w
/<
A STUDY OP PROLONGED FASTING
BY
FRANCIS GANO BENEDICT
WASHINGTON, D. C.
Published by the Carnegie Institution op Washington
1915
Carnegie Institution of Washington
Publication No. 203
PRESS OF GIBSON BROTHERS, INC.
WASHINGTON, D. C.
PREFACE.
The research reported in this book on the metabolism during pro-
longed fasting is a continuation and amplification of the investiga-
tions reported in "The influence of inanition on metabolism" (Carnegie
Institution of Washington Publication No. 77, 1907).
The opportunity to conduct this series of scientific observations on a
man living for 31 days without food and drinking only distilled water
would have been of little value without the cooperation of a large
number of scientific associates and computers. Certain co-workers
kindly assumed the responsibility not only for the accumulation of the
data but also for the preparation of a report of their respective findings.
In this book special reports are made by Dr. H. W. Goodall on the
physical condition of the subject during the fast, his subjective impres-
sions and mental attitude toward the fast, and the microscopy of the
urine and the tests for albumin; by Dr. J. E. Ash on the blood; by Dr.
H. S. Langfeld on the psycho-physiology of the fast; by Dr. A. I. Kendall
on bacterial intestinal flora; and by Mr. H. L. Higgins on alveolar air.
Aside from those who shared directly in the responsibility of the
studies, I am indebted to numerous scientific authorities for counsel and
advice, both during the experiment and during the preparation of the
material for publication. Those not specifically mentioned in the text
are Professors Luciani of Rome, Fano of Florence, Zuntz of Berlin,
Tangl of Budapest, Tigerstedt of Helsingf ors, and Lusk of New York.
In no undertaking of the Nutrition Laboratory have the concentration
and the unification of resources and assistants been so intensely applied
and to the whole staff of the Laboratory my warmest thanks are due.
Their interest and conscientious, painstaking work alone made sure
the collection of the data reported in the following pages. The labor
of the final preparation of the material has fallen in no small part into
the excellent hands of my editorial associates, Mr. W. H. Leslie and
Miss A. N. Darling.
Nutrition Laboratory of the Carnegie Institution of Washington,
Boston, Mass., July 29, 191 4.
3
CONTENTS.
PAGE.
Introduction 11
Previous observations of prolonged fasts 13
Observations on Succi 16
Research on metabolism in prolonged fasting at the Nutrition Laboratory 19
Problems to be studied 19
Selection of subject 20
Proofs of physical fitness 21
Autobiographical notes 22
General characteristics of subject 28
General history of fasting experiment 29
Program for research 31
Daily records of fasting experiment 32
Preliminary period 32
Fasting period 37
Re-alimentation period 49
Physical condition of the subject during the fast 53
Results of physical examination 54
Summary as to physical condition 63
Photographic study of subject 66
Anthropometric measurements 67
Body-weight 69
Routine of observations 69
Daily losses in body-weight in fasting experiments 71
Total loss in body-weight *. 80
Analysis of losses in body-weight 82
Insensible perspiration 83
Drinking water 85
Body-temperature 88
Changes in temperature rhythm 89
Observations of the body-temperature in the night period 92
Average body-temperature 92
Range in body-temperature 94
Observations of the body-temperature in the day period 95
Constancy in body-temperature at a given hour 97
Pulse-rate 99
Records of pulse-rate obtained in earlier fasting experiments 100
Records of pulse-rate obtained in the experiment with subject L 103
Pulse-rate in the night periods 106
Pulse-rate in the day periods 110
Comparison of pulse records obtained in experiments with the bed calori-
meter and the respiration apparatus 112
Influence of body position 115
Influence of the work of writing 116
Influence of breathing an oxygen-rich atmosphere 116
Diurnal rhythm 117
Irritability of the heart 117
Blood pressure ' 119
The blood 124
Correlation of literature 124
Erythrocytes 125
Haemoglobin 132
Leucocytes 136
Physico-chemical changes 144
Observations on L's blood 148
Discussion and conclusions 156
Mechanics of respiration 158
Typical graphic records of respiration 158
Method of calculating the total ventilation of the lungs 160
Method of calculating the volume per inspiration 161
Results of observations on the mechanics of respiration 162
Respiration-rate 163
Ventilation of the lungs per minute 164
5
6 A STUDY OF PROLONGED FASTING.
PAGE.
Research on metabolism in prolonged fasting at the Nutrition Laboratory — Continued.
Mechanics of respiration — Continued.
Results of observations on the mechanics of respiration — Continued.
Volume per inspiration 164
Influence of changes in body position 165
Influence of the work of writing 165
Influence of breathing an oxygen-rich atmosphere 166
Maximum expiration of the lungs 166
Alveolar air 168
Significance of alveolar air 168
Methods of determining the alveolar air 169
Haldane method 169
Plesch method 171
Method of calculating alveolar air from respiration experiments 172
Conditions of taking alveolar-air samples " 174
Discussion of results 175
Size of dead space in fasting 175
Difference in mechanics of respiration in morning and evening 178
Significance of change in the alveolar air during the fast 180
Conclusions 181
Subjective impressions and mental attitude toward the fast 182
Subjective impressions 182
Mental attitude of the subject toward the fast 187
The psycho-physiology of a fast 191
Memory for words 193
Tapping tests 194
Strength tests 196
Tactual-space threshold 198
Rote memory for digits 199
Association tests 200
Cancellation test 206
Visual acuity 207
Later tests 208
Correlations 211
General summary and conclusions 212
Appendix I. Dreams 222
Appendix II. Complete series of association tests 222
Feces 230
Observations upon the bacterial intestinal flora of a starving man 232
Excretion through the skin 233
Urine 236
General routine of collection and sampling 236
Composition of the urine prior to the fasting experiment 238
Physical characteristics of the fasting urine 238
Volume of urine 240
Specific gravity 242
Total solids 243
Day and night urines 245
Chemical constituents of fasting urine 247
Total nitrogen 247
Comparison of total nitrogen excretion of L. with that of other fasting
subjects 247
Daily excretion of nitrogen 250
Nitrogen excretion per kilogram of body-weight 252
Comparison of methods for determining total nitrogen and ammonia-
nitrogen 253
The partition of the nitrogen excretion 254
Urea 254
Ammonia 257
Uric acid 259
Creatinine 262
Rest-nitrogen 268
Acid radicles 268
Chlorine 268
CONTENTS. /
Research on metabolism in prolonged fasting at the Nutrition Laboratory — Continued. page.
Urine — Continued.
Chemical constituents of fasting urine — Continued.
Phosphorus 273
Sulphur 277
Total acidity 281
/8-oxybutyric acid 282
Mineral metabolism 285
Relationships of the mineral constituents 287
Reducing power 291
Carbon in urine 293
Carbon-nitrogen ratio 295
Energy of urine 297
Calorie-nitrogen ratio 297
Calorie-carbon ratio 298
Microscopy of urine and tests for albumin 300
Detailed results 300
Summary 302
The respiratory exchange 304
Apparatus and methods used in the calorimeter experiments 305
Absorption of water-vapor and carbon dioxide 305
Analysis of chamber air at the end of periods 306
Tension equalizer 309
Argon in oxygen from liquid air 310
Graphic registration of degree of muscular repose of subject inside the respi-
ration calorimeter 311
Methods used in experiments with the respiration apparatus 315
Studies with the bed calorimeter 320
Atmospheric conditions inside the chamber 320
Measurement of the respiratory exchange inside the bed calorimeter 322
Periodic changes in the metabolism 323
Total metabolism 327
Respiratory quotient 330
Relationships of pulse-rate, body-temperature, and metabolism 331
Studies with the universal respiration apparatus 333
Variations in the metabolism as the fast progressed 335
Relationship between the pulse-rate and the metabolism 336
Diurnal variations in metabolism 337
External influences upon metabolism 338
Effect of changes in body position 338
Influence of the work of writing 340
Influence of breathing an oxygen-rich atmosphere 341
Influence of sleep 343
Metabolism per unit of weight and surface 351
Metabolism per kilogram of body-weight 353
Metabolism per kilogram of body-weight in calorimeter experiments . . 359
Metabolism per kilogram of body-weight in respiration-apparatus ex-
periments 362
Conclusions regarding the metabolism per kilogram of body-weight . . . 364
Metabolism per square meter of body-surface 366
Metabolism per square meter of body-surface in the calorimeter experi-
ments 369
Metabolism per square meter of body-surface in the respiration-appa-
tus experiments 370
Conclusions regarding the metabolism per square meter of body-surface . 370
Summary of results regarding the metabolism per kilogram of body-weight
and per square meter of body-surface 372
Elimination of water through lungs and skin 373
Calorimetry 379
Direct calorimetry 379
Indirect calorimetry 384
Balance of income and outgo 392
Total katabolism per 24 hours 392
Daily activity 393
Total carbon-dioxide production and oxygen consumption per 24 hours .... 395
8 A STUDY OF PROLONGED FASTING.
PAOE.
Research on metabolism in prolonged fasting at the Nutrition Laboratory — Continued.
Balance of income and outgo — Continued.
Character of the katabolism 399
Protein katabolism 400
Apportionment of non-protein katabolism between carbohydrate and fat. . . 401
Carbon dioxide produced and oxygen consumed in the katabolism of
carbohydrate and fat 402
Significance of the non-protein respiratory quotients 403
Energy derived from katabolism of carbohydrate and fat 405
Amounts of carbohydrate and of fat katabolized 406
Loss of water from the body 407
Loss of preformed water 408
Total loss of original body-substance 412
Total energy loss 413
ILLUSTRATIONS
PAGE.
Plate 1. A. Characteristic pose of L., sitting in the balcony, during the day, writing at
his desk. B. Use of universal respiration apparatus for studying the respira-
tory exchange while writing 10
Plate 2. C. Respiration experiment made by T. M. Carpenter on the universal respiration
apparatus. D. Weighing the subject on the thirty-first day of the fast. . . 18
Plate 3. E. The subject L. ascending the stairs of the balcony on the thirty-first day of
the fast. F. Clinical examination by Dr. H. W. Goodall 30
Plate 4. Views of subject Levanzin on first day of 31-day fast 64
Plate 5. Views of subject Levanzin on last day of 31-day fast 64
Fig. 1. Body-weight curves for fasting experiments with Succi 74
2. Body-weight curve for Levanzin 75
3. Body-weight curves for prolonged fasting experiments with dogs 77
4. Body-temperature curves during the night and early morning for the second and
fourth to eighth days of fast 90
5. Body-temperature curves during the night and early morning for the ninth to the
sixteenth days of fast 91
6. Body-temperature curves during the night and early morning for the seventeenth
to twenty-second days of fast 92
7. Body-temperature curves during the night and early morning for twenty-third to
twenty-ninth days of fast 93
8. Body-temperature curves during the night and early morning for thirtieth and
thirty-first days of fast and second and third days with food 94
9. Body-temperature curves for approximately 24 hours on twenty-fourth and
twenty-fifth days of fast 96
10. Body-temperature curves showing change from lying to sitting position 97
11. Body-temperature curve showing change from lying to sitting position 97
12. Pulse-rate chart of subject L for days preceding fast 104
13. Pulse-rate chart of subject L. for first to fifth days of fast 105
14. Pulse-rate chart of subject L. for sixth to eleventh days of fast 106
15. Pulse-rate chart of subject L. for twelfth to eighteenth days of fast 107
16. Pulse-rate chart of subject L. for nineteenth to twenty-fifth days of fast 108
17. Pulse-rate chart of subject L. for twenty-sixth to thirtieth days of fast 109
18. Pulse-rate chart of subject L. for thirty-first day of fast and three subsequent
days with food 110
19. Chart showing blood pressure, pulse pressure, and pulse-rate of subject L 121
20. Chart I. Relation of haemoglobin to erythrocytes. Chart II. Composite curve of
the polynuclears compared with one of mononuclears 152
21. Charts III and IV. Relation of total to differential leucocyte counts 153
22. Specimen respiration curves for subject L. when lying on couch in experiments
with the respiration apparatus 159
23. Memory tests 193
24. Tapping tests 195
25. Strength tests 196
26. Strength tests 197
27. Strength tests 198
28. Tactual-space threshold and visual acuity 199
29. Free association tests 201
30. Association tests. Reactions to verbs and nouns 201
31. Association tests. Reactions to adjectives 202
32. Association tests. Reactions to abstract nouns 202
33. Reproduction tests and mean variations 203
34. Controlled association tests 205
35. Cancellation tests 206
36. Specimen records of change in volume of the spirometer on the bed calorimeter
during last 5 minutes of periods in experiment with L 310
37. Method for obtaining graphic record of activity in bed calorimeter 312
38. Specimen pneumograph records of movements of bed calorimeter lever mattress
support in night experiments with L 314
39. Schematic outline of universal respiration apparatus 316
9
10 A STUDY OF PROLONGED FASTING.
PAGE.
Fig. 40. Spirometer for studying the mechanics of ventilation 318
41. Curves showing oxygen consumption, carbon-dioxide production, and respiratory
quotient during night periods in the bed calorimeter for the four days preceding
the fast and the first to the fourth days of the fast 323
42. Curves showing oxygen consumption, carbon-dioxide production, and respiratory
quotient during night periods in the bed calorimeter for the fifth to the fifteenth
days of the fast 324
43. Curves showing oxygen consumption, carbon-dioxide production, and respiratory
quotient during night periods in the bed calorimeter for the sixteenth to the
twenty-fourth days of the fast 325
44. Curves showing oxygen consumption, carbon-dioxide production, and respiratory
quotient during night periods in the bed calorimeter for the twenty-fifth to the
thirty-first days of the fast and the second and third food days 326
45. Complete metabolism chart of fasting dog (Awrorow No. 2) 356
46. Complete metabolism chart of fasting dog (Awrorow No. 3) 357
47. Metabolism chart of the most important factors measured on subject L. through-
out the fast 416
FASTING
PLATE 1
A. Characteristic pose of L. sittine in the Balcony, during the day, writing at his deslc.
B. Use of Universal Respiration Apparatus for studying the Respiratory Exchange while writing.
INTRODUCTION.
Prolonged fasting has formed a part of religious ceremony for centu-
ries. In early times the ascetic, in his efforts to subdue all carnal
desires, believed it necessary to withdraw from the distractions of daily
life and to abstain either wholly or in part from food, particularly the
flesh of animals; by thus refraining from material things, he hoped to
be free for spiritual thought and philosophical introspection.
Periodic fasting still constitutes a part of the rites of some religious
bodies, particularly among the Hebrews, but in modern times a pro-
longed fast is usually undertaken either in the hope of curing or alle-
viating some ailment or for pecuniary gain. When a fast is resorted
to for its supposed therapeutic value, information as to its history and
results usually appears in one of the numerous books published by the
advocates of peculiar dietetic regimes. When a fast is made by a
so-called " professional faster" for pecuniary gain, the subject is exhib-
ited to the public as an attraction to the lovers of sensational amuse-
ments. Three decades ago such exhibitions were not uncommon and
in many instances the subjects consented (possibly in the hope of
increasing the interest in their performance) to more or less strictly
controlled observations of their fasts. Not infrequently the observa-
tions made in these professional fasting exhibitions have contributed
materially to the sum of human knowledge, since there is an intense
physiological interest in the vital processes during such prolonged
abstinence from food.
When one considers the complex activities which make up the life of
man, it will be seen that no mechanism thus far invented approximates
the high organization of the vital processes which are necessary to the
life of even the simplest of the warm-blooded animals ; and yet sufficient
experimental evidence has been accumulated to show that under normal
conditions of life, and with similar routine, there are no marked vari-
ations in the life processes of normal individuals. Under varying
conditions of life, however, we find that the vital activities are carried
on with a greater or less intensity, this being true even of the normal
individual. We thus see that there may be definite, well-established
planes of vital activity. For example, when the average healthy indi-
vidual is lying in bed asleep, there is no intellectual activity and no exter-
nal muscular activity, the vital activity being only sufficient for simple
maintenance. When he is lying quietly in bed awake, the plane of
vital activity is higher, and as we study the metabolism under the vary-
ing conditions of sitting, standing, walking, and doing muscular work,
we find an increasing intensity in the vital processes, with an increase
in productive capacity and often an increase in efficiency.
11
12 A STUDY OF PROLONGED FASTING.
The average normal man represents the mean between the two
extremes of the emaciated, half-starved individual, disinclined to phys-
ical or mental work, and the over-fed, obese epicure, both extremes
being relatively low in vital activity and in productivity. Further-
more, if we consider the metabolism under pathological conditions, we
find even greater variations in the different levels of vital activity.
Thus a sick person, much emaciated, lying in bed without food, and
with subnormal temperature, has unquestionably a low cellular activity.
On the other hand, a sick person with a high fever, even when asleep
and without extraneous muscular activity, may have a greatly increased
cellular activity. It will be seen, therefore, that from the standpoint
of both normal physiology and pathology, a study of human individuals
under different conditions and with different planes of activity is of
fundamental importance.
For such study it is essential to determine the basal or fundamental
metabolism, when the activities are on a low plane, to be used as a basis
of comparison with other values. We may ask, then, "What is the
lowest plane of vital activity which is compatible with life?" Unques-
tionably there have been severe pathological cases, with emaciation and
muscular atrophy, in which life has been maintained at a plane far
below that which can be reached by the average normal man, but it has
been the prime object of most investigators in metabolism to concen-
trate their efforts upon securing, with normal individuals, physiological
values which may withstand criticism, since these constitute the only
true basis of comparison.
Taking into consideration the influence upon metabolism of muscular
activity, of the ingestion of food, and the state of being awake, we may
assert that the lowest metabolic plane would be found for an individual
during deep sleep in bed, with complete muscular repose, and without
food in the alimentary tract. As a matter of fact, with most people
such a condition is usually closely approximated each day about 4 a. m.
While in general no food is taken by an individual for about 10 or 12
hours during the night, yet for a considerable period of time after the
evening meal nutrients are being absorbed from the ingested food mate-
rials and carried to different parts of the body, there to be oxidized or
deposited. It is furthermore true that certain molecular fragments,
probably acid in nature, maybe absorbed from the food materials which,
when carried to the various parts of the body, may actually stimulate
metabolism to a greater intensity, these being the so-called katabolic
stimuli. Usually the influence of the ingestion of food ceases from
6 to 8 or 10 hours after the meal, particularly if the food ingested is not
protein-rich. Accordingly, for one or two hours prior to rising in the
morning the normal man is probably living at his lowest metabolic
plane.
PREVIOUS OBSERVATIONS OF PROLONGED FASTS. 13
As is well known, the normal body is liberally provided with reserve
material, a fact which has been strikingly brought out and emphasized
by Meltzer.1 Consequently there is always a plethora of available
material stored in the body for drafts in emergencies. In the normal
life of man, the demands for nutrition are usually met by periodic
feeding. When the demands are not met, body reserves must be drawn
upon. Under such conditions it is of particular interest to note what
kind of body-material is first used, the rapidity of its depletion, and the
proportions of the various body constituents disintegrated as the drafts
continue. It is to study these problems that observations are made
upon fasting individuals. Furthermore, since many prominent clin-
icians are inclined to consider disease as closely allied to the various
stages of inanition, data secured in a study of metabolism during fasting
have a great pathological importance for interpreting the transforma-
tions of matter in disease.
PREVIOUS OBSERVATIONS OF PROLONGED FASTS.
The literature giving the results of observations during fasts has been
reviewed at some length in a previous publication,2 special emphasis
being laid upon the results obtained in the earlier stages of a fast. In
this publication it seems desirable to give a review of the longer fasts
which have been more or less scientifically controlled and whose results
can be considered as worthy of careful consideration.
The longer fasts have almost without exception been made by pro-
fessional fasters who, for purposes of exhibition, have purposed going
without food for a definite length of time. While such a purpose would
of itself seem to show an abnormal mental condition, yet the majority
of professional fasters who have been used in these experiments are for
the most part physically strong, and the results may usually be looked
upon as of physiological importance, not complicated by pathological
lesions of any measurable magnitude. This is particularly fortunate,
as many fasts reported in the daily press are undertaken as a therapeutic
measure to overcome some more or less definitely localized organic
or functional trouble. It is obvious, however, that such experiments
are of physiological importance when the subjects are normal individ-
uals, voluntarily fasting under strict scientific control.
Many professional fasters have made experiments of longer or shorter
duration and have been studied by various investigators, but none have
been so carefully studied and had so many experiments made with
them and of such long duration as the Italian, Succi. Indeed, the
classical work of Luciani on Succi emphasized perhaps more than any
Seltzer, The Factors of Safety in Animal Structure and Animal Economy. Harvey Society
Lectures, New York, N. Y., 1906-1907, p. 139.
2Benedict, Carnegie Inst. Wash. Pub. No. 77, 1907.
14 A STUDY OF PROLONGED FASTING.
other piece of research the importance of studying prolonged fasting.
In this review of the literature on long fasts, therefore, brief descriptions
of the fasts made by subjects other than Succi will first be given chrono-
logically, these being followed by descriptions of experiments made
with the Italian subject. Such discussion of the results as may be
necessary will be reserved for later chapters.
Observations by Paton and Stockman.1 — An experiment was made in
the fall of 1888 by Paton and Stockman on the professional faster
Jacques and continued for 30 days. The body-weight was recorded,
but unfortunately the urine was analyzed by the old hypobromite
method. Furthermore, the values for total nitrogen output were un-
doubtedly disturbed by the singular fact that the subject drank from
60 to 300 c.c. of his own urine each day. Since the volume of fluid
taken per day varied greatly, the body-weight fluctuated considerably,
actual gains in weight being shown on some days. No feces were
passed during the fast.
Observations by Lehmann, Mueller, Munk, Senator, and Zuntz} —
Although this research was hardly long enough to be called a study of
prolonged fasting, the two experiments made by Lehmann, Mueller,
Munk, Senator, and Zuntz, one on Cetti of 10 days and one on Breit-
haupt of 6 days, present a study of metabolism during fasting which
has never been excelled in accuracy for this length of time. The
experimental plan adopted in this research has been followed with but
minor changes by practically all succeeding investigators. It was the
intention to continue the experiments with these subjects for 20
or 30 days, but they were unavoidably shortened, owing to the condi-
tion of the subjects. The experiment on Cetti was made in March
1887, and the observations secured in this experiment were of such
importance that the experimenters took advantage of an opportunity
occurring in March 1888 to make an experiment with the professional
faster Breithaupt. Unfortunately this experiment continued only
six days. Observations were made in both experiments of the body
functions, body measurements, pulse-rate, urine, feces, and respiratory
exchange, and the computations and conclusions are of fundamental
importance. They will be continually referred to in connection with
this report.
Observations by van Hoogenhuyze and Verploegh.3 — In a study made of
the urine excreted by a professional fasting woman, van Hoogenhuyze
and Verploegh gave especial attention to the creatinine content. The
experiment began on June 11, 1905, and ended June 25, 1905; the
iPaton and Stockman, Proc. Royal Soc. of Edinburgh, 1888-1889, 16, p. 121.
"Lehmann, Mueller, Munk, Senator, and Zuntz, Archiv f. path. Anat. u. Physiol, u. f. klin.
Med., 1893, 131, Supp., p. 1.
3Van Hoogenhuyze and Verploegh, Zeitschr. f. physiol. Chem., 1905, 46, p. 415.
PREVIOUS OBSERVATIONS OF PROLONGED FASTS. 15
following constituents of the urine were determined: Total nitrogen,
urea, creatinine, uric acid, chlorides, phosphoric acid, indigo, and total
acidity.
Observations by Brugsch, Mohr, Bonniger, Baumstark, and Hirsch.1 —
An experiment made on a fasting woman by Brugsch, Mohr, Bonniger,
Baumstark, and Hirsch was continued from March 10 to March 25,
1906. The observations included loss in body-weight, total nitrogen,
and especially acetone in the breath and acetone and /3-oxybutyric acid
in the urine. The ammonia-nitrogen was likewise determined. The
research is of peculiar importance in that special emphasis was laid
upon the relationship between acidosis and fasting.
Observations by Cathcart.2 — An experiment, carried out by E. P.
Cathcart on the professional faster Victor Beaute, with the strictest
surveillance, was designed primarily to study the effect of fasting upon
the partition of the nitrogen, the new methods introduced by Folin
being used. The experiment was made in Glasgow in 1907 and con-
tinued 14 days. An especial study was made of the mineral matters
excreted and the creatine and creatinine content of the fasting urine.
An interesting complement to the fasting experiment was a study made
at the end of the effect of the ingestion of the starch-cream diet of
Folin, i. e., a low nitrogenous diet, which was continued for a few days.
During this time the uric acid and purine-nitrogen were accurately
determined and the chlorine, phosphorus, and the several forms of
sulphur were carefully estimated. This investigation represents the
most comprehensive and exact observation of the constituents of urine
passed while fasting to be found in the literature. Cathcart's associate,
Charteris, published his blood findings somewhat later.3
Observations on Gayer. — An uncontrolled fast of 30 days was made in
New York on a professional faster, Gayer, continuing from May 16 to
June 14, 1910. Although the attending physicians are by no means
unanimous in their opinions regarding the genuineness of the fast, the
body-weights reported in a non-scientific publication4 indicate a loss
in weight not unlike that experienced in accredited fasting experiments.
The accurate blood examination made by Dr. I. S. Wile6 inspires confi-
dence in the report of this fast.
Observations by Grafe. — Although complicated by abnormal psychical
conditions, by an error on the part of the nurse in giving a rectal enema
on the seventh day, and by considerable variations in muscular activity,
brugsch and Hirsch, Zeitschr. f. exp. Path. u. Therapie, 1906, 3, p. 638; Bonniger and Mohr,
ibid., p. 675; Baumstark and Mohr., ibid., p. 687.
"Cathcart, Biochem. Zeitschr., 1907, 6, p. 109; Journ. Physiol., 1907, 35, p. 500; Cathcart
and Fawsitt, Journ. Physiol., 1907, 36, p. 27.
'Charteris, Lancet, 1907, 173, p. 685.
*Long, Physical Culture, August 1910, p. 190.
'Personal letter received from Dr. I. S. Wile, dated December 24, 1912.
16 A STUDY OF PROLONGED FASTING.
the experiment of Grafe1 with the Jaquet respiration apparatus2 at the
Medical Clinic in Heidelberg is of interest in throwing light upon the
gaseous exchange and the character of the katabolism during prolonged
inanition and on the ratio of carbon to nitrogen in fasting urine. Fur-
thermore, it substantiated the observations made by Brugsch and
others on the acidosis during fasting, as indicated by the excretion of
acetone and ,8-oxybutyric acid.
Observations at Wesleyan University, Middletown, Connecticut. — With
a special view to studying the drafts upon body-material during fasts
of 24 to 168 hours, a lengthy series of experiments was undertaken at
Wesleyan University, Middletown, Connecticut, the results of which
have already been published.3 These experiments threw much light
upon the character of the drafts upon body-material during the experi-
mental periods and showed that the organism and particularly the
storage of glycogen in the body may be greatly affected by even a
short fast. It has furthermore been shown that glycogen — the body-
material which is first and most heavily drawn upon during fasting — may
be considered as one of the most quickly realizable assets, the removal
of which affects profoundly one of the safety factors of the human body.
OBSERVATIONS ON SUCCI.
Fast in Florence, 1888. — Although a short account of the experiment
on Cetti made by Lehmann, Mueller, Munk, Senator, and Zuntz was
published in 1887,4 the details of their investigation did not appear
until 1893,5 and the first extensive report of a prolonged fasting investi-
gation was that made by Luciani of the fasting experiment with Succi
in Florence during the spring of 1888. The Italian report of this fast
was published in 1889,6 but the work is best known to other than
Italian readers by Fraenkel's translation.7
Luciani's study of Succi included an extensive series of observations.
Unfortunately, since the partition of the nitrogen in the urine was at
that time imperfectly understood and as the gaseous exchange was
studied under conditions affecting seriously the accuracy of the results,
Luciani's observations are more especially of value as indications of
the general body functions of a fasting man than as measurements of
^rafe, Zeitschr. f. physiol. Chem., 1910, 65, p. 21.
2Geheimrat W. His, on a recent visit to the Nutrition Laboratory, informed us that a fasting
experiment with a professional faster, a woman, continuing 4 weeks, had been carried out not
long before in his clinic in Berlin by Professor Staehelin, in which the Jaquet respiration apparatus
had been used. Owing to the indisposition of the subject, the experiment was less successful
than had been hoped, and Professor Staehelin's removal to Basel has indefinitely postponed the
publication of the results.
"Benedict, Carnegie Inst. Wash. Pub. No. 77, 1907.
4Lehmann, Mueller, Munk, Senator, and Zuntz, Berliner klin. Woch., 1887, pp. 290 and 425.
"Lehmann, Mueller, Munk, Senator, and Zuntz, Archiv f. path. Anat. u. Physiol, u. f. klin.
Med., 1893, 131, Supp., p. 1.
•Luciani, Fisiologia del digiuno; studi sull' uomo. Florence, 1839.
7Luciani, Das Hungern. Translation by M. C. Fraenkel. Hamburg and Leipsic, 1890.
PREVIOUS OBSERVATIONS OF PROLONGED FASTS. 17
specific chemical transformations. Succi's peculiar psychical condi-
tion, a condition which seems to be characteristic of the ascetic who
subjects himself to a fast of 30 days or more, is interestingly commented
upon in extenso by Luciani. The research as a whole was a model in
plan, and as a painstaking record of cooperative research in fasting it is
equaled only by the experiments of the Berlin investigators. Luciani's
study unquestionably stimulated the considerable number of experi-
ments subsequently carried out with Succi, at least 7 experiments, each
continuing 20 or more days, being made with him by different investi-
gators and in different places.
Fasts in Milan and Paris, 1886. — In reporting the results of the
Florence fast, Luciani refers to two fasts said by Succi to have been
made previously, one in Milan in August and September 1886, and a
second in Paris in the latter part of November and the early part of
December 1886. The short time between the fasts is of special interest.
Little is known regarding these two fasts, but Luciani considered the
records of the body- weights obtained from Succi' s notebooks sufficiently
reliable to include in the published report of his research and he plotted
curves from them showing the loss in body- weight during the fasts.
Fast in London, 1890. — In 1890 Succi carried out a 40-day fast in
London, which began on March 17.1 Although observations were
made of a number of factors during this fast, the controls were so
incomplete that, aside from the body-weight, the observations have but
little value at the present time. The body-weights were apparently
recorded with a great degree of accuracy and form the basis of a curve
which will be discussed later. No statements accompanied the records
of the pulse and respiration as to whether the subject was lying, sitting,
or standing, so that they can have but little significance; fluctuations
in the pulse-rate give evidence of marked changes in the muscular
activity at times. Strength tests were made with a hand dynamometer
each day, showing practically no alteration in the strength. The
axillary records of the body-temperature indicate a lowering of the
temperature toward the end of the fast.
Fast in New York, 1890. — According to Succi's own statements,
substantiated by newspaper reports, Succi carried out a large number
of fasts which were not scientifically controlled. One of the most
important of these was made in New York City about 8 months after
the London fast.2 The New York fast began on November 6, 1890,
and was said to have continued 45 days. Correspondence with several
of the physicians who attended this fast shows a diversity of opinion as
to its authenticity. On the other hand, the body-weights recorded, if
correct, indicate about the usual loss in weight, the records being 147.4
British Med. Joum., 1890, pp. 764, 819, 876, 935, 996, and 1056, also p. 1444.
2The New York Daily Tribune, November 6, 1890, and December 21, 1890.
18 A STUDY OF PROLONGED FASTING.
pounds (66.86 kilograms) at the beginning of the fast and 104.75 pounds
(47.52 kilograms) at the end.
Fast in Naples, 1892. — The next scientifically controlled fast with
Succi was in Naples, beginning August 7, 1892. Observations were
made by Ajello and Solaro,1 most of these being on the urine. The
body-weight was likewise carefully recorded as the fast progressed, as
well as the amounts of water taken. The determinations made on the
urine which are of interest at this time are those of the chlorine and
phosphoric and sulphuric acids. On the second day of the fast, 2
grams of feces were passed and on the eleventh day, 317 grams.
Fast in Rome, 1893. — A number of observations were made on Succi
by Dutto and Lo-Monaco2 during a 20-day fast in Rome beginning
December 16, 1893. The body- weight was recorded each day, also
the amount of water taken. Analyses were made of the urine excreted,
these being much more complete than in any of the earlier fasts, as the
nitrogen was determined by the Kjeldahl method. Determinations
were also made of the acidity of the urine and the content of sulphur,
ethereal sulphates, neutral sulphur, chlorine, phosphorus, sodium, and
potassium.
Fast in Vienna, 1896. — The urine excreted by Succi in a 21-day fast
was studied by E. and 0. Freund3 in Vienna in 1896, an extensive
partition of the nitrogen being attempted for the first time. The
observations as to Succi' s condition, including the body- weight, were
unfortunately lost.
Fast in Zurich, 1896. — During a 21-day fast of Succi in Zurich,
beginning September 13, 1896, Daiber4 studied the urine and obtained
the body-weight. The body-temperature, the amount of water taken,
and the chlorides in the urine were all determined with sufficient accu-
racy to make them of value at the present day.
Fast in Hamburg, 1904-. — The last recorded experiment on Succi was
made in Hamburg in March 1904. During the last 10 days of this
30-day fast, the urine was examined by Brugsch,6 who determined the
partition of the nitrogen. Special emphasis was laid upon the acidosis.
^Ajello and Solaro, La Riforma Medica, 1893, 2, p. 542.
2Dutto and Lo-Monaco, Policlinico, 1895, 2, p. 1.
3E. and O. Freund, Wiener klin. Rundschau, 1901, 15, pp. 69 and 91.
♦Daiber. Schweiz. Woch. f. Chem. u. Pharm., 1896, 34, p. 395.
5Brugsch, Zeitschr. f. exp. Path. u. Therapie, 1905, 1, p. 419.
FASTING
PLATE 2
C. Respiration experiment made by T. M. Carpenter, on the Universal Respiration Apparatus.
These experiments were made each morning, just after the Subject left the Respiration Calori-
meter, and before he stood up.
D. Weighing the Subject on the Thirty-first day of the Fast. At the right is shown the Bed on
which he has just finished the Respiration Experiment ; the Universal Respiration Apparatus
is shown at the extreme right.
RESEARCH ON METABOLISM IN PROLONGED FASTING AT
THE NUTRITION LABORATORY.
PROBLEMS TO BE STUDIED.
In the research on metabolism during short fasting periods, which
was carried out at Wesleyan University, Middletown, Connecticut, the
changes incidental to the first days of fasting were, it is believed, ade-
quately studied. On the other hand, it was desirable to supplement the
earlier observations by a study of the metabolism during prolonged
fasting, since many points regarding the course of the metabolism
after the body had adjusted itself to the fasting condition had not been
established. For instance, as the fast progresses it is important to
know whether the gross metabolism alters either per kilogram of body-
weight or per square meter of body-surface, also whether the acidosis
is extreme or whether there is an acquired tolerance of it, and what
effect the acidosis, if present, has upon the metabolism. Since the
carbon, the ammonia, and the heat of combustion of the urine, also
the composition of the alveolar air, give indications as to acidosis, a
study of prolonged fasting should include determinations of all of these
factors. In the earlier fasting study determinations were made of a
number of the constituents of the urine, including total solids, nitrogen,
creatine and creatinine, phosphorus, sulphur, and chlorine. In the
longer research it would be necessary to elaborate these determina-
tions, studying also the composition of the feces, should any be passed
during the period. Furthermore, the relationship between the pulse-
rate and the metabolism, the character of the respiration as shown by
graphic records, the variations in the body-temperature, and the changes
in the composition of the blood, all have sufficient significance to
warrant investigation. Since muscular activity has so great an influ-
ence upon metabolism, the experiments of Zuntz on Breithaupt should
be duplicated with more modern technique. Comparison should be
made of the metabolism in selected periods with constant external
conditions instead of with changing activity as in the earlier research,
and experiments in which the subject breathed a high oxygen atmos-
phere would also be desirable.
The Nutrition Laboratory was especially fitted to carry out a research
of this kind, being well equipped with apparatus for determining the
respiratory exchange and the heat output, as well as for measuring the
pulse, respiration, and muscular activity. It was therefore of funda-
mental importance to have ready a carefully prepared plan for studying
the metabolism during prolonged fasting which could be used when-
ever an opportunity offered for conducting such a research. On the
other hand, it was not desirable to make undue haste in beginning the
19
20 A STUDY OF PROLONGED FASTING.
study, inasmuch as the equipment of the Laboratory was steadily being
increased. The chemical technique was also being rapidly perfected,
the development of the new micro methods of Professor Folin being of
especial value in studying the relatively small volumes of urine excreted
during prolonged fasting.
SELECTION OF SUBJECT.
While no particular effort was made to secure a subject for this
research, advantage was taken of a visit to New York by Succi to confer
with him. His age and his somewhat unreasonable demands for a
large compensation made an arrangement with him undesirable. Fur-
thermore, he would not have cooperated readily in the great number of
tests that were included in the plan for the fasting research. A number
of individuals, stimulated by the report of the earlier study, offered
themselves to the Nutrition Laboratory as subjects for a fasting experi-
ment. A large majority of these were either sufferers or imagined that
they were sufferers from "nervous disease," and were therefore patho-
logically or psychologically undesirable. Furthermore, none of them
had a clear conception of a scientifically controlled fast and of the
importance of the observations which would be included in such a
research. They were therefore not seriously considered.
In the spring of 1911, a letter was received from A. Levanzin of Malta,
offering himself as a subject for a long fasting experiment to be carried
out at the Nutrition Laboratory. The letter was voluminous, but very
intelligently written, and showed an appreciation of the scientific value
of such a research. As Professor Luciani, of Rome, who had made
the classical study with Succi, later expressed his confidence in the
ability of Levanzin to carry out a fast of this length, it seemed probable
that the subject desired for the research had been found. It was sub-
sequently learned that Professor Luciani's acquaintance with Levanzin
was through correspondence only, but his recommendation went far
to convince us of the desirability of attempting an experiment with
this man. Accordingly an exact statement was sent A. L. of the duties
involved in a research of this nature and an arrangement was entered
into for him to come to Boston for the purpose. In accordance with
his own proposition, the agreement was made to cover his expenses, with
a bonus if the experiment was successfully completed, and every
attempt was made to minimize anxiety on the part of the subject. The
risk of protracted illness incidental to the journey from Malta to Boston,
to the change in climate, and possibly as a result of the fasting experi-
ment, had to be considered, and a sworn statement exonerating the
Nutrition Laboratory from any responsibility for illness of more than
4 days' duration was obtained from L. before he left Malta.
PROOFS OF PHYSICAL FITNESS.
21
PROOFS OF PHYSICAL FITNESS.
It was necessary to assure us as far as possible of the fitness of this
man for the research, and he was requested to send us a physician's
certificate as to his health. These proofs were supplied and were as
follows:
Ratnapoora, Sliema, Malta, 10th January, 1912.
I hereby certify that Mr. A. Levanzin, B. A., is in good health. He does not
suffer from any disease and his organs are healthy.
(Signed) Robt. Samut,
Professor of Physiology of Malta University.
Examination of Urine submitted by Mr. Levanzin on Feb. 10, 1912.
Quantity in 24 hours: Unknown; taken as 1500 c.c.
Color:
Light amber.
Odor:
Sui generis.
Reaction :
Acid.
Specific gravity:
1018.
Total solids:
41.9
Deposit :
None.
Urea:
2.1 per cent.
Uric acid:
0.03 per cent.
Chlorides:
1 per cent.
Phosphates:
0.35 per cent.
Indican :
Nil.
Microscopical examination: Negative.
(Signed)
Abnormal constituents.
Albumin
Peptone
Globulin
Glucose
Acetone
Blood r None.
Bile
Pus
Mucus
Diazo reaction
Robt. Samut, Edinb.
Roseville, 39 Strada Ghar-id-dud, Sliema, Malta,
January 20, 1912.
This is to certify that I have to-day physically examined Mr. Agostino
Levanzin and that I have found him in good health and free from organic
disease.
(Signed) Jos. S. Galigia, M. D.
52 Victoria Terrace, Sliema, Malta, November 7, 1911.
I hereby certify to have examined A. Levanzin, Esq., B. A., Ph. Ch., P. L.,
and have found him in a good state of health. The urine was normal in every
respect. Specific gravity, 1025. No traces of albumen nor those of glucose,
etc., have been detected. His height is 5 feet 6 inches. His skeleton and
muscles are normally developed. His gross weight is 152 lbs. I am of
opinion that he could undergo quite easily under ordinary circumstances a
period of prolonged fasting without detriment or danger to his health, and that
under ordinary conditions he is not liable to suffer from any illness that might
upset the experiment or entail any hindrance to same.
I know Mr. Levanzin since many years and in fact I am his family doctor.
I might add that he has already fasted for a long period without suffering any
serious bad after-effects; indeed, I was astonished at his rapid recovery there-
from and return to his normal state of health. * * *
(Signed) Dr. P. P. Agius, B. A., Ph. Ch., M. D.
22 A STUDY OF PROLONGED FASTING.
AUTOBIOGRAPHICAL NOTES.
On the twenty-ninth, thirtieth, and thirty-first days of his fast at the
Nutrition Laboratory, L. wrote a sketch of his life. This is reproduced
verbatim, since it shows many of the interesting features of the life,
education, and habits of thought of the subject.
12th of May, 19 1 2 (29th day of my fast).
More than one hundred years ago, Gabriele Avanzino, a Sicilian, settled in
Malta. Gradually the surname was corrupted into Levanzin. My mother,
Lorenza Borg, living and aged about 58, descends from pure and noble Maltese
blood since 400 years. Her grandfather's uncle was the famous Vincenzo
Barbara, the daring sea-captain of one of the French battle-ships who was by
Botta and other historians falsely accused of having betrayed Marat when he
landed him to take possession of Naples on behalf of Napoleon. Barbara was
the right arm of Napoleon to plot and get rid Malta from the yoke of the
Knights of St. John and he was also the first Grand-Master of Free-Masons
in the Island. Her grandfather was Joseph Borg, another sea-captain who
came to America in the time of the Revolution, volunteered with the insurgents
and fought for the American independence many battles as in his portrait that
we keep he has on his breast from seven to eight medals. That is why I
probably love so much freedom, independence of thought, and sympathize
keenly with America.
My father, Paolo, living and aged about 68, is also the son of a sea-captain,
Agostino, who was drowned when my father was only 3 years of age and so
could not nave a liberal education. He learned the art of ship-building which
was very flourishing in those commercial times, but now being disabled from
both his hands through two accidents that happened to him during his work,
he is carrying a grocery-shop in a village as my mother is carrying a confec-
tionery and toy one in the same place. They are both very honest and hard-
working people and although they have sufficient property to keep them up
comfortably during their old days, they do not want to give up their business
as they want "to leave us something after their death." I have a sister,
Teresina, 20 years, living with my mother and a married one to an engineer,
Ursola, 28 years.
I was born in the Citta Cospicus of Malta, on the 23rd of May, 1872 — 40
years ago. At 6 years of age I went to Egypt with my mother where my
father was working but came back after two years as the hot climate did not
suit us. Frequented the public free-schools and at ten had my first prize — a
five shilling piece — for writing the best essay against "Cruelty to Animals."
Then prizes for drawing as I am very fond of art especially of music and
painting. At 12 I entered the free Dockyard Schools and had several prizes.
At 14 I was admitted by competitive examination as shipwright apprentice as
I wished to follow my father's career, then promoted to draughtsman and then
to clerk. From infancy I was always inclined to hard study and sometimes
during the night I used to steal out of bed to read some interesting book
because my parents did not like to see me overstrain my already weak eyes.
During the time that I served my apprenticeship in the Dockyard I published
two weekly papers, successively, in Maltese the "Habil ta Cullhadd" (The
Friend of All) and " Is-Sengha" (Art) to educate and enlighten the working
classes that live in a very miserable condition and are totally forsaken by the
Government, but both papers failed after a few months through lack of sub-
scribers. At 17 I felt inclined to follow the ecclesiastical career to devote
AUTOBIOGRAPHICAL NOTES. 23
myself entirely to study and oratory, that I like so much, and became a cleric,
but after four years, through matter of convictions and bigoted tyranny of the
superiors, I put off my black robe and entered the Lyceum to prepare myself
for a professional career.
At 20 (1892) I passed my matriculation examination and took up the medical
courses. At the same time I was contributing literary and political contribu-
tions to our best papers and published several poems in Italian that were very
favorably appreciated by the press. I started also the publication of a
University Magazine "Lo Studente Maltese" to stimulate the other students to
contribute literary and scientific articles and I published in English and
Italian a study on Shakespearean drama and some biographies of eminent
Maltese personages. The paper dragged a stinty existence for two years and
perished through lack of funds. At the same time I was conducting two other
political papers in vernacular (Maltese), the " Cottonera" and the "Habil ta'l
Poplu," and it was one of the articles contributed to the "Cottonera" that
provoked against me my first libel and was tried by jury.
My father was still working in Dockyard and as his foreman used to take
bribes from his employees and borrow from them money that he never used to
return back, and as my father did never like to satisfy him in this because he
fulfilled always all his duties honestly and regularly he became his scapegoat
and was always ordered to do the most dangerous and hard kind of work.
Twice he was hurt, twice amputations had been operated on fingers of both
hands, with peril to his life, till he became a disabled man. I protested to the
superiors and they answered that they did not care a bit about it and so, at
last, I published in the "Cottonera," in 1895, a violent article in English in
which I enumerated with details the many bribes and irregularities that were
continually committed in H. M. Dockyard, signed the article and defied the
Admiral Superintendent that I was ready to prove in court all my assertions.
The article provoked a great scandal and the Admiral was obliged to arraign
me before the criminal courts to prove my assertions. The penalty demanded
against me was six months of hard labor imprisonment and a fine of £500.
All my assertions were proved to the very hilt after a fierce fight and I was
triumphantly acquitted, unanimously, by the jury. As I was defending the
cause of thousands of leech-bled victims against a few vampires I was triumph-
antly carried on the shoulders of the workmen, with bengala-fires and bands
playing, but the next morning my father was discharged from the Dockyard
and lost his bread that was keeping us!!! I felt the shock tremendously but
did not discourage myself. I put myself in correspondence with Mr. Labou-
chere of the "Truth" of London, who not only published my contributions in
his very influential paper but brought the matter before Parliament, being an
M. P., and fought it out very bravely. A Commission was sent to Malta
and all my statements have been found to be true, my father was put to work
again, and several important reforms were introduced. But after a few
months my father was discharged again and forever!!! under the free and
glorious banner of liberal Britain!!!
The libel took place on the 7th of August, 1895. In September of the same
year, I took my degree of Bachelor of Arts from the Malta University after
obtaining for three years a 50 per cent in higher mathematics, physics, natural
history, philosophy, Latin, English and Italian literature and history. But
my father about that time was out of work and so I had to add to my already
overstraining work private lessons after my lectures, sometimes till 10 p. m.,
and plodded on in this very hard and anxious life for about two years in which
I have followed successfully the Anatomical, General and Pathological, the
Dissectional, the Physiological, the Obstetrical, the Surgical, the General
24 A STUDY OF PROLONGED FASTING.
Pathology, the Chemistry, the Bacteriology, the Materia Medica, the Thera-
peutical and the Pharmaceutical Courses. But as at that time I was under the
false impression that as I was working mentally very hard I had to eat more
and more, I used to stuff myself with a lot of meat and eggs and milk and these,
added to the great overstrain, shattered my nervous system down with a severe
shock of neurasthenia. My professors gave me the good advice to take a long
rest and to suspend my studies for a prolonged period of time. But my family
could not afford that for my father was not working all the time and I had to
work to live. So I took the warrant as a Pharmaceutical-Chemist after a
severe examination and was employed as director of the most important
pharmacy in Valletta (the capital of Malta), called "Mizzi's Dispensary."
I lived there for a year and my neurasthenia got a little better through enforced
rest.
But I was living away from my family and had to run into many expenses
to have my meals in hotels and I was always sleeping in the pharmacy not to
cross the sea late in the night and go home. So I employed myself in a phar-
macy at Cospicua, very near home, and lived there for about two years. My
father and mother at the same time started their business and were progressing
very prosperously. My wife, Lucia, lived just opposite, and we loved each
other. I married her on the 24th of April, 1900. She is the eldest daughter
of Doctor G. F. Inglott, Medical Officer to Government, Knight of the Pope,
and member of several literary and scientific academies and is considered as
the most clever obstetrician and gynecologist in Malta, enjoying a very wide
practice. So I was determined by him to start a pharmacy of my own, which
I did and the result was a very successful one, but a short time after the
Transvaal War broke out and as he is well conversant with the English lan-
guage was called by the military authorities in charge of the Military Hospital
and so all his time was absorbed in these exacting duties and could not take
care any more of his private practice. This lasted for over two years and at
last the pharmacy broke down and I had to remove to a wealthy country
district called Birchircara.
18th of May, 1912 (80th day of my fast).
Before going to live in Birchircara I had fought two great battles — one on
behalf of down-trodden and neglected Democracy and the other one advocating
the maintenance in our tribunals of the Italian language that has been the
means of our civilization since about 600 years. I have founded the first
"Malta Trade Union,,, of which I was elected President, with 700 members,
free schools, lectures, honest amusements, band, and carried it on successfully
for some time, but political intrigue not to encourage a labor party and not to
enlighten the lower class made it dwindle into nothingness and all my " love's
labor was lost."
Then I went to Italy, at my own expense, for about a month, to lecture
against Mr. Chamberlain's (England's Prime Minister at that time) edict that
the Italian language had to be cleared off from our courts within a lapse of
fifteen years. The movement had some good effect, because all the Italian
press was awakened and protested loudly and vigorously and Mr. Chamberlain
had to give up his Order in Council.
After creating in Birchircara a prosperous practice for my pharmacy I
wished to provide for the future, and as my neurasthenia was progressing
through the very close and sedentary life that I was conducting, shut up from
7 a. m. to 10 p. m., including holidays, I determined to secure a more easy
career — law. I entered the legal course and succeeded to obtain a warrant.
But to continue to carry on the pharmacy, to keep up my family, and to follow
AUTOBIOGRAPHICAL NOTES. 25
a difficult university course was a very severe test on my already shattered
nerves, and always under the false idea that to work very hard I had to overeat
and to stuff myself with as much protein as possible, I ruined my health to such
an extent that I was compelled to give up my pharmacy forever and dedicate
myself to the practice of law that offered more leisure and also better prospects
for me as I was and am still very popular and beloved by the people. Fortu-
nately enough to help me at the start of my legal career, I was offered at Sliema
(a beautiful summer resort in Malta) the management of a pharmacy with a
very good salary with the permission to absent myself during the morning
hours to go to court and plead my cases. So I went to live there and Miranda
Cordelia was born, while Jolanda Beatrice was born in Birchircara. My legal
practice prospered so rapidly that after a year I had to give up my pharmacy
management and dedicate myself entirely to the legal career that I am still
following at present.
When I thought to have fixed a solid basis for my family's subsistence I tried
again to do some good work for the cause of our trampled down and utterly
neglected lower classes. It has been always my ideal to enlighten them, to
help them to push themselves forward as the workmen of other more progres-
sive countries do, because although I have parted from their class my demo-
cratic soul was always with them. So I started the publication of a weekly
paper entitled "In Nahla" (The Bee) the scope of which was to instruct
in scientific, artistic, historical, and literary knowledge, as plainly and as enter-
tainingly as possible. The effort was a brilliant success because I had immedi-
ately the greatest circulation ever attained by any paper published in any
language in Malta. My wife cooperated herself very effectively because she
contributed, every week, some interesting article about the rearing up of
babies, hygiene, against the marriage of consumptives or between relatives, etc.
I have published in the same paper a historical novel "Is Sahhar Falzon"
(The Wizard Falzon) in which I have treated fully and faithfully all the
history of the first 60 years of the dominion of the Knights of Malta over the
Island from Lisleadam to La Cassiere. My intent was to teach to the people
its history not in the usual pedantic and monotonous way but enhancing it by
intermingling to it the attractive episodes of chivalry and love. In the third
part of the novel I have tried the scientific novel trying to popularize science in
a delectable and easy way as I have done with history, and as Falzon was a
Roman Catholic priest who was burned up alive accused of witchcraft, I
developed all the up-to-date positive knowledge about psychical science of
which I am an ardent and keen student. In many notes I have suggested the
best books and authors and described the most authoritative experiment for
those who wished to delve deeper into the matter. All the facts about Falzon
were gathered through a lot of poking in our archives amongst very rare
manuscripts of those past, dark, and barbarous ages. This novel was a great
success because I had to publish separately in three volumes comprising over
650 large pages and the edition was sold out very rapidly.
In the "Nahla" I have not only tried to instruct the lower classes but I have
fought hard also to defend their rights and to uplift my voice for the injustices
committed against them. Twice I was tried by jury for libelous articles but
twice I was triumphantly acquitted. The first time was on the 28th of
October, 1909, when Antonia Azzopardi, a murderer, was hanged. The
doctors in charge had executed their post-mortem examination so carelessly
that there was doubt that the man was buried alive only one hour after the
execution! I accused them of that in a very violent article, and all Malta
was in a devilish row about it. The Governor ordered the Chief Medical
Officer, who was responsible, to libel me and after a very hard struggle before
26 A STUDY OP PROLONGED FASTING.
the jury I have succeeded to prove that there were no positive and scientific
facts to prove that the executed man was dead when buried. This result
provoked a new law in Malta and now instead of burying the executed men
after only one hour from the execution as before, they watch them keenly for
24 hours, and as I protested also that it was barbarous to bury them in a sack
after that Justice had made its cold vengeance on a creature of God against
whose life she has no right at all, now they bury them in a cheap coffin.
The second trial was provoked by this fact. To communicate by means of
telephone in Malta you have to pay 60 cents, and the telephones are at the
Police Station. Poor people are supplied gratuitously by Government with
doctors, midwives, and medicines. At a village called Zeitum a very poor
woman was dying through post-partum hemorrhage. The midwife sent for
the doctor for assistance as she thought the case a fatal one. The doctor
happened to be in another village, and the policeman refused to call him
immediately before levying the tax of the telephone. The poor woman had
not the 60 cents to pay for it, and more than an hour was spent till they got
them from a distant sister. When the doctor arrived there was no more hopes
to save her and the poor victim of human brutality died leaving a husband and
six orphans. I published a violent attack against the police accusing them of
manslaughter and was libelled, but having proved to the hilt all the facts
stated, I was again acquitted triumphantly by the jury.
As you can see my "Bee" was really a "busy" one and played very well and
smartly her humanitarian and democratic mission. At the same time we did
not miss to advocate, and very ardently, " Fletcherism " and the Fasting Cure
for the cure of disease as also many other important dietetic reforms. Many
articles were also published on behalf of the idea of an international Language.
A lecture in Italian that I delivered in Malta several years ago advocating
Esperanto was published in it. About 25 years ago I learned Schleyer's
" Volapuk" that broke down, substituted by "Idiom Neutral," a more national
system. I follow my friend Rosenberger of St. Petersburg and learned it also
but had very little success. Then my dear friend's Dr. Zamenhof of Warsaw
"Esperanto" came in vogue and I learned it and took up arms in favor of it
very ardently. I have given in Malta free courses in the University, lectures,
founded societies and succeeded also to start the first female course in the
University in any branch of knowledge. Mrs. Levanzin was a great help to
me in this movement and now she is the first woman in Malta to enter the
University to follow a medical career. She is trying with all her efforts not
only to enlighten the female classes of Malta that are yet shrouded in mediaeval
darkness by publishing very instructive articles but also by setting them the
good example of opening for them new and prosperous careers. Esperanto
had a great vogue in Malta; I, with Mrs. Levanzin, took part in the Inter-
national Congress of Barcelona and there I was elected "President of the
International Association of Pharmaceutical Esperantists," editor of the
scientific Esperanto monthly, "La Vocho de Farmacustoj," and Corresponding
Member of the "Colegio des Farmaceuticos " (the oldest one in the world and
where the first pharmacopaeia was published) after my lecture in Esperanto on
the "Fungus Melitensis" by colleagues of over 30 different nationalities.
Now I am advocating "Ido" or Simplified Esperanto as I find that it is easier,
more logical, and cropped of all the errors and incongruities contained in Dr.
Zamenhof's system.
14th of May, 1912 (31st and last day of my fast).
I have also at the same time fought hard against much ridicule and prejudice
to found the first "Society of Psychical Studies and Research" in Malta of
which I am President. Honorary Members are Prof. Crookes, Russell
AUTOBIOGRAPHICAL NOTES. 27
Wallace, Lodge, Maxwell, Richet, Lombroso, Morselli, Carrington, etc. Now
another battle for Science and Humanity — Fasting. About two and a half
years ago, while I was over-eating, obese, neurasthenic, pessimistic and with a
shattered nervous system, I chanced to read in the "Contemporary Review" an
article about fasting. It was a flash of light that struck me vividly. It indicated
to me the right path to health and happiness and I followed immediately its
dictates with enthusiasm. I fasted for 8 days with very great benefit. Then
I procured all the possible literature in several languages about fasting and
prepared myself thoroughly for a whole year for a long and "conquest" fast.
I started that on the 1st of March, 1911, and Mrs. Levanzin did the same as
she had been suffering since several years from severe dyspepsia and insomnia
through over-eating. She broke her fast on the 33rd day and I on the 40th
with immense benefit to our health because our ailments disappeared. We
continued all our usual occupations during our fast and did never feel any bad
effects.
In the following August, cholera broke out in Malta and as a preventive
precaution I fasted again for 12, Mrs. Levanzin for 17, and my daughters for
several days each. I have cured Jolanda from a severe case of small-pox by
17 days of fasting and Miranda from a severe case of fever with 8 days.
