In all, 103 Ss were given the series of buildings, landscapes, and colors. Over half of the tests were given by the writer, and the remainder by an assistant experienced in psychometric methods. The group was a rather highly selected one. It comprised members of the professional class or students, salesmen, and clerical workers or executives, with a few semi-skilled and unskilled workers. The amount of education and training in aesthetic appreciation must necessarily have had an influence on the choice reactions.

The following table, Table I, shows the distribution of ages according to decades:

Between the ages of 20 and 30 there were 5 more men than women, but in all other decades women predominated.

Distribution curves of the reaction-times for the total group and for the men and women separately were made, which, however, were largely skewed. Scores were widely scattered, possibly because of too small a group or too selected a group. Then distribution curves were made according to decades, and only in the decade from 20-30, which was also that comprising the largest homogeneous group, was there anything resembling a Gaussian curve. Median times and the mean deviations are shown in Table II.

In general the total reaction-times tend to decrease with each group of stimuli, possibly from increasing familiarity with the task. The sepaFor help in the development and administration of the test, the writer is indebted to Mrs. Esther C. Whitman.

rate reaction-times for the men and women respectively show a slight exception to this, the men taking a somewhat longer time for landscapes than for buildings, and with no decrease in time between the blue and green colors. The reaction-times of the women were consistently lower with each group except for the last stimulus, shades of red, whose time was somewhat longer than that for the preceding stimulus, yellow. There was, from a superficial point of view, more difference between men and women on the reactions to red than any other color. Subjective interpretation given by certain individuals seemed to point to the fact that the men experienced an immediate sensation of pleasure from the standards of comparison, with a subsequent quick relegation of the remaining shades to the pile designated as worse.

Because of the wide range of reaction-times, the preponderance of older women, and the greater variations which appeared among the women, the median time by decades of age was worked out as shown in Table III.

TABLE III

Distribution of Median Times and Mean Deviations
according to Age Groups

Age Groups

Cards

20-30

30-40

40-50

M.T. M.D. Buildings 29.3 17.0 Landscapes 29.4 12.3

M.T. M.D. M.T. M.D.

50-60 M.T. M.D.

27.5
24.3

13.1 60.0
13.2 58.3

28.0

22.7

Green

12.5 6.2

Blue

II.4

5.6

12.5 13.3

Yellow

12.0

Red

11.3

12.7 43.8 12.2 45.0 7.0 17.5 7.1 17.5 6.0 14.4 8.3 II. I 4.2 12.5 8.7 15.0 4.6 20.0 6.7 12.8 5.7 15.0 II.2 20.0 13.9

As one can see readily, the times appear to increase fairly consistently with the older Ss. However, with the small number of individuals tested and the wide range of times, this cannot be regarded as particularly significant.

The percentage of preference for every individual stimulus was determined, but with little appreciable difference in affectivity between the men and women.

That the men gave shorter reaction-times than the women is contrary to Wells's findings. Possibly the extremely long times recorded of women in the later decades, which tended to raise the median, may account in part for this. Also in the experiment of Wells the number of Ss was smaller by far.

Distribution curves were also made on the basis of educational achievement, dividing the Ss into three groups-college, high school, and grammar school-and including within each group those Ss who had education equivalent to the standard named. There seemed quite a definite differOp. cit., 76.

ence in the reaction-times among the groups. The high school group showed considerably longer times than the college group, and the grammar school group longer times than the high school group. Here again, however, the disproportion of numbers should be considered. Sixty-two of the individuals tested were in the college group, 32 in the high school group, and only 9 in the grammar school group. The median times and median deviations are shown in Table IV.

TABLE IV

Distribution of Median Times and Mean Deviations According to Educational Achievement

Here again the mean deviations are so great as to make none but the most superficial conclusions justifiable. The difference in times suggested by the table was quite reasonably to be anticipated, for the degree of cultural education has in most cases attuned the individual to a certain degree of aesthetic appreciation and the emotional set is ready for response.

