixing. It will have been clear 
 that the whole web of deductions depended from the fact 
 that, in general, the eifect of a given stimulus or combination 
 of stimuli was constant, so that we might attribute to the 
 stimulus a causal value. In other words, we have regarded 
 the sensations of colour as signs or differentice of processes 
 initiated by the stimuli ; the resulting theory was accordingly 
 not in any real sense a theory of visual sensations, but one of 
 visual stimuli. 
 
 If the estimate I formed of the value of this process 
 be in any measure just, it would seem to follow that its 
 weakness lay rather in what was left unsaid than in any 
 positive error. One cannot but notice in the theory as 
 at present advanced a certain dryness, a detachment from 
 one's experience of colour on the subjective side, which is 
 repellent. All this possibly amounts to saying that the theory 
 is not very simple nor very direct, objections which can be 
 urged against many theories of wider intellectual value than 
 that of Young and Helmholtz in other fields of thought, 
 objections which are not by any means fatal. Recognising 
 the existence of these difficulties, however, it is well to see 
 whether we can obtain a more satisfactory solution of our 
 problem by approaching it in a somewhat different way. 
 This chapter will be devoted to the study of another pro- 
 posed solution of the matter, that of Ewald Hering. 
 
 This other line of investigation practically resumes the pro- 
 blem at the point at which the Greeks left it, and dates from 
 the publication in 1810 of Goethe's Farbenlehre. The com- 
 paratively slight direct influence of this work on the develop- 
 
190 PHYSIOLOGY OF THE SPECIAL SENSES 
 
 ment of modern physiological thought respecting the nature 
 of physiological processes is due to causes well worthy of 
 attention. The physical analysis of white light into mono- 
 chromatic constituents by Newton had naturally attracted 
 the chief, almost the exclusive, attention of those who occupied 
 themselves with the study of colour vision. Goethe, however, 
 with characteristic intellectual insight saw that the difficulties 
 of the problem were not to be overcome by vague references 
 to physical experiments. He saw that the problem was one 
 of sensations, and he approached it from the sensational 
 standpoint. Had he contented himself with this, with an 
 analysis of sensations of colour, his work must have had an 
 enormous influence; but he went further. The prevailing 
 tendency to over-estimate the significance of the physical 
 side of the question led Goethe into the opposite error. He 
 sustained the thesis that the Newtonian analysis was physically 
 incorrect, and that the alleged decomposition of white light was 
 not in general possible. Consequently, much of his work is 
 devoted to an attack upon the Newtonian doctrine from the 
 physical side, an attack which signally failed. Unfortunately 
 the failure of this attack involved more valuable parts of 
 Goethe's work in discredit, and his book is not well known 
 even to professional physiologists. I can best give an idea 
 of the valuable parts by quoting a few passages which are 
 specially relevant for us. 
 
 "With regard to the German terminology, it has the 
 advantage of possessing four monosyllabic names no longer 
 to be traced to their origin, viz. yellow (Gelb), blue, red, 
 green. They represent the most general idea of colour to 
 the imagination, without reference to any very specific modi- 
 fication. If we were to add two other qualifying terms to 
 each of these four, as thus, red-yellow and yellow-red, red- 
 blue and blue-red, yellow-green and green-yellow, blue- 
 green and green-blue, we should express the gradations 
 of the chromatic circle with sufficient distinctness; and 
 if we were to add' the designations of light and dark, and 
 again define, in some measure, the degree of purity or its 
 opposite by the monosyllables black, white, grey, brown, we 
 should have a tolerably sufficient range of expressions to 
 
HEEING'S THEORY OF VISUAL SENSATIONS 191 
 
 describe the ordinary appearances presented to us, without 
 troubling ourselves whether they were produced dynamically 
 or atomically." ^ 
 
 " Considered in a general point of view, colour is deter- 
 mined towards one of two sides. It thus presents a contrast 
 which we call a polarity, and which we may fitly designate 
 by the expressions plus and minus. 
 
 Plus. 
 
 Minus. 
 
 Yellow. 
 
 Blue. 
 
 Action. 
 
 Negation. 
 
 Light. 
 
 Shadow. 
 
 Brightness. 
 
 Darkness. 
 
 Force. 
 
 Weakness. 
 
 Warmth. 
 
 Coldness. 
 
 Proximity. 
 
 Distance, 
 
 Repulsion. 
 
 Attraction. 
 
 Affinity with acids. 
 
 Affinity with alkalis." ^ 
 
 We see here formulated the conception of certain colour 
 sensations as occupying unique places in our sensational 
 field, and presenting also, as it were, a species of sensational 
 contrast one with another. 
 
 Of course in many respects this idea is an old one, and 
 has perhaps always been realised by painters. In a dialogue 
 on colours by Ludovico Dolce, published in 1565, the fol- 
 lowing passage occurs : " He who wishes to combine colours 
 that are agreeable to the eye will put grey next dusky 
 orange, yellow-green next rose colour, blue next orange, dark 
 purple, black, next dark green, white next black, and white 
 next flesh colour." ^ 
 
 Titian, according to his biographer Ridolfi, Avas fond of 
 opposing red and blue to his flesh tints, and Rubens con- 
 trasted a bright red with his "still cooler flesh colour" 
 (Eastlake). 
 
 Here, as elsewhere, the greatest art is to conceal the art. 
 A study of the works of Rembrandt, whose skill in con- 
 trastive effects has never been equalled, reveals the fact that 
 
 1 Goethe's Theory of Colours, Eastlake's translation, London, 1840, pp. 
 243-4. 
 
 ' Goethe, op. oil., p. 276. ' Ibid., Note C. 
 
192 PHYSIOLOGY OF THE SPECIAL SENSES 
 
 the most striking effects are produced by light-dark contrasts 
 rather than specific colour oppositions. The remarkable 
 specific brightness {vide infra) of yellow is, however, well 
 seen in the (so-called) " Nachtwache " ; the extraordinarily 
 vivid effect produced by the yellow dress of one of the central 
 figures is balanced by general shadow, without, so far as I 
 could judge, any obvious use of the " Gegenfarbe." These 
 remarks seem to me to apply also to the " Staalmeesters." I 
 hope that some day a physiologist with a competent know- 
 ledge of art will undertake a study of the master works of 
 art from the standpoint of the physiology of vision. 
 
 The point to mark in the preceding passages is the 
 general agreement that certain of our sensations of colour 
 are really singled out from the whole group as presenting 
 sharply defined, special characters. All sensational theories 
 are primarily concerned with the definition of these characters, 
 and secondarily with an attempt to describe the data in 
 terms of a physiological hypothesis. Of such attempts the 
 views developed by Professor Ewald Hering of Leipzig and 
 his pupils during the last five-and-thirty years are the most 
 valuable results. Whatever may be our ultimate conclusion 
 as to the validity of these theories, no one can doubt that 
 they have greatly advanced our knowledge of visual physi- 
 ology, and their study cannot be neglected by any one desirous 
 of acquiring even a superficial idea of modern conceptions. 
 Whenever I come upon a new writer dealing with colour 
 vision, I turn to his account of Hering's theories ; if he dis- 
 misses them as " obviously " incorrect or absurd, I am con- 
 fident that his own contribution to the subject is a very 
 small one. 
 
 According to Hering's method of analysis, our whole 
 visual world can be resolved into six elementary qualities of 
 sensation ; white, black, the toneless, and blue, yellow, green, 
 red, the toned or bright (hunte) colours. If one considers the 
 tone-free qualities, they can be arranged to form a series of 
 shades or gradations passing from the intensest white to the 
 deepest black. If one attends to the toned colours, they can 
 be arranged in a circle with four divisions. 
 
 " If we choose in such a colour circle any colour as starting 
 
HEEING'S THEORY OF VISUAL SENSATIONS 193 
 
 point — for instance, a red similar to that with which a spectrum 
 usually begins at the long- waved end — we see the red colours 
 arranged in one direction gradually becoming more yellowish, 
 while the redness of the colours correspondingly, diminishes, 
 until finally, passing through orange and golden-yellow, we 
 arrive at a yellow which contains no trace of the red which 
 is still so apparent in the orange. To this yellow succeed 
 other yellow colours which play more and more into the 
 green (sulphur-yellow, canary-yellow) ; further on (as in sap- 
 green) the yellowishness recedes more and more behind the 
 steadily increasing greenishness, until we finally reach a 
 green which seems to be entirely free from yellow. To this 
 succeed green colours which already play into blue (water- 
 green); further on the bluishness of the colours becomes 
 increasingly stronger, the greenishness weaker, until we 
 finall}' reach a blue exhibiting no more greenishness at all. 
 To this blue succeed blue colours of increasing reddishness 
 and correspondingly diminishing bluishness (blue-violet, red- 
 violet, purple-red), until the last trace of bluishness vanishes 
 in a definite red." ^ 
 
 If we define a pure green as a sensation free from ad- 
 mixture with that of blue and yellow, and the other three 
 sensation qualities in the same manner, we see that our 
 pure colours (from this point of view) can be arranged in 
 two pairs, yellow and blue forming one, and red and green 
 the other. The members of each pair can be placed oppo- 
 site one another in a diagram, because we can only pass 
 from yellow to blue or from red to green by traversing the 
 province of a member of the other pair, as just explained. 
 There is no pure yellowish-blue or reddish-green sensation 
 quality. 
 
 But there is yet another contrast, from the standpoint of 
 sensation quality, between yellow and red on the one hand, 
 and blue and green on the other. Somehow, in a manner diffi- 
 cult to express in words, yet of universal experience, the two 
 former colours are associated with a certain heightening and 
 increased vividness of sensation tone. This finds its ex- 
 
 ^ Ewald Hering, Griindziige der Lehre vom Zichtsinn, Leipzig', 19C5, 
 p. 41. 
 
194 PHYSIOLOGY OF THE SPECIAL SENSES 
 
 pression in the classification by artists of colour into warm 
 and cold. Goethe has emphasised these points : — 
 
 "We find from experience again that yellow excites a 
 warm and agreeable impression. Hence in painting it be- 
 longs to the illumined and emphatic side. 
 
 "This impression of warmth may be experienced in a 
 Yery lively and emphatic manner if we look at a landscape 
 through a yellow glass, particularly on a grey winter's day. 
 The eye is gladdened, the heart expanded and cheered, a 
 glow seems at once to breathe towards us." ^ 
 
 Of blue he says : " This colour has a peculiar and almost 
 indescribable effect upon the eye. As a hue it is powerful, 
 but it is on the negative side, and in its highest purity is, 
 as it were, a stimulating negation. Its appearance, then, is 
 a kind of contradiction between excitement and repose. 
 
 " Rooms which are hung with pure blue appear in some 
 degree larger, but at the same time empty and cold. 
 
 "The appearance of objects seen through a blue glass is 
 gloomy and melancholy." ^ 
 
 In the opinion of Hering, the facts are most satisfactorily 
 described by saying that in the sensation-complex blue and 
 green produce a darkening and yellow or red a brightening 
 effect ; the toneless colours, black and white, also contribute 
 respectively in a negative or positive sense to the sum-total 
 of effects. We have, therefore, the brightness of a colour 
 defined in strictly sensational terms. 
 
 We have now reached the conception of six primary 
 sensation qualities arranged in three pairs — white-refl, red- 
 green, yellow-blue. The first member in each case increases, 
 the second diminishes the subjective intensity or brightness 
 of a sensation complex of which it forms part.^ 
 
 "The brightness or darkness of a toned (bunte) colour 
 is, according to this view, the result of the inherent bright- 
 ness or darkness (Eigenhell vmd Eigendunlcel) of its consti- 
 tuent pure colours, which as the pure constituents of that 
 colour agreeably to their respective distinctness determine 
 
 1 Goethe, op. cit., p. 307. ^ Ibid., p. 310. 
 
 ^ Seusation-eomplex is a term used merely to indicate the supposed multi- 
 plicity of infra-conscious representatives, not in reference to perception. 
 