Several other friends and parents underwent the cure of fasting under my advice
with marvelous effects. Enthused by these beneficial results, I determined to
fix a scientific basis to it by undergoing a thorough and seriously controlled
experiment under the direction of a physiologist of high repute and great
experience. I submitted the case to my friend, Professor Luciani, of Physi-
ology, of Rome, who studied Succi and published a good book on the "Physi-
ology of Fasting," and he suggested to me to come over to Boston at the
Carnegie Institution, * * * as the Institution was the best equipped in
the world for such an important experiment. I took up his suggestion and
crossed over 5,000 miles to undergo my fast, refusing any pecuniary remunera-
tion, only the expenses being defrayed for it. To-day is the 31st day and last
day of it, and I can simply tell you that it is a complete success. I am feeling
very well, very uplifted, and I wished to prolong it further, at least to 40 days,
because I do not feel yet any trace of hunger at all. But Professor Benedict
thought it already very expensive and fatiguing and bid me to break it to-
morrow. He only allowed me to prolong it for a day more, simply to beat the
record of the longest controlled scientific fast ever made. During the fast I did
not feel the least uncomfortable sensation except the bad taste of my coated
tongue, and the catarrh and congestion of my eyes that I had at the start
have nearly disappeared. I hope that a great benefit to my health shall
accrue from it.
GENERAL CHARACTERISTICS OF SUBJECT.
As will be inferred from his biographical notes, L. was a propagandist
with pronounced views on all subjects. He had had some legal training
and was inclined to be exceedingly contentious. His chirography was
excellent. He also had a good command of the English language, as well
as of Italian, French, Spanish, Maltese, and Esperanto. His familiarity
with the vagarious literature on fasting was astonishing, and led him
to make many suggestions indicative of a mind working upon a propa-
ganda for the supposed benefits to mankind to be derived from fasting,
instead of an appreciation of the true scientific value of a prolonged
fasting experiment. As an example of this, while he was unwilling to
undergo a series of carefully planned strength tests, he nevertheless
attempted some sensational strength tests of which he had read, such
as lifting up a man and holding him suspended for a moment or two.
He was a moderately well-nourished man, but his flesh was soft and
flabby. This was natural, as he was decidedly sedentary in his habits
and much averse to any muscular effort. It was hoped that measure-
ments of the fasting metabolism during muscular work could be made
with this subject by having him take a moderate amount of exercise
daily on the bicycle ergometer, but he absolutely refused to mount the
ergometer. He said he never rode the bicycle and thought it beneath
his dignity, and that although the bicycle was used in Malta, it had not
been employed by his people. As L. showed so strong an objection to
muscular activity, we were obliged to omit these valuable observations.
This subject called himself a vegetarian and frequently made a
statement to that effect during the fast, but his practice did not wholly
bear out his claim. He admitted that he ate meat in the European
restaurants and on the boat during his trip to Boston, but said that it
made him sick and uncomfortable, and that he was obliged to eat the
meat, since he could not get the food he wished. Of considerable sig-
nificance in this connection is his selection of food on the days preceding
the fasting period. On his arrival in Boston he was taken to a hotel
by one of the laboratory assistants and when given his choice of food
from the menu, he ordered a large steak covered with onions; on other
occasions he ordered salmon, pork, and lamb chops. During his stay
in the hospital after the fasting experiment was over, he again called
for a beefsteak. While probably not an excessive eater of meat, he
was by no means a vegetarian for several weeks prior to the fast.
The nitrogen found per day in the urine during the ten days preceding
the fast indicated that he was living on a fairly high protein level.
During the food days in Boston his diet was unrestricted and he was
repeatedly told that he could have whatever he wished to eat during
this preliminary period, except that it was preferred that the last meal
of the day should not be excessively high in protein to avoid the long
duration of the specific katabolic action of the protein.
28
GENERAL HISTORY OF FASTING EXPERIMENT.
L. left Malta the latter part of February 1912, visiting Rome,
Florence, Paris, and London, on his way to Liverpool. From the
latter city he came direct to Boston, arriving at the Nutrition Labor-
atory on the evening of April 10, 1912. Previous to his leaving for
Boston, he had been asked to collect in 24-hour periods the urine
passed during the trip across the ocean. While the conditions under
which he would be living were necessarily abnormal, it was hoped by
means of these specimens of urine to obtain some idea of the daily
nitrogen outgo of the body. Specific instructions were given him
as to the measurement, sampling, and preservation of the urine; as
he had had a thorough training in pharmacy, he was well qualified to
carry out the routine intelligently. The collection of the urine proved
somewhat troublesome, as his roommates on the steamer could not
appreciate the importance of the scientific test that he was to undergo.
During his stay in the laboratory, he lived the entire time in the
calorimeter room, except when he was taken out for a ride or to the
roof for a change of air and scene. The calorimeter room is large and
well-lighted and contains several calorimeters and respiration apparatus
of various models, thus permitting a considerable number of observa-
tions. The subject made his headquarters in a balcony of this room,
but during the night he slept in a sealed calorimeter.
The balcony in which he spent his time when he was not in the respi-
ration chamber or on the respiration apparatus was supplied with a
comfortable sofa, chair, and desk. A bottle containing a liter of distilled
water was given him, also a drinking glass, two urine jars, and a vessel
for defecation. The balcony floor is 2.5 meters from the calorimeter
room floor; the stairs leading to it have a rather sharp inclination, with
12 steps. (See Plate 3, figure E, page 31.) Before arranging to have
him occupy this balcony, he was asked if he would be likely to become
dizzy as a result of going up and down the stairs, but he replied that he
was never dizzy during his fasts. Indeed, he seemed to like the idea
of living in this balcony, as it gave him considerable freedom and yet
made it possible to watch him. Plate 1, figure A, gives a view of him
in a characteristic pose, writing at his desk in the balcony.
Throughout the fast he was under constant surveillance by various
responsible members of the staff and there were nearly always two or
three assistants on duty in the room. It was therefore impossible for
him to leave the balcony or to obtain food without its being known
at once. The watching problem in this experiment was very simple. As
L. was unknown in this country, he had no friends who would attempt
to bring food to him and all of those who came into communication
with him had a scientific interest in having the fast carried out to the
end of the time planned. While he might have drunk his urine, as did
the faster Jacques, it would have been practically impossible for him
to do this without its being known. Moreover, he had too much inter-
est in the fast to do anything of the kind, and we firmly believe that if
he had been surreptitiously offered food, he would have refused it.
29
30 A STUDY OF PROLONGED FASTING.
The three days preceding the fast — the so-called preliminary period —
were used to accustom the subject and the staff of assistants to the
apparatus and to the general routine, in order that the program could
be carried out as smoothly as possible, and without too great a demand
upon the time of the physicians and co-workers who made observations
upon the subject. His diet and daily life were under constant obser-
vation during this period, but he was free to choose his food and to
arrange his time as he desired when no tests were being made upon him.
It was necessary to be certain that L. was physically and psychically
a fit subject for the long fasting experiment. He was accordingly given
several rigid physical examinations by Dr. H.W. Goodall, of the Harvard
Medical School, and also underwent a psychical examination by Prof.
E. E. Southard, director of the Massachusetts Psychopathic Hospital.
The results of these examinations gave us every assurance that L.
was a suitable subject for this long fasting experiment.
The body- weight of this subject when he reached Boston was some-
what smaller than the initial body- weight reported for his earlier fast.
L. stated that his body-weight at the beginning of the previous fast
was excessive and that he desired to begin this experiment with his
normal body-weight. While this reasoning was scientifically correct,
his small weight caused us considerable anxiety, as it was feared that
he would be unable to endure a 31-day fast. Inasmuch as a fast of
7 to 10 days' duration would be of practically no value to us except as
a duplication of the earlier work, every possible arrangement was made
to adjust the conditions so as to prolong the fast; the subject quickly
found that if he made the statement that any particularly distasteful
routine or test would tend to "shorten the fast," it would be omitted.
On the other hand he took an intense interest in the outcome of the
experiment and had an almost religious belief in the benefits to
humanity to be derived from it. He enjoyed the distinction of having
so many observers studying him, and his peculiar appreciation of the
scientific value of the observations enabled us frequently to induce him
to waive his objections to any routine by a summary refusal to go on
with that particular test unless the routine were carried out. To his
credit it must be said that whatever idiosyncracies he exhibited at
times, he would, after reflecting on the importance of the experiment,
beg for the continuation of the complete routine.
To secure as much information as possible regarding the normal
metabolism of L., he was asked to sleep inside the respiration calo-
rimeter immediately on arriving in Boston. He was provided with a
comfortable bed, air mattress, and bed clothing, the bed comparing
well in size and comfort with a berth on an ocean steamer. Important
data regarding the normal metabolism of this subject were thus secured
for several nights before the actual fast began. Fortunately L. slept
very quietly, and when not asleep he remained very quietly in the
same position for long periods of time, thus greatly facilitating the
accurate measurement of the metabolism.
PLATE 3
E. L. on the Thirty-first day of the Fast, ascending the Stairs of the Balcony. The picture required
20 seconds ; there is no Evidence of Unsteadiness.
F. Clinical Examination by Dr. H. W. Goodall. This photograph was taken on the Thirty- first day
of Fasting, upon the Balcony occupied by L. during the day.
GENERAL HISTORY OF FASTING EXPERIMENT. 31
PROGRAM FOR RESEARCH.
The many observations and the large number of co-workers and
assistants made a carefully prepared program absolutely essential, so
as to use the actual available time of the subject and the co-workers to
the best advantage. The observations planned for each day were the
weighing of the subject after he had urinated and arisen; blood tests;
measurements of blood-pressure and the alveolar air; test for acetone
in the breath; records of rectal temperature and of pulse-rate; the
careful collection, measurement, and subsequent complete analysis
of the urine; and the apportionment and measurement of the water
taken. The subject entered the bed calorimeter about 8 o'clock each
night, remaining there until 8 o'clock the next morning, during which
time the respiratory exchange, water vaporized, and heat produced
were continuously measured. He was then taken out and his respira-
tory exchange was observed in three experimental periods by means
of the universal respiration apparatus. (See Plate 2, fig. C, page 19.)
Respiration experiments were also frequently made with the subject
at other times of the day and in varying body positions. The respira-
tory exchange when the subject was breathing an oxygen-rich atmos-
phere was determined several times, and a series of respiration experi-
ments was made by Mr. T. M. Carpenter when the subject was writing.
(See Plate 1, figure B, page 11.) In addition to the regular routine,
there was a rigid clinical examination by Dr. Goodall every second
day (see Plate 3, figure F), psychological tests were made by Dr.
H. S. Langfeld, and anthropometric measurements were taken by
Professor W. G. Anderson once a week. Every five or six days a
complete series of photographs was made of the naked subject. (See
Plates 4 and 5, p. 65.) Once a week his body was washed with distilled
water, the water used being preserved and analyzed. Among the many
incidental observations carried out during the experiment was a series
of X-ray plates on the thirtieth day of the fast by Dr. F. H. Williams
and a study of the flora in the colon on the thirty-first day by Dr. A. I.
Kendall. In clear, pleasant weather the subject was taken to the
roof or more frequently given a drive through the park system of Boston.
The program for a typical day — that of May 7-8, 1912 — appears below:
May 7. 7h 46m a.m. Bed calorimeter experiment ended.
Respiration experiment (three periods.)
Weighed.
Photographs taken.
Blood sample taken.
Blood pressure tests. Alveolar air.
Respiration experiment made with subject
writing (two periods.)
Psychological tests.
Respiration experiment (two periods).
Bath of distilled water; underwear changed.
Entered bed calorimeter.
Calorimeter experiment begun.
May 8. 7 50 a.m. Bed calorimeter experiment ended (5 con-
secutive periods.)
7h46ir
1 a.m.
8 15
a.m.
to 9h 17m a.m.
9 28
a.m.
10 00
a.m.
10 30
a.m.
1 40
p.m.
3 43
p.m.
to 4h 14m p.m.
5 00
p.m.
7 01
p.m.
to 7h 44m p.m.
7 50
p.m.
8 23
p.m.
9 34
p.m.
7 50
a.m.
32 A STUDY OF PROLONGED FASTING.
It will be noted that the regular clinical examination did not take
place on this day, as these examinations were made only on alternate
days. Furthermore, no drive was taken. As will be seen from this
typical program, the subject found himself fully occupied by the vari-
ous observations; in all of these he took a keen personal interest.
DAILY RECORDS OF FASTING EXPERIMENT.
Although a definitely arranged program was prepared, and for the
most part rigidly followed, the daily routine was varied by a large
number of extraneous observations, particularly in regard to the feel-
ings and moods of the subject, as well as observations made by co-
workers. To present these adequately, it seems desirable to give
them in the form of a daily record beginning with the arrival of the
subject at the Nutrition Laboratory. This record will be in the nature
of a "log-book," which will simply give the general history of the
experiment from day to day, with no attempt to describe the technique
or discuss the results.
The experimental day for most purposes ended with the completion
of the respiration experiment at about 9h 30m a. m. The last meal
was eaten at 6 p. m. on April 13, 1912; thus the true fasting period
began at 9h 30m a. m., April 14, or about 15 hours after the last meal.
In this daily history the personal observations of the subject on his
experiences during the night are always given in the notes for the next,
day, but all events up to the moment of entering the calorimeter are
recorded on the date of occurrence. While much that is said regarding
the previous fasts of the subject must, from a strictly scientific stand-
point, be considered as worthless, yet the trend of thought is not
without interest in interpreting the mental make-up of the subject.
PRELIMINARY PERIOD.
April 10, 1912. — L. arrived at the Nutrition Laboratory about 8 p. m.,
coming directly from the steamer and leaving his baggage on the wharf, but
bringing with him the samples of urine which he had collected for several days
on his passage across the ocean. He showed himself to be heartily in sympathy
with the plan of the experiment, and appeared to be a subject who would
cooperate fully in the experimental routine; he placed himself entirely in our
hands. As he had had no evening meal, he was sent to a hotel with one of the
laboratory assistants. This meal, which was of his own selection, consisted
of a large beefsteak with onions, one boiled potato, one portion chocolate ice
cream, and a glass of water.
He returned to the laboratory at 10 o'clock, and then reported that he had
had a very rough passage on the steamer, there being but a few hours of smooth
sailing on the third day out and a few hours on the last day. On the other
hand, he appeared to be in very good condition and showed no bad effects from
the discomforts of the trip. He maintained that during the previous year he
had lived almost exclusively on a vegetarian diet, taking occasionally milk and
cheese, very rarely eggs, and no meat. On the steamer, however, the menu
did not include the food he was accustomed to and he was compelled to eat
GENERAL HISTORY OF FASTING EXPERIMENT. 33
meat and "highly seasoned sauces" which he did not particularly care for.
During the previous year he had been living upon one meal a day, which was
eaten about noon, using the juice of an orange to satisfy his thirst when
necessary. He claimed that this limited dietary had been very beneficial to
his health. During the fast he wished to drink distilled water. He did not
especially like it, and usually drank hot water, but since some people believed
that there was nutriment in water, he wished to use distilled water so that there
could be no question as to his obtaining nutriment in this way.
In his previous fasts it had been his custom to carry on his regular business
and to go into court and plead his cases as usual, thus engaging in a not incon-
siderable amount of muscular activity. He snowed a decidedly intelligent
interest in the experiment, as was indicated by his asking if the eyes should not
be examined by an eye specialist, for he had found that as a fast progressed,
the eyesight improved considerably, though normally he had very poor eye-
sight. He also thought it important to study the blood and seemed much
gratified when he was told that both eye and blood tests would be made. The
important role which he would play in the experiment was emphasized to him
and he was shown that the efforts of the laboratory staff would be of no avail
without his full cooperation. His attitude toward the experiment and under-
standing of the requirements showed him to be by far the most intelligent
man that has ever been studied as a fasting subject.
When discussing the question of defecation during a fast, he made the state-
ment that in some of his long fasts he had defecated only once or twice. Often
he defecated shortly after the beginning of the fast and then not again until
after the fast was over, but after beginning eating he defecated quite regularly.
In one fast he said that he did not defecate until the twenty-seventh day.
As there was no time that evening to discuss with him at length his past
hiBtory and the details of the fasting experiment, he was taken down to the
calorimeter laboratory, where he urinated, removed all but his underclothing,
and prepared to go into the calorimeter. The stethoscope was adjusted, and
the rectal temperature taken with a clinical thermometer, which was left in the
rectum three minutes. He drank a glass of water and was then placed inside
the chamber of the calorimeter. After he had been shown how to use the tele-
phone and the signal bell, a black cloth was placed before the window so that
the electric light would not disturb him and the calorimeter was then sealed.
Even on this first day the subject was inclined to talk about the method of
breaking his fast, saying that he was accustomed to do this by taking the juice
of one or two lemons, and afterwards orange juice, to which he sometimes
added sugar. As was seen later, such a method for breaking the fast proved
to be disastrous to our predetermined plan of securing data after the fast.
April 11, 1912. — The calorimeter experiment for the previous night was
uneventful and ended at 8h 02m a.m. The measurements were made in three
consecutive periods, the idea being to secure observations during the latter
part of the night and thus eliminate, if possible, the influence of food taken
during the evening. The subject kept very quiet most of the time, and proved
exceedingly tractable and intelligent. The importance of lying quietly inside
the respiration chamber had been impressed upon him and we have rarely had
a subject who lay so quietly for so long a time. About 5h 30m a. m. he tele-
phoned to ask if everything were all right, reporting that he felt very well but
had been awake for some hours. He was instructed to ring an electric bell
every few minutes by pressing a small push-button inside the chamber, so
as to show us that he was awake. He rang this bell regularly throughout
the rest of the experiment, beginning at 5h 30m a. m.
When the calorimeter was opened, a strong odor of onions was apparent,
34 A STUDY OF PROLONGED FASTING.
doubtless due to the fact that he had eaten beefsteak and onions the evening
before. L. reported that for the first two hours after he entered the chamber
he was very warm, but later became cool and slept comfortably. Since the
temperature inside the calorimeter seldom varies by 0.1° C, such an observa-
tion serves excellently to illustrate the futility of placing any weight on personal
impressions. The interior of the respiration chamber reminded him of his
cabin on the steamer. It was impossible to obtain records of the pulse-rate
for about one hour after 5h 30m a. m., this being due to some change in the
position of the stethoscope. Later the subject was able to readjust it and the
records were obtained thereafter.
When L. came out of the chamber, shortly after 8 a. m., he urinated and
immediately the experiment with the universal respiration apparatus was
begun. (See Plate 2, figure C, page 19.) This experiment consisted of three
15-minute periods.
Almost the entire day was spent by the subject in familiarizing himself with
the experimental routines. After the respiration experiment was over, several
tests of the blood pressure were made, and samples of the alveolar air taken by
the Plesch and Haldane methods. Professor W. G. Anderson made a series
of physical measurements and attempted the routine strength tests, but
was not able to obtain these, owing to the disinclination of the subject. It did
not seem advisable to complicate the program by taking photographs on this
day. An examination of the blood was made, also a most careful clinical
examination. L. then took a hot bath and went out with Mr. H. L. Higgins
for the first meal of the day. This meal, selected a la carte, consisted of one
portion of scallops and tartare sauce, one portion of roast lamb and mint sauce,
two portions of mashed potato, three rolls, two portions of butter, and one
portion of custard pie.
On returning to the laboratory in the afternoon, he occupied himself in
writing letters and in talking with different members of the staff until about
4h 30m p. m., when the first series of psychological tests was made. The visual
acuity test was not very successful, as L. has a very short vision and the letters
used were so small that new ones had to be secured. The chief value of the
test on this day was to familiarize the subject with the routine. L. continued
to have a keen interest in the success of the experiment and cooperated in
every way except in the strength test.
In order to make sure that the last meal of the day should contain only a
small amount of protein, I went personally with him to the restaurant. His
supper at this time consisted of half a grapefruit, to which he added quite a
little sugar, a plate of split-pea soup (this containing practically all of the
protein in the whole meal), two or three slices of bread and butter, one portion
of stuffed tomatoes, one of fried sweet potatoes, another of white potatoes,
one dish of strawberries and cream, and some strawberry ice cream. He
returned to the laboratory about 9h 30m p. m. and prepared for the night in
the calorimeter.
While I was with him in the afternoon and evening, L. gave me considerable
information regarding his previous fasts. Although the unscientific nature of
these personal impressions is recognized, it seems desirable that they should
be recorded. His observations for the most part had to do with defecation,
feelings of hunger, and changes in body-weight. During his fast of 40 days,
which he reports as having been broken on April 10, 1911, he defecated twice
during the first two days, and again, according to his remembrance, on the
twenty-fifth day. The feces on the latter day consisted of a small amount of
blackish or very dark brown material, with a yellowish white mucus. He
defecated again the evening of the day on which he broke the fast. He
GENERAL HISTORY OF FASTING EXPERIMENT. 35
noticed that there was a large amount of gas in the intestines. His theory was
that the lower portion of the bowels was clogged with feces, the rest of the
intestines being filled with air, and that when he ate, the air was compressed
by the food passing along the intestines, this compressed air distending the
bowels and producing much colic. Thinking that it might be more advan-
tageous for him to empty the lower bowels by an enema before beginning the
fast, I suggested this to him, but he preferred not to take an enema unless
it were scientifically necessary, as he believed in natural rather than forced
movements. He had never had any distress from defecation or from inability
to defecate such as that experienced by the subject of the fasting experiments
made in Middletown, Connecticut. He also said that he usually had no hunger
pains, but felt somewhat hungry. His weak point was his throat, which fre-
quently troubled him considerably. On the twentieth day of his long fast
his throat was very dry and slight traces of blood appeared. On the second
day of the same fast he noticed that he was a little irritable. In the first
part of a fast he was usually somewhat depressed, but not to any great degree,
and after the third day he would be very happy, with no desire for food.
During one of his Malta fasts he went to the table every day and watched
his children eat, and, in fact, prepared delicacies for them. The first indication
of a desire to break the fast was usually shown by an intense craving for an
acid and it was his custom to take first a lemon and then an orange. All
throughout his earlier Malta fast the color of his tongue was unnatural.
The changes in body-weight during his previous fasts were most significant.
The initial weight of the 40-day Malta fast was much greater than when he
came to Boston and even at the end of the fast he weighed more than he did
on this date. His contention was that when a man fasts with a large amount of
fatty tissue, he is "really not fasting but simply draws upon body tissue, so
that the only true fasting is when a man begins the fast with a normal weight."
If I had had the decision of the matter, I should have preferred to have him
begin his fast with a larger amount of fatty tissue than he had. Nevertheless
the results have more interest from the fact that he did not have this excessive
amount of fat.
He continued to keep up his interest in the results and wished to know them
from day to day, but I pointed out to him that as the fast progressed, if the
results were abnormal in any way and he should be told of them, he might
instinctively and unintentionally attempt to alter the conditions so as to meet
the variations that we should find. It was suggested to him that he should
spend all of his energy and interest on familiarizing himself with the technique
and pay little attention to the results that we found, until the fast was over.
On this day he discussed extensively the writers on fasting, more especially
those which would be designated as the semipopular writers. He pointed out
rathei naively that most of the writers on fasting wrote of the experiences of
others, but never fasted themselves. In speaking of the fasts carried out by his
wife, he said that during a prolonged fast menstruation disappeared entirely.
L. was quite impressed with the increased mental activity and power of
working during a fast and maintained that any student who is to take an
examination should know better than to take an excessive amount of food.
He cited instances where he wrote poetry and continued his literary work
during his fast and said that he was conscious of a considerably increased
efficiency. According to his experience, the hours of sleep decreased somewhat,
and although he began work every morning during the fast at 5 o'clock, he
felt but little fatigue.
He gave no special attention to the amount of muscular exercise taken, but
walked to and from his business, pleaded cases in court, went to his club, and
walked about the street as usual, this exercise continuing some six or eight
36 A STUDY OF PROLONGED FASTING.
hours each day, but he took no very long walks. He said that at one time,
when he was being jeered at in regard to his fasting by one of his clubmates
who said that he was losing strength, he suggested that they test his strength
with a hand dynamometer, and he was able to show more pressure on the hand
dynamometer than the man who ridiculed him. He did not believe in Succi's
theory, however, that there is an increase in muscular strength during a fast.
April 12, 1912. — The records of the night observers showed that L. slept
rather quietly from 9h 40m p. m. on April 11 until 3h 33m a. m. on April 12,
after which time he rang the bell at intervals from 5 to 10 minutes. At 5>,30m
a. m. he telephoned that it was necessary to urinate. For this he used a 500
c.c. urine jar which had been placed inside the calorimeter the night before,
but notwithstanding its size it was not sufficiently large to contain all of the
urine he desired to pass. The subject reported no particular discomfort from
the fact that he could not completely empty the bladder.
At noon on this day, L. went to dinner with Mr. T. M. Carpenter. His
dinner, selected by himself, consisted of one portion of broiled salmon and a
small quantity of green peas, two pork chops saut£, three heaping tablespoon-
fuls of mashed potato, a dish of sliced cucumbers and tomatoes, three small
rolls, two small pieces of butter, one portion of strawberry ice cream, and two
glasses of water.
At 7 p. m., he took supper, likewise with Mr. Carpenter, eating one large
portion of macaroni, with apparently but little cheese in it, one large portion of
fried sweet potatoes, and one large portion of fried eggplant. In addition, he
ate two slices of French bread, two portions of butter, one portion of chocolate
ice cream, one-half dozen macaroons, and drank five glasses of water. He
seemed to enjoy this meal very much.
Foi the first time on this day he showed apprehension in regard to the
experiment, maintaining that the number of tests made with him would tend
to shorten the fast, as it required concentration on his part to cooperate with
the different observers and this concentration used a certain amount of his
energy. On the other hand, he suggested that a series of anthropometric
measurements should be added to the tests planned for him in order to show
that he was a normal individual, citing the fact that one objection that is
made to professional fasters and men who fasted any great length of time is
that they are not normal people intellectually and for that reason the results
obtained with them could not be considered normal. He considered these
measurements of the greatest importance and thought they could be obtained
by taking a photograph showing the angle of the face, width of the head, etc.
Since he desired them, a series of these photographs was subsequently taken.
April 13, 1912. — The records for this day show that the subject entered the
calorimeter chamber at 9h 40m p. m., April 12, and apparently was quiet until
12h 17m a. m., April 13, when he began ringing the bell and continued this more
or less regularly, i. e., several times an hour, throughout the remainder of the
night. He telephoned at 4h llm a. m., and again at 5h 10m a. m., stating
that he had slept little but felt very comfortable. He urinated at 4h llm a.m.,
passing 595 c.c. of urine. On coming out of the calorimeter, he had consider-
able to say regarding the dreams that he had had; between 9h 40m p. m. and
12h 17m a. m., he slept very well without dreaming. After that period his
sleep was much broken, and he had a number of dreams, one of which was
accompanied by a seminal emission. He explained this by saying that his
supper the night before contained too large a proportion of carbohydrates,
which heated his blood and made him very uncomfortable. He almost imme-
diately remarked that he had been very comfortable inside the chamber and,
in fact, rather enjoyed being there.
GENERAL HISTORY OF FASTING EXPERIMENT. 37
A considerable number of photographs were taken of the nude subject, a
procedure which evidently gave him much pleasure (see Plates 4 and 5).
At llh 16m a. m., he urinated and defecated, and later went out to his midday
meal with Mr. Higgins. He selected pea soup with three crackers, one gen-
erous portion of fish (finnan haddie), mashed potatoes, two lamb chops, one
portion of French fried potatoes, four slices of bread with butter, and one
portion of strawberry ice cream and fruit syrup (college ice).
In the forenoon of the third day I told the subject that it was important
that we should have some idea as to how long he expected to fast, for he had
frequently made the statement that so many tests upon him would tend to
shorten the fast. He was shown that a short fast would have no interest for
us. He replied that he expected to fast until his body-weight fell below 100
pounds (45.4 kilograms). Inasmuch as his initial weight was 134 pounds
(60.6 kilograms), he thought he would lose about a pound a day, which would
make the fast approximately 30 days long.
He brought me two sealed bottles of the liquor that Succi uses in his fasts to
allay the pangs of hunger at the beginning of a fast. One of these bottles he
presented to the Laboratory, and asked me to keep the other for him until the
fast was concluded, as he did not wish any one to say that he took the liquor
and that it helped him to carry out the fast. The subject sympathized fully
with the strictest surveillance and was much impressed with the fact that so
many co-workers were watching and studying him.
In order to obtain an expert opinion regarding the mental state of the subject
at the beginning of the fast, arrangements were made by which Dr. E. E.
Southard, of the Massachusetts Psychopathic Hospital, could examine him
at 3h 30m p. m. on this day.
L. was excessiuely voluble, continually emphasizing the importance of making
measurements of his face and head, the length of the ears, and similar measure-
ments, as he desired that a careful study should be made in order to prove
that he was a normal man, and not erratic and abnormal. He also advocated
with great persistency the study of the influence of fasting on the sexual organs.
At 6 p. m. he went with Mr. Carpenter to a local restaurant for the last meal
before beginning the fast. This meal consisted of bananas and cream, straw^
berry shortcake, ice cream, and three glasses of water.
In the evening he took a hot bath, after which he was sponged with distilled
water, and then put on a union suit and a pair of white stockings, both of which
had been previously thoroughly washed and rinsed in distilled water and dried.
The union suit absorbed the perspiration, so that a measure could be obtained
of the nitrogen eliminated through the skin in the form of urea or organic
nitrogenous material. After drinking a glass of water, he entered the bed
calorimeter at 9h 44m p. m.
FASTING PERIOD.
April 14,1912 (first day of fast). — According to the experimental records,
L. rang the bell to show that he was awake from time to time during the night,
there being one hour of quiet between 10h 17m p. m. and llh 29m p. m. and
another hour between llh 48tt p. m. and 12h 49m a. m., but from 2h 06m a. m.,
April 14, he rang the bell more or less regularly until 5h 25m a. m., when for
two hours he remained quiet and was apparently asleep. At 2h 06m a. m. he
telephoned that he was compelled to urinate. This disturbed him, as he had
been sleeping very soundly. Later he suggested that if he could drink water
during the day and not have to urinate during the night, he would feel more
comfortable. His program was therefore arranged subsequently, to include
the taking of the greater part of the water during the daytime.
38 A STUDY OF PROLONGED FASTING.
When he came out of the calorimeter, he reported that he felt fairly com-
fortable. Some of the air had been let out of his air mattress, so that he found
it easier than formerly. After the respiration experiment was over and he had
washed his hands and face, he went up into the balcony and a thorough
examination was made of all his clothing and baggage. This was done to
make sure that nothing was hidden, such as food tablets, or alkaloids of any
kind, in a form that might be sewed into his clothing, concealed in hollow
books, in the tips of his shoes, or in some similar place. As he had had previous
experience as a pharmacist, it seemed desirable that such precautions should
be taken. Even the linings of his clothing were examined, but nothing was
found, only two small cakes of soap being removed from the balcony. Besides
his clothing and numerous testimonials from many of his Malta associates, a
large part of his luggage consisted of so-called "fasting literature"; of particu-
lar interest to us were the materials he had collected in his visit to Succi on his
way to Boston.
He had never been in the habit of using his eyes for reading or studying in
the evening, but finished his work by 6 or 7 o'clock and rose early in the morning.
It had been his custom to spend the evening with his family or go out for a
walk, or to some place of amusement. Anticipating the tediousness of the
fast, we sought to interest him in some simple game as solitaire, checkers, or
whist, but he refused all of these and preferred to retire early, entering the
respiration calorimeter even earlier than we had planned.
Much of the first day he talked about how well he would feel as a result of
his fast, how happy he was to think that he was beginning it, how he would be
relieved from the necessity of eating and drinking, so that the time he now
spent in this way could be devoted to higher mental work. As he usually felt
so much better during a fast, he expected that these fasting days would be
among the happiest of his life. He considered that he bore the tests on this
day much better than previously, showing a greater power of concentration.
In the psychological test he observed that the ticking of the metronome
seemed louder than in previous tests and attributed this to the fact that during
fasting his hearing was always more acute. Inasmuch as this observation was
made when he had omitted but one meal — the meal in the middle of the day —
the weight which should be given to this observation is easily estimated.
Although the weather was dull, which usually depressed him, he was perfectly
certain that throughout this day he felt much better both mentally and physi-
cally than the day before. During the evening he was unusually lively and
cheerful, sang and whistled quite a little, and said he felt like dancing.
Although the temperature on the balcony was 22° C, he began to complain of
the cold and in the afternoon wore a blanket wrapper and bedroom slippers in
addition to his regular clothing. (See Plate 1 , figure A, page 11.) Subsequently
he wore his heavy-weight suit of underclothes over the union suit and stockings,
which had been washed in distilled water.
He said that he never used alcohol in any form and in the afternoon com-
plained of having to drink too much water, remarking that it would shorten
the fast if he drank so much, as it would wash out the salts. While there was
a legitimate foundation for the statement that the distilled water might affect
the salt metabolism, it was evident that L. had discovered an efficacious way of
obtaining anything that he wanted by bringing forward the argument that it
would "shorten the fast." It was decided that if the amount of water given
him to drink caused him discomfort, he could lessen the amount. He pre-
ferred to drink only when thirsty, but that morning had taken water without
feeling the need for it, and had had absolutely no hunger all the day. The
water tasted much better than he had expected it to taste, as he usually dis-
liked the taste of distilled water.
GENERAL HISTORY OF FASTING EXPERIMENT. 39
In preparing him for the respiration chamber at night, it was difficult to
adjust the stethoscope so as to hear the pulse-beats through it clearly. The
best results were obtained by placing the stethoscope about 1 cm. above the
left nipple and 2\ cm. toward the center line of the chest. He urinated at
6h 20m p. m. and thought he would not urinate again during the night. By
this time he had become thoroughly accustomed to the calorimeter, showing
no anxiety regarding his stay in it and having the greatest confidence in those
who had charge of the experiment. He entered the respiration chamber at
8h 48m p. m.
April 15, 1912 {second day of fast). — From the experimental records and the
report of the subject, L. evidently had a very comfortable night, ringing the
bell only occasionally and sleeping much better than he had any night since
he had come to Boston. When he left the apparatus, he reported himself to
be in excellent condition. There was no noticeable odor when the calorimeter
was opened. He felt no pain or sensation of hunger, and very little thirst.
During the forenoon he wrote busily and several long letters were mailed for
him at noon. L. is ambidextrous, using his right hand for writing and left
hand for work, that is, in taking the dynamometer test, he always used the left
hand first. In conversation, he invariably led up to the discussion of the
innumerable "popular" books on fasting, with which he was remarkably
familiar. An ingenious argument against fasting, which he reported as having
been given him by a prominent American vegetarian, was that in a fast a man
became a flesh-eater, as he existed upon his own flesh.
At 4 p. m., 250 c.c. of water were taken from the 1 liter of distilled water
contained in his bottle, reducing his apportionment of water to 750 c.c.
Subsequently this amount was given to him daily. The rectal thermometer
was used for the first time during the night of the second day of fasting, L.
inserting the thermometer himself. Previous to entering the calorimeter, the
subject was in unusually good spirits, singing and talking a great deal.
April 16, 1912 {third day of fast). — On coming out of the apparatus in the
morning, L. said that he wore the rectal thermometer all night and suffered no
distress, although at times it troubled him somewhat. The observer reported
a good series of temperature measurements throughout the night. L. evidently
slept more soundly than usual, ringing the bell only occasionally. No special
odor was noted when the chamber was opened.
After the subject had been weighed, several photographs were taken of him
nude in various positions, corresponding approximately to those taken on
April 13. After he was dressed, at his request several photographs were taken
of the head to show the facial characteristics. His innumerable suggestions
displayed a worthy interest in the experiment, though many of them had to
be disregarded.
On this day the drinking water and the urine bottle were placed on a table
at the foot of the stairs leading to the balcony, the subject notifying the
assistant whenever he wished water or the urine bottle. In this way all
possibility of his drinking the urine, as was done by the subject of Paton and
Stockman, was eliminated. Although L. was less cheerful than he was the
day before, he was by no means depressed, spending a considerable part of the
forenoon in writing in his diary. He was particular ly cautioned against writing
any of his experiences or sending out information unauthorized, as misuse
would be made of the material which he gave out before the fast was completed.
We had expected that he would lie down occasionally and had provided a
well-upholstered couch for his use, but he sat up practically all day. It seemed
desirable, therefore, to study his metabolism in this position. Accordingly in
the afternoon he came down from the balcony and a respiration experiment was
40 A STUDY OF PROLONGED FASTING.
made with him in two experimental periods, while he was sitting in a com-
fortable chair.
During the day the subject reported that he had no bad feelings of any kind.
He felt slightly irritable, but considering the confinement, he was remarkably
free from this feeling of irritation, much more so than in his last fast. The
disinclination to exert himself in any way which might cause the slightest
strain was shown on this day when he complained that the dynamometer hurt
his right hand and he did not dare to press it as hard as he would have liked
to. He passed no urine from 8h 05m a. m., when he came out of the bed
calorimeter, until 8h 05m p. m., just before entering the chamber, although he
drank all of the 750 c.c. of distilled water, taking the last portion at 5h 50m p. m.,
when he finished the psychological tests.
In the evening, just before entering the calorimeter, he said that the rectal
thermometer irritated him considerably and had kept him awake more or less
the night before. He was aware of its presence every time he woke up and
during the day the anus had been somewhat irritated. He asked if the temp-
erature could not be taken every other night instead of every night, as had been
planned, and the thermometer was accordingly not used on the night of
April 16-17.
April 17, 1912 {fourth day of fast). — Both the experimental records and L.'s
report indicate that he slept better on the night preceding this day than he had
any night thus far passed in the chamber. The day was uneventful. He said
that he had no great thirst and he drank the distilled water more from a sense
of duty than from any desire for it. He did not leave the balcony during the
day except for the psychological test at 4h 50m p. m. At one time during the
forenoon he was seen to hold up a light man, weighing but 125 pounds (56.7
kilograms) for a few seconds. He appeared pleased with this supposedly
"sensational feat" and was quite unappreciative of the caution to conserve
his strength for tests of muscular strength that could be measured, tests that
he had repeatedly refused to make.
On this day he wore the stethoscope most of the day, an assistant counting the
pulse-rate at a distance from him and unknown to him for the greater part of
the time. It was hoped that more or less continuous records of the pulse-rate
could be obtained in this way. In preparing him for the calorimeter at night,
I personally inserted the rectal thermometer and he reported that it did not
hurt him in any way. He believed that, when putting it in himself the first
night, he must have irritated the anus somewhat, thus causing the subsequent
discomfort. The subject entered the calorimeter for the night at 8h 19m p. m.
April 18, 1912 {fifth day of fast). — The subject reported in the morning that
he slept quite well during the night, but not so well as he did the night before.
He thought he went to sleep about an hour after entering the calorimeter and
woke up several times during the night. He said he was entirely comfortable,
the rectal thermometer giving him no trouble. The spring supporting the
bed had been somewhat weakened in order to obtain a greater sensitivity for
the apparatus recording the muscular activity. The records show, for the
most part, a remarkably quiet night, so that long calorimeter experiments of
10 or 11 hours will be perfectly comparable so far as the muscular activity is
concerned. In this respect the subject was exceptionally well adapted for an
experiment of this kind.
Notwithstanding the dull weather, L. said that he felt very well, with no
loss of strength, headache, feeling of hunger, or of apprehension. He was not
usually troubled with headache during a fast. When his measurements were
taken, his height was found to be 1.707 meters. He said that his maximum
weight was 14 stone or 196 pounds (88.9 kilograms). His initial weight for
GENERAL HISTORY OF FASTING EXPERIMENT. 41
the Malta fast was 12 stone and 3 pounds, or 171 pounds (77.6 kilograms).
During this fast he lost 37 pounds (16.8 kilograms), and thus vveighed at the
end of the fast 134 pounds (60.8 kilograms). These measurements included
the weight of the clothing worn and were taken with ordinary scales, but
probably represent average weights.
In the evening L. talked about his work in Malta and his interest in the
numerous fasting books which he had read. He became quite excited in
talking of three legal cases in which he was involved in Malta, speaking with a
good deal of vigor and enthusiasm. There was not the slightest evidence of
his having fasted for 4§ days. After inserting the rectal thermometer, the
subject entered the bed calorimeter about the usual time.
April 19, 1912 (sixth day of fast). — The subject was sleeping very soundly
when the calorimeter experiment was ended and on leaving the apparatus,
reported that he had had a very comfortable night. He seemed very bright
and quite like himself. According to the records of the observers, the subject
slept very well throughout the night until 5h 30m a. m., when he rang the bell
intermittently for a short time. He still continued to have no discomfort, no
sensations of hunger, and no apprehensions, and felt very much inclined to
mental work. He said he did not expect to have such an excellent fast and
seemed to be content with the progress of the experiment. On this day the
subject spent considerable time in writing. To provide data for subsequent
computation of the total energy transformation during the 24 hours, it seemed
desirable to measure the respiratory exchange while the subject was writing.
This type of experiment, which had never been attempted in this laboratory
before, proved to be very successful, the subject sitting in a comfortable chair
and writing as usual. (See Plate 1, figure B, page 11.)
April 20, 1912 (seventh day of fast). — When the calorimeter was opened, no
odor of acetone was apparent, and no unpleasant odor of any kind. Thus far
in the series of calorimeter experiments, the only unpleasant odor which has
been noted was the strong smell of onions on the morning following his evening
meal of beefsteak and onions. L. said that he was in excellent condition;
intellectually he was very acute. He slept well throughout the night with no
discomfort and rang the bell only a few times. With the sunny weather the
subject's spirits rose and he reported himself as feeling very happy and
uplifted; he sang at times during the morning. A barber cut his hair and
beard on this day, the trimmings being saved for analysis.
After taking a bath in the evening, he entered the calorimeter. He told
one of the assistants that he felt very drowsy and expected to go to sleep
almost immediately.
April 21, 1912 (eighth day of fast). — On coming out of the apparatus, the
subject reported that he did not go to sleep as quickly as he had expected to;
he lay awake for 2 hours and also woke up in the night once or twice. He
slept well, however, and had several dreams which he reported to Dr. Langfeld.
L. asked if he did not have fever during the first 2 or 3 days of the fast, stating
that his wife during the first days of her fast had quite a high fever. When
he was told that he had had no fever he was much surprised. His theory was
that "in the first days of the fast, the blood rid itself of all impurities, the
burning process producing a high fever."
In the afternoon, the subject went up to the roof, using the elevator; he
remained there for about 1? hours, sitting in the sun and chatting on general
subjects. He said considerable about his experience with Esperanto and about
many of his Maltese customs. His voice was strong, and he seemed to be
very well and bright. He came down from the roof by the elevator to the
office and dictated into the dictaphone a statement regarding his feelings.
The dictated report is as follows:
42 A STUDY OF PROLONGED FASTING.
"I felt very bright this morning because the sun was shining very brightly
and the day is so fine and so sunny that it looks very much like one of my
Maltese days for which I am very desirous. This afternoon I have been on
the roof of the laboratory so as to enjoy the sun, which I love so much because
I was born and bred in it, and I do not feel very fatigued at all or any appre-
hension or any fear. I feel very hopeful that my fast will be not only a
very long one but a very successful one, and I feel no pangs of hunger whatever
or any other sensations in my stomach, and I think that my health is progress-
ing because the catarrh of the head in my nose and in my pharynx before the
starting of the fast has nearly disappeared and I can aspirate very easily
through my nostrils, a thing which I could not do very easily before, and also
I feel that my eyes are getting less congested and better. I do not feel any
dizziness at all, and I have never felt it, neither when I lean down to get up
something that falls from the hand; in fact, I can come down the steep stairs
of my balcony apartment very easily without taking hold of the railing and
without any necessity of being helped by any one. My legs feel very strong
and I do not feel any numbness in them."
The subject went from the office to the calorimeter laboratory on the elevator
and drank a glass of water, but did not return to the balcony, as it was almost
time for the psychological tests. The rectal thermometer was used during
the bed calorimeter experiment, observations being made every 10 minutes
during the night.
April 22, 1912 {ninth day of fast). — On opening the calorimeter chamber no
odor was noticeable. The subject reported a very comfortable night and said
that he felt very well, except that he had had a bad taste in his mouth for
several days. He slept very quietly, ringing the bell only three times.
During the day he had visits from a number of scientific men, with whom he
talked considerably. He was particularly interesting and vivacious in dis-
cussing his experiments with Professors Cannon and Folin and with Dr.
Cathcart. He reported that he did not now feel so much inclined to mental
work, explaining this fact on the ground that "the brain was occupied in
eliminating the impurities from the blood."
April 23, 1912 (tenth day of fast). — This morning L. said that the preceding
night was the best that he had had in the calorimeter. He was very quiet
during the night and rang the bell but a few times. He still complained of the
bad taste in his mouth.
In the afternoon a respiration experiment of two periods was made with the
subject sitting in a chair. In this experiment certain abnormal values were
obtained for the pulse-rate and the respiratory quotient, and Mr. Carpenter
thought that there was a slight irregularity in the action of the heart. Since
the man was essentially different as to his total metabolism from the day
before, Dr. Goodall was asked to examine him during Dr. Langfeld's tests.
Dr. Goodall noted the heart rate before, during, and after the dynamometer
test, and pronounced the subject to be in an excellent condition.
L. had been less self-assertive for the week previous, much more inclined
to cooperate in the experimental routine, and a little more reconciled to the
fact that the observers were familiar with their technique. He was more
quiet and less argumentative than formerly and was at times considerably
depressed, spending much of the time in thinking of home. In conversation
he seemed very intelligent and interested in everything, and by no means
lethargic. He professed to have no discomfort from the rectal ther-
mometer, although he stated frankly that he feared its use very much before
coming to America. He also was becoming accustomed to the respiration
chamber and did not find it at all irksome.
GENERAL HISTORY OF FASTING EXPERIMENT. 43
During the respiration experiment in the afternoon he spoke of the delightful
sensation that he had in these experiments, saying that the sound of the rotary
blower did not disturb him, but on the contrary it seemed soothing. This
effect is not unusual with other subjects of respiration experiments, the purring
sound of the blower frequently producing drowsiness, so that it is difficult for
the subjects to keep awake.
April 24, 1912 (eleventh day of fast). — Although L. rang the bell but once an
hour throughout the night, he maintained that he did not sleep well and conse-
quently felt somewhat irritated when he came out of the apparatus. Although
the weather was fine, he was quite depressed during much of the day. He said
that it was the twelfth anniversary of his marriage, and he spent the whole day
thinking of his family and his home in Malta.
In the respiration experiment in the afternoon, with the subject sitting,
the nose-pieces troubled him somewhat, and he became very irritable before
the experiment was finished. He was, in fact, quite intractable for a time,
complaining bitterly about the rapidity with which the experiments were
made and how frequently he was passed from one observer to another. In
discussing the matter, he said that it was not the experiments themselves which
troubled him and he did not particularly object to the respiration experiments,
but that when he had finished and opened his eyes, he saw the next observer
waiting for him, and he thought the tests made upon him were not far enough
apart. On beginning certain of the psychological tests, he stopped and
refused to go any farther, saying that he wanted to have 15 minutes' rest.
He was very irritable and said that irritability was bad for him. At such
times, the best method to use with him was to say that a part of the experi-
mental routine would be omitted. For instance, he was told that the alveolar
air test would be omitted, but he was much distressed and insisted that it be
made as usual. Again, the evening before, one of the assistants was directed
not to insert the rectal thermometer unless the subject insisted upon it, but to
say to L. that we did not wish to tire him with too many observations. The
subject was much disturbed by this omission and insisted that the thermometer
should be inserted. He was frequently inconsistent, at one time saying that
there were too many experiments, too much going on, too many people about
him, too many questions asked as to his feelings, and then almost immediately
he would complain of being left too much alone to brood and to think of his
family. He complained bitterly of his inability to get out of doors. He
enjoyed very much his stay on the roof on April 21, but the weather had been
raw and cold and, as he seemed very sensitive to cold, it did not seem wise to
give him an outing.
April 25, 1912 (twelfth day of fast). — The records of the observer showed
that L. slept very well the night before, as the subject rang the bell only once
or twice throughout the whole night. He reported himself as being very
comfortable. To lessen the number of observations so as not to overtire him,
the tests of the blood-pressure and the alveolar air were omitted. In the
afternoon L. went for a ride in a closed carriage; this he seemed to enjoy very
much indeed. On this day, also, Dr. Southard examined him again, having a
long talk with him. A letter which he received from his wife stated that his
brother-in-law was very ill and later in the day he read in a Maltese paper of
his brother-in-law's death. This depressed him very much.
April 26, 1912 (thirteenth day of fast). — L. slept fairly well in the calorimeter,
ringing the bell about once an hour throughout the night. No odor was
apparent when the chamber was opened. L. said that he was very comfort-
able; he seemed bright and in good spirits. In the afternoon he again went
up on the roof, and subsequently a respiration experiment, with two periods,
44 A STUDY OF PROLONGED FASTING.
was made with him while he was sitting in a chair. He was very drowsy and
thought he could sleep throughout the whole night, as the previous night he
had not slept as well as formerly. As he does not read in the evening, he found
it somewhat tedious to sit idle until 8 o'clock, when he usually went into the
calorimeter, and he asked to go into the apparatus an hour earlier. A respira-
tion experiment was made early in the evening just before he went into the
calorimeter, and this experiment was made a part of the experimental routine
on subsequent days. He entered the bed calorimeter at 9 p. m.
April 27, 1912 {fourteenth day of fast). — The subject reported that he had
had a very comfortable night; he rang the bell but twice during the whole time
he was in the calorimeter chamber. About 10h 45m a. m., L. was taken in the
elevator to the third floor of the laboratory, where he was shown about and
several pieces of apparatus were explained to him. He then walked down-
stairs to the second floor and when about a third of the way down, he com-
menced to move rapidly and finally actually ran down the stairs. In the office
he looked at some photographs and dictated the following statement:
"To-day is the fourteenth day of my fast. I feel exceedingly well; I feel
cheerful and hopeful of the grand success. I have slept several good hours
during the night, and am enjoying with great enthusiasm all the experiences
that are carried on me by the professor. I have to-day with Professor
Benedict gone around the upper floors of the laboratory and I have admired
all the great formalities that I have seen."
Subsequently he walked out into the hall and down-stairs to the calorimeter
room, where he sat down in an arm chair and was the subject of a respiration
experiment.
In the afternoon I spent about an hour with L. on the balcony. He was very
cheerful and talked a great deal, gesticulating freely with his hands. He
seemed to enjoy talking about the "beautiful island of Malta." He wore
very much lighter clothing to-day, not using his heavy blanket wrapper at all.
In fact, while walking about the laboratory, he carried his blanket wrapper
over his arm, and although the office was a little cool, he did not put it on, but
sat near the window.
One of the striking features of this fast is the fact that it has been so different
from what he expected in many ways, a good illustration of the unreliability
of personal impressions. For instance, he had expected to feel very chilly,
but while he said he felt chilly the first two or three days, it is possible that
this may have been due to his imagination, as subsequently he was not nearly
so chilly and on this day asked that the temperature of the room should be
lowered, which was done. He also said one of the first days that he was in
the laboratory that he would have a large amount of phlegm as the fast pro-
gressed, but this had not been the case up to this date. While talking with
Dr. Ash on this day, L. said that he had expected that he would become very
hoarse during the fast, but thus far his voice had not changed, except to grow
clearer. In the evening he was given a bath, his body being sponged with
distilled water, after which he entered the calorimeter.
A )ril 28, 1912 {fifteenth day of fast). — The experimental records show that
the subject slept very well throughout the night and he himself reported a
good night. In the afternoon he was taken out for a carriage drive, with the
windows of the carriage open. He was well wrapped up, and said that he was
perfectly comfortable and did not get too tired. He walked down to the
basement and out to the carriage and on his return walked from the carriage
up-stairs again into the calorimeter laboratory. He then sat down for several
minutes and later a photograph was taken of him, with the blackboard showing
GENERAL HISTORY OF FASTING EXPERIMENT. 45
his body-weight curve as a background. He seemed unusually strong and
active, walked with certain feet, and did not seem to be in any way exhausted
or tired. He talked very excitedly during the whole of the drive; in fact, he
talked continuously. He said that about 6 o'clock that morning he had had
a normal seminal emission, which did not irritate him as did the one just prior
to the fast. His observation was verified by Dr. Goodall's subsequent exami-
nation of the urine, a large number of spermatozoa being found.
April 29, 1912 {sixteenth day of fast). — When the calorimeter was opened in
the morning, the usual absence of odor was recorded, and the subject reported
himself as very comfortable and as having slept very well. Inasmuch as it
was not L.'s custom to use a toothbrush — at least no toothbrush was found in
his luggage and he had not brushed his teeth since his arrival in Boston — it
seemed desirable to obtain some cultures of the micro-organisms of the mouth.
These were secured by Dr. Kendall, of the Harvard Medical School, but the
examination of them showed only adventitious organisms. L. was much inter-
ested in this test, especially as such a test had never been made in a previous
fast. Early in the forenoon, he was very irritable and complained of the exces-
sive amount of drinking water; he was especially disturbed because he had not
heard from home. While Mr. Carpenter was making the respiration experi-
ment in the afternoon, the subject did not seem quite so well as on other days.
April SO, 1912 (seventeenth day of fast). — When he came out of the chamber
in the morning, L. reported that the night had been one of the best he had
had since coming to the laboratory. The observer said that there was no odor
and that the subject was very comfortable. Aside from the taking of a series
of new photographs, the day was uneventful.
May 1, 1912 (eighteenth day of fast). — The observers' records show that the
subject scarcely moved during the night, and that the bell rang only once or
twice. There was no odor when the apparatus was opened. In the afternoon
of this day, L. was taken for a drive.
May 2, 1912 (nineteenth day of fast). — L. reported a very comfortable night
in the calorimeter; no odor was apparent when the apparatus was opened.
At this stage of the fast, it had become very clear to the observers that L. had
changed since the beginning. While he walked slowly and steadily and gave
no special evidence of weakness, he was much less talkative, less inclined to
offer advice, and more quiet and subdued. He was by no means so active a
man as at the beginning of the fast. He complained that for several mornings
past he had "felt his bones" during the night and was so uncomfortable that
he waked up and turned over. It had become increasingly difficult to adjust
the stethoscope so as to obtain a clear record of the pulse-rate. On this day
so much difficulty was experienced that an assistant was detailed to watch the
movement of an artery in the neck. His voice was somewhat weaker, though
not husky, but it occasionally rose to its original tone. Furthermore, although
he had maintained that about this period of the fast he would be intellectually
better, he was in reality disinclined to read, write, or talk as much as at the
beginning. This was well brought out by the fact that when he was told on
this day that he ought to spend his time studying the principles of the appa-
ratus and the technique rather than to trouble himself about the results, he
stated that he did not want to spend so much time in this way and that he did
not find himself very much interested in them, but that he would be better
later on. This appears to contradict the statement that he would be more
active intellectually as the fast progressed. He was measured by Dr. Anderson
on this day, who remarked that, as he saw him only once a week, it was very
obvious to him that he had changed very much since the beginning of the fast.
His loss in flesh was of course much more apparent to Dr. Anderson than to
46 A STUDY OF PROLONGED FASTING.
those of us who saw him daily. In the afternoon he was taken to the roof
again, where he stayed for a little more than an hour. When he returned to
the calorimeter laboratory, he sat down in a chair until the psychological tests
began. He still had a great interest in the various tests and made the sug-
gestion that the rectal thermometer be inserted before the evening respiration
experiment, so as to get the temperature changes during this experiment.
May S, 1912 {twentieth day of fast). — Although the subject reported that he
did not sleep so well as he did the night before, he nevertheless had a fairly
comfortable night. There was no odor apparent on opening the chamber.
As the weather was exceptionally fine, L. was taken for a drive in the afternoon.
May 4, 1912 {twenty-first day of fast). — The subject reported this morning
that he had had "several good hours of sleep." He was very comfortable,
except that he had to change his position several times during the night. A
number of unsuccessful attempts were made to take the pulse-rate with the
string galvanometer, but the apparatus was broken. In the afternoon the
subject was taken out driving. In the evening much time was spent in finding
a suitable location for the stethoscope, and when the place was finally found
it was so sharply localized that a movement of the stethoscope bell I cm. in
any direction would pi event the taking of good records. It was then agreed
upon that if during the night the pulse records were not obtainable, a signal
to L. would be given, who would place the bell of the stethoscope as nearly as
he could in the place decided on. Fortunately, on this and also on the fol-
lowing night the pulse-rate was secured with considerable regularity.
May 5, 1912 {twenty-second day of fast). — The experimental records showed
that the subject remained very quiet throughout the night, ringing the bell
only once or twice. In the afternoon of this day he was taken out for a carriage
drive of two hours. He enjoyed this drive very much indeed, speaking enthu-
siastically about the beauty of Boston and its suburbs, and how it elevated him
to see it. When he returned, he claimed that he was neither tired nor cold,
although the air was cooler than when he went out the day before.
He had changed considerably during the preceding week, a fact which was
noted by several persons who had not seen him during that period. He
walked much more deliberately, but appeared to be perfectly sure of his footing.
When he came down-stairs to take the psychological tests, he took the last
three or four steps quite rapidly and stepped off briskly to greet Dr. Langfeld.
On the other hand, at the end of the psychological test, he sat rather dejectedly
in his chair. When Dr. Langfeld told him that the tests were over, however,
he rose from the chair immediately, and returned to the balcony, but stopped
half-way up the stairs to look at the string galvanometer with which we were
working, and seemed as intellectually keen and bright as any one.
May 6, 1912 {twenty-third day of fast). — On coming out of the chamber, L.
reported that he had a fairly comfortable night, sleeping very well. He com-
plained, however, that as he had lost so much adipose tissue, he found it
rather difficult for him to sleep on one side for any length of time, and he was
obliged to turn from one side to the other frequently. The observers also
found it extremely difficult to get the records of the pulse-rate. L. was much
stimulated as a result of visitors in the forenoon. He exhibited a much
sharper intellectual activity than had been apparent for several days, which
might possibly prove his statement that he would become more intellectually
keen as the fast progressed. Other than that we could see little evidence of
the so-called intellectual activity that he had referred to frequently during
the fast.
May 7, 1912 {twenty-fourth day of fast). — The subject was fairly quiet
throughout the night, moving only occasionally. When the calorimeter was
GENERAL HISTORY OF FASTING EXPERIMENT. 47
opened, no odor was observed. Thus far in the fast, no acetone odors had
been noted. In the forenoon a number of visitors came into the laboratory
and several small photographs were taken of L. In the afternoon respiration
experiment, when the subject sat writing, he was reported as being very
obstinate and intractable. He seemed to have some difficulty in finding a
position to suit him, and the nose-pieces had to be inserted a second time to
make sure that they were well fitted. He also complained that the writing
paper was too heavy, so that he would have to write on both sides of the paper
to save postage. He appeared to be very difficult to please. At 5 o'clock on
this day, L. reported that he was feeling very much depressed, owing to the
weather, and that he spent the whole time thinking about home and was
much worried.