The question arose whether or not the presentation of familiar buildings did not carry some emotional value for different individuals. For this reason pictures of churches had been eliminated. The introspective comments of various persons are interesting as relevant to this point. One young woman who quite obviously was about to choose the picture of a Y. M. C. A. building as worse, suddenly changed her mind, saying, “Oh, that should be counted as better. Think of all the good it does." Again, it was noticed that people invariably designated the State House as better, although the view shown was not particularly attractive, and several times the remark was made that the State House was a beautiful building. It seemed probable that the Ss were supplementing the picture presented with their own memories of the actual structure. Among the landscapes there was a post-card showing a glimpse of two church spires among the trees, and on looking at this, one individual observed, "I always think churches are pretty." It would seem a better plan, in an experiment of this sort, to present pictures of unfamiliar buildings, and also with the titles eliminated. This would, to a certain extent, limit the opportunity to weight the stimulus with emotional values.

This article has assumed the theory that sensitivity is expressed in shorter time of response. That may be a dangerous premise, and it may be that the more sensitive individual takes longer to give a decision than the one whose aesthetic appreciations are less delicate. Wells observed in the small number of cases used in his experiment that shorter reactiontimes were shown by the women, commenting that this result might be connected with the greater affective sensibility recognized to exist among women, whereby the pictures would make a greater affective impression on them. But throughout the study described here, in practically every group the men showed on the average shorter times than the women. This can be regarded as arguing against the theory that higher sensitivity is expressed in shortened times of responses, or as indicating that for the men the pictures presented greater affective differences. Or it may be concluded that neither statement is quite justified, because of the mean deviation, which is so wide as to invalidate any positive inference.

AUDITORY THEORY: A CRITICISM OF PROFESSOR
BORING'S HYPOTHESES

By H. BANISTER, Cambridge University, England

In an interesting paper entitled "Auditory theory with special reference to intensity, volume and localization" Professor Boring has endeavoured to "bring the known facts [of auditory theory] into reasonable relation, to point out clearly the lacunae in our knowledge and the discrepances that remain in even the best interpretation, and to formulate the crucial experiments that might resolve these discrepances.' He considers that the evidence points to a non-resonance theory of hearing (p.158), and he suggests, further, that a non-resonance theory is easily to be understood if we assume "some kind of integration in the brain where multiplicity of impulses is translated back into degree of excitation" (p.161), and if we accept Köhler's theory of the electrical brain field.

The paper calls for criticism on two grounds, (a) because very doubtful physics is made the basis for wide hypotheses; and (b) because the claim that the resonance theory fails to explain the experimental facts is not fully substantiated.

The non-acceptance of the resonance theory forces the explanation of auditory phenomena back to more central processes. This does not, however, appear to Boring merely another way of confessing complete ignorance. Apparently, if we assume that cortical excitation establishes electrostatic fields, all is clear. I for one am absolutely incapable of accepting such reasoning. It seems to me that with our present knowledge of the functioning of the cortex such explanations are bad physiology and worse psychology.

Boring considers the resonance theory inadequate because (p.158) he does not see how it can explain graduations in intensity: also, in his opinion, the non-resonance theory promises a solution of the 'primary problem' of localisation; the resonance theory, apparently, offers no solution. Hartridge' has suggested an explanation of the facts of intensity, while I have shown how the facts of auditory localisation can be explained, on the basis of present-day physiological knowledge, in accordance with a resonance theory.

I do not propose, in this brief communication, to criticise Boring's argument in detail, but shall content myself with pointing out the difficulties in the way of the acceptance of his views. I will begin with those difficulties he himself acknowledges, and then I shall go on to others he has not mentioned.

The first class of difficulties-those which Boring points out-includes (a) Ohm's Law; (b) the length of the refractory period of the auditory nerve fibres; and (c) tonal lacunae.

(a) As regards Ohm's Law, Boring confesses (p.177) that "the hearing by an observer, presented with a complex periodic stimulus, of the tones that correspond approximately to the simple harmonic components of the stimulus as given by Fourier's analysis" is the "great stumbling block to a frequency i.e. non-resonance] theory of hearing.' He admits that the law

can be explained immediately on a resonance theory.