HERING'S THEORY OF VISUAL SENSATIONS 195 
 
 the quality of the colour. In any colour really existent for 
 us is a definite inherent degree of brightness and darkness, 
 and in accordance with whether the brightness or the dark- 
 ness be the more distinct, we call the colour bright or 
 dark. 
 
 " A toned colour may generally be regarded as made up 
 of four primary components, two toned and two tone-free 
 •(black and white). Only in colours of the tone of a 
 pure colour is one toned constituent present by itself. In 
 any red-yellow colour, e.g. orange, we have accordingly to dis- 
 tinguish three bright, pure components (red, yellow, white), 
 and one dark (black); but in any green-blue, three dark 
 {green, blue, black) and one bright. The red-green and green- 
 yellow colours would contain, however, two bright and two 
 dark pure components. 
 
 "From what has been said, the following rules can be 
 deduced : — 
 
 " If two colours of equal tone and equal purity differ in 
 brightness, this is due to a difference in their black-white 
 components. 
 
 " Two colours differing in tone may, notwithstanding equal 
 degrees of purity and equality as regards their black-white 
 components, differ in brightness. 
 
 " With equality of conditions as to the black- white com- 
 ponents, a yellow, a red, or a yellow-red colour is so much 
 the brighter, a blue, a green, or a blue green so much the 
 darker the more distinct the colour tone in comparison with 
 the black- white constituent." ^ 
 
 This quotation describes what is often called the theory 
 of the " Specific Brightness of Colours." The whole is, or 
 claims to be, a faithful analysis of our visual sensations with- 
 out reference to any hypothesis whatever. That this is a 
 perfectly legitimate process I have already attempted to 
 show ; the next step is to translate these facts, or supposed 
 facts, into terms of a physiological hypothesis. Such a 
 translation can be readily effected. 
 
 It is supposed that somewhere in the retino-cerebral 
 a,pparatus, in the infra-conscious sphere, four distinct sub- 
 ^ Hering, op. cit., p. 61. 
 
196 PHYSIOLOGY OF THE SPECIAL SENSES 
 
 stances exist. Each of these substances can undergo a 
 building up, or anabolic, and a breaking down, or katabolic, 
 change. External stimuli will, depending on their natures, 
 induce either an anabolic or a katabolic change in these 
 substances, and are associated with definite sensations of 
 colour. The building up of the black-white substance corre- 
 sponds to a sensation of blackness, its breaking down to a 
 sensation of whiteness ; anabolism of the red-green sub- 
 stance is associated with green colouration, its katabolism 
 with red colouration ; similarly in the third substance yellow 
 is katabolic in origin, blue anabolic. 
 
 Before discussing these views in detail, I must warn 
 the reader against some popular misconceptions. Certain 
 opponents have asserted or suggested that the facts upon 
 which Hering founded his theory differ in some perverse 
 way from those data which are ordinarily called facts of 
 experiment. This is not the case. The facts the hypothesis 
 attempts to describe are as legitimately objects of inquiry 
 as any others within the purview of physiological science. 
 It is further to be noted that the four physiological " sub- 
 stances" have just as much and just as little real existence 
 as the three components of our other theory. It is idle to 
 say that the postulated anabolic and katabolic processes are 
 essentially unlike any chemical mechanisms with which we 
 are acquainted. It is equally vain to object that stimulation 
 processes in animate nature are to all appearance bound up 
 with katabolic changes; this would only be a valid objection 
 if we attempted to identify the hypothetical substances with 
 any known retino-cerebral constituent. No such identifi- 
 cation is attempted; the suggestion that Hering and his- 
 school identify the black- white substance with visual purple 
 is entirely unjustified. The fact is, that this theory can only 
 be judged on the grounds of scientific expediency. Does. 
 Hering's method give us a better account of the phenomena 
 of colour vision than that based upon stimulus relations ? 
 This is the only question worth the physiologist's while to 
 answer. 
 
 One notices directly that the hypothesis, in the form in. 
 which it has just been presented, offers two considerable^ 
 
BERING'S THEORY OF VISUAL SENSATIONS 197 
 
 advantages. Firstly, it deals with the immediate data of 
 vision, the sensations of colour, directly, not merely in terms of 
 stimulation magnitudes. In the second place, it is essenti- 
 ally easy to comprehend, which cannot, perhaps, be said of 
 the components' hypothesis. Let us first of all attempt 
 to express the facts of successive contrast, after-images, in 
 terms of this hypothesis. After resting the eye on a white 
 object we should expect, under certain conditions, a positive, 
 under others a negative after-effect, and experiment agrees 
 with theory. For example, if the eye be stimulated with 
 green light, anabolism will occur in the red-green substance, 
 an anabolism which will lead to the formation of a large 
 quantity of "material." If, now, red light stimulates the 
 retina it produces not only katabolism of the normal 
 quantity of substance, but the new formed material also falls 
 to pieces, and the correlative sensation is greatly enhanced. 
 That a positive after-image will sometimes be produced 
 might be expected. After the stimulus is withdrawn, the 
 anabolic (or katabolic) change induced by it will continue 
 for some short space of time owing to a species of chemical 
 inertia in the substance. In fact, all the obvious phenomena 
 of after-images are well enough described in terms of Bering's 
 theory ; it is when we come to the details that trouble arises. 
 It was noticed by v. Kries that the responsiveness of the 
 eye to monochromatic light was markedly altered by previous 
 exposure to white. He found that the stimulus value of the 
 red he employed was diminished in the ratio of about one 
 to four by previously retuning the eye with white. The 
 validity of v. Kries' conclusions has been contested by 
 Bering, who objected that the colour used by v. Kries as 
 a reactor (180° blue, 180° white on a disc) was not sufficiently 
 saturated, and brings forward the following experiment : Two 
 discs, A and B, are arranged. A consists of a black centre 
 surrounded by a white ring, B of a centre composed of 120° 
 blue and 240° black, encircled by a ring containing 356° blue 
 and 4° white. A point upon the internal margin of the 
 white ring in A is fixated for a given time, and the ex- 
 perimenter then turns his eye to an exactly corresponding 
 point in B. Bering always found that the outer ring of 
 
198 PHYSIOLOGY OF THE SPECIAL SENSES 
 
 B, under these conditions, looked more saturated than the 
 centre. 
 
 As the term saturated is not used by Hering in the 
 physical sense, the exact bearing of his objection to v. Kries' 
 experiment is not clear. Presumably he holds tha,t the 
 " blueness " of v. Kries' disc was not distinctly separable from. 
 its " whiteness," with the result that the apparent change in 
 the latter on retuning with white was mistaken for a change 
 in the former. If this objection is admissible, it is to be 
 observed that in Bering's experiment the difference between 
 the amounts of blue in reacting and comparison fields was so 
 great that an enormous diminution in responsiveness over 
 the retuned area would be necessary for the production of 
 a good match between the centre and the periphery of B. 
 It is contended that after white retuning a reacting blue and 
 white can be made to match a pure blue in brightness by 
 adding more white ; adding more blue will always make the 
 reactor too saturated. Hering's experiment actually demon- 
 strates that a diminution of some 66| per cent, in apparent 
 intensity cannot be attained by retuning with the white he 
 employed. That the apparent brightness was correspondingly 
 reduced is, of course, evidence in his favour, but the difiiculties 
 of specific brightness comparison are under such conditions 
 not small. 
 
 The conclusion of Hering's memoir is so important from 
 the theoretical standpoint, that I quote it verbatim. 
 
 " To the change of state experienced by an element of the 
 somatic visual field when acted upon by, e.g., blue light, with 
 which the blue sensation is associated, the whole somatic 
 visual field reacts by a change in the opposite direction 
 which corresponds to the oppositely coloured or yellow 
 sensation, and any light that now falls on the retina acts, 
 in consequence of this chromatic retuning of the visual field, 
 as if its yellow valency were increased and its blue valency 
 correspondingly diminished. This retuning is maximal in 
 the immediate vicinity of the element acted upon by the 
 blue light, and diminishes with its distance from the same. 
 ... A white light falling on the neighbourhood of the region 
 which has been stimulated with blue seems therefore more 
 
HERING'S THEORY OF VISUAL SENSATIONS 199 
 
 or less yellowish, but a white light which. falls together with 
 blue on the spot that has been stimulated with blue, see- 
 ing that it behaves as a more or less yellow-valent light, 
 neutralises the blue valency of the blue light so much the 
 more the greater be the quantity of it mixed with the latter. 
 This explains the striking fact that the chromatic quality of 
 a saturated colour is so extremely quickly extinguished by 
 increasing the amount of white mixed with it. . . . 
 
 " When V. Kries, therefore, supposed that, according to the 
 theory of opposite colours (Gegenfarben), the same result 
 would be obtained from a fatigued and an unfatigued area if 
 the same quantity of blue were allowed to fall on both, but 
 in addition on the fatigued area a suitably chosen quantity 
 of white light, he was in .error. In such a case the blue 
 valency of the blue light at the unfatigued area is unaltered, 
 since no other light is mixed with it ; over the fatigued area 
 the blue valency of the blue light is partly neutralised by the 
 admixture of white. Accordingly the blue at this latter area 
 must appear less saturated than at the former. In fact, a 
 transitory equality in brightness and saturation between the 
 two areas is only obtained when white is indeed mixed with 
 blue for the fatigued area, but, on the other hand, the blue 
 light for the unfatigued area is suitably diminished. In 
 general, for reasons already given, an equality in colour 
 tone for blue can only be obtained under exceptional cir- 
 cumstances when the character of the daylight is specially 
 favourable, and the tone of the blue just right." ^ 
 
 It is clear, therefore, that while white tuning is held by 
 Hering not to affect colour valency, chromatic retuning does 
 markedly affect white valency. In other words, what is, 
 sensationally, a pure white has a definite colour valency ; 
 pure Avhite, in terms of the physiological process in the 
 white-black " substance," is hypothetical. We see again in 
 Hering's theory the weak point we noticed in the hypothesis 
 of Young and Helmholtz, viz. the necessity of complicating 
 the theory in order to make it cover the facts, a necessity 
 which brings the theory out of touch with the immediate 
 
 1 Hering, "Ueber die von der Farbenempfindlichkeit unabhangige Aen- 
 derung der Weissempfindlichkeit," Pflilg. Arch., xolv. (1903), 533. 
 
200 PHYSIOLOGY OF THE SPECIAL SENSES 
 
 data of experience. It is rather interesting to notice tliat 
 Hering's theory becomes more difiicult to follow by develop- 
 ing in a direction opposite to that followed by the other 
 theory ; the latter became unsatisfactory, to some, by tending 
 to be too general, the former by multiplying its detailed 
 sub-hypothesis. 
 
 I cannot too strongly urge on the reader the considera- 
 tion that such difficulties will always arise; they are bound 
 up with the progress of knowledge. An explanation which was 
 satisfactory at the date of its publication ceases to be so as 
 the volume of scientific output in the subject of which it treats 
 widens. In this sense all theories are creatures of a day. 
 
 Hering's account of partial colour-blindness is subject 
 to similar, but perhaps stronger, objections. He originally 
 taught that in partial colour-blindness the red-green sub- 
 stance was absent. In view of the fact that there are two 
 forms of partial colour-blindness further explanation was 
 necessary, and a most ingenious conception, a conception 
 which in all probability contains at least some truth, was 
 evolved. We have seen that trichromatic systems exist 
 which differ appreciably from the normal in responsiveness 
 to yellow and blue light. We have ako seen that pigmenta- 
 tion of the retina and the lens influence appreciably colour 
 matches. Hering suggested, and supported his suggestion 
 by some observations on subjects of the two forms of 
 trichromatic anomaly, that the two forms of partial colour- 
 blindness were, in a sense, extreme forms of yellow and blue 
 sightedness combined with extremely marked or extremely 
 slight pigmentation. The protanopes would centre round the 
 blue anomaly (relatively blue sighted), and the deuteranopes 
 round the yellow anomaly (relatively yellow sighted). That 
 the conception of continuous variation in colour sense is 
 antecedently probable must be admitted; there is much 
 reason to think that the occurrence of discontinuous varia- 
 tion in biological characters is far less common than certain 
 enthusiasts would have us believe. It must, however, bo 
 said that the observations of v. Kries and others, which 
 were alluded to in an earlier chapter, have made it difficult 
 for us to believe that the difference between protanopia 
 
BERING'S THEORY OF VISUAL SENSATIONS 201 
 
 and deuteranopia is not more definite than this explana- 
 tion would suggest. Tschermak/ a prominent supporter of 
 Bering's views, seems more or less definitely to abandon 
 this attempt to describe dichromatic vision, and, so far as 
 I know, an adequate expression of the facts in terms of 
 Hering's hypothesis has not yet been found. 
 
