Popular Science Monthly/Volume 81/September 1912/A Consideration of the Nature of Hunger

by Walter Bradford Cannon

Layout 4

A CONSIDERATION OF THE NATURE OF HUNGER^[1]

By Professor W. B. CANNON

LABORATORY OF PHYSIOLOGY IN THE HARVARD MEDICAL SCHOOL

"WHY do we eat?" This question, presented to a group of educated people, is likely to bring forth the answer, "We eat to compensate for body waste, or to supply the body with fuel for its labors." Although the body is in fact losing weight continuously and drawing continuously on its store of energy, and although the body must periodically be supplied with fresh material and energy in order to keep a more or less even balance between the income and the outgo, this maintenance of weight and strength is not the motive for taking food.

Primitive man, and the lower animals, may be regarded as quite unacquainted with notions of the equilibrium of matter and energy in the body, and yet they take food and have an efficient existence, in spite of this ignorance. In nature, generally, important processes, such as the preservation of the individual and the continuance of the race, are not left to be determined by intellectual considerations, but are provided for in automatic devices. Natural desires and impulses arise in consciousness, driving us to action; and only by analysis do we learn their origin or divine their significance. Thus our primary reasons for eating are to be found, not in convictions about metabolism, but in the experiences of appetite and hunger.

 

Appetite and Hunger

The sensations of appetite and hunger are so complex and so intimately interrelated that any discussion is sure to go astray unless at the start there is clear understanding of the meanings of the terms. The view has been propounded that appetite is the first degree of hunger, the mild and pleasant stage, agreeable in character; and that hunger itself is a more advanced condition, disagreeable and even painful—the unpleasant result of not satisfying the appetite.^[2] On this basis appetite and hunger would differ only quantitatively. Another view, which seems more justifiable, is that the two experiences are fundamentally different.

Careful observation indicates that appetite is related to previous sensations of taste and smell of food. Delightful or disgusting tastes and odors, associated with this or that edible substance, determine the appetite. It has therefore important psychic elements in its composition, as the studies by Pawlow and his collaborators have so clearly shown. Thus, by taking thought, we can anticipate the odor of a delicious beefsteak or the taste of peaches and cream, and in that imagination we can find pleasure. In the realization, direct effects in the senses of taste and smell give still further delight. We now know from observations on experimental animals and on human beings, that the pleasures of both anticipation and realization, by stimulating the flow of saliva and gastric juice, play a highly significant rôle in the initiation of digestive processes.^[3]

Among prosperous people, supplied with abundance of food, the appetite seems sufficient to ensure for bodily needs a proper supply of nutriment. We eat because dinner is announced, because by eating we avoid unpleasant consequences, and because food is placed before us in delectable form and with tempting tastes and odors. Under less easy circumstances, however, the body needs are supplied through the much stronger and more insistent demands of hunger.

The sensation of hunger is difficult to describe, but almost every one from childhood has felt at times that dull ache or gnawing pain referred to the lower mid-chest region and the epigastrium, which may take imperious control of human actions. As Sternberg has pointed out, hunger may be sufficiently insistent to force the taking of food which is so distasteful that it not only fails to rouse appetite, but may even produce nausea. The hungry being gulps his food with a rush. The pleasures of appetite are not for him—he wants quantity rather than quality, and he wants it at once.

Hunger and appetite are, therefore, widely different—in physiological basis, in localization and in psychic elements. Hunger may be satisfied while the appetite still calls. Who is still hungry when the tempting dessert is served, and yet are there any who refuse it, pleading they no longer need it? On the other hand, appetite may be in abeyance while hunger is goading.^[4] What ravenous boy is critical of his food? Do we not all know that "hunger is the best sauce"? Although the two sensations may thus exist separately, they, nevertheless, have the same function of leading to the intake of food, and they usually appear together. Indeed the cooperation of hunger and appetite is prohably the reason for their being so frequently confused.

 

The Sensation of Hunger

In the present paper we shall deal only with hunger. The sensation may be described as having a central core and certain more or less variable accessories. The peculiar dull ache of hungriness, referred to the epigastrium, is usually the organism's first strong demand for food; and when the initial order is not obeyed, the sensation is likely to grow into a highly uncomfortable pang or gnawing, less definitely localized as it becomes more intense. This may be regarded as the essential feature of hunger. Besides the dull ache, however, lassitude and drowsiness may appear, or faintness, or violent headache, or irritability and restlessness such that continuous effort in ordinary affairs becomes increasingly difficult. That these states differ much with individuals—headache in one, and faintness in another, for example—indicates that they do not constitute the central fact of hunger, but are more or less inconstant accompaniments, and need not for the present engage our attention. The "feeling of emptiness," which has been mentioned as an important element of the experience,^[5] is an inference rather than a distinct datum of consciousness, and can likewise be eliminated from further consideration. The dull pressing sensation is left, therefore, as the constant characteristic, the central fact, to be examined in detail.