(b) Boring notes that his theory requires that the refractory period be very low, and he thinks "we may assume that there is a reasonable possi

This JOURNAL, 37, 1926, 157-188.

H. Hartridge, A vindication of the resonance hypothesis of audition, Brit. J. Psychol., 12, 1922, 362-382.

H. Banister, ibid., 17, 1926, 142-153.

bility that the refractory period of the human acoustic fibres is as low as 0.050" (p.181). This value is about 1/40 the usual estimation. I believe no physiologist would be willing to make the assumption.

(c) Tonal lacunae present a serious obstacle also. Boring surmounts it by suggesting that "recent research, however, indicates that there really are no tonal lacunae in the strict sense. Sensitivity for certain regions of frequency may become diminished, but even here the tones can be heard if they are intense enough" (p.185). It may be a fact that individuals who are normally deaf to certain tones can hear them if the tones are made sufficiently intense, but who is able to say that these intense tones do not affect other fibres of the basilar membrane? It is an experimental fact that even if the fundamental frequency of a complex tone is completely eliminated the tone is heard. It is identified by its over-tones. May not this be the explanation of the fact quoted by Boring? The deduction, that "there really are no tonal lacunae in the strict sense" is scarcely justified.

In this connection some experiments performed by Andréev, in Pavlov's laboratory, are of interest. Five dogs were taken and the left cochlea of each was extirpated. Two of the dogs died almost immediately as a result of the operation. Conditioned reflexes, to tones varying from 100 d.v./sec. to 11,600 d.v./sec., were then established in the remaining three dogs; after which the bone towards the upper end of the right cochlea was opened and the membranous portion of the organ of Corti was injured by means of a thin neddle. On recovery from the second operation the reflexes gradually reappeared. Full details regarding one of these dogs are given. On the fifth day high notes, e.g. 8,000 and 14,000 d.v./sec., gave positive results, and gradually responses to lower notes appeared. On the 114th day a tone of 580 d.v./sec. was effective, and 28 days later responses were made to a tone of 360 d.v./sec. Numerous experiments have been made since then, but, so far, 3 years after the second operation, no response has been obtained to a tone below 317 d.v./sec. The second dog is found to be deaf to tones in the middle of the scale. No mention is made in the paper of the third dog. One awaits with interest the results of the histological examination of the cochleae, which will show exactly what portions of the basilar membrane have been damaged.

The second class of difficulties—those not noted by Boring-have to do largely with the facts of localisation. These facts are not so simple as Boring would appear to assume. He points out (p.164) that there are three theories of localisation: the intensity theory, the time theory, and the phase theory. The last theory may be considered as a special case of the second, but the intensity theory, "that localization is towards the side of the ear in which the intensity of the sound is greater" (p.165), is not strictly true, in spite of its popularity among psychologists. If the intensities at the two ears are not equal an observer may judge that a tone of low pitch is localised towards the side at which the intensity is the greater, but, with more careful observation, two sounds will, as a rule, be discriminated, vide the experiments of Myers and Wilson, Stewart,' and myself. Boring's statement (p.166): "an intensive difference can be equivalent to a time difference and the two can cancel each other in effect," in support of which Stewart's results' are quoted, quite overlooks the fact that Stewart found

4-Andréev, Pavlov's Jubilee Volume, 1925, 339-362.

"Personal communication by Dr. Anrep who is in close touch with Pavlov's laboratory. H. A. Wilson and C. S. Myers, The influence of binaural phase differences on the localisation of sounds, Brit. J. Psychol., 2, 1908, 370 f.

7G. W. Stewart, The function of intensity and phase in binaural location of pure tones, Phys. Rev., 15, 1920, 427 f.

$H. Banister, Three experiments on the localisation of tones, Brit. J. Psychol., 16, 1926, 266-279.

Stewart, op. cit., 438