 Our general conclusions, therefore, may perhaps be ex- 
 pressed in the following way : — 
 
 The theory associated with the name of Hering is an 
 attempt to arrive at a general conception of the visual 
 processes by an analysis and comparison of sensations of 
 colour. In this way prominence is given to many important 
 facts neither specially nor adequately resumed in the theory 
 founded on stimulus relations. If, however, we attempt to 
 build up upon such data an hypothesis adequate to the 
 task of describing all the experimental facts, difficulties are 
 encountered not less formidable than those associated with 
 the Young-Helmholtz theory. 
 
 In attempting to meet their respective difficulties the 
 two theories became unsatisfactory in different ways. The 
 stimulus hypothesis becomes too general, the sensational 
 hypothesis too detailed. Examples have been seen in the 
 former's treatment of dichromatic vision, and in the latter's 
 account of after-images. 
 
 The most tempting reconciliation of the modes of analysis 
 is to suppose that the theories are in some sense comple- 
 mentary — a view akin to that of Donders ^ — that they both 
 contain some measure of truth, surveying the vast complex 
 of phenomena from different points of views. Neither 
 theory is wholly true nor yet wholly false, nor does 
 adhesion to the one imply total rejection of its ostensible 
 rival. 
 
 Of the numerous other theories of colour vision I do 
 not, for reasons already given, intend to speak ; many, especi- 
 ally perhaps those of Schenck and Ladd-Franklin,^ are highly 
 
 ^ Tschermak, Ergebn. d. Physiologic, 1st Jahrg. (1902), second part, p. 795. 
 
 * Donders, Archif. f. Opthalm., 1881, xxvli. part 1, p. 55; ibid., 1884, xxx. 
 part 1, p. 15. 
 
 ' Schenck, Pflilg. Arch., cxviii. (1907), p. 161. Ladd-Franklin, Z.P.P.S.O., 
 iv. (1893), 211. 
 
202 PHYSIOLOGY OF THE SPECIAL SENSES 
 
 ingenious ; all contain some points of importance, but none 
 can be compared in interest and educational value with those 
 we have examined in the last two chapters. 
 
 RECOMMENDED FOR FURTHER STUDY 
 
 The best account in English, and written with judicial impartiality, 
 is that of Dr. Rivers, Schafer's Text-book of Physiology, Edinburgh, 
 1900, vol. ii. p. 1026. 
 
 In Support of The Yodng-Helmholtz Theory — 
 Helmholtz, Handb. d. Physiol. Optik., second edition. 
 V. Kries, Nagel's Handb., vol. iii. p. 109. 
 
 In SnppoET of Hering's Theory — 
 Hering, Griindziige der Lehre vom Lichtsinn (Sonderabdruck a. 
 d. Handbuch d. Augenheilkunde, 1 Teil, xii. Kap.). Leipzig, 
 1905-7, Engelmann. (This is an admirably cleax account so 
 far as it has gone, but the separate publication appears to be 
 delayed. ) 
 
CHAPTER XX 
 
 SIMULTANEOUS CONTRAST 
 
 It may be regarded as a very general truth of sense physiology 
 that the sum of the effects produced by two simultaneous 
 stimuli is not equal to the total effect produced by two 
 stimuli applied successively. The mutual effect exerted 
 in this way by two synchronous excitations is revealed to 
 us subjectively as contrast. Although this phenomenon 
 of contrast is observed in each department of sensation, 
 its study is most conveniently effected in the case of vision, 
 and a rich harvest of results has been garnered by workers in 
 this field. For this reason I shall confine myself to a study 
 of visual contrast ; the generalisation of the results, together 
 with the modifications necessary in special cases, is an 
 exercise which may profitably be left to the reader. 
 
 Although it is perfectly true that the fundamental ex- 
 periments in simultaneous contrast are readily performed, 
 the subject in its detailed examination is not free from 
 difficulty, and a satisfactory theory has not yet been pro- 
 pounded. In summarising the experimental methods and 
 results I shall follow the order adopted by Tschermak,i 
 whose memoir is both lucid and exhaustive. The theoretical 
 considerations which I shall finally submit, although not 
 in any true sense novel, will be found to differ slightly from 
 the usual interpretations. The first class of experiments 
 comprises Surface or Areal Contrast, in which the alteration 
 is appreciable over a comparatively large surface. These ex- 
 periments can be subdivided again into brightness or black- 
 white contrasts and colour contrasts. 
 
 The simplest instance in the first group is the varying 
 brightness of a scrap of grey paper in accordance with the 
 whiteness or blackness of the background against which 
 it is viewed. Bering has demonstrated this in his elegant 
 " Doppel-Zimmer " experiments. A circular aperture in an 
 
 ' A. V. Tsohermak, " Ueber Kontrast und Irradiation," Ergebnisse d. Fhysiol., 
 1903, second part, pp. 726-798, with exhaustive bibliography. 
 
 203 
 
204 PHYSIOLOGY OF THE SPECIAL SENSES 
 
 opaque partition is kept constantly illuminated ; its apparent 
 brightness is, however, seen to vary with changes in its 
 environment. Generally the two contrasting surfaces are 
 concentric, but this is merely for convenience of experiment, 
 as Aubert pointed out. 
 
 Areal colour contrast has long been known, the simple 
 fact that a colour tends to produce an apparent com- 
 plementary tinging of a neighbouring field having been 
 pointed out by Leonardo da Vinci. The ordinary laboratory 
 method of demonstration is known as Meyer's experiment, 
 although the principle is re?Jly due to Johannes Mtiller. A 
 sheet of green (for example) paper is spread out and a scrap 
 of grey paper placed in the middle, the whole being covered 
 with tissue paper. The grey scrap appears of a distinct rose 
 hue. Another method depends on the colouring of an 
 objectively colourless mirrored light or shadow which is 
 caused by a coloured background. Goethe noticed that the 
 image of window bars reflected from the upper surface of 
 a piece of green glass was purple. 
 
 The best form of the experiment is due to Ragona Scina, 
 and has been slightly modified by Hering. The observer 
 looks through a sheet of coloured glass inclined at an angle 
 of 45° to a horizontal sheet of paper with black and white 
 figures on it — for instance, rings of about 1 cm. in breadth ; 
 a second sheet of white paper is placed vertically opposite 
 the inclined plate.^ The black rings on the horizontal 
 sheet, upon which colourless light is mirrored by the vertical 
 sheet, are brightly tinged by contrast. The white rings 
 appear in the saturated and the ground in the unsaturated 
 colour qi the glass. Cobalt glass gives, I think, the best 
 effect, but ordinary red glass does very well. 
 
 Probably the most striking effect of all is that of coloured 
 shadows, the existence of which was recognised by Leonardo 
 da Vinci and Otto von Guericke. For demonstration pur- 
 poses no special apparatus is necessary, and I find the foUow- 
 
 ^ This sheet, which should be at right angles to that seen through the 
 glass, ought to have black and white rings on it alternating with those on 
 the other sheet, so that the colour of the white rings seen through the glass 
 is not weakened by mirroring of white light from corresponding rings on 
 the vertical sheet. 
 
SIMULTANEOUS CONTEAST 205 
 
 ing arrangement quite good. Set up vertically on a table 
 a large white screen, and lean against it a lecturer's pointer. 
 Darken the room except for one window slit, which gives 
 a sharp shadow on the screen. On the side of the table 
 away from the window stand an ordinary electric table lamp, 
 so that a second shadow is formed on the screen. The 
 window shadow is diffusely illuminated by the yellow electric 
 light, and looks yellow ; the lamp shadow, which is diffusely 
 illuminated by the window light and should appear grey, 
 is strongly blue. The contrastive nature of the appearance 
 is demonstrated by switching off the electric light, when 
 the shadow at once loses its colour. The most exact 
 application of the method is due to Hering, and is, in prin- 
 ciple, as follows: Two parallel slits are made in a vertical 
 shutter; one is covered with ground glass, the other with 
 coloured glass, the double shadow being focussed on a white 
 screen. Another good method is that of Helmholtz. A 
 small black disc is brought upon the dividing line of a 
 half- white, half-green field and looked at through a fragment 
 of Iceland spar. The extraordinary image of polarisation of 
 the green falling on the ordinary white image forms an 
 unsaturated green middle stripe. Internal to this the 
 ordinary image of the right side of the black disc loses its 
 whiteness and appears saturated green; the extraordinary 
 image of the left half of the disc also appears internal, but is 
 illuminated with white, and appears in the contrast colour. 
 
 Another form of contrast observations relates to the 
 changes observed at the margin of a field. For instance, 
 a grey disc on a black ground appears brighter at the 
 periphery than at the centre, and a white trellis work on 
 a dark ground seems darker at the points of crossing than 
 at other parts. This emphasis, as it were, of the margin is 
 responsible for some of the beauties of mountain scenery, the 
 graduation of grey values in mountain peaks with intensi- 
 fication of the contour lines. 
 
 We must now consider some details of the contrastive 
 action. 
 
 (1) The Spatial Extent of the Effect. 
 
 The intensity of the contrast diminishes rapidly as we pass 
 
206 PHYSIOLOGY OF THE SPECIAL SENSES 
 
 out from the excited area, and a very slight separation of 
 the contrasting fields makes a considerable difference in the 
 result. A small black line separating two fields markedly 
 diminishes the contrast, a fact to which Helmholtz attributes 
 much importance. A somewhat important observation is 
 that of Charpentier, that the liminal stimulus value for a 
 given retinal area is the same whether light be completely 
 excluded from that area or contrast blackness induced by 
 stimulating a neighbouring point. If true, this opinion tells 
 strongly against the theory of contrast advanced by Hering 
 (vide infra), but many observations are difficult to reconcile 
 with it, especially those of Brewster, Meyer, and Aubert. 
 (2) The temporal Succession of Contrast Effects. 
 The optimum contrast efi'ect is attained after a measur- 
 able but very short period of time (Exner), and it then falls 
 off rapidly. Hering showed " that any true contrastive effect 
 is most marked at the beginning of the observation (apart 
 from the very short period of development), and diminishes 
 rapidly, only remaining distinct for a very short interval." 
 He attributes contradictory results to the occurrence of after- 
 images associated with movements of the eye. With long 
 continued fixation the contrast colour fades, giving place 
 to its complementary (Hering's Simultaneous Coloured or 
 Colourless Induction). 
 
 By employing a colourless disc in a coloured field, when 
 the contrast has faded, the complementary colour gradually 
 appears, increasing in saturation until finally disc and field 
 fuse together. Accordingly during fixation of the object it- 
 self the negative after-image may result, a sudden diminu- 
 tion in the amount of light being a specially favouring 
 circumstance. It is, therefore, a difficult matter to observe 
 accurately the time relations of simultaneous contrast with 
 the exclusion of after-images. Absolutely steady fixation, 
 utilisation of the first moment of appearance of the induced 
 colour, and short duration of the whole experiment are 
 essential conditions. 
 
 The relations between simultaneous and successive con- 
 trast were exhaustively studied in Hering's laboratory by 
 Kuhnt. He used colourless discs on a moderately illuminated 
 
SIMULTANEOUS CONTRAST 207 
 
 background of coloured glass or paper. With red and green 
 the simultaneous contrastive effect endures for some seconds, 
 with yellow or blue the effect is but momentary. In the end, 
 in all cases, disc and background become indistinguishable. 
 
