method is by far the most accurate. In Gundobin's clinic the rectal method is strictly forbidden because he considers that the danger of breaking the thermometer is great. Others object because of the ease of carrying infection and the difficulty of sterilizing the thermometer. It has been recommended to introduce the thermometer into the rectum so that the end entirely disappears, i. e., about three centimeters, holding it there. for five minutes. A reading is made and the themometer is pushed in two centimeters deeper. The difference between these two readings is more marked when the body temperature is low than when the child suffers from fever. In newly born infants it is well to introduce the thermometer into the rectum to a uniform depth of three to five centimeters. If it be introduced further the thermometer shows a proportionately higher temperature. With a good standard fever thermometer the temperature can be obtained at five centimeters in one minute. It appears that a number of factors must be considered in judging the accuracy of rectal thermometry. Some of the factors which influence the accuracy of the readings are the age of the patient, sex, condition of bodily nutrition, relaxation of the anal sphincter, and presence of feces in the rectum. When taken with an accurate, though not a self-registering, chemical thermometer, the rectal temperature obtained at seven to fourteen centimeters was found to be 0.2°-1.3°higher than a reading made at two to six centimeters. This is not only true of patients with fever but also those without fever and in normal health. Where the bowel is filled with feces, the deep and more superficial temperatures are more nearly equal. These facts were determined in children of different ages, and it was found that the same variations occurred in the six-day-old child as in the seven-month or eighteen-month-old child. It was thought possible that if the thermometer were allowed to lie in the superficial parts of the rectum for a longer time, it might show a higher temperature, but the difference between the superficial and deep temperatures was constant no matter how long the thermometer remained in this superficial position.

The axillary temperature, of course, is lower than that of the rectum or vagina. Occasionally one observes in premature infants that the axillary temperature is higher than that of the

rectum. This applies chiefly to children kept in incubators. It is explained by the fact that the surface of the body under these conditions is warmer than the internal temperature because of the artificial heat. Wunderlich noted that in the vagina and in the feces-free rectum the readings were 0.2°-0.7° higher than those of the axilla. These observations were at various times corroborated. One observer used a perfectly accurate bulb thermometer, and the axilla was tightly closed down upon the bulb. By allowing it to remain one or two minutes after it had reached its highest point (eight to ten minutes) he found a variation from the rectal temperature of 0.9° to 1.4° though in weaklings it was only 0.4°. The difference in these temperatures would probably be more marked in fever. It is obvious, however, that the rectal temperature is the more accurate of the two and is the method most frequently employed by clinicians of experiFilatoff suggests a method of taking the axillary temperature which is unique. He claims for it the advantage that it is more rapid. He uses a chemical, not a self-registering, thermometer and by rubbing or closely applying the thumb and forefinger to the bulb he raises it to 104°. He quickly introduces it into the axilla. In a minute or two it has fallen and records the axillary temperature. He maintains that the reading obtained by this method is 0.2°-0.4° less than the rectal reading.

ence.

The heat regulating mechanism of newly born. infants is incompletely developed and unstable. Marked temperature variations occur even in health. The temperature curve shows greater daily variations in younger infants than in older children and adults. Transitory elevations of temperature frequently occur which seem to depend upon bodily movements, taking of food, and the external temperature. Similarly, sudden sinking of temperature may depend upon sudden chilling, cool baths or insufficient clothing. It seems certain proof that a baby is delicate and undeveloped when its temperature is low or considerably under normal. Babies who are receiving insufficient food as well as those receiving insufficient fluids show a loss in temperature. It may be noted that a baby may show a temperature immediately after nursing which is several tenths higher than the interval between feedings. On the other hand, infections may be present which produce no elevation in temperature and, indeed,

may be characterized by a subnormal temperature. Naturally very little is known of the fetus in utero. Still, some facts have been developed which throw light on heat production in the fetus. Any knowledge that might be obtained about the temperature of the fetus in utero will throw light on tissue metabolism in the unborn infant.

In experimental animals the temperature of the fetus falls and rises with that of the mother. In the human, increased rapidity of the fetal heart tones probably points to febrile reaction in the mother. Clinical observations have taught that the fetus will die when the maternal temperature reaches 107.6° and sometimes at 104°. It is difficult for the fetus in utero to lose heat. Consequently it is to be expected that its temperature will be higher than that of the maternal organism. It is scarcely conceivable that the fetus in utero has a well-developed heat regulating mechanism if it possesses one at all. Since it is protected from outside changes it has no demand made upon it by its environment. We would be interested to know whether the fetus itself has the power to produce heat. An opportunity presents itself in breech presentations to obtain the rectal temperature of the infant and the vaginal temperature of the mother. almost invariably found that the temperature of the fetus is higher than that of the mother. The new born mature infant shows a slightly higher temperature than that of the mother immediately after its expulsion. It has been found that there may be a difference of 0.3°-0.5° between the temperature of twins. It has also been observed that well-developed infants have a slightly higher temperature than feeble infants directly after birth. These observations indicate that the fetus in utero receives heat from the mother but may be to some extent a producer of heat itself.