Hunger can evidently be regarded from the psychological point of view, and discussed solely on the basis of introspection; or it can be studied with reference to its antecedents and to the physiological conditions which accompany it—a consideration which requires the use of both objective methods and subjective observation. This psychophysiological treatment of the subject will be deferred till the last. Certain theories which have been advanced with regard to hunger and which have been given more or less credit must first be examined.

Two main theories have been advocated. The first is supported by evidence that hunger is a general sensation, arising at no special region of the body, but having a local reference. This theory has been more widely credited by physiologists and psychologists than the other. The other is supported by evidence that hunger has a local source and therefore a local reference. In the course of our examination of these views we shall have opportunity to consider some pertinent new observations.

 

The Theory that Hunger is a General Sensation

The conception that hunger arises from a general condition of the body rests in turn on the notion that, as the body uses up material, the blood becomes impoverished. Schiff advocated this notion, and suggested that poverty of the blood in food substance affects the tissues in such manner that they demand a new supply. The nerve cells of the brain share in this general shortage of provisions, and because of internal changes, give rise to the sensation.^[6] Thus is hunger explained as an experience dependent on the body as a whole.

Three classes of evidence are cited in support of this view.

1. "Hunger Increases as Time Passes"—a Partial Statement.—The development of hunger as time passes is a common observation which quite accords with the assumption that the condition of the body and the state of the blood are becoming constantly worse, so long as the need, once established, is not satisfied.

While it is true that with the lapse of time hunger increases as the supply of body nutriment decreases, this concomitance is not proof that the sensation arises directly from a serious encroachment on the store of food materials. If this argument were valid we should expect hunger to become more and more distressing until death. There is abundant evidence that the sensation is not thus intensified; on the contrary, during continued fasting hunger wholly disappears after the first few days. Luciani, who carefully recorded the experience of the faster Sucei, states that after a certain time the hunger feelings vanish and do not return.^[7] And he tells of two dogs that showed no signs of hunger after the third or fourth day of fasting; thereafter they remained quite passive in the presence of food. Tigerstedt, who also has studied the metabolism of starvation, declares that although the desire to eat is very great during the first day of the ordeal, the unpleasant sensations disappear early, and that at the end of the fast the subject may have to force himself to take nourishment.^[8] The subject, "J. A.," studied by Tigerstedt and his co-workers, reported that after the fourth day of fasting, he had no disagreeable feelings.^[9] Carrington, after examining many persons who, to better their health, abstained from eating for different periods, records that "habit-hunger" usually lasts only two or three days and, if plenty of water is drunk, does not last longer than three days.^[10] Viterbi, a Corsican lawyer, condemned to death for political causes, determined to escape execution by depriving his body of food and drink. During the eighteen days that he lived he kept careful notes. On the third day the sensation of hunger departed, and although thereafter thirst came and went, hunger never returned.^[11] Still further evidence of the same character could be cited, but enough has already been given to show that after the first few days of fasting the hunger feelings cease. On the theory that hunger is a manifestation of bodily need, are we to suppose that, in the course of starvation, the body is mysteriously not in need after the third day, and that therefore the sensation of hunger disappears? The absurdity of such a view is obvious.

2. "Hunger may be Felt though the Stomach be Full"—a Selected Alternative.—Instances of duodenal fistula in man have been carefully studied, which have shown that a modified sensation of hunger may be felt when the stomach is full. A famous case described by Busch has been repeatedly used as evidence. His patient, who lost nutriment through the fistula, was hungry soon after eating, and felt satisfied only when the chyme was restored to the intestine through the distal fistulous opening.^[12] As food is absorbed mainly through the intestinal wall, the inference is direct that the general bodily state, and not the local conditions of the alimentary canal, must account for the patient's feelings.

A full consideration of the evidence from cases of duodenal fistula can not so effectively be presented now as later. That in Busch' s case hunger disappeared while food was being taken is, as we shall see, quite significant. It may be that the restoration of chyme to the intestine quieted hunger, not because nutriment was thus introduced into the body, but because the presence of material altered the nature of intestinal activity. The basis for this suggestion will be given in due course.

3. "Animals may Eat Eagerly after Section of their Vagus and Splanchnic Nerves"—a Fallacious Argument.—The third support for the view that hunger has a general origin in the body is derived from observations on experimental animals. By severance of the vagus and splanchnic nerves, the lower œsophagus, the stomach and the small intestine can be wholly separated from the central nervous system. Animals thus operated upon nevertheless eat food placed before them, and may indeed manifest some eagerness for it.^[13] How is this behavior to be accounted for—when the possibility of local stimulation has been eliminated—save by assuming a central origin of the impulse to eat?