 It is generally held that contrastive effects are better 
 marked in indirect than in foveal vision, which naturally 
 suggests that light-dark adaptation would be of importance 
 in this connection. Some experiments support this con- 
 clusion, but no systematic inquiries have, I think, been 
 published. 
 
 Quantitative Laws of Simultaneous Contrast 
 
 (1) White-Black Contrast. 
 
 It has been generally held that the subjective bright- 
 ness of a measurably illuminated surface varies when seen 
 against a background of varying illumination. Lehmann, 
 as the result of numerous experiments, concluded that the 
 maximum contrastive effect was attained when the illumina- 
 tions of background and disc were in a constant ratio (about 
 4"76) to one another. Ebbinghaus' results were somewhat 
 complicated. He asserted that contrastive whiteness was 
 proportional to the difference between the intensities of 
 illumination of the contrasting fields and independent of 
 their absolute magnitudes. 
 
 Contrast darkening, on the other hand, was said to be 
 proportional to the difference of the illuminations multi- 
 plied by their quotient, and accordingly depended on absolute 
 intensity. 
 
 Hess and Pretori, who repeated the work with special 
 precautions, agree with Ebbinghaus in respect of contrastive 
 whiteness, but disagree with his rule for contrast darkening, 
 which, they hold, follows the same law as was found in the 
 opposite case. 
 
 (2) Colov/r Contrast. 
 
 This has been studied somewhat exhaustively by Pretori 
 and Sachs in Hering's laboratory. In their first series of 
 experiments they worked in the following way. In the 
 contrast-exciting field they used a coloured paper and a grey 
 paper, which looked exactly like it to a dark adapted eye. 
 
208 PHYSIOLOGY OF THE SPECIAL SENSES 
 
 In the contrast-sufFering field (for brevity, I shall write 
 c.e. and c.s. fields) a definite grey was used, formed by mixing 
 black and white, while the initial contrastive tinging was 
 sought to be eliminated by the introduction of a coloured 
 sector. In the experiments the coloured sector in the c.e. 
 field was continuously increased, that in the c.s. field kept 
 constant, the contrastive tinging being obliterated by using 
 a larger and larger black sector. They found that with a 
 constant white valency in the c.e. field, combined with in- 
 creasing colour valency, the same magnitude in colour 
 contrastive effect is reached with a simply proportional 
 diminution of white valency in the c.s. field. 
 
 If the c.e. light were kept constant, and the amount 
 of white in the c.s. field increased, the amount of contrast 
 effect rises from zero to a certain optimum amount. 
 
 In the second series of experiments the coloured sector 
 of the c.e. field was kept constant and the grey varied. 
 The coloured sector of the c.s. field was maintained con- 
 stant and the white sector varied until the contrastive 
 tinging disappeared. Under these conditions the optimum 
 contrastive effect was attained with a proportionally higher 
 white valency of the c.s. field. In the third series of 
 observations, both coloured and colourless components of 
 the c.e. field were varied, but in such ratio that the white 
 valency increased at the same rate as the colour valency. 
 In the c.s. field the white sector was varied. 
 
 In general, they found no increase in the contrastive 
 effect when both white and coloured components of the c.e. 
 field were thus varied {i.e. with a constant saturation of the 
 colour), but some experiments indicated an increase in con- 
 trastive effect with the intensity of the c.e. field up to a 
 definite maximum. 
 
 The Colour of the Stimulus and that of the 
 
 CONTEASTING FlELD 
 
 It has long been known that, under ordinary conditions, 
 the contrast colour is not exactly complementary to the 
 exciting colour, and a similar, more striking, discrepancy 
 has been observed in ordinary after-images. The fact is 
 
SIMULTANEOUS CONTRAST 209 
 
 undoubted, but the explanation is obscure. Since, in general, 
 the variation takes the form of an apparent addition of 
 blue-red to the true complementary, and is either slight or 
 absent after prolonged dark adaptation, Hering has supposed 
 that ordinary daylight possesses a certain (yellowish) colour 
 valency, so that the eye is not, in his terminology, in a 
 condition of neutral tuning. The explanation is undoubtedly 
 ingenious, but its discussion would be out of place in an 
 elementary work, a remark which also applies to individual 
 and pathological anomalies in contrast effect. 
 
 Binocular Contrast 
 
 The earliest experiments on this subject were due to 
 Smith, Brewster, Fechner, and Meyer, A white object 
 against a dark background was binocularly viewed, and one 
 eye was illuminated through the sclerotic with a beam of 
 yellowish-red light. To that eye the object appeared blue- 
 green, to the other, of the same tinge as the objective side 
 light. The most striking demonstration of binocular con- 
 trast is probably that devised by Hering. One eye looks 
 through a red, the other through a blue glass of not very 
 different brightness, both glasses being sloped obliquely 
 from the nasal to the temporal side to allow saturation to 
 be suitably regulated by mirroring white light from side 
 screens. A black stripe on a white ground is doubled by 
 increasing or diminishing the ocular convergence. Although 
 the observed background is here and there patchy, now red, 
 now blue, owing to " retinal rivalry," or sometimes a uniform 
 whitish-violet, the stripe seen through the red glass looks 
 green, and the image for the eye looking through the blue 
 glass appears yellow. 
 
 The Theories of Simultaneous Contrast 
 
 In early times much difference of opinion prevailed as 
 to whether contrast colour was objective in nature or a 
 subjective phenomenon, but the problem cannot be said to 
 have attained much importance in the theory of sensory 
 processes before the time of Helmholtz. The theory ac- 
 
210 PHYSIOLOGY OF THE SPECIAL SENSES 
 
 cepted by Brticke, and elaborated by Helmboltz, denies that 
 the basis of contrast action is a physiological alteration 
 in any of the percipient structures, and has consequently 
 been termed a psychological theory. It must, however, be 
 thoroughly grasped from the outset that consciousness was 
 not thought by Helmholtz to have any part in the genesis 
 of contrast. It is the more needful to realise this, because 
 certain passages in the treatise of Helmholtz are so worded 
 that it is very difficult to avoid the inference that he is 
 invoking a process of conscious judgment to account for 
 the facts. Some of the more superficial students of Helm- 
 holtz have undoubtedly fallen into this error. As to this, 
 it is sufficient to remark that the experiment in Binocular 
 Contrast just described is alone sufficient to disprove the 
 possibility of consciousness intervening in the matter at all, 
 since, if we make such an assumption, we should have to 
 believe in the possibility of two incorrect judgments re- 
 specting the same objects, and inconsistent one with the 
 other, being simultaneously entertained, which is evidently 
 absurd. With this caution, I shall enter upon a description 
 of the BrUcke-Helmholtz theory. 
 
 The basis of pure simultaneous contrast, especially of 
 marginal contrast, is not a change in the mechanism of 
 sensation, but a change in our infra-conscious interpretation 
 of the sensation. Through the continuous and predominat- 
 ing influence of one colour, the standard of what we call 
 white undergoes a change. Great errors in judgment as 
 to colour are obviated by the existence of retinal light (the 
 sensation experienced when the eyes are closed), but if an 
 object is viewed under circumstances which render com- 
 parison with an objective standard difficult or impossible, 
 then errors arise. Helmholtz performed the experiment of 
 Ragona Scina, already described in a slightly different way. 
 Vertical and horizontal sheets of paper were white, on the 
 vertical plate is a black, and on the horizontal a white and 
 a black scrap of paper. The black spot is seen in the con- 
 trast colour, rose-red, when the glass plate is green. Helm- 
 holtz explains this in the following terms : — 
 
 " One judges that the black spot on the lower horizontal 
 
SIMULTANEOUS CONTRAST 2U 
 
 sheet is rose-red, but also judges that one sees this spot with 
 its rose-red colour like the whole sheet through the green 
 glass, and that the green colour given by the glass extends 
 itself without a break over the whole underlying surface, 
 including the dark spot. One believes, therefore, that at 
 this place one sees simultaneously two colours, viz. green, 
 which one ascribes to the glass plate, and rose-red, which 
 one ascribes ' to the paper, and both together give, as a 
 matter of fact, the true colour of this part, namely white. 
 In fact, an object which, when viewed through green glass, 
 sends white light to the eye, like this spot, must be rose-red. 
 If we now bring above the glass plate a distinctly perceived 
 white object, there is no more reason for splitting the colour 
 of the object in two, and it looks to us white." ^ 
 
 He describes Meyer's experiment in almost identical 
 language : — 
 
 " It is the same when coloured surfaces are covered with 
 tissue paper. If the underlying surface is green, the paper 
 appears itself to be of a greenish colour. If, now, the sub- 
 stance of the paper extends without any visible interruption 
 over the grey placed underneath, one thinks one sees an 
 object shining through green paper, and such an object 
 must necessarily be rose-red in order to give white light. 
 But if the white part is definitely marked out, continuity 
 with the greenish part of the surface is destroyed, and one 
 regards it as a white object lying on that surface." ^ 
 
 Helmholtz also laid some stress on the observation of 
 Osann, that the coloured shadow effect (see above) persisted 
 when the shadow was looked at through a tube so that the 
 contrasting field was not seen. 
 
 This theory and the arguments adduced in its favour 
 have long been the object of Hering's attacks, attacks which 
 must, I think, be regarded as successful, at any rate in the 
 sense that the theory as outlined in the preceding quotations 
 can hardly be maintained. 
 
 Hering demonstrated that recognition of the grey disc (in 
 Meyer's experiment) as not forming a part of the green field 
 does not destroy the contrast. He also pointed out that 
 1 Helmholtz, Phys. OptiJc., p. 560. ^ ji^_ 
 
212 PHYSIOLOGY OF THE SPECIAL SENSES 
 
 marking off the grey disc with a black pencil line, or using 
 glazed paper (either of which, according to Helmholtz, 
 diminished the contrast effect) also diminish after-images, 
 which Helmholtz agreed to be of physiological origin. With 
 regard to Kagona Scina's experiment, he took even stronger 
 ground. He showed that in this experiment, as modified by 
 himself, the effect is just as marked when the glass plate is 
 not seen at all, or even its existence known to the observer. 
 By rotating the glass plate and the horizontal sheet the 
 rings can be seen at different distances, and even in reversed 
 positions, without change in the contrast effect. Osann's 
 observation is attributed by Hering to the disturbing 
 influence of after-image formation. Hering accounts for 
 simultaneous contrast on the general lines of the theory 
 explained in a previous chapter, anabolism or katabolism 
 in a visual substance leading to a reciprocal activity in the 
 immediate vicinity of the excited substance. In what region 
 this reciprocal effect is produced, Hering does not decide. It 
 cannot be in the retina itself, because simultaneous contrast 
 has been obtained on a complete central scotoma due to 
 retinitis of the papulo-macular bundle (Tschermak), and 
 many workers have obtained both colourless and colour 
 contrast at the blind spot. 
 
 It would be a somewhat thankless task to examine this 
 controversy in detail, but I venture to put forward for the 
 reader's consideration the following attempt at reconciling 
 the two theories. 
 
 No afferent impulse when it passes beyond the physio- 
 logical level into what one may term the infra-conscious 
 region is simplex, its entrance implies the simultaneous 
 entrance of something not itself. The basis of a colour 
 sensation, red, say, may be, perhaps must be, a unique 
 chemico-physical change in the receiving organ, but there 
 is at the next stage no such thing as red per se, only red 
 in contrast to what is not red. Supposing, therefore, the 
 products of, for instance, stimulation with white light and 
 stimulation with red light pass on into the infra-conscious 
 region we shall really have four things — red, not-red, white, 
 uot-white. Since the not-red and the not-white are, as it 
 
SIMULTANEOUS CONTRAST 213 
 
 were, solely infra-conscious products, they will tend to pass 
 away from each other towards the immediate products of 
 activity in the physiological field. There will therefore be 
 an intensification of the opposition between the " real " red 
 and the " real " white. The red and the white will have their 
 mutual difference exaggerated. But if we remember Hering's 
 analysis of the colour field — an analysis which, as the reader 
 will remember, depends in no way upon any physiological 
 hypothesis, but is the starting point of such an hypothesis — 
 the opposite to white is black and to red is green. Hence 
 we have presented in the infra-conscious sphere a (darkened) 
 red and a green. In this way one can account for the facts 
 of simultaneous contrast. Evidently the introduction of 
 some third element, which may serve as a standard of 
 reference, will destroy this simple opposition, and this is 
 what appears to happen. 
 