May 8, 1912 (twenty-fifth day of fast). — The subject reported a veiy comfort-
able night, moving only once or twice during the night. In order to control
the possible influence of small muscular movements on the temperature of the
air in the chamber at the end of an experimental period, the height of the
spirometer attached to the calorimeter was recorded on a smoked paper drum,
a routine which proved very helpful on subsequent nights. On this morning
L. was very sullen and disagreeable when spoken to, especially with Mr.
Carpenter. He said that he had expected there would be a yellow pigmen-
tation of the skin as in other fasts, but this did not appear. In the afternoon
I asked him how much longer he wished to fast. He replied by asking how
long I wanted him to fast. As I had previously told him on several occasions
that I should like him to fast 30 days, it was obvious that he wished to be
asked to stop fasting rather than to break the fast himself. I told him that
we should probably wish him to fast for 30 or 31 days. The diet of the
re-alimentation period was then discussed. He believed very strongly that
the fast should be broken with acid fruit juices alone. The rectal thermometer
remained in position throughout the day and the temperature records were
secured for the greater part of the time.
May 9, 1912 (twenty-sixth day of fast). — The subject spent an unusually
quiet night, ringing the beU only three times throughout the night. In the
afternoon, during an active conversation which I had with him, in which
general questions were discussed, he seemed to be very spirited, lively, and
interested. His pulse-rate was 86 at this time. In the evening, when he came
down-stairs to take the psychological test, he appeared to be quite unsteady
on his feet and said that he was light-headed. He thought that his unsteadi-
ness was due to the fact that he had lost so much weight that he put unusual
strength on his foot, more than he needed to take his weight, and that this
tended to unbalance him. Mr. Carpenter also noticed that when walking, L.
was quite uncertain in his steps and thereafter we watched him more closely
when he was on his feet. He said considerable on this day about the method
of breaking the fast, emphasizing the fact that he wished to be able to say that
the fast was broken by request and not by his desire. It was arranged to
allow him to fast for 31 days, and then to begin taking food. He also expressed
a preference for breaking his fast on a diet consisting of boiled rice and honey,
with the juice of oranges and lemons and grape juice. Obviously such a diet
should have many scientific points of interest, but it was questionable
whether, after fasting so long, the stomach should be filled with the acids of
oranges, lemons, and grapes. He was firmly convinced that this was the diet
that should be used and could not be dissuaded from it. Subsequent experi-
ence proved the undesirability of this kind of a diet for breaking a long fast.
May 10, 1912 (twenty-seventh day of fast). — The subject reported that he did
not sleep so well during the night as he had the night previous. He was not
48 A STUDY OF PROLONGED FASTING.
uncomfortable in any way, but simply could not sleep. There was no notice-
able odor. In a discussion with some physicians who visited him to-day, L.
emphasized the fact that his tongue remained coated throughout the whole
fast and that he had more or less of a bad taste in his mouth. His theory was
that the waste products were not all eliminated from the body and that "they
were trying to find their way out by the mouth." He thought that if he had
defecated he would have had less discomfort. Although the subject went for
a drive on the afternoon of this day, he complained that the weather had
changed and that it was not pleasant. He also complained bitterly to Dr.
Langfeld regarding Mr. Carpenter, saying that he would like to break every
bone in his body. This would pronounce against fasting for amiability.
May 11, 1912 {twenty-eighth day of fast). — L. reported that it was late before
he fell asleep and that he had not been able to sleep much or soundly during
the night. In the early part of the evening, after he had been sealed into the
calorimeter, he signaled that he was obliged to urinate and he had to be taken
out. When he was sealed up again shortly after 10h 30m p. m., he thought he
did not go to sleep until 12 o'clock. He moved several times during the night.
During the afternoon of this day he spent some time on the roof, reading, and
seemed to enjoy himself very much, appearing to be in good spirits all of the
afternoon. But when he came down from the roof, he was extremely stubborn
when taking his tests with Dr. Langfeld. For example, in a list of ten words
which was read to him for memorizing, the eighth word was "wart." L. did
not understand what it meant and asked to have it explained. Dr. Langfeld
tried to keep him quiet until the list was read a second time, but he would not
listen and insisted upon knowing what the word was. When Dr. Langfeld
read the list to him a second time, he said that he had not heard the words, as
he was thinking of what the word "wart" meant.
May 12, 1912 (twenty-ninth day of fast). — The subject said that he slept very
well the previous night and in the morning seemed very bright. He busied
himself much of the forenoon in writing a history of his life. (See p. 22.)
A respiration experiment was made with him in the morning in which the
subject breathed an atmosphere containing a large percentage of oxygen.
In the afternoon I took him for a drive of 2 hours. Duiing the drive he
was bright and very talkative, telling a good deal about his family, his per-
sonal history, and his difficulty in getting an education. He became more
quiet before he returned to the laboratory and in the psychological test he
told Dr. Langfeld that the weather depressed him very much. He also
told Dr. Langfeld that he was very sorry that I wanted him to break the fast
and that he could easily fast for 10 days more; he did not like to break his
fast until his tongue had become clear. On the other hand, he said that he
would be out of the laboratory on May 23 and would be a free man again and
"have free ice cream." It was evident that he would be difficult to control.
There has been none of the pigmentation of the skin which he had expected
and very little phlegm.
May 18, 1912 (thirtieth day of fast). — L. reported that he had slept well. He
was in excellent spirits on this day and very lively and jolly. His movements
were vigorous and he was unusually bright and active. All of his spare time
was occupied in writing his autobiographical notes, which at this time covered
13 closely written pages. As usual he felt quite able to be photographed.
At noon I asked him how he was and he replied that he felt very well indeed,
saying: "We have seen the lighthouse and are now entering the harbor and
will see land shortly."
In the afternoon I took him to the Boston City Hospital, where Dr. Francis
H. Williams made an extensive series of X-ray photographs. L. was intensely
GENERAL HISTORY OF FASTING EXPERIMENT. 49
interestedin this, and also thoroughly enjoyed the ride to and from the hospital.
He was very much pleased with the whole day's program. He said that he
was "now dropping anchor and therefore at the end of the voyage." His
mental condition seemed to make a great difference in his whole make-up.
On some days his faculties were veiy much keener than on others. For in-
stance, during the test for visual acuity, he answered the position of the
letter "E" with great strength of voice and with no hesitation whatever.
May 14, 1912 {thirty-first day of fast). — On this day, which was the last day
of the fast, the program was very full, including an extensive series of photo-
graphs (see Plate 5) and the body measurements. The first word that L.
spoke when the calorimeter was opened was in regard to having his photograph
taken. In response to an invitation issued by the Nutrition Laboratory, a
number of medical men from the vicinity of Boston came to see the subject
between 2 and 3 o'clock in the afternoon. L. talked very rapidly and in a
lively manner for nearly 40 minutes, setting forth his views in a more or less
lucid manner to his audience. In the middle of his talk, I took his pulse-rate and
found it to be 82. The sub j ect was particularly desirous of having his photograph
taken during the afternoon and wished to be photographed with each individual
who had worked with him. In this afternoon talk with the physicians, L.
made many conflicting statements. For instance, he said that he never used
wine or alcohol in any form, while he had repeatedly told me that he frequently
drank Malta wine and was not a teetotaler. He also reported that he was
invariably a vegetarian, fruitarian, and nutarian, eating no meat; this was
strikingly in contrast with the fact that he ate meat several times in the days
just preceding the fast. During the visit of the physicians, he became very
much excited and enthusiastic, evidently enjoying the opportunity of speaking
to his audience. During the respiration experiment in the evening, he talked
considerably about the end of the fast, saying that while he was not exactly
tired, yet there was much emotion connected with his fast and he was thinking
of the effect produced at home when the news reached them that he had
completed the fast of 31 days successfully. Malta was awaiting the news and
all of the people would be discussing his wonderful feat.
Figures D, E, and F in Plates 2 and 3 (pages 19 and 31) are from photo-
graphs secured on this last day of fasting. Figure E is of special interest, as
it shows L. posing for 20 seconds while the exposure was being made. There
is no evidence of unsteadiness.
RE-ALIMENTATION PERIOD.
May 15, 1912 (first day with food). — The bed calorimeter experiment was
ended at 7h 56m a. m. When the apparatus was opened there was no odor
apparent. The observer reported that the subject was comparatively quiet
throughout the night, with but one movement between llh44m p. m. and
5h 28m a. m. L. complained, however, that he was kept awake by acid fumes
and noise in the calorimeter room. Photographs were taken of the subject
while he was upon the respiration apparatus, and also later in various positions.
When he was weighed after the respiration experiment was over, it was found
that he had lost 13.25 kilograms of his initial weight. He seemed much
pleased that the fast was ended and spoke of the excitement there would be at
his home when the cablegram was received announcing the successful com-
pletion of his fast.
The method of breaking the fast had been thoroughly discussed with the
subject previously, but he insisted upon using lemons, oranges, grape juice,
and honey. It was finally arranged that he should follow his own choice in
the matter and, accordingly, lemons, oranges, boiled rice, and honey were
50 A STUDY OF PROLONGED FASTING.
supplied him at his request. At 9h 38m a. m. he peeled and ate two lemons,
using neither water nor sugar. In addition, three oranges, about 300 grams
of honey, and approximately 1 liter of grape juice were taken in portions
during the day. To study the flora of the colon, a rectal injection of sterile
salt water was given him at lh 10m p. m. About 4h 30m p. m. he complained of
severe colic, but was somewhat relieved about 5h 30m p. m. by a copious move-
ment of the bowels. Subsequently he again had colic, with a second move-
ment of the bowels and vomiting. He appeared to be utterly wretched and
weak, and to have entirely lost his courage. The contrast between his high
spirits of the day before and his condition on this day was very striking.
No respiration experiments were attempted on this day after the ingestion
of food, but a sample of the blood was taken, and records were made of the
blood pressure and the alveolar air. Under the circumstances, it did not seem
wise to have the subject sleep in the calorimeter. Arrangements were there-
fore made for him to sleep on a couch in the calorimeter laboratory, with a
physician in constant attendance. While no records of the metabolism were
obtained during the night, the pulse-rate was recorded by means of the
stethoscope, the assistant sitting behind a screen out of sight of the patient.
May 16, 1912 (second day with food). — The subject passed a restless night,
having another attack of colic about 2 a. m. and defecating. The pulse-rate
through the night was somewhat higher than on previous nights. As he
seemed weak and sick in the morning, Dr. Goodall, who came to see him early
in the day, urged him to take some weak beef tea or clam broth, also toast,
but he utterly refused to take anything of this nature, saying that such food
would poison him. He furthermore said that as we were now through with
him, we wished to poison him. He was as firmly convinced as ever that his
method of breaking the fast was the correct one, but thought that he took food
too soon and that he should have waited until his tongue had cleared, as he
considered it dangerous to break the fast sooner. He told the physician in
attendance during the night that a man might have to fast 100 days before
the natural hunger would return. L. lay on the couch until about 10 a. m.,
when he went to the balcony and dressed. He continued his diet of fruit juices
and honey throughout the day, but, at our urgent request, diluted the lemon
juice with distilled water, taking first 77 c.c. of lemon juice, an equal amount
of distilled water, and 11 grams of honey. Between 9h 30m a. m. and 7h 30m
p. m., he drank at intervals a mixture containing the juice of 6 oranges, 367 c.c.
of water, and 128 grams of honey, making a total volume of about 1,400 c.c.
About 7h 30m p. m., he took the juice of another orange, saying that he was
hungry but felt very well. During the day Dr. Goodall made the usual
physical examination; the blood pressure and the alveolar air were also
observed. The subject entered the bed calorimeter at 8h 14m p. m. for the
usual night experiment.
May 17, 1912 (third day with food). — At 10 p. m. the subject telephoned and
urinated, but after that time the experimental records show that he was very
quiet. When he left the apparatus in the morning, he said that he had had a
very comfortable night, sleeping better than he had for years. He thought
he must have had at least 10 hours of dreamless sleep. He appeared to be in
good spirits. The usual respiration experiment was made with him this
morning. Throughout the day, he continued taking at intervals a mixture
of orange juice and honey, diluted with water; he was also supplied with
raisins and dates. The usual physical examination was made by Dr. Goodall;
also tests of the blood pressure, the alveolar air, and for acetone in the breath.
A sample of the blood was taken by Dr. Ash. In the afternoon he went up on
the roof for a time, but was ill-humored, very difficult to please, and full of
GENERAL HISTORY OF FASTING EXPERIMENT. 51
complaints. The day before I had taken exception to his statements that
we were to blame for his illness on taking food and had shown him that he
had chosen his own diet and had refused to be guided by the advice of the
physicians. He was greatly offended at this, and was disagreeable to every
one in consequence all of this day. When Dr. Langfeld came at about 5
o'clock for the psychological tests, he burst forth in a long tirade, complaining
of nearly every member of the staff, and of his treatment at the laboratory.
He claimed that throughout the night, while in the calorimeter, he was left in
charge of "boys," who paid no attention to him, and that he might die there
without any one knowing of it. He had said nothing about this to me, but
asked Dr. Langfeld to tell me.
In the evening L. again had colic, with diarrhea and vomiting. He entered
the calorimeter for the regular experiment at 8h 35m p. m., but only on the
condition that Mr. Emmes or Mr. Carpenter should stay in the calorimeter
room all of the night. Accordingly Mr. Carpenter slept on a couch in the
balcony, so as to be near if needed during the night, the usual experienced
observers being on duty. At 10 p. m. it was reported to me that the subject
had a pulse-rate of 104 and body-temperature (rectal) of approximately 38° C.
I went to the calorimeter laboratory and remained there for a time, thinking
that he might be suffering from colic and would have to be taken out. About
HP 30m p. m. he telephoned that he wished to defecate. He was accordingly
taken out of the calorimeter. After defecating, he returned to the apparatus
and the experiment began at llh 57m p. m.
May 18, 1912 {fourth day with food). — The subject had a restless night,
being evidently suspicious, unhappy, and discontented. He called for Mr.
Carpenter about 3 a. m. and asked if something had not fallen upon the calo-
rimeter, as he had felt a jar. He was assured that nothing of the kind had
happened, both of the observers having been sitting quietly at their respective
posts when he called. He was taken out of the apparatus at 5h 45m a. m. and
said that he had not slept all night. His pulse-rate and respiration-rate were
high ; the body-temperature was also slightly elevated. He seemed apprehen-
sive and was apparently mentally unbalanced, talking in an irrational way with
Dr. Goodall. The usual respiration experiment was made with him on this
morning and he was weighed, but no other observations were taken.
After the respiration experiment, he complained of exhaustion, but finally
returned to the balcony and began to eat; his attitude towards nearly every
member of the staff was suspicious and antagonistic. He insisted that the
British consul should be asked to come to the laboratory to confer with him,
which was done. It was finally decided that considering his condition and
general attitude, he would more quickly recover his balance if cared for at a
hospital. Arrangements were made with the Massachusetts General Hospital
to defray the expense of his care and he was taken there in a carriage by one
of the members of the staff, and placed in a private room. On May 19 he
was seen by Dr. Goodall, who found him very contrite. He asked Dr.
Goodall to find me at once and to apologize for his behavior the day before.
By permission, a copy of the hospital records is appended herewith, giving
the history of his stay in the Massachusetts General Hospital.
"May 18. — Patient states that he completed a 30-day fast under observation
at the Carnegie Nutrition Laboratory three days ago, and that he was sent
here to-day through a misunderstanding and that he should not be in a
hospital. History and physical examination not attempted by order of Dr.
Langnecker, superintendent on duty.
"May 19. — Seems contented, cheerful, and comfortable. Eats with good
appetite and relish. Says the bed is not as comfortable as the calorimeter.
52 A STUDY OF PROLONGED FASTING.
"May 20. — Patient was very angry on coming into the hospital, first,
because he was in a public institution, whereas he had expected to go into a
'convalescent home for wealthy people.' Second, because the fact of his
going to a hospital after his fast would detract from the renown and interest
of his feat.
" Yesterday he appeared very happy. He seemed to realize and accept that
there were certain disadvantages here. He was asked what diet he wanted
and requested straight house diet, and this was given him in accordance with
the ideas of Dr. Benedict. He ate this with relish. He appeared a little weak,
but able to be about. He spent the day in writing and in sorting out various
articles and pictures, of which he had a half of a valise full.
"Last night he was nauseated and vomited several times, but was not ill
enough to be reported. This morning he was again very angry. He com-
plained that oatmeal was very bad for a man recovering from starvation.
He repeated his complaints of Saturday (May 18) and said that he was being
held by conspiracy and wanted to summon or go to the British consul. He
was therefore discharged at his own request.
"While at the front door waiting to have his bill arranged and his valuables
procured, he complained loudly that his possessions were being withheld from
him. While waiting for a carriage to be summoned, he called out loudly many
times that he would get out of here if he had to crawl on his hands and knees.
"While in the wards he was shown every attention by the nurses, and
especially all food and drink brought to him as desired.
"Discharged to own M. D. (Not treated)."
Subsequent to his leaving the Massachusetts General Hospital, L. was seen
by Drs. Langfeld and Goodall and a second set of X-ray photographs was
taken by Dr. Williams. Such data as were secured are included in the
individual reports.
PHYSICAL CONDITION OF THE SUBJECT DURING THE FAST.
By Harry W. Goodall,
Department of Physiological Chemistry, Harvard Medical School.
A complete physical examination of the subject was made 60 hours
before the beginning of the fast, on the first day of the fast, and every
alternate day thereafter. So far as possible, the same conditions as to
time of day, posture, methods, etc., were observed at each examination.
The percussion outline of the various organs was marked with pencil and
these lines were not disturbed until variations in the size of the organs
made it necessary. Owing to the thinness of the subcutaneous tissue
the percussion note could be sharply defined and the favorable mental
attitude of the subject permitted complete muscular relaxation.
Family history. — Father 68 years of age, mother 64 years, both living and
well. Two sisters, one 27 years of age, the other 19 years, both living and
well. One sister died in infancy of malaria. One brother died in childhood
of croup. Twin brother died in infancy, cause unknown.
Past history. — Was a delicate child, the most pronounced characteristic
being a sensitive nervous organization. Does not remember about diseases
of childhood. In his twentieth year (20 years ago), while in the university,
he had a severe nervous breakdown and had to give up his studies. Has
suffered from neurasthenia since. Has been under the care of several physi-
cians at different periods without receiving any benefit. Being discouraged
with these experiences, he became interested in certain popular articles advo-
cating fasting as a cure for diseases, including neurasthenia, and in April 1910
he underwent his first fast. So great was the improvement in his nervous
condition following this fast that he has since devoted much of his attention
to the study of fasting. Previous to the present experiment he has undergone
the following fasts:
April 1910. Fasted 8 days, taking nothing but water; not under obser-
vation.
March 1911. Fasted 40 days, taking nothing but water; under partial
observation.
August 1911. Fasted 12 days, taking nothing but water; under partial
observation.
November 1911. Fasted 5 days, complete fast; not under observation.
These experiences have convinced him that, in health, food not only
makes the individual susceptible to diseases and causes disease, but
also interferes with the proper exercise of the mental faculties. He
states that during the fasting period the mind is clear and alert and
that there is a strong desire for study. It is now the practice of his
wife, his two children, and himself to abstain entirely from food during
any illness, and he is convinced that the severity of any disease is
reduced, the subjective symptoms made less disagreeable, and the course
of the disease shortened if such a course is pursued.
53
54 A STUDY OF PROLONGED FASTING.
His personal experiences and his study have been so convincing that
he was desirous of undergoing a scientific experiment, under the most
perfect conditions, for the benefit of humanity. Having learned that
the most perfect equipment in the world was at the Nutrition Labor-
atory, he applied for the privilege of undergoing this experiment there.
He expressed his pleasure at the cordial manner in which he had been
received at the laboratory, and stated that he was very happy in the
thought that he was about to undertake the fast, not alone on account
of the scientific value of the experiment, but also because of the im-
provement which he anticipated in his own well-being. He expressed
his pleasure in going into minute details as to his subjective feelings.
RESULTS OF PHYSICAL EXAMINATION.
April 11, 1912 (60 hours before beginning the fast):
A well-developed, well-proportioned, and fairly well-nourished man.
Height, 170.7 cm.; weight, 60.1 kilograms; age, 40 years. Stands
erect. Walks with body erect, with no abnormalities in the gait.
Hair of head and beard dark. Skin has a muddy yellowish tinge, but
is soft and moist. Slight conjunctivitis of both eyes. Very moderate
amount of subcutaneous fat. Muscles of moderate size but rather
soft. Has a small infected papule on left alae nasse. No pulsations
noted in the neck, chest, or abdomen. No visible abnormalities.
Mouth: Mucous membrane of the lips and cheeks of good color and moist.
Tongue moist with a slight coating, especially on the central and poste-
rior portions. Teeth in fair condition. There is a slight deposit of
discolored tartar at the base of the teeth and two teeth have temporary
soft fillings. No particular odor to the breath. No enlargement of
the tonsils. Pharynx is reddened, the blood vessels dilated, and there
are a few blebs on the posterior pharyngeal wall with a little mucus
adhering.
Glands: Cervical, axillary, and epitrochlea glands not palpable. A few
small glands in both groins.
Reflexes: Pupils equal and react normally to light and distance. Abdom-
inal, cremasteric, patella, Achilles, and plantar reflexes normal.
Chest: Symmetrical, well formed, some sinking in of the supra- and infra-
clavicular spaces. Good expansion with inspiration. No bulging
of the praecordia, and apex beat of heart is not visible or palpable.
Lungs: Percussion of the right lung shows normal resonance to the upper
border of the fifth rib in the nipple line, to the lower border of the
fifth rib in the axillary line, and to the eleventh rib in the back. On
the left normal resonance to the eighth rib in the mid-axillary line
and to the eleventh interspace in the back. Vocal fremitus is
slightly increased and expiration slightly prolonged at the right apex,
extending to the second rib in front and the spine of the scapula
behind. There were no rales and the lungs were otherwise negative.
Heart: The area of superficial cardiac dullness (light percussion) was
measured from a perpendicular line through the mid-sternum and a
horizontal line drawn at the level of the nipples. The upper border
of cardiac dullness was at the third interspace. The left border of
cardiac dullness was 9.5 cm. from the mid-sternum, 1.2 cm. inside
the nipple. The right border of cardiac dullness was 1.2 cm. from
PHYSICAL CONDITION OF SUBJECT DURING FAST. 55
April 11, 1912— Continued.
the mid-sternum. The total width of cardiac dullness was 10.7 cm.
The cardiac sounds were somewhat distant, but of good quality and
regular rhythm. There were no murmurs to be heard. The aortic
and pulmonic second sounds were of equal intensity.
Pulse: The pulses were equal, regular at the rate of 82 per minute.
Rhythm regular, volume fair. No sclerosis of the vessels noted.
Abdomen: The abdomen was symmetrical, rather prominent when
standing, but flat when reclining. It was soft, tympanitic, but with
no distension. There was no tenderness on palpation. Nothing
abnormal was felt.
Liver: The upper border of liver dullness was at the lower border of the
fifth rib in the nipple line. The lower border of liver dullness was
I cm. below the costal margin. Total width of dullness 11.5 cm.
The edge of the liver was indistinctly palpable, soft, and without
irregularities.
Stomach: The measurements of the stomach were determined as accu-
rately as possible by means of auscultatory percussion. The lines
of measurement were the median line of the body and a horizontal
line through a point half-way between the tip of the ensiform and the
umbilicus. Tympany in the median line extended from the tip of the
ensiform to a point 3.5 cm. above the umbilicus, a total distance of
II cm. The left border of tympany extended to a point 16 cm. from
the median line. Faint rhythmic sounds were to be heard with the
stethoscope. There was no splashing with palpation.
Spleen: The upper border of splenic dullness was at the eighth rib. The
area of splenic dullness was vaguely determined as 7X5 cm. The
spleen was not felt.
Kidneys: Neither kidney was palpable.
Genital organs : Aside from a long prepuce and a slight left variococele,
the penis and testicles were normal.
April 14, 1912 (first day of fast):
Abdomen: Not as prominent. Flat to percussion everywhere except
over the area of stomach tympany.
Stomach: No rhythmic sounds heard.
No change noted in general appearance, mouth, glands, reflexes, chest,
lungs, heart, pulse, liver, spleen, kidneys, or genital organs.
April 16, 1912 (third day of fast):
Mouth : No change noted, with the exception of a pronounced odor to the
breath.
Reflexes : No change noted except patella reflex not as active.
Liver: Lower border of liver dullness at costal margin. Total width of
dullness 10.5 cm. Edge not felt. (First change noted.)
Stomach: Tympany in the median line extends from a point 3 cm. above
the tip of the ensiform to 7 cm. above the umbilicus. Total width
10.5 cm. Left border of tympany 18 cm. from the median line.
Active peristalsis with pronounced rhythmic sounds.
No change noted in general appearance, glands, chest, lungs, heart, pulse,
abdomen, spleen, kidneys, or genital organs.
April 18, 1912 (fifth day of fast):
Mouth: No change noted. Odor to breath still pronounced and tongue
more heavily coated.
Reflexes: Patella reflexes only obtained with reenforcement.
56 A STUDY OF PROLONGED FASTING.
April 18, 1912— Continued.
Abdomen: Not as prominent while standing, retracted while reclining.
Dull to percussion, except over the area of stomach tympany.
Marked visible pulsation of the aorta.
Liver: No change from note of third fasting day.
Stomach: Tympany in the median line from a point 3.5 cm. above the
tip of ensiform to 7.5 cm. above umbilicus. Total width, 10.5 cm.
Left border of tympany 16 cm. from median line. Rhythmic sounds
were audible but not marked.
No change noted in general appearance, glands, chest, lungs, heart, pulse,
spleen, kidneys, or genital organs.
April 20, 1912 {seventh day of fast):
General appearance : No change noted, except that the features are slightly
drawn and subject moves about a little more slowly.
Mouth : No change from note of fifth fasting day.
Reflexes: No change in pupillary and plantar reflexes. Patella reflexes
obtained with difficulty. Achilles reflex very slight.
Heart: Left border of cardiac dullness 8.5 cm. and right border 1 cm.
from mid-sternum. Total width, 9.5 cm. (first change noted). No
change in character of sounds.
Abdomen: No change from note of fifth fasting day.
Liver: No change from note of third fasting day.
Stomach : Tympany from a point 3 cm. above the tip of the ensiform to
8 cm. above the umbilicus. Total width 9.5 cm. Left border 14
cm. from median line. Rhythmic sounds heard.
No change noted in glands, chest, lungs, pulse, spleen, kidneys, or genital
organs.
April 22, 1912 (ninth day of fast):
General appearance: Conjunctivitis not quite so marked. Infected
papule on the nose has disappeared. Otherwise no particular change.
Mouth : Mucous membrane of tongue and mouth dry. Tongue slightly
less coated. Odor to breath not so pronounced.
Reflexes : Patella and abdominal reflexes absent. Achilles reflex obtained
only with difficulty. Cremasteric and plantar reflexes normal.
Heart: Left border of cardiac dullness 8 cm., right border 1 cm. from
mid-sternum. Total width 9 cm. (second change in size of heart
noted). No change in character of sounds.
Abdomen: No change from note of fifth day.
Liver : No change from note of third day.
Stomach: Tympany from a point 4 cm. above tip of ensiform to 8 cm.
above the umbilicus. Total width, 10.5 cm. Left border tympany
14 cm. from median line.
No change noted in glands, chest, lungs, pulse, spleen, kidneys, or genital
organs.
April 24, 1912 (eleventh day of fast):
General appearance: Features somewhat drawn. Muscles not quite so
firm. Skin more elastic. Seborrhea sicca of entire scalp. Walks
without evidence of weakness. No unsteadiness when standing with
eyes closed.
Mouth: Mucous membrane of mouth and tongue dry. Lips dry and
desquamating. Odor to breath less marked.
Reflexes : No change from note of ninth day of fast, except Achilles reflex
absent.
Chest : Some sinking in of supra- and infra-clavicular spaces.
PHYSICAL CONDITION OF SUBJECT DURING FAST. 57
April 24, 1912— Continued.
Heart : No change noted in measurements of heart from ninth day of fast.
Faint systolic souffle heard all over the praecordia, loudest at the apex.
Not transmitted, and not related to the respiratory murmur. Inten-
sity not influenced by posture. Heart sounds not so distinct.
Pulse : Volume of pulse not so good.
Abdomen: No change from note of fifth day of fast.
Liver: No change from note of third day of fast.
Stomach : No change from note of ninth day of fast. Rhythmic sounds
heard.
No change noted in glands, lungs, spleen, kidneys, or genital organs.
April 26, 1912 {thirteenth day of fast):
General appearance: Features not so drawn. Conjunctivitis somewhat
improved. Otherwise no change from note of eleventh fasting day.
Mouth : No change from note of eleventh fasting day.
Reflexes: No change from note of eleventh fasting day.
Chest : No change from note of eleventh fasting day.
Heart: No change from note of eleventh fasting day.
Pulse: No change from note of eleventh fasting day.
Abdomen: Markedly retracted when reclining.
Liver: No change from note of third fasting day.
Stomach: Tympany from a point 3 cm. above tip of ensiform to 8 cm.
above the umbilicus. Total width 10.5 cm. Left border of tympany
12.5 cm. from median line. Rhythmic sounds heard.
Kidneys: Pole of right kidney just palpable. (First change noted.)
No change noted in glands, lungs, spleen, or genital organs.
April 28, 1912 (fifteenth day of fast):
General appearance: Odor to breath less pronounced. Stands erect.
Has normal gait but moves about more deliberately. Seborrhea
sicca much improved.
Mouth : Mucous membrane of mouth and lips moist. No desquamation
of lips. Tongue slightly less coated.
Reflexes: No change from note of eleventh fasting day.
Chest: No change from note of eleventh fasting day.
Heart: No change in measurements of heart from note of ninth fasting
day. The systolic souffle which appeared on the eleventh fasting
day is no longer heard. Intensity of heart sounds not increased.
Pulse: No change noted from note of eleventh fasting day.
Abdomen: Retracted. Dull to percussion everywhere except over the
area of stomach tympany. No tenderness on palpation except slight
tenderness over the pulsating aorta. The large intestines appear to
be about the size of a thumb and can be rolled under the finger
extending from the caecum up to the right hypochondrium and from
the left hypochondrium down to the brim of the pelvis. There is no
gurgling with pressure.
Liver: The upper border of liver dullness is 1 cm. lower than at the
beginning of the fast. Total width 9.5 cm. (second change in size).
Stomach: Tympany from a point 8 cm. above the umbilicus. Total
width 9.5 cm. Left border of tympany 13 cm. from the median line.
Rhythmic sounds heard.
Kidneys: No change from note of thirteenth fasting day.
No change noted in glands, lungs, spleen, or genital organs.
58 A STUDY OF PROLONGED FASTING.
April SO, 1912 {seventeenth day of fast):
General appearance: Features not so drawn, otherwise no change from
note of fifteenth fasting day.
Mouth : Desquamation of lips entirely disappeared, otherwise no changes
from" note of fifteenth fasting day.
Reflexes: No change from note of eleventh fasting day.
Chest: No change from note of eleventh fasting day.
Heart: Left border of cardiac dullness 7.5 cm., right 0.5 cm. from median
line. Total width 8 cm. (third change in the size of heart noted.)
No further change in quality of heart sounds.
Pulse : No change from the note of the eleventh fasting day.
Abdomen: No change from the note of the fifteenth fasting day.
Liver : No change from the note of the fifteenth fasting day.
Stomach : No change from the note of the fifteenth fasting day.
Kidneys : No change from the note of the thirteenth fasting day.
No change noted in glands, lungs, spleen, or genital organs.
May 2, 1912 {nineteenth day of fast):
General appearance : No change from note of seventeenth fasting day.
Mouth : No change from note of seventeenth fasting day.
Reflexes: No change from note of eleventh fasting day.
Chest: No change from note of eleventh fasting day.
Heart: Left border of cardiac dullness 7 cm. from the median line, right
border at median line. Total width, 7 cm. (fourth change in meas-
urements of heart noted). No change in quality of sounds from the
note of the fifteenth fasting day.
Pulse : No change from note of eleventh fasting day.
Abdomen: No change from note of fifteenth fasting day.
Liver: No change from note of the fifteenth fasting day.
Stomach: Tympany from a point 1.5 cm. above tip of the ensiform to 8.5
cm. above the umbilicus. Total width, 8.5 cm. Left border of tym-
pany 11 cm. from median line. No rhythmic sounds heard.
Kidneys: No change from note of the thiiteenth fasting day.
No change noted in glands, lungs, spleen, or genital organs.
May 4, 1912 {twenty-first fasting day):
General appearance: The soft parts of the extremities not so firm and
not so large. Otherwise no change from the note of the seventeenth
fasting day.
Mouth : No change from the note of the seventeenth fasting day.
Reflexes: No change from the note of the eleventh fasting day.
Chest: No change from the note of the eleventh fasting day.
Heart: No change from the note of the nineteenth fasting day.
Pulse: No change from note of the eleventh fasting day.
Abdomen: No change from the note of the fifteenth fasting day.
Liver: Upper border of liver dullness at sixth rib, lower border at costal
margin. Total width 9.0 cm. (third change noted in size of liver).
Stomach : Area of tympany the same as noted on the nineteenth fasting
day. Rhythmic sounds heard.
Kidneys: Right kidney palpable, pole of left kidney (for first time) just
felt with deep inspiration.
No change noted in glands, lungs, spleen, or genital organs.
May 6, 1912 {twenty-third day of fast):
General appearance : No change from note of the twenty-first fasting day.
Mouth: Mucous membrane moist. Very slight coating on tongue.
Very little odor to breath.
PHYSICAL CONDITION OF SUBJECT DURING FAST. 59
May 6, 1912— Continued.
Reflexes: No change from note of eleventh fasting day.
Chest: No change from note of eleventh fasting day.
Heart: Measurements of cardiac dullness the same as noted on the
nineteenth fasting day. The cardiac sounds are rather more distant.
The aortic second sound is a little louder than the pulmonic sound.
Pulse: No change from note of eleventh fasting day.
Abdomen: Very slight gurgling (gas and liquid) with palpation of the
right hypochondrium.
Liver : No change from note of twenty-first fasting day.
Stomach : No change from note of twenty-first day, except that rhythmic
sounds are heard.
Kidneys: No change from note of twenty-first fasting day.
No change noted in glands, lungs, spleen, or genital organs.
May 8, 1912 (twenty-fifth day of fast):
General appearance : No change from note of twenty-first fasting day.
... Mouth: Mucous membrane moist. Good color. Tongue nearly clean.
Very little odor to breath.
Reflexes : No change from note of eleventh fasting day.
Chest : No change from note of eleventh fasting day.
Heart : Left border of cardiac dullness 6.8 cm. from mid-sternum. Right
border 0.3 cm. to left of mid-sternal line. Total width 6.5 cm.
(fifth change noted in measurements of heart). First sounds of the
heart not distinct.
Pulse : No change from note of eleventh fasting day.
Abdomen: Retracted. Dull to percussion except over area of stomach
tympany. No gurgling on palpation and no tenderness except over
the pulsating aorta.
Liver : Lower border of liver dullness unchanged. Total width of dullness
8.6 cm. (fourth change noted in size of liver).
Stomach: No change in area of stomach tympany from note of twenty-
first fasting day. No rhythmic sounds heard.
Kidneys: No change from note of twenty-first fasting day.
No change noted in glands, lungs, spleen, or genital organs.
May 10, 1912 (twenty-seventh day of fast):
General appearance : No special change from note of twenty-first fasting
day.
Mouth: Tongue nearly clean. Odor of breath not specially marked.
Reflexes : No change from note of eleventh fasting day.
Chest : No change from note of eleventh fasting day.
Heart : No change from note of twenty-fifth fasting day.
Pulse : No change from note of eleventh fasting day.
Abdomen : No change from note of twenty-fifth fasting day.
Liver: No change from note of twenty-fifth fasting day.
Stomach : No change from note of twenty-first fasting day.
Kidneys: No change from note of twenty-first fasting day.
No change noted in glands, lungs, spleen, or genital organs.
May 12, 1912 (twenty-ninth fasting day):
General appearance: No change from note of twenty-first fasting day.
Mouth : No change from note of twenty-seventh fasting day.
Reflexes : No change from note of eleventh fasting day.
Chest: No change from note of eleventh fasting day.
Lungs: The percussion note at both apices and just below the clavicle is
slightly higher pitched. No change noted in the respiratory murmur.
60 A STUDY OF PROLONGED FASTING.
May 12, 1912— Continued.
Heart: No change from note of twenty-fifth fasting day.
Pulse : No change from note of eleventh fasting day.
Abdomen: Aside from a slight gurgling of gas and liquid in the right
hypochondrium with palpation no change noted from twenty-fifth
fasting day.
Liver: Total width of liver dullness 7.4 cm. (fifth change noted in size
of liver).
Stomach: No change from note of twenty-first fasting day. Rhythmic
sounds heard.
Kidneys: No change from note of twenty-first fasting day.
No change noted in glands, spleen, or genital organs.
May 14, 1912 (thirty-first day of fast):
General appearance: The features have a drawn appearance. Subject
stands erect. Walks with body erect and shows no abnormalities in
gait but moves about slowly as if fatigued. The skin has a muddy
yellowish appearance but is soft and moist. The skin is relaxed and
is easily picked up, as if there were but little subcutaneous fat. The
muscles are rather small and not firm. There is a slight conjunc-
tivitis of both eyes. No abnormal pulsation noted in the neck or
chest, but the pulsation of the abdominal aorta is marked, especially
when reclining.
Mouth : Mucous membrane of the lips and cheeks of good color and moist.
Tongue moist, with a very slight coating on the posterior part. Teeth
show a slight deposit of discolored tartar at the base. The breath
has a very slight odor. No enlargement of the tonsils. The pharynx
shows some injection of the blood vessels and a few blebs on the
posterior pharyngeal wall. There is a little mucus on the posterior
naso-pharynx.
Glands: Cervical, axillary and epitrochlea glands not palpable. A few
small glands in both groins.
Reflexes : Pupillary, cremasteric, and plantar reflexes normal. Abdominal,
patella and Achilles reflex not obtained. No ankle clonus. No
Rhomberg.
Chest: Symmetrical. Ribs prominent. Marked sinking in of the supra-
and infra-clavicular spaces. Good expansion with inspiration.
Lungs: Percussion of the right lung shows normal resonance to the lower
border of the sixth rib in the axillary line, to the eleventh rib in the
mid-axillary line and the eleventh rib in the back. On the left normal
resonance to the eighth rib in the mid-axillary line and to the eleventh
interspace in the back. The percussion note is high pitched at
both apices. Vocal fremitus is slightly increased at both apices,
more especially on the right. The respiratory murmur is normal
throughout.
Heart: The upper border of cardiac dullness is at the fourth rib; the left
border of cardiac dullness 6.8 cm. from the mid-sternum; the right
border 0.3 cm. to the left of the mid-sternal line. Total width, 6.5
cm. The cardiac sounds are distant; the first sound of the heart
is not clear. The aortic second sound is somewhat sharper than the
pulmonic second.
Pulse: The pulses are equal and regular, but small volume. No sclerosis
of vessels noted.
PHYSICAL CONDITION OF SUBJECT DURING FAST. 61
May 14, 1912— Continued.
Abdomen: Symmetrical. Flat when standing. Much retracted when
reclining. Marked visible pulsation of the abdominal aorta. The
abdomen is everywhere flat to percussion except over the area of
stomach tympany. There is no tenderness except over the abdominal
aorta. Intestines about the size of a thumb ; can be rolled under the
finger extending from the caecum to the right hypochondrium and
from the left hypochondrium down to the brim of the pelvis. There
is slight gurgling with pressure in the right hypochondrium.
Liver: The upper border of liver dullness is from the lower border of the
sixth rib to the costal margin. Total width of liver dullness 6 cm.
Edge not felt.
Stomach: Tympany from a point 2 cm. above tip of the ensiform to a
point 8 cm. above the umbilicus. Total width 9.5 cm. Left border
of tympany 13 cm. from median line. Rhythmic sounds heard.
Spleen: The upper border of splenic dullness is at the eighth rib. Area
of splenic dullness vaguely determined as 6X5 cm. The spleen
was not felt.
Kidneys: Right kidney readily palpated. The pole of left kidney is felt.
Genital organs: Aside from a long prepuce and a slight variococele on the
left, the penis and testicles are normal.
May 15, 1912 {two hours after breaking fast) :
General appearance: Features slightly drawn. Walks erect but delib-
erately, as if fatigued. No unsteadiness in gait. The tongue is clean
and there is no odor to the breath. Otherwise no change in the
physical examination from notes of the thirty-first and last day of fast.
May 16, 1912 (24 hours after breaking fast) :
General appearance: Features quite drawn. Walks hesitatingly and a
little unsteadily. Marked lassitude. Voice weak and faltering.
Mouth: Mucous membrane of lips and tongue dry. Tongue slightly
coated posteriorly.
Abdomen: When subject is standing, abdomen is symmetrical and slightly
more prominent than on last day of fast. When reclining, the
abdomen is not so retracted. The right half of the abdomen is
somewhat more prominent than the left. The right half of the
abdomen is tympanitic; the left half flat to percussion. Palpation
of the right half elicits some rumbling of gas and liquid and causes
slight pain. Rhythmic sounds are heard on the right half.
Liver: Upper border of dullness at upper border of sixth rib, lower border
at the costal margin. Total width, 7.5 cm. (1.5 cm. wider than on
last day of fast). Edge not palpable.
Stomach: Tympany from a point 2.5 cm. above ensiform to a point 8 cm.
above the umbilicus. Total width, 10 cm. Left border of tympany
17 cm. from median line. Rhythmic sounds heard. Otherwise the
physical examination is the same as on the last fasting day.
May 18, 1912 (3 days after breaking fast) :
General appearance : Features appear thin but not drawn. Walks slowly
but with no unsteadiness. Voice more natural.
Mouth: Tongue is clean.
Heart: The left border of cardiac dullness is 7 cm. from the mid-sternal
line, the right border at the mid-sternal line. Total width 7 cm.
(0.5 cm. larger than on last fasting day.) Both sounds of heart
distant but clear and distinct.
62 A STUDY OF PROLONGED FASTING.
May 18, 1912— Continued.
Pulse : Good volume to pulse.
Abdomen : Slightly more prominent when standing. Abdomen full when
reclining. Symmetrical. Tympanitic all over, with no difference
between the two sides.
Liver: Upper border of liver dullness at fifth rib, lower border at costal
margin. Total width, 11.5 cm. (the same as at the beginning of the
fast). Edge not felt.
Stomach: Tympany from a point 4.5 cm. above the tip of the ensiform
to a point 5.5 cm. above the umbilicus. Total width, 13.5 cm. The
left border of tympany extends to a point 16 cm. from the median
line. Rhythmic sounds heard. Otherwise no change from the last
fasting day noted in the physical examination.
May 19, 1912 (4 days after breaking fast) :
General appearance: Features drawn. Conjunctivae injected (from weep-
ing). Forehead bathed in cold perspiration. Marked tremor to
hands. Gait unsteady. Walks as if quite weak physically. (Gen-
eral condition that of hysteria.)
Tongue: The tongue is clean.
Heart: Left border of cardiac dullness 7 cm. from median line. Right
border 0.5 cm. to right of mid-sternum. Total width, 7.5 cm.
Sounds clear and more distinct. Aortic and pulmonic second
sounds of equal intensity.
Stomach : Tympany from a point 4.5 cm. above the tip of the ensiform to
2.5 cm. above the umbilicus. Total width, 15.5 cm. Left border
of tympany 16 cm. from the median line. Rhythmic sounds heard.
Otherwise no change from the last fasting day noted in the physical
examination.
October 19, 1912 (5 months after breaking fast) :
General appearance: General aspect not that of a well-nourished man.
Weight 126 lbs., 15 ounces (57.6 kilograms) (nude). Stands erect.
Gait normal. Skin has a muddy yellowish tinge, but is soft and
moist. Slight conjunctivitis of both eyes. Small amount of sub-
cutaneous fat. Muscles moderate in size and rather soft. No pul-
sation noted in neck, chest, or abdomen.
Mouth: Mucous membrane of lips and cheeks moist and of good color.
The tongue has a slight coat, especially on the posterior portion.
Deposit of discolored tartar on the teeth. Slight odor to breath.
Pharynx is reddened, the blood vessels injected, and some mucus
adherent.
Glands: Cervical, axillary and epitrochlea glands are not palpable. A
few small glands in both groins.
Reflexes : Pupils equal and react normally to light and distance. Abdom-
inal, cremasteric, patella, Achilles, and plantar reflexes normal.
Chest: Symmetrical, well formed. Some sinking in of the supra- and
infra-clavicular spaces. Good expansion with inspiration. No bulg-
ing of the praecordia and apex beat of the heart is not visible or
palpable.
Lungs: Percussion of the right lung shows normal resonance to the upper
border of the fifth rib in the nipple line, to the lower border of the
fifth rib in the axillary line, and to the eleventh rib in the back. On
the left normal resonance to the eighth rib in the mid-axillary line
and to the eleventh interspace in the back. Vocal fremitus is slightly
increased and expiration slightly prolonged at the right apex, ex-
PHYSICAL CONDITION OF SUBJECT DURING FAST. 63
October 19, 1912— Continued.
tending to the second rib in front and the spine of the scapula behind.
There were no rales and the lungs were otherwise negative.
Heart: Left border of cardiac dullness 9 cm. from mid-sternum. Right
border 2 cm. to right of mid-sternum. Total width, 11 cm. Sounds
clear. No murmurs. Aortic and pulmonic second sounds of equal
intensity.
Pulse : The pulses were equal, regular at rate of 82 per minute. Rhythm
regular, volume fair. No sclerosis of the vessels noted. Systolic
blood pressure, 120 mm. Hg., diastolic, 85 mm. (Riva-Rocci instru-
ment, sitting position.)
Abdomen : The abdomen is symmetrical, rather prominent when standing
but flat when reclining. It is soft, tympanitic, but no distension.
There is no tenderness on palpation. Nothing abnormal felt.
Liver: The upper border of liver dullness at the upper border of the sixth
rib. Dullness extends to 2 cm. below costal margin. Total width
of dullness 11 cm. Edge indistinctly palpable, and with no irregu-
larities.
Stomach : Tympany in the median line extended from the tip of the ensi-
form to a point 3.5 cm. above the umbilicus, a total distance of 11 cm.
The left border of tympany extended to a point 16 cm. from the
median line. Faint rhythmic sounds are to be heard with the stetho-
scope. There is no splashing with palpation.
Spleen: The upper border of splenic dullness is the eighth rib. Area of
splenic dullness vaguely determined as 7X5 cm. Spleen not felt.
Kidneys: Neither kidney palpable.
Genital organs: Aside from a long prepuce and a slight left variococele,
the penis and testicles are normal.
SUMMARY AS TO PHYSICAL CONDITION.
General appearance: The general appearance of the subject remained
good throughout the period of observation. A gradual loss of body
tissue was evident, but the changes were not marked from day to day.
The most pronounced change was in the abdomen, which became flat
as soon as he ceased to take food and was distinctly retracted after
the fifth fasting day. This for the most part appeared to be due to
the prompt disappearance of gas in the intestines. The actual loss
in body tissue appeared to be quite evenly distributed over the body,
but was most noticeable in the tissue of the back of the neck, in the
sinking in of the supra- and infra-clavicular spaces, and in the promi-
nence of the ribs.
The muscles of the extremities, which were but moderately firm at the
beginning of the fast, appeared to have softened to a slight degree.
The muscular movements became less active after the seventh fasting
day, but the impression was that of muscular fatigue rather than
weakness. The gait was always steady and there was no swaying
of the body while standing with the feet together and the eyes closed.
The features frequently appeared drawn after the first week, but this
was present, as a rule, only during periods of mental depression.
The tremor of the hands, the weakness of the muscular movements,
and the changes in the voice noted at the end of the fast and after
breaking the fast were apparently a part of his hysterical condition.
The muddy yellowish tinge to the skin did not change throughout
64 A STUDY OF PROLONGED FASTING.
The conjunctivitis present at the beginning improved slightly after
the ninth day of the fast, but was even more marked after the fast
was broken, probably because of weeping.
Mouth : The color of the mucous membrane remained good throughout.
No change was noted in the teeth (the teeth were not brushed through-
out the period of observation). The slight coating on the tongue
became more pronounced until the ninth fasting day, when it began
to disappear slowly. The tongue did not become entirely clean
until the third day after food was taken. On the third day the odor
of the breath was offensive, becoming fetid. After the ninth day this
was less pronounced and gradually decreased until the twenty-third
day, after which time very little odor was noticed.
On the ninth fasting day the mucous membrane of the mouth and lips was
dry. On the eleventh day the lips were desquamating (at the same
time the seborrhea of the scalp appeared) and this continued until
the fifteenth day of the fast and did not occur again. The decrease
of these signs promptly followed the prescribed intake of larger
quantities of water. No change was noted in the chronic naso-
pharyngitis.
Glands : No change whatever was noted in the cervical, axillary, epi-
trochlea glands, or the glands in the groins.
Nervous system: During the first week of the fast the mental attitude
was a cheerful one. He was enthusiastic about the experiment but
very opinionated. From this time on to the end of the fast he was
frequently depressed and sometimes irritable. On these days he was
disinclined to talk, his physical movements were more deliberate,
and he was more sensitive to any physical discomfort, such as
pressure of the hands. He attributed this to the depressive effects
of the rain and cloudy weather, but the impression was that he felt
actual physical fatigue. On the last day of the fast and for the
remaining period of observation he exhibited varying mental states
of depression, irritability, and sullenness, weeping at times. Taking
the period as a whole there was a gradual increasing depression and
irritability, which, with the onset of abdominal pain, manifested
itself in hysterical conditions. There was no outward demonstration
of any mental improvement as a result of the fast.
Reflexes: The pupillary, plantar, and cremasteric reflexes were normal
through the fast. The patella reflex gradually diminished and on
the fifth day could only be obtained by re-enforcement and on the
ninth day and during the remainder of the fast could not be obtained.
The abdominal reflexes were absent on and after the ninth day, and
the Achilles on and after the eleventh day. Five months after, the
reflexes were normal.
Chest : No change was noted in the chest, with the exception of the gradual
sinking in of the supra- and infra-clavicular spaces and the promi-
nence of the ribs as the subcutaneous tissues disappeared.
Lungs: As the loss of tissue in the supra-clavicular space progressed, the
percussion note of the apices became slightly higher pitched. The
lower border of resonance of the right lung followed the decrease in
the size of the liver. No change was noted in the respiratory murmur.
Heart: There was a gradual decrease in the percussion border of the heart
noted as follows, the total diminution in size of heart during fast
being 4.2 cm. :
c. 5
FASTING
PLATE 5
PHYSICAL CONDITION OF SUBJECT DURING FAST.
65
From mid-sternum.
Total width.
Left border.
Right border.
cm.
9.5
8.5
8.0
7.5
7.0
6.8
7.0
8.0
9.0
cm.
1.2
1.0
1.0
0.5
0.0
0.3 to left
0 . 5 to right
0 . 5 to right
1.5
cm.
10.7
9.5
9.0
8.0
7.0
6.5
7.5
8.5
10.5
Seventeenth day of fast
Nineteenth day of fast
Twenty-fifth day of fast
Third dav after fast
Fourth day after fast
5 months after fast
Heart sounds: No change was noted in the character of the heart sounds
until the eleventh fasting day, when the sounds were less distinct and
a systolic souffle was heard all over the praecordia. This souffle was
not heard after the fifteenth day, but the sounds remained more dis-
tant throughout the fast and after the twenty-fifth day the first
sound of the heart was not distinct. After the twenty-third day the
aortic second sound was more distinct than the pulmonic second.
The systolic souffle appeared at the same period as the seborrhea,
the dry mouth, and desquamating lips and disappeared when the
water intake was increased.
Abdomen: On the first fasting day most of the gas disappeared from the
intestines. The abdomen became retracted and by the fifth day had
reached its maximum, the pulsation of the abdominal aorta being
pronounced. After this time it was everywhere flat to percussion
except over the area of gastric tympany. After the fifteenth day of
the fast the contracted large intestines could be palpated along the
course of the ascending and descending colon. On the twenty-third,
twenty-ninth, and thirty-first days there was a slight gurgling of gas
and liquid in the right hypochondrium with pressure.
Twenty-four hours after breaking the fast the ascending colon was dis-
tended with gas, but the descending colon was still contracted. On
the third day after breaking the fast the whole abdomen was tym-
panitic and symmetrical.
Liver: A gradual decrease in width of liver dullness as measured in the
nipple line was noted, as shown in the table. The total diminution
of the width of liver dullness in the nipple line was 5.5 cm. or 47.8
per cent of the total width at the beginning of the fast. On the third
day after taking food, the liver was found to be of the same width
as at the beginning of the fast.
Total width.
First day of fast
11.5 cm., edge palpable.
10.5 cm., edge not palpable.
9.5 cm.
9.0 cm.
8.6 cm.
7.4 cm.
6.0 cm.
7.5 cm.
11.5 cm.
11.0 cm., edge palpable.
Third day of fast
Fifteenth day of fast
Twenty-first day of fast
Twenty-fifth day of fast
Thirty-first day of fast
Three days after breaking fast . . .
Five months after breaking fast. .
Spleen. No change was noted in the spleen during the fast.
66 A STUDY OF PROLONGED FASTING.
Kidneys: On the thirteenth day of the fast the right kidney was palpable
and on the twenty-first the left was palpable and both remained so
during the rest of the fasting period. Five months later neither
kidney was palpable.
Genital organs: No change noted in the genital organs during the fast.
Observations of the physical condition of Breithaupt, who fasted for 6 days,
of Cetti, who fasted for 10 days, and of Beauts, who fasted for 14 days,
failed to show any change in the size or position of the organs.
PHOTOGRAPHIC STUDY OF SUBJECT.
The most striking external evidence of prolonged inanition with a
fasting subject is the degree of emaciation. In order to visualize this
as much as possible as the fast progressed, an extensive series of photo-
graphs was taken practically once a week. At these times the calo-
rimeter laboratory was specially warmed with gas stoves, particularly
in the part of the room where L. was to pose, and screens were put in
place for the background. The subject undressed and put on a small
loin cloth; he was then posed on a low pedestal, which was covered
with black cloth. In the selection of poses we had the valuable advice
of Professor W. G. Anderson, of Yale University.
Probably no routine throughout the whole fast pleased the subject
more than this series of photographs, as he seemed obsessed with the
desire to have himself photographed. We were accordingly able to
obtain a large number of photographs of the subject in a variety of
poses. Several of those obtained on the first and last days of the fast
are given in Plates 4 and 5. In the latter part of the fast the subject
became somewhat less sure of his footing and rested lightly against a
wooden frame. A rough approximation of the measurements of this
man may be made by using 640 mm. as the inside distance between
the wooden uprights. Lack of time prevented our adjusting the
accurate mirror arrangement of Friedenthal1 for securing photographs
that could subsequently be measured. It should be considered,
however, that the chief reason for taking this series of photographs
was to visualize the general appearance of emaciation and not to furnish
material for exact measurements of the loss of tissue. This was sup-
plied by the accurate measurements made according to the regular
schedule by Professor W. G. Anderson. (See p. 68.)
In addition to the anatomical photographs, a great many photo-
graphs were taken of L. at his own desire, since this seemed to be
the one thing which would amuse him at any time. Accordingly the
camera was pointed at him several hundred times throughout the
course of the fast, although admittedly many of these were false expos-
ures. A considerable number of photographs were thus obtained which
show him in his environment, some of which are deemed worthy of
reproduction (see Plates 1, 2, and 3, pages 11, 19, and 31).
friedenthal, Med. Klinik, 1909, No. 19, p. 1.
ANTHROPOMETRIC MEASUREMENTS.
The importance of careful anthropometric measurements for noting
the diminution in size of the body as the fast progressed has been recog-
nized by all writers on fasting. Fortunately Professor William G.
Anderson, of Yale University, was in Boston during the period when
this fasting experiment was being made and he kindly offered to make
a series of anthropometric measurements of the subject. In this he
was assisted by Dr. W. L. Anderson. These measurements were made
approximately once a week, Dr. W. L. Anderson making a special
trip from New Haven to complete the series at the end of the fast.
Professor Anderson reports that "the measurements of the forearm
are taken with the hand tightly closed and the wrist slightly flexed.
The measurements of the upper arm are taken at the largest part after
the elbow is completely flexed and all flexors and extensors contracted
to their utmost. In taking the measurements of the calf and thigh, we
select the largest part after the man has contracted the muscles as
well as he can while in the standing position."
The measurements for each week are given in table 1, the total
decreases in the various girths for the whole fast being given in the last
column. As would be expected, the largest change in girth was at the
waist, there being a decrease of 153 mm. The girth of the abdomen
decreased 119 mm., while a large decrease is shown for both thighs
and for the chest. Certain of the measurements were not made until
the second or third examination; the losses are therefore inclosed in
parentheses to indicate that the series of measurements was not
complete. The distinct loss in practically all measurements is obvious.
In the hope of securing some evidence in regard to the muscular
strength of the subject, Professor Anderson brought with him his
dynamometers to test the strength. To our great surprise, the subject
even before the fast began refused to carry out any of these tests,
stoutly maintaining that he was not an athlete but a professional
gentleman and that he was not accustomed to doing muscular work of
any kind. This was wholly in line with his attitude toward other
muscular- work tests which were contemplated, but which it was neces-
sary to omit, greatly to our regret. The only evidence that we have
regarding the muscular strength of the subject is the material obtained
in the dynamometer tests which were secured every afternoon by Profes-
sor Langfeld. Even regarding these we are somewhat uncertain as to
whether the subject exerted his greatest strength in all the tests. The
pressure which he placed upon the dynamometer was clearly influenced
by his fear that such pressure might give him a little pain, to which he
was strongly averse.
67
68
A STUDY OF PROLONGED FASTING.
Table 1. — Measurements of subject L.
Measurement.
April
11,
1912.
April 18,
5th day
of fast.
April 25,
12th day
of fast.
