I. INTRODUCTION

A few of the experiments recorded in this monograph were carried out in the Animal Breeding Research Department of the University of Edinburgh, and thanks are due to the Director, Dr. F. A. E. Crew, for the hospitality of his department. Most of the experimental work, however, was done in the Department of Psychology, and most cordial thanks must be expressed to Dr. J. Drever for facilities afforded and for his kind interest and help. The subjects who kindly volunteered for the experiments, frequently at some personal inconvenience, are too numerous to mention individually, but their services are much appreciated. The work was further made possible by the tenure of a Carnegie Research Fellowship, and the financial aid given by the Carnegie Trust for the Universities of Scotland is most gratefully acknowledged. In 1924, the manuscript was accepted by the Faculty of Science (Edinburgh University) as a Thesis for the degree of Doctor of Philosophy.

Whilst studying a particular psychological aspect, it became necessary to obtain a view, even a very imperfect view, of the science of smell as a whole. Because of the abundance of references to chemical, physical, morphological, physiological, psychological, medical, and technological data concerning the science of smell in its widest sense, or osmics (l'osmique, osmica, Osmik) a convenient term-it became advisable to deal with these references in a special publication from time to time. The first number of "Osmics" appeared in 1922, the second in 1924, each containing 500 bibliographical items fully indexed, which are obtainable from the publishers, Messrs. Oliver & Boyd, Edinburgh, at cost price (2 shillings each). Some of the references are rearranged, with a few additions, at the end of the monograph.

Facilities for verifying references were kindly given by the library staffs of Edinburgh and Glasgow Universities, while special indebtedness must be recorded to Mr. Alfred Ela, of Boston, Mass., and others for having supplied important references which might otherwise have escaped notice.

II. SUBJECTS, ODORS, AND METHODS EMPLOYED

Sixty-three persons, of whom twenty-nine were male and thirty-four female, kindly volunteered to act as subjects for the experiments, and are designated by numbers, 1 to 29 for male and from 30 to 63 for female subjects. In recording reactions by the same individual on different dates a letter is placed after the number.

The majority were Scottish university students, between eighteen and twenty-four years of age, and all were European. In some cases special characteristics, such as hair and eye color, presence of nasal or pharyngeal catarrh, color blindness, and other conditions were noted. It may be mentioned in passing that the presence of eye defects such as myopia, astigmatism, etc., was ascertained in as many as twenty-three out of the sixtythree subjects.

The odorous substances (osmyls) employed, twenty-six in all, were as follows: Amyl acetate, amyl alcohol, asafoetida, benzole, butyric acid, camphor, carbon bisulphide, cassia oil, cedarwood oil, citronella oil, clove oil, ethyl alcohol, eucalyptus oil, lavender oil, menthol, musk, origanum oil, orris root, pine oil (ol. pin. sylv.), rose oil (otto of roses), sandalwood oil, terebene, Tonka bean, ferric valerian, vanillin, and xylol.

For various reasons not all the subjects could be tested with all the osmyls, nor were the same series of odorous substances invariably employed in any repeat experiment with the same subject. Moreover, some subjects were tested with very few osmyls only, and some osmyls were used as stimuli only with very few subjects, so that the comparison of reactions by presenting all the reactions in percentages is not expedient in many cases. The results summarized in the tables, therefore, include reactions given on different dates by the same subject, as well as the cases in which the subject could be tested on one occassion only. In the aggregate, however, the number of reactions given is relatively large-1,444 affective and 1,296 associative reactions

are tabulated—and bearing in mind the necessary reservations, it is possible to induce quite definite conclusions with regard to many points at issue.

The odorous substances were contained in small bottles or tubes, and the subjects were requested to smell the contents and to state the affect or association produced. To preclude reactions arising from visual stimuli they were further asked to close their eyes during each test, a necessary precaution indeed. Of course, no mention was made of the substances to be employed, and these were not administered in any particular order.

Owing to the phenomena of fatigue and after-smells a sufficient interval was allowed between successive stimulations. Although the stimuli were selected at random and a recuperative interval of at least two minutes duration was deemed sufficient, the adequacy of the precautions taken was tested by means of a table which clearly showed that there was no correlation whatever between the affect produced by the stimulus and the affect caused by the preceding osmyl. As was to be expected, however, the content of an association was not rarely influenced by the content of a previous association to a different stimulus in the same series of experiments, or even of associations given on a previous date.

In some cases a stop-watch was used to measure the time between the first inspiration of the stimulus and the associated word reaction, the subject holding his breath till the signal (touch or sound) was given to sniff. In a few instances the psychogalvanic reflex was observed by means of a D'Arsonval galvanometer.

Even at the risk of appearing trite, of restating obvious truisms, such as the fact that otto of roses generally produces a more pleasant sensation than does asafoetida, or that olfactory memories are often vivid, it was deemed best to undertake and interpret the experiments as far as is possible in a spirit unbiased by previous observations and conclusions, and to record the results irrespective of whether they constitute a discovery or merely a rediscovery.

III. THE AFFECTIVE OLFACTORY SYNDROME

By affect is meant here the pleasantness or unpleasantness experienced owing to an odorous stimulus. The pleasantness or unpleasantness may be due to the olfactory sensation itself or concomitant sensation (synaesthesia), or be due to a previous, associated experience or association content, or arise from a blending of past and present. Only in the case of unfamiliar odors, experienced for the first time in the course of the experiment, or where the subject was definitely aware of a difference between the direct and the associated affect, is it possible to refer the affect to the sensation experimentally produced. Even then, it is not a case of a 'pure' direct affect, since it would seem impossible to find an odor which is not somehow associated with a previous experience, of olfactory, gustatory, or other origin.

In most cases the affective judgment given by the subjects represents a blending of present and past feeling-tone. Nor is in all cases the affect static. An odor, e.g., of musk, may be voted pleasant at the first whiff and then become unpleasant, a phenomenon largely, but not entirely, due to intensity, and also to concomitant sensations, for instance due to stimulation of the endings of the fifth cranial nerve. The affects recorded here represent the final judgment by the subject as to whether the odor is pleasant or unpleasant, whatever elements form the basis for the judgment, so long as the subject refers to the odor itself.

The affects are here recorded in four categories, comprising two degrees of pleasantness, entirely pleasant (++) and not entirely pleasant (+-), and two degrees of unpleasantness, entirely unpleasant (——) and not entirely unpleasant (—+). Even if there was any balancing on the knife edge of neutrality, it was found that this balancing was transient and that in all instances the affective judgment was relegated to one of the four categories.