 I think this way of looking at the facts better than that 
 indicated in the quotations from Helmholtz, because it is 
 hard to avoid the intervention of true consciousness when 
 we use what looks like a rather complicated process of 
 inference, as in his account of Meyer's experiment. Strictly 
 speaking, all that is done is to transfer Hering's theory of 
 reciprocal action to the infra-conscious sphere — I do not 
 speak of sub-consciousness, because that term has a fairly 
 precise psychological meaning, not corresponding to my idea 
 — from the physiological level. This transference is, I think> 
 necessary. In the first place, it is very difficult to form a 
 clear 'conception of the physiological mechanism by which 
 the reciprocal action can be supposed to be effected. In the 
 second place, the idea of a simple colour sensation without 
 any existent opposite sensation is — to me at least — unattain- 
 able. I can conceive of a single physiological change pre- 
 ceding the existence of the sensation of redness, but I 
 cannot isolate that sensation when it comes into existence 
 above the physiological level. The existence of redness 
 seems to me to imply necessarily the simultaneous existence 
 of not-redness. 
 
 I claim no originality for this attempted reconciliation of 
 the rival theories, although I am not consciously borrowing 
 
214 PHYSIOLOGY OF THE SPECIAL SENSES 
 
 it from another writer. The objections to it are numerous. 
 It is psychologically crude and very incomplete. I suspect, 
 however, that investigation along these lines might well 
 enable a competent psychologist to reconcile the apparently 
 opposite theories of Hering and Helmholtz, although it may 
 reasonably be doubted whether the end would justify the 
 trouble involved. 
 
 RECOMMENDED FOR FURTHER STUDY 
 
 The reader will find a clear account of all the important researches 
 and a full bibliography of the subject in A. Tsc/iermak's Ueber Kontrast 
 und Irradiation (Ergebnisse d. Physiol., 1903, II. Abth., pp. 726-798). 
 
CHAPTER XXI 
 
 THE PHYSIOLOGY OF "SPACE" 
 
 A SCIENTIFIC poet in the noble language which never failed 
 him has set forth the conception of space as an objective 
 reality which was once accepted by philosophers, and is still 
 believed by those who have neither opportunity nor inclina- 
 tion to consider the subject with attention.^ If, he teaches, 
 we consider the nature of things, we find that there are 
 bodies and empty space. Were there no empty space, bodies 
 could have neither position nor movement. This space must 
 likewise be intangible — 
 
 " Cui si tactus erit quamvis levis exiguusque, 
 Augmine vel grandi vel parvo deuique, dum sit, 
 Corporis augebit numerum summamque sequetur." 
 
 Beyond this space and the bodies moving in it nothing has 
 existence — 
 
 " Nam quae cumque cluent, aut his conjuncta duabus 
 Rebus ea invenies aut horum eventa videbis." 
 
 The speculations of many of the greatest philosophers 
 of past time, pre-eminently perhaps those of Berkeley, have 
 moulded thought into a very different form, and from the 
 scientific side — apart from speculative metaphysics — we may 
 be content to say that " Space and Time are not realities of the 
 phenomenal world, but the modes under which we see things 
 apart. They are not infinitely large nor infinitely divisible, 
 but are essentially limited by the contents of our perception." ^ 
 
 It is, however, no part of a physiologist's business to 
 discuss these obscure riddles. We have to recognise the 
 existence of a faculty of separating simultaneous sense 
 
 ' Lucretius, De Rer. Nat., Bk. I., 418 et seq. 
 
 - The Grammar of Science, by Karl Pearson, second edition, 1900, 
 p. 191. 
 
216 PHYSIOLOGY OF THE SPECIAL SENSES 
 
 impressions or, to speak rather unscientifically, of perceiv- 
 ing spatial relations between stimuli, and we are to inquire 
 which of the sense organs are specially concerned in this 
 process. 
 
 It has always been, at least in modern times, a moot 
 point whether this power or faculty is innate or acquired, and, 
 if the latter, which sense mechanism is to be deemed of 
 primary importance. Without expressing any opinion 'at all 
 as to which school is abstractly correct, it must be admitted 
 that, physiologically, the work of the empiricists is of much 
 greater importance, and the doctor of this school whose work 
 deserves special attention is undoubtedly Berkeley. I shall, 
 therefore, examine his views at some length. 
 
 Berkeley's contribution to the subject is contained in An 
 Essay towards a New Theory of Vision, which was first 
 published in 1709, when its author was only twenty-four years 
 of age, being perhaps the most important contribution to 
 modern philosophy which has been made by so young a man. 
 Berkeley's doctrine can be summarised in a few words. 
 Distance and magnitude are not directly determined by the 
 nature of the images formed on the retina, nor by changes in 
 the inclination of the optic axes. They are ideas which we 
 form as the result of experience, and are not, strictly speaking, 
 sensations at all. 
 
 " I know it is a received opinion that, by altering the 
 disposition of the eyes, the mind perceives whether the angle 
 of the optic axes, or the lateral angle comprehended between 
 the interval of the eyes and the optic axes, are made greater 
 or lesser ; and thus, accordingly, by a kind of naural geometry, 
 it judges the point of their intersection to be nearer or 
 farther ofT. But that this is not true I am convinced by my 
 own experience, since I am not conscious that I make any 
 such use of the perception I have by the turn of my eyes. 
 And for me to make these judgments, and draw these con- 
 clusions from it, without knowing that I do so, seems 
 altogether imcomprehensible." ^ 
 
 The position is defined in the following admirable passage : 
 " From what we have shown, it is a manifest consequence 
 1 Berkeley's Works, vol. i. p. 83 (Sampson's Edition, London, Bell, 1897). 
 
THE PHYSIOLOGY OF. "SPACE" 217 
 
 that the ideas of space, outness, and things placed at a distance 
 are not, strictly speaking, the objects of sight ; they are not 
 otherwise perceived by the eye than by the ear. Sitting in 
 my study I hear a coach drive along the street; I look 
 through the casement and see it ; I walk out and enter it. 
 Thus, common speech would incline one to think I heard, 
 saw, and touched the same thing, to wit, the coach. It is 
 nevertheless certain the ideas intromitted by each sense are 
 widely different and distinct from each other; but having 
 been observed constantly to go together, they are spoken of 
 a,s one and the same thing. By the variation of the noise 
 I perceive the different distances of the coach, and know that 
 it approaches before I look out. Thus, by the ear I perceive 
 distance j ust after the same manner as I do by the eye. I 
 do not nevertheless say I hear distance in like manner as I 
 say that I see it — the ideas perceived by hearing not being 
 so apt to be confounded with the ideas of touch as those of 
 sight are." ^ 
 
 The concluding sentence of this passage will suggest to 
 the reader the weak point of Berkeley's reasoning. Although 
 he is careful to point out that there is no necessary but only 
 a habitual connection between visible and tangible magni- 
 tude,2 yet in his detailed applications of the theory he is 
 constantly accounting for visual sensations of magnitude by 
 their connection with tangible magnitudes. Thus he con- 
 cludes a discussion of units of length with the words: 
 " From all which it is manifest that the judgments we make 
 of the magnitudes of objects by sight are altogether in re- 
 ference to their tangible extension." But there is, in theory, 
 no reason why visual space should be subordinate to tangible 
 space ; the contrary proposition could equally well be main- 
 tained. We might, for instance, imagine that our unit of 
 length is not a tangible impression involving some number 
 of touch areas, but a retinal impression affecting a certain 
 number of conoidal areas. I take it the reason why tangible 
 magnitude is taken to be primitive is that cases of congenital 
 blindness with normal tactile sensibility are common, while 
 cases of total cutaneous anaesthesia of congenital origin com- 
 
 1 Op. eit., p. 98. ^ Ibid., p. 102. 
 
218 PHYSIOLOGY OF THE SPECIAL SENSES 
 
 bined with normal vision either never occur or have never 
 been described. It would, therefore, seem that Berkeley's 
 treatment of the subject, admirable though it is, does not 
 give an adequate account of spatial perception in general ; 
 its physiological importance is its insistence on a study of 
 the empirical facts. 
 
 A somewhat similar objection has been urged by Professor 
 James against the speculations of an even greater thinker on 
 the subject, Helmholtz. I shall now, without further re- 
 ference to the theory of the matter, enumerate the physio- 
 logical data which have to be taken into consideration, 
 beginning with the information derived from the cutaneous 
 mechanisms. 
 
 It is a matter of common knowledge that we possess the 
 power of localising the point of application of a tactile 
 stimulus, and that the fineness of this power varies in dif- 
 ferent persons and in different skin areas in the same subject. 
 
 Experimentally, the matter can be investigated by the 
 use of some form of sesthesiometer. The most useful form is 
 an instrument constructed on the principle of a compass, the 
 points being covered in such a way that pain is not produced 
 by their application, and that the surface of contact with the 
 skin is identical in both limbs. Since if two stimuli are 
 applied together, we can when they are more than a certain 
 distant apart recognise that two points are stimulated, and 
 since when two different points are successively touched, we 
 can recognise that the second stimulus was not applied to 
 the same region as the first, two sets of experiments are 
 to be planned. In the first set we apply the stimuli to- 
 gether and determine the shortest distance which corresponds- 
 to (a) a duplicate sensation, (b) a single sensation having 
 lineal extension. In the second set the stimuli are 
 successively employed, and we ascertain the shortest distance 
 enabling the subject to perceive (a) that the second stimulated 
 point is not Identical with the first (b) exactly where the 
 second poiny is situated. The general conclusion to be 
 drawn from such experiments is, that the liminal distance 
 for simultaneous is far greater than for successive stimulation, 
 V. Frey concludes that if two neighbouring touch spots are 
 
THE PHYSIOLOGY OF "SPACE"- 219 
 
 stimulated successively, one can always, under favourable 
 experimental conditions, recognise that the second stimulus 
 was not applied to the same point as the first. If the spots 
 are touched at the same time, then a duplicate sensation 
 does not arise, the latter being only produced if one un- 
 stimulated touch area lies between the points of application. 
 (The actual liminal distance for simultaneous stimulation varies 
 from about 1*1 mm. on the tip of the tongue to over 
 65 on the upper arm or thigh. In the case of the limbs 
 the direction of the line joining the two points is not in- 
 different, the liminal distance being greater for stimulation 
 parallel with the long axis of the limb than for transverse 
 application. The actual values obtained depend very greatly 
 on the conditions of the experiment, the influence of fatigue 
 being particularly appareny Local vascular conditions also 
 play a part, although different workers are not agreed as to 
 their exact influence. (^It has also been generally asserted 
 that practice diminishes the liminal distance, but this 
 statement has not, I think, been entirely substantiated/ 
 That any change in the usual relations of the part stimulated 
 greatly affects one's power of localisation is well seen in any 
 of the modifications of Aristotle's experiment, the best being 
 that of Henri. When the middle and ring fingers are 
 crossed and the subject is given a diagram of the fingers 
 in their normal relation on which he is to mark the point 
 touched, stimuli applied to the radial side of the crossed 
 finger tend to be localised towards the ulnar margin. 
 Without discussing the theoretical interpretation of these 
 results, we can pass to the visual data for space perception. 
 
 The physiological basis of spatial perception, in so far as 
 it depends on visual sensations, is to be sought in move- 
 ments of the eyes. The musculature of the eye and the 
 exact shape of the eyeball, together with the direction and 
 force with which each muscle pulls, have been elaborately 
 studied. Of the methods adopted, two are of special interest. 
 