It is

The mature baby's temperature immediately after birth is usually 99.7°-100.6°. Shortly after birth the temperature falls, particularly in the first hour, and in two or three hours is 2.7° -3.6°F. lower than at birth. This is, then, the first occasion when the heat regulating mechanism is called into play. The diminution in temperature depends upon the atmospheric conditions and the protection of the baby by clothes and blankets. The less protected the baby or the cooler the room, the lower the temperature sinks. The first

bath leads to a loss of about one degree. This loss in bodily heat is characteristic of the first few days of life. After the minimum is reached (on the first or second day) the temperature gradually rises more slowly, however, than the initial decline. After the second day the temperature begins to approach normal, but it requires eight or ten days before the heat regulating function is of the same constant kind as in older children. In more delicate babies the rise does not occur until the third day.

The daily variations of the temperature in the first ten days of life have been made the object. of study by numerous observers. Forster, Sommers, Jürgensen, Wolff, and Gundobin state that the highest temperature occurs about noon and the minimum at various times during the night. On the contrary, Eross and Feiss found the maximum at night and the minimum at noon. To further confuse the situation Jundell decides that some children reach their high point early in the morning and some in the late afternoon. To sum up the observations, Muhlmann concludes that it is not possible to establish any rule for twenty-four hour variations. He emphasizes the influence of frequent food taking, crying, restlessness, motion, sleep, and the external factors such as the room temperature and the infant's clothing.

Undoubtedly the most marked variations occur in delicate and premature babies. The temperature changes are markedly exaggerated in the premature infant. The temperature is subnormal shortly after birth. It has been known to fall to 86° or lower. We have long since learned that in order to maintain the life of the premature infant it is necessary to guard against chilling and to supply external heat in the form of an incubator or a warm room. Further proof of the instability of the heat regulating mechanism is furnished by the ease of the development of hyperthermia. If the room is warm the temperature may rise to 100° or 102° or even higher. This is particularly true, however, of somewhat better developed children who are slightly older than a week. During the first days of life infants do not usually react with hyperthermia. These prematures who become infected frequently run an afebrile course. There are other factors besides the insufficient heat regulating mechanism in the skin which may explain the tendency to

low temperatures. The skin surface in prematures is relatively greater in proportion to the volume of body than in older children. The poorly developed subcutaneous fat enhances the heat liberation factor. Furthermore, prematures take only small amounts of food and so limit the heat production. Under ordinary circumstances, however, the temperature of the human body is not dependent upon the food intake. Gundobin considers that the temperature of the premature is a criterion of its development. Incidentally, it should be mentioned that Budin believes that sclerema, which is due to the coagulation of the subcutaneous fat at low temperatures, is infrequent in modern times because the temperature of the premature infant is carefully maintained. In the newly born, infections of the umbilicus or the umbilical vessels sometimes proceed with a normal or subnormal temperature. This may occur in well-developed and mature infants as well as in the premature. Of 1,665 newly born infants which he studied, Smudzinski found 14 who fell under this classification 7 of whom died and the remainder recovered.

It has been suggested by Devilliers that a certain number of cases of fever in newly born infants are due to obstetrical traumatism. He thinks that the application of forceps, a rigid birth canal compressing the head, prolonged labor as occurs in brow, face, transverse, and breech positions, hemorrhage into the sternocleidomastoid or cephalohematoma may be sufficient to produce a febrile reaction. Indeed, the hyperthermia may be the principal symptom of the traumatism. The interpretation of these cases shows that the obstetrical traumatism has been relatively severe due to traction and prolonged period of expulsion. These cases occur most commonly in primipara. The fever is not protracted and rarely lasts longer than two days. In thirteen out of twenty-one cases it lasted one day; in five, two days. It usually reaches 100° and rarely exceeds 102°. To make these observations accurately it is necessary to make frequent readings. There are numerous lesions following the use of the forceps the most serious of which are fractures and meningeal hemorrhage. In such cases there are paralyses, focal symptoms, coma, and rigid protuberant fontanelles. The fever is prolonged and the condition is more grave than in the cases with contusions of the scalp, peri

cranium and muscles of the neck, and hematomata. Recently it has been pointed out that autopsies on infants, particularly those extracted with forceps, show small hemorrhagic foci in the central nervous system, most frequently in the cervical cord, medulla and pons. The series referred to by. Devilliers considered for the most part the mild lesions with a short, febrile course. Nevertheless, the facts are of interest and sustain the opinion that a forceps operation is a potential traumatism to the fetus. This doctrine of aseptic fever is concurred in by Mendelsohn who thinks that the fever following fractures, hematomas, and tissue contusions in children is due to the absorption of assimilable body proteins.