May 2,
19th day
of fast.
May 8,
25th day
of fast.
May 14,
31st day
of fast.
Total
loss.
Height:
Standing
Sitting
Girth:
Neck
Chest, normal
full
empty
Ninth rib, full
empty
Tenth rib
Abdomen
Waist
Hips
Right biceps, extended . .
flexed
Right forearm, extended .
Left biceps, extended1 . . .
flexed
Left forearm, extended . .
Right thigh
Right calf
Left thigh
Left calf
Breadth:
Shoulders
Chest, full
empty
Hips
Depth:
Chest, full
empty
Abdomen
mm.
1707
376
879
930
828
874
787
800
780
251
279
254
254
284
262
488
335
488
345
mm.
1707
371
871
937
823
881
787
785
785
749
241
269
244
251
272
254
465
335
470
338
254
188
251
183
mm.
1707
874
368
856
904
813
866
775
770
742
696
234
259
236
241
267
246
450
323
457
320
419
279
254
312
241
211
163
mm.
1704
884
361
825
876
792
848
767
762
757
673
813
226
249
234
229
262
241
432
310
432
315
424
272
254
307
241
211
160
mm.
1707
884
338
805
864
787
825
759
749
686
648
805
221
236
229
224
246
241
427
305
406
302
417
269
254
282
229
201
170
mm.
1707
881
335
800
851
782
820
754
749
681
627
792
211
239
224
218
239
229
394
300
399
295
419
254
239
279
229
203
152
mm.
0
(—7)
41
79
79
46
54
33
(36)
119
153
(21)
40
40
30
36
45
33
94
35
89
50
(0)
(25)
(15)
(33)
25
(8)
36
xThe subject was left-handed.
BODY-WEIGHT.
To the ordinary individual the most striking index of the severity of
a prolonged fast is the loss in body-weight, the abstinence from food
resulting in great emaciation. The fact that even in the short space
of 24 hours the body-weight changes considerably is not so patent, and
a consideration of these changes is of interest. If the body- weight is
determined each hour throughout the day, it will be seen that while
sudden fluctuations accompany the ingestion of food, the voiding of
urine, or the passing of feces, there is a general tendency for a regular
fall in the body-weight from hour to hour amounting, with adults, to
not far from 40 grams per hour.1 During the night, the decrease in
the body-weight is regular, although not quite so rapid. Since such a
tendency is shown in the course of 24 hours, it would normally be
expected that it would be more especially evident in the 20 or 30 days
of a fast and that the body-weight would decrease steadily as the fast
progressed.
ROUTINE OF OBSERVATIONS.
The losses in body-weight have usually been recorded in every
reported fast, whether scientifically controlled or not. Unfortunately,
however, the observations vary in value, as the weighings have not
always been made under constant conditions. At times they even
show a gain rather than a loss. Comparable results in such observa-
tions may be secured by the following routine :
The weights should be taken at approximately the same time each day.
If the subject is not weighed nude, the clothing worn should be approxi-
mately of the same weight, and its weight should be deducted from
the total weight recorded, thus giving the true value for the body-
weight of the subject.
The bladder should always be emptied immediately or a short time before
the weighing.
The amount of drinking water taken prior to the weighing should be
constant.
No water should be taken for some hours before the observation is made.
The weighings should always be made on the same carefully calibrated
scales and should be checked by a second observer.
The environmental temperature and the muscular activity should be
approximately constant throughout the whole period of the fast.
As usually fasting subjects are very captious, investigators are ordi-
narily content to control them only in so far as the collection of the
excreta and the abstinence from food are concerned, without rigor-
ously insisting upon their remaining in a quiet, closed room during
'Benedict and Carpenter, Carnegie Inst. Wash. Pub. No. 126, 1910, p. 113. Benedict and
Joslin, Carnegie Inst. Wash. Pub. No. 176, 1912, p. 90.
69
70 A STUDY OF PROLONGED FASTING.
the whole period of the fast, with a constancy in the muscular activity.
In the long fasting experiment with L., however, the routine for weigh-
ing previously outlined was followed very closely.
The subject emptied the bladder immediately after leaving the bed
calorimeter each day about 8 a. m. A respiration experiment of three
or four 15-minute periods was next made with him. This was usually
finished about 9h 30m a. m. He was then carefully weighed on a
calibrated platform balance, the weighings and records being made by
Mr. Carpenter and checked by a second observer. (See plate 2,
figure D, page 19.) The scales used were the so-called "silk" scales,
capable of weighing 150 kilograms with a sensitivity of 10 grams with
a full load. The temperature of the calorimeter room was rarely
below 20° C, but as the subject was used to a warmer climate he was
especially sensitive to cold. He was therefore not weighed nude, but
in a cotton union suit and socks which had been washed in distilled
water. He also wore over this union suit his heavy woolen underwear.
The exact weight of this clothing was known and deducted from the
weight shown on the scales. It was not practicable to make the weigh-
ing directly after he had emptied the bladder, as it seemed undesirable
to have him stand so long before the respiration experiment began.
No water was taken during the night, so that when the subject was
weighed he had been without water for some 12 hours. Furthermore,
the amount of water taken during the day was approximately constant
in quantity, i. e., for the first 10 days of the fast 750 c. c. and for the
remaining days 900 c. c. During the night the subject had remained
in the calorimeter chamber under constant temperature conditions,
and as he usually lay very quietly, the activity was at a minimum.
While the temperature conditions and muscular activity necessarily
varied somewhat during the day, they were fairly constant, especially
as the subject was by nature averse to muscular activity.
As L. was extremely interested in the records of the body-weight
from day to day, the change in weight was computed and roughly
plotted daily in the form of a curve on the blackboard in the calorimeter
laboratory. Any irregularities in the curve would be instantly detected
and verifications made if necessary. As a matter of fact, such verifi-
cation never indicated a discrepancy and we have the fullest confidence
in this series of weights.
While the time relations were not theoretically ideal, they were as
nearly so as was practicable with the large number of observations
necessary to be made simultaneously upon this man. A sample day's
computation of the loss of weight is as follows :
Body-weight in cotton underwear, socks, and heavy woolen underwear, kilos.
9h 35m a.m., April 20, 1912 57.37
Weight of cotton underwear and heavy woolen underwear. 1 .48
Naked body-weight, 9h 35m a.m., April 20, 1912 55 . 89
Naked body-weight, 9h 40m a.m., April 19, 1912 56.37
Loss in body-weight, April 19-20, 1912 0.48
BODY- WEIGHT. 71
DAILY LOSSES IN BODY-WEIGHT IN FASTING EXPERIMENTS.
The loss in body-weight in fasts of short duration has been exten-
sively discussed in a former publication.1 Since the appearance of this
book, several other fasts have been reported which were but super-
ficially mentioned there. We have accordingly gathered together in
table 2 the records of body-weight obtained in a considerable number of
fasting experiments continuing 14 days or more. The largest number
of fasting experiments with any one man has been made with the pro-
fessional faster, Succi. The scientific aspects of these experiments
have become world-renowned by means of the classical research of
Luciani,2 which was carried out in Florence in 1890 and has never been
equaled as a careful analytical study of prolonged fasting. In the
course of this report it will be occasionally necessary to call into ques-
tion Luciani's conclusions, but the reader is particularly requested
to consider that since the publication of Luciani's work a quarter of a
century has passed and that the criticisms raised must be chiefly of the
technique rather than of the interpretation of the results by the Italian
master.
Table 2 includes not only the data for the seven fasts of Succi, but
also the records of the body-weights secured for three other individuals,
i. e., Jacques, Beaute, and the fasting woman Schenk. In examining
these records, it will be seen that in some of the experiments Succi
had a body-weight some 13 or 14 kilograms greater than in others.
Several of these observations also show actual gains in weight, as, for
instance, two records in Succi's fast in Florence, one record in the
Naples fast, and five records for the fasting man Jacques. Usually
the protocols for the experiments explain these apparent gains as being
due to changes in the amount of water consumed or in the time of
weighing.
An examination of the losses of weight in these fasts shows that in
general the larger losses were found in the first days of the fast, although
on the twelfth day of the Paris fast and the eleventh day of the Milan
fast, Succi lost more than 1 kilogram of weight. The usual losses in
the later days of the fast were from 300 to 400 grams in a day. Although
occasionally records are found of a loss of only 100 grams or less in a
day, such values are open to suspicion and are generally accounted for
by errors in weighing or lack of control of conditions. These minimum
losses are by no means a 'priori evidence that the fast was not genuine
so far as abstinence from food was concerned, since irregularities in the
amount of drinking water and particularly in the length of time inter-
vening between the drinking of water and the weighing, irregularities
in the voiding of urine as compared with the time of weighing, and
Benedict, Carnegie Inst., Wash. Pub. No. 77, 1907, p. 301.
2Luciani, Fisiologia del digiuno. Florence, 1889. Das Hungern. Translation by M. C.
Fraenkel. Hamburg and Leipsic, 1890.
72
A STUDY OF PROLONGED FASTING.
changes in the environmental temperature or muscular activity will
of course increase or decrease the regular loss of material.
Table 2 also gives the records of the body-weights obtained in the
fasting experiment with our own subject L. The greatest loss shown
during the 31 days of the fast was 1.04 kilograms on the first day and
the smallest loss in weight was 0.11 kilogram on the thirteenth day.
Table 2. — Losses of body-weight by fasting subjects, with initial weight and weiglii
on each day of fast.
(Weight given in kilograms.)
Day of
fast.
Succi.
Paris,
1886.
Milan,
1886.
Florence,
1888.
London,
1890.
Naples,
1892.
Rome,
1893.
Zurich,
1896.
Wt.
Loss.
Wt.
Loss.
Wt.
Loss.
Wt.
Loss.
Wt.
Loss.
Wt.
Loss.
Wt.
Loss.
Initial wt.
1st
2d .. ,.
3d
4th
5th
6th
7th ... .
8th
9th
10th. , .
11th... .
12th,
13th
14th
63.30
62.40
61.00
69.80
59.90
59.30
58.65
58.15
57.65
57.25
56.70
56.25
55.60
55.25
54.85
54.60
54.30
54.10
53.65
53.20
52.80
52.60
62.25
51.85
51.45
51.50
51.30
51.25
51.05
50.45
65.10
64.70
64.00
63.30
62.80
61.20
60.90
60.70
69.30
59.10
59.00
59.00
58.90
58.55
58.20
58.00
57.80
57.60
57.50
57.05
56.50
0.40
.70
.70
.50
1.60
.30
.20
1.40
.20
.10
.00
.10
.35
.35
.20
.20
.20
.10
.45
.55
71.70
63.00
59.40
59.00
58.00
57.40
57.10
57.00
56.90
56.00
55.70
55.30
54.00
53.20
3.60
.40
1.00
.60
.30
.10
.10
.90
.30
.40
1.30
.80
61.30
59.75
58.95
58.20
57.70
56.85
56.30
56.10
55.60
55.40
54.40
54.30
54.00
53.60
53.10
52.85
52.60
52.10
51.35
51.15
50.90
50.90
50.60
50.15
49.70
49.40
49.00
48.70
48.50
48.20
1.55
.80
.75
.50
.85
.55
.20
.50
.20
1.00
.10
.30
.40
.50
.25
.26
.60
.75
.20
.25
.00
.30
.45
.45
.30
.40
.30
.20
.30
6.90
1.40
1.20
+ .10
.60
.65
.50
.60
.40
.56
.45
.65
.35
.40
.26
.30
.20
.45
.45
.40
.20
.35
.40
.40
+ .05
.20
.05
.20
.60
55.80
54.90
53.90
52.80
52.10
51.50
51.00
50.40
50.10
49.80
49.70
49.30
48.90
48.70
48.30
48.10
47.55
47.25
47.10
46.80
46.50
46.30
45.90
45.60
45.40
45.20
44.90
44.60
44.30
44.20
44.10
43.80
43.70
43.50
43.20
43.00
42.75
42.60
42.30
41.70
0.90
1.00
1.10
.70
.60
.50
.60
.30
.30
.10
.40
.40
.20
.40
.20
.55
.30
.15
.30
.30
.20
.40
.30
.20
.20
.30
.30
.30
.10
.10
.30
.10
.20
.30
.20
.25
.15
.30
.60
63.60
61.80
60.60
59.80
59.10
58.20
57.50
57.10
56.90
57.20
56.60
55.90
55.50
55.40
55.30
55.25
54.60
54.00
53.50
53.00
52.40
1.80
1.20
.80
.70
.90
.70
.40
.20
+ .30
.70
.60
.40
.10
.10
.05
.65
.60
.50
.50
.60
68.50
68.00
67.55
67.25
66.55
3.201
.50
.45
.30
.70
65.70
65.40
65.00
64.55
64.05
.85*
.30
.40
.45
.50
15th
16th...
17th. ,
18th...
19th...
20th
21st ,
22d
53.00
52.70
52.50
52.20
51.90
51.60
51.20
.20s
.30
.20
.30
.30
.30
.40
63.50
63.00
62.75
62.50
62.20
61.90
.552
.50
.25
.25
.30
.30
23d
24 th...
25th... ,
26th...
27th, ,
28th
29th...
30th, .
31st
50.75
50.20
50.10
50.00
49.65
49.50
49.25
48.75
.45"
.55
.10
.10
.35
.15
.25
.50
32d
33d
34th
35th
36th
37th
38th
39th
40th
1 Loss for 3 days.
2 Loss for 2 days.
BODY- WEIGHT.
73
Table 2. — Losses of body-weight by fasting subjects, with initial weight and weight
on each day of fast — Continued.
(Weight given
in kilograms.)
Jacques
, 1888.
Beaute, 1907.
Schenk, 1906.
L., 1912.
Day of fast.
Weight.
Loss.
Weight. ]
jOSS.
Weight.
Loss.
Weight.
Loss.
Initial wt. . .
62.01
65.61
56.3
60.64
1st
60.68
1.33
64.57
1.04
54.4
1
.90
69.60
1.04
2d
59.74
.94
63.72
.85
53.6
.80
58.68
.92
3d
59.23
.51
62.77
.95
53.2
.40
57.79
.89
4th
59.24
+ .01
61.96
.81
52.5
.70
57.03
.76
5th
58.98
.26
61.41
.55
51.9
.60
56.37
.66
6th
58.35
.63
60.83
.58
51.2
.70
55.89
.48
58.55
+ .20
60.23
.60
50.9
.30
55.50
.39
8th
56.68
1.87
60.04
.19
50.6
.30
55.08
.42
9th
56.23
.45
59.80
.24
50.5
.10
54.63
.45
10th
56.23
.00
59.11
.69
50.2
.30
54.13
.50
11th
55.80
.43
58.64
.47
49.9
.30
53.88
.25
12th
55.60
.20
58.64
.00
49.5
.40
63.56
.32
13th
54.67
.93
58.37
.27
49.1
.40
53.45
.11
14th
54.67
.00
57.78
.59
48.8
.30
53.15
.30
15th
55.04
+ .37
48.4
.40
52.84
.31
16th
54.98
.06
48.2
.20
52.26
.58
17th
55.06
+ .08
51.79
.47
18th
54.81
.25
51.50
.29
19th
53.93
.88
51.11
.39
20th
53.93
.00
50.93
.18
21st
53.82
.11
50.49
.44
22d
53.36
.46
50.13
.36
23d
53.00
.36
49.96
.17
24th
52.74
.26
49.62
.34
25th
52.46
.28
49.33
.29
26th
52.40
.06
49.02
.31
27th
51.89
.51
48.70
.32
28th
52.23
+ .34
48.46
.24
29th
51.86
.37
48.10
.36
30th
51.69
.17
47.69
.41
31st
47.39
.30
Every effort was made to secure uniformity of conditions throughout
this fast, and probably in no long fast with man have these ideal con-
ditions been so nearly approached. Yet, even with this care, it will be
seen that the losses were by no means regular from day to day, although
the variations are not so great as in the other fasts referred to in table 2.
Unquestionably the loss in weight in a strictly controlled fast may
be considered a good general measure of the intensity of metabolic
processes; yet with such wide fluctuations as are shown for L. it hardly
seems probable that the body-weight can be looked upon as an accurate
index of the total tissue change. But attempts have frequently been
made by investigators to establish a mathematical relationship between
the daily loss in weight and the length of the fast. Luciani, basing his
conclusions upon the results of his study with Succi in Florence and
particularly upon two long experiments with dogs, was confident that
such a mathematical relationship existed. To study the possibilities
74
A STUDY OF PROLONGED FASTING.
1
(IIOS
3
s
7
9
II
1?
IS
n
19
21
23
2S
27
29
3f 33 35 37
3<
\
)AY i OF
Fast
670
8U
\
\
64i
HA
\
v
1
\
ZURICH
\
\
v 1896
\
\
\
IV
\ 1
\
\
\
1
\
V
\\
V
\
\
ROMI
WO
\
\
\ \
V
v^
\
89
3
\
\
0
\
1,
\
\
\
\
\
s
L 1
89
2
'10
\
\
\
\
\\
V
1
\
V
^
m
\
\
\
\
V
\
'.FLORENCE
sin
V
\
\_
IIS
88
k
\\
\
/
M
V.
\
N
1
\
\
\
w
kRI
)
1
411
\
\
M
IL
1
J8
\
6^
\
470
4U
mti
LONDON
440
1890
Fig. 1. — Body -weight curves for fasting experiments
with Succi.
BODY- WEIGHT.
75
3 5
7
9
ii 13 is n
19 2
23 25 27 29 31
\ILOS
610
D.
kYS
OF 1
"AST
!
\J
60.0
590
5fl0
570
\
S
\
560
\
^
\\
JV>0
\
MO
•no
s?n
51.0
*\
son
440
480
47.0
\
Fig. 2. — Body-weight curve for Levanzin.
76 A STUDY OF PROLONGED FASTING.
of this relationship, curves have been plotted showing graphically the
changes in weight during each of Succi's fasts. (See figure 1.) A
similar curve has been plotted showing the records of body-weights
during the fasting experiment with L. (See figure 2.)
If we analyze the components which make up the loss in weight
of the body, we find it is due not only to the loss of body tissue which is
oxidized to supply material for the maintenance of the body activity,
but to the loss of preformed water, i.e., the water existing in the tissues
oxidized. According to observations in some of the earlier fasting
experiments in Middletown, Connecticut, this preformed water, which
varies widely in amount, appears to be rapidly discharged in the first
days of fasting. We would consequently expect to find that the curves
for a fasting experiment would indicate a rapid fall in weight at the
beginning of the fast, the percentage loss becoming gradually smaller,
until the body-weight curve tends to become a straight line. If the
curves for Succi and Levanzin are examined, this tendency will be seen.
On the other hand, while all the curves have the same general trend, a
careful mathematical analysis shows no regularity that would justify
the use of a mathematical expression by means of which losses of weight
may be predicted during prolonged fasts. When one considers that
only the Florence fast was controlled by Luciani, and that the others
were made in different years, at different seasons, and in different
countries, it will be seen that but little can be expected from a com-
parison of these curves.
Nevertheless, the semblance of mathematical regularity shown in
the records of body-weight obtained in Succi's Florence fast and in the
experiments with dogs led Luciani to seek the aid of his associate,
Bufalini, who computed that the body-weight curves, especially those
obtained in experiments on dogs, (see P and P' on figure 3) had a ten-
dency to represent an equilateral hyperbola. Reasoning from the
equilateral hyperbola equation obtained with dogs, Luciani computed
the probable curve for Succi's weights during the Florence fast and
found that the loss in weight was very much less than he would have
expected. He interpreted this as being due to the fact that Succi
drank much larger amounts of water than did the dogs and that water
apparently acted as a nutrient, thus sparing the tissues.
Since a reasonable regularity was also shown in the course of the
curve obtained for L., a probable curve for this subject was developed
by Mr. E. H. Lange, physicist of the Nutrition Laboratory. (See
curve in light line in figure 2.) Using W to represent the weight in
kilograms and T the time in days, the weight for any given day is
found by the formula:
TF=3.20 (10)-0143r-0.32477+57.43
BODY-WEIGHT.
77
A similar equation has been worked out for Succi's London fast, as
follows :
TT=4.98 (10)-00692r-0.22277+50.75
(See curve in light line in fig. 1.) Obviously such a complicated curve
can not in any wise be considered a simple mathematical relationship.
0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 SO 85 90 95 100 105 110 115 120
PRO
Kfts.
?7 0
DA
fS 0
f FA
ST
26X1
\\
25.0
v
\HA
M
1 FAE
T
?4.0
\
v
?30
HAW
1"
A
FAST
22.0
?1.0
V
20.0
\a
WROR
»3
3W
10.
I
19.0
Kgs. ~*
1
0J>
N
iao
\a
WROR
V4
DW
i *,n
l
170
7.0
^
\
16.0
AWRO
S«2
ROW
«!<5 —
v.
i<sn
•v «S.fl
1
met
\p
Nl
4*0
13j0
\
3
0
V
^A
Vlu
V
JANI
ex>
6.0
\
7.0
\
Fia. 3. — Body-weight curves for prolonged fasting experiments with dogs.
Ideal conditions for studying the loss in weight during fasts would
be those in which the subject had a constant amount of drinking water,
emptied the bladder completely at a definite period each day, remained
in an absolutely constant environment with a constant temperature,
78 A STUDY OF PROLONGED FASTING.
and the metabolism pursued a course entirely unaffected by extraneous
conditions. Such constant conditions are impracticable with human
beings, but are more easily obtained with animals. Several remarkably
long fasting experiments have been made with dogs, but the records are
for the most part not easily accessible. Curves showing the records of
body-weight in these animal experiments are given in fig. 3. The body-
weights obtained by Luciani on two dogs, one in Siena and another
in Florence, are represented by the curves designated as P and P\
These dogs were catheterized each day, were given exactly 150 c. c.
of water daily, and were kept in a room free from disturbance of any
kind and in a temperature of approximately 12° to 13° C. It will
be seen that the loss in weight follows a fairly regular course, save on
the last few days of each fast, none of the irregularities shown in the
body-weights of Succi and our subject L. being apparent.
Luciani's two experiments with dogs continued 34 and 43 days
respectively, but still longer experiments have been made by P. B.
Hawk, in which a dog fasted in two experiments of 117 and 104 days
respectively. The complete results of these experiments have not yet
been published,1 but the investigator has kindly given me the privilege
of using certain of the data in this connection. The body-weights
obtained during these fasts are also shown in figure 3. The amount
of drinking water given to the dog was constant, but the animal was
not kept in a chamber with even temperature and the other conditions
were not so well controlled as in Luciani's experiments, as the purposes
of the fasts were entirely different. Aside from slight fluctuations,
however, the curves follow a reasonably constant course. These
curves are of particular value owing to the extraordinary length of the
fasting periods.
A series of observations made by Awrorow on dogs with complete
fasting is of even more interest in studying this specific problem.
The dogs were confined in the Pashutin respiration chamber in the
Imperial Medical Academy in St. Petersburg, receiving neither food
nor water. They were catheterized daily, weighed at a regular hour,
and spent 22 or more hours out of the 24 hours in the quiet and isolation
of the respiration chamber, during which the carbon-dioxide production
was carefully measured. The period of fasting with the four dogs
continued for 16, 44, 60, and 66 days respectively. Under these con-
ditions one would expect a most regular progress in the metabolism,
with constant loss of water and organic material, and changes in the
body-weight. That this is true to a marked extent is shown from an
examination of the three curves for the dogs which fasted the longest,
i. e., 44, 60, and 66 days. The striking regularity of these curves bears
out completely Professor Luciani's view that if such experimental con-
'Howe and Hawk, Am. Journ. Physiol., 1912, 30, p. 174; Howe, Mattill, and Hawk, Journ.
Biol. Chem., 1912, 11, p. 103.
BODY-WEIGHT. 79
ditions can be secured in a fasting experiment, the curve will be extra-
ordinarily regular.1
As an effort was made to secure constant conditions in the experiment
with our fasting subject, it was hoped that a curve approximating the
regularity of the curve for Awrorow's dogs could be obtained, but an
examination of the plotted values for the daily body-weights shows
that this was far from being the case. It would be practically impos-
sible to carry out a lengthy fasting experiment with a man with envi-
ronmental conditions as constant as were those of the dogs used by
Awrorow. We know also that the loss of material varies greatly both
in amount and in kind with the progress of the fast. Thus there is
always greater metabolism, greater activity, and a greater disintegra-
tion of material in the first few days of fasting. As the fast progresses,
the effect becomes more or less noticeable, and the subject becomes
disinclined to active muscular work, thus naturally conserving his
energy. Furthermore, human subjects, by covering themselves with
extra clothing and preferring warm rooms, attempt to conserve their
calorific output.
The character of the katabolism may also vary greatly, particularly
during the first few days of the fast, so that there is unquestionably
a rapid depletion of the glycogen storage in the body in this period.
The course of the curve would therefore vary according to whether the
subject of the experiment was well nourished, poorly nourished, or
obese. The character of the foregoing diet may likewise play a role in
this connection. It will be seen later, however, when a study is made of
the gaseous metabolism during the fast, that the possibility for analyz-
ing the daily losses in a fast such as that carried out by L. are much
greater than in any fasting experiment thus far observed with man, and
that no mathematical relationship between the length of time of a
human fast and the loss in body-weight can be expected.
JThe data for these curves were secured from the large work on fasting dogs written by Professor
Awrorow. Curves showing the percentage loss of body-weight for two of these dogs are shown on
two lecture charts kindly given me by Professor Awrorow during my 1907 visit to St. Petersburg,
and reproduced in figures 45 and 46 on pages 356 and 357 of this report.
While this report on the long fasting experiment made in the Nutrition Laboratory is designed
primarily to deal only with the influence of prolonged fasting on human metabolism, it will not
be out of place here to emphasize the fact that a colossal amount of research on fasting animals
has been accumulating in the laboratories in St. Petersburg for a number of years, chiefly under
the direction of Albitsky. A considerable portion of this material was incorporated in the second
volume of Pashutin's experimental pathology, of which 840 pages are devoted to the discussion
of fasting. This material has been considered so important to workers in metabolism that it has
been translated in the Nutrition Laboratory and typewritten copies of the translation have been
deposited in the library of the Office of the Surgeon-General of the Army, in Washington, D. C,
the John Crerar Library in Chicago, Illinois, and the New York Public Library in New York City,
the fourth copy being retained by the Nutrition Laboratory.
In addition to this material in Pashutin's book, much of which is recorded for the first time,
there are a number of large monographs published by Awrorow, Kartaschefsky, Albitsky, and
Likhatcheff , which deal with the abstract problems of metabolism and which have been translated
in whole or in part for this laboratory. It is greatly to be regretted that this collection of data,
which far exceeds both in quality and amount the total accumulation of earlier investigators on
fasting animals, should be so inaccessible to American, English, and Continental readers. It is
certainly true that no one working on the gaseous metabolism of animals can at the present day
afford to overlook this wonderful collection of Russian material.
80
A STUDY OF PROLONGED FASTING.
TOTAL LOSS IN BODY-WEIGHT.
An examination of the data in table 2 for the subjects other than L.
show apparent discrepancies in the initial weights and in the loss of
weight on the first day. The exact length of the fasting period and
the weight on the last day are also frequently doubtful. The Florence
weights were all taken from the plates at the end of Luciani's report
"Fisiologia del digiuno." It is obviously important to note whether
the initial weight was taken immediately after the meal or before
the meal, or what was the condition of the alimentary tract. With
our subject L. we believed it to be necessary to obtain an accurate
weight at the beginning of the fast ; consequently the initial weight was
taken approximately 12 hours after the last meal, several hours after
drinking water, and a definite time after urinating. Such precautions
were not taken, we believe, with any of the other subjects given in
table 2, with the possible exception of Cathcart's subject, Beaute\
The total losses in the various fasts, particularly when computed as
percentages of the initial weights, have certain features that are not
without interest. For comparison we give in table 3 the total loss and
the percentage loss for each subject at the end of 14, 16, 20, 29, 30, 31,
and 40 days respectively (using the data recorded in table 2). Obvi-
ously a comparison can be made for all of the subjects for only 14 days,
with all the subjects but one for 16 days, with all the subjects but two
for 20 days, and finally with but one subject for 40 days. At the end
of 14 days the average percentage loss was 12.6 per cent. The lowest
loss was with Succi, in the Rome fast, of 10.6 per cent; and the highest
loss was 15.7 per cent with the same subject, in the Paris fast. For
Table 3. — Summary of losses of body-weight by fasting subjects.
Subject.
14 days.
16 days.
20 days.
29 days.
30 days.
31 days.
40 days.
Kilos.
Per
cent.
Kilos.
Per
cent.
Kilos.
Per
cent.
Kilos.
Per
cent.
Kilos.
Per
cent.
Kilos.
Per
cent.
Kilos.
Per
cent.
Levanzin . .
Succi:
Paris. . . .
Milan . . .
Florence .
London. .
Naples . .
Rome . . .
Zurich.. .
Jacques. . . .
V.BeautS..
Schenk ....
Average .
7.49
9.90
7.70
8.45
7.10
8.20
6.90
7.95
7.34
7.83
7.50
12.4
15.7
12.6
13.3
12.7
12.9
10.6
11.1
11.8
11.9
13.3
8.38
10.30
8.45
9.00
7.70
8.35
7.30
8.70
7.03
13.8
16.3
13.8
14.2
13.8
13.1
11.2
12.1
11.3
9.71
11.40
10.15
10.50
9.00
10.60
8.60
9.80
8.08
16.0
18.1
16.6
16.6
16.1
16.7
13.2
13.7
13.0
12.54
13.75
12.80
12.85
11.50
20.7
21.8
20.9
20.3
20.6
12.95
14.25
13.10
21.4
22.6
21.4
13.25
21.9
11.60
20.8
11.70
21.0
14.10
25.3
10.15
16.4
10.32
16.6
8.1
14.4
12.6
13.4
15.6
20.1
20 6
21.5
25.3
BODY-WEIGHT. 81
16 days the average percentage loss was 13.4 per cent, the lowest again
appearing with Succi in the Rome fast of 11.2 per cent and the highest
16.3 per cent in the Paris fast. For 20 days the average percentage
loss was 15.6 per cent, the lowest being with Jacques of 13 per cent,
and the highest 18.1 per cent with Succi in the Paris fast. For 29
days the average loss was 20.1 per cent, the lowest again being with
Jacques of 16.4 per cent and the highest 21.8 per cent in the Paris fast
of Succi. For 30 days the average value was 20.6 per cent, the lowest
being with Jacques, of 16.6 per cent, and the highest 22.6 per cent in the
Paris fast of Succi. For 31 days only two experiments were com-
parable, the percentage loss in both of these being about 21 per cent,
while in the 40-day experiment the percentage loss was 25.3 per cent.
Of special interest is the fact that, aside from the fast with Jacques, in
which the weights of the drinking water and urine were perhaps less
trustworthy than in the other fasts, it can be said that at the end of 30
days of fasting, 21.5 per cent of the initial body- weight was lost. This
is strikingly regular in the fasts of both Succi and Levanzin. The
maximum loss of weight recorded in any controlled fasting experiment
with man was in the London fast with Succi, when at the end of 40
days a loss of 25.3 per cent of the initial body-weight was shown.
In contrast with these values found with man are the losses found
with animals, when the degree of emaciation has been carried to an
extreme and, indeed, in some instances to the point of death. In
Hawk's first fasting experiment, in which the dog fasted 117 days,
62.9 per cent of the initial body-weight was lost. The dog recovered,
was fed, and later underwent a second fast of 104 days, in which he
lost 52.5 per cent of the initial body-weight and then suddenly died.
Hawk's dog lived in the laboratory and was given a definite amount of
water, but Awrorow's dogs received neither food nor water, and the
fasting was carried to the point of death. With dog No. 2 the fast
lasted 44 days, with a loss of 55 per cent of the initial body-weight.
With dog No. 3 the fast continued 60 days, with a loss of 61.6 per cent,
while with dog No. 4 the fast was 66 days in length, with a loss of 62.0
per cent of the initial body- weight. Still other values were obtained
by Luciani with the two dogs which fasted under special experimental
conditions. As shown in his curve,1 the dog P lost 43.5 per cent of his
initial body-weight in a 43-day fast, while the dog P' lost 45.5 per cent
in a 34-day fast.
Incidentally it should be mentioned that Gayer, in his 30-day fast in
New York in 1912, was said to have lost 17.4 per cent of his initial
weight of 210 pounds (95.3 kilograms), while Penny, in his self-con-
trolled fast of 30 days, lost 19 per cent of his initial body- weight of
137.5 pounds (62.4 kilograms). While both these values are somewhat
luciani, Das Hungern, Hamburg and Leipsic, 1890, plate n. See also this publication, fig.
3, p. 77.
82 A STUDY OF PROLONGED FASTING.
less than the average loss found with L. and Succi, they are sufficiently
close to imply that in all probability no measurable amounts of food
were taken during these two uncontrolled fasts.
ANALYSIS OF LOSSES IN BODY-WEIGHT.
An analysis of the factors influencing the body-weight shows that
there may be a retention of water in the body due to the drinking of
more water than is excreted in the urine; a loss due to feces; a regular
loss due to the oxidation of organized material, the carbon burning
to carbon dioxide and the hydrogen to water; and a further loss of solids
in the urine. The amount of organized material oxidized in the body
will be influenced in large part by the muscular activity of the subject,
and if the activity is constant, the loss due to oxidation will progress in
a reasonably regular manner.
Considering the body as a living organism, therefore, we see that in
a fasting experiment the intake consists of drinking water and oxygen
from the air. The output consists of water-vapor and carbon dioxide
given off from the lungs and skin and the urine and feces excreted.
In this particular fast, however, the subject did not defecate during the
experiment.
The water vaporized from the lungs and the skin and given off in the
urine undoubtedly contains a large amount of preformed water which
was taken with the water drunk each day. It also contains water
which has been stored in the body and is given off as a result of the
breaking down of the protein, i. e., muscular tissue. There is likewise
a small amount of water due to the combustion of the organic hydrogen
of the body with the oxygen taken from the air. Without estimations
of the carbon-dioxide excretion, there are at present no known means of
satisfactorily computing these separate factors in the measurement of
the water output. When it is possible to have the subject live the
entire time inside the respiration chamber, as was done in the experi-
ments at Wesleyan University,1 the complete income and outgo may
be determined, including the income of oxygen and water and the out-
put of carbon dioxide, water- vapor, water in urine, solids in urine, and
an analysis of the solids. An approximate apportionment may then
be made of the water leaving the body as oxidized organic hydrogen
and as preformed water in the body. This has already been done
for the 7-day experiment reported in the earlier publication.2 From
the computed amounts of carbon dioxide excreted and the probable
organic hydrogen oxidized, a similar apportionment of the water loss
has been made for this experiment (see page 407 of this report).
Inasmuch as the body consists in large part of water — some 60 per
cent or more — it will be seen that there may be an addition to or loss
Benedict, Carnegie Inst. Wash. Pub. No. 77, 1907.
"Benedict, Carnegie Inst. Wash. Pub. No. 77, 1907, p. 469.
BODY-WEIGHT. 83
from the storage of water in the body, as, for instance, 200 grams,
without materially affecting the total percentage of water. It is easy
to see, therefore, that the changes in weight noted from day to day with
a fasting subject have only an indirect and passing influence.
INSENSIBLE PERSPIRATION.
In the long fasting experiment with L., the subject was not kept
inside the respiration chamber for the entire time of the fast, so that the
complete output of water-vapor was not determined. On the other
hand, a study of the so-called "insensible perspiration, " which has been
of great interest ever since the days of Sanctorius, shows some facts of
value.
The body loses in weight regularly as a result of the elimination of
carbon dioxide and water-vapor. It loses weight spasmodically by
the passing of urine and it gains in weight spasmodically by the drink-
ing of water. By correcting for the amount of water taken and the
weight of urine passed, the degree of insensible loss, or the "insensible
perspiration," may be accurately calculated. This has been done in
table 4, which gives for each day of the fast the loss of body-weight
in grams, the weight of the urine passed, the weight of the drinking
water taken, and the insensible perspiration. The excretion of urine
was always less than the amount of the drinking water with one excep-
tion, that of April 29-30. The insensible perspiration is therefore
readily obtained by finding the difference between the amount of water
taken and the weight of urine excreted and adding it to the observed
loss in body-weight.
A fact of special interest in this connection is that while the losses
in body- weight fluctuate considerably, the losses as shown by the insen-
sible perspiration are reasonably regular, the lowest being 371 grams
on May S-A; after the first 10 days, the highest value was 691 grams on
April 25-26. Theoretically this insensible perspiration should give us
a reasonable clue to the progress of the fast and should be an index of
the loss of water and carbon dioxide. On the other hand, while the loss
of preformed water is a real quantitative loss, the carbon dioxide and
water of oxidation are not, as they are in large part made up of oxygen
which is taken from the air.
In the later part of the fast it will be seen, from table 4, that this man
had on the average an insensible perspiration of not far from 20 grams
per hour. The insensible perspiration of the subjects in the fasts
described in the earlier publication, particularly S. A. B., was inad-
vertently not reported. Subsequently, Benedict and Carpenter,1 in
discussing the metabolism of healthy men, computed the insensible
perspiration of all of the fasting subjects. From their figures it will
be seen that the fasting subject S. A. B., who spent 24 hours of each day
1Benedict and Carpenter, Carnegie Inst. Wash. Pub. No. 126, 1910, p. 114.
84
A STUDY OF PROLONGED FASTING.
inside the respiration chamber, had an insensible perspiration of not
far from 25 to 27 grams per hour. When it is considered that this
represents the first 5 to 7 days of fasting, it will be seen that the results
obtained for L. are quite in accordance with those secured with the
earlier subject, in that they indicate a distinct tendency for the insen-
sible perspiration to decrease as the fast progressed. It should be
Table 4. — Insensible perspiration during fasting experiment with subject L.
Date.
Day of
fast.
Loss of
body-weight.
A
Urine.
B
Drinking
water.
C
Insensible perspiration.
Per 24 hours
a+(c-b).
D
Per hour.
E
1912.
Apr. 14-15
15-16
16-17
17-18
18-19
19-20
20-21
21-22
22-23
23-24
24-25
25-26
26-27
27-28
28-29
29-30,..
Apr. 30-May 1 . .
May 1-2
2-3
3-1
4-5
5-6
6-7
7-8
8-9
9-10
10-11
11-12
12-13
13-14
14-15
1st
2d ,,
3d
4th
6th. .. .
6th. . . .
7th. . . .
8th....
9th
10th....
11th....
12th....
13th....
14th....
15th....
16th....
17th....
18th....
19th....
20th
21st . . . .
22d,
23d.
24th
25th....
26th
27th
28th
29th
30th....
31st
grams.
1040
920
890
760
660
480
390
420
450
500
250
320
110
300
310
580
470
290
390
180
440
360
170
340
290
310
320
240
360
410
300
grams.
674
482
681
731
683
624
537
601
622
678
677
529
674
660
768
902
861
669
740
709
717
795
566
760
722
737
663
663
706
780
675
grams.
720
750
750
750
750
750
750
750
750
760
900
900
900
900
900
900
900
900
900
900
900
900
900
900
900
900
900
900
900
900
900
grams.
1086
1188
1059
779
727
606
603
569
578
672
573
691
436
540
442
578
509
521
550
371
623
465
504
480
468
473
657
477
554
530
625
grams.
45
60
44
32
30
25
25
24
24
28
24
29
18
23
18
24
21
22
23
15
26
19
21
20
19
20
23
20
23
22
26
borne in mind, however, in making any comparisons between the results
obtained in these two fasting experiments, that while the subject
S. A. B. remained in a respiration chamber the whole period of the fast
and consequently had an approximately constant temperature envir-
onment and minimum muscular activity, the subject L. was not in a
respiration chamber, but was for certain days partially naked for some
time while being photographed, measured, or clinically examined,
occasionally had a bath, and several times went out for a carriage
BODY-WEIGHT. 85
drive. His temperature environment and muscular activity were
therefore more variable than those of the subject of the earlier fasting
experiment.
An exact explanation of the variations in the insensible perspiration
from day to day is difficult, particularly for those of May 3^1 and 4-5,
when the lowest value of the fast was found on the day before a very
much higher value was found. It is always possible that there may
have been an error in the weighing, but on the other hand these weigh-
ings were very carefully made and recorded. Furthermore, an attempt
to explain the variations on account of a difference in the activity is
somewhat difficult, since the subject had a drive on May 3-4 and a bath
and drive on May 4-5, and on other days when he was given a carriage
ride the insensible perspiration was much greater than on May 3-4.
It is obvious, therefore, that these figures should not be considered
individually, but only as a general picture, this showing that the insen-
sible perspiration had a tendency to decrease as the fast progressed.
The increased value for May 14-15 may without doubt be explained
by the fact that this being the last day of the fast, there was much
greater excitement and muscular activity on the part of the subject.
On this day he talked vigorously to a group of medical men for some
40 or 50 minutes; on this day, also, he was nude for a time while photo-
graphs were being taken and during a series of physical measurements.
Considering the values generally, however, it will be seen that the insen-
sible perspiration is a far more scientific basis for estimating the loss of
body-substance during a fast than is the mere record of body-weight,
which considers neither fluctuations in drinking water nor the volume
of urine passed.
DRINKING WATER.
The intake of a fasting man is confined to water and oxygen of the
air. Of these the water may be readily measured. Such measure-
ments are of great importance in intelligently interpreting the losses
in weight from day to day. Accordingly special care was taken to
insure accurate records of the water consumed.
The selection of the kind and amount of drinking water for use in a
long fast is by no means simple. On the one hand, there is the belief
that distilled water is dangerous in that it washes out the salts from the
body, while on the other there is the fact that in many fasts the subjects
took either ordinary tap water or, as in Succi's fasts, various alkaline
or spring waters containing large amounts of salts, sometimes of a
distinctly purgative character. It has been believed by some that
these mineral waters have an actual nutritive value due to the salts
contained in them, if not to the organic matters. Furthermore, the sup-
position is reasonable that the salts interfere seriously with the mineral
metabolism, and it is obviously impossible in a metabolism experiment
to make an intelligent study of the output of sulphur, phosphorus, or
86 A STUDY OF PROLONGED FASTING.
chlorine in the urine if at the same time the subject is taking a large
amount of water containing sulphates, phosphates, or chlorides.
In the fasting experiments at Wesleyan University, one of the
subjects, S. A. B., preferred distilled water.1 Similarly Penny2 records
that he used only distilled water during his fast. Our subject, L.,
himself suggested that he be given distilled water during the fast, as
otherwise it might be said that tap water was either not pure or con-
tained mineral or organic matters which would contribute to his sus-
tenance. Although an experiment in which the man used distilled
water only was somewhat unusual, the desirability of being able to
study the mineral metabolism without the conflicting factor of the
ingestion of salts was, of course, apparent and arrangements were
therefore made for supplying L. with distilled water throughout the fast.
Dr. E. P. Cathcart was at this time a Research Associate of the
Nutrition Laboratory and advised that L. be given a constant amount
of drinking water each day, since in his observations on Beauts he had
experienced considerable difficulty with the volumes of the urine.
Our subject was first given 1,000 c.c. of distilled water, but was able
to take but 720 c.c. on the first day. L. then suggested that he be
given only 750 c.c. This amount was continued for a number of days,
when it was increased to 900 c.c, at which volume it continued through-
out the remaining days of the fast.
The amount of water taken each day by the subject is given in
table 4. Since the body excretes so large an amount of water, it is
perhaps somewhat unfortunate that the volume taken by the subject
was not constant for the whole period of the fast, although it is much
more nearly uniform than in any long fast heretofore reported. In any
discussion of the body-weight or the volume of urine, it is obviously
necessary to consider these fluctuations in the intake of water.
The subject was very inconsistent in his comments regarding the
water. On some days he said it was very good, but on other days
considered it to be very bad, although exactly the same amount was
given him and from the same glass carboy. On some days, also, he
found the amount given him was not enough and again not infrequently
complained that he was given too much water. He recognized the
importance, however, of maintaining the volume of urine so that a large
number of analyses could be made.
The daily allotment of distilled water was supplied to the subject in
a bottle and from this he poured out the amount desired. Early in the
fast he found that it was desirable to drink as large a portion of water
as possible during the first part of the day, so that it would not be neces-
sary for him to urinate during the night. The records show that he rarely
urinated during the night. Furthermore, as in the later part of the
Benedict, Carnegie Inst. Wash. Pub. No. 77, 1907, pp. 136 and 140.
'Penny, British Med. Journal, 1909, p. 1414.
BODY-WEIGHT. 87
fast he divided the urine into day and night periods, it provided a
particularly satisfactory method for studying the constituents of the
urine separately for these periods. Usually the last of the water was
taken a short time before the subject entered the bed calorimeter for
the night experiment.
At the end of the first 10 days of the fast, during which L. had taken
but 750 c.c. of water daily, the attending physician, Dr. H. W. Goodall,
expressed the opinion that there was a distinct physiological need of
water in the body. The lips of the subject were parched, his skin was
dry, and dandruff appeared. At Dr. Goodall's suggestion, L. was
prevailed upon to increase the amount of drinking water to 900 c.c.
daily. Two days later he reported to Dr. Goodall that for the first
time he felt thirst. Unfortunately some of the statements of the
subject were so inconsistent at this time that it is difficult to say
whether or not there was a physiological need for water which was not
felt by the subject but which was observed by the physician.
Aside from the objective indications noted by Dr. Goodall, the need
for water in the body may be inferred, though not scientifically proved,
by the figures given in table 4, the difference between the water taken
and the urine excreted being considerably increased when the water
intake was changed from 750 c.c. to 900 c.c. The subject had pre-
viously excreted in the urine about 600 c.c. of water daily, but when
the intake of water was increased, the amount given off in the urine was
actually decreased for several days, so that an average of over 340 c.c.
of water was retained per day for three days. This increase in the
difference between the water taken and urine excreted would imply a
distinct physiological need, since in the earlier experiments at Wesleyan
University, in which the subjects fasted for a shorter period and the
intake of water fluctuated widely, the variations in the amount of the
urine followed very closely the variations in the amount of water
ingested by the subject.
Finally, it is significant that at no time was there any indication of a
toxic effect in using distilled water, and we are able to sustain the con-
tention of Winckler1 that distilled water is without deleterious effect.
^inckler, Zeitschr. f. diat. u. physikal. Therapie, 1905, 8, p. 567.
BODY-TEMPERATURE.
The profound alterations in metabolism in the body of a fasting man
would lead one to expect some disturbance between thermogenesis and
thermolysis. Body-temperature, which is the index of the resultant
of these two factors, may obviously be affected by the disturbance of
either. If there is a decrease in thermogenesis with no change in the
thermolysis, there will be a fall in body-temperature. Conversely, if
there is an increase in thermolysis with constancy of thermogenesis,
there will again be a fall in temperature.
In this laboratory body-temperature measurements have a dual
significance : first, the value per se of the actual fluctuation, which indi-
cates a disturbance in the relationship between thermolysis and thermo-
genesis, and second, the importance of knowing body-temperature
changes for the accurate computation of the heat production in the
body. To determine the heat production it is not sufficient simply
to measure the heat radiated from the body and to add to this value
the heat of vaporization of water, for if in a given experimental period
the body-temperature has decreased, there has been a loss of heat from
the body unaccompanied by a production; hence the heat production
is measured only after correcting for the body-temperature changes.
In the series of body-temperature measurements in the short fasts at
Wesleyan University, the average body-temperature did not alter
noticeably, although there was distinct evidence of a flattening out of
the curve showing the daily rhythm. In the prolonged fasting experi-
ment with our subject L., we attempted to measure with the greatest
degree of refinement all the factors. It seemed important, therefore,
that frequent and careful records of the body-temperature should be
made in connection with this fasting observation.
The subject of body-temperature has been given special attention
in this laboratory for a considerable period, and a year previous to this
fasting experiment an extensive study of the fluctuations of the tem-
perature in the various parts of the human body was reported.1 As a
result of this research, it became evident that the only suitable place
for measuring body-temperature is deep in the body trunk, preferably
in the rectum. The identical apparatus used in the study referred to
was available for this fasting experiment and consequently body-tem-
perature measurements were secured as frequently as possible. A
detailed description of this apparatus and the tests made with it were
published in the report cited. Briefly, the apparatus consists of a
thermal element which is inserted about 7 cm. in the rectum, this
thermal element being connected with another thermal junction in a
constant-temperature bath. By means of this apparatus, it is possible
Benedict and Slack, Carnegie Inst. Wash. Pub. No. 155, 1911.
88
BODY-TEMPERATURE. 89
to determine the rectal temperature of a subject within 0.01°C, and
records can be made as frequently as desired.
Records of the rectal temperature were obtained nearly every night
while the subject was in the calorimeter chamber, the junction being
inserted in the rectum of the subject, connections made with the binding
posts inside the chamber, and observations taken on an average of
every 15 minutes throughout the night. On some nights records were
taken every 5 or 6 minutes. Observations were also made at various
times during the day and on at least two days continuous records were
secured for nearly the whole day-period. The apparatus was frequently
controlled by comparison with a standard thermometer, so we believe
that these observations represent the absolute temperature changes
of this individual.
As in most of the measurements taken during the fast, the subject
cooperated heartily in these body-temperature observations. After
the first night, and, in fact, after the thermometer had been inserted
a few moments, he experienced no particular difficulty and expressed
himself as being very much pleased that this routine gave him no dis-
comfort. It is clear that the use of the thermometer did not interfere
in the slightest with his sleeping. The observations were therefore made
under normal conditions, so far as it was possible to control them.
The body-temperature measurements made in this fasting experi-
ment may be considered in two ways: first, as to the alteration in the
regular rhythm of the temperature as the fast progressed, and second,
as to the effect of the fast upon the average of the temperature meas-
urements.
CHANGES IN TEMPERATURE RHYTHM.
In order to study the first of these problems, namely, the changes in
the temperature rhythm, curves have been plotted giving the tempera-
ture values for the period beginning about 8 p.m. and ending about 10
a. m. the following day. During this time the subject was in the
calorimeter chamber from 8 p.m. until about 8 a.m., and then, without
leaving the bed, he was withdrawn from the apparatus and was for the
next two hours the subject of the morning respiration experiments.
Accordingly, he was lying on the same bed in the same position for
the entire time, the only change being that in the last two hours he
was in the calorimeter room instead of inside the calorimeter chamber.
Since the temperature of the calorimeter laboratory was essentially that
of the respiration chamber, there was practically no alteration in the
temperature environment throughout the whole period covered by the
observations shown by the curve.
It is, furthermore, of value to note that this period includes what is
normally found to be the maximum diurnal change, for with normal
individuals it has been shown that the lowest temperatures are found
about 3 a. m. and the highest about 5 p. m. From 5 p. m. until
90
A STUDY OF PROLONGED FASTING.
about 11 p. m., or until bedtime, the temperature usually remains
approximately constant. The temperature, as a rule, falls rather
rapidly after one goes to bed, reaching the minimum about 2 or 3
a. m. With the fasting subject, the maximum temperature undoubt-
edly was reached prior to his entering the chamber at 8 p. m., as he
usually lay on the couch for an hour or more previous to that time.
The body-temperature was unquestionably falling continuously during
this preliminary period, so that the range for the night would be some-
what less than the actual daily range.
2:00 A.M. 4:00
37.iT:
369
367
363
363
361
376
37.4
37.2
37.0
368
366
364
362
360
37.1
369
367
369
363
361
v^
"©*'
— •.,
■v
\
\
APR. 15-
APR. 17-
18
\©
A
\
S '
"*\
/
<N
ft
/
.''A-
S
***
M-
\^
APR. 18
APR. 19
ZO
\®
.......
.. ^
N
%*%
X'
"•v^
j/.
** — *
\
<3L
,^\
>®
X
APR. 2<
APR.2
22
•%
.»*»..
•«./'
2^zs>
Fig. 4. — Body-temperature curves during the night and early morning
for the second and fourth to eighth days of fast.
BODY-TEMPERATURE .
91
The temperature curves for the period from 8 p. m. to about 10 a. m.
for every day of the experiment, with but three exceptions, are given
in figures 4 to 8. In order to save space, it has been necessary to plot
the curves in pairs, but the observations are of such interest that it
appears unwise to plot them in larger groups. The day of the fast is
indicated by a number in a circle on each curve. It can be seen that the
general trend of the curves remains essentially the same throughout the
entire fast. There is a noticeable fall in the evening, the minimum
being reached not far from 3 or 4 a. m. This is followed almost inva-
riably by a distinct rise, which continues until the end of the record.
e-.oo p.m.
37.0°C
*N,
^
A.
r
te«
APR. 2;
APR. 2:
r
j~
36.8
36.6
X
,.-''
36.4
\
..-"•-.
/
36.2
37.3
(0
5jr
Z\,
APR. Zt
APR. 2!
1-25
37.1
36.9
■^x
^S.
36.7
A
./\
'{""
.-— ''
36.5
rW
37.0
Lp>
/
v
\®
y\ APR. 2(
\ APR. 2"
!
36.8
36.6
•
V
36.4
*•*
^
/
36.2
\
— -. *
j — X
r*-*».
--''
36.0
\
-..--'
/
37.0
\
.,©
36.8
#
5.
^v
APR. 2
APR. 2
36.6
36.4
nN
.
36.2
s^
-rS^
/ s
^.♦s^
36.0
Fig. 5. — Body-temperature curves during the night and early morning
for the ninth to the sixteenth days of fast.
92
A STUDY OF PROLONGED FASTING.
OBSERVATIONS OF THE BODY-TEMPERATURE IN THE NIGHT PERIOD.
AVERAGE BODY-TEMPERATURE.
The greatest interest, at least to the clinician, lies not in the course
of the temperature curve throughout the night, but in the average tem-
perature values as the fast progresses. These are recorded in table
5 (page 95) for nearly every night of the fasting experiment, only the
time that the subject was inside the bed calorimeter being included in the
values. These averages were taken directly from plotted curves and,
except in the values for the tenth to the fourteenth nights of the fast.
37.0C
v£2>
368
0s-
\
,
MAY 1-2
366
364
\
1 '•
/
\
^/"
t
If.?
360
V"
******
.WO
>j,
M
MAY 2
MAY 3
_
;
'
368
Ififi
\©
\..
^
J
fS
VI 4
\
..
,
...-•'
1
\f
36?
\
*•%.--"*
»"*
360
Mil
369
N^
•
.1*7
@X-
. \
361
\
.
1
\/"
\
MAY 4
MAY 5
_
fj
363
361
V «
•'
319
V
A /
— >
357
\
\,
35.5
V"
Fig. 6. — Body-temperature curves during the night and early morning
for the seventeenth to twenty-second days of fast.
BODY-TEMPERATURE .
93
show a general tendency for the temperature to remain reasonably-
constant up to the seventeenth night of fasting. The temperature then
fluctuated, falling as low as 35.88° C. on the twenty-fourth night of
fasting and rising as high as 36.37° C. on the twenty-eighth night of the
fast. The maximum average value for the body-temperature observed
in any night during the fast was 36.85° C. on the twelfth night and the
minimum value was 35.88° C. on the twenty-fourth night. On the
last night of the fast, the average body-temperature was 36.14° C.
8:00RM.
10:00
12:00
2:00AM.
4:00
6:00
8:00
10:00
36.5°C
**„
**-.-''
^
MAY 6
MAY 7
-7
36.3
36.1
^
\
\/,
35.9
V
-■
'"/
35.7
•• \f
Vw
— "s/"
*w
<s
37.2
37 0
/*
\©
36.8
MAY 8
MAY 10
n
/
.16 6
36 4
-"\.'\
W
\
36.2
©N
Vn
36.0
"•—.
-Vf
\
t
'
35.8
»-^*-
V.
,-*"'
V
368
\J
""""'
\ f
\p>
MAY II
MAY 12
/
yf
36.6
36.4
@^
/ V
.(
y4
t
1
36.2
^
■
36.0
\..,
../
"V/
"**\
" V
r
35.8
wv
t
Fio. 7. — Body-temperature curves during the night and early morning for twenty-third to
twenty-ninth days of fast.
94
A STUDY OF PROLONGED FASTING.
With the resumption of food the average temperature increased on
May 16-17 to 36.79° C. and on May 17-18 to 37.53° C.
RANGE IN BODY-TEMPERATURE.
The maximum and minimum temperatures and the range in the
temperature for each night are likewise recorded in table 5. The maxi-
mum temperature observed on any fasting night was 37.45° C. on the
fifth night; the minimum temperature was 35.61° C. on the twenty-
second and twenty-third nights. The difference between the minimum
and maximum values, or the range in temperature, is also recorded in
table 5. The average range was not far from 0.90° C. The maximum
range, 1.27° C, was observed on the fifth night of fasting; the minimum
36.7 C
36.5
36.3
36.1
35.9
35.7
37.1
36.9
36.7
96.5
37.9
377
37.5
37.3
V
— *"""*>
N@,
MAY 13
MAY 14
\
*^^w. *'
*\
-\
/'
••
— \
i
Li -V
^jS'
• **
V
v
^o .
MAY 16-17
V
%
*\
MAY 17-18
f
>"*N
A
fS
/-
V*-»i
/
/
V
V
^
"^\
U^
Fig. 8. — Body-temperature curves during the night and early morning for thirtieth and thirty-first
days of fast and second and third days with food.
BODY-TEMPERATURE .
95
range was 0.50° C. on the twelfth night of the fast. It is important to
note, however, that the variations in the range have not even the
semblance of regularity.
OBSERVATIONS OF THE BODY-TEMPERATURE IN THE DAY PERIOD.
On two days (May 7-8 and May 8-9) the body-temperature was
measured almost continuously throughout the entire 24 hours. Curves
showing the fluctuations in body-temperature on these days are given
Table 5. — Body-temperature (rectal) of subject L. during experiments in the bed
calorimeter at night.
Date.
Day of
fast.1
Max.
Min.
Range.
Average.
Observation
at 7 a.m.
1912.
°C.
°C.
°C.
°C.
°C.
Apr. 15-16
2d
36.92
35.97
0.95
36.41
36.44
17-18
4th
37.23
36.01
1.22
36.55
36.36
18-19
5th
37.45
36.18
1.27
36.58
36.50
19-20
6th
36.85
36.14
.71
36.44
36.42
20-21
7th. . . .