The fallacy of this evidence, though repeatedly overlooked, is easily shown. We have already seen that appetite as well as hunger may lead to the taking of food. Indeed, the animal with all gastrointestinal nerves cut may have the same incentive to eat that a well-fed man may have, who delights in the pleasurable taste and smell of food and knows nothing of hunger pangs. Even when the nerves of taste are cut, as in Longet's experiments,^[14] sensations of smell are still possible, as well as agreeable associations which can be roused by sight. More than fifty years ago Ludwig pointed out that, even if all the nerves were severed, psychic reasons could be given for the taking of food,^[15] and yet because animals eat after one or another set of nerves is eliminated, the conclusion has been drawn by various writers that the nerves in question are thereby proved to be not concerned in the sensation of hunger. Evidently, since hunger is not required for eating, the fact that an animal eats is no testimony whatever that the animal is hungry, and therefore, after nerves have been severed, is no proof that hunger is of central origin.

Weakness of the Assumptions Underlying the Theory that Hunger is a General Sensation.—The evidence thus far examined has been shown to afford only shaky support for the theory that hunger is a general sensation. The theory, furthermore, is weak in its fundamental assumptions. There is no clear indication, for example, that the blood undergoes, or has undergone, any marked change, chemical or physical, when the first stages of hunger appear. There is no evidence of any direct chemical stimulation of the gray matter of the cerebral cortex. Indeed, attempts to excite the gray matter artificially by chemical agents have been without results;^[16] and even electrical stimulation, which is effective, must, in order to produce movements, be so powerful that the movements have been attributed to excitation of underlying white matter rather than cells in the gray. This insensitivity of cortical cells to direct stimulation is not at all favorable to the notion that they are sentinels set to warn against too great diminution of bodily supplies.

Body Need may Exist without Hunger.—Still further evidence opposed to the theory that hunger results directly from the using up of organic stores is found in patients suffering from fever. Metabolism in fever patients is augmented, body substance is destroyed to such a degree that the weight of the patient may be greatly reduced, and yet the sensation of hunger under these conditions of increased need is wholly lacking.

Again, if a person is hungry and takes food, the sensation is suppressed soon afterwards, long before any considerable amount of nutriment could be digested and absorbed, and therefore long before the blood and the general bodily condition, if previously altered, could be restored to normal.

Furthermore, persons exposed to privation have testified that hunger can be temporarily suppressed by swallowing indigestible materials. Certainly scraps of leather and bits of moss, not to mention clay eaten by the Otomacs, would not materially compensate for large organic losses. In rebuttal to this argument the comment has been made that central states as a rule can be readily overwhelmed by peripheral stimulation, and just as sleep, for example, can be abolished by bathing the temples, so hunger can be abolished by irritating the gastric walls.^[17] That comment is beside the point, for it meets the issue by merely assuming as true the condition under discussion. The absence of hunger during the ravages of fever, and its quick abolition after food or even indigestible stuff is swallowed, still further weakens the argument, therefore, that the sensation arises directly from lack of nutriment in the body.

The Theory that Hunger is of General Origin does not Explain the Quick Onset and the Periodicity of the Sensation.—Many persons have noted that hunger has a sharp onset. A person may be tramping in the woods or working in the fields, where fixed attention is not demanded, and without premonition may feel the abrupt arrival of the characteristic ache. The expression "grub-struck" is a picturesque description of this experience. If this sudden arrival of the sensation corresponds to the general bodily state, the change in the general bodily state must occur with like suddenness or have a critical point at which the sensation is instantly precipitated. There is no evidence whatever that either of these conditions occurs in the course of metabolism.

Another peculiarity of hunger, which I have noticed in my own person, is its intermittancy. It may come and go several times in the course of a few hours. Furthermore, while the sensation is prevailing, its intensity is not uniform, but marked by ups and downs. In some instances the ups and downs change to a periodic presence and absence without change of rate. In making the above statements I do not depend on my own introspection alone; psychologists trained in this method of observation have reported that in their experience the temporal course of the sensation is distinctly intermittent.^[18] In my own experience the hunger pangs came and went on one occasion as follows:

Came	Went	Came	Went
12-37-20	38-30	43-20	43-35
40-45	41-10	44-40	45-55
41-45	42-25	46-15	46-30

and so on, for ten minutes longer. Again in this relation, the intermittent and periodic character of hunger would require, on the theory under examination, that the bodily supplies be intermittently and periodically insufficient. During one moment the absence of hunger would imply an abundance of nutriment in the organism, ten seconds later the presence of hunger would imply that the stores had been suddenly reduced, ten seconds later still the absence of hunger would imply a sudden renewal of plenty. Such zig-zag shifts of the general bodily state may not be impossible, but from all that is known of the course of metabolism, such quick changes are highly improbable. The periodicity of hunger, therefore, is further evidence against the theory that the sensation has a general basis in the body.