One of the phenomena peculiar to the first days of life is the so-called "inanition" fever of Holt and Crandall, the "thirst" fever of Eric Muller, and the "desiccation" fever suggested by Von Reuss. This is observed on the first to the fifth day, lasting a few hours with remissions rarely lasting five days. The temperature reaches 100°-102° though if long continued it may go to 103°. Cold packs reduce the temperature though it rises following removal. The child may be quiet and drowsy, or it may be fretful and cry a good deal. The appetite is poor, and the child nurses little even when the breasts contain an abundance of milk. The loss in weight is relatively marked, and usually the time of the fever and the loss in weight coincide although the weight loss may be more marked after the temperature has subsided. The stools may have the hunger appearance. It is peculiar that welldeveloped strong children are usually attacked. In the preaseptic period this transitory fever was thought to be an unlocalized sepsis and was compared to that occurring in puerperal women. Eross thought that the umbilical, gastro-intestinal, and pulmonary processes were complications of the original infection. There are many facts which oppose the theory of infection. The attack always occurs on the third to the fifth day. Premature and weak children, who are predisposed to overheating, are not affected. Under identical conditions only a few children develop the fever. It is difficult to understand how an infection through the navel will disappear in a few days without reappearing. The theory has been advanced that there is a contest between the meconium and the milk flora occurring at

this time. Chossat noted that starving animals attempted to maintain their temperature. In cats and dogs before death the temperature shows a marked increase. Shortly before death this power is lost and the temperature rapidly sinks until death ensues. L., F. Meyer and Rosenstein showed there is a distinct tendency towards a sinking temperature after eight to ten days in infants who are starved. Experiments must be carefully controlled because applying external heat may elevate the otherwise low temperature. After long continued starvation the administration of diluted milk tends to produce a fall in temperature. The temperature is kept low by salt-poor malt soup while butter and skimmed milk tend to keep the temperature on a level. Eric Muller states the greater the loss in weight, the higher the temperature. He noted that fifty-five per cent. of infants who lost from 500 to 720 grams had fever. There may be a specific toxic product, which is probably the result of protein metabolism, retained by the body because of imperfect elimination. This theory would place transitory fever in the group of autointoxications where it probably belongs.

Fevers associated with digestive disorders of infancy have been variously explained. For the most part the possible bacterial origin of the fevers has been difficult to reconcile with bacteriological and pathological knowledge. Finkelstein and his students have attempted to explain this fever phenomenon in a unique way. They maintain that the salt content of whey whether administered as pure whey or milk damages the intestinal mucosa. The carbohydrates, particularly the sugars, whether milk or malt sugar, undergo abnormal fermentation, which results in the production of toxic products. The damaged intestinal wall no longer acts as a barrier between the intestinal lumen and the organs beyond which are concerned in intermediary metabolism. Consequently, the toxic products readily pass through the intestinal wall and exert their deleterious influences producing that long train of symptoms which Finkelstein calls alimentary intoxication, not the least important symptom of which is high fever. Finkelstein's theory rests upon the fact that salt injected subcutaneously or taken by mouth is capable of producing febrile reactions and that sugar in the presence of whey salts in the intestinal tract

is capable of producing fever in the manner described. Various observers have reported the occurrence of temperatures following hypodermic or intravenous injections of salt solutions. These results have been called into question, and the experiments have been repeated with sterile salt solutions without producing any temperature. Conclusions have been expressed that the temperatures following the hypodermic injection of salt solutions were not to be ascribed to the salt but to saprophytic impurities in the water used in the solution which contained bacterial proteins not destroyed in the ordinary cooking process.

The effect of external conditions have been previously alluded to. Normal and previously healthy children may suffer from heat stroke. Very high temperatures may be produced by a complete breaking down of the heat regulating center. Children who have suffered frequent insults from gastro-intestinal diseases during the summer are liable to acute exacerbations of alimentary intoxication associated with high fever and marked losses in weight. These are the patients who contribute to the marked increases of infant morbidity and mortality in the hot seaRest and sleep also exert their influence on the temperature levels. In these conditions there is an absence of the stimuli acting on the nervous system tending to elevate the temperature. It has been noted that the temperature of an animal rudely awakened may rapidly rise.

son.