36.74
36.18
.56
36.42
36.28
21-22
8th
37.17
36.25
.92
36.55
36.53
22-23
9th
37.12
36.11
1.01
36.50
36.41
23-24
10th
36.96
36.39
.57
36.64
36.66
24-25
11th....
37.28
36.52
.76
36.80
36.54
25-26
12th
37.19
36.69
.50
36.85
36.81
26-27
13th
37.04
36.13
.91
36.62
36.63
27-28
14th
36.60
35.99
.61
36.30
36.28
28-29
15th
36.95
36.11
.84
36.43
36.54
29-30
16th
36.92
36.16
.76
36.40
36.31
Apr. 30-May 1...
17th....
37.05
36.04
1.01
36.42
36.45
May 1-2
18th....
36.95
35.97
.98
36.30
36.31
2-3
19th
36.86
35.82
1.04
36.21
36.40
3-4
20th
36.88
36.19
.69
36.51
36.61
4-5
21st
36.96
35.70
1.26
36.12
36.04
5-6
22d
36.67
35.61
1.06
36.10
36.31
6-7
23d
36.48
35.61
.87
35.98
35.84
7-8
24th
36.38
35.68
.70
35.88
35.78
8-9
25th
37.00
35.99
1.01
36.31
36.36
10-11
27th
36.27
35.76
.51
36.03
35.98
11-12
28th
36.69
36.04
.65
36.37
36.10
12-13
29th
36.75
35.92
.83
36.23
35.92
13-14
30th
36.57
35.79
.78
36.06
35.94
14-15
31st
36.74
35.74
1.00
36.14
35.96
16-17
37.15
37.78
36.50
37.40
.65
.38
36.79
37.53
36.58
•37.68
17-18*
'For the duration of the period during which these observations were made, see table 44.
2The maximum temperature on this night was observed near the end of the calorimeter period.
•This observation was obtained during the morning respiration experiment, the subject lying
on the couch after leaving the calorimeter.
in figure 9. These curves, which were obtained on the twenty-fourth
and twenty-fifth days of the fasting experiment, show that even late
in the fast there was a very large diurnal variation.
It has been stated that the range in temperature during the night
was not a criterion of the probable total range throughout the 24-hour
96
A STUDY OF PROLONGED FASTING.
day, for before the night observation began the subject had been in a
condition of rest for one or more hours. On these two days the maxi-
mum temperature observations occurred during the daytime, as, for
instance, on May 7-8, when the maximum temperature observed was
37.10° C. at about 5 p. m., the normal hour of the day. The minimum
record was 35.68° C. at 4 a. m., the entire range being 1.42° C, a value
exceeding any range given in table 5. Similarly on May 8-9, the high-
est temperature recorded was in the daytime at about 12h 15m p. m.,
when a temperature of 37.51° C. was recorded. The minimum value
was 35.99° C, at 2h 40m a. m. Thus the range was 1.52° C, exceeding
even that of the preceding day. These curves give a general picture
of what was probably the average daily course of the body-temperature
BOOPM
10 00
K
00
2:00 AM.
4:00
6.00
8:00
I0O0
12:00
2:00 P.M.
4:00
e.<
T7-4.V.
, /
/•V-^
37?
f
y
r
V^\/'
/
'X
370
**
\©
I
\
MAV 7
MAV 8
_
J
?
*\
36.8
36.6
^
/
\
36.4
\ ^-.
S
36.2
@\
\r-.
/
360
• l
sy-'
1
~/
r
35.8
/
356
J
Flo. 9. — Body-temperature curves for approximately 24 hours on twenty-fourth and twenty-fifth days of fast.
of this subject throughout the 31 days of fasting. Here, again, there is a
distinct tendency for the maximum temperature to appear in the late
afternoon and the minimum temperature in the early morning, this not
being affected by many days of fasting.
On three other days, body-temperature records were obtained for a
part of the morning. On May 1 and 4, observations were made during
respiration experiments in which L. was sitting up and writing. (See
plate 1, fig. B.) These values are given in figure 1 0. This experiment
followed the regular morning respiration experiment, in which the
subject lay upon a couch; the observations recorded for the lying
position are also given in figure 10 for comparison. Of particular
interest is the fact that the change in position from lying to sitting
did not greatly alter the rate of the morning rise in temperature.
On May 15, when the subject first began to eat, the observations
commenced shortly after the end of the regular respiration experiment
BODY-TEMPERATURE.
97
and continued until noon. During this time the subject was sitting up
and eating. The curve given in figure 11 for both the respiration
experiment and the eating period shows that when the subject was
sitting and eating the ascent is somewhat more noticeable than in the
lying period, but it is evident that even eating after a 31-day fast did
not materially disturb the course of the rectal temperature curve.
CONSTANCY IN BODY-TEMPERATURE AT A GIVEN HOUR.
Since at 7 a. m. the subject had been living under constant conditions
of quiet and rest inside the chamber for 8 or 10 hours, a comparison
may be made of the values for the body-temperature obtained at this
time from day to day. This comparison is the more important since
373oc8:ooa.m
37.1s
9:00
10:00
36.9"
3G.7
36.5
MAY 1
MAY 4
J>,
•
1
/
•
/ /
/ /
/ /
f
Fig. 10. — Body-temperature curves showing change from lying to
sitting position.
37.(f c 8:00 A.M.
36.8*
36.6°
36.2°
1200
Fig. 11. — Body-temperature curve showing change from lying to
sitting position.
98 A STUDY OF PROLONGED FASTING.
in many fasts the body-temperature measurements are made but once
each day and usually at a given hour. Accordingly, in table 5 the
temperature records obtained at or near 7 a. m. have been given for
each day. In general, the variations in the temperature at 7 a. m. are
not markedly different from the variations in the average temperature
throughout the night, since the maximum and minimum records for
this time were found on the same days as the maximum and minimum
average temperatures. The maximum value found at 7 a. m. was
36.81° C, at the end of the twelfth night of fasting, and the minimum
value 35.78° C, at the end of the twenty-fourth night of fasting. No
uniformity in the values is apparent.
With the fluctuations in the body-temperature varying as they do
it will be seen that the difficulties in securing an average temperature
throughout the fasting period by means of one or two observations
during the day have been overlooked. Only by securing average values
throughout the entire night or throughout several hours at approxi-
mately the same time each day can a true picture of the average temper-
ature change of the body as affected by inanition be secured.
The well-known influence of muscular activity on body-temperature
makes it the more regrettable that certain experiments with muscular
activity could not have been carried out with this subject, as the effect
of a moderate amount of muscular exercise upon the temperature regu-
lation as the fast progressed would have great theoretical interest.
This is one of the problems that should certainly be studied in any
subsequent fasting experiment.
The observations of body-temperature on other fasting individuals
have frequently been made without reference to the preceding muscular
activity or the general condition of the sub j ect . Obviously those made in
the morning, just before the subject rises, have by far the greatest value.
It is a characteristic of practically all the fasts heretofore reported — in
which the temperature observations have been made for the most part
in the axilla or in the mouth (both localities unsuited for physiological
experiments) — that there has not been sufficient disturbance in the
temperature regulation to be recorded by this method of thermometry.
PULSE-RATE.
In practically all of the fasting experiments with which we are
familiar, the method of taking the pulse-rate from the radial artery has
been used. In the fasting experiments made at Wesleyan University,
in which the subject remained in the calorimeter during the whole
period, it was at first necessary to rely upon the subject's own obser-
vations of the radial pulse. This method was by no means ideal and,
in a later series of 2-day fasting experiments with seven individuals,
the method was improved upon, in that a small tube-pneumograph was
placed about the chest. The pulse-beats were thus superimposed upon
the respiration movements of the tambour and could be counted by an
observer outside of the chamber.
The striking relationship between pulse-rate and metabolism, which
has been regularly noted in this laboratory for many years, not only
with men but with animals and more recently with infants, led us to be
especially interested in the pulse-rate of our fasting subject. For a
study of the pulse-rate during the fasting experiments, it was necessary
to select a method by which continuous records could be made, as the
pulse-rate gives a reasonably accurate index of the metabolism at the
time the pulse record is made. The method of recording the pulse-rate
from the radial artery, either by an observer or by the subject himself,
has distinct disadvantages in that the knowledge that the observation
is being made has a psychical influence which is undesirable. Con-
tinuous records, therefore, could not be obtained by this method.
Furthermore, while the pneumograph method may properly be used
in a short experiment, its use in a long-continued experiment is objec-
tionable. The wearing of the pneumograph for a considerable period
of time may cause the subject much discomfort, as the traction becomes
wearisome, and if he changes his position during the experiment inside
the calorimeter the pneumograph may possibly press into the flesh and
be somewhat painful. The transmission tube may also become twisted
and thus interrupt the record.
It was hoped that this continuous record of the pulse-rate could be
obtained by photographic registration with the string galvanometer,
but although Professor W. B. Cannon kindly loaned us the string gal-
vanometer belonging to the Harvard Medical School, it was impossible
to install and test it suitably in time for its use in this experiment.
In our previous experimenting we had found it advantageous to fasten
the bell of a Bowles stethoscope over the apex beat of the heart and by
using long transmission tubes very satisfactory counts of the pulse-rate
were obtained. Accordingly, since records could not be made by
photographic registration, the stethoscope was used for nearly all of the
observations in this fasting experiment. The stethoscope is much less
99
100 A STUDY OF PROLONGED FASTING.
disturbing to the subject than feeling of the radial pulse, but a few
additional records were obtained by the latter method. In the later
days of the fast, when the apex beat of the heart became fainter, it was
occasionally necessary for the observer in the respiration experiments
to note the pulsations of the carotid artery.
The pulse-rate records may be classed in two groups. The first in-
cludes a large number of perfectly comparable observations : those made
throughout the night, while the subject was inside the bed calorimeter,
and those during the 1§ or 2 hours of the morning respiration experiment.
Usually the period of continuous observation extended from 8 p. m. to
9h 30m or 10 a. m. ; during this time the subject was lying quietly upon
a couch. These records were made regularly every day of the fast.
While the subject was in the bed calorimeter, the records were made
by the regular chemical assistant as often as possible, the frequency of
the observations obviously depending somewhat upon his other duties.
Occasionally when the subject moved inside the calorimeter, so as to
slightly displace the bell of the stethoscope, the pulse beats could not
be heard and there would consequently be a break in the records until
the subject again changed his position so as to bring the bell to its
former location. During the morning respiration experiment a special
observer was detailed to count the pulse-rate continuously throughout
the whole period. (See plate 2, figure C, page 19).
The second group of observations consists of those taken at various
times throughout the day, a part of which were continuous, while others
were individual records. This group includes the observations in the
miscellaneous respiration experiments, such as those made in the even-
ing before the subject entered the bed calorimeter, while the subject was
writing, or when he was breathing an oxygen-rich atmosphere. During
the latter part of the fast, the pulse-rate was also recorded twice when
the daily record of the blood pressure was taken, and occasionally when
other special tests were made. At times the subject wore the stetho-
scope throughout the whole day, so that the observations were more or
less continuous for the 24 hours. On the days when the continuous
observations were made, the subject was followed by an assistant who
kept out of sight but made the records regularly and also noted the
changes in body-position. These records were frequently verified by
a second observer.
RECORDS OF PULSE-RATE OBTAINED IN EARLIER FASTING EXPERIMENTS.
Before giving the records of the pulse-rate obtained in the fasting
experiment with L., it will be of interest to cite those secured in fasting
experiments made by other investigators. In discussing such obser-
vations, two essentially different comparisons can be made, first, the
influence of a prolonged fast upon the pulse-rate determined under any
given conditions, and second, the variations in pulse-rate incidental to
the changes in position or mental activity. Usually in fasting experi-
PULSE-RATE. 101
ments observers have contented themselves with taking the morning
pulse-rate and occasionally the evening pulse-rate. No particular em-
phasis has been placed upon these individual observations, aside from
the general fact that the pulse may have altered as the fast progressed.
Not recognizing the great significance of the pulse-rate in relation to the
metabolism, experimenters have not ordinarily taken especial precau-
tions (as did Luciani) to keep the suoject lying quietly while the pulse-
rate was being observed and, indeed, for some time previous to the
observation. This probably explains difficulties found in comparing
the records, in that some observers note a continually decreasing pulse-
rate during the fast, while others find marked irregularities. As would
be expected, the more recent observations take into account the factors
influencing the pulse-rate and the records are thus more trustworthy.
Of the pulse records obtained in Tanner's fast, we have been able to
find only those given in the British Medical Journal.1 On the thirty-
seventh day of this fast, the pulse, respiration, and temperature are
reported as having been "normal." On the twenty-fifth day the pulse-
rate is given as 75, the respiration as 15, and the temperature of the
mouth as 98.4° F. (36.89° C). On the thirtieth day the pulse-rate was
reported as 84 and slightly more regular, the temperature as 98.8° F.
(37.11° C), and the respiration as 14, with the general statement that
he was weaker than on any previous day. "On the twenty-ninth day,
two of the experts attending him reported that there was no material
alteration in the vascular pressure indicated by the heart's impulse,
while its volume was scarcely less than in health."
Paton and Stockman2 report that the pulse-rate of their subject
averaged between 50 and 60 and the respiration usually between 23 and
30, but no continuous records of the pulse-rate are given.
The most extensive series of continuous observations of the pulse-
rate of a fasting subject is that reported by Hoover and Sollmann.3
In this 5-day fast, the pulse was counted and recorded once every hour
by relays of watchers. The initial record of the pulse-rate was 75, the
lowest value of 37 being recorded on the last day, thus showing a dis-
tinct tendency for the pulse-rate to decrease as the fast progressed.
Unfortunately the fast continued for only 5 days and, in the opinion
of the authors, the pulse records are vitiated by the fact that they were
obtained with a hypnotic subject and that the pulse-rate was purposely
lowered by suggestion.
In reporting a fast carried out by Succi in New York in December
1890, and said to have continued for 45 days, a newspaper states4 that
on the last day of the fast Succi's pulse-rate was 62. Unfortunately no
scientific record of this fast was ever published.
British Med. Journ., 1880, 2, p. 171.
2Paton and Stockman, Proo. Roy. Soc, Edinburgh, 1888-1889, 16, p. 121.
'Hoover and Sollmann, Journ. Exp. Med., 1897, 2, p. 403.
«N. Y. Daily Tribune, December 21, 1890.
102 A STUDY OF PROLONGED FASTING.
In a fast carried out by Succi in London in 1890, which continued
40 days, the pulse-rates, taken every day at noon,1 varied from 82 on
the second day of the fast to 52 on the thirty-fifth day. The degree of
irregularity noted in all conditions of the fast, however, shows that
proper attention had not been paid to secure uniform quiet before the
observations were made. The respirations varied from 16 on the
thirty-first day of the fast to 28 on the sixth day of the fast. Here
again the irregularity noted on all days implies variations in muscular
activity prior to the observations, no general trend of the respiration
rate being apparent.
In the 10-day fasting experiment with Cetti,2 the pulse-rate ranged
from 68 on the morning of the fourth day to 92 in the afternoon of the
seventh day. The high pulse-rate was accompanied by abdominal
pains. In certain of the respiration experiments carried out with Cetti,
the pu]se-rate was likewise recorded. In one instance it was noted that
the pulse-rate changed from 88 while the subject was lying down to 120
while he was walking about the room. In another experiment the
pulse-rate changed from 86 while he was lying down to 98 when he was
sitting, smoking, and talking. The great increase in the heart action
of this subject was commented on at some length by these authors.
In an experiment with Breithaupt, continuing for 6 days, the same
authors record a minimum pulse-rate of 47 on the last day of the fast
and a maximum rate of 66 on the morning of the second day. Taking
advantage of the fact that their subject performed muscular work on
the ergostat, the authors made some interesting notes upon the increase
in the heart-beat with a definite amount of work and the return of the
pulse-rate to normal after the work ceased. In their general conclusions
they maintained that with Cetti, who was of an excitable temperament,
the pulse-rate in the resting condition was not noticeably changed by
fasting, but that it slowly decreased with Breithaupt, who was quiet
and phlegmatic. They also emphasize the fact that during the fast
there was a distinct tendency to a considerable increase in the irri-
tability of the heart, slight muscular activity producing a great increase
in the pulse-rate.
Luciani contends that, during his experiment with Succi, the pulse-
rate remained strictly inside the physiological limits, rising to 70 but
twice and only occasionally falling below 50. He also points out that
the pulse-rate, as well as the temperature and the respiration, were
always measured during complete muscular rest, as the subject was
lying in bed. An interesting observation on the irritability of the heart,
as indicated by the rise in the pulse-rate after exercise, was likewise
made by Luciani, who was fortunate in having a fasting subject who
freely indulged in muscular activity.
British Medical Journal, 1890, pp. 764, 819, 876, 935, 996, 1056, and 1444.
*Lehmann, Mueller, Munk, Senator, and Zuntz, Archiv f. path. Anat. u. Physiol, u. f. klin. Med.,
1893, 131, Supp., p. 1.
PULSE-EATE. 103
The pulse-rates were also recorded in a long fast made by Penny.1
This fast was less strictly controlled than the previous fasts cited,
but Penny states that his observations of the pulse-rate were verified by
another doctor. The morning observations were made about 9 o'clock
before he rose from his bed; the evening pulse-rate was taken about the
time of retiring, i. e., 10 or 11 o'clock. The records for the morning
ranged from 59 on the second day of the fast to 39 on the thirteenth,
fourteenth, fifteenth, and eighteenth days of the fast. The evening
records ranged from 80 on the last day to 44 on the eleventh, fifteenth,
and sixteenth days of the fast.
Far less confidence can be placed in the observations reported for
Gayer, who was said to have carried out a 30-day fast in New York in
1910. My only justification for calling attention to these observations
in the report of this fast is the personal assurance of Dr. Ira S. Wile,
of New York, who, while not vouching for the authenticity of the fast,
is inclined to believe that the records are for the most part trustworthy.
These show a pulse-rate ranging from 54 to 80, but, as the writer points
out, the maximum observation was taken after the subject had come in
from a 2-mile walk and on the very next day a pulse-rate of 61 was
noted when the subject spent the morning lying down.
Cathcart2 recorded both the morning and evening pulse-rates of his
subject, Beaute. Charteris3 also recorded the pulse-rates on this indi-
vidual, but obviously at a slightly different time of day, as his records
do not agree with those of Cathcart. Nevertheless both authors draw
the conclusion that there was a general tendency for the pulse-rate to
fall as the fast progressed. Charteris furthermore points out that the
subject was well aware of this fact from his previous experience, as he
was a professional faster. Cathcart's morning observations ranged from
70 on the seventh day of the fast to 58 on the twelfth and fourteenth
days of the fast. The highest observation secured in the evening was
71 on the second day, and the lowest was 57 on the tenth day of fast-
ing. The records obtained by Charteris show a range from 68 on the
second day of fasting to 58 on the twelfth day of the fast.
RECORDS OF PULSE-RATE OBTAINED IN THE EXPERIMENT WITH SUBJECT L.
The number of observations obtained with L. was sufficient to justify
their presentation in the form of 24-hour curves, as shown in figures 12
to 18. In these curves the day begins with 8 p. m., when the subject
entered the respiration calorimeter, ending 24 hours later. Continuous
records were secured for every night experiment and frequent records
were made during the day. The values are perfectly comparable for
each day between 8 p. m. and 10 a. m. and also for the most part
throughout the rest of the day, as the daily routine of the subject was
^enny, British Med. Journ., 1909, p. 1414.
"Cathcart, Biochem., Zeitschr., 1907, 6, p. 109.
JCharteris, Lancet, 1907, 173, p. 685.
104
A STUDY OF PROLONGED FASTING.
more or less regular. The pulse records which were obtained in the
evening respiration experiments may logically be attached to the
records for the bed-calorimeter experiments as preliminary periods, but
the fact that the evening experiments were made only in the latter
part of the fast complicated their presentation in this manner. As a
matter of fact, the record of the pulse-rate from the beginning of the
evening respiration experiment, i. e., about 7 p. m., until the close of
100
8:00 P.M. 10:00
12:00
2:00A.M. 4:00
6:00
8:00
10:00
Fig. 12. — Pulse-rate chart of subject L. for days preceding fast.
PULSE-RATE.
105
the respiration experiment the next morning at 9h 30m or 10 a. m., was
continuous, as the subject did not rise from his couch during that period.
He urinated lying on the side. Since the conditions of activity were
essentially uniform from the time the subject entered the bed calorimeter
at about 8 p. m. until the end of the morning respiration experiment
at 9h 30m or 10 a. m., the records of the pulse-rate taken during this
period on every day of the fast are more comparable. It is therefore
) P.M. 10:00 IfcQO 2:00A.M. 4:00 8:00 8:00 10:00 12:00 2:00 P.M. 4:00 6:00 8:00
H£-
A.x,\www
o:
^
k
4^\
fiL
V
\
1
V
®
Fio. 13. — Pulse-rate chart of subject L. for first to fifth days of fast.
106
A STUDY OF PROLONGED FASTING.
permissible to discuss these observations first, and later consider the
more or less heterogeneous observations taken during the day when the
activity might vary.
PULSE-RATE IN THE NIGHT PERIODS.
The relatively large fluctuations in the pulse-rate that are apparent
in the first two or three nights are naturally to be explained by the fact
that the subject was a stranger in America, and was experiencing for
if
JOPU. w
00 12:
X) M
0AM «■
» «(
10 S:
30 W:
>0 tt:
90 1:C
K)».M. 4.
30
6.
X) 8.001
T«i
J
\h
M
\®.
1
pa. i»-
M
flju
■
J*
H
V*
\J
to
PO
m
A
Va
.
j
y^~
^,
A
m jo-j
1
,l 1
V^
' V
f
y
-^ n
^X
1
ri
M
APR. 21
■22
L /^
ss
<sj^
\rS-J
v
7(1
h
fiO
*\
*v
A
PR. 22-
A 7\
23
N
»y
-
■
^*-
•XI
vy
V
AJW|
W)
nji
tl)
l
A
«0
/>
©
APR.
23-24
»
\k
i/^
.
1
1^
so
V
*"N
vAV
y^J
VVy
N
f\
H
(£V
APR. 2
♦ -25
^
4S
Fig. 14. — Pulse-rate chart of subject L. for sixth to
eleventh days of fast.
PULSE-RATE.
107
the first time the novel sensation of being inclosed in the respiration
chamber. On the night of April 14-15 (figure 13), we find reasonably-
constant pulse-rates until 4 a. m., when the observer's records show
that he woke up, then dozed for the rest of the night. Usually the
pulse-rate showed a tendency to fall prior to midnight, thereafter to
continue fairly low until it rose again in the morning, although the
period of minimum pulse-rate might continue for several hours. As
the fast progressed, there was a marked tendency for the amplitude of
^8<X)P.M.IO:00 12:00 2:00A.M. 4:00 6:00 8:00 10:00 12.00 2:00P.M. AM 6:00 OflO
60
\
"\
1
APR 2
5-26
/\l\
. l
ij/t
Ijl
A
SO
dr
K
Ja
fi\
jj
7
Ir-
\
u
^
H
^V
^v
APR.
26 -Zf
-1*
\ K
*/
r
#*
7,1
\_j-r°r
SJ
W*
65
/v
is
rv
^©
APR.
27-28
tA
/W/rV
-
r
45
1
"V
t*
y
4\s*
70
1
*0
•
/i
p
SO
d\
APR. 21
1-29
jJ
V
I
60
/v
SO
©\
1
kPR. 29-
30
_A--
K-J
W*
r
70
\
\
60
J
V
\
\ ,/•
10
PR. 30
l L
-MAV 1
A_
^J
V
/
fil
—rb
IS
S
-y$2>
«av i-
2
Ar
„
/
45
Fig. 15. — Pulse-rate chart of subject L. for twelfth to
eighteenth days of fast.
108
A STUDY OF PROLONGED FASTING.
the curve to become less and less — that is, the fluctuations from maxi-
mum to minimum throughout the night were less and the periods of
reasonably constant pulse values grew longer and longer.
On the night of May 14-15, the last night of the fast (figure 18),
special attention was given to the pulse-rate, the records being made
frequently throughout the whole night. Although the curve is in con-
Jjgjl 10:00 tt-OO MO*-* +OQ fcOO fcOO 10-00 IfcOO jjOMj, 4c00 6:00 8:00
0 — "I
1
\
■a
©V^
HAY 2-
3
N
v*
'1
»
«nL
1
IAY i-4
,v
i
\
■a
A@L
JU
-A.
/
^i
w*
«s
w\
*t
MAY
*-s
Hi
iH
X
J1
ta
V
MJ
u
A^
IH,
/
'V
I
«
Ji©
HAY 5-
«
./
/
1
/*
4S
r*
.-vA-*
TD
1
1
-A
Ml
&
Mj
lY 6-7
A 1
i
I
1
1
/
SO
\
W.
a
w^>
V
uw
l>
■K
M
-v\
n®
MA
r- 7-8
A /-l
}j
\/
**
«n
V-\
V^-v
^V^\
y^1
~JU
«n ~
^
1
HAY 8-
i
1
/
■I
\A.
\J
^
v\
JW
V*1
1
Fia. 16. — Pulse-rate chart of subject L. for nineteenth to twenty-fifth days of fast.
PULSE-KATE.
109
sequence irregular in shape, the general trend is not markedly different
from those for the preceding and following nights. Even on May 15-16,
the first night following the ingestion of food, although the subject was
in such distress that he did not go inside the chamber, but lay on a couch
outside, the frequent records of the pulse-rate did not show extraordi-
narily large fluctuations. On the night of May 16-17 relatively few
records of the pulse-rate were taken; and also on May 17-18, but on
this night we find a greatly increased amplitude. The general deduc-
tion is, therefore, that the amplitude of the fluctuations of the pulse-
rate during the night decreased regularly as the fast progressed, showing
a tendency upon the resumption of feeding to return to the variations
commonly experienced.
00P.M. 10:00 12:00 2:00A.M. 4:00 6:00 8:00 10:00 12:00 2:00P,M. 4:00 6:00 8:00
60
50
^N^AjA/^'
Fig. 17. — Pulse-rate chart of subject L. for twenty-sixth to thirtieth days of fast.
110
A STUDY OF PROLONGED FASTING.
PULSE-RATE IN THE DAY PERIODS.
From 10 a. m. until 7 p. m., the records are naturally much less com-
plete than the series obtained during the night; nevertheless on certain
days reasonably complete records of the pulse-rate were obtained
throughout the day.
8:00 P.M KMX) 12.00 200 AM 4:00 6:00 8:00 KMW IZ.00 2:00A.M. 4:00 6:00 8:00
Fiq. 18. — Pulse-rate chart of subject L. for thirty-first day of fast and three subsequent days
with food. The point when the subject took food on May 14-15 is indicated by a heavy
vertical line.
PULSE-RATE. Ill
On April 16-17 (figure 13), the records were made for nearly the whole
day, these values probably being fairly typical of records which would
have been obtained if the observations had been more complete on other
days. The minor fluctuations shown on April 16-17 are obviously due
to changes in the activity of the subject when talking or moving about.
The high values obtained about 5 p. m. are coincidental with the hand
dynamometer test, in which there was some muscular exertion by the
subject, the highest record at this time being 102. After the dyna-
mometer test was over, the pulse-rate immediately fell again to an
approximately normal level. Until 5h 30m p. m., the subject was lying
on a couch and from 6h 05m p. m. was asleep in his chair for half an
hour, the low level of the pulse-rate being apparent at this time. The
records shown by the last portion of the curve were obtained during
the evening respiration experiment. For the greater part of this day,
the pulse-rate was on the average not far from 10 beats per minute
above that which would ordinarily be found when the subject was lying
upon a couch, although during muscular exertion, and especially after
the dynamometer test, the pulse-rate at times tended to rise consider-
ably above this value. The increase as a result of the dynamometer
test may also be noted on April 18-19, April 19-20, and April 20-21.
Beginning with April 30-May 1, records were made each day not
far from 1 p. m., at the time of the blood-pressure test. These records
are of unusual interest, inasmuch as they indicate the values while
the subject was sitting and again immediately afterwards when he lay
down upon the couch. Thus, on April 30-May 1 (figure 15) the record
for the sitting position was 68 and that for the lying position 63. These
records, which appear with but few exceptions in the curves for the
latter part of the fast, are of special interest, as they show the change in
the pulse-rate due to change in position. This subject will be con-
sidered in a later section.
On May 13-14 (figure 17) a number of observations were made in the
afternoon, one at 2h 30m p. m., while the subject was talking in a lively
manner to an assembly of medical men. That the after-effect of this
stimulus continued for some time is shown by the curve which follows.
The fall in the pulse-rate, due to a change in position from sitting while
writing to lying down upon the couch, is likewise shown, as between
6h 20m p. m. and 7 p. m. the subject was sitting up and writing and just
afterward lay down upon the couch for a respiration experiment.
On the first day of realimentation, the curve (figure 18) shows very-
great fluctuations in the pulse-rate. These are in part due to the inges-
tion of food and in part to the pain and distress incidental to the colic
resulting from the taking of a large quantity of acid material into the
stomach and intestinal tract. As a matter of fact, the highest observa-
112 A STUDY OF PROLONGED FASTING.
tions on this day were obtained at a time when there was a reasonably
small amount of muscular activity. In a series of observations from
2h45mp.m. until 4 p.m., which were made while the subject was sitting
quietly eating an orange or drinking grape juice, a value of 112 was
found. Even an hour after eating, when the subject had colic and was
in much distress, the pulse-rate was considerably lower than during
the period of eating, while the average value obtained when the subject
was lying on the couch during a respiration experiment in the early
morning was about 59. Evidently the process of eating or drinking,
immediately following a prolonged period of inanition, increased the
pulse-rate very greatly. On May 15-16, the second day of this period,
the sharp rise in the pulse-rate incidental to taking food was likewise
noted at 9h 40m a. m. and again at llb 46m a. m. On May 17-18 the
values obtained from 6 a. m. to 9 a. m. were unusually high, this being
due in part to the fact that the subject was extremely excited and after
the experimental period was over broke out into abusive language.
There was undoubtedly a great increase in the psychic disturbance.
COMPARISON OF PULSE RECORDS OBTAINED IN EXPERIMENTS WITH THE
BED CALORIMETER AND THE RESPIRATION APPARATUS.
While an examination of the general trend of the pulse curves shows
admirably the tendency for the amplitude during the night to fall to
a lower level, a comparison of the average values obtained under varying
conditions can best be made in tabular form. Accordingly, in table 6
the average values are given for observations made when the subject
was lying in the bed calorimeter and also the average of the records
obtained when the pulse-rate had reached its lowest level during the
calorimeter period. The values for the experiments with the respira-
tion apparatus are likewise given, including those made in the morn-
ing, in the evening, when the subject was sitting quietly and also
when writing. Furthermore, for purposes of comparison the pulse-rate
records taken during the blood-pressure tests are included for both
positions of sitting and lying. A number of important comparisons can
thus be made.
During his stay in the bed calorimeter the subject was probably asleep
for the greater part of the time — at least on many nights. On every
night he had periods of wakefulness, which at times may have been of
considerable length. Consequently not all of the values obtained in
the bed-calorimeter experiments can be taken as actually obtained
during sleep, but by examining the curves for these experiments it is
relatively easy to select a value which probably represents the average
minimum pulse-rate for this subject during sleep. These values are
given in column b in table 6. Both the average night pulse-rate and
the average minimum pulse-rate have a distinct tendency to decrease, as
the fast progresses, until about the twenty-second fasting day. From
PULSE-RATE.
113
that time until the end of the fast the pulse records usually rise, so that
at the end of the observations the average values are 3 or 4 beats higher
than they were at the minimum point.
Table 6. — Average pulse-rate of subject L. at different times of the day and with, varying activity.
During blooc
-pressure tests
(about lh30mp.m.).
Bed calorimeter
(usually 10 p.m. to
Respira-
Increase
Decrease
Date.
Day of
fast.
8 a.m.).
tion ap-
with
subject
with
ing subject
paratus
Sitting.
T,y
(subject
awake
lying
Average.
(subject
asleep).
awake).1
(c-b).
(b— f).
A
B
C
D
E
F G
1912.
Apr. 10-11. .
82
76
78
70
68
76
70
73
64
64
272
273
272
273
74
-4
3
-1
9
10
11-12
12-13
13-14 . .
14-15
1st. .
15-16
2d...
66
63
73
10
16-17
3d...
62
60
70
10
17-18
4th..
65
58
68
10
18-19
5th..
63
59
67
8
19-20
6th..
60
57
64
7
20-21
7th..
59
56
64
8
21-22
8th..
61
58
65
7
22-23
9th..
59
67
63
6
23-24
10th.
57
55
63
8
. ,
24-25
11th..
67
54
61
7
25-26
12th..
58
56
61
5
26-27
13th..
66
64
59
5
27-28
14th..
53
51
58
7
28-29
15th..
53
51
67
6
29-30
16th..
53
62
58
6
i
H
Apr. 30-Mayl..
17th..
52
49
57
8
68
i
>3 5
May 1-2
18th..
52
51
56
5
2-3
19th..
52
50
67
7
69
<
>2 7
3-4
20th..
52
51
58
7
68
i
\2 6
4-5
21st . .
54
61
69
8
t
»6
5-6
22d. . .
63
61
59
8
67
t
►8 9
6-7
23d. . .
56
53
58
5
71
e
»3 8
7-8
24th..
55
53
59
6
76
c
»1 15
8-9
25th..
55
53
60
7
66
e
0 6
9-10
26th..
66
54
61
7
68
e
4 4
10-11
27th..
57
55
62
7
71
c
2 9
11-12
28th..
59
67
61
4
72
€
2 10
12-13
29th..
58
55
63
8
73
I
7 6
13-14
30th . .
58
55
59
4
69
€
3 6
14-15
31st. . .
57
54
60
6
383
37
3 10
15-16*
68
64
90
66
60
84
«72
'84
12
0
76?
99
7
8
3 3?
9 10
8
16-17
17-18
xThe respiration experiments in the morning were usually made between 8h 30m and 9b SO™.
2During the respiration experiments in the morning on April 11, 12, 13, and 14 the subject
was without breakfast.
'The subject had broken his fast by means of fruit juices during the morning.
4During the night of May 15-16 the subject lay on the couch in the calorimeter laboratory.
'During the morning respiration experiments on May 17 and 18 the subject was without
breakfast.
114
A STUDY OF PROLONGED FASTING.
Table 6. — Average pvhe-rate of subject L. at different times of the day and with varying
activity — Continued.
Date.
Day of
fast.
Respiration apparatus.
Sitting.1
Lying (usually 7 to
7h 45m p.m.).
Period.
Average.
H
Average.
I
Increase
over
lying
in the
morning
(i-c).
J
1912.
Apr. 15-16
16-17
17-18
18-19
19-20
20-21
21-22
22-23
23-24
24-25
25-26
26-27
27-28
28-29
29-30
Apr. 30-May 1 . .
May 1-2
2-3
3-4
4-5
5-6
6-7
7-8
8-9
9-10
10-11
11-12
12-13
2d
3d
4h 00m p.m.
to 4h35mp.m,
82
80
62
69
60
68
68?
69
65
69
75
4th
5th....
6th
4 10 p.m.
4 43 p.m.* . . .
7th
8th
9th....
10th....
11th
3 52 p.m.
3 58 p.m.
4 28 p.m, . ,
4 57 p.m
12th
13th....
14th
3 13 p.m.
12 14 p.m.
4 11 p.m
12 48 p.m
62
*59
59
61
59
61
62
60
57
63
60
63
63
66
66
66
67
1
0
1
4
1
4
6
3
-2
4
2
4
3
5
4
5
4
15th....
16th
3 23 p.m.
3 56 p.m.*. . .
17th....
18th
9 31 a.m.
10 04 a.m.*
19th
20th
21st. . . .
9 35 a.m.
10 10 a.m.*....
22d
23d.
24th
3 43 p.m.
4 14 p.m.*....
25th
26th
27th
28th
29th
13-14
30th
6 32 p.m.
7 02 p.m.*
71
12
1Periods indicated by an asterisk (*) were obtained with the subject sitting, writing.
The average pulse-rate for a period 3h 16™ p.m. to 3h 51m p.m. on this day with the subject
lying on the couch was 61 per minute.
Any important deductions from average values for the night are out
of the question on account of the irregularity in the number of the pulse
records during the night and the impossibility of recording accurately
the time when the subject slept and when he woke. If, however, we
compare the values for the average pulse-rate with those for the aver-
age minimum pulse-rate, we find that about the middle of the fast
the difference is only 1 or 2 beats. The greatest variation between
these two series of averages is on the fourth day of fasting, when the
average during the night was 65 and the average minimum value was
58. For further purposes of comparison, it is obviously more logical
to use the average minimum values.
PULSE-RATE. 115
The pulse-rate during the respiration experiments is recorded with
great frequency and regularity. Those obtained in the morning respi-
ration experiments during the first period with food, i. e., in the period
preceding the fasting period, were invariably lower than the records
obtained during the night experiments with the bed calorimeter, with
the single exception of the fourth record, when the average pulse-rate
for the night had a minimum of 64 and the average obtained in the
respiration experiment was 73. Unquestionably the pulse-rate obtained
during the night was influenced to a considerable extent by the food
taken the day before, especially for the evening meal. In this con-
nection it is of special interest to note that on the last night of this
period the food taken in the evening meal was of such a character that
the effect would be less prolonged. From this time on, the pulse-rate
in the experiment with the respiration apparatus was invariably higher
than the minimum pulse record obtained during the night. The differ-
ence in the early part of the fast was 10 beats, but then decreased until
about the middle of the fast, when it fell as low as 5 beats. In the
latter part of the fast it showed a tendency to increase again, although
on the twenty-eighth and thirtieth days the difference was but 4 beats.
This difference in the pulse records is of great significance, indicating
clearly an increased heart action, an increased muscle tonus, and,
according to all previous experience in this laboratory, an increased
metabolism. The question as to the influence of sleep on the pulse-rate
and the metabolism has received considerable attention in this labora-
tory for a number of years and we find ourselves quite out of harmony
with many European writers who maintain that sleep per se has no
influence upon the pulse-rate and the metabolism. A subsequent
examination of the records of the metabolism for this fasting subject
bears out definitely our contention that there is a great difference in the
pulse-rate and the metabolism in the waking over the sleeping condition.
INFLUENCE OF BODY POSITION.
In the latter part of the fast, the pulse-rate was recorded during the
blood-pressure tests for both the lying and sitting positions. While the
comparisons between the calorimeter experiments and the morning
respiration experiments considered only the values secured when the
subject was lying asleep or lying awake, these records obtained at noon
indicate the change in the pulse-rate due to the change in the position
from sitting to lying. The decrease in the pulse-rate is considerable,
varying from 4 beats on the twenty-sixth day to 15 beats on the twenty-
fourth day. A number of pulse records for the sitting position were
also obtained with the subject in two morning and nine afternoon
respiration experiments. In the morning experiments and also in four
of the afternoon experiments, the subject was writing. On the after-
noon of the second day of the fast, when the subject was sitting quietly,
116
A STUDY OF PROLONGED FASTING.
the pulse-rate was as high as 82; the lowest observation was 60, which
was recorded on the twelfth day of the fast.1
INFLUENCE OF THE WORK OF WRITING.
As the subject spent a considerable portion of the day sitting up
writing, an effort was made to study the pulse-rate during a period of
writing. Hence on a number of days the subject was connected with
the respiration apparatus, the metabolism was studied, and records of
the pulse-rate were taken. On two of the days these tests were carried
out in the morning, immediately after the regular series of respiration
experiments. On these two days — the seventeenth and twentieth days
of the fast — the combined influence of this position and occupation was
to increase the pulse-rate over the records obtained in the lying position,
this increase being 12 beats on the seventeenth day and 7 beats on the
twentieth day. Four experiments of this character were also made in
the afternoon, but the pulse-rate showed little increase over records
obtained in similar afternoon experiments when the subject sat quietly
without writing.
INFLUENCE OF BREATHING AN OXYGEN-RICH ATMOSPHERE.
In an earlier series of observations of the pulse-rate in which normal
subjects breathed an oxygen-rich atmosphere, there was a distinct
tendency for the pulse-rate to decrease. Since it was of interest to note
if this tendency to a decrease in the pulse-rate would be greater in
prolonged fasting than under normal conditions, respiration experiments
were made on the twenty-eighth, twenty-ninth, and thirtieth days of
the fast, in which the subject breathed an oxygen-rich atmosphere and
pulse records were simultaneously obtained. These experiments imme-
diately followed the morning respiration experiments for those days in
which the subject breathed air with a normal content, the average
pulse records for the latter experiments being 61, 63, and 59 respectively.
The pulse records in the experiments with the oxygen-rich atmosphere
were as shown herewith.
Date.
Time of day.
Pulse-rate.
May 12....
13....
14....
9h 04ma.m.-9h 48m a.m.
9 06 a.m.-9 52 a.m.
9 16 a.m.-9 49 a.m.
62
61
58
From these results it appears that the pulse-rate in the high-oxygen
experiment on the first day increased by 1 beat, on the second day
decreased by 2 beats, and on the third day decreased by 1 beat. In
former experiments2 with normal individuals, the average decrease in
xThe designation of the days for part of the data in table 6 is not in strict accordance with our
method of ending the experimental day with the end of the morning respiration experiment,
but the same numbering is used throughout for comparison purposes.
2Benedict and Higgins, Am. Journ. Physiol., 1911, 28, p. 1.
PULSE-RATE. 117
the pulse-rate with the oxygen-rich atmosphere was not far from 5 to
6 beats, but with this fasting subject the inhalation of an oxygen-rich
atmosphere apparently produced no change in the pulse-rate.
DIURNAL RHYTHM.
Beginning with the twelfth day of the fast, respiration experiments
were made each evening, just before the subject entered the bed calori-
meter. Comparing the records of the pulse-rate for these experiments
with those obtained in the morning experiments, it will be seen that
the pulse-rate in the evening was almost invariably higher than in the
morning. While the difference in the middle of the fast is very slight, it
tends to increase as the fast progresses, until on the thirtieth day there is
a difference of 12 beats. On the twenty-first day the evening rate was
2 beats lower than that observed in the morning, the lowest pulse
records for the evening experiments being found on this day.
IRRITABILITY OF THE HEART.
In all of the comparisons of the pulse records it is difficult to find
any very definite indications of the so-called " irritable heart," espe-
cially emphasized by the Berlin investigators in their study of Cetti.
It is true that the change from lying asleep to lying awake resulted in an
increase in the pulse-rate, the difference in the fasting period being not
far from 7 to 8 beats, with a maximum of 10 and a minimum of 4 beats.
It is further true that the change from a sitting to a lying position, as
noted at the time of the blood-pressure tests in the latter part of the
fast, tended to decrease the pulse-rate from 4 to 15 beats per minute,
averaging not far from 7 to 8 beats. Likewise, the pulse-rate in the
evening respiration experiments averaged a few beats more than in the
morning respiration experiments. But no great indication of an irrita-
bility of the heart was noted with any of these minor changes in position
and activity.
On the food days, however, there was a great increase in the pulse-
rate at the time the food was taken. Furthermore, the records obtained
at the time of the blood-pressure observations show that the pulse-rate
did not quickly return to the minimum after changing from a sitting
to a lying position for the pulse-rate in the lying position is usually a
few beats higher than the value obtained in the morning respiration
experiments. Since the value for the lying position was found almost
immediately after the test, it is hardly possible that the body had time
to adjust itself to the position, but the tendency to reach the minimum
found in the morning is worthy of note. Even with the slight activity
due to the blood-pressure tests and changing the position, the pulse-
rate is not so high as the pulse-rate obtained in the evening respiration
experiments, showing that the prevailing diurnal variation was greater
than that obtained with slight activity earlier in the day.
118 A STUDY OF PROLONGED FASTING.
A point of considerable importance is the fact that there was a
distinct tendency for the pulse-rate to reach a minimum between the
fourteenth and twenty-second days of the fast, that is, during the third
week. Consequently, all of the pulse records, including the average
minimum records inside the bed calorimeter, the average for the calori-
meter experiments, and the average for the morning respiration exper-
iments, show a tendency to increase in the fourth week. The same
tendency may be noted in the pulse records obtained at various times
throughout the day. It is clear, therefore, that during the latter part
of the fast the heart of the subject was in a somewhat more irritable
condition than during the third week.
The observations on the pulse-rate in this fasting experiment have
great significance when compared with the simultaneous measurements
of the metabolism, as is done in subsequent sections of this publication.
The records have therefore been presented in extenso, as they show
strikingly that the pulse-rate may legitimately be used as an index of
the metabolism. The total metabolism was measured only during the
times when the subject was on the respiration apparatus or inside the
bed calorimeter. For the remainder of the day, especially during those
times when the subject was most active, the metabolism was not meas-
ured, so that the probable metabolism at these times must be estimated
in so far as possible from the records of the pulse-rate. We therefore
have little evidence of the effect of muscular activity upon the heart,
except as shown by the few pulse records taken following slight activity.
These do not indicate a distinctly irritable heart.
The muscular activity of the subject was probably greater at the time
of the dynamometer tests than during any other observations. The
relatively few records made before and after these tests show a distinct
rise in the pulse-rate incidental to the dynamometer test, but also a very
rapid return. Unfortunately they were not taken with sufficient
regularity for us to note positively any indication of the increased or
decreased irritability of the heart as the fast continued. It was the
intention to study in this experiment the influence of light muscular
activity upon the heart beat and the metabolism, but here again the
unwillingness of the subject to engage in muscular activity of any kind
prevented valuable observations originally planned for.
BLOOD PRESSURE.
As it seemed desirable to supplement the observations of the pulse-
rate throughout the fast with observations of the blood pressure, the
determinations were made by Mr. H. L. Higgins. In accordance
with the advice of Dr. E. P. Cathcart, who had previously experimented
with Beauts, the values were secured for both the lying and sitting
positions, as it was quite possible that later in the fast it might be
advantageous to keep the subject in bed. The records were made
almost invariably shortly after noon and immediately alter the subject
had taken a glass of water. Employing the auscultatory method with
the Erlanger sphygmomanometer, both the systolic and diastolic pres-
sures were obtained.
In former fasts emphasis has been chiefly laid upon the determina-
tions of the systolic pressure. Luciani, in using the sphygmomanom-
eter of von Basch with an aneroid manometer, found that the varia-
tions in the blood pressure at the radial artery had a tendency to
decrease as the fast continued, the decrease being from 220 mm. on
the first day of fasting to 120 mm. on the twenty-sixth day. On the
last four days of the fast there was a slight tendency for the pressure to
increase. Luciani also states that there were daily fluctuations from
morning to evening, which frequently, but not always, corresponded
to changes in the temperature and the pulse-beat.
Cathcart,1 using the C. J. Martin modification of the Riva-Rocci
apparatus, found with Beaute" a continual fall in the maximum pressure
during the 14 days of the fast, his values being 108 for the first day and
for the subsequent days 96, 98, 98, 92, 94, 88, 90, 88, 92, 88, 94, 90,
and 88. These observations were always taken in the evening. Cath-
cart's determinations were controlled by Charteris,2 who used the same
instrument and presumably made the records at about the same time.
The values differ but little from those reported by Cathcart, showing
an excellent agreement of blood-pressure determinations by two obser-
vers. Charteris concludes that the pulse-wave became shorter and
weaker, but remained regular in rhythm, and that the arterial pressure
gradually sank so that at the end of the fast the fall amounted to
almost 25 per cent of the normal reading. He noted a rapid recovery
after the fast, the pressure being again practically normal after the first
week of food.
In a 30-day fast reported by Penny,3 in which he was himself the
subject, the author states that the blood pressure was taken by a
Martin's modification of the Riva-Rocci sphygmomanometer and fell
Cathcart, Biochem. Zeitschr., 1907, 6, p. 109.
2Charteris, Lancet, 1907, 173, p. 686.
'Penny, British Medical Journal, 1909, p. 1414.
119
120 A STUDY OF PROLONGED FASTING.
steadily during the fast from 1 10 to 90 mm. The daily observations are
not recorded.
No direct blood-pressure measurements were made by the Berlin
investigators on their fasting subjects, although from an examination
of the pulse curves made with a Marey sphygmograph at the radial
artery, Senator and Mueller1 concluded that there was a noticeable
decrease in the arterial tension which produced not only a dicrotism
but likewise a decrease in the elasticity. This was more noticeable
with a subject who fasted 10 days than with another who only fasted
6 days, notwithstanding the fact that on the last fasting day the pulse-
rate was so weak that they could not secure a suitable curve.
The observations made of the blood pressure in the fasting experi-
ment with our subject L. are given graphically in figure 19, the curves
representing the systolic, diastolic, and pulse pressures for both posi-
tions of lying and sitting. With these are compared curves showing
the average pulse-rate secured in the bed-calorimeter experiments
throughout the night and also in the morning respiration experiments.
During the latter part of the fast, the pulse-rate was likewise secured
at the time the blood pressure was taken, in both positions of lying and
sitting. These values are given in table 6 (pages 113 and 114).
As is usually the case, the systolic blood pressure when the subject
was lying down was invariably somewhat higher than when the subject
was in a sitting position, with a general tendency for the difference
between the two to become greater as the fast progressed. On the last
day of the fast, however, the difference was not much greater than at
the first of the fast. The curves for the diastolic pressure also show
higher values for the lying position, although the difference is not so
great as with the systolic pressure, for up to about the fifteenth day
the two curves are approximately the same.
The systolic pressure for the lying position falls quite rapidly through
the first half of the fast, fluctuating considerably in the last half above
or below the average value of 100 mm. of mercury. The records ranged
from 134 mm. on April 16 to 94 mm. on April 30. An even lower value
was obtained on May 16, 26 hours after taking food, namely, 92 mm.
Two days later, however, it had increased to 124 mm.
The curve for the systolic pressure for the sitting position is nearly
parallel to that of the systolic pressure for the lying position, although
in the latter part of the fast the values were considerably lower and the
fluctuations were not so great. The range in values was from 123 mm.
on April 16 to 83 mm. on May 8. During the latter part of the fast
it averaged not far from 90 mm.
The diastolic pressure for the lying position shows a marked fall from
the third to the fourth day. Subsequently there is a gradual fall to the
middle of the experiment, with a distinct tendency in the latter part of
^ehmann, Mueller, Munk, Senator, and Zuntz, Archiv f. path. Anat. u. Physiol, u. f. klin.
Med., 1893, 131, Supp. p. 101.
BLOOD PRESSURE.
121
APRIL MAY
II 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 I 2 3 4 5 6 7 8 9 10 1 1 12 13 14 15 16 17 18
DAYS Or DAYS OF FASTING DAYS OF
F00D I 2 3 4-5 6 7 8 9 10 1 1 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 F00D
Fig. 19. — Chart showing blood pressure, pulse pressure, and pulse-rate of subject L.
122 A STUDY OF PROLONGED FASTING.
the fast for the general course of the curve to rise. The highest record
was 100 mm. on April 16 and the lowest was 73 mm. on May 6. Twenty-
six hours after food was first taken the diastolic pressure in this position
fell to 74, but it rose on the third day to 102 mm.
A similar course is followed by the curve for the diastolic pressure in
the sitting position, although the fall from the third to the fourth day is
not so pronounced. The slight tendency to rise in the last part of the
fast is also shown in this curve. The highest record was 90 mm. on
April 16 and the lowest was 70 mm., recorded on April 30, May 1, and
May 2.
The general trend of all four curves for the systolic and diastolic
pressures in the two positions is a distinct decrease in blood pressure
during the first 15 days of the fast, followed either by an average con-
stant value or a slight tendency for the pressure to rise in the last part
of the fast.
With these observations of the systolic and diastolic pressure, it was
possible to obtain the pulse pressure, this being the difference between
the maximum and minimum blood pressures. According to some
writers, the pulse pressure is of more significance than the blood-pres-
sure observations themselves. An examination of the curves shows
that the pulse pressure in the lying position was, except on one day,
higher than the pulse pressure for the sitting position. The curves are
by no means parallel, however, as there are great differences in the
levels. The pulse pressure for the lying position decreases until the
fifteenth or sixteenth day of the fast, and then fluctuates considerably,
but averages approximately 23 mm. The pulse pressure for the sitting
position shows a much more marked similarity to the trend of the blood-
pressure curves, i. e., a falling pressure until about the fifteenth day of
the fast, followed by a period of approximately constant values, and
finally a distinct tendency to increased pressure towards the end of
the fast.
In considering the blood-pressure records, it is of interest to compare
them with the average pulse-rate curves obtained in the bed calorimeter
experiment during the night and on the respiration apparatus in the
morning. These curves are given in the lower part of figure 19, and
show a general parallelism with each other. Curves are also given
showing the pulse-rate records for both positions obtained in the latter
part of the fast at the time the blood pressure was taken. A fact of
special interest in connection with these curves is that the average
pulse-rate in the latter part of the fast has a tendency to rise during the
calorimeter experiment and also in the morning respiration experiment.
During the same period the blood-pressure curves remain essentially
constant, showing only a slight tendency to rise, the pulse pressure for
the sitting position alone having a tendency to follow more closely the
average pulse record. From this it may be concluded that the pulse-
BLOOD PRESSURE. 123
rate, which is considered in this laboratory as an index to the total
metabolism, is not so indicative of the actual work of the heart as of
the general metabolic tonus of the whole body.
The distinct decrease in blood pressure as the fast progressed may be
due to one or both of two causes. The first is the decrease in the con-
tractibility of the heart muscles as the fast continued and the second
is the decrease in the general tone of the peripheral vessels. Luciani
points out that theoretically the heart muscles should decrease more
than any other muscles of the body, as they are continually at work, and
that such a decrease has been found with fasting subjects. As will be
seen from Dr. GoodalPs report,1 there was evidence of a distinct and
regular decrease in the size of the heart during the fast. It is true that
the decrease in the size of the heart observed by Dr. Bianchi, on Succi
in his Florence fast, was not so great as that found by Dr. Goodall with
L. ; nevertheless the general conclusion in both series of observations is
the same, namely, that there is a considerable diminution in the size
of the heart as the fast progresses.
Somewhat in opposition to the belief that the size of the heart is the
determining factor in the decrease in the blood pressure is the fact that
on the third day after the fast the systolic and diastolic pressures had
both returned to their normal level; the complete regeneration of
the heart muscles in this time is difficult to conceive. A disturbing
factor entered into the observations of the last day as the subject was
laboring under intense psychical excitement.
^ee page 65.
THE BLOOD.
By J. E. Ash, M. D.,
Department of Pathology, Harvard Medical School.
CORRELATION OF LITERATURE.
The blood has frequently been studied during inanition, but an
exhaustive search through the literature brought to light only a few
records of systematic examinations covering so long a period of fasting
in man as the case that forms the basis of this report. Fasts of man
conducted under scientific supervision and including blood examina-
tions are limited in number, though the work on animals has been
rather prolific. Before presenting the findings in Levanzin's blood,
there is given a correlation of abstracts from this literature, of interest
not only from an historical standpoint, but in demonstrating the diver-
sity of results obtained by the various observers.
Among the earliest references are those to the work of Valentin1 in
1838, who concluded that there was no alteration in the relation of
blood-weight to body-weight as a result of exhaustive starvation, and
that of Bidder and Schmidt,2 in 1852, though these latter do not report
anything more specific than an increase in the solid constituents in
the blood of a starving cat. With reference still to the blood as a
whole, London,3 much later studying a series of 8 rabbits from which
both food and drink were withheld, found a loss in total quantity,
proportional, though, to loss of body-weight. Pashutin4 concludes,
as the result of the work of Heidenhain, Panum, and Voit, that the
blood is not impoverished by fasting; on the contrary, in certain periods,
the organism is plethoric. He holds it as remarkable that the number
of erythrocytes increase — probably, however, only because of the rapid
decrease in plasma. This latter fact is demonstrated markedly in the
dogs observed by W. Muller and Buntzen,6 though in none of their
animals did a loss of more than 15 per cent in body-weight occur.
Luciani6 holds that, aside from water-content, the blood exhibits a
resistance similar to the nervous system and that the apparent fluctua-
tions, in corpuscular content at least, depend chiefly on the amount of
water consumed. Chossat, however, quoted by Pashutin,4 considered
Valentin, Repel, f. Anat. u. Physiol., 1838, 3, p. 156.
2Bidder u. Schmidt, Die Verdauungsf ahrte in der Stoffwechsel, Mitau u. Leipzig, 1852, p. 328.
3London, Note sur la question du changement de la quantite generate et de l'alcalinite du sang
dans le jeune absolu. Arch, des Sciences Biol., 1895-96,4, p. 516. (Abstract by Miihlmann. See
footnote 4, p. 125.)
4Pashutin, Pathological Physiology, Inanition. 1902, 2, pt. i, p. 81 (Russian).
6Miiller and Buntzen, Transfusion and plethora. Christiania, 1875.
•Luciani, Fisiologia del digiuno, Firenze, 1889. Authorized translation by M. O. Fraenkel.
Das Hungern, Studien u. Experimente am Menschen, Hamburg u. Leipzig, 1890.
124
THE BLOOD. 125
that, next to the fat, the blood suffered the greatest loss, amounting even
to 75 per cent of its former weight. This view is not tenable in the light
of practically all other work and was evidently the result of faulty
technique or observations. Pashutin1 quotes Valentin as noting
striking general changes during hibernation — in part, that the blood
putrefies from 2 to 4 times more slowly, that the arterial blood is not so
bright a red, and that the venous blood is not so dark as normally, due
to disturbance of oxygen interchange.
ERYTHROCYTES.
Considering more specifically the blood elements and beginning with
the erythrocytes, we find that as early as 1843 Schultz2 studied starving
animals and found these cells atrophic, attributing the death of the
animals to the inability of the shrunken cells to bind oxygen. Jones,3
but a few years later (1856), observed that the corpuscles in dogs'
blood appeared to have undergone "partial decomposition." Others
since then have noted these striking alterations in shape and size of the
corpuscles, among them Manassein,4 Andral-Gavarret,4 Laptschinski,4
and especially Kagen,6 who studied dogs and rabbits. He found little
change in the first days, but as the fast progressed the red cells became
smaller and crenated ones appeared more frequently, until at the end
many "star forms" were seen and microcytes predominated. Liu-
boumdrow6 also found variations in the character of the red cells —
macrocytes, microcytes, and nucleated cells being common, especially
the large form, which reached 20 to 30 per cent of the total red count.