 The Method of After-images. 
 A coloured ■ piece of paper is affixed to a sheet of grey 
 paper ruled in squares of known size. The after-image of 
 
220 PHYSIOLOGY OF THE SPECIAL SENSES 
 
 the coloured scrap appears on different parts of the ruled 
 surface in accordance with the way in which the eye has 
 been moved; the ruled squares enable the position to be 
 accurately defined. 
 
 The Substitution Method of Hering 
 Probably the best form of this method is that adopted by 
 Bonders in his Isoscope (Fig. 23). RR is a fixed framework 
 on which a movable parallelogram is fixed. D is a thread 
 attached to the frame EE, two threads attached to the 
 parallelogram and moving with it. The parallelogram is mov- 
 
 FiG. 23. 
 
 able round an antero-posterior axis, and supplied with an index 
 pointer F reading on a scale G. With this instrument one can 
 measure the inclination of EE and D to one another which 
 gives parallel double images when both eyes fixate a given 
 point. In this way one can study the apparent situation of 
 images when the relation of the two retinse is altered. 
 
 It is usual to call the arrangement of the eyes which 
 obtains when the head is vertical and the visual axes parallel 
 to the plan of the ground the " primary . position." To 
 change from this position either directly up or down or 
 
THE PHYSIOLOGY OF "SPACE" 221 
 
 directly to the right or the left, the eyes never rotate round 
 an antero-posterior axis, like, for instance, the steering-wheel 
 of a ship. The last sentence contains the teaching of Listing 
 on this point, and is usually termed Listing's " Law." Another 
 "Law," first propounded by Bonders, is that for identical 
 directions of the eyes the orientation of the retinae is 
 identical, however the eyes may have been rotated into 
 the final position. Both these statements only apply to 
 cases in which the visual axes are approximately parallel 
 and movements are effected without undue muscular strain. 
 It is of mechanical and physiological advantage for rotation 
 to be effected in accordance with Listing's principle; the 
 mechanical advantage is that the movement follows the 
 shortest path and, other things being equal, can be effected 
 in the shortest time; the physiological advantage (Hering) 
 is that apparent rotation of an external object, owing to 
 wheel-like movements of the eye, does not occur, and 
 that the number of corresponding retinal points (vide 
 infra) is maximal. 
 
 Numerous researches, mainly by the method of after- 
 images, seem to demonstrate that even for parallel, axes 
 the law of Listing is not strictly true, but the deviations, at 
 least for normal eyes, are not large. 
 
 MoNocuLAE Vision 
 
 The first question which arises is with regard to visual 
 acuity, or the power of distinguishing between the images 
 formed by different objects or by different parts of one 
 object. The simplest case is an answer to the question. How 
 far apart must two points be in order to be distinguished? 
 The answer depends upon several circumstances, of which 
 the most important is the amount of contrast between the 
 points and the background (e.g. black points on a white 
 surface) against which they are viewed, so that the minimum 
 distance varies enormously in accordance with the experi- 
 mental conditions. I cite a few examples on the following 
 page. 
 
222 PHYSIOLOGY OF THE SPECIAL SENSES 
 
 Object. 
 
 Visual Angle. 
 
 Observer. 
 
 Parallel threads . . ... 
 
 50 sees. 
 
 Hirschniann. 
 
 Black points (on a white ground) 
 
 1 min. 4 sees. 
 
 Hueck. 
 
 White squares (on a blackground) 
 
 55 sees. 
 
 Aubert. 
 
 Parallel black lines alternating 
 with equally broad white ones . 
 
 52 secs.-lmin. 15 
 sees. 
 
 Bergmann, 
 
 Jupiter's satellites 
 
 More than 2 mins. 
 
 A. Humboldt. 
 
 With good conditions of contrast and moderate illumina- 
 tion, Snellen's assumption that a normal eye can distinguish 
 points separated by an angular distance of 1 minute is probably 
 accurate, enough for most purposes. Calculations made on 
 the basis of such experiments as to the actual size of the 
 retinal images suggest that, as in the case of touch, one 
 unstimulated retinal element must always intervene for 
 stimuli affecting two other elements to be separately sensed. 
 It is generally held that, in the case of foveal vision, the 
 retinal elements in question are the cones; whether the 
 inner or the outer limbs of the cones are involved is a moot 
 point, but Hensen has advanced somewhat cogent evidence 
 in favour of the outer limbs. 
 
 Eor practical purposes, visual acuity at the fovea is tested 
 by the method invented by Jager and improved by Snellen, 
 which is too well known to need description. The hypo- 
 thesis of Snellen that the legibility of his test letters is pro- 
 portional to the visual angle in every direction of the letter 
 has not escaped question. According to Guillery, there is 
 no uniform relation between power to recognise an object 
 and the size of its retinal image in simple objects, and, a 
 fortiori, none in the case of such complicated figures as 
 letters. The ease with which Snellen's types can be recog- 
 nised is greater than one's power of isolating the squares of 
 a chess-board pattern seen under a visual angle of 1 minute for 
 each square and with effective contrast. The fact is, that 
 in reading letters of a familiar alphabet we fill in the letter 
 mentally from seeing a small part of it, j ust as in a Persian 
 
THE PHYSIOLOGY OF "SPACE" 223 
 
 manuscript the letters are not written out in full, only the (to 
 a native) characteristic portions of each. Snellen's method 
 is, therefore, a test of the oninimwm of visual acuity. 
 
 Snellen's method has been modified by Cohn, who sub- 
 stitutes combinations of lines for the letters so that it can 
 be used for illiterate persons. Cohn, who has investigated 
 the normal acuity of uncivilised peoples (Beduins, Egyptians, 
 etc.), is of opinion that the general belief that civilised peoples 
 have lower visual acuity than uncivilised races is unfounded. 
 His data have not, however, been submitted to rigorous 
 statistical analysis. Cohn reports some cases of visual acute- 
 ness far transcending the normal limits; thus an Egyptian 
 boy of sixteen had an acuity eight times the normal value, 
 and a girl of eleven, six times the normal. 
 
 It is improbable in view of Pearson and Barrington's 
 work that unfavourable environment — for instance, too close 
 application to books in early childhood — is of such importance 
 in deteriorating eyesight as various popular authors have 
 alleged, but the subject must still be considered as of a con- 
 troversial nature. 
 
 Hering has pointed out that methods such as those 
 described, which rest on the determination of the just-notice- 
 able distance between two objects, do not reveal the fineness 
 of the visual "Raumsinn" for the determination of "the 
 slightest difference in situation or magnitude which the eye 
 can appreciate." This is analogous to the skin sense which 
 is able to distinguish two points successively touched when 
 closer together than two points simultaneously stimulated 
 — i.e. contiguous elements of the retinal surface give rise to 
 different sensations or have " local sign." 
 
 So far we have considered foveal acuity; as we pass 
 towards the periphery of the retina, the acuteness of vision 
 diminishes, under ordinary circumstances, fast, but the change 
 is not spatially symmetrical. The falling off is more rapid 
 in the vertical than in the horizontal meridian, and in the 
 latter is more rapid externally than internally. Important 
 researches on the comparative acuteness of vision in light 
 and dark adapted eyes have been carried out by v. Kries and 
 his pupils. V. Kries found that in dark adaptation visual 
 
224 PHYSIOLOGY OF THE SPECIAL SENSES 
 
 acuity between tlie centre of the fovea and the blind spot, at 
 a distance of 4 to 12 degrees from the fovea, was practically 
 constant, while at the fovea itself the test objects were imper- 
 ceptible. From the blind spot outwards, acuity is the same 
 for light and dark adapted eyes. Koester taking visual 
 acuity as zero at the fovea ( " dark " eye), found it rapidly 
 rising at an eccentricity of 5 to 10 degrees, and from thence 
 fairly constant towards the periphery. At an eccentricity of 
 30 to 40 degrees, " dark " visual acuity is greater than that of 
 the " light " eye. 
 
 Bloom and Garten obtained somewhat different results; 
 according to them,neither centrum nor periphery of the "dark" 
 eye attains the degree of visual acuity possessed in light 
 adaptation. Central and peripheral acuity are affected in 
 the same way by dark adaptation, but not to the same ex- 
 tent ; only when the illumination is very feeble does the dark 
 adapted retina give the better response. 
 
 Perception of Depth with the help of one Eye 
 
 Strictly speaking, we can with the help of one eye only 
 perceive differences of directions in lines along which points 
 are situated, but accumulated experience enables us to dis- 
 tinguish depth and form with moderate accuracy. How 
 large a part is played by past experience will be clear when 
 we recall the numerous illusions of form which beset us when 
 we examine distant objects for the first time. Factors which 
 influence our judgment of relief (either in the case of mono- 
 cular or binocular vision) are (1) the distribution of shadows, 
 as seen in the fine sculpturipg of snow-clad hills and valleys 
 when the sun is low on the horizon; (2) the state of the 
 atmosphere, aerial perspective, which causes the citizen to 
 misjudge distances greatly when in clear mountain air. 
 
 For the judgment of the distances of near objects the 
 state of accommodation, although of importance, does not 
 give so much help as might be expected. Wundt has ex- 
 perimented on this point. The subject looks at a black 
 thread, which can be moved nearer to or farther from a 
 white screen, through an opening. 
 
THE PHYSIOLOGY OF "SPACE" 
 
 225 
 
 Although the absolute distance of the thread could not 
 be gauged, yet, within limits, it could be ascertained in which 
 direction the thread had been moved. 
 
 Distance of Thread. 
 
 Liminal Movement. 
 
 Wlien moved Forwards. 
 
 When moved Backwards. 
 
 Metres. 
 
 2-5 
 
 2-2 
 
 2 
 
 1-8 
 
 1 
 •8 
 ■5 
 ■4 
 
 Centimetres. 
 
 12 
 
 10 
 8 
 8 
 8 
 5 
 
 4-5 
 4-5 
 
 Centimetres. 
 
 12 
 
 12 
 
 12 
 
 12 
 
 11 
 7 
 
 6-5 
 4-5 
 
 These experiments show sufficiently well how inexact are 
 the data afforded by monocular vision for judging distance 
 with the help of accommodation, and they are really too 
 favourable, for information is afforded by changes in the 
 brightness and distinctness of the threads. Hillebrand and 
 others, by methods in which these latter sources of infor- 
 mation were excluded, have found that the data afibrded 
 by changes in accommodation (or, to speak more precisely, 
 changes in accommodation and convergence) are even less 
 complete than Wundt's work suggests. 
 
 We all know that changes in the position of the eye — 
 or head — with respect to the object improve our judg- 
 ment of solidity. One reason why this should be so is 
 plain. When the eye is moved the rapidity with which a 
 retinal image changes its position depends on the distance 
 of the object, and is inversely proportional to it. If there 
 be two objects in the field of vision, and one moves faster 
 than the other when the head is moved, we conclude the 
 former to be nearer to us. Some experiments carried 
 out by Bourdon are of interest in this connection. Two 
 points, separated in one experiment 6 degrees, in the other 
 1 degree apart, were placed at about 6 metres from the 
 eye, one point being higher than the other. The observer 
 was allowed to move his head but not his body. In the 
 
 p 
 
226 PHYSIOLOGY OF THE SPECIAL SENSES 
 
 case of the higher point being the farther off, a correct 
 answer was given for both distances; when the lower 
 point was the farther off, the results were not so good 
 but generally better for the smaller angular distances. 
 
 Judgment of Shape 
 
 This part of the subject, which includes a study of the 
 numerous geometrical illusions, such as that of Zollner and 
 Poggendorf, can only be dealt with in a few words. Many 
 experiments have been performed to determine how much 
 two straight lines must differ in order that we may per- 
 ceive the difference, but the numerical values of different 
 observers are discordant. Horizontal lines can be more 
 accurately compared than vertical ones, and a comparison 
 of horizontal and vertical distances leads to still greater 
 errors.' If one attempts to draw a square on a plane at 
 right angles to the line of sight, the vertical sides may 
 be as much as one-fifth shorter than the horizontal ones, 
 the least error being one-sixtieth to one-thirtieth. According 
 to Feilchenfeld, when a rectangular cross is approximated 
 to the eye one increasingly over-estimates the nasal limb, 
 a result which he attributes to the greater extent of the 
 temporal visual field. If one freely chooses the fixation 
 point in monocular vision, the temporal half of a line to be 
 bisected is over-estimated because (Feilchenfeld) the point 
 which lies along the fixation line, and is too much towards 
 the temporal side, is erroneously taken to be the centre. 
 These explanations appear to me to leave the experimental 
 facts pretty much where they were before. 
 