It has frequently been shown that adults show a slight rise in temperature after exercise, the rise varying with the violence of the procedure. Jürgensen showed that individuals sawing wood for four hours exhibited a rise in temperature of two degrees. Penzoldt states that tuberculous individuals show a characteristic rise in temperature after exercise which was higher than that found in normal individuals. The correctness of his observations, however, has been doubted.. Soldiers who have been subjected to a long march frequently show a rise in rectal temperature which may be 100° or higher. After they have rested for thiry minutes the temperature tends to fall to normal or nearly so. This occurs in normal individuals although accentuated in the tuberculous subject. It is likewise present in neurotics and those convalescing from acute infection and tonsillar hypertrophy. This obser

In

vation also applies to older children. The fact has already been referred to that the temperature of young infants is influenced by bodily exercise, restlessness and crying. In older children this fact comes out in a more striking manner. one girl after thirty minutes' exercise, the temperature rose to 99.9°. After a half hour more it rose to 100.4°. After she rested thirty minutes the temperature fell. Such children show a concomitant increase in pulse rate and a moderate increase in the leucocyte count. Young children who undergo athletic training not infrequently show an increase in temperature. On the other hand, those who have weak muscles tend to show some increase in temperature after long physical effort. It is not always necessary that the actual exercises be carried out, but activities such as tossing in bed or dressing violently may produce a moderate rise in temperature. One observer noticed that children who came to his dispensary with nutritional disturbances or anemias presented themselves with a temperature of 99°. After a rest it fell to a point 0.5°-0.7° lower. If they were sent out to jump or run up and down stairs the temperature would again rise 0.5°-0.7°. He observed that diseased conditions seemed to exert no influence on this rise; weak or strong, normal children or those infected with tuberculosis, the change was the same. He found that definitely neuropathic children responded more readily with an increased bodily temperature after exercise than normal children. oscillations in temperature are so universal that they should be considered normal manifestations based upon a somewhat unstable heat regulating mechanism in young children.

These

From personal observations I have been able to follow many children through infancy and childhood who thrive and seem to be in perfect health and constantly show a temperature of 99, 99.5° and even 100.5°. They have frequently been a source of great anxiety to their parents, and in the first years of my practice I confess that I participated in their fears. I finally came to the conclusion that a certain group of children show a temperature somewhat higher than normal, and after every possible pathological condition is excluded, such as tonsillitis, pyelitis, and indeed any acute or chronic infection, I advise that the thermometer should be rarely if ever used. Strange to say, not a

few of these cases occur in the families of physicians, former nurses, and in neurotic families where it was a common practice to resort to the thermometer.

THE IMPORTANCE OF THE ANAEROBIC BACTERIA TO MAN.*

W. L. HOLMAN, M. D.,

PITTSBURGH, PA.

(Continued from July issue, page 289)

The most important anaerobes met with in war wounds are-1. B. welchii (Welch 1892), known as B. aerogenes capsulatus in America and as B. perfringens in France This is the anaerobe with which we are all most familiar. It is very widespread, extremely common in feces and is comparatively easy to cultivate. It is a large Gram positive frequently diplobacillus, non-motile and capsulated. It gives large colonies in and on media under a variety of anaerobic conditions. On account of its strong powers of fermentation it makes itself very evident in mixed cultures in milk, meat and similar media. Injected into the muscle of guinea pigs it is pathogenic and produces a marked breaking down of the muscle which is probably the result of the carbohydrate fermentation and possibly autolytic ferments as suggested by Dernby's studies since B. welchii shows in the test tube only the most minimal attack on solid proteins such as meat. It spores readily in feces, not infrequently in wounds, and with comparative difficulty in artificial media. Acid production interferes with sporulation. It is found in the vast majority of deep lacerated wounds and is the most frequent cause of gas gangrene. Its toxine has been discovered by Bull and Pritchett, Weinberg and others and an antitoxine prepared and used effectively, both experimentally and practically.

2. Vibrion septique (Pasteur 1872) is much more difficult to isolate. It is a rather long, moderately stout, motile bacillus with usually subterminal spores. Its colonies are very small, both surface and deep, so that it may remain in mixtures of other anærobes and not be recognized. It is very pathogenic when injected into the muscles of animals and gives rise in man to a serious and often fatal form of gas gangrene. It is probably the second most common cause of this condition. *Read before the Chicago Medical Society, March 12, 1919.