In the roundabout way, we get from Pashutin7 an abstract from
Wratsch, 1881, p. 78, quoting from foreign journals (not specified) a
reference to Dr. Tanner's blood after his 40-day fast in 1880. This was
a public exhibition, but well controlled and was absolute for the first 15
days. The plasma and white cells presented nothing unusual. The red
cells, however, were somewhat smaller than normal, being ^Vinr m°h m
diameter instead of 40100 to -ginnr inch.
Curtis8 made systematic observations of Griscom's blood during
his 45-day fast in 1880. This constitutes the longest period with blood
examinations of which any record could be found.
'Pashutin, Pathological Physiology, Inanition, 1902, 2, pt. i, p. 81 (Russian).
2Schultz, Beitr. x. phys. u. path. Chem. von Simon, 1884 (quoted by Muhlinann. See footnote
4, this page.)
3Jones, Smithsonian Cont. to Knowledge, 1856.
4Muhlmann, Russisch Literatur iiber die Pathologic des Hungerns. Centralblatt f. allgem.
Path., 1899, 10, p. 160.
'Kagen, Blood and blood pressure in starving organisms. Dissert. , St. Petersburg, 1884, Russian.
From the Laboratory for General and Experimental Path., Prof. V. Pashutin, St. Petersburg.
6Liuboumdrow, Changes in the blood and organs in starvation. 71 Dissert., 1893, Russian.
From the Path. Anat. Laboratory, Prof. W. Winogradow, St. Petersburg.
7Pashutin, Pathological Physiology, Inanition, 1902, 2, footnote p. 605 (Russian).
'Curtis, A study of blood during a prolonged fast. Proc. Am. Ass. Adv. Science, 1881, 30,
pp. 95-105.
126 A STUDY OF PROLONGED FASTING.
Curtis describes the morphology of the erythrocytes as follows :
The first examination, made just after Griscom's last meal, showed the cells
in abundance, of bright color, regular, smooth of outline, solid in appearance,
and of usual size — y-rVir inch.
On the third day they were paler and apparently not so firm.
Fourth day: The change had progressed. There were two sorts of cells to be
seen, one pale and large, the other deeper in color and contracted. Some of
the former were almost invisible, appeared soft and sticky, enveloping objects
encountered in flow. Their shape was altered to a round rim with abrupt
descent to a flat floor. They averaged -ohm inch. The other sort were
deeper in color, less transparent than normal, and covered with nodules like
blunt cones (evidently crenated). The cells had lost their usual concavity, and
seemed as though acted upon by an astringent, being much smaller than
normal — tjVtt inch.
Fifth day: The soft pale cells had disappeared, the smaller variety seemed
larger and nodular. Irregularities in shape were first noticed, some cells being
elongated, others lemon or club shaped, and still others had pointed ends.
Sixth to ninth days: The large soft form appeared and persisted in greater
or less numbers. Later small colored bodies like red corpuscles appeared,
measuring T¥Vv inch. All the erythrocytes at this time were small. The
extremely small ones continued to increase in number and diminish in size.
Sixteenth day: Corpuscle-like bodies observed as small as yuVxr inch. Those
of nfon to tuVtt inch diameter were like normal red cells. Others were nodu-
lated or of a chestnut-burr appearance.
Thirty-sixth day: " Saw an erythrocyte undergo direct division. From this
day on, the red cells changed for the worse." They became pale, ragged and
shrivelled. At this time the subject showed signs of weakened circulation —
vertigo, numbness of hands and feet.
Thirty-eighth day: He fainted on rising from bed.
Thirty-ninth day: There was scarcely a normal corpuscle to be seen.
Fortieth day: After an excursion of 2\ hours on the lake, there was a remark-
able change in the blood picture. The ragged, pale, and broken corpuscles all
disappeared and all the erythrocytes became smooth in outline and bright in
color. They seemed quite normal, except that they were smaller, averaging
7TVtf inch. After this, they again retrograded, became soft, pale, and sticky,
but never so bad as just before the lake excursion. Certain minute granules
were seen in this blood, granules which, in the author's experience, exist in all
other persons, except one, whose blood was examined. They were small red
points Tu^irTr inch in diameter and highly refractile. They are found also in
lymph and cow's milk. They existed in the blood in great numbers at first,
decreasing till after the eighth day; then disappearing until the twenty-fourth
day, when a few pale ones appeared. They then increased in number, but
only returned to their normal abundance after the fast was broken. (These
were apparently the platelets that he was observing.)
This report is given in detail to illustrate not only the painstaking
care with which the observations were made, but somewhat the com-
parative crudity of the methods employed at that time. As will be
seen, it is only in the earlier reports that the appearance of macrocytes,
microcytes, crenated and distorted cells are recorded, and it has
occurred to the writer that with the improvement of blood technique
the occasion for their presence was eliminated.
THE BLOOD. 127
In 1887 Senator1 reported finding a number of microcytes on the
13th day of a " Schlafsucht" in which a 54-year-old woman lay for about
7 weeks receiving as nourishment only a small amount of milk and wine.
In Succi's2 blood, late in the 40-day fast made as a public exhibition
in London in 1890, numbers of imperfect blood disks were observed.
Charteris,3 on the other hand, could find no alteration in shape, size,
or staining qualities of the erythrocytes in the blood of his human
subject during the fast of 14 days in 1907, except that a few nucleated
ones did appear during the last 4 days. Two examinations were made
on Gayer's blood by Dr. Wile.4 His was a public fast of 30 days,
undertaken largely for advertising purposes. He drank water ad lib.,
and though he was fairly well guarded, there was opportunity for his
obtaining food secretly. However, at the beginning of the fast his
weight was 210 pounds and at the end 174f pounds, a loss of over 35
pounds, which proves rather conclusively that the experiment was con-
ducted in good faith. On the eighteenth day, the red cells numbered
5,192,000 and on the thirtieth, 5,776,000, a slight rise. The absence
of anisocytosis or any degeneration of cells is mentioned specifically in
the report. The subject refused to allow further examinations, particu-
larly after breaking the fast, as desired by Dr. Wile in order to determine
a normal picture. In none of the other reports are the characteristics
of the individual red cell noted, so it is most likely that alterations of
importance did not occur.
Continuing the consideration of the effects of inanition on the numer-
ical estimation of the erythrocytes, the results are found to be at rather
wide variance.
Senator1 found no significant variation during the long period of
almost complete inanition already mentioned above. He attributed
little value to the small number of counts he made, but concluded
there was probably a slight diminution. During Cetti's 10-day and
Breithaupt's 6-day fasts, studied by Senator, Lehmann, et ah? an
increase was noted, amounting, in the former subject, to a million.
Cetti's normal count of 5,720,000 was above the average. (Table 7.)
Dup6ri66 also claims that a considerable increase in number occurs.
From the study of Succi's blood during a 30-day fast supervised by
Luciani,7 the latter concludes that the variations noted in numbers are
Senator, Ueber einen Fall von sog. Schlafsucht mit Inanition. Charit6-Annalen, 1887, 12, p.316
2The fasting man. Brit. Med. Journ., 1890, 1, p. 1444.
3Charterid, Record of changes observed in the blood count and in the opsonic power of a man
undergoing a prolonged fast. Lancet, 1907, 2, p. 685.
*Gayer's fast: A private communication from Dr. Wile, of New York City.
5Lehmann, Mueller, Munk, Senator, and Zuntz, Untersuchungen an zwei hungernden Men-
schen. Archiv f. path. Anat. u. Physiol., Virchow's, 1893, 131, supphft., p. 1.
6Duperie, Sur les variations physiologiques dans l'etat normal des globules du sang. Paris,
1878. Cited by Rollett in Hermann's Handbuch d. Physiolog., 4, (1).
7Luciani, Fisiologia del digiuno, Firenze, 1889. Authorized translation by M. O. Fraenkel.
Das Hungern, Studien u. Experimente am Menschen, Hamburg u. Leipzig, 1890.
128
A STUDY OF PROLONGED FASTING.
only relative, depending on the concentration or dilution of blood from
alteration in water-content. While on the twenty-seventh day the
greatest loss was noted, there followed on the twenty-ninth day a rise
that brought the number to the level of the first day. Andreesen,1
Malassez,1 and Lepine1 found that while in the beginnings of the fasting
periods there would be an increase in number, in the later days a
decrease occurred.
Table 7. — Red-cell counts on Ceiti and Breithaupt.
Day.
Cetti.
Breithaupt.
3d fast
6,720,000
5^285,000
6,830,000
6,660,000
6,730,000
4,953,200
6,184,000
'4,801,000
'4,820,000
4,812,000
4th fast
6th fast
9th fast
2d diet
2 weeks later
'Before and after first meal on the sixth day.
Clinical records furnish us with the two following reports of relevant
interest. The first is of one of Landouzy's patients studied by
Malassez.2
"A boy of 18 lived 3 months and 20 days with a stricture of the esophagus,
the result of swallowing H2S04. He obtained practically no nourishment, as
he vomited food administered by tube. Ten days before death, he began to
take a small quantity of milk and meat. The red cells numbered 3,600,000
20 days before death, and a week before the end 2,600,000, a decided loss
from normal. Two days before the boy died, the count had risen to 3,200,000.
A transfusion was performed immediately after this examination, followed in
20 minutes by another count which showed a rise to 3,500,000. The next day
they had returned to 3,200,000."
This is a striking loss, and while it is not possible to rule out a toxic
influence in this case, the almost complete inanition was no doubt the
prominent factor.
"The second case (reported by Brouardel)3 was a man of 48 years, who lived
4 months 12 days after an experience similar to that just quoted. But one
blood examination was made, and that 2 days before death, when the erythro-
cytes were 4,849,000 and the leucocytes 7,852."
While to the present writer these would be considered as practically
normal counts, the author of the report concludes that they demonstrate
a concentration of the blood.
'Miihlmann, Russiche Literature iiber die Pathologie des Hungerns. Centralblatt f. allgem.
Path., 1899, 10, p. 160.
2Malassez, Bull, et mem. de la soc. med. des hopitaux de Paris, 1874, 11, p. 124.
"Brouardel, Union Med., 1876, ser. 3, 22, p. 408.
THE BLOOD.
129
Von Noorden1 has found the corpuscular content normal in five cases
of gastric ulcer with emaciation. He remarks that in spite of the anaemic
appearance presented by patients suffering from various conditions caus-
ing malnutrition, their blood is usually normal. This of course does not
hold in those cases where the cause of the malnutrition has a direct
Table 8a. — Estimations of red cells during Griscom'sfast (Curtis).
Day.
Estimation
of red cells.
Remarks.
4th
4,320,000
5th
4,485,000
6th
2,370,000
8th
4,860,000
10th
3,260,000
11th
4,720,000
12th
3,790,000
13th
4,480,000
14th
4,210,000
15th
2,800,000
18th
5,790,000
19th
6,770,000
20th
6,500,000
Flatulence. Patient felt quite ill ;
21st
5,600,000
took enema, causing stool.
22d
2,100,000
23d
5,460,000
24th
5,420,000
25th
3,920,000
These figures are not entirely des-
26th
4,160,000
titute of symmetry.
27th
2,540,000
28th
3,130,000
29th
3,180,000
Counts on 6 intervals of 6 days.
30th
3,180,000
10 " " 4 "
31st
3,360,000
15 " " 5 "
32d
4,420,000
22 " " 7 '*
33d
3,600,000
27 " " 5 "
34th
3,900,000
37 " " 10 "
35th
3,700,000
40 " " 3 "
36th
3,810,000
44 "4 "
37th
3,520,000
Pointing to the opinion held of a
38th
4,080,000
certain limited duration of life
39th
4,200,000
of red blood corpuscles.
40th
3,200,000
41st
3,390,000
42d
3,590,000
43d
3,490,000
44th
3,150,000
45th
5,390,000
influence on the blood, as in infections. He discusses this practical phase
of the subject and gives many references to observations of the effects
on the various properties and constituents of the blood of "clinical"
inanition.
xVon Noorden, Metabolism and practical medicine, Anglo-American issue, Chicago, 1907, 2,
p. 28.
130 A STUDY OF PROLONGED FASTING.
Returning to Curtis's1 article, we find the following protocol (table 8 a)
of the 38 numerical estimations made of the red cells during the 45-day
fast of Griscom, with the former's comments.
"The subject was at his worst physically and mentally between the twenty-
seventh and fortieth days, and during this period the counts were consistently
low. On the fortieth day he took the excursion on the lake, which was appar-
ently the cause of the drop of 1,000,000 from the count of the preceding day.
It will also be noted that for the few days before a decided fall in number there
was usually a rise. The corpuscles on the days of these low counts always
appeared healthier than at other times. On the last day Mr. G. drank no
water and the high count of that day may have been due to concentration."
Kagen,2 in 1884, claimed that the ordinary methods of determining
the cell content of blood were open to so many sources of error that the
results were not dependable. He limited his observations, therefore,
to the direct estimation of the solid constituents, the specific gravity
(by pyknometer), and haemoglobin content (Malassez's hsemochromom-
eter) . He examined 6 dogs and found in the early days an increase in all
three factors, attributing the changes to concentration through water
loss. The amount of solid constituents, he claims, can equal even at the
end of the fast that present under normal conditions. Liuboumdrow3
noted, as an average of observations on 17 dogs, a slight increase in
erythrocytes till the loss of body- weight amounted to 10 to 15 per cent,
then a steady decrease till death, the diminution amounting to as high
as 32 percent on the twenty-eighth day. Nasse4 also found an increase
in number in a dog after 11 days of complete fasting. As proof that
this was due to variation in water-content, he states that he obtained a
reaction in the opposite direction when the animal was again allowed
water.
Pol6taew5 studied 8 dogs that received neither food nor water, dying
after loss of 50 per cent in body-weight. These all showed an increase
in red cells until late in the fasts, after a loss of 30 percent body- weight,
when there was a gradual decrease till death. Pol6taew is not satisfied
with the explanation of this finding simply on the grounds of concen-
tration, for he found an increase also in the dogs that were allowed
water. He holds that while there may be interference with blood
formation, there is also less destruction for bile formation. Tauszk,6
Curtis, Physiology of autonutrition ; A study of blood during a prolonged fast. Am. Ass.
Adv. Science, 1880, 30, pp. 95-105.
2Kagen, Blood and blood pressure in starving organisms. Dissert. St. Petersburg, 1884,
Russian. From the Laboratory for General and Experimental Path., Prof. V. Pashutin, St.
Petersburg.
3Liuboumdrow, Changes in the blood and organs in starvation. 71 Dissert., 1893, Russian.
From the Path. Anat. Laboratory, Prof. W. Winogradow, St. Petersburg.
4de Martigny u. Nasse, Ueber den Einfluss der Nahrung auf daa Blut. Marburg u. Leipsic,
1850.
6Poletaew, The morphologic composition of the blood in complete and incomplete starvation
in dogs. Dissert. 97, 1894, St. Petersburg (Russian). From the Laboratory of Path. Anat.,
Prof. Uskow, Riv. internaz. d'ig., Roma, '95, 6, p. 129, and Arch. d. sc. Biol. St. Petersburg,
1893. 2, p. 794.
6Tauszk, Jahrsb. tiber d. Fortschr. der Thier-Chemie, 1894, 24, p. 147; abstracted from Orvoia
hetilap, Budapest, 1894, p. 512. Also Haematologische Untersuchungen am hungernden mens-
chen. Wien. klin. Rundschau, 1896, 10, p. 306.
THE BLOOD.
131
in a study of Succi's blood during his 30-day fast in 1894, found, after a
short interval of decrease, a moderate increase in the red cells. (See
table 8 b.) The form of the cells remained normal to the end.
Daiber1 in 1896 drew his conclusions as to the effect of inanition on
the blood from his findings in Succi's urine during a 20-day fast. There
was an increase in urobilin and earthy alkaline phosphates, both of
which were to be accounted for by assuming an enormous destruction
of erythrocytes, though he does concede that the phosphates might
have come from tissue destruction elsewhere. As proof of the adapta-
bility of the blood to altered conditions, he presents the decrease in
urobilin and disappearance of phosphate sediment noted after the
fifth day, showing an acquired resistance to the previously destructive
influence of fasting. The urobilin was distinctly demonstrable through-
Table 8 b. — Succi's red-cell counts.
Ratio
Day.
Red cells.
of red to
white cells.
Third
5,246,000
1 :545
Eighth
4,840,000
1 :584
Thirteenth. . .
4,932,000
1 :684
Seventeenth . .
5,136,000
1 :744
Twenty-first. .
5,160,000
1 :938
Twenty-fifth. .
5,268,000
1 : 1097
Thirtieth
5,472,000
1 : 1302
out the fast, though greatly reduced, but the phosphate sediment was
replaced by one of urates. The plasma remained intact, as no transu-
dation of its constituents, particularly albumen, through the kidneys
could be demonstrated. Daiber concludes that the conditions during
inanition must resemble those present in continued fevers in which there
is usually red-cell destruction sufficient to give rise to a demonstrable
anaemia. In these cases urobilin is present in the urine in distinctive
amounts.
In the dog which died on the twenty-fifth day, after a loss of 52 per
cent in body-weight, Hayem2 reports an increase till the eighteenth
day from 4,200,000 to 5,500,000. There was then a slight decrease,
though at the end the erythrocytes numbered 4,800,000, still above the
original count. The hsematoblasts decreased continually during the
fast. Reyne3 found a progressive increase in the number in a dog
dying on the twenty-fifth day of starvation. Charteris,4 on the other
'Daiber, Beitrag zur Kenntniss des Stoffwechsels beim Hungern. Schweitzer Wochenschr. f .
Chem. u. Pharm., Zurich, 1896, 34, p. 395.
2Hayem, Lecons sur les modifications du sang. Paris, 1882, p. 382.
3Reyne, quoted by E. Bardier in his article on Inanition, in Dictionnaire de Physiologie,
Charles Richet, 9, p. 99.
4Charteris, Record of changes observed in the blood count and in the opsonic power of a man
undergoing a prolonged fast. Lancet, 1907, 2, p. 685.
132 A STUDY OF PROLONGED FASTING.
hand, in the case of his already mentioned, could find no suggestive
variation, though there was some daily fluctuation.
Gordon1 studying the blood of Martin, a medical student who under-
went a 9-day fast with the uniform daily water consumption of 24
ounces, could find practically no variation in red-cell count, except that
on the sixth day of refeeding it was about 1,000,000 below the normal.
At the end of the first week of Succi's fourth fast, one of 40 days con-
ducted in London,2 the red cells numbered 6,500,000, an increase of
1,000,000 over the average normal individual's count. It may be more
or less in this particular case, as the normal count is not given. A. R.
Diefendorf , however, found a slight diminution during and a relatively
rapid rise immediately following each of the two fasts of a man of 7 and
4 days interrupted by a feeding period of 19 days, which formed the
basis of Benedict's3 report.
Three counts were made on the blood of Dr. Penny,4 who fasted for
30 days, in 1909, drinking only distilled water. They demonstrated a
moderate increase till the twentieth day and a loss of 1,000,000 during
the remaining 10 days. Here again no normal count was obtained.
The results for the three counts were for the twelfth day, 6,600,000;
twentieth day, 7,000,000; thirtieth day, 6,000,000. Ronsse and van
Wilder5 hold there will always be a slow increase in erythrocytes if
water as well as food is withheld.
Though the conditions are not altogether analogous, it is interesting
to note that in hibernating animals there is a decided decrease in erythro-
cytes, as is reported by Ranke.6
HAEMOGLOBIN.
Senator, using the v. Fleischl method, noted a moderate increase in
his woman subject7 and in Breithaupt;8 a loss, however, of about 20
per cent in Cetti8 in 9 days. (Table 9.)
Liuboumdrow9 (with Malassez's method) found a slight increase in
the blood of dogs until a loss of 10 to 15 per cent of body-weight had
occurred, when a decrease was recorded that progressed until the
animals died.
Benedict3 reports a slight loss during, with a rise after, the fasts,
corresponding to the fluctuations of the erythrocytes. The v. Fleischl
and Tallqvist methods were used.
Gordon, A prolonged fast. Montreal Med. Journ., 1907, 36, p. 482.
2The fasting man. Brit. Med. Journ., 1890, 1, p. 1444.
»Benedict, Carnegie Inst. Wash. Pub. 77, 1907, p. 322.
*Penny, Notes on a thirty-day fast. Brit. Med. Journ., 1909, 1, p. 1414.
'Ronsse et van Wilder, Variations du nombre des globule3 rouges et du taux de l'hemoglobine
au cours de l'inanition chez le lapin. Arch, intern, de Pharm. et Ther., 1903, 11, p. 301.
6Ranke, Grundziige der Physiolog., 3d ed.t p. 380.
"Senator, Ueber einen Fall von sog. Schlafsucht mit Inanition. Charite-Annalen, 1887, 12, p. 316.
8Lehmann, Mueller, Munk, Senator, and Zuntz, Untersuchungen an zwei hungernden Men-
schen. Archiv f. path. Anat. u. Physiol., Virchow's, 1893, 131, supphft., p. 1.
'Liuboumdrow, Changes in the blood and organs in starvation. 71 Dissert., 1893, Russian.
From the Path. Anat. Laboratory, Prof. W. Winogradow, St. Petersburg.
THE BLOOD.
133
Martin's1 haemoglobin was 90 per cent the first day, 95 per cent the
fourth, and 90 per cent again on the last day, the ninth, dropping to 80
per cent 6 days after resuming food.
Penny2 showed an increase of 8 per cent during his 30-day fast, going
from 104 to 112 per cent.
Table 9. — Haemoglobin estimations on Cetti and Breithaupt.
Day.
Before fast
Second day
Third day
Fourth day
Sixth day (before first meal)
(2 hours after first meal)
Ninth day
Second day of diet
Cetti.
Per cent.
115-118
110
85-90
Breithaupt.
Per ant.
107
114
114
110
130
116
114
In Luciani's3 report is to be seen a variation synchronous with that
of the erythrocytes, except on the eleventh and thirteenth days, when
the percentage failed to rise with them. There was a small increase on
the third and twenty-first days and it was lowest on the thirteenth and
twenty-sixth days. The highest estimation was 90 per cent, the lowest
72 per cent (v. Fleischl apparatus). He concludes that there is an
actual loss of haemoglobin. Quoting his studies with Bufalini, carried
out in Sienna in 1882, on a dog that lived 53 days without food, he states
there was a rapid rise during the first 6 days. (Modified Bizzozero's
method used). This he explains as being due not only to concentration
from loss of water, but to the more rapid consumption of plasma than
corpuscles. The initial rise was followed by a gradual decrease con-
tinuous until the last 12 days of the fast, during which the percentage
was constant.
Gayer' s4 percentage rose from 80 on the eighteenth day to 100 on the
thirtieth, when his fast was broken.
Charteris,6 reporting the 14-day fast, notes a drop from 110 per cent
to 96 per cent, after remaining unaffected for the first few days. The
loss was not recovered until several days after breaking fast.
Subbotin,6 using Preyer's method (spectroscope), found a decrease
of haemoglobin in a case fed on nitrogen-free diet. By the twenty-sixth
day it had fallen from 13.80 per cent to 1 1.65 per cent and on the thirty-
'Gordon, A prolonged fast. Montreal Med. Journ., 1907, 36, p. 482.
2Penny, Notes on a thirty-day fast. Brit. Med. Journ., 1909, 1, p. 1414.
3Luciani, Fisiologia del digiuno, Firenze, 1889. Authorized translation by M. O. Fraenkel.
Das Hungern, Studien u. Experimente am Menschen, Hamburg u. Leipzig, 1890.
4Gayer's fast: A private communication from Dr. Wile, of New York City.
6Charteris, Record of changes observed in the blood count and in the opsonic power of a man
undergoing a prolonged fast. Lancet, 1907, 2, p. 685.
6Subbotin, Zeitschr. f. Biol., 1871,7, p. 187.
134
A STUDY OF PROLONGED FASTING.
eighth it was 9.52 per cent. In starving rabbits, however, there was
an increase due (he concludes) to decrease in water-content of the blood.
In a dog that starved for 38 days there was very little variation, from
13.80 at the beginning to 13.33 per cent at the end.
Though the inanition was only partial, the experiment of v. Hoss-
lin1 is suggestive and his conclusions are interesting. He observed
two growing dogs, one of which, (a), weighing 3.2 kilograms, was given
only one-third the nourishment that (6), weighing 3.1 kilograms,
received. Table 10 presents the results:
Table 10. — Results of v. Hdsslin's observations.
Dog.
66th day.
124th day.
18 months.
Kilos.
Percent-
age of
haemo-
globin.
Kilos.
Percent-
age of
haemo-
globin.
Erythro-
cytes.
Kilos.
Percent-
age of
haemo-
globin.
Erythro-
cytes.
6.5
11.6
11.2
10.2
8.5
23.4
16.0
14.9
7,970,000
6,820,000
9.5
30.3
15.5
17.6
7,300,000
8,300,000
The difference in haemoglobin content of the two dogs, in spite of
extreme emaciation in dog (a), is still within physiological limits,
proving how independent of the amount of nourishment the haemoglobin
really is. He claims a greater influence on the blood constituents
through the nature than the amount of nourishment, long-continued
impoverishment in albumen, e. g., causing a decrease in both haemo-
globin and red cells. We do wrong, he claims, to consider as a result
of the malnutrition the apparent or real anaemia seen in poorly nourished
individuals. Either it is only (1) apparent, the haemoglobin and red-
cell content remaining high while the patient appears anaemic from
contraction of peripheral vessels in an effort to compensate for the
lessened thermogenesis, or it may be (2) actual anaemia, the result,
however, of the condition that is responsible for the malnutrition, e. g.,
long-continued fever, cancerous ulcers, repeated haemorrhages, intes-
tinal parasites, etc. The amount of nutrition, however, has a great
influence on blood formation, so that anaemia, from whatever cause,
will clear up much more readily under good than poor nourishment.
The effect of uncomplicated inanition, therefore, is to reduce the total
quantity of blood as it does the muscle and organ volume generally,
rather than disturb the individual constituents. On refeeding, this
total quantity is restored quickly, so that there appears an anaemia, in
spite of the better nourishment — a relative condition, however, due to
V. Hosslin, Ueber den Einfluss ungeniigender Ernahrung auf die Beschaffenheit des Blutes.
Gesellschaft f. Morphologie u. Physiol., Munich, 1890, p. 119.
THE BLOOD. 135
more rapid restoration of fluid than haemoglobin and cellular elements
and apparent till the normal relation is established.
Gallerani1 found that the mean resistance of the haemoglobin of
dogs and frogs to solutions of NaClof various percentages was increased
during fast; the resistance to the high percentage solutions decreased,
while to the low it increased. The former was due, he claimed, to the
absence of newly formed haemoglobin, which is more resistant to the
stronger solution and the latter to the absence of very old or much used
haemoglobin, which is less resistant to the weak solution.
Hermann2 in discussing the subject, says that the changes found
with the ordinary methods of examination can well be due to concen-
tration, because the water is the most variable of the blood constituents.
Results are conclusive, therefore, only when they deal less with the
haemoglobin content than with the relation of haemoglobin to the total
quantity of solid constituents.
Groll's3 work was carried out on this line with rabbits, cats, and one
dog. Their fasts were absolute, no water being allowed. They existed
under these conditions for from 1 to 22 days. He estimated the haemo-
globin quotient by dividing the percentage of haemoglobin, as deter-
mined by the v. Fleischl apparatus, by the percentage of solid con-
stituents. The latter was obtained by heating measured quantities of
blood to 110° C. till the weight remained constant. Withfew exceptions
the color quotient thus found was increased. The simple haemoglobin
per cent showed a rise in all the animals, being, however, only a relative
increase, due to concentration of the blood. The diminution in the
total solids was due to the greater susceptibility to destruction of the
other solid constituents. During the period of restitution, there was
a diminution in the haemoglobin. This, again, was largely a relative
change, occurring as a result of dilution from increased water intake,
though no doubt the haemoglobin is slower in regenerating as well as
being more resistant to destruction. Groll concludes that the haemo-
globin is more stable in starving conditions than any of the other solid
constituents of the blood.
It would seem from these data that the red cells and haemoglobin
are particularly resistant, though in the long fasts there is no doubt a
slight loss of both elements. The consensus of opinion appears to be
that concentration of the blood through water loss is responsible for
the increase found during the fast and that dilution from more rapid
return of water than the other elements accounts for the decreases
found immediately after the fasting.
1Gallerani, Resistenz des Haemoglobins im Hunger. Jahrsb. u. die Fortschr. der Thier-
Chemie, 1894, 24, p. 120. (Abstracted from Ann. di chim. Farmacol., 1892, 15, p. 3.)
2Hermann, Untersuchungen iiber den Hsemoglobin-Gehalt des Blutes bei vollstandiger Inani-
tion. Dissert., Kdnigsberg i. Pr., 1887.
sGroll, Untersuchungen iiber den Haemoglobin Gehalt des Blutes bei vollstandiger Inanition.
Dissert., Konigsberg i. Pr., 1887.
136
A STUDY OF PROLONGED FASTING.
LEUCOCYTES.
More attention has been paid to the white cells than to any of the
other blood constituents and reports are more at variance as to just what
does happen to them during states of inanition. Almost every possible
change, especially numerical, has been observed at one time or another.
Morphologic alterations are recorded by Luciani,1 who noted an early
decrease in size, so that by the fifth day all the leucocytes were smaller
than the red cells. They recovered their normal size by the ninth
day, however.
Charteris,2 on the other hand, mentions specifically that he observed
no alteration in size.
Manassein3 reports the presence in the leucocytes of fasting rabbits
of refractile bodies that are not affected by acetic acid.
Kallmark4 observed, in his rabbits, rarefaction in the basophiles,
with agglutination and peripheral arrangement of the granules.
Curtis,5 in 1880, observed peculiar bodies, resembling leucocytes but
larger, consisting of spherules too small to measure. These cells meas-
ured ttsW inch in diameter and exhibited amoeboid movement. When
these very indefinite bodies were most abundant, the granules were
absent from the leucocytes. Curtis does not speculate as to whether
these were altered white corpuscles or a foreign cell entering the blood
from the tissues.
Hayem6 concludes that there is no essential change in the leucocytes
during starvation, at least in dogs.
Considered numerically both as to total and differential estimations,
the following results are reported from studies of fasts in man :
Table 11. — Cetti' 8 and Breithaupt' 8 white-ceU count.
Cetti fasted 11 days, Breithaupt 6.
Day.
Cetti.
Breithaupt.
Before fast
Normal.
4,800
4,200
12,300
7,950
6,500
6,870
7,000
Fourth day
Sixth day
Ninth day
Broke fast
Second day of diet ....
luciani, Fisiologia del digiuno, Firenze, 1889. Authorized translation by M. O. Fraenkel.
Das Hungern, Studien u. Experimente am Menschen, Hamburg u. Leipzig, 1890.
2Charteris, Record of changes observed in the blood count and in the opsonic power of a man
undergoing a prolonged fast. Lancet, 1907, 2, p. 685.
3Muhlmann, Russiche Literature iiber die Pathologie des Hungerns. Centralblatt f. allgem.
Path., 1899, 10, p. 160.
4Kallmark, Zur Kenntniss des Verhaltens der weissen Blutkorperchen bei Inanition. Folia
Hsemat., 1911, 11, pt. 1, p. 411.
*Curtis, Physiology of autonutrition : A study of blood during a prolonged fast. Am. Ass. Adv.
Science, 1880, 30, pp. 95-105.
6Hayem, Lecons sur les modifications du sang. Paris, 1882, p. 382.
THE BLOOD.
137
With both Cetti and Breithaupt1 a moderate decrease was observed
during, with a considerable rise for the first few days following, the
fasting period. (See table 11.)
Senator2 concludes that there is a lively new formation of leucocytes
on refeeding.
Luciani3 records a marked diminution in the early period of Succi's
30-day fast, dropping from 14,536, the count on the first day, to 861 on
the seventh. The count then rose to 1,550, where it remained until the
twenty-ninth day with slight fluctuations due to concentration and
dilution of blood. He attributes the marked diminution to the diges-
tive action of trypsin, which evidently enters the blood as such during
the cessation of intestinal digestion. He bases this theory on the work
of Albertoni, who by intravenous injection of trypsin got almost a
complete disappearance of leucocytes. The trypsin apparently has
no effect on the erythrocytes. It is quite possible, also, that there may
exist in the early days of the hunger period some special destructive
condition, evidenced also by the loss of haemoglobin. Two other factors
at work, he argues, are, first, the disappearance of the lymphocytes, they
no longer being required to alter the assimilated products of digestion
in the blood plasma (after the work of Schaeffer, Hofmeister and
Zawarykins); secondly, the leucocytes may have lost their "Wander-
lust." There would, then, not only be a failure of "outwandering"
from the blood, but of "inwandering" from the tissues as well, and the
latter would exert the greater influence on the number. The white
cells practically disappeared from Succi's blood during his fourth fast,
till late, when small and ill-formed corpuscles were found.
Tauszk4 notes a decrease in total count during one of Succi's 30-day
fasts. As will be seen in table 12 a this was due to loss of mononu-
Table 12 a. — Succi's total and differential white-cell counts.
Day.
Total
white
cells.
Poly-
morphs.
Mono-
cytes.
Eosino-
phils.
Third
Eighth
Thirteenth. . .
Seventeenth. .
Twenty-first .
Twenty-fifth.
Thirtieth
9,600
8,300
7,200
6,900
5,500
4,800
4,200
p. ct.
64.1
68!5
79!2
p. ct.
33.1
27 A
16.6
p. ct.
2.7
3.9
4.7
1Lehmann, Mueller, Munk, Senator, and Zuntz, Untersuchungen an zwei hungernden Men-
schen. Archiv f. path. Anat. u. Physiol., Virchow's, 1893, 131, supphft., p. 1.
'Senator, Bericht tiber die Ergebnisse des auf Cetti ausgefuhrten Hungersversuchs. Berlin
klin. Wochenschr., 1887, 24, p. 427.
3Luciani, Fisiologia del digiuno, Firenze, 1889. Authorized translation by M. O. Fraenkel.
Das Hungern, Studien u. Experimente am Menschen, Hamburg u. Leipzig, 1890.
4Tauszk, Jahrsb. iiber d. Fortschr. der Thier-Chemie, 1894, 24, p. 147, abstracted from Orvosi
hetilap, Budapest, 1894, p. 512. Also Hsematologische Untersuchungen am hungernden Men-
schen. Wien. klin. Rundschau, 1896, 10, p. 306.
138
A STUDY OF PROLONGED FASTING.
clear cells, including lymphocytes. The eosinophiles and polymorphs
were increased. Neubert1 found the opposite changes, that is, an
increase in mononuclear cells and decrease of the eosinophiles and
polymorpho-nuclears. His studies were made on cases of carcinoma
and pulmonary tuberculosis, so that the inanition was not simple.
There was little change in the white-cell content of Martin's2 blood
during his 9-day fast, except that on the second and ninth days they
rose to 10,000. As will be seen in table 12 6, there was a very slight
progressive loss in the polymorphs, while the lymphocytes were
increased somewhat on the sixth and ninth days. There was no
differential count made to specify the total rise noted on the second
day.
Table 12 b. — Martin's differential white-cell counts.
Day.
Poly-
morphs.
Small
mono-
nuclears.
Large
mono-
nuclears.
Transi-
tionals.
Baso-
phils.
Eosino-
philes.
Sixth day
p.ct.
68
60
65
59
58
p. ct.
22
30
18
32
35
p.ct.
6
7
23
9
6
p. ct.
4
2
1
p.ct.
3
1
p. ct.
3
Ninth day
Sixth day after ....
The results of the two examinations of Gayer's3 blood are given in
table 13. The striking points in this case are the very low total count
on both occasions, the increase in the small and the drop in the large
lymphocytes at the end of the fast. No explanation of these changes is
offered by Wile, who made the observations.
Table 13. — Gayer's total and differential white-cell counts.
Day.
Total
white
cells.
Poly-
morphs.
Large
lymph-
ocytes.
Small
lymph-
ocytes.
Eosin-
ophiles.
Baso-
phils.
Thirtieth
2,600
2,800
p.ct.
54
51
p.ct.
12.0
5.4
p.ct.
30.4
42.0
p. ct.
1.6
1.0
p.ct.
2.0
.6
In Professor Benedict's subject, whose blood was studied by Dr.
Diefendorf,4 there was evidently normally a high white-cell count.
The results are given in table 14.
During the first fast of 7 days there was a progressive diminution till
the last day, when there was a slight rise. After an interval of 19 days
there was a second fast of 4 days. During this period a gradual rise
xNeubert, Ein Beitrag zur Blutuntersuchung, speciell Phthisis pulm. und carcinom, Dorpat,
1889.
2Gordon, A prolonged fast. Montreal Med. Journ., 1907, 36. p. 482.
3Gayer's fast: A private communication from Dr. Wile, of New York City.
^Benedict, Carnegie Inst. Wash. Pub. 77, 1907, p. 322.
THE BLOOD.
139
amounting to about 2,000 was noted. The polymorphs averaged high
during both fasts, but at no time could the number be considered
distinctly pathological. The small lymphocytes averaged low. The
large lymphocytes were high during the last 2 days of the first period
and throughout the last. The eosinophiles were low and the baso-
phils high in both fasts.
Table 14. — S. A. B.'a total and differential white-cell counts.
Total,
white
cells.
Poly-
morphs.
Small
lymph-
ocytes.
Large
lymph-
ocytes.
Eosino-
philes.
Mast
cells.
1905.
Prior to fast:
Mar. 3 , , .
Fasting:
Mar. 61
7.,.
8,..
9
10
11
Diet:
Mar. 12
13...
14
15 , .
20 ...
Apr. 3
7 ,
Fasting:
Apr. 8
9
10....
11
Diet:
Apr. 12 ... .
25 ...
26
May 5
23,000
15,750
13,000
12,000
10,000
10,000
10,000
11,250
13,250
13,750
14,500
15,000
11,750
13,000
13,250
13.250
13,250
13,750
14,500
15,000
5,500
8,100
p. ct.
63
76
73
76
77
64
75
66
66.5
69
78
58.5
54
63
69
73
76
68.5
62
50
49
p. ct.
27.0
15.5
17.5
18.0
10.5
25.0
10.3
17.5
20.5
17.0
14.5
22.5
25.0
14.5
15.0
18.3
21.0
27.3
33.5
35.5
45.0
p. ct.
7.3
7.3
8.0
5.1
11.5
9.1
12.7
13.4
14.4
10.8
6.9
15.7
17.0
13.9
12.9
11.8
15.4
9.1
3.1
10.9
5.8
p. ct.
2.0
0.5
1.6
0.4
1.2
1.4
0.45
1.6
0.0
1.65
1.8
2.4
1.8
1.0
1.3
1.6
0.0
0.8
1.6
1.8
0.25
p. ct.
0.6
.4
.5
.6
.4
.6
.2
.2
.2
.4
1.0
1.0
1.0
1.0
1.0
.4
.4
.4
1.0
xThe first fasting day was March 4-5.
Charteris,1 in 1907, found a moderate leucocytosis, reaching 14,000
on the sixth day from 5,300, the count before the fast. He noted
further a gradual increase of the eosinophiles to 7 per cent, a condition
that had never been noted before in human blood during inanition.
His subject went 14 days without food, receiving a constant quantity of
water, 1 liter per day.
Penny's2 blood showed also a slight leucocytosis, with a return to
normal at the end. Only three counts were made, these with the
1Charteris, Record of changes observed in the blood count and in the opsonic power of a man
undergoing a prolonged fast. Lancet, 1907, 2, p. 685.
2Penny, Notes on a thirty-day fast. Brit. Med. Journ., 1909, 1, p. 1414.
140
A STUDY OF PROLONGED FASTING.
results shown in table 15. The noteworthy features of the differential
counts are the high polymorph percentage, the very marked falling-off
of the lymphocytes, and the increase in the large mononuclears.
Reyne1 could demonstrate no influence on the leucocytes in his dog
that fasted for 25 days.
Howe and Hawke2 studied two men during 7-day fasting periods
with uniform water allowance. There was an increase in the poly-
morphs at the beginning, followed by a decrease to below normal by
the end of the fast. The small lymphocytes presented the reverse
picture, while the large lymphocytes increased during the early days.
One subject showed a moderate increase in eosinophiles. The blood
of both men returned to normal after several days of diet.
Table 15. — Penny's total and differential white-cell counts.
Day.
Total.
Poly-
morphs.
Large
mono-
nuclears.
Lymph-
ocytes.
Eosino-
philes.
Twelfth
Twentieth.. . .
Thirtieth, . . ,
10,000
11,000
8,800
p. ct.
76
76
70
p. ct.
12
18
20
p. ct.
12.0
6.0
7.5
p. ct.
1.5
Even in animals where conditions can be comparatively readily con-
trolled, there has been a striking lack of harmony in the findings.
Rabbits and dogs have been the animals of most frequent choice and
Okintschitz's3 report of his work on the leucocytes of the former in
1892 is one of the earliest. He was concerned only with the differential
variations. Normally in rabbits the eosinophiles constitute about 50
per cent, lymphocytes 25 per cent, large round cells and polymorphs each
12.5 per cent of the total white-cell content. Following Professor Luk-
janow's classification of the inanition period, he divides it into four parts :
(1) the stage of indifference; (2) that of excitation; (3) of depression;
(4) paralysis of functions. The animals were allowed no water. He
found a diminution of the lymphocytes and polymorphs and an increase
in eosinophiles and large round cells. During the middle periods, the
relative diminution was not so rapid as in the first and last periods.
The lymphocytes and mononuclears showed their respective alterations
in the first period, while the polymorphs and eosinophiles were not
affected till later. The polymorphs showed the most marked diminu-
tion and on refeeding they were increased, seeming, therefore, to be the
1Reyne, quoted by E. Bardier in his article on Inanition, in Dictionnaire de Physiologie,
Charles Richet, 9, p. 99.
2Howe and Hawke, Fasting studies, No. IX. On the differential leucocyte count during pro-
longed fasting. Am. Journ. Physiol., 1912, 30, p. 174.
3Okintschitz, Ueber die Zahlenverhaltnisse verschiedener Arten weisser Blutkorperchen bei
vollstandiger Inanition undbei nachtraglicher Auffilterung. (Versuche an Kaninchen). Archiv f.
exp. Path. u. Pharm., 1892-93, 31, p. 383.
THE BLOOD. 141
form most affected by food. The disturbance in blood picture was
still evident, even when the animals had almost completely regained
their body-weight. Hayem,1 much earlier, in 1882, could see no differ-
ence in the variations of the leucocytes in the dog studied by him during
its 25-day fast, and those which occur under normal circumstances.
Argaud and Billard2 found about the same alterations in the blood
of the two rabbits they studied, as did Okintschitz. They report a
marked hypoleucocytosis with an inversion of the formula, there being
present 3 mononuclears to every polymorph. The recovery in these
animals, however, was more rapid, for in a few days the blood picture
had resumed the normal.
Kallmark3 also studied rabbits during periods of complete starvation
varying from 7 to 14 days. He noted a primary fall in lymphocytes
and polyneutrophiles, followed by a rise until the seventh day, slow in
the latter form, more rapid in the lymphocytes. In the longer fasts
this rise was followed by fluctuations, none of which, however, went as
high as the normal counts. The basophiles showed a marked rise on
the third day of the longer fasts. On refeeding, the polymorphs showed
a more rapid rise than the lymphocytes, duplicating the experience of
Okintschitz. Kallmark concludes that the lymphocytes are supplied
in greater abundance during inanition, most probably by the thymus,
which in so doing atrophies. The primary fall and the post-inanition
rise in the leucocytes occur, he believes, before compensation for the
disturbance of equilibrium has been established or the organism has
adapted itself to the altered conditions. When this has been accom-
plished, the changes in the blood are not so much different from those
noted under normal circumstances.
Rieder4 reports finding a marked hypoleucocytosis in the dogs he
studied in 1892.
Liuboumdrow5 found that the leucocytes of his 15 dogs decreased
gradually at the beginning of their fasts or until a loss of 20 per cent
body- weight was reached. A gradual rise was then noticed, except in
6 of them, frequently reaching normal. The lymphocytes showed a
diminution persisting to the end, most marked early, dropping from
15 per cent to 3 per cent or less. The monocytes reappeared to a
certain extent, after the primary fall. The polymorphs were propor-
tional throughout to the total count. Eosinophils appeared early in
those animals that did not show them before fasting, and in most cases
there was an increase of 7 to 8 times, which lasted until a loss of from
^ayem, Lecons sur les modifications du sang. Paris, 1882, p. 382.
2Argaud et Billard, Inversion de la formule leucocytaire sous l'influence de l'inanition. Compt.
rend. Soc. Biol., 1911, 70, p. 746.
'Kallmark, Zur Kenntniss des Verhaltens der weissen Blutkorperchen bei Inanition. Folia
Haemat., 1911, 11, pt. 1, p. 411.
4Rieder, Beitrage zur Kenntniss der Leukocytose u. s. w., Leipsic, 1892.
6Liuboumdrow, Changes in the blood and organs in starvation. 71 Dissert., 1893, Russian.
From the Path. Anat. Laboratory, Prof. W. Winogradow, St. Petersburg.
142 A STUDY OF PROLONGED FASTING.
10 to 30 per cent in body-weight had occurred, when they began to
diminish. By the end of the fast they had disappeared altogether.
Pol^taew1 and Reyne2 both observed great variations in the number
of white cells in the dogs they studied, the former concluding that there
was evidently a diminution in all the forms until a loss of from 30 to
40 per cent in weight and then an increase toward the end, of the
younger elements, including lymphocytes.
Uskow1 interprets these results as follows: In the beginning the
entrance of young leucocytes into the blood is retarded, as is also the
transition of the young into ripe forms. In the later period, however,
the lymph tissue, probably stimulated into increased activity by the
products of degeneration, sends more cells into the blood, and further,
there is probably a more rapid development of the young forms already
present into ripe cells.
Keuthe3 noted a decrease in polymorphs and an increase in lympho-
cytes during the first days and a reversal of this relation in the later
days of fasting.
Pashutin,4 on the other hand, concludes that the fast has practically
no effect on the leucocytes, that they show very little alteration.
Howe and Hawke5 observed the following changes in four dogs
receiving only a constant quantity of water: Three of them, fasting 117,
15, and 30 days, respectively, showed a decrease in polymorphs with
an increase in the small lymphocytes. The basophiles, eosinophiles,
and transitional forms showed no noteworthy changes. The fourth
dog, fasting for 48 days, was already anaemic. His blood presented the
reverse picture, the polymorphs increased, the lymphocytes decreased.
An early-developing eosinophilia disappeared. Two of these four
animals showed an increase in large lymphocytes, while the other two
showed a fairly constant decrease in the same variety of cell. During
later fasting periods of 15 and 30 days in these dogs, the results were
quite different, all the forms remaining practically constant, save the
large lymphocytes.
The work of Mann and Gage6 is of interest, though it is concerned with
the effects of food rather than of starvation on the morphology and stain-
ing properties of the leucocytes. They conclude that during digestion
there is a marked increase in the intensity of staining in the nuclei; the
1Pol6taew, The morphologic composition of the blood in complete and incomplete starvation
in dogs. Dissert. 97, 1894, St. Petersburg (Russian). From the Laboratory of Path. Anat.,
Prof. Uskow., Riv. internaz. d'ig., Roma, 1895, 6, p. 129. and Arch. d. sc. Biol., St. Petersburg,
1893, 2, p. 794.
2Reyne, quoted by E. Bardier in his article on Inanition, in Dictionnaire de Physiologie,
Charles Richet, 9, p. 99.
3Keuthe, Ueber die funktionelle Bedeutung der Leukocyten im Zirkulierenden Blute bei
verschiedener Ernahrung. Deutsch. med. Wochenschr., 1907, 33, p. 588.
4Pashutin, Pathological Physiology, Inanition, 1902, 2, pt. i, p. 81 (Russian).
6Howe and Hawke, Fasting studies, No. IX. On the differential leucocyte count during pro-
longed fasting. Am. Journ. Physiol., 1912, 30, p. 174.
•Mann and Gage, On the changes induced in blood by feeding, etc. Lancet, Lond., 1912, 2,
p. 1069.
THE BLOOD. 143
rim of cytoplasm in the lymphocytes becomes narrower; the granules in
the leucocytes decrease both in size and number; and the entire cell may
show a diminution in size.
Considering again the question of hibernating animals as throwing
some light on the changes during simple starvation, the work of Hanse-
mann1 is interesting. Killing the animals during their hibernating
state and examining the various organs, no evidence of cell division
could be found. He concludes that the physiologic cell division occurs
as a result of the mechanical wearing out of the tissue. If this is
eliminated, as it is under these circumstances when muscular and diges-
tive activity and the general vital processes are practically in abeyance,
there is no stimulus for cell division. The reduced activity, the absence
of intestinal digestive processes or products, and the presence of per-
verted products of parenteral digestion during fasting, no doubt would
be the important factors in influencing the blood picture.
Argaud and Billard2 examined the blood in hibernating dormice and
only a few monocytes were found after careful search, the other forms
having apparently disappeared.
Valentin, quoted by Pashutin,3 had the same experience, finding
only a few white cells. He explains their absence to the lack of lymph,
which, he claims, introduces the leucocytes into the blood.
Interesting, also, in view of the findings in some of the cases of inani-
tion quoted above, are the changes noted in bone marrow by Roger and
Josue.4 In rabbits that were completely starved for 6 or 7 days the
marrow showed the presence of many giant cells. Neutrophilic granu-
lar myelocytes predominated, though there were many polymorpho-
nuclear cells. Eosinophilic cells were rare. The fat was largely re-
placed by a granular, albuminoid substance, not mucin. On refeeding,
the eosinophiles were even less in evidence, but there were many very
large giant cells and numbers of nucleated red cells, some of them
polynuclear. Not before 24 days of feeding did the marrow return
to its normal state. This picture is not altogether consistent with
the blood findings reported, especially the eosinophilia described by
Okintschitz5 and the scarcity of polymorphs observed by Argaud and
Billard.2
As already stated, the thymus has been found to atrophy during
inanition. Kallmark6 not only noticed diminution in the size of the
1Hansemann, Ueber den Einfluss des Winterschlafes auf die Zellteilung. Archiv f. Physiol.,
1898, 5 and 6, p. 262.
2Argaud et Billard, Inversion de la formule leucocytaire sous l'influence de l'inanition. Compt.
rend. Soc. Biol., 1911, 70, p. 746.
3Pashutin, Pathological Physiology, Inanition, 1902, 2, pt. i, p. 81 (Russian).
4Roger et Josue, Des modifications histologiques de la raoelle osseuse dans l'inanition. Compt.
rend. Soc. Biol., 1900, 52, p. 417.
6Okintschitz, Ueber die Zahlenverhaltnisse verschiedener Arten weisser Blutkorperchen bei
vollstandiger Inanition und bei nachtraglicher Auffuterung (Versuche an Kaninchen). Archiv
f. exp. Path. u. Pharm., 1892-93, 31, p. 383.
"Kallmark. Zur Kenntniss des Verhaltens der wie3sen Blutkorperchen bei Inanition. Folia
Hsemat., 1911, 11, pt. 1 p. 411.
144 A STUDY OF PROLONGED FASTING.
organ, but in the number of mitotic figures, an evidence of its inactivity.
He quotes v. Friedleben as having made the same observations as early
as 1859, Hammar in 1905, and v. Jonson in 1909. These are the
only references that could be found mentioning the histological appear-
ances of the haematopoietic organs during starvation. This is a neg-
lected feature of the subject that would seem to offer a rich field for
investigation. Curran1 and Jolly and Levin2 write of the general patho-
logical changes. The latter carried out their studies on rats and describe
particularly the changes in the lymphatic tissue, essentially, atrophy,
particularly of the Malpighian bodies.
PHYSICO-CHEMICAL CHANGES.
Specific gravity. — The question of influence of food and drink and the
abstinence from them on the density of the blood has been rather
frequently the subject of investigation. We find that, as early as 1834,
Thacrah3 noted an increase in the specific gravity during hunger peri-
ods. J. Davy4 obtained the same result by depriving his subject only
of water, and Nasse,5 starving dogs but allowing water, found that a
decrease occurred in specific gravity after 3 to 4 days, but that by the
eleventh day the blood had returned to or even exceeded its normal
density.
Liuboumdrow,6 using the pyknometer, detected fluctuations in den-
sity as marked, comparatively, as those noted in the number of red
cells, but a complete agreement between specific gravity and erythro-
cyte count was not found.
Castellino,7 studying starving rabbits, found an increase in density
and at the same time a decrease in the serum content of their bloods.
Popel8 also reports an increase, though a slight one. He studied
both rabbits and dogs, using Hammerschlag's method. In the former
the increase did not exceed 1.6 per cent and it was still less in the dogs.
After ligation of the ureters, there was the slight fall of 9.11 per cent
from normal in rabbits, while the dogs showed a rise of 0.72 per cent,
a rather unexpected result, if taking only the water content of the blood
into consideration.
1Curran, The pathology of starvation, Med. Press and Circ, London, 1880, n. s., 29, pp.
210 and 229.
2Jolly et Levin, Sur les modifications histologiques de la rate a la suite du jeune. Compt. rend.
Soc. Biol. Paris, 1912, 72, p. 829.
3Thacrah, An inquiry into the nature and properties of the blood as existent in health and
disease. London, 1819-1834.
4Davy, Physiolog. and Anat. Researches. London, 1839.
Bde Martigny u. Nasse, Ueber den Einfluss der Nahrung auf das Blut. Marburg u. Leipsic, 1850.
•Liuboumdrow, Changes in the blood and organs in starvation. 71 Dissert., 1893, Russian.
From the Path. Anat. Laboratory, Prof. W. Winogradow, St. Petersburg.
7Castellino, La Suscettibilita infettiva nella inanizione lenta. Riv. d'Igiene e Sanita Pub.,
Roma, 1893, 4, No. 3, p. 461.
8Popel, Sur les variations de la densite du sang dans le jeune absolu, simple, ou complique de
la ligature des ureteres. From the Laboratory of General Pathology, Prof. S. Lukjanow, Arch, dea
eci. Biol., 1895-96, 4, p. 354.
THE BLOOD. 145
London's1 findings do not agree with those above. He also used
Hammerschlag's method, but reports a slight diminution in the rabbits
that starved for from 5 to 14 days, the average dropping from 1.048
to 1 .043. The animals in both the above series were deprived of water.
There was evidently a considerable fall in Martin's2 blood, for while
no preliminary estimation was made, on the sixth day of his fast the
specific gravity was 1.026; on the eighth it rose to 1.031, and on the
ninth and last day it had dropped to 1.021. One week after breaking
fast it was 1.043, still very low if we consider the normal to be 1.059 to
1.060.
Lloyd Jones is quoted by Lyonnet3 as finding, on the tenth day of
one of Succi's fasts, a specific gravity of 1.061 that rose to 1.063 on the
thirty-ninth day. In speaking of the influence of food and drink on
the specific gravity, Lyonnet holds that there is usually, though not
invariably, a diminution after the intake of water, the change being
but very temporary. Abstinence from all liquid causes an increase,
but not of so marked a degree as one would suppose. (In this he is
quoting Lichtheim.) Food apparently has some effect, in that after
meals there is a decrease to be found that lasts for an hour or so.
Coagulability. — Very little mention is made in the literature of the
influence of inanition on the coagulation time. Vierordt4 was among
the first to refer to this feature of the subject, having made the obser-
vation, in 1878, that an acceleration of the process occurred as a result
of starving.
Arnold6 and Collard de Martigny6'6 both noticed that the clot was
larger than usual in relation to the amount of serum, and the latter in
1850 found a decrease in fibrin content.
Jones5 also noted that the water and fibrin decreased more rapidly
than the solid constituents.
Kallmark7 noticed that in rabbits, after the fifth or sixth day of
starvation, the blood coagulated more rapidly, but no estimations of
the time are given.
Tria8 reports quite recently that he could detect very little variation
during short fasts in rabbits and dogs.
1London, Note sur la question du changement de la quantity generate et de l'alcalinite du sang
dans le jeune absolu. Arch, des Sciences Biol., 1895-96, 4, p. 516. (Abstract by Miihlmann.
See footnote 5, this page.)
2Gordon, A prolonged fast. Montreal Med. Journ., 1907, 36, p. 482.
3Lyonnet, De la density du sang, sa determination clinique, ses variations physiologiquea et
pathologiques. Paris, 1892, p. 73.
4Vierordt, Arch. d. Heilk., 1878, 14, p. 193.
'Miihlmann, Russisch Literatur tiber die Pathologie des Hungerns. Centralblatt f. allgem.
Path., 1899, 10, p. 160.
6de Martigny u. Nasse, Ueber den Einfluss der Nahrung auf das Blut. Marburg u. Leipsic, 1850.
7Kallmark, Zur Kenntniss des Verhaltens der weissen Blutkorperchen bei Inanition. Folia
Hsemat., 1911, 11, pt. 1, p. 411.
8Tria, Propriet6s chimico-physiques du sang durant l'inanition. Archiv. ital. de biol., Pise,
1911, 55, p. 49. (Arch, di farmacol. sper., Roma, 1909, 8, p. 359.)
146 A STUDY OF PROLONGED FASTING.
Valentin1 noticed a marked retardation of coagulation in hibernating
animals.
No report of the specific estimation of coagulation time during fast-
ing in man could be found. Dr. Wile2 reports that on both examina-
tions of Gayer's blood, made on the eighteenth and thirtieth days,
there was apparent decrease in the platelets, but that coagulation was
accelerated. He says, of the last examination, that "the blood was
thick, dark red, and did not flow easily." Aside from such general
conclusions without data to demonstrate them, the only clue as to what
might be expected in man are a few observations that have been made
relative to meal times.
Coleman3 found the longest coagulation time an hour after the
principal meal and the shortest before breakfast.
Cohen,4 using the method devised by himself, determined that the
average time before meals was 7-g- minutes and after meals 9 minutes,
while Mercier, quoted by Cohen in the above article, constantly found
the coagulation more rapid after meals than before, and Addis5 claims
that food has no influence on the process.
Cohen4 quotes A. E. Wright as crediting fluids with a greater influ-
ence on coagulability of the blood than food, but that hunger does
retard the process, a view not upheld by the observations of Coleman
and Cohen. Increased consumption of liquids lengthens the time and
withholding them has the opposite effect.