The Theory that Hunger is of General Origin does not Explain the Local Reference.—The last objection to this theory is that it does not account for the most common feature of hunger—namely, the reference of the sensation to the region of the stomach. Schiff and others who have supported the theory^[19] have met this objection by two contentions. First they have pointed out that the sensation is not always referred to the stomach. Schiff interrogated ignorant soldiers regarding the local reference; several indicated the neck or chest, twenty-three the sternum, four were uncertain of any region, and two only designated the stomach. In other words, the stomach region was most rarely mentioned.

The second contention against the importance of local reference is that such evidence is fallacious. An armless man may feel tinglings which seem to arise in fingers which have long since ceased to be a portion of his body. The fact that he experiences such tinglings and ascribes them to dissevered parts, does not prove that the sensation originates in those parts. And similarly the assignment of the ache of hunger to any special region of the body does not demonstrate that the ache arises from that region. Such are the arguments against a local origin of hunger.

Concerning these arguments we may recall, first, Schiff s admission that the soldiers he questioned were too few to give conclusive evidence. Further, the testimony of most of them that hunger seemed to originate in the chest or region of the sternum can not be claimed as unfavorable to a peripheral source of the sensation. The description of feelings which develop from disturbances within the body is almost always indefinite. As Head and others have shown, conditions in a viscus which give rise to sensation are likely not to be attributed to the viscus, but to related skin areas.^[20] Under such circumstances we do not dismiss the testimony as worthless merely because it may not point precisely to the source of the trouble. On the contrary, we use such testimony constantly as a basis for judging internal disorders.

With regard to the contention that reference to the periphery is not proof of the peripheral origin of a sensation, we may answer that the force of that contention depends on the amount of accessory evidence which is available. Thus if we see an object come into contact with a finger, we are justified in assuming that the simultaneous sensation of touch which we refer to that finger has resulted from the contact, and is not a purely central experience accidentally attributed to an outlying member. Similarly in the case of hunger—all that we need as support for the peripheral reference of the sensation is proof that conditions occur there, simultaneously with hunger pangs, which might reasonably be regarded as giving rise to those pangs.

 

Objections to Some Theories that Hunger is of Local Origin

With the requirement in mind that peripheral conditions be adequate, let us examine the state of the fasting stomach to see whether indeed conditions may be present in times of hunger which would sustain the theory that hunger has a local outlying source.

Hunger not Due to Emptiness of the Stomach.—Among the suggestions which have been offered to account for a peripheral origin of the sensation is that of attributing it to emptiness of the stomach. By use of the stomach tube Nicolai found that when his subjects had their first intimation of hunger the stomach was quite empty. But, in other instances, after lavage of the stomach, the sensation did not appear for intervals varying between one and a half and three and a half hours.^[21] During these intervals the stomach must have been empty, and yet no sensation was experienced. The same testimony was given long before by Beaumont, who, from his observations on Alexis St. Martin, declared that hunger arises some time after the stomach is normally evacuated.^[22] Mere emptiness of the organ, therefore, does not explain the phenomenon.

Hunger not Due to Hydrochloric Acid in the Empty Stomach.—A second theory, apparently suggested by observations on cases of hyperacidity, is that the ache or pang is due to hydrochloric acid secreted into the stomach while empty. Again the facts are hostile. Nicolai reported that the gastric wash-water from his hungry subjects was neutral or only slightly acid.^[23] This testimony confirms Beaumont's statement, and is in complete agreement with the results of gastric examination of fasting animals reported by numerous experimenters. There is no secretion into the empty stomach during the first days of starvation. Furthermore, persons suffering from absence of hydrochloric acid (achylia gastrica) declare that they have normal feelings of hunger. Hydrochloric acid can not therefore be called upon to account for the sensation.

Hunger not Due to Turgescence of the Gastric Mucosa.—Another theory, which was first advanced by Beaumont, is that hunger arises from turgescence of the gastric glands.^[24] The disappearance of the pangs as fasting continues has been accounted for by supposing that the gastric glands share in the general depletion of the body, and that thus the turgescence is relieved.^[25] This turgescence theory has commended itself to several recent writers. Thus Luciani has accepted it, and by adding the idea that nerves distributed to the mucosa are specially sensitive to deprivation of food he accounts for the hunger pangs.^[26] Also Valenti declared two years ago that the turgescence theory of Beaumont is the only one with a semblance of truth in it.^[27] The experimental work reported by these two investigators, however, does not necessarily sustain the turgescence theory. Luciani severed the previously exposed vagi after cocainizing them, and Valenti merely cocainized the nerves; the fasting dogs, eager to eat a few minutes previous to this operation, now ran about as before, but when offered food, licked and smelled it, but did not take it. This total neglect of the food lasted varying periods up to two hours. The vagus nerves seem, indeed, to convey impulses which affect the procedure of eating, but there is no clear evidence that those impulses arise from distention of the gland cells. The turgescence theory, moreover, does not explain the effect of taking indigestible material into the stomach. According to Pawlow, and to others who have observed human beings, the chewing and swallowing of unappetizing stuff does not cause any secretion of gastric juice.^[28] Yet such stuff when swallowed will cause the disappearance of hunger, and Nicolai found that the sensation could be abolished by simply introducing a stomach sound. It is highly improbable that the turgescence of the gastric glands can be reduced by either of these procedures. The turgescence theory, furthermore, does not explain the quick onset of hunger, or its intermittent and periodic character. That the cells are repeatedly swollen and contracted within periods a few seconds in duration is almost inconceivable. For these reasons, therefore, the theory that hunger results from turgescence of the gastric mucosa can reasonably be rejected.