 The various ways in which the eye can be deceived with 
 regard to form and direction are too well known through the 
 medium of popular magazines to need reproduction, and I 
 confine myself to a diagram of the Mliller-Lyer illusion 
 (Fig. 24). The part of the horizontal line towards which 
 the short lines approach appears shorter than the exactly 
 equal portion from which the lines recede. Painstaking 
 attempts have been made to interpret this and similar 
 illusions, with indifferent success. Some regard the basis 
 
THE PHYSIOLOGY OF "SPACE" 227 
 
 as directly sensational, others as secondary or dependent 
 upon a process of subconscious or unconscious inference. 
 
 Fie. 24. 
 
 Helmholtz has interpreted many of these illusions on 
 the basis of a " law of contrast," according to which clearly 
 perceived differences are judged to be greater than indis- 
 tinctly perceived ones, a reversal of the old dictum, omne 
 ignotum pro nnagnifico. He explains in this way the well- 
 known illusion that a distance divided by dots appears greater 
 than the same space not divided, the former being more 
 distinctly seen, and therefore judged to be greater. In other 
 cases, irradiation and ocular movements are thought to be 
 influential. 
 
 Hering has suggested that the apparent distance between 
 two points depends on the linear distance of their retinal 
 image points. Hence, small distances are more accurately 
 judged than large ones, since the length of the chord of a 
 small arc approximates to that of the arc. Einthoven attri- 
 butes certain illusions to differences of distinctness between 
 images formed at the fovea and on the peripheral retina. It 
 may, perhaps, be said that since our knowledge of all the 
 factors which are involved in judgment of size is, even at the 
 physiological level, imperfect, the explanations put forward 
 are rather of interest as intellectual exercises than as com- 
 plete interpretations of the phenomena. One interesting 
 problem must be mentioned, viz. the reason why the moon on 
 the horizon looks larger than when she is at the zenith, a 
 question which has been touched on by such proficients as 
 Berkeley, Gauss, and Helmholtz. The experiments of Zoth 
 appear to prove that the influence of aerial perspective, of the 
 comparison with objects of known size, and of the apparent 
 curvature of the sky, are not the main determining factors in 
 the illusion. Zoth holds that the apparently smaller size of 
 the moon at the zenith depends on the necessary elevation 
 of the eye. Guttman's experiments, which demonstrate an 
 
228 PHYSIOLOGY OF THE SPECIAL SENSES 
 
 apparent diminution in size of 3^ to 3| per cent, in objects 
 25 to 36 cm. distant placed so that the line of vision has an 
 elevation of 40 degrees, confirm this view, but only push the 
 difficulty one stage farther back. The association of upward 
 gaze and diminution in size may possibly be due to the con- 
 vergence which occurs when the eyes look upwards (Zoth). 
 
 Binocular Vision 
 
 The most obvious advantage of using both eyes is the 
 increased power we notice of observing the solidity of objects, 
 but before we examine the physiological basis of this we 
 must consider some of the data which help us to understand 
 the joint working of the eyes. Since whenever we turn our 
 eyes to a point a separate image of it is formed on each 
 retina, it is a remarkable thing that we see not two points, but 
 one only. 
 
 This remarkable and, so far as we are concerned, ultimate 
 fact can be expressed by the statement that the retinse con- 
 tain areas or points which " correspond," viz. the simultaneous 
 excitation of which calls up a single sensation. We also 
 know that when the ordinary mutual arrangement of the eyes 
 is disturbed, e.g. by pushing one eyeball with the finger, a 
 single object gives rise to two images. Hence, any two 
 retinal areas do not indifferently " correspond " ; some retinal 
 points are disparate. The first problem is, accordingly, to 
 determine how the "corresponding" points lie with reference 
 to some fixed retinal point taken as origin, or, in other words, 
 to map out the binocular field of single vision. 
 
 The locus of the external points, images of which fall 
 on " corresponding " points of the two retina, is called the 
 Horopter, or, more strictly, the Point Horopter, and it can be 
 investigated in numerous ways. Theoretically, it might seem 
 an easy task to map out the horopter experimentally ; a point 
 might be fixed and a small object moved round on a peri- 
 meter, the positions being noted in which it was seen as a 
 single point or as two points. Again, a haploscopic arrange- 
 ment can be used, mirrors capable of rotation on different 
 axes being adapted to the purpose. Unfortunately these ex- 
 
THE PHYSIOLOGY OF "SPACE" 229 
 
 periments are very difficult to carry out ; they require a highly 
 trained subject as well as an experienced observer, for recog- 
 nition of double images (diplopia) is only easy in extreme 
 cases. It has, therefore, been more usual to determine the 
 horopter theoretically, using a few simple experimental data 
 as a starting point. The data used are the following : — 
 
 Simple observation appears to show that symmetrical 
 points of the two retinae "correspond," e.g. that when a 
 point on the right retina 1 millimetre to the left of the 
 fovea in the horizontal meridian and a point 1 millimetre to 
 the left of the left fovea in the same meridian are stimulated, 
 that a single image is seen. Accepting this datum, the pro- 
 blem reduces itself to finding the locus of external points, the 
 images of which are formed on symmetrical areas of the 
 retina and becomes purely geometrical. The geometrical 
 problem is indeed somewhat complex, because inter alia the 
 eyes can be moved in so many directions, but it is one which 
 a highly skilled geometer could solve without ever leaving 
 his study. It will be observed that two assumptions are 
 made in stating the problem in this form, viz. (1) the stimula- 
 tion of symmetrical points is always attended by single 
 vision ; (2) the stimulation of asymmetrical points does not 
 call up a single image. Of these assumptions, (1) is probably 
 but not certainly true, (2) is almost certainly false. It thus 
 follows that the geometrical horopter does not correspond 
 accurately to the physiological one, but it is true that the 
 divergence is not, so far as we know, very considerable. I do 
 not think, however, that I should be justified in reproducing 
 the difficult investigation by which the theoretical horopter 
 has been obtained, and it is possible that even a statement of 
 its results will not be quite clear to all readers. Helmholtz 
 found that the point horopter is in general a curve of double 
 curvature, and can be regarded as the section of two surfaces 
 of the second order, which have in addition a common straight 
 line. In a particular case only is the point horopter a plane 
 surface, viz. when the fixation point lies at infinity in the 
 mesial plane, and the horizontal meridia of the retinse are 
 in the visual plane. Under these conditions the horopteric 
 plane is either perpendicular to the line of sight and at 
 
230 PHYSIOLOGY OF THE SPECIAL SENSES 
 
 infinity, or parallel to the visual plane and passes through the 
 line of intersection of the planes of the apparently vertical 
 meridian ; for normal eyes the line of section and the 
 horopteric plane nearly coincide with the plane of the ground 
 on which the observer stands. When we consider not points 
 but lines the problem becomes still more complex, and can- 
 not be even summarised in an elementary treatise. 
 
 Fig. 25. 
 
 We must now touch on the rules which appear to govern 
 the projection of objects viewed with both eyes. The most 
 important empirical law is that worked out by Hering, and is 
 very simple to understand. 
 
 Let A and B represent the right and left eyes, and A^ and Bj 
 their nodal points, C an imaginary "cyclopean" eye, and Cj^ 
 its nodal point. Let D be the point to which the eyes 
 converge, so that its images are formed on corresponding 
 points. Join DAp DBp and DCp and produce to intersect 
 
THE PHYSIOLOGY OF "SPACE' 
 
 231 
 
 the retinae. Then the point D appears along the line DC^. 
 In a similar way the localisation of double images is clear.^ 
 
 We now come to the factors which render the two eyes 
 working together more efficient than the separate organs for 
 discriminating sizes. Of these, convergence and binocular 
 parallax are the most important. With regard to the former, 
 the experiments of Wundt, Bourdon, and others appear to 
 justify the following statement : Convergence affords a better 
 means of measuring distance than do changes in the state of 
 
 ^B 
 
 Fig. 26. 
 
 accommodation, yet the information so obtained does not rank 
 in importance with that due to parallax. 
 
 The importance of parallax, the apparent difference in an 
 object viewed from two different points in the special case of 
 binocular vision, was first described by Wheatstone in 1838. 
 Let AB (Fig. 26) be the horizontal section of a plane at right 
 angles to the plane of the paper, which we will regard as 
 the visual plane, P and Q the mid-points of the lines of vision 
 of the two eyes, and S a point looked at. E and T are 
 the projections of S on AB. Let a be the distance between 
 the eyes and c the distance from R to T, & the distance 
 
 1 When D is fixated, the images of the point E are formed at Ej and Ej. 
 (Fi) and (Ej) are the points which correspond to Fj and Ej upon the Cyclopean 
 Eetina. Join (Fj) and (Ej) to G and produce to intersect EFj and EE^. The 
 image due to the right eye is projected along (Fi)H, i.e. there is Uncrossed 
 Diplopia. In a similar way, we find the image of the nearer point F due to the 
 right eye is projected to the left — Crossed Diplopia. 
 
232 PHYSIOLOGY OF THE SPECIAL SENSES 
 
 of AB from PQ, / the distance from S to PQ, and d the 
 
 distance from S to AB. 
 
 Since the triangles PSQ and EST are similar, 
 
 c a , ad 
 
 d=f andc = j. 
 
 a — c = e, can be called the stereoscopic distance (Helmholtz). 
 Further, 
 
 b a — c be bn 
 
 j= — or e=j- = -f. 
 a c d, J 
 
 Both c and e diminish when / increases, and vanish when / 
 is infinite ; in other words, there is no binocular parallax for 
 very distant objects. Both c and e increase with the inter- 
 ocular distances, c increases as the distance of the object 
 from the plane of projection, e as the distance of the projec- 
 tion plane from the eyes. 
 
 Many experimental determinations of the fineness of 
 depth perception have been carried out. Helmholtz measured 
 how far the central one of three needles must be moved 
 
 Bourdon's Experiments on Judgment of Distance. 
 
 [The Plane of the Fixed Needles was 2 meters from the Eyes.] 
 [The Needles were 35 mm. apart.] 
 
 
 Results of Twenty Trials. | 
 
 Position of the 
 Middle Keedle. 
 
 
 
 
 
 
 
 
 "Nearer." 
 
 The same 
 Distance." 
 
 " Farther." 
 
 r 3-0 
 
 20 
 
 
 
 
 
 
 2'5 
 
 20 
 
 
 
 
 
 In front of 
 the others 
 
 2-0 
 1-5 
 1-0 
 
 19 
 
 18 
 17 
 
 1 
 
 2 
 3 
 
 
 
 
 
 
 •5 
 
 11 
 
 9 
 
 
 
 [ -0 
 
 2 
 
 18 
 
 
 
 r -5 
 
 3 
 
 15 
 
 2 
 
 
 1-0 
 
 
 
 13 
 
 7 
 
 Behind the 
 
 1-5 
 
 
 
 11 
 
 9 
 
 others ' 2-0 
 
 
 
 3 
 
 17 
 
 2'5 
 
 
 
 
 
 20 
 
 i 3-0 
 
 
 
 1 
 
 19 
 
 forwards or backwards for a difference to be noticed. He 
 found that a needle 34 centimetres from the eyes when 
 
THE PHYSIOLOGY OF "SPACE" 233 
 
 moved through half a millimetre, approximately its thick- 
 ness, backwards or forwards from the plane of the others, 
 could be detected with certainty as being no longer in 
 that plane. Bourdon's experiments reveal even greater 
 accuracy. Calculation on the basis of Bourdon's work shows 
 that a diiference in position of the retinal images corre- 
 sponding to five angular seconds is appreciable. In ordinary 
 observation of a solid body we are accustomed to change the 
 fixation point frequently, and it has been thought that these 
 eye movements are of great importance in spatial perception. 
 It is, however, fairly certain that although such movements 
 are not without value, their importance in comparison with 
 binocular parallax is secondary. 
 