Immunity. — There have been a few studies made of the effect of
starvation on immunity in general and the immune body-content of
the blood specifically, but the data are scarcely sufficient to warrant
definite conclusions.
In 1890 Canalis and Morpurgo6 studied the effect on the natural
immunity pigeons exhibit toward anthrax. They were found con-
stantly to lose this resistance if the fast were begun immediately after
the injection of the organisms, or a day or so before. They regained
it, however, on refeeding, if the inanition period had not been too long.
This same natural immunity possessed by chickens7 was not lost
unless they were starved for more than 8 days. If starved before
inoculation they proved more susceptible. These workers were unable
to make rats susceptible to anthrax by starving.
Valentin, Repel, f. Anat. u. Physiol., 1838, 3, p. 156.
2Gayer's fast : A private communication from Dr. Wile, of New York City.
'Coleman, The coagulation of the blood and the effects of certain drugs upon it. Biochem.
Journ., 1906-7, 2, p. 184.
4Cohen, Coagulation time of the blood as affected by various conditions. Arch. Int. Med.,
1911. 8, pp. 684 and 820.
6Addis, The coagulation time of the blood in man. Quart. Journ. Exp. Physiol., 1908, 1 , p. 305.
6Canalis and Morpurgo, Ueber den Einfiuss des Hungers auf die Empf anglichkeit fur Inf ections-
krankheiten. Fortschr. der Medicin, 1890, 8, p. 693.
7Canalis and Morpurgo, ibid., 8, p. 729.
THE BLOOD. 147
P. Castellino1 concluded, from his studies on rabbits in 1893, that
there was a diminution in resistance to infection.
A few years later, in 1899, Meltzer and Norris2 could demonstrate no
difference in the bactericidal action against the typhoid bacillus of the
blood of starved, under- or overfed dogs.
Roger and Josu6,3 on the other hand, having observed an increase
in the resistance to the colon bacillus in fasting rabbits, suggest that
some possible benefit may be derived from fasts. This is the only bit
of experimental evidence — with reference to the blood, at least — that
speaks for the value or advisability of this procedure as a general thera-
peutic measure. This increase in resistance they attribute to hyper-
activity of the bone marrow, whereby there is a more rapid proliferation
of the defensive cells.
Charteris4 noticed a wide daily variation in the opsonic index of the
blood of his human subject during the latter's 14-day fast, but, as he
obtained a similar result with his own blood, he was led to conclude
that the changes during fasting were not significant, due, rather, to the
use each day of a fresh emulsion of bacteria. Martin's5 blood, however,
showed a gradual lowering of the index, returning to normal 4 days after
the fast was broken. In this case the Staphylococcus aureus was used.
Bizzozero6 studied the natural hemolytic power of the blood serum
of 8 chickens that starved for from 8 to 17 days and could find practically
no alteration. He concludes that the hemolysins are not concerned in
the defense of the organism against bacterial invasion, because, as we
have seen from the work of Canalis and Morpurgo, starving does lower
the resistance to infection.
Among the studies on other properties of the blood are to be men-
tioned those of Tria,7 on the viscosity and electro-conductivity in rab-
bits and dogs. He found little alteration, some decrease, in both early.
He concludes from his entire study that the body is able to compensate
pretty well for the disturbances in nutrition, thus permitting of long
fasts without serious consequences. The investigations of Deter-
mann,8 and of Maran6n and Saristan9 dealt especially with the viscosity.
Castellino, La Suscettibilita infettiva nella inanizione lenta. Riv. d'Igiene e Sanita Pub.,
Roma, 1893, 4, No. 3, p. 461.
2Meltzer and Norris, On the influence of fasting upon the bactericidal action of the blood.
Journ. Exp. Med., 1899, 4, p. 131.
3Roger et Josue, Influence de l'inanition sur la resistance a l'infection colibacillaire. Compt.
rend. Soc. Biol., 1900, 52, p. 696.
4Charteris, Record of changes observed in the blood count and in the opsonic power of a man
undergoing a prolonged fast. Lancet, 1907, 2, p. 685.
6Gordon, A prolonged fast. Montreal Med. Journ., 1907, 36, p. 482.
"Bizzozero, Pouvoir hemolytique naturel, pulet dans l'inanition aigu6. Arch. ital. de biol.,
Turin, 1904-5, 42, p. 212.
7Tria, Proprietes chimico-physiques du sang durant l'inanition. Archiv. ital. de biol., Pise,
1911, 55, p. 49. (Archiv. di farmacol. sper., Roma, 1909, 8, p. 359.)
8Determann, Die Beziehung der Viskositat des Blutes zu den Korperfunktionen. Veroffentl.
d. balneol. Gesellsch. in Berl., Berlin u. Wien, 1910, pp. 259 and 270.
9Marafi6n y Saristan, La viscosidad de la sangre humana en varios estados patal6gicos. Rev .
Ibero Am. de cien. m£d., Madrid, 1911, 26, p. 244.
148 A STUDY OF PROLONGED FASTING.
A decrease in alkalescence was noticed by Tauszk1 in Succi's blood,
by Castellino2 in rabbits, and Benedict reports the same change in his
subject.3 A very moderate decrease was also observed by London4 in
his eight rabbits. Castellino2 found also a decrease in NaCl content
and in the bulk of serum.
For additional data as to the effects of inanition on the physico-chem-
ical properties, reference can be made to the work of Githens,6 Schce-
neich,6 Fria,7 Lattes,8 Robertson,9 Bierry and Fandard,10 Daddi,11 Moroz-
off,12 and Weber13 (who includes an exhaustive correlation of references
to the literature of the entire subject of inanition). Manca14 and Mac-
alum15 confined their investigations to the cold-blooded animals.
OBSERVATIONS ON L.*S BLOOD.
There is little danger of one's opinions being biased by the diverse
results above correlated. We can therefore take up the consideration
of our subject's blood either with an open mind free from preconceived
ideas, or with confused expectations, ranging from absolutely negative
findings to very grave disturbances, with the confidence that we have
precedent for almost any picture that may present itself. The coagu-
lation time and specific gravity were investigated, but the examina-
tions were concerned principally with the red and white cell and haemo-
globin content, the technique for which follows. That employed in the
tauszk, Jahrsb. tiber d. Fortschr. der Thier-Chemie, 1894, 24, p. 147, abstracted from Orvosi
hetilap, Budapest, 1894, p. 512 : also Hamatologische Untersuchungen am hungernden Menschen.
Wien. klin. Rundschau, 1896, 10, p. 306.
2Castellino, La suscettibilita infettiva nella inanizione lenta. Riv. d'Igiene e Sanita Pub.,
Roma, 1893, 4, No. 3, p. 461.
3Benedict, Carnegie Inst. Wash. Pub. 77, 1907, p. 322.
4London, Note sur la question du changement de la quantity g£nerale et de Palcalanite du sang
dans le jeune absolu. Arch, des Sciences Biol., 1895-96,4, p. 516. (Abstract by Miihlmann. See
footnote 5, p. 145.)
KJithens, Influence of hunger and haemorrhage on the composition of the blood plasma. Proc.
Phila. Count. Med. Soc, Philadelphia, 1904-5, 25, p. 279.
6Schosneich, Beschaffenheit des Blutes unter verschiedenen Bedingungen. Ztschr. f. exp. Path,
u. Therap., 1905, 2, p. 419.
7Fria, Alcune ricerche comparative sul sangue di animali nutriti naturalmente ed innaturalmente.
Folia clin., chim. et micros., Salsomaggiore, 1910-11, 3, p. 44.
8Lattes, Ueber den Fettgehalt des Blutes des Hundes unter normalen u. unter verschiedenen
experimentellen Verhaltnissen (Verdauung, Hungern, etc.). Arch. f. exp. Path. u. Pharmacol.,
Leipzig, 1911, 66, p. 132.
9Robertson, Studies in the blood relationship of animals, etc. 1. A comparison of the sera of
the horse, rabbit, rat, and ox with respect to their content of various proteins in the normal
and in the fasting condition. Journ. Biol. Chem., Baltimore, 1912, 13, p. 325.
10Bierry et Fandard, Variations de la glycemie pendant l'inanition. Compt. rend. Acad. d. sc,
Paris, 1913, 156, p. 2010.
uDaddi, Surles modifications du poids de l'extrait 6th6r6 du sang durant le jeftne de longue
dur6e. Arch. ital. de Biol., Turin, 1898, 30, p. 439; also Sulle modificazioni del peso dell estratto
estereo del sangue durante il digiuno di lunga durata. (Sperimental.) Arch, di biol., Firenze,
1898, 52, p. 43.
^Morozoff , On the effect of fasting for a short time on the morphologic composition of the blood.
(Russian) Vrach. St. Petersburg, 1897, 18, p. 1081.
13Weber, Ueber Hungerstoffwechsel. Ergebnisse der Physiologie (Biochemie), 1902, 1 Abt.,
p. 702.,
14Manca, Le cours de l'inanition chez les animaux a sang froid. Arch. ital. de biol., Turin,
1895, 23, p. 243, and 1896, 25, p. 299; also Chemical researches on animals (cold-blooded) during
inanition (Italian). Arch. ital. de biol., Turin, 1903, 39, p. 193.
15Macalum, The inorganic composition of blood plasma in the frog after a long period of inani-
tion. Rep. Brit. Assoc. Adv. Sc, London, 1911, 80, p. 766.
THE BLOOD. 149
coagulation and specific-gravity estimations is given under these head-
ings. The attempt was made to determine the opsonic index, but the
subject was so far from an incubator and centrifuge that it was found
impossible to obtain accurate results. There was a possibility, also, of
this additional manipulation having a disturbing influence on the physi-
ological investigations, so this feature was abandoned, though still recog-
nized as one of the most important lessons to be learned from the blood.
Technique. — For the three days just preceding the fast, Levanzin's
blood was examined to determine a normal picture with which to
compare the results later obtained. During the fast, with the excep-
tion of the first day and three days scattered through the period,
daily examinations were made and also on the first and third days of
refeeding. The time of the day did not vary more than half an hour
throughout, all the specimens being obtained between 10 and 10h 30m
a. m., so that there was a constant relation to the general routine of
the subject's daily activities — that is, immediately after he had finished
with the respiration experiment, had been weighed, had washed his
face and hands, and climbed the short flight of stairs to his balcony.
It was his habit to take about half a glass of water before submitting
to the lancet prick. (It may be well to mention here that, for the first
10 days of the experiment, the subject received the constant quan-
tity of 750 c.c. of distilled water per day, and thereafter, 900 c.c.
per day.) The specimens of blood were obtained from alternate
fingers of the left hand and occasionally from the ear. Deep pricks
were made, so as to obtain sufficient blood without squeezing. For
counting the red and white cells the Thoma-Zeiss apparatus was
used, diluting with fresh salt solution for the former and 1 per cent
acetic-acid solution tinged with gentian- violet for the leucocytes. The
usual precautions were taken to insure uniform suspension of corpuscles
and the even filling of the counting chamber. In the case of the ery-
throcytes, 80 small squares were counted and 4 ciphers added to the
total. The average of two or more such figures was taken as the final
result. In estimating the leucocytes, the whole cross-ruled field was
counted and the result multiplied by 200. The average of three or
more such figures was taken as the final estimation. The use of gentian-
violet in the acetic-acid solution makes the counting of the leucocytes
much easier and the likelihood of mistaking foreign particles for cells
practically impossible. The smears for the differential counts were
stained by the Wright method; 200 or more cells were examined
by the use of the tV objective and No. 1 eye-piece, and classified as fol-
lows : Polymorphonuclear neutrophile, eosinophile, and basophile ; small
and large lymphocyte ; monocyte ; transitional cell. As the classification
of the last three forms is so mooted a question, it will be necessary to
go into some detail as to just what cells were placed under these heads.
Under large lymphocyte was classified the mononuclear cell, con-
siderably larger than the red corpuscle, with a round or typically
150 A STUDY OF PROLONGED FASTING.
indented nucleus (like that of the small lymphocyte though not staining
as deeply), small amount of cytoplasm in proportion to nucleus, the
former not being markedly basophilic and usually containing a few
faintly-staining granules. The mononuclear was considered the cell
whose nucleus was more indented, the proportion of cytoplasm was
greater and granules more evident, the latter basophilic and the whole
cell staining more deeply than the large lymphocyte. The occasional
large mononuclear cell of the endothelial type was also counted in with
these cells, though not considered as being associated with them generi-
cally. The transitional cell showed a pale-staining, usually kidney or
horse-shoe shaped nucleus, surrounded by cytoplasm, pale and free from
granules. This classification follows generally that of Pappenheim.1
In estimating the haemoglobin percentage the Tallqvist2 scale was
used throughout the series. This method, it is true, is open to some
criticism in that it is scarcely possible to detect differences of less than
3 per cent. My experiences, however, have developed a confidence in
it that has been justified by comparative readings with other methods.
In the present case, from the nineteenth to the twenty-fourth day,
estimations were made with the Sahli apparatus and the results were
practically the same as those obtained for the same days with the
Tallqvist scale. Care must be taken to follow exactly the same tech-
nique for each reading, particularly in the matter of light; the results
will then be relatively correct, even if the method is comparatively
weaker than some of the others.
The results of these examinations are contained in table 16. They are
correlated and presented more graphically by curves given in figures
20 and 21 . No. I shows the relation of haemoglobin to red cells, and No.
Ill the relation of the total white-cell count to the differential. To avoid
confusion of curves the transitional, eosinophile, and basophile are
plotted separately in No. IV and the scale enlarged. In No. II the
composite curve of the polynuclears, that is, neutrophiles, basophiles,
and eosinophiles and one of the mononuclear cells, large and small
lymphocytes, transitionals, and monocytes, are given for comparison.
Erythrocytes. — The subject's normal count apparently was high; the
three preliminary estimates range well above 6,000,000. It maintained
this high figure throughout the test, going below it on only two occasions,
the tenth day of the fast and the third day following. In the early part
of the fast there is daily variation, ranging under 1,000,000. This
becomes less evident toward the end. The general impression given
by the curve is that of a very moderate decrease. There were no alter-
ations in the characteristics of the individual cells as to size, shape, and
staining properties and no nucleated red cells were found at any time.
1Pappenheim u. Ferrata, Ueber die verschiedenen lymphoiden Zellformen des normalen und
pathologischen Blutes. Folia Haemat., 1910, 10, p. 78.
2Tallqvist, Ueber die Anwendung des Filtrirpapiers in Dienst der praktischen Hsematologie,
Berl. klin. Wochenschr., 1904, 41, p. 926.
THE BLOOD.
151
Haemoglobin. — The haemoglobin average ranges rather consistently-
above 85 per cent, the most marked variations being consistent with
those of the erythrocytes. The low period is between the tenth and
sixteenth days, from then on showing a very moderate rise.
Table 16. — Levanzin's cell counts and haemoglobin percentage.
Days.
Haemo-
globin.
Total
erythro-
cytes.
Total
leuco-
cytes.
1
5
p^
gjj
>>
a
CO
PQ
\_
Days before fasting:
3d
2d
1st
Days of fasting:
2d
3d
4th
6th
6th
__7th
9th
10th
11th
12th
13th
14th
16th
17th
18th
19th
20th
21st
22d
23d
24th
, ^25th
^ 27th
28th
29th
30th
31st
Days of diet:
1st
2d
3d
p. ct.
85
85
95
90
92
92
91
90
88
90
88
85
85
85
87
85
88
87
88
90
88
87
90
90
92
93
92
92
8,984,000
6,632,000
7,000,000
6,100,000
7,200,000
6,120,000
7,170,000
6,250,000
6,450,000
7,000,000
5,950,000
6,750,000
6,480,000
6,580,000
6,280,000
6,010,000
6,600,000
6,700,000
6,250,000
6,250,000
6,450,000
6,130,000
6,630,000
6,000,000
6,250,000
6,240,000
6,350,000
6,190,000
6,050,000
6,170,000
5,900
6,400
6,000
8,400
12,400
8,400
9,400
8,600
7,000
8,600
7,000
8,400
7,000
6,800
7,900
9,000
6,200
6,600
6,200
6,400
6,400
6,100
6,600
6,900
6,500
5,800
7,800
6,000
6,000
8,000
p. ct.
67.5
66.0
59.5
73.0
79.0
66.0
68.0
67.5
60.0
66.0
68.0
68.0
74.0
64.0
64.0
72.5
64.5
65.5
62.5
61.0
62.5
61.0
58.5
65.0
65.5
63.0
61.5
63.5
64.5
60.0
6,280,000 7,200 70.0
Too ill for examination.
5,960,000 I 6,600 | 60.5
p. ct.
23.5
22.0
32.0
20.0
16.5
24.0
26.5
19.0
24.0
26.0
22.0
23.0
15.0
24.5
26.5
16.5
29.0
24.0
22.5
28.5
20.5
26.5
28.0
25.0
27.0
26.0
30.0
27.5
26.5
32.0
21.5
32.5
p. ct.
2.5
1.5
3.0
3.0
.6
4.0
2.0
4.5
4.0
3.6
3.0
2.0
2.0
7.6
4.0
3.5
3.0
6.5
5.5
3.0
5.0
3.6
4.5
2.5
1.5
3.0
1.0
1.5
2.0
2.5
2.5
3.0
p. ct.
4.5
7.5
4.0
3.0
3.0
4.0
3.0
3.0
3.5
1.0
3.0
5.0
4.5
1.0
2.6
1.5
.5
2.0
3.6
3.5
6.5
6.5
4.5
4.5
6.0
6.5
5.5
5.0
5.0
5.0
4.5
3.0
p. ct.
1.5
2.0
0
1.5
0
p. ct.
0.5
.5
1.5
. 5
.5
.6
0
.5
2.0
.5
1.0
0
1.5
.6
.5
.5
0
1.0
.5
1.0
1.5
0
.6
.6
0
0
0
0
.5
0
0
.5
p. ct.
0
.5
0
0
0
0
0
.6
0
0
.6
0
0
.6
0
.6
0
0
.5
.6
0
0
1.0
0
0
0
0
.5
0
0
0
0
Leucocytes. — It is in the total leucocyte count that we have the most
striking change of the whole series of examinations. There was a rapid
rise at the onset of the fast, reaching 12,400 on the third day. On the
fourth day, however, it immediately fell to 8,400, after which there was
a consistent daily variation of about 1,000 until the sixteenth day,
when it reached approximately the preliminary count, after which there
will be noticed a more marked daily variation. It was only possible
to continue the examinations for 3 days following the fasting period,
so that the count had not settled down to normal when the subject
went out from under observation.
152
A STUDY OF PROLONGED FASTING.
APRIL MAY
II 12 1314-15 16 17 1819 2021 22 23242S 2627 282930 I 2 3456 7 S 9 101 1 12 I3I4I5I6I7
DAY 3
2
1
I \Z
3
4
5
6
7
S
9
10
II
12
13
4
5
6
7
8
19,2021
222324
25
26
2282919
31
112 3
7,000,000
l
/
i
/
re
/
\
/
/
,
ERYTHROCYTES
/
v
/
\
/
\
\
-fc
/
\
\
f
'
/
\
^
S
/
\
/
IE
/
\
/
/
\
/
s
N
/
\
7\
\
\
■*
s
X
6<000,000
i \
1
1
N
'
^
\
/
-->
^
'\
06*
/
v:
/
x
■~s
:^!
p
/
S
\
S
s
r
HEMOGLOBIN
1
i
^
r"
^
1
65%
i
\
/
«v
i
/
BE
FC
Rl\
IN
A
in
10
N
PC
R1C
D
+01ET
DAY 3
?
1
HISSSSZ
3Si5u.I2i3!4i5
gi7igi9 2S
21222324
jiESSI
3303! M2 3
1 -t
9,000
T i
1 1
JL i
1
1 1
1
8,000
1 1
1
1 1
7,000
1
1
1
1
J
1 A
4 I
6,000
tu tX
tT I
T7 t^
M-
J^L /
4t L.
4t4A t
J- t
txt-Z t
5,000
x i
AL X—T
1
i 4
1 XI
- ZK
x H
t At
j[
- 44A
4^ H
r -*f-
_ _j
V 4
, 4^V
POLY NUCLEAR
M
ft
,_ 7
V 4-
\ t ! A
4,000
V
tt. it
Z!SI
21 £
A-t-
t !
tt
~xu
VL-
-
1
i
\
3.000
- Jl '
4\
- 4W--
7 rr is
2
h 44^-
V"7r /
I.
k ft^
J£-V4
^TX-
V V
2.000
W
- r
-¥-
MONO NUCLEAR
T
1,000
3£F0
!Ef
Fig. 20. — Chart I. Relation of haemoglobin to erythrocytes.
Chart II. Composite curve of the polynuclears
compared with one of mononuclears.
THE BLOOD.
153
APRIL MAY
II 12 13 14- 15 16 17 18 I9Z02I 222334252627282930 I 2 3 4 5 6 7 8 9 10 II l2 1314 15 16 17
DAY 3i
J I
ii2 34:
5 6 7 8910,
1 1 12 13
141516171819$
2i222J
24
25
26272
8 29303
1 112 3
1 2,000
1
1
1
1
1
1
1 1,000
1 T"
1 -i-
' -
1 1
lO.ooo.
1
1
1
1
1
jl i
9,000 .
1 II
tX
1 4U
.%
/
1 111
j
/
1 tV-
J
i
7
j
/
8,000
Ml".
- 1- tl
n
mX -
- 4 LI
/
1 1
nrt :
- 44 4
/ \
1
\l
it -
- - J 4
/ \
7
1
v
it--
- t J
/ \
r
V
7,000 .
±t
r
'
t
k
j~ ...
s
/
,
1
i s
jt ..
/
\
i --
JE
i
A ,
/
1
A
tr j.
t -
/
/ \/
\
TOTAL 6,000 1
\
_T J
i
/
A^
V.
LEU KOCYT ES — * -
i /
■j -
;
\
ZlI
/ i
it .
- ^c
— IL-
'
it .
- - tx
\
5,000 .
(
- x t\
/ I
J
i .
- 44 J
/ \
44
\
A
i- -
- -4 -1
' \
I
■JU -
i \
jr .
- t
\
£-
A
'\
IE .
. t _
a
/
\
\ .
i V
4.000 1
\
tr .
t V
/
V
\ /
i >
POLY NEUTRO-"*-
\
ft .
^
A.
f
VJ
i '
1
□C -
^7*
\f
3.000 -
iu. _,
ji_
ct
2.000 .
i ~Z.
-C Zw
/\ i
'\' /
"4 /
-A v \
A
7^
1
\ /
v /
/
-J Z. 4
f i
L.
V,
/N> /
^ /
■^
*>J
-42
SMALL LYMPH.
/
\ /
\'
V
\/
^v.'
V
\/
1,000.
V
LARGE MONO-n. ,
K
-, — -*
t
LARGE. LYMPH-J.
\
mkrfl'
-
" — h
7<>
S"
i
o:
~K
1 "x'
~-f\
*,*-*"
_a»c
*—- '
' _r~*
DAY 3
> 1
1 12 34 J
> 6 7 8 9 10
II 12 13
14 1
3161
7*1819 20
2J2223
?A
?5
2627?
129303
1112 3
- Jti-
-Z-t
j
4-00.
--1 4
/
-4 4
/
\/
-4 4
/
V
"t n
/
-4 L
300.
-4 t
1
-t 4
/
-t 4
1 li
r i
;
1
\
200.
: l
t
A
31
t
/
\
, i
1 \
z:
-i-4
/
1 A
/ \
t-
1
1 J
r !J "i
/ y
r~
\
\
(\
100 ,
i 7
1
\
\
/
\
TRANSITIONAL I
1
, i '
i
\
,
\
y
A
y
X /
i r
H
cX.
/
\
/\
EOSIN^.
..J- 'r
TT't1
..
/
VI \
BASO-s ;
i\ \
"A V
/I
ti-v~
s • /
\-
1 V
o\
^=-^
nIV
c;
1
O
VJP
\
£N
•,..i-- \
Fig. 21. — Charts III and IV. Relation of total to differential leucocyte counts.
154 A STUDY OF PROLONGED FASTING.
Polymorphonuclear neutrophiles. — These ran throughout rather con-
sistently with the total count and the marked variations in this latter
were quite apparently due to the change in polynuclear content.
Small lymphocytes. — There are fluctuations in the count of this cell,
except during the period between the fourteenth and twenty-seventh
days, when they were comparatively constant, as were also the total and
polymorpho-neutrophile counts. It will be noted that the majority of
the rises and falls are the opposite of those seen in total and polymor-
pho-neutrophile curves. This is particularly the case on the second,
fourth, thirteenth, fourteenth, sixteenth, seventeenth, twentieth, and
twenty-first days of fast and the first day of refeeding. It seems safe
to conclude, therefore, that these fluctuations are only relative, due
really to the fluctuations in polymorphs, and that their number was
practically constant throughout the fast.
The other forms of leucocytes present no distinctive features except the
transitional, which was subject to several rises, namely, on the seventh,
sixteenth, nineteenth, and twenty-first days, when they were above 5
per cent. Only an occasional eosinophile was found during the last
10 days of the fast and they had not returned to their usual number
when examinations were discontinued. By examination of chart No.
11 (figure 20), there appears to be a very slight increase in the combined
mononuclear cells throughout the fasting period, the average being
raised by the fluctuations in the transitional form, for the other types,
i. e., monocyte, large and small lymphocyte are practically constant
throughout, except the variations already noted in the latter.
Coagulation time. — Toward the end of the second week of inanition
it was noticed that the blood coagulated more rapidly than it had during
the earlier days. This became more noticeable each day, so that if the
mixing pipettes were not filled very rapidly the drop would coagulate
or the blood would clot in the tubes. It is certain that this was not due
to any physical alteration of the patient's environment. The tempera-
ture of the balcony where the subject stayed and where the estimations
were made was practically constant. This is a very important factor,
for most experimental evidence goes to show that variations in temper-
ature have decided influences on the coagulation time. Addis, Fox,
and Wright, quoted by Cohen,1 and the latter himself, all showed that
rise in temperature accelerates and cold retards the process. Hartmann2
also notes that the higher the temperature the shorter the coagulation
time, and Rudolf,3 determining the effect more specifically, states that
in general each degree of rise and fall between 15° and 20° C. decreased
and increased, respectively, the time one minute.
On the seventeenth day the estimations of the coagulation time were
begun. Until the twenty-fifth day the McGowan4 method was used.
1Cohen, Coagulation time of the blood as affected by various conditions. Arch. Int. Med.,
1911, 8, pp. 684 and 820.
2Hartmann, Zur Frage der Blutgerinnungszeit. Munch, med. Wochenschr., 1909, 56, p. 796.
"Rudolf, Tr. Assoc. Am. Phys., Philadelphia, 1910, 25, p. 504.
4McGowan, A clinical method for estimating the coagulation time of the blood. Brit. Med.
Journ., 1907, 2, p. 1580.
THE BLOOD.
155
Capillary tubes of uniform caliber were filled with blood escaping after
the specimens for the other examinations had been obtained, practically
always the same relative drop, the third. Small sections were broken
off at intervals of from 10 to 30 seconds. When a fine filament of
fibrin was observed between the carefully separated ends of the tube
and fragment, the time elapsed since the appearance of the drop used
was taken as the reading. For the remainder of the examinations, the
Boggs1 apparatus was used. This consists of a truncated cone of
glass that sets into a chamber into the side of which a small metal tube
is inserted, connected with a rubber bulb, in such a way that a stream
of air can be directed against the blood that is placed on the polished
undersurface of the cone. As soon as coagulation has occurred in the
drop, the mass of corpuscles that moved readily in one direction under
agitation of the air-current merely vacillate. This method is more
accurate than the first procedure, but is open, as are all the other
devices for measuring coagulation time, to sources of error that, if not
avoided, will cause wide variations in results. This refers particularly
to the matter of temperature, the particular drop used, and the pres-
ence of foreign particles in the blood or on the receiving surface of the
cone, such as hair or lint. The first drop appearing after the prick will
clot much slower than the subsequent ones, when the platelets will have
accumulated about the edge of the wound. It is necessary, therefore, to
use the same relative drop on each examination, preferably the second,
though, as has been stated, in this case the third was used.
Table 17. — Coagulation limes of Levanzin and a control.
Day.
Levanzin.
J. F.
Remarks.
17th
1' 5"
0' 3"
McGowan method.
18th
1 20
2 50
Do.
19th
1 20
2 00
Do.
20th
1 5
Do.
21st
0 55
1 20
1 5
1 5
1 55
1 35
4 20
Do.
Do.
Do.
Boggs used on control.
22d
23d
24th
25th
4 50
From this day on, Boggs
27th
3 20
4 +
used on both.
28th
2 50
29th
2 20
30th
1 50
3 30
31st
2 30
2 45
6 20
First diet
Note: '= minutes; " = seconds.
The figures as obtained by these two methods, together with those
found in a normal individual examined on the same days with the same
methods and under approximately the same conditions, are collected in
table 17. The last record on the control was made in a temperature at
'Boggs, Johns Hopkins Hosp. Bull., Baltimore, 1904, 15, p. 174.
156 A STUDY OF PROLONGED FASTING.
most 2° C. cooler than that of the subject's environment, but a deduc-
tion of 2 minutes, as correction for this, still leaves the time distinctly
longer than that obtained on the same day in Levanzin's blood. A
comparison of the two series of results will demonstrate a distinct
increase in the coagulability of the starving man's blood, more notice-
able toward the end of his fast.
Specific gravity. — The specific gravity was determined only twice,
while the subject was eating his first meal after fasting and on the
third day of refeeding. The first time it was 1.0612, the second it was
1.0618. The estimations were made by the Hammerschlag1 method,
the specific gravity of the mixture of chloroform and benzol being
determined by the pyknometer. As no figures were obtained, either
before or during the fast, these two examinations are of little value,
except that from them it may be assumed there is a very slight in-
crease in density, taking 1.059 to 1.060 as the average normal specific
gravity.
DISCUSSION AND CONCLUSIONS.
The results of the above studies are conspicuous rather from the
absence than the presence of striking alterations in the blood picture.
Really the only prominent features are the early rise in polymorpho-
nuclear neutrophiles and the decrease in coagulation time. The leuco-
cytes, i. e., the neutrophiles, at least, are the most sensitive of the blood-
cells to changes in body conditions, and we know that apparently slight
disturbances will call forth a recognizable increase in these cells — a cold
bath, for example. They seem always to be on the alert, ready at the
least evidence of disturbance to rush forth in defense of the organism.
It is scarcely to be wondered at, therefore, that in response to such an
unusual condition as starvation there should be an outpouring of the
reserve supply, at least for a day or so, or until the organism has had an
opportunity to adapt itself to the altered conditions. The variations
in water-content of the blood can not be considered a factor in this rise,
involving as this does only the one form of cell. The only explanation
that suggests itself, therefore, and frankly not a particularly scientific
one, is this alertness of the polymorpho-neutrophile and its ever-readiness
to be on the defense for the organism. The products of the somewhat
perverted metabolism may excite them into this early activity and
later fail to do so, but there is no evidence to prove this supposition.
It is not easy to understand why they should respond to toxic products in
the early days and not during the later as well, unless we assume they
acquire a tolerance for them, which seems improbable when we compare
the reaction to infections, in which their fight is evident throughout the
disease if the organism is to conquer. It is further possible that an insig-
nificant, obscure source of bacterial infection happened to develop at
this particular time; if so, there was no other evidence.
hammerschlag, Wien. klin. Wochenschr., 1890, 3, p. 1018.
THE BLOOD. 157
As to the effect of the starving on the total quantity of blood, it does
seem evident that there are fluctuations, at least in the first two weeks.
By comparing the curves of the white- and red-cell counts it will be noted
that the variations are synchronous; that on the third, fifth, ninth, and
eleventh days particularly the noticeable increases in the one are accom-
panied by equally frank rises in the other. It would seem that this
could only be due to variations in water-content. The specific gravity
would have gone far toward proving this point, but unfortunately this
was not determined during this period. Taking these fluctuations as
indication of variations in water-content, it appears that during the mid-
dle period of the fast, at least, the equilibrium of intake (including that
drawn from tissues) and output was pretty well established. The last
counts made, namely, on the third day after the fast was broken, are
considerably lower than those of the last day of fast. This no doubt is
a relative decrease, due to increase in water-content. While the diet
was of course limited, there was an increase in the intake of fluids.
The haemoglobin appears to be particularly resistant, the percentage
on the last day being within 2 per cent of the highest estimation,
found on the day before fast began, though there was a moderate
decrease during the second 10 days.
It is difficult to account for the only other marked change in the
blood — the acceleration of coagulation. Loss of water could be respon-
sible, but there is no evidence that this occurred. The very slight
increase in density at the last, determined by the specific gravity, would
certainly not demonstrate a sufficient concentration. There were no
estimations of the platelet content made, but it is possible that the
explanation lies with these. An increase in them could be responsible.
The final conclusions as to the effects of uncomplicated starvation on
the blood to be drawn from the results of examinations on Levanzin are:
1. There is a slight actual loss in haemoglobin, more marked during
the second 10 days.
2. There are moderate fluctuations in water-content, particularly
during the first half of the period, and an increase after breaking fast,
evident till after the third day, at least.
3. There is a decided rise in polymorpho-neutrophiles in the early
days.
4. There is an increase in coagulability, especially after the first two
weeks.
5. In an otherwise normal individual, whose mental and physical
activities are restricted, the blood as a whole is able to withstand the
effects of complete abstinence from food for a period of at least 31 days,
without displaying any essentially pathological change.
MECHANICS OF RESPIRATION.
A physiological study of the human body during a prolonged fast
would be incomplete without a careful investigation of the influence of
inanition upon the mechanics of respiration. Fortunately, it was pos-
sible to obtain such data in all of the experiments with the respiration
apparatus, as the spirometer gave a graphic record of the respiration,1
from which accurate data regarding the respiration-rate, the ventilation
of the lungs per minute, and the volume of air per inspiration could be
obtained. Such data are available for the morning respiration experi-
ment for every day of the fast, for the experiments made in the evening
before the subject entered the bed calorimeter, for experiments made
on several occasions when the subject was sitting quietly or sitting
writing, and also for experiments in which he breathed an oxygen-rich
atmosphere.
TYPICAL GRAPHIC RECORDS OF RESPIRATION.
The graphic records obtained by the spirometer method have a
special interest in connection with the fasting experiment in that they
indicate the character and rate of the respiration as the fast progressed.
Out of 200 or more records obtained with this subject, four typical
curves have been selected for reproduction in figure 22, i. e., one each
for April 17 and April 30, and two for May 14, 1912. From these
curves it will be seen that at each inspiration the pointer on the spiro-
meter rises and at each expiration falls. The experiments were so
conducted that the communication between the subject and the spiro-
meter was made at exactly the end of a normal expiration; consequently
the first deviation from the straight line is that due to an inspiration.
Similarly, at the end of the experiment the communication with the
spirometer was cut off at the exact end of the normal expiration.
From this record the respiration-rate can easily be counted.
Immediately below the record of the respiration is the line showing
the time in minutes; the lowest line indicates the number of revolutions
of the recording device — the so-called work-adder wheel — from which
the total volume of ventilation is calculated. Since a record of the
muscular activity is essential for all intelligent comparison of the results
of respiration experiments, a method was followed similar to that used
for the bed calorimeter, the bed upon which the subject lay being
provided with a pneumograph, tambour, and pointer, by which a record
of the degree of muscular repose was obtained. This record is shown
in the line directly above the respiration curve. Frequent testing has
shown that this form of bed2 is extremely sensitive.
1See description and schematic outline of spirometer on p. 317 and figure 40.
2See arrangement of bed inside respiration calorimeter, figure 37, page 312.
158
MECHANICS OF RESPIRATION.
159
In the curve for April 17, 1912, which was obtained near the beginning
of the fast, it will be noted that the subject took a deep breath every
2 or 3 minutes, but in general the vertical height of the various lines
indicates a fair regularity in the volume of air inspired. The record
of the degree of muscular repose shows that during the whole period of
15 minutes the subject did not make a movement which could be
recorded. As the recording device is so sensitive, it can be confidently
asserted that the subject was in absolute muscular repose throughout
I. PERIOD 2. APR. 17, 1912
lllttltflkiiiiiiiiiw
TIME IN MINS.
VtNTILATION
J
llfflWIIHWinhiii.lli
E. PERIOD I. APR. 3ft 1912,
HI
mmmm
jm
m. PERIOD 3. MAY 14. 1912
***timmmtmm**Hm
JZ. PERIOD 4. MAY 14. 1912
Fig. 22. — Specimen respiration curves for subject L. when lying on couch in experiments
with the respiration apparatus.
160 A STUDY OF PROLONGED FASTING.
the period so far as external muscular activity is concerned, although
it is obvious that no idea of the muscular tonus can be obtained by
this method.
The second curve, that for April 30, 1912, was obtained about the
middle of the fasting period and is typical of many obtained about this
time. In this respiratory record but two abnormally deep breaths
are noted.
The third curve was obtained on May 14, 1912, at the end of the
thirtieth day of the fast. In this record a greater frequency of respi-
ration may be noted, with less amplitude, this being clearly shown even
without measurement. The great sensitivity of the device for record-
ing the degree of muscular repose is shown in the original kymograph
curve by a wave-like line above the respiration record indicating the
slight disturbance in the center of gravity of the body due to the res-
piratory movements . While thi s may be very plainly seen in the original
curve, it is lost in the reproduction.
Immediately after the third curve was obtained, the apparatus was
rilled with pure oxygen, so that the subject breathed an atmosphere
containing 95 per cent oxygen. One deep respiration is shown in the
curve obtained (curve IV). The rate is apparently a little slower than
in the preceding curve and the volume somewhat larger. The line
above the respiration record again shows that the subject was abso-
lutely quiet throughout the whole period, as was usual with this man.
METHOD OF CALCULATING THE TOTAL VENTILATION OF THE LUNGS.
The construction of the spirometer bell is such that each millimeter
length corresponds to a volume in the bell of 23c.c. ; hence by measuring
the vertical distance between the bottom and top levels of the record
made on the kymograph drum by the pointer at the beginning and end
of every inspiration or expiration, the apparent volume of air inhaled
or exhaled may be computed. By measuring all the rising portions
of the respiration curve and subsequently multiplying the result by the
known factor, the total ventilation of the lungs during the experimental
period can be obtained. To simplify this calculation, a recording
device has been added to the spirometer which is somewhat in the
nature of a work-adder wheel1 and permits the accumulative measure-
ment of the movements of the spirometer bell in one direction. Each
revolution of this wheel corresponds to a rise in the spirometer bell of a
certain number of millimeters, and from the record of the number of
revolutions of this wheel the apparent volume of air passing through
the lungs can be calculated.
In these experiments, the apparent volume obtained by this calcula-
tion was converted to standard conditions of temperature and pressure
by multiplying it by the fraction =^j in which p represents the baro-
xBenedict, Deutsch. Archiv f. klin. Med., 1912, 102, p. 176.
MECHANICS OF RESPIRATION. 161
metric reading corrected for scale correction and diminished by 5 mm.
This correction of 5 mm. was found desirable as a result of experiments
in which the humidity of the air inside the spirometer bell was found to
be usually about 30 per cent. As a matter of fact, calculations showed
that the difference due to using an assumed value for complete satu-
ration or partial saturation is not more than 1 or 2 per cent. In addi-
tion to the correction for the pressure, the usual correction for tempera-
ture was made. The total volume as reported is therefore the total
ventilation per minute, corrected for 0° C. and 760 mm. and likewise
for an average value of 5 mm., corresponding to the probable humidity
of the air inside the spirometer bell.
METHOD OF CALCULATING THE VOLUME PER INSPIRATION.
The method of calculating the volume per inspiration is not so simple
as it at first appears. Instead of simply dividing the total ventilation
per minute by the number of respirations, most writers have been
accustomed to calculating the volume per inspiration from the volume
of the air converted to the conditions which exist in the lungs, that is,
the prevailing atmospheric pressure less the tension of water-vapor at
37° C. and corrected for the temperature of the lungs at 37° C. There
has been considerable discussion, particularly in connection with the
experiments of Galeotti,1 and Loewy and Gerhartz,2 as to whether the
temperature conditions should be taken as 37° C, and whether the air
is saturated at this temperature or not. This value is, however, most
commonly used, and, indeed, we are not far in error in doing this,
although, as was shown in an earlier publication,3 the correct determi-
nation of the temperature of the air in the lungs and the degree of
saturation will obviously affect these computations somewhat.
The method used for calculating our results is as follows : The total
ventilation of the lungs, which has been reduced to standard conditions
of 0° C. and 760 mm. pressure, is divided by the number of respirations.
This value is then converted to the pressure existing in the lungs, which
is the atmospheric pressure less the tension of aqueous vapor at 37° C,
or 46.7 mm. It is subsequently converted to the temperature of the
lungs by the usual calculation. A sample calculation will serve to
show the method used : In the morning respiration experiment on April
11, the ventilation of the lungs was 5.32 liters per minute at 0° C. and
760 mm. The observed barometer was 758.7 mm. and the number of
respirations per minute was 12.2. The volume per inspiration would
therefore be
760X (273+37) X5.32
(758.7-46.7)X273Xl2.2
= 529 c.c.
'Galeotti, Biochem. Zeitschr., 1912, 46, p. 173.
1Loewy and Gerhartz, Biochem. Zeitschr., 1912, 47, p. 343.
3Benedict, Carnegie Inst. Wash. Pub. 77, 1907, p. 436.
162
A STUDY OF PROLONGED FASTING.
RESULTS OF OBSERVATIONS ON THE MECHANICS OF RESPIRATION.
The data secured by these methods regarding the respiration-rate,
the ventilation of the lungs per minute, and the volume per inspiration
give material for an interesting study of the effect of prolonged fasting
upon the mechanics of respiration. These data are given in table 18,
which shows two extensive series of values, one for the morning respi-
ration experiments made directly after the subject came out of the
Table 18. — Ventilation of lungs in experiments with L. at different times of the day, and with
varying activity. (Respiration apparatus.)
Date.
Day of
fast.
Lying.
Usually 8h30m a.m. to 9h30m a.m. T.
rsually 7 p.m. to 7h 45m p.m.
Respira-
tion-rate.
Lung ven-
tilation per
minute.1
Volume -n
per inspi- ^
ration.*
spira-
n-rate.
Lung ven-
tilation per
minute.1
Volume
per inspi-
ration.2
1912.
liters.
C.C.
liters.
c.c.
Apr. 11....
12.2
5.32
529
12
9.6
9.6
10.6
9.3
5.21
5.19
4.79
4.97
655
650
539
639
13
14
15....
1st
16....
2d
10.9
5.18
576
17....
3d
11.3
5.24
562
18....
4th
9.8
4.77
591
19....
5th....
11.8
4.88
507
20. . . .
6th
12.0
4.70
473
21....
7th....
11.8
4.79
489
22....
8th....
10.7
4.67
530
23....
9th
12.1
4.65
476
24. . . .
10th
10.9
4.55
504
25. . . .
11th
10.1
4.40
522
26....
12th
12.8
4.64
429 1
L2.8
5^24
488
27....
13th
12.8
4.63
437 ]
14.9
35.35
"437
28....
14th
12.4
4.61
448 ]
L4.7
5.32
437
29. . . .
15th
12.3
4.55
446 ]
L4.6
5.83
483
30....
16th
13.1
5.00
462 1
L4.6
6.01
497
May 1
17th
12.3
4.81
471 ]
14.5
5.79
482
2....
18th
13.2
4.61
422 ]
L5.1
5.81
465
3....
19th
12.8
4.78
449 1
L4.7
5.66
465
4....
20th
14.3
4.90
413
5....
21st
10.0
4.43
532 ]
L5!l
5.76
458
6....
22d
13.5
4.91
436 ]
L4.9
5.69
460
7....
23d
14.0
4.76
410 ]
L6.7
6.03
438
8....
24th
13.7
4.69
417 ]
L5.5
5.77
456
9....
25th
14.2
4.95
428 ]
L4.4
5.74
490
10. . . .
26th....
12.8
4.75
454 ]
L3.6
5.58
502
11....
27th....
12.8
4.89
461 ]
L4.fi
5.82
483
12....
28th
14.8
5.04
410 ]
L4.7
5.76
473
13....
29th
14.1
4.96
426 1
L3.9
5.72
501
14....
30th
14.8
4.80
391 ]
L3.9
5.92
514
15....
31st
13.3
4.84
438
17
9.9
14.0
3.93
5.72
485
494
18
xThe lung ventilation observed is here reduced to 0° C. and 760 mm. pressure.
2Calculated to the pressure existing in the lungs and to 37° C.
3During the period 3h 16m p.m. to 3h 51m p.m., with subject in lying position, the observa-
tions were: respiration-rate, 13.4; lung ventilation, 5.14 liters; volume per inspiration, 466 c.c.
MECHANICS OF RESPIRATION.
163
Table 18. — Ventilation of lungs in experiments with L. at different times of the day, and with
varying activity. (Respiration apparatus.) — Continued.
Date.
Day of
fast.
Sitting.1
Period.
Respira-
tion-rate.
Lung ven-
tilation per
minute.2
Volume
per inspi-
ration.3
1912.
Apr. 16
19....
23....
24....
26. . . .
27. . . .
29....
May 1
4....
7....
14....
2d. . . .
5th...
9th...
10th...
12th...
13th...
15th...
17th...
20th...
23d. . . .
30th...
4h 00m p.m. to 4h 35m p.m , ,
4 10 p.m. 4 43 p.m.* . . .
3 52 p.m. 4 28 p.m...
3 58 p.m. 4 57 p.m
3 13 p.m. 4 11 p.m, ..
12 14 p.m. 12 48 p.m
3 23 p.m. 3 66 p.m.* . . .
9 31 a.m. 10 04 a.m.*. . .
9 35 a.m. 10 10 a.m.*. . .
3 43 p.m. 4 14 p.m.*
6 32 p.m. 7 02 p.m.*. ..
10.3
17.9
16.7
14.6
15.8
12.8
18.7
14.6
15.3
16.1
17.8
liters.
5.58
7.54
5.48
5.83
5.37
5.55
7.88
6.57
6.22
7.62
8.05
c.c.
660
517
402
484
404
525
510
542
490
573
546
1Periods indicated by an asterisk were obtained with the subject sitting, writing.
2The lung ventilation observed ia here reduced to 0° C. and 760 mm. pressure.
'Calculated to the pressure existing in the lungs and to 37° C.
calorimeter and representing every day of the fast, and another series
for the evening respiration experiments made each day during the
latter part of the fast just before the subject entered the calorimeter.
RESPIRATION-RATE.
An examination of the respiration-rate for the morning period shows
that there was a distinct tendency for it to increase as the fast con-
tinued, the lowest rate being observed on the first day of the fast, i. e.,
9.3 respirations per minute and the highest rate of 14.8 respirations, on
the twenty-eighth and thirtieth days of the fast. The values obtained
in the evening began on the twelfth day of the fast and indicate a
reasonably constant respiration-rate, averaging not far from 15 respi-
rations per minute. The evening rate in practically all cases was
slightly higher than the morning rate.
Cathcart states that with his subject there was no change in the
character of the respiration as the fast progressed, and his figures show
a tendency for the morning respiration-rate to remain constant or to
decrease slightly. Our observations, on the contrary, indicate a tend-
ency to increase in the morning. Our findings also differ from those
recorded by Luciani for Succi, as his curve indicates a tendency to fall
towards the end of the fast. While on a number of days in Succi's fast
the evening respiration-rate was higher than that obtained in the
morning, in a large number of instances the reverse was true. In view
of Succi's excitable temperament and the fact that his daily routine was
not absolutely constant, it is more than likely that the discrepancies
appearing between our results and Luciani's may be easily explained
by the fact that in the experiment with Levanzin the routine was
164 A STUDY OP PROLONGED FASTING.
rigidly adhered to each day, the subject lying very quietly for some time
while the respiration-rates were being recorded.
The experiments of the Berlin investigators with Cetti and Breit-
haupt were not sufficiently long to make them comparable with this
31-day experiment and they were likewise complicated considerably
by the fact that the subjects suffered from cold and colic.
From an examination of all of the kymograph records obtained
with L., it is clear that while prolonged fasting tended to increase
the average respiration-rate, there was great regularity of respiration
throughout each 15-minute period. Occasionally a deep breath was
taken, but there was nothing like the great irregularity noted by Zuntz
and his co-workers on the two Berlin fasters, an irregularity which may
again be explained by the complications of cold and colic.
VENTILATION OF THE LUNGS PER MINUTE.
The actual amount of air passing through the lungs was measured on
the spirometer and its recording attachment. The ventilation of the
lungs per minute, which is given for each experiment in table 18,
followed a somewhat singular course. In the morning observations
the ventilation per minute showed a persistent, though slight, tendency
to decrease during the first 4 days with food; it then rose perceptibly
in the first 3 days of the fast and subsequently decreased until the low
value of 4.4 liters was reached on the eleventh day. This was closely
approximated on the twenty-first day, when a value of 4.43 liters was
obtained. The lowest value in the experimental period was found on
the second day with food after the fast, when the ventilation was 3.93
liters. Few deductions can be drawn from these figures for the lung
ventilation per minute, save that on certain fasting days the values
were very low as compared with the four days preceding the fast,
although, as has already been pointed out, the minimum value was
obtained on the second day with food after the fast. Here again the
values for the evening observations show an increase, the ventilation
being invariably greater than during the morning experiments, rising
at times as high as 6 liters. The average value was 1 liter higher than
those obtained during the morning observations.
VOLUME PER INSPIRATION.
In discussing the values for the volume per inspiration given in table
18, it must again be stated that these were not obtained by dividing
the total ventilation of the lungs by the number of respirations, but
by using the volumes changed to the conditions in the lungs, as is com-
monly done by other writers of the present day. These figures show
that there is a distinct tendency for the volume per inspiration to de-
crease as the fast progressed, although certain high values are found on
the eighth, eleventh, and twenty-first days of fasting. On the other
MECHANICS OF RESPIRATION. 165
hand, the lowest value recorded in the morning experiments — 391 c.c. —
was on the thirtieth day of fasting. In the evening series we note that
while the values in general are somewhat higher than in the morning,
this increase seems to become greater toward the end of the fast. Thus,
on the thirtieth day of the fast it was 391 c.c. in the morning and 514 c.c.
in the evening. While, therefore, there is a positive average difference,
inasmuch as in the evening the volume per inspiration is greater than
in the morning, the difference has a tendency to become very much
greater in the last week of the fast.
INFLUENCE OF CHANGES IN BODY POSITION.
On a number of days the subject was studied when sitting in his
chair, either resting or writing. The values obtained are given in table
18 for comparison with those found while the subject was lying on a
couch. During the sitting experiments, when the subject was not
writing, the respiration-rate was in practically all cases slightly higher
than the values obtained in the morning respiration experiments while
the subject was lying quietly. On the ninth day of the fast it increased
from 12.1 to 16.7 respirations per minute. The ventilation of the
lungs per minute also increased perceptibly in every instance, the in-
crease being not far from 0.8 liter. On the other hand, the volume per
inspiration varied considerably. In two instances there was a per-
ceptible increase, on two other days it decreased, while on another day it
remained essentially constant. This difference is not so apparent when
the results are compared with the records for the evening respiration
experiments. Unfortunately the sitting experiments were not suffi-
ciently extended to draw any definite conclusions regarding the effect
of the change in body position; furthermore, the whole study lacks
suitable normal values for comparison.
INFLUENCE OF THE WORK OF WRITING.
On 6 of the fasting days the metabolism of the subject was studied
while he sat in a chair and wrote actively. On 2 of these days he was
studied in the forenoon and on 4 days in the afternoon. Since there
is a tendency towards a diurnal variation in the mechanics of respiration
between morning and evening, as shown by the increase in the respira-
tion-rate and the ventilation of the lungs per minute, and the tendency
for the volume per inspiration to increase, it is necessary to take this
fact into consideration in discussing the results. In the two forenoon
experiments there was in both instances an increase in the respiration-
rate, a marked increase in the ventilation of the lungs per minute, and
a great increase in the volume per inspiration. Inasmuch as the writing
was accompanied by distinct, though perhaps slight, muscular effort,
these findings are only what would be expected. In the afternoon
experiments there was an increase in the respiration-rate much more
166 A STUDY OF PROLONGED FASTING.
noticeable than in the experiments in the forenoon. The ventilation
of the lungs per minute showed a large increase, the values averaging
about 7.75 liters per minute. There was also usually a measurable
increase in the volume per inspiration.
From these results it can be inferred that the slight muscular work of
writing letters perceptibly affected the mechanics of ventilation in that
the respiration-rate was somewhat increased and the ventilation of the
lungs per minute noticeably so. So far as we know, no study has been
made with normal individuals in which the ventilation of the lungs per
minute and the volume per inspiration were so carefully observed as
were those of our fasting subject, and hence we have no comparable
values which will show to what extent the factors affecting the mechanics
of respiration were influenced by prolonged fasting. It is reasonable
to suppose, however, that muscular exercise of any kind would require
a greater effort in the later stages of inanition. It is of particular
interest that the lung ventilation per minute in the afternoon experi-
ments was perceptibly greater than when essentially the same amount
of work was carried out in the forenoon.
INFLUENCE OF BREATHING AN OXYGEN-RICH ATMOSPHERE.
On three days during the fast an experiment was made directly after
the morning respiration experiment, in which the subject breathed an
atmosphere containing from 95 to 75 per cent of oxygen. The influence
of this increased amount of oxygen was distinctly noticeable with the
ventilation of the lungs per minute and the volume per inspiration,
although the respiration-rate changed but little. On the first day on
which these experiments were made (May 12, 1912) the volume per
inspiration with normal air was 410 c.c. and with the oxygen-rich
mixture it was 487 c.c, an increase of 16 per cent. The values obtained
in these experiments were as follows : Twenty-eighth day of fast, respi-
ration-rate, 13.6, lung ventilation, 5.50, volume per inspiration, 487;
twenty-ninth day, respiration-rate, 14.0; lung ventilation, 5.44, and
volume per inspiration, 471 ; on the thirtieth day, respiration-rate, 14.2;
lung ventilation, 5.34; volume per inspiration, 454. It is clear, there-
fore, that with this subject the breathing of oxygen-rich mixtures
resulted in a considerable increase in the ventilation of the lungs per
minute, and while the respiration-rate was not materially affected,
there was a considerable increase in the volume per inspiration.
MAXIMUM EXPIRATION OF THE LUNGS.
As an index of a possible change in the volume of the lungs and par-
ticularly in the strength of the chest muscles, observations in regard to
the maximum expiration of the lungs were made by Mr. Carpenter on
5 days during the fasting period. For these observations a long rubber
tube was attached to a 10-liter Bohr meter. The subject stood up and,
MECHANICS OF RESPIRATION.
167
while holding his nose, inhaled as deeply as he could, then placed the
end of the rubber tube in his mouth and exhaled into the meter to the
smallest possible volume of the lungs. The difference between the
beginning and end readings on the meter gave the maximum apparent
expiration. In computing the true volume of the expiration, these
figures were corrected for the temperature of the air in the gas-meter,
the barometer, and the temperature of the air in the lungs, the latter
being assumed to be 37° C. The results are given in table 19.
Table 19. — Maximum expiration of subject L
/. during fasting.
Date.
Day of fast.
Time of
observation.
Volume
observed.
Barometric
pressure.
Volume
exhaled
(computed).1
1912.
Apr. 15
20
May 7
8
14, ,
2d
7th
24th
25th
31st
4h 35m p.m.
3 00
3 04
3 30
/4 30
\4 36
liters.
3.45
3.60
2.90
3.00
2.50
2.45
mm.
764.2
762.0
759.4
752.6
761.6
761.6
liters.
3.74
3.91
3.15
3.24
2.71
2.66
1In computing these values it was assumed that the temperature of the air when exhaled
was 37° C, and when passing through the meter it was between 23° and 23.8° C.
Although the volume on the seventh day of the fast (3.91 liters) was
higher than that on the second day, a distinct tendency is shown for the
volumes to decrease as the fast progressed. On the twenty-fourth and
twenty-fifth days the volumes were essentially the same. Six days
later — on the thirty-first day of the fast — the volume again materially
decreased, as duplicate readings show values of 2.71 and 2.66 liters
respectively.
One can hardly ascribe this marked loss in the volume of air expired,
amounting to about 30 per cent, exclusively to change in the volume of
the lungs as a result of the fasting or exclusively to the inability of the
weakened muscles of the chest to compress the chest walls further. In
all probability both factors contributed to this change in the volume of
the expiration. The readiness with which the lungs respond to arti-
ficial atmospheric conditions leads one to believe that there may have
been an absolute diminution in the available lung volume during the
fasting period. On the other hand, there was unquestionably a falling
off in the strength and general tone of the subject and he may not have
been able to compress the lungs sufficiently to force out a large volume
of air at the end of the fast.
ALVEOLAR AIR.
By Harold L. Higoins.
Observations were made of the carbon-dioxide percentage of the
alveolar air nearly every day throughout this fasting experiment. This
offered an index as to the acidity of the blood and also an opportunity
to study the control and mechanics of respiration throughout the fast.
Alveolar air is the air which is in or comes from the alveoli of the
lungs. As the active exchange of carbon dioxide and oxygen between
the blood and the lungs takes place in the alveoli, it is readily seen that
the tension or partial pressures of the different gases in the alveoli
(carbon dioxide, oxygen, and also nitrogen, argon, etc.) will be very
nearly the same in the alveoli as in the blood leaving the lungs. Inas-
much as the quantity of a gas dissolved in a liquid is proportional to
the partial pressure, and not to the percentage of the gas, the compo-
sition of alveolar air is therefore probably better expressed in tensions
or partial pressures than in percentages.
SIGNIFICANCE OF ALVEOLAR AIR.