 

Hunger the Result of Contractions

There remain to be considered, as a possible cause of hunger-pangs, contractions of the stomach and other parts of the alimentary canal. This suggestion is not new. Sixty-six years ago Weber declared his belief that "strong contraction of the muscle fibers of the wholly empty stomach, whereby its cavity disappears, makes a part of the sensation which we call hunger."^[29] Vierordt drew the same inference twentyfive years later (in 1871),^[30] and since then Ewald, Knapp, and Hertz have declared their adherence to this view. These writers have not brought forward any direct evidence for their conclusion, though Hertz has cited Boldireff's observations on fasting dogs as probably accounting for what he terms "the gastric constituent of the sensation."^[31]

The Empty Stomach, and Intestine Contract.—The argument commonly used against the gastric contraction theory is that the stomach is not energetically active when empty. Thus Schiff stated "the movements of the empty stomach are rare and much less energetic than during digestion."^[32] Luciani expressed his disbelief by asserting that gastric movements are much more active during gastric digestion than at other times, and cease almost entirely when the stomach has discharged its contents.^[33] And Valenti stated only year before last "we know very well that gastric movements are exaggerated while digestion is proceeding in the stomach, but when the organ is empty they are more rare and much less pronounced," and therefore they can not account for hunger.^[34]

Evidence opposed to these suppositions has been in existence for many years. In 1899, Bettmann called attention to the contracted condition of the stomach after several days' fast.^[35] In 1902, Wolff reported that after forty-eight hours without food the stomach of the cat may be so small as to look like a slightly enlarged duodenum.^[36] In a similar circumstance I have noticed the same extraordinary smallness of the organ, especially in the pyloric half. The anatomist His also recorded his observation of the phenomenon.^[37] Six years ago Boldireff demonstrated that the whole gastro-intestinal tract has a periodic activity while not digesting.^[38] Each period of activity lasts from 20 to 30 minutes, and is characterized in the stomach by rhythmic contractions 10 to 20 in number. These contractions, Boldireff reports, may be stronger than during digestion, and his published records clearly support this statement. The intervals of repose between periodic recurrences of the contractions lasted from one and a half to two and a half hours. Especially noteworthy is Boldireff's observation that if fasting is continued for two or three days, the groups of contractions appear at gradually longer intervals and last for gradually shorter periods, and thereupon, as the gastric glands begin continuous secretion, all movements cease.

Observations Suggesting a Relation Between Contractions and Hunger.—When Boldireff's paper was published I was studying auscultation of abdominal sounds. Repeatedly there was occasion to note that the sensation of hunger was, as already stated, not constant, but recurrent, and that its momentary disappearance was often associated with a rather loud gurgling sound, as heard through the stethoscope. That contractions of the alimentary canal on a gaseous content might explain the hunger pangs seemed probable at that time, especially in the light of Boldireff's observations. Indeed, Boldireff himself had considered hunger in relation to the activities he described, but solely with the idea that hunger might provoke them; and since the activities dwindled in force and frequency as time passed, whereas, in his belief they should have become more pronounced, he abandoned the notion of any relation between the phenomena.^[39] Did not Boldireff misinterpret his own observations? When he was considering whether hunger might cause the contractions, did he not overlook the possibility that the contractions might cause hunger? A number of experiences have led to the conviction that Boldireff did, indeed, fail to perceive part of the significance of his results. For example, I have noticed the disappearance of a hunger pang as gas was heard gurgling upward through the cardia. That the gas was rising rather than being forced downward was proved by its regurgitation immediately after the sound was heard. In all probability the pressure that forced the gas from the stomach was the cause of the preceding sensation of hunger. Again the sensation can be momentarily abolished a few seconds after swallowing a small accumulation of saliva or a teaspoonful of water. If the stomach is in strong contraction in hunger, this result can be accounted for as due to the inhibition of the contraction by swallowing.^[40] Thus also could be explained the prompt vanishing of the ache soon after we begin to eat, for repeated swallowing results in continued inhibition.^[41] Furthermore, Ducceschi's discovery that hydrochloric acid diminishes the tonus of the pyloric portion of the stomach^[42]-may have its application here; the acid would be secreted as food is taken and would then cause relaxation of the very region which is most strongly contracted.