 Stereoscopy 
 
 I shall conclude this chapter with a few words descriptive 
 of a matter which, although not a part of the physiology 
 of visible space, is germane to it, viz. the theory of the 
 stereoscope. 
 
 If it be true that binocular parallax is the main founda- 
 tion of what we call solidity or depth, it should follow that 
 when two pictures of an object taken from the point of view 
 of each eye are so presented to the binocular combination 
 that the images are made to fuse, an impression of solidity 
 is produced. This deduction, first made by Wheatstone, 
 has been the foundation of optical experiments which have 
 produced instruments of great interest and practical use. 
 
 Photographic pictures for stereoscopy are taken with 
 objectives somewhat farther apart than the average distance 
 between the eyes, and accordingly represent such effects as 
 would be produced naturally by rather nearer objects than those 
 which they represent, i.e. there is an exaggeration of the relief. 
 Owing to the fact that the observed points lie in a plane, the 
 state of accommodation associated with the convergence of 
 the optic axes is not quite appropriate ; there is not a com- 
 plete reproduction of the natural state of affairs. To unite 
 the two pictures Wheatstone employed an arrangement of 
 mirrors, but Brewster's device of two prisms (Fig. 27) has 
 
234 PHYSIOLOGY OF THE SPECIAL SENSES 
 
 proved more satisfactory. The older instruments were 
 mostly box-shaped, light being admitted through a slit 
 in the front wall, but these have been largely replaced 
 
 D 
 
 P^ 
 
 J>, 
 
 ^-' 
 
 A and B The Eyes. 
 
 C and Ci Prisms. 
 
 D and D, Stereoscopic Pictures. 
 
 E Partition. 
 
 F Apparent situation of 
 Blended Picture. 
 
 Fig. 27. 
 
 by open instruments, such as the admirable hand stereoscope 
 of the Zeiss works. 
 
 Very useful applications of the stereoscope have resulted 
 from modifications of the distance between the eyes; the 
 principle of one of these is shown in the diagram (Fig. 28). 
 
 Supposing, now, that one places in the image planes of a 
 binocular telescope two glass plates with a series of corre- 
 sponding signs, in accordance with the ocular distance 
 (modified by and a constant of the instrument) ; some object 
 is fixated, and one notices at which point in the series of 
 marks it is seen stereoscopically. Each of the pairs of marks 
 corresponds to a definite distance (depending on the cor- 
 rected ocular distance), and in this way the distance of the 
 object can be directly read off. This is the principle of the 
 Stereotelemeter made by Zeiss ; the accuracy of their smaller 
 
THE PHYSIOLOGY OF "SPACE" 235 
 
 instrument is ± 5 cm. for a distance of 20 metres, +31-3 m. 
 for 500 metres. Their larger instrument reads to 440 m. in 
 10 kilometres. 
 
 If two absolutely identical pictures or impressions on a 
 plane surface are examined with a stereoscope, the fused 
 impression, as we should expect, is once more that of a flat 
 surface. If, however, the two patterns are not absolutely 
 
 B and J5, The Eyes. 
 
 B and B, Small Mirrors inclined at an 
 angle of 45° to the mesial 
 
 ©plane of the observer. 
 C and C| Large Mirrors parallel to E 
 A and £?,. 
 
 -^ By this arrangement an effect is produced 
 much as if the eyes were situated at A 
 and B. 
 
 Fig. 28. 
 
 6 
 
 J3 
 
 identical, then, in accordance with the principles of stereo- 
 scopic vision an imperfect appearance of relief, an obliquely 
 vaulted image is obtained. Medals stamped with the same 
 die but from different metals, although indistinguishable to 
 touch or by the naked eye, have been found to produce this 
 effect (Dove) ; different editions of printed books and forged 
 bank-notes can also be detected in this way. Eor such 
 purposes, a modified stereoscope, Pulfrich's Stereocom- 
 parator, has been introduced with satisfactory results ; indeed, 
 the adaptations and modifications of Wheatstone's instru- 
 ment are numerous, and will in time be more so. 
 
 The brief sketch presented in this chapter will, perhaps, 
 help the reader to form an idea of the diverse fields of 
 intellectual research which are touched on by the physio- 
 logical study of spatial magnitude ; he will find that almost 
 any of the branches of inquiry which each part of the subject 
 
236 PHYSIOLOGY OF THE SPECIAL SENSES 
 
 suggests can be made the source of great pleasure and in- 
 struction. 
 
 RECOMMENDED FOR FURTHER STUDY 
 
 (i) The masterly article on Space in the second volume of Professor 
 William Jomes' Principles of Psycholocry should first te mastered. Next 
 come (2) Helmholtz's treatise, and (3) Hering's article in Hermann's 
 Handb. d. Phys., Bk. iii. part 1. (4) An admirable summary of the 
 physiological data relating to visual space is given in Zotlt'a article in 
 Nagel's Handb., vol. iii. 
 
INDEX 
 
 Acuity — 
 
 Auditory, 72 
 
 Visual, 222-3 
 Adaptation — 
 
 Theory of Visual, 118 
 
 Thermal, 15 
 
 Visual, 101-112 
 After-images, 155 
 
 Burch on, 163 
 
 Goethe on, 156 
 
 Hering on, 198 
 
 Positive, 156 
 
 White Values of, 162 
 Analysis, Cochleal, 81 
 Answers, Method of Right 
 Wrong, 4 
 
 Beat Tones, 76 
 
 Beats, 75 
 
 Blue Sightedness, 137 
 
 Canals, Semicircular, 58 
 Cochlea, 80 
 
 Analytical Theory of, 81 
 Coefficient Rule, 159 
 Colour — ■ 
 
 Absorption by Macula 
 Lens, 136 
 
 Complementary, 135 
 
 Memory, 125 
 
 Mixing, 128 
 
 Real and Accidental, 125 
 
 Specific Brightness of, 
 195 
 
 Triangle, 131 
 
 Vision, 124-214 
 Colour-blindness — 
 
 Artificial, 184 
 
 and 
 
 and 
 
 194, 
 
 Colour-blindness (continued) — 
 
 Blue- Yellow, 152 
 
 Goethe on, 139 
 
 Green-Red, 141 
 
 Tests for, 146 
 
 Theories of, 182, 189 
 
 Total, 107, 108, 111 
 Cones, Vision of, 120 
 Contrast — 
 
 Binocular, 209 
 
 Colour, 207 
 
 Smell, 44 
 
 Spatial Efi'ect of, 205 
 
 Taste, 37 
 
 Temporal Succession of, 206 
 
 Theory of, 209 
 
 Visual, 203-214 
 
 White-Black, 207 
 Corti's Organ, 81 
 Cutaneous Sensations, 8-24 
 Cyclopean Eye, 230 
 
 Deaf Mutism, 59 
 Depth, Perception of, 224 
 Deuteranopia, 142 
 Dichromatism, 139 
 Difference, Smallest Percep- 
 tible, 4 
 Diplopia, 229 
 Direction, Sense of, 49 
 Donders' Law, 221 
 Dove, Vision of, 117 
 
 Ear, Anatomy of, 77. See also 
 
 Hearing 
 Epicritic Sensations, 28 
 Error, Mean, Method of, 4 
 Eustachian Tube, 80 
 
 237 
 
238 
 
 INDEX 
 
 Fechner's Law, 5 
 Fehlpunkt, 150, 153 
 "Flicker "Method, 103 
 
 Hearing, 64-85 
 
 Anaxagoras on, 65 
 
 Aristotle on, 66 
 
 Democritus on, 65 
 
 Empedocles on, 65 
 
 Plato on, 65 
 Horopter, 228 
 
 Illusions — 
 
 Of Position and Movement, 
 52, 55 
 
 Visual, 229 
 Image — 
 
 After-, 155 
 
 Satellite, 113 
 
 Secondary, 113 
 
 Tertiary, 113 
 Induction — 
 
 Simultaneous, 203 
 
 Successive, 155 
 
 Labyrinth — • 
 
 Ewald's Experiments, 58 
 Goltz, Theory of, 60 
 Mach-Breuer Theory, 60 
 
 Leech, Vision of, 92 
 
 Limen. See Threshold Values 
 
 Listing's Law, 221 
 
 Mach-Breuer Theory, 60 
 Meniere's Disease, 59 
 Meyer's Experiment, 204 
 Movement, Sense of, 54 
 Muller's Law, 2 
 Muscular Sense, 55 
 
 Neutral Points, 145 
 Night Blindness, 119 
 
 Olfactometer, 39 
 Otolithic Organs, 51 
 Owl, Vision of, 117 
 
 Pain — 
 
 Points, 19 
 
 Referred, 21 
 
 Visceral, 21 
 Parallax, Binocular, 231 
 Perception — 
 
 Of Pitch, Limiting Values, 71 
 
 Spatial, Berkeley on, 216, 217 
 Position, Sense of, 52 
 Protanopia, 142 
 Protapathic Sensation, 28 
 Purkinje's Phenomenon, 101 
 Purple, Visual, 98 
 
 Eetina— 
 
 Chemical Reaction of, 97 
 Electric Response of, 96 
 Phototropic Response of, 97 
 
 " Retuning," 158 
 
 Rod Vision, 118, 120 
 
 Sensations — 
 
 Auditory and Labyrinthine, 
 48-84 
 
 Cutaneous, 8-18, 24r-35, 70- 
 86 
 
 Gustatory, 34-39 
 
 Olfactory, 39-47 
 
 Visceral and Deep, 21, 29 
 
 Visual, 101-163, 203-209 
 Shape, Judgment of, 226 
 Smell- 
 Antagonistic Stimuli, 43 
 
 Biological Importance, 46 
 
 Examination of, 39 
 
 Memory and, 46 
 
 Paths of Afferent Impulses, 45 
 Snellen's Test, 222 
 Space — 
 
 Lucretius on, 215 
 
 Tactual, 218 
 
 Visual, 219-236 
 Spectrum — 
 
 Twilight Value of, 110 
 
 Use in Colour Testing, 128, 
 134, 141 
 
INDEX 
 
 239 
 
 Stereocomparator, 235 
 Stereoscope, 233 
 Stereotelemeter, 234 
 
 Taste — 
 
 Examination of, 36 
 
 Gasserian Ganglion and, 38 
 
 Greek Theories of, 35 
 
 Paths of Afferent Impulses, 
 38 
 Temperature Sensations — 
 
 Localisation, 14 
 
 Theories of, 14, 15, 26-33 
 Threshold Values, 3 
 Tones — 
 
 Beat, 76 
 
 Compound, 74 
 Touch, Sensations of — 
 
 Aristotle on, 8 
 
 Examination of, 9, 10 
 
 Localisation of, 11 
 
 Organs of, 10 
 Triangle, Colour, 131 
 Trichromatism, 133 
 
 Unlust, 19 
 
 Vertigo, 55 
 
 Vision {see also Contrast, Colour- 
 blindness, Space) — 
 
 Adaptation in, 118 
 
 Anaxagoras on, 169 
 
 Aristotle on, 171 
 
 Binocular, 228 
 
 Comparative Physiology of; 86- 
 95 
 
 Cone, 118, 120 
 
 Dichromatic, 139 
 
 Empedocles on, 167 
 
 Foveal and Peripheral. See 
 Adaptation 
 
 Bering's Theory of, 189-203 
 
 Monocular, 221 
 
 Plato on, 170 
 
 Recurrent, 112 
 
 Rod, 118, 120 
 
 Trichromatic, 124 
 
 Twilight, 110 
 
 Young-Helmholtz Theory of, 
 177-189 
 
 Weber's Law, 3 
 
 Yellow Sightedness, 137 
 
 THE END 
 
 Printed by Ballantyne, Hanson &' Co. 
 Edinbursb &* London