The Concomitance of Contractions and Hunger in Man.—Although the evidence above outlined had led me to the conviction that hunger results from contractions of the alimentary canal, direct proof was still lacking. In order to learn whether such proof might be secured, one of my students, Mr. A. L. Washburn, determined to become accustomed to the presence of a rubber tube in the oesophagus.^[43] Almost every day for several weeks Mr. Washburn introduced as far as the stomach a small tube, to the lower end of which was attached a soft-rubber balloon about 8 cm. in diameter. The tube was thus carried about each time for two or three hours. After this preliminary experience the introduction of the tube, and its presence in the gullet and stomach, were not at all disturbing. When a record was to be taken, the balloon, placed just below the cardia, was moderately distended with air, and was connected with a water manometer ending in a cylindrical chamber 3.5 cm. wide. A float recorder resting on the water in the chamber permitted registering any contractions of the fundus of the stomach. On the days of observation Mr. Washburn would abstain from breakfast, or eat sparingly; and without taking any luncheon would appear in the laboratory about two o'clock. The recording apparatus was arranged as above described. In order to avoid the possibility of an artifact, a pneumograph, fastened below the ribs, was made to record the movements of the abdominal wall. Between the records of gastric pressure and abdominal movement, time was marked in minutes, and an electromagnetic signal traced a line which could be altered by pressing a key. All these recording arrangments were out of Mr. Washburn's sight; he sat with one hand at the key, ready whenever the sensation of hunger was experienced to make the current which moved the signal.

Sometimes the observations were started before any hunger was noted; at other times the sensation, after running a course, gave way to a feeling of fatigue. Under either of these circumstances there were no contractions of the stomach. When Mr. Washburn stated that he was hungry, however, powerful contractions of the stomach were invariably being registered. As in the experience of the psychologists, the sensations were characterized by periodic recurrences with free intervals, or by periodic accesses of an uninterrupted ache. The record of Mr. Washburn's introspection of his hunger pangs agreed closely with the record of his gastric contractions. Almost invariably, however, the contraction nearly reached its maximum before the record of the sensation was started (see Fig. 1). This fact may be regarded as evidence

Fig. 1. One half the original size. The top record represents intragastric pressure (the small oscillations due to respiration, the large to contractions of the stomach); the second record is time in minutes (ten minutes); the third record ia Mr. W.'s report of hunger pangs; the lowest record is respiration registered by means of a pneumograph about the abdomen.

that the contraction precedes the sensation, and not vice versa, as Boldireff considered it. The contractions were about a half-minute in duration, and the intervals between varied from thirty to ninety seconds, with an average of about one minute. The augmentations of intragastric pressure in Mr. Washburn ranged between 11 and 13 in twenty minutes; I had previously counted in myself eleven hunger pangs in the same time. The rate in each of us was, therefore, approximately the same. This rate is slightly slower than that found in dogs by Boldireff; the difference is perhaps correlated with the slower rhythm of gastric peristalsis in man compared with that in the dog.^[44]

Before hunger was experienced by Mr. Washburn the recording apparatus revealed no signs of gastric activity. Sometimes a rather tedious period of waiting had to be endured before contractions occurred. And after they began they continued for a while, then ceased (see Fig. 2). The feeling of hunger, which was reported while the contractions were recurring, disappeared as the waves stopped. The inability of the subject to control the contractions eliminated the possibility of their being artifacts, perhaps induced by suggestion. The close concomitance of the contractions with hunger pangs, therefore, clearly indicates that they are the real source of those pangs.

Boldireff's studies proved that when the empty stomach is manifesting periodic contractions, the intestines also are active. Conceivably all parts of the alimentary canal composed of smooth muscle share

Fig. 2. One half the original size. The same conditions as in Fig. 1. (Fifteen minutes.) There was a long wait for hunger to disappear. After x, Mr. W. reported himself "tired but not hungry." The record from y to z was the continuance on a second drum of x to y.

in these movements. The lower œsophagus in man is provided with smooth muscle. It was possible to determine whether this region in Mr. Washburn was active during hunger.

To the œsophageal tube a thin-rubber finger-cot (2 cm. in length) was attached and lowered into the stomach. The little rubber bag was,

Fig. 3. One half the original size. The top record represents compression of a thin rubber bag in the lower œsophagus. The pressure in the bag varied between ft and 13 cm. of water. The cylinder of the recorder was of smaller diameter than that used in the gastric records. The œsophageal contractions compressed the bag so completely that, at the summits of the large oscillations, the respirations were not registered. When the oscillations dropped to the time line, the bag was about half inflated. The middle line registers time in minutes (ten minutes). The bottom record is Mr. W.'s report of hunger pangs.

distended with air, and the tube, pinched to keep the bag inflated, was gently withdrawn until resistance was felt. The air was now released from the bag, and the tube further withdrawn about 3 cm. The bag was again distended with air at a manometric pressure of 10 cm. of water. Inspiration now caused the writing lever, which recorded the pressure changes, to rise; and a slightly further withdrawal of the tube changed the rise, on inspiration, to a fall. The former position of the tube, therefore, was above the gastric cavity and below the diaphragm. In this position the bag, attached to a float-recorder (with chamber 2.3 cm. in diameter), registered the periodic oscillations shown in Fig. 3. Though individually more prolonged than those of the stomach, these contractions, it will be noted, occur at about the same rate. It is probable that the periodic activity of the two regions is simultaneous, for otherwise the stomach would force its gaseous content into the oesophagus with the rise of intragastric pressure.

What causes the contractions to occur has not been determined. From evidence already given they do not seem to be directly related to bodily need. Habit no doubt plays an important rôle. For present considerations, however, it is enough that they do occur, and that they are abolished when food, which satisfies bodily need, is taken into the stomach. By such indirection, as already stated, are performed some of the most fundamental of the bodily functions.

Peculiarities of Hunger Explained by Contractions.—If these contractions are admitted as the cause of hunger, most of the difficulties confronting other explanations are readily obviated. Thus the occurrence of hunger at meal times is most natural, for, as the regularity of defecation indicates, the alimentary canal has habits. Activity returns at the usual meal time as the result of custom. By taking food regularly at a definite hour in the evening for several days, a new hunger period can be established. Since at these times the œsophagus and the empty stomach strongly contract, hunger is aroused.

The contractions furthermore explain the sudden onset of hunger and its peculiar periodicity—phenomena which no other explanation of hunger can account for. The quick development of the sensation after taking a cold drink is possibly associated with the well-known power of cold to induce contraction in smooth muscle.

The great intensity of hunger during the first day of starvation, and its gradual disappearance till it vanishes on the third or fourth day, are made quite clear, for Boldireff observed that the gastric contractions in his fasting dogs went through precisely such alterations of intensity, and were not seen after the third day.

In fever, when bodily material is being most rapidly used, hunger is absent. Its absence is understood from an observation reported four years ago, that infection, with systemic involvement, is accompanied by a total cessation of all movements of the alimentary canal.^[45] Boldireff observed that when his dogs were fatigued the rhythmic contractions failed to appear. Being "too tired to eat" is thereby given a rational explanation.

Another pathological form of the sensation—the inordinate hunger (bulimia) of certain neurotics—is in accordance with the well-known disturbances of the tonic innervation of the alimentary canal in such individuals.

Since the lower end of the œsophagus, as well as the stomach, contracts periodically in hunger, the reference of the sensation to the sternum by the ignorant persons questioned by Schiff was wholly natural. The activity of the lower œsophagus also explains why, after the stomach has been removed, or in some cases when the stomach is distended with food, hunger can still be experienced. Conceivably the intestines also originate vague sensations by their contractions. Indeed the final banishment of the modified hunger sensation in the patient with duodenal fistula, described by Busch, may have been due to the lessened activity of the intestines when chyme was injected into them.

The observations recorded in this paper have, as already noted, numerous points of similarity to Boldireff's observations on the periodic activity of the alimentary canal in fasting dogs. Each period of activity, he found, comprised not only wide-spread contractions of the digestive canal, but also the pouring out of bile, and of pancreatic and intestinal juices rich in ferments. Gastric juice was not secreted at these times; when it was secreted and reached the intestine, the periodic activity ceased.^[46] What is the significance of this extensive disturbance? Recently evidence has been presented that gastric peristalsis is dependent on the stretching of gastric muscle when tonically contracted.^[47] The evidence that the stomach is in fact strongly contracted in hunger—i. e., in a state of high tone—has been presented above.^[48] Thus the very condition which causes hunger and leads to the taking of food is the condition, when the swallowed food stretches the shortened muscles, for immediate starting of gastric peristalsis. In this connection the recent observations of Haudek and Stigler are probably significant. They found that the stomach discharges its contents more rapidly if food is eaten in hunger than if not so eaten.^[49] Hunger, in other words, is normally the signal that the stomach is contracted for action; the unpleasantness of hunger leads to eating; eating starts gastric secretion, distends the contracted organ, initiates the movements of gastric digestion, and abolishes the sensation. Meanwhile pancreatic and intestinal juices, as well as bile, have been prepared in the duodenum to receive the oncoming chyme. The periodic activity of the alimentary canal in fasting, therefore, is not solely the source of hunger pangs, but is at the same time an exhibition in the digestive organs of readiness for prompt attack on the food swallowed by the hungry animal.

1. Presented to the Harvey Society, New York City, December 16, 1911. The results here stated were published in the American Journal of Physiology, 1912, XXIX., pp. 41-454.
2. Bardier, Richet's "Dictionnaire de Physiologie," article "Faim," 1904, VI., p. 1. See, also, Howell, "Text-book of Physiology," fourth edition, Philadelphia and London, 1911, p. 285.
3. Pawlow, "The Work of the Digestive Glands," London, 1902, pp. 50, 71.
4. See Sternberg, Zentralblatt für Physiologie, 1909, XXII., p. 653. Similar views were expressed by Bayle in a thesis presented to the Faculty of Medicine in Paris in 1816.
5. See Hertz, "The Sensibility of the Alimentary Canal," London, 1911, p. 38.
6. Schiff, "Physiologie de la digestion," Florence and Turin, 1867, p. 40.
7. Luciani, "Das Hungern," Hamburg and Leipzig, 1890, p. 113.
8. Tigerstedt, Nagel's "Handbuch der Physiologie," Berlin, 1909, I., p. 376.
9. Johanson, Landergren, Sonden and Tigerstedt, Scandinavisches Archiv für Physiologie, 1897, VII., p. 33.
10. Carrington, "Vitality, Fasting and Nutrition," New York, 1908, p. 555.
11. Viterbi, quoted by Bardier, loc. cit., p. 7.
12. Busch, Archiv fur pathologische Anatomie und Physiologie und für klinische Medicin, 1858, XIV., p. 147.
13. See Schiff, loc. cit., p. 37; also Duceeschi, Archivio di fisiologia, 1910, VIII., p. 579.
14. Longet, "Traité de physiologie," Paris, 1868, L, p. 23.
15. Ludwig, "Lehrbuch der Physiologie des Menschen," Leipzig and Heidelberg, 1858, II., p. 584.
16. Maxwell, Journal of Biological Chemistry, 1906-7, II., p. 194.
17. See Schiff, loc. cit., p. 49.
18. I am indebted to Professor J. W. Baird, of Clark University, and his collaborators, for this corroborative testimony.
19. See Schiff, loc. cit., p. 31; Bardier, loc. cit., p. 16.
20. Head, Brain, 1893, XVI., p. 1; 1901, XXIV., p. 345.
21. Nicolai, "Ueber die Entstehung des Hungergefühls," Inaugural-Dissertation, Berlin, 1892, p. 17.
22. Beaumont, "The Physiology of Digestion," second edition, Burlington, 1847, p. 51.
23. Nicolai, loc. cit., p. 15.
24. Beaumont, loc. cit., p. 55.
25. A better explanation perhaps is afforded by Boldireff's discovery that at the end of two or three days the stomachs of fasting dogs begin to secrete gastric juice and continue the secretion indefinitely. (Boldireff, Archives biologiques de St. Petersburg, 1905, XI., p. 98.)
26. Luciani, Archivio di fisiologia, 1906, III., p. 54. Tiedemann long ago suggested that gastric nerves become increasingly sensitive as fasting progresses. ("Physiologie des Menschen," Darmstadt, 1836, III., p. 22.)
27. Valenti, Archives italiennes de biologie, 1910, LIII., p. 94.
28. Pawlow, loc. cit., p. 70; Hornborg, Skandinavisches Archiv für Physiologie, 1904, XV., p. 248.
29. Weber, Wagner's "Handwörterbuch der Physiologie," 1846, III²., p. 580.
30. Vierordt, "Grundriss der Physiologie," Tübingen, 1871, p. 433.
31. Knapp, American Medicine, 1905, X., p. 358; Hertz, loc. cit., p. 37.
32. Schiff, loc. cit., p. 33.
33. Luciani, loc. cit., p. 542.
34. Valenti, loc. cit., p. 95.
35. Bettmann, Philadelphia Monthly Medical Journal, 1899, I., p. 133.
36. Wolff, Dissertation, Giessen, 1902, p. 9.
37. His, Archiv für Anatomie, 1903, p. 345.
38. Boldireff, loc. cit., p. 1.
39. Boldireff, loc. cit., p. 96.
40. See Cannon and Lieb, American Journal of Physiology, 1911, XXIX., p. 267.
41. The absence of hunger in Busch 's patient while food was being eaten (see p. 295) can also be accounted for in this manner.
42. Ducceschi, Archivio per le Scienze Mediche, 1897, XXI., p. 154.
43. Nicolai (loc. cit.) reported that although the introduction of a stomach tube at first abolished hunger in his subjects, with repeated use the effects became insignificant.
44. See Cannon, American Journal of Physiology, 1903, VIII., p. xxi; 1905, XIV., p. 344.
45. Cannon and Murphy, Journal of the American Medical Association, 1907, XLIX., p. 840.
46. Boldireff, loc. cit., pp. 108-111.
47. Cannon, this journal, 1911, XXIX., p. 250.
48. The "empty" stomach and œsophagus contain gas (see Hertz, Quarterly Journal of Medicine, 1910, III., p. 378; Mikulicz, "Mittheilungen aus dem Grenzgebieten der Medicin und Chirurgie," 1903, XII., p. 596). They would naturally manifest rhythmic contractions on shortening tonically on their content.
49. Haudek and Stigler, Archiv für die gesammte Physiologie, 1910, CXXXIII., p. 159.