Haldane and Priestley1 have shown that carbon dioxide is the pre-
vailing stimulus to respiration under normal conditions. Thus, if the
carbon-dioxide tension in the respiratory center falls below a certain
level, apncea is the result; and if, on the other hand, it rises above this
level, the respiration volume is greatly increased and hyperpncea sets
in. In other words, the respiratory center by respiratory impulses
automatically keeps its carbon-dioxide tension constant. But the
carbon-dioxide tension of the respiratory center is largely controlled by
that of the arterial blood and the latter is, as mentioned previously,
essentially that of the alveolar air. Haldane has therefore introduced
the use of alveolar carbon-dioxide tension and shown that in any one
individual it is practically constant under normal conditions, although
the normal values of individuals may differ markedly from each
other. It has been discovered that when there is an increased
acidity of the blood, as in diabetic acidosis, or with reduced barometric
pressure, as in high altitudes, the alveolar carbon-dioxide tension is
lower than normal, and a smaller tension of carbon dioxide stimulates
respiration. This has led to the presentation of the theory,2 now
quite satisfactorily established, that the H-ion concentration of the
blood rather than the carbon-dioxide tension is the predominating
factor in the control of respiration. Thus, when the H-ion concentra-
tion (or degree of acidity) of the blood coming to the respiratory center
reaches a certain level, impulses are sent out from the center to increase
haldane and Priestley, Joum. Physiol., 1905, 32, p. 225.
2Winterstein, Archiv f. die ges. Physiol., 1911, 138, p. 167.
168
ALVEOLAR AIR. 169
the respiration so that the net result is always the same H-ion concen-
tration in the center.
The acidity of the blood may be divided into two parts, that due to
carbon dioxide and that due to other acids. As the total acidity
necessary to cause respiration must always be the same, it is readily
seen that if the other acids in the blood increase in amount, less carbon
dioxide is necessary to raise the acidity to the point of stimulation of the
respiratory center. Thus, one may say that the quantity of carbon
dioxide will vary conversely from that of the other acids of the arterial
blood. Since alveolar carbon-dioxide tension represents so closely the
carbon-dioxide tension of the arterial blood, it affords a very good
index of the acidity of the blood. It was mainly for this reason that
the alveolar carbon dioxide in the experiment with L. was so closely
followed. In fact, it seems that this index of the degree of acidosis is
much more satisfactory and important than the urinary tests for acidity
NH — N
(as /3-oxybutyric acid, — i — » total titratable acidity, etc.), because
the former represents the acid actually in the blood, while the latter
only represents that excreted from the body. The other factors which
affect the alveolar carbon-dioxide tension, such as the absorption of
food and varying postures, were avoided with L., and thus one is able to
study the results almost purely from the point of view of blood acidity.
METHODS OF DETERMINING THE ALVEOLAR AIR.
HALDANE METHOD.
Haldane's method1 for determining the alveolar carbon-dioxide ten-
sion is the oldest and probably theoretically the most sound of any of
the methods now in use. By it one collects two samples of alveolar air
from different phases of the respiratory cycle and averages their carbon-
dioxide content. The two phases chosen are immediately at the end
of an inspiration, which is approximately when the alveolar carbon-
dioxide tension is lowest, and at the end of an expiration, when the
alveolar carbon-dioxide tension is nearly at its highest point. The
subject breathes normally for some time; then at the end of a normal
inspiration he makes a rapid, deep expiration through a tube about
2 cm. in diameter and about 150 cm. long, sealing with his tongue the
end he has just breathed into. A sample of the air in the tube near
the mouth is then taken. This sample is considered to be alveolar air,
as the air in the dead space of the respiratory passages and in that part
of the tube from which the sample is taken has previously been pushed
out by the air from the alveoli. Similarly a sample is taken of air
forced through the tube from the lungs at the end of a normal expira-
tion. Instead of sealing off the end of the tube with the tongue, use
^aldane and Priestley, Journ. Physiol., 1905, 32, p. 225.
170 A STUDY OF PROLONGED FASTING.
has been made in our laboratory of a simple Siebeck1 valve, which puts
the subject under much less strain, as he does not have to hold his tongue
to the tube while the sample is being taken into a gas-sampler. The
average of the two analyses gives very closely the composition of the
alveolar air.
The Haldane method requires considerable attention on the part of
the subject and, as it was feared that possibly in the course of the long
fast the subject would not be physically able to co-operate very satis-
factorily, the method used in these tests was modified somewhat. In
view of what we now know of the condition of the subject throughout
the fast, we may feel assured that this method would have been very
successful ; but as several samples are often required to be sure of good
agreement, and as it was probable that the subject's time would be
much occupied, it was decided to modify the method somewhat to be
sure of better agreement on fewer samples.
It has been observed that, in the Haldane method, holding the breath
for several seconds before the expiration does not cause the percentag ;
of carbon dioxide in the alveolar air to increase with very great rapiditye
this is naturally to be expected, for as the carbon-dioxide tensions of
the alveolar air and the blood coming to the lungs approach the same
figure, the increase in the former is slower. Furthermore, it appears
that if a subject has previously been breathing somewhat abnormally
for not over three or four respirations, the percentage of carbon dioxide
in the alveolar air, after holding the breath for a few seconds, will be
nearer that of the alveolar air when the breath is held similarly after
normal respiration than is the percentage of carbon dioxide in the
alveolar air of the same two cases when the breath is not held. Thus,
it would seem that small deviations from the normal, such as appear in
conscious respiration, would not be disturbing to agreeing results and
that in a very small number of determinations (seldom more than two)
figures can be obtained which are very good duplicates and which will
bear a constant relation to the true alveolar carbon-dioxide tension
when the subject is in the same position (sitting quietly).
The modified method used in these tests, which later is called the Hal-
dane method, is as follows: The subject began by breathing normally
into the room through a short (5 cm.) tube connected with the Siebeck
valve. Then at the end of an inspiration, selected by the observer who
was watching the respiration, the subject was told to hold his breath.
At the end of 5 seconds, timed by the observer, during which the valve
had been opened, the subject breathed out rapidly and deeply through
the long tube as in the Haldane method. After the expiration the valve
was again closed. Usually two samples were taken each day in which
the subject held his breath 5 seconds and two in which he held his breath
8 seconds. The results obtained when the subject held his breath 8
xDr. R. Siebeck, of Heidelberg, has devised an ingenious slide-valve for this purpose, which
may be secured of Universitats Mechaniker Runge in Heidelberg.
ALVEOLAR AIR. 171
seconds average a trifle higher than when the breath was held for 5
seconds, but the agreement is so close that one could not satisfactorily
select the individual determinations of each class if the results were put
together and not labeled. The results were averaged, therefore, with-
out distinction as to time. The determinations made in duplicate by
this method agreed in general to 1 part in 20 and usually closer. To
determine how close the results were to the figures which would have
been obtained by the Haldane method, we experimented by both
methods on ten different subjects sitting; the average result when the
breath was held 5 seconds was 8.3 per cent higher and when held 8
seconds was 9.4 per cent higher than with the Haldane method. The
averages of the 5 second samples and the 8 second samples are thus
about 9 per cent higher than the Haldane figures. Excluding two of
the ten cases (5 per cent and 19 per cent), none showed differences of
more than 3 per cent (6 per cent to 12 per cent) from the average
difference (9 per cent). Thus, for comparing the daily observations
with each other, it appears that the values obtained with the 5-second
and 8-second methods in the experiment with L. are practically as
significant as if the Haldane method were used.
PLESCH METHOD.
Use was also made of the Plesch method1 applied by Porges, Leim-
dorfer and Markovici2 to clinical cases. By further modification of
the method I have been able to get very constant duplicates with a
minimum amount of attention by the subject. The apparatus used in
this method consists of a woman's rubber bathing cap (pure gum),
which is fastened to the bottom of an inverted shallow copper pan
(about 20 cm. in diameter). On the other side of the pan is soldered a
f-inch (2 cm.) three-way valve; by means of a rubber- tube connection
the subject may breathe back and forth through this valve, either from
the room or from the bag made of the bathing cap and the pan. A
small brass stop-cock is attached to the pan, from which a sample of the
gas in the bag may be obtained. In a determination, the bag was first
emptied and 600 c.c. of the room air was admitted, the measurement
being made by a meter. The subject then began breathing room air
through the rubber tube and three-way valve, closing the nose with
the thumb and forefinger of the hand holding the apparatus. At the
end of a normal expiration the observer turned the valve and the
subject breathed in all of the 600 c.c. of air in the bag. He then
breathed back and forth at the rate of one complete respiration in
5 seconds, the time being followed by the observer, who instructed
the subject when to breathe in and when to breathe out. At the end of
4 complete respirations, %. e., 20 seconds, the three-way valve was
turned and a sample taken in the gas-analysis apparatus for analysis.5
Plesch, Zeitschr. f. exp. Path. u. Therapie, 1909, 6, p. 380.
2Porges, Leimdorfer and Markovici, Zeitschr. f. klin. Med., 1911, 73, p. 389.
3AU the analyses of alveolar air were made on a portable Haldane apparatus. (Haldane, Methods
of Air Analysis, London, 1912.)
172 A STUDY OF PROLONGED FASTING.
This method is probably the most adaptable for use with the average
patient, when the condition of acidosis is being compared from day to
day, or when a gross picture of the degree of acidosis is desired. The
method does not give the carbon-dioxide tension of the arterial blood,
but seems rather to approach the carbon-dioxide tension of the venous
blood, because, as the same air is rebreathed, it is obvious that the
alveolar air, the arterial blood, and venous blood will all have eventu-
ally the same carbon-dioxide tension, namely, that of the venous blood,
because it is the highest. For this reason, especially as it is the normal
carbon-dioxide tension of the arterial blood, which is the important
factor in the regulation of the respiration, this method theoretically is
not so important as the Haldane method.1 But with the subject in
the same position and with the same amount of previous activity, we
have found that the carbon-dioxide tension determined by this means
bears a very constant relation to that of the Haldane method; this was
assured from numerous comparisons of the different methods on many
normal individuals, the results being about 20 per cent higher than the
values obtained with the Haldane method. The same relation may
also be observed with L., as the results in table 20 show that, excepting
on the first few days, the difference between the 5- and 8-second Haldane
method and the modified Plesch method is about 10 per cent.
METHOD OF CALCULATING ALVEOLAR AIR FROM RESPIRATION
EXPERIMENTS.
The morning and evening respiration experiments, which were made
with the universal respiration apparatus, included the determinations
of the carbon-dioxide production, oxygen consumption, respiratory
quotient, pulse- and respiration-rates, and inspiratory ventilation of
the lungs. These experiments also give some data regarding the alve-
olar carbon-dioxide tension and the dead space of breathing, which are
of interest in considering the other alveolar-air determinations. The
dead space in respiration is the air that is inspired and again expired
without entering the alveoli, in which active gaseous exchange takes
place, and thus is unchanged. The following formula for calculating
the percentage of carbon dioxide in the alveolar air is therefore readily
understood :
A1 . ^~ C02 V = Total volume of air ex-
Alv. per cent C02 = - — —-- — — • 7 •
V-(DSXR) piredmcx.
C02 = C02 production in c.c. DS = Dead space in c.c.
R = Number of respirations.
Naturally these factors must be measured for the same unit of time
and under the same conditions of pressure, temperature, and aqueous
tension. The unit of time chosen in the experiments with L. has been
*The same criticism applies to the 8-second and 5-second Haldane method as used in these
experiments, although probably to a less degree.
ALVEOLAR AIR. 173
1 minute, while the gas volumes considered in the application of the
above formula have, for the sake of simplicity in calculation, been taken
at 20° C, 760 mm., and dry.
Elaborating this general formula for use in connection with and cal-
culation from the respiration experiment, we get the following:
A1 + rn CO,X 1.075
Alv.p.ct.C02= . p TpTT ~r\ _nr\ Z
v(m) -°-015 v(t1>) " ™<rLTBt) ~ <Rx m
C02 = C02 production in c.c. per minute at 0° C. and 760 mm. ;
and C02X 1.075 = C02 production at 20° C, 760 mm.
02 = 02 consumption in c.c. per minute, 0° C, 760 mm.
V = inspiratory ventilation of lungs per minute at barom-
eter, temperature, and humidity prevailing in spirom-
eter of respiration apparatus (i. e., about 20°C. and
33 per cent or 66 per cent humidity).
v(^) -0.015 V(^) - 1.075(°2~C°2) -expiratory ven-
tilation of lungs per minute at 760 mm., 20° C, dry.
DS = dead space in c.c. (20° C, dry, 760 mm.)
R = respiration-rate per minute.
P = barometric pressure.
1.075 = factor to convert gas volumes from 0° C. to 20° C.
In a respiration experiment the inspiratory ventilation is obtained at
the pressure, temperature, and humidity of the air in the spirometer on
the apparatus. The temperature is not taken in each experiment, but
as it is probably very close to 20° C, this temperature is assumed in
P
each calculation. The term =^r obviously reduces the ventilation to
/ P \
760 mm. The term —0.015 V (;™) corrects for moisture in the
spirometer; this moisture comes from two sources, namely, from the
lungs of the subject and from the moistener used to prevent the air in
the respiration apparatus from becoming too dry for comfortable respi-
ration. Two kinds of moisteners were used during the fasting experi-
ment. In the first part of the series (until April 22) a moistener con-
structed of the lower part of a Kipp gas-generator was employed. In
this the air before coming to the nosepieces bubbles through water;
the humidity with this form of moistener has been found to be 66 per
cent saturated; 1.5 per cent of the recorded ventilation was therefore
water- vapor and accordingly subtracted. The other moistener was a
piece of moist cheese-cloth in the tube leading to the nosepiece. When
this was used, the humidity in the spirometer was found to be only 33
per cent, so that in these experiments 0.75 per cent of the recorded
volume was subtracted and not 1.5 per cent. The expiratory volume
174 A STUDY OF PROLONGED FASTING.
is smaller than the inspiratory volume, because the amount of oxygen
consumed is greater than the amount of carbon dioxide produced.
Accordingly we subtract 1.075 (02 — C02) from the inspiratory venti-
lation. But during inspirations in a respiration experiment one-half
of the carbon dioxide produced is absorbed in the soda-lime bottles,
so that the recorded inspiratory volume is correspondingly increased;
on the other hand a volume of oxygen equal in volume to one-half the
volume consumed is added to the respiration apparatus during an
inspiration and thus makes the recorded inspiratory volume corre-
spondingly too small. Thus, instead of 1.075 (02 — C02), the factor
for changing the inspiratory volume to the expiratory volume becomes
1.076 (9q£%
The only other factors in the equation are Alv. p. ct. C02 and DS.
With either one known, the other may easily be determined. The
formula for calculating the dead space is as follows :
R
When the experiments on L. were made, we had not considered the
possible use of this equation and so have not the complete data for
calculating either of these factors, but we still have sufficient material
to draw some interesting conclusions.
Assuming the personal dead space of breathing for the subject L.,
together with that of the nosepieces, etc., to be 120 c.c, the percentage
of carbon dioxide in the alveolar air has been calculated for all of the
morning and evening respiration experiments, and the results for each
series have been averaged as shown in table 20. Also, making use of
the alveolar-air figures found on the same day, the respiratory dead
space in each experiment has similarly been calculated and averaged.
As the alveolar air was not taken at the time of the respiration experi-
ments and as the dead space might possibly have changed in size
during the fast, fixed differentiation of the results is difficult. These
results will be discussed later.
CONDITIONS OF TAKING ALVEOLAR-AIR SAMPLES
The samples were taken by the Haldane and Plesch methods with
the subject sitting in an armchair. After the fast had begun, they were
taken at about lh35m p.m. to 2 p.m. Between the taking of the samples,
each of which was analyzed before another was taken, the subject was
sitting quietly and usually reading. On several days he had visitors
while the experiment was in progress. On one of these days, April 22,
while talking with a visitor, he became quite excited. On the other
days there was no marked excitement while the samples were being
taken. On April 22, it is interesting to note that the alveolar carbon-
dioxide tension by the Haldane method was very low — in fact, much
ALVEOLAR AIR. 175
lower than that calculated from the respiration experiments on the
same day would seem to indicate it should normally have been. On
2 days, April 25 and 26, the alveolar air was not determined.
On the food days preliminary to the fast, the alveolar air was sampled
with the subject sitting, immediately after the respiration experiment
of the morning and the taking of the body-weight. On the morning
of the first food day, April 11, the subject was unused to the apparatus
and the tests were unduly hastened ; the results obtained can not there-
fore be considered so reliable as on later days. The samples of alveolar
air in the food days subsequent to the fast were taken at approximately
lh 35m to 2 p. m., as during the fast.
DISCUSSION OF RESULTS.
The results of the determinations made by the Plesch and Haldane
methods are expressed in table 20, columns f and g, as tensions (milli-
meters of mercury). The tensions are calculated from the carbon
dioxide obtained by the analyses. From the prevailing barometric
pressure is subtracted the figure 46.7 mm., which is the aqueous tension
of air saturated at 37° C. (the air in the alveoli being saturated with
water-vapor at this temperature), and the resulting pressure is multi-
plied by the percentage of carbon dioxide found. Before discussing
from a physiological point of view the results obtained by these methods
it seems desirable to summarize first the results gathered from the
respiration experiments.
SIZE OF DEAD SPACE IN FASTING.
A diminution in the size of the heart, liver, and other organs, as well
as in the size of the muscular tissue, having been observed during the
fast, the question was raised by Dr. Benedict as to whether or not the
dead space in breathing also changed in size during the fast, and at his
suggestion use was made of the formula given previously, and the data
available, to calculate so far as possible a figure for the size of the dead
space for each morning and evening experiment. To get a value for
the alveolar carbon-dioxide percentage to use in these calculations, cer-
tain corrections have been made in the values obtained by the 5-second
and 8-second methods given in column g of table 20. These corrections
are necessary for two reasons : first, because the alveolar air was taken
with the subject sitting, while in the respiration experiments the sub-
ject was lying on his back; second, because the results obtained by the
Haldane method give a carbon-dioxide percentage about 9 per cent
of the total less than that when the subject held his breath from 5 to 8
seconds. As shown in a recent paper,1 the alveolar carbon dioxide
lying is about 106 per cent of that sitting. Thus the alveolar percent-
age of carbon dioxide which should be used in calculating the dead
space in the respiration experiments is 97 per cent of the figure from
which column g is calculated, and is given in column a. In using
Wiggins, Am. Journ. Physiol., 1914, 34, p. 114.
176
A STUDY OF PROLONGED FASTING.
this value for each day, it was necessary to assume that the alveolar
percentage of carbon dioxide had not changed between 8 a. m. or
8 p. m., when the respiration experiments were made, and 2 p. m. of
the same day, when the alveolar percentage of carbon dioxide was
taken. It would seem, however, that if there were a change in the
Table 20. — Alveolar-air and dead-space determinations in
experiment with L.
Day
Alveolar air.
Alveolar
CO,
Computed alveolar COj (dead
space = 120 c.c ).
Date.
of
(Haldane),
From morning respiration
From evening respiration
fast.
corrected
to lying
position.1
experiments.
experiments.
By periods.
Average.
By periods.
Average.
A
B
C
D
£
1912.
p. ct.
p.ct.
p.ct.
p.ct.
p.ct.
mm. Hg.
p.ct.
p.ct. 1 p.ct.
p. a
. mm.Hg.
Apr. 11
4.76
5.11
4.94
35.2
....
12
4.89
4.70
4.84
4.75
4.76
34.0
13
4.94
4.80
4.89
4.99
4.89
35.1
14
5.10
4.94
5.25
5.12
5.10
36.9
IS
1st
4.43
4.76
4.91
4.60
4.76
34.3
16
2d
4.29
4.68
4.63
4.42
4.58
32.6
....
17
3d
4.37
4.29
4.30
4.30
4.30
30.6
18
4th
4.36
4.42
4.35
4.37
4.38
31.1
19
6th
4.33
4.50
4.65
*4.43
4.47
31.5
20
6th
4.30
4.39
4.51
4.43
4.44
31.7
21
7th
4.38
4.49
4.43
4.48
4.47
32.0
22
8th
3.94
4.36
4.31
4.36
4.34
30.9
23
9th
4.45
4.44
4.30
4.32
4.35
30.3
24
10th
4.29
4.23
4.31
4.21
4.25
30.4
25
11th
4.43
4.23
4.19
4.28
30.8
26
12th
4.42
4.42
4.27
4.37
31.9
3.65
3.64
3.63
3.64
[ 26.4
27
13th
3.88
4.43
4.33
4.40
4.39
31.4
3.71
3.64
3.68
3.6*
\ 26.1
28
14th
3.95
4.17
4.10
4.15
4.14
29.6
3.92
3.58
3.45
3.61
26.1
29
15th
3.92
4.27
4.14
4.04
4.15
29.7
3.29
3.21
3.21
3.24
23.1
30
16th
3.74
3.79
3.75
3.76
3.77
26.9
3.04
3.05
3.07
3. OS
21.8
May 1
17th
3.87
3.82
3.81
3.74
3.79
27.2
3.03
3.14
3.14
3.1C
22.2
2
18th
3.90
4.10
3.87
3.83
3.93
28.1
3.08
3.07
3.15
3.1C
22.2
3
19th
3.75
3.87
3.83
3.72
3.81
27.3
3.13
3.11
3.15
3.13
22.4
4
20th
3.64
3.82
3.75
3.67
3.75
26.9
5
21st
3.87
3.92
3.79
3.92
3.88
27.9
3.03
3.05
3!o4
3.04
21.9
6
22d
3.77
3.74
3.63
3.54
3.64
26.2
3.07
2.96
3.21
3.08
22.1
7
23d
3.78
3.71
3.95
3.75
3.80
27.2
3.12
2.90
3.01
21.4
8
24th
3.68
3.97
3.80
3.86
3.88
27.5
3.12
2.98
3! 14
3.08
21.7
9
25th
3.77
3.79
3.74
3.60
3.71
26.1
2.95
2.97
3.00
2.97
20.9
10
26th
3.84
3.74
3.76
3.92
3.81
26.9
3.11
3.14
3.15
3.13
22.2
11
27th
3.72
3.64
3.81
3.73
26.7
2.92
3.01
3.03
2.99
21.4
12
28th
3.78
3.78
3.71
3.48
3.66
26.2
2.92
2.96
3.00
2.96
21.1
13
29th
3.84
3.73
3.62
3.65
3.67
26.1
2.91
2.95
2.96
2.94
20.8
14
30th
3.77
3.83
3.87
3.69
3.80
27.2
2.89
2.88
2.89
20.7
15
31st
34.33
3.60
3.64
3.47
3.57
25.6
16
34.36
17
34.83
4.73
4.68
24.'77
4.70
33^2
18
4.06
4.15
24.06
4.11
29.3
1Obtained by taking 97 per cent of the percentages alveolar CO2 from which the figures in column a
■were calculated.
Calculated percentages for the fourth period on April 19, 4.31; May 17, 4.60; May 18, 4.18.
3The subject ended his fast with the taking of fruit juices and honey on the morning of May 15, afte
the conclusion of the respiration experiments.
ALVEOLAR AIR.
177
alveolar air during that time, it would be proportional on each day;
on certain days subsequently specified there is good reason to believe
that the change was not proportional. The results of these calcula-
tions are given in column h for the morning experiments and j for the
evening experiments, with the averages in columns i and k, respectively.
Table 20. — Alveolar-air and dead-space determinations in experiment with L. — Continued.
Alveolar air.
Volume of dead space (using alveolar CO2 in column A).
COj tension for
sitting.
Computed from morning
respiration experiments.
Computed from evening
Day
respiration experiments
.
of
Modifica-
Haldane
Date.
tion of
method
fast.
Plesch
method.
(breath held
5"to 8").
By periods
1.
Aver-
age.
By periods.
Ave
age
r-
F
G
H
I
J
K
mm. Hg.
mm. Hg.
c.c.
c.c.
c.c.
c.c.
c.c.
c.c.
c.c.
c.c
1912.
31.9
31.7
Apr. 11
35.4
36.0
141
124
132
132
12
35.1
36.5
135
124
116
125
13
35.7
37.5
130
110
118
119
14
33.7
32.8
85
80
98
88
1st
15
33.7
31.3
90
84
108
94
2d
16
33.1
32.1
127
126
126
126
3d
17
35.4
31.9
115
121
120
119
4th
18
34.1
31.4
109
96
^ll
109
5th
19
34.7
31.6
114
105
111
110
6th
20
35.3
32.3
112
117
113
114
7th
21
34.4
28.7
86
86
81
84
8th
22
34.3
32.1
120
130
128
126
9th
23
34.0
31.5
126
118
126
123
10th
11th
12th
24
25
26
32^8
28 A
84
88
82
85 1
32
136 ]
133 ]
34
[ 13th
27
32.4
29.1
103
109
107
106 3
22
145 ]
56 ]
41
14th
28
31.6
28.9
96
105
111
104 ]
69
175 1
L79 ]
74
L 15th
29
30.8
27.5
116
119
118
118 ]
78
184 ]
L77 ]
8C
1 16th
30
30.0
28.5
124
125
130
126 1
90
176 ]
L79 1
85
17th
May 1
31.4
28.7
107
122
125
118 1
83
179 ]
179 1
8C
> 18th
2
29.3
27.6
111
114
122
116 1
68
172 ]
66 ]
OS
19th
3
30.1
26.9
107
112
118
112
20th
4
30.0
28.7
115
127
116
119 1
79
187 ]
82 1
83
21st
5
30.5
27.9
122
130
138
130 1
76
181 ]
64 1
74
22d
6
29.5
27.8
125
110
122
119 1
65
185
1
75
23d
7
29.1
26.8
100
112
108
107 1
64
175 1
60 1
6C
24th
8
29.7
27.3
119
122
131
124 1
89
185 1
82 1
8fi
25th
9
29.2
28.1
127
126
114
122 1
85
179 1
77 1
8C
26th
10
29.4
27.5
126
113
120 1
91
178 1
75 1
81
27th
11
29.9
27.9
120
125
140
128 1
90
184 ]
83 1
86
28th
12
29.1
28.1
127
135
133
132 1
97
196 ]
93 1
95
29th
13
29.7
27.8
116
114
125
118 2
08
192
2
0G
30th
14
233.2
231.8
31st
15
235.0
232.0
16
238.0
235.1
126
129
*124
128 '.
17
18
'Calculated volume of dead space for the fourth period on April 19, 121 c.c; May 18, 133 c.c.
2The subject ended his fast with the taking of fruit juices and honey on the morning of May 15, after
tJ e conclusion of the respiration experiments.
178 A STUDY OF PROLONGED FASTING.
Considering each series by itself, the conclusion may be drawn from
both the morning and the evening experiments that there is no constant
change in the size of the dead space of breathing as a result of the fast.
Although there is more or less fluctuation from day to day, yet the
general average of dead-space volumes found at the beginning of the
fast is much the same as that toward the end of the fast. A number
of very low values may doubtless be explained quite well as follows:
On April 15 and 16 there is a markedly lower value for the dead space;
this is probably due to an especially large drop in the alveolar air
during the day, as one might perhaps expect, these two days being the
first two of the fast. As previously mentioned, the subject was much
excited on April 22, when the Haldane samples were taken, and as a
result the value obtained for the alveolar air was probably low; thus
the calculation makes the dead-space figures also too low. On April 27
and 28 there is also an indication of a lower dead space; this is like-
wise probably due to a change in the alveolar air, as it will be noted that
table 20 shows a marked fall in the alveolar carbon-dioxide tension
about this date.
DIFFERENCE IN MECHANICS OF RESPIRATION IN MORNING AND EVENING.
On comparing the values for the dead space calculated from the
evening experiments with those for the morning experiments, one finds
a constantly higher dead space, which is, on the average, 55 c.c. This,
of course, is based on the assumption that the alveolar percentage of
carbon dioxide is the same in the morning as it is at night. Since some
physiologists believe that the dead space is always essentially the same,
it seems desirable to consider how large a difference in the alveolar
percentage of carbon dioxide must have existed between the morning
and evening experiments to indicate so marked a change in the venti-
lation. Assuming for the size of the dead space the figure 120 c.c,
which represents approximately the mean value previously calculated
for the dead space in the morning experiments, the alveolar air has
been computed for each respiration experiment, as shown in columns
b and d and the averages in columns c and e. The alveolar percentage
of carbon dioxide (carbon-dioxide tension) in the morning experiments
shows, in general, the same changes that the alveolar carbon-dioxide
tensions by the other methods have indicated. In fact, the alveolar
carbon-dioxide tensions obtained in this manner very satisfactorily
supply the values for April 25 and 26, when the alveolar air was not
taken by the other methods. The close agreement of the values for
the alveolar percentage of carbon dioxide in the several duplicate experi-
ments gives evidence of the even and normal respiration of the subject
and the great care in making the respiration experiments.
If the alveolar carbon-dioxide tensions of the morning and evening
experiments of the same day are compared, an average difference is
ALVEOLAR AIR. 179
found of 5.4 mm. (0.7 per cent). It is difficult to interpret accurately
this change in the mechanics of ventilation, but it is clear that it exists,
for with a given output of carbon dioxide, the respiration volume is
much larger in the evening than in the morning. Two causes for this
are possible — one, a lower percentage of carbon dioxide in the alve-
olar air, the other an increased volume in the dead space of respira-
tion. Both causes may be in part responsible for the difference. If
the alveolar air had been taken in the evening and in the morning at the
same time as the respiration experiments, the exact cause could have
been located; but unfortunately the data are not available. However,
this change seems of sufficient importance to summarize possibilities.
A lower alveolar carbon-dioxide tension in the evening will mean,
perhaps, a higher acidity of the blood toward evening, or possibly a
respiratory center more sensitive to a given stimulus. FitzGerald and
Haldane1 have noted that the alveolar carbon-dioxide tension falls as a
subject becomes mentally tired. Ordinarily this fall would not be
noticed during the day, as the food eaten tends to raise the alveolar
carbon-dioxide tension and thus renders the figures uncertain; but in
a one-day fasting experiment with myself as subject, I failed to find
any fall in the alveolar carbon-dioxide tension up to 4h 30m p. m., when
the experiment stopped.2 Changes in the dead space have been reported
by Douglas and Haldane,3 who state that with muscular work a
larger dead space is found, which would lead to the general conclusion
that the dead space increases with increasing metabolism. Krogh,4
however, using different experimental methods, maintains that the
dead space is practically always the same. As the metabolism in the
evening experiments is only about 10 per cent higher than in the morn-
ing, it seems unlikely that this higher metabolism would of itself cause
the change in the dead space, especially as the subject was at complete
rest and in the same position in both cases. In experiments carried
out by Dr. J. H. Means, of the Massachusetts General Hospital, and
myself, we have found changes in the dead space as the result of drugs
which correspond very closely to such changes as may have occurred
here. It seems quite possible that the bronchi became dilated in the
latter part of the day, having lost some of their tone with increasing
fatigue. Any changes in the dead space which may have occurred
in the experiments with L. are of such size that they can readily be
explained by dilation of the bronchi.5
In this connection it may be well to state that the dead space given
in table 20 includes the volume of the inspired air, measured at 760 mm.
and 20° C, dry, which does not reach the alveoli, where active exchange
FitzGerald and Haldane, Journ. Physiol., 1905, 32, p. 486.
2Higgins, Am. Journ. Physiol., 1914, 34, p. 116.
•Douglas and Haldane, Journ. Physiol., 1912, 45, p. 235.
4Krogh, Journ. Physiol., 1913, 47, p. 30.
BSiebeck, Skand. Archiv f. Physiol., 1911, 25, p. 81.
180 A STUDY OF PROLONGED FASTING.
of gas takes place. As L. was breathing through the three-way valve
and nosepiece connected to the apparatus, the figures given are 30 c.c.
higher than his actual personal dead space, since the subject with
each respiration drew from the air system of the apparatus 30 c.c.
which did not reach either the respiratory passages or the lungs.
SIGNIFICANCE OF CHANGE IN THE ALVEOLAR AIR DURING THE FAST.
Finally, it is advisable to compare the data of the alveolar carbon-
dioxide tensions as the fast progressed. Columns c, f, and g, in table
20, serve this purpose the best. As one would naturally expect, there is a
drop in the alveolar carbon-dioxide tension with the increased acidosis
of the fast. The subject ate the last meal before fasting on the evening
of April 13; on the afternoon of April 15 the fall in the alveolar carbon-
dioxide tension is first noted. This fall is about 4 to 5 mm. (2 mm.
Plesch method) ; after this the alveolar air continues at about the same
level until April 27 or 28, when a second quite sharp drop, also of about
4 to 5 mm., is apparent. The new level is maintained with only slight
fluctuations until the end of the fast. With the taking of food again,
there is a rise in the alveolar carbon-dioxide tension, as would be
expected with the resulting diminution of acidosis. On the morning
of May 18 a slight fall in the alveolar carbon-dioxide tension is again
noted in connection with the respiration experiments before breakfast.
Possibly a part of the rise on May 15 and subsequent days with the
Plesch and Haldane methods may have been due not only to the dimi-
nution of acidosis, but to an effect similar to that since noticed in this
laboratory with the ingestion of food.1
In connection with the second sharp fall in the alveolar carbon-
dioxide tension, it is of interest to note several parallel experimental
findings. The chlorine excretion in the urine on April 27 and 28
dropped from a previously high level to a lower figure, at which it
continued for the remainder of the fast. A rise in the total volume of
the urine occurred also at about this time. The daily nitrogen excre-
tion in the urine for the first 10 or 12 days of the fast was slightly
over 10 grams; there was then a fall to about 8 grams, which was
maintained throughout the rest of the fast. It may be noted that this
drop in the nitrogen excretion is simultaneous with the drop in the
alveolar carbon-dioxide tension.
Since alveolar air is intimately associated with the acidosis of the
subject, one naturally looks for simultaneous changes in the factors of
the urine which are taken as indicators of acidosis. Thus the /3-oxy-
butyric acid seems to show a rise to a high level about the twelfth day
of the fast; such a change is difficult to judge, however, as the /3-oxy-
butyric-acid variations from day to day are quite large. The other
index, the ratio of ammonia nitrogen to total nitrogen, also shows a
Wiggins, Am. Journ., Physiol., 1914, 34, p. 117.
ALVEOLAR AIR. 181
simultaneous rise by reason of the increase of the ammonia and the
decrease of the total nitrogen.
From these data one may safely conclude that there is a marked
increase in the acidity of the blood in the fast, beginning on the second
day; the acidity then did not change markedly until about the four-
teenth day of the fast, when another decided increase in blood acidity
occurred. The recovery to normal acidity in the blood begins to be
evident in the first few days after the fast.
CONCLUSIONS.
The results for the alveolar air and dead space may be summarized
as follows:
(1) On the second day of the fast, the carbon-dioxide tension in the
alveolar air showed a drop from the normal value. It remained at this
new level until about the fourteenth day of the fast, when there was a
second rather sharp fall, after which no further marked change occurred.
Each of these falls is about 4 mm. Thus the blood acidity may be said
to have markedly increased on the second day of the fast and to
have remained at this higher level until the fourteenth day, when a
second increase occurred; there was no further change until the end
of the fast.
(2) There is no sign of an accumulative change in the size of the dead
space from day to day as the fast progressed.
(3) A change in the mechanics of respiration on the respiration
apparatus between morning and evening experiments during the fast
shows that there was either a marked change in the alveolar air or
else a change in the size of the dead space during the course of each
day. If the former, the alveolar air fell about 6 mm. between morn-
ing and evening, returning during the course of the night to essen-
tially the morning figure. If the dead space changed, it increased in
size about 55 c.c. and again became normal by the next morning.
SUBJECTIVE IMPRESSIONS AND MENTAL ATTITUDE TOWARD
THE FAST.
By Harry W. Good all, M. D.
The freedom of speech characterizing this subject, his preconceived
ideas on fasting and on the humanitarian service of his fast led to excep-
tionally full comments on the whole project and specifically his subjec-
tive impressions. His habit of thought and introspection probably
make them of average value, though admittedly they are recorded not
as scientific observations, but as general indices to his mental makeup,
his personal experiences and his beliefs, as outlined freely to the writer
on each visit.
SUBJECTIVE IMPRESSIONS.
April 14, 1912 (18 hours after beginning the fast) :
The subject states that he is very happy in the thought that the fast has
actually begun. The value of the experiment to the world can not
be estimated, and after a few days the mind will be clear and active.
In explaining the influence of fasting upon the mind, he stated that
on his long journey from Malta he had been obliged to eat food that
was poisonous, more especially animal foods, and that his body was
saturated with this poisonous waste, making the mind dull and causing
a kind of physical fatigue. "When these waste matters are elimi-
nated the mind will be clear, and I will feel buoyant and hopeful."
In referring to the nocturnal emission which occurred during the
night of April 12-13, he stated that one of the most important things
noted in connection with his fasts was the behavior of the sexual
organs. During fasting there is a reversion to the animal type, a
periodicity of sexual desire at monthly intervals. In speaking of his
subjective sensations since beginning the fast, he states that he feels
perfectly well, has had no sensations of hunger, and no thought of
food. He has been a mouth breather for years. Is troubled with
naso-pharyngitis and tinitus aurium. Both these conditions always
improve with fasting.
April 16, 1912 (third day of fast) :
Feels perfectly well. Has had no sensation of hunger, and no abdominal
sensations, aside from slight rumbling of gas in the intestines. Has
passed very little odorless gas by rectum; there has been no belching
of gas, no sense of fatigue. Mind is not yet clear enough for active
mental work.
April 18, 1912 (fifth day of fast) :
Feels perfectly well. No sensation of hunger. No longing for food, but
occasionally thinks of the agreeable taste of ice cream. No sense of
muscular weakness or fatigue. Has passed a little odorless gas by
rectum. Naso-pharyngitis and tinitus better.
April 20, 1912 (seventh day of fast) :
Mentally depressed yesterday and to-day. He attributes this to the rain
and cloudy weather, as he always feels depressed when the sun does
not shine. Feels as well physically as usual. Expressed his satis-
faction at the manner in which the fast was being conducted.
182
SUBJECTIVE IMPRESSIONS AND MENTAL ATTITUDE. 183
April 22, 1912 (ninth day of fast) :
Feels well and hopeful again. Only complaint is the sensation of a dry
coating of the pharynx and a bad taste in the mouth. No sensation of
hunger. Not conscious of his stomach. Has passed a little odorless
gas.
April 24, 1912 (eleventh day of fast) :
Is conscious of slight muscular weakness but otherwise feels well. No
loss of ambition. No sensation of hunger and reading about food
does not stimulate a desire to eat. There is no dryness in the pharynx
today. Still passes a little odorless gas by rectum. Says he is thirsty
for the first time.
April 26, 1912 (thirteenth day of fast) :
Still conscious of slight muscular weakness. Has no special inclination
for mental work. Still has the sensation of thirst. No sensation of
hunger. Still passes a little odorless gas by rectum. Expressed
satisfaction at the progress of the fast.
April 28, 1912 (fifteenth day of fast) :
No depression to-day. Mind is clear now. Desires to study. No sen-
sation of hunger. Mouth, which has tasted bad since the ninth day
of the fast, is now improving. Passes very little odorless gas by
rectum. Had nocturnal emission at 6 a. m.
April 30, 1912 (seventeenth day of fast) :
Mind is clear. Feels well. Ambitious to study. Conscious of slight
muscular weakness. No sense of hunger.
May 2, 1912 (nineteenth day of fast) :
Feels somewhat weaker physically, but mind is clearer and can do better
mental work. Thinks the poisons of the food ingested previous to
the fast are about eliminated now. Complains of a "bilious taste"
in the mouth to-day. No sensation of hunger.
May 4, 1912 (twenty-first day of fast) :
Feels practically the same as he did May 2, except that he is a little
depressed by the cloudy weather and by remaining indoors.
May 6,1912 (twenty-third day of fast) :
Feels a little brighter to-day. Conscious of physical weakness. Mouth
does not taste so bad. No desire for food.
May 8, 1912 (twenty-fifth day of fast) :
Feels very well. No desire for food. Still passes a little odorless gas by
rectum.
May 10, 1912 (twenty-seventh day of fast) :
No change from last note. Still has a bad taste in his mouth.
May 12,1912 (twenty-ninth day of fast):
Sense of physical weakness, but to-day is very ambitious for his studies
and writing. Feels more hopeful and clearer mentally. Pleased
with the progress of the experiment. No desire for food.
May 14, 1912 (thirty-first day of fast) :
Depressed mentally because he has to break his fast tomorrow. States
that the fast should not be broken until the tongue has become clean
and a desire for food has returned. This in his opinion would take
several more days. To break the fast at this time is injurious.
Aside from this he is pleased with the manner in which the fast has
been conducted, and congratulates himself that he has been able to
go through with it successfully. Says the time has passed very
quickly. To-day he feels well. Has no particular weakness, but has
been conscious of some muscular fatigue since the eleventh day of
184 A STUDY OF PROLONGED FASTING.
May 14, 1912— Continued.
the fast. This, however, is no more marked than it is many days
when he is taking regular meals. The most pronounced physical
change is a sensation that his body is very light. This has gradually
developed as his weight has decreased, and necessitates his measuring
his steps when he walks. Says that at no time has he felt like
reclining on account of body fatigue. His neurasthenia is greatly
improved, and during the entire fast has only shown itself as slight
despondency and some irritability on the rainy days. His mind has
been much clearer throughout the fast than it is when food is taken.
To-day his mind is clear. He has better imagination. He is full of
hope and courage. Is ambitious to do mental work. Has had no
sensation of hunger, no sensations of faintness. He had no desire
for food, aside from the pleasant thought of ice cream on the third
and fourth days of the fast, and has been glad that he did not have
to eat. Has had no abdominal pain or discomfort. His naso-
pharyngitis is much improved, and his tinitus has practically dis-
appeared.
Maylld, 1912 (two hours after breaking fast) :
Extremely depressed and despondent because he had been obliged to break
his fast before his body was prepared for it. Felt very weak physi-
cally. The foods that he selected in breaking his fast were concen-
trated orange and lemon juice, grape juice, and honey. Experience
had taught him that these were the only natural, rational foods to be
taken at this time. Meat broths and other animal foods were poi-
sonous. In his opinion these natural foods should be followed first
by cooked fruits, then vegetables, and later a return to the ordinary
diet. Stated that he had no appetite, and that nothing tasted good,
although the lemon juice was not unpleasant.
10 a. m. Food was first ingested. This immediately caused a sensation
of warmth in the stomach, and he was conscious of a pulsation in the
epigastrium. There was no desire to belch gas, no nausea or other
symptoms until —
llh 45m a. m. when he began to have distress in the abdomen, starting in
the epigastrium and radiating towards the right hypochondrium.
The pain was dull, intermittent in character, but not colicky. He
describes it as a sensation of distension of the stomach accompanied
by rumbling and belching of gas. Furthermore, he believed he could
feel the progress of the distension of the alimentary tract as the
ingested food moved along the intestines, and he felt by —
12u 10m p. m. that the food had proceeded as far as the right iliac fossa.
Then at—
1 p. m. the rumbling and belching of gas was much more marked and the
pain was somewhat more severe. Marked lassitude and depression.
Soon after pain became easier, until —
4h 15m p. m., when the character of the pains changed from the "pains of
distension" to intermittent attacks of cramp-like or colicky pain,
which became very severe. These were experienced about the um-
bilicus and in the lower abdomen. Between the attacks he was very
drowsy. The colic gradually increased in severity until —
5 p. m., when he defecated for the first time. This was accomplished
without difficulty or pain. There was some gas, and the movement
had a very bad odor. After this he was free from pain until —
5h 45m p. m., when the cramps began again, increasing in severity, the
location being the same. This time the pain made him extremely
SUBJECTIVE IMPRESSIONS AND MENTAL ATTITUDE. 185
May 15, 1912— Continued.
weak. There was profuse perspiration and intense thirst. This
continued until —
7h 15m p. m., when he had a second movement of the bowels, liquid in
character. After this he was nauseated, and the pain continued
severe until —
9h 4&m V- m-> when he vomited and then began to feel better, and at —
11 p. m. the pain stopped.
May 16, 1912 (second day after breaking fast) :
About 2 a. m. began to have severe abdominal cramps again, continuing
until —
2h 30m a. m., when he had a third movement of the bowels, liquid in
character. After this he was comfortable, but passed gas at frequent
intervals. The odor of the gas was very disagreeable, but the passage
was not accompanied by pain. Was comfortable until —
8 p. m., when he stated that he was in no physical distress, but that he
was very despondent. He appeared hysterical, crying a good deal.
He was offended because he had been obliged to break the fast, and
complained that this, together with a few disagreeable experiences
with some of the men in charge of the experiment, and the lack of
fresh air and exercise throughout the fast, was responsible for his
physical and mental weakness. He attributed the abdominal pain
to the presence of gas. Inasmuch as he had not defecated during the
fast, "a hard plug of intestinal secretion had accumulated in the
rectum, and when he took food the advancing bolus compressed the
gas present in the intestines, causing pain."
May 17, 1912 (third day after breaking fast) :
Feels very weak, but no abdominal pain for past 24 hours. Passed a
fairly comfortable night, but did not sleep as well as usual. Is still
depressed and emotional, but to a much less degree since the pain
stopped. Is taking the same kinds of food, but well diluted with
water, as was suggested to him. Feels much better physically,
but has no ambition for mental work. Late yesterday afternoon
he was very thirsty for about an hour, drinking 5 glasses of water.
Up to the present time he has eaten the juice of two lemons diluted
with equal parts of water and sweetened with a teaspoonful of honey,
also 1,500 c.c. of orange juice diluted one-third with water, to which
enough honey was added to make it sweet. Says his general condi-
tion is not nearly so good as it was on the last day of the fast. His
appetite is just beginning to return, but he is not hungry yet.
May 18,1912 (fourth day after breaking fast) :
Extreme mental depression. Marked excitability, hysterical in character.
Weeps when spoken to, and his voice is scarcely audible. Complains
of general weakness, almost prostration. Sensation of trembling all
over the body. Did not sleep one moment during the entire night.
Beginning at 5 a. m. he had three loose movements of the bowels
preceded by severe colicky pains. He attributes this diarrhea to
worry. Can not bear the thought of food to-day. Explains his
mental state as being due to his general dissatisfaction at his treat-
ment at the laboratory. Later in the day, at his own request, he
was sent to a private room in the Massachusetts General Hospital.
Upon admission his condition was fully described by the writer to
the resident physician and the visiting physician.
186 A STUDY OF PROLONGED FASTING.
May 19, 1912 (fifth day after breaking fast) :
Was seen at the Massachusetts General Hospital at 9 a. m. At the time
he was sitting on one of the verandas reading. While he still appeared
somewhat emotional, he said he was no longer depressed, but was
in a most cheerful state of mind. He expressed his regret for the
trouble he had caused at the laboratory, and said he was so nervous
and irritable that he did not realize just what he was doing. He sent
his apologies to Professor Benedict, and wished me to say that he
would be glad to go back to the laboratory and go into the calorimeter
again, if it would add anything to the value of the experiment.
When asked about his comfort at the hospital, he said he was very
much pleased with the care and attention he was getting. His
appetite had returned, and he had relished his breakfast of "Boston
baked beans," which he ate for the first time.
May 20, 1912 (sixth day after breaking fast) :
The hospital reported that he had voluntarily left the institution. His
reason for leaving could not be learned, but in so far as could be
determined he had no grievance.
October 19, 1912 (five months after breaking fast) :
On account of the sensational stories which were circulated, an effort was
made to obtain as much information as possible regarding the personal
experiences of the subject after leaving the hospital. He was first
questioned as to the reason for leaving the hospital so abruptly.
First of all he stated that he was not sincere in his remark regarding
the hospital, as noted under the date of May 19. His reasons for
misrepresenting the true condition was a fear that he would not get
good treatment if he made any complaints. He was quite dissatisfied
at the hospital. In the first place, his pride was injured in that he
was not given the attention which was due an individual of his stand-
ing. One grievance was the fact that he was recorded on the chart
as "A laborer from the island of Java." Another was the fact that,
although he was in a room by himself, there were a good many rooms
in the ward and he was obliged to use the common toilet and bath-
room. Still another was the fact that the physicians at the hospital
had no experience in fasting and did not order the proper food for
him. The day following his admission he was given milk and eggs,
and he believed that these animal foods would poison his blood and
spoil all the benefits of the fast. The ingestion of the food was
followed by cramps and diarrhea, and he thought he was delirious.
The following morning (May 20) he looked at his chart, and finding
his temperature recorded at 99 degrees and his pulse-rate 90, he
knew he was being poisoned and hastily left the institution.
When he reached the street he did not know where to go, inasmuch as he
was entirely unfamiliar with Boston. Being a newspaper editor
himself, he thought he would find trustworthy advice as to where to
go by consulting an editor of one of the Boston papers. He then
took a cab to the office of one of the newspapers, the editor of which
engaged a room for him at one of the large hotels. In his room at
the hotel he was visited by some of the staff of the paper and talked
the situation over with them. He was surprised to find an account
of his fast, with photographs of himself, in the paper the following
morning, as he thought he was speaking in confidence and he was
not aware that photographs were being taken.
SUBJECTIVE IMPRESSIONS AND MENTAL ATTITUDE. 187
October 19, 1912— Continued.
After spending one night at the hotel, he was disinterestedly ( ! ) advised
by the editor to go to some secluded place to avoid annoyance from
reporters from other papers. In accordance with this advice the
editor arranged for his care at Bridge water, Mass., agreeing to take
care of his expenses for three weeks, and later start him on a lecture
tour in return for information concerning the experiment and his
treatment at the laboratory. He went to Bridgewater, accompanied
by a reporter who visited him each day, and was delighted with the
place. On the way down he talked freely with the reporter. The
next morning he was given a copy of the paper, and upon reading the
article which concerned himself, he became angry and excited,
because he did not know the things he said were to be published, and
declared that he never made some of the remarks pertaining to
Professor Benedict. He told the reporter that his action was dis-
honorable, and that if such reports were to be continued he would go
to another paper with his story. It was then agreed that the paper
should only publish what he himself wrote and signed.
After this the paper was not sent to him, and it was 5 days before he could
obtain the copies. To his surprise and anger he found they had not
published the best part of his stories, and had put in things that he
did not write and which were detrimental to those in charge of the
experiment. He again protested, saying it was not honest treatment,
and finally refused to have anything more to do with the paper. The
newspaper defrayed his expenses for 9 days, and then he went to the
house of a fellow countryman in East Boston. After this he gave
some lectures, and later was taken up by the Esperanto Society.
Later still a gentleman became interested in him, and offered to defray
his expenses while studying medicine. At the present time, i. e.,
date of this examination October 19, 1912, he is a student at the
Harvard Medical School. Each morning on his way to the School
he passes Professor Benedict's window in an automobile, which, he
said, with much satisfaction, is the " irony of the case." He concluded
his remarks by saying he had encountered a good deal of trouble with
various persons he had come in contact with since arriving in America,
but he felt now that for the most part it was due to misunderstanding.
MENTAL ATTITUDE OF THE SUBJECT TOWARD THE FAST.
The mental attitude of the subject was noted at each visit during
the entire period of observation. The predominating idea with him
throughout was that fasting is always beneficial. He believed that
it is followed by physical and mental well-being in normal persons
and is the rational treatment for diseased conditions. His own expres-
sions were: "Food impedes the body and mind and animal food is
poison. * * * I am anticipating the fast with much pleasure, as
the poison of the food I have eaten will be eliminated, my body cleaned
of its impurities, and my mind will be free and active. * * * The
neurasthenia and depression from which I have suffered for years will
leave me and I shall feel free and light and full of hope." Any dissent
from these ideas was promptly resented with such a remark as: "My
188 A STUDY OF PROLONGED FASTING.
dear doctor, you know nothing about fasting, while I have made it a
scientific study." He considered himself an authority, qualifying the
assertion by saying that he is now and always has been a student, that
he has repeatedly fasted, carefully watching the effects of fasting, and
that he has studied all the available writings upon the subject. He
stated that his object in undertaking the present experiment was to
accomplish the most complete fast yet undertaken, under the best
scientific conditions possible to obtain, not for his own enlightenment,
but to demonstrate to the world beyond doubt the truth of his theories.
He said : " The experiment I am about to undertake will be of the great-
est benefit to mankind."
Upon careful questioning it was learned that he had never undertaken
a fast under strict scientific observation and that his reading had been in
the main confined to non-scientific works, largely magazine articles.
On the occasion of the first visit it was evident that he was not only
willing but anxious to assist in every way possible the work that was
being done. He was very cheerful and was plainly pleased with atten-
tion shown him. This cheerful attitude continued for the first week.
He was interested in what was being done and apparently tried to
describe his subjective feelings with exactness. If any attempt was
made to oppose his ideas as to fasting, his usual smile disappeared
quickly and he assumed a sober, slightly injured air.
On the seventh day he appeared downcast, his movements were
less active, and he was decidedly depressed mentally. He attributed
this to the cloudy weather and rain and in so far as could be determined
no other reason for the change existed. He expressed no displeasure
at the manner in which the experiment was being conducted. On the
ninth and eleventh days he was again cheerful, but not so enthusiastic
as he had been during the first week, and he moved about as though
he felt some physical fatigue. He admitted that he experienced
muscular weakness. Any conversation, however, which was pleasing
to him would arouse his enthusiasm for a short time. From this time
up to the twenty-ninth day of the fast he was frequently depressed,
but always courteous and ready to submit to the examination. On the
days with sunshine he was always more cheerful, but at no time was he
as enthusiastic as during the first week. He appeared to be slightly
fatigued most of the time after the first week. His movements in
preparing for the examinations were more deliberate, and any attempt
to hurry him was politely resented. After the first week he gradually
became more sensitive to discomfort and pain, complaining of any
unusual pressure of the stethoscope or pressure of the hands in palpat-
ing the organs. He frequently spoke of the annoyance of the rectal
thermometer and of the adhesive tape used in retaining the stethoscope
on the chest wall. This annoyance was plainly shown in the expression
of his face and in the careful manner in which he moved about when the
SUBJECTIVE IMPRESSIONS AND MENTAL ATTITUDE. 189
thermometer was in the rectum. During the periods of depression he
was frequently disinclined to talk, and was sometimes irritable. At no
time, however, did he object to the examination, and he always seemed
willing to do whatever was necessary for the success of the experiment.
On the fifteenth day he said that his mind was beginning for the first
time to become clear; on the nineteenth day he began to feel the desire
to do mental work; and from this time on he declared that his mind was
continually growing clearer. Certainly there was no outward evidence
of the truth of this statement. He appeared fatigued mentally, and
he neither understood nor answered the questions put to him so
promptly as he did early in the fast. As far as could be determined
there was no such stimulation for study as he had predicted. On the
contrary he seemed to be less occupied with his books and papers.
On the last day of the fast he appeared to be very sober and assumed
rather an injured air. His general attitude was a complaining one.
The fast was being broken contrary to his judgment, as it was harmful
to take food before the desire for food had returned. Then for the first
time he complained of his medical care, saying that while the examina-
tions were made in the most careful and painstaking manner no atten-
tion had been paid to his physical exercise and he had not been allowed
to go out in the fresh air as much as he should have.
Nothing was observed at any time which would lead one to suppose
that the subject experienced any sensation of hunger or any feeling of
distress in the abdomen throughout the entire fast.1
Two hours after breaking the fast he was seen seated at a table
where he was slowly eating his fruit juice and honey. His expression
was downcast and his features drawn. His voice was weak and he
spoke with deliberation. Notwithstanding his resentment at the break-
ing of his fast he willingly submitted to examination. He complained of
pain in the abdomen and sudden spasmodic changes in his expression
occurred at the time he said he was having pain. Palpation of his
abdomen, however, only slightly intensified the discomfort.
Twenty-four hours after breaking the fast the general expression of
depression was more marked. The voice was weak and he moved
about very slowly. At this time there was no evidence of any physical
discomfort other than lassitude.
On the third day after partaking of food he appeared in decidedly
better spirits. There was no expression of discomfort and he smiled
frequently during the examination. His movements were not so
deliberate. He attributed his bad feelings of the previous days to the
pain and discomfort he had suffered and did not appear to entertain
'Cetti on the fifth day of the fast complained of pain in the epigastrium from time to time, on
the sixth day of belching gas and of distress in the abdomen, on the eighth day of severe abdominal
pain, which disappeared on the ninth day after the bowels moved ; on the tenth day he complained
of feeling very weak physically and of being nauseated. No note is made of the gastro-enteric
condition in the case of Breithaupt and Beaute.
190 A STUDY OF PROLONGED FASTING.
such a strong feeling of resentment because he had been obliged to
break his fast.
On the fourth day after taking food his mental attitude had entirely
changed. He was very emotional, his voice scarcely audible. He
wept as he talked. His hands trembled and his face was bathed in
perspiration. He appeared weak physically, and while he made no
objection to the examination, any undue haste or unusual pressure
on the body made him complain. He evidently felt that he had not
been given the proper medical attention during the entire period,
although he had refused, at all times, to accept suggestion, except that
he dilute his fruit juice. This was because "the physicians had no
experience in fasting." He demanded that he be sent to a hospital,
where he could get the attention he needed in his present sick, weakened
condition.
The following morning, when visited at the hospital, his entire
attitude had again changed. He was sitting on a veranda reading, and
appeared delighted at seeing me. He was in a very cheerful state of
mind, but was still emotional. He moved about quickly, showing none
of the prostration of the previous day. He appeared to be very sorry
that he had been so unreasonable at the laboratory, was apologetic,
and expressed his willingness to return.
When seen five months after the fast was broken, he appeared rather
unhappy. In telling his story it was evident that he had had differ-
ences with nearly every person he had come in contact with. He was
plainly disappointed because the world had not given him the recogni-
tion due him for the sacrifice he had made for the benefit of mankind.
THE PSYCHO-PHYSIOLOGY OF A FAST.
By Herbert Sidney Langfeld,
Instructor in Psychology, Harvard University.
The subject of these tests was a man 40 years of age, of medium
height and slender. When not in conversation his manner was languid
and it is perhaps due partly to this that he seemed to lack physical
strength and vigor. In temperament he was of the decidedly emotional
southern type, sensitive, quick to anger, loquacious, credulous, and
fertile in imagination. This last characteristic is probably responsible
for the fact that the unusual appealed to him. Once having espoused
a cause or entertained an idea, he would hold to it tenaciously. He was
a man of a few fixed ideas or complexes, which formed the basis of his
mental life.1
The tests herein described lasted from April 11 to May 15 inclusive.2
Food was taken on April 10, 11, 12, and 13 and again on May 15.
The intervening 31 days were fast days. The psychological tests were
made at 5 p. m. each day and lasted 1 hour. During the half hour
before the tests the subject rested.
The psychological tests were made under unusually accurate and
complete control of diet and occupation. One factor important to
mental measurement was found to vary, that is, the mood of the subject.
As far as L.'s willingness to cooperate is concerned, there was nothing
to indicate to the experimenter a change in this attitude or that his
general interest in the work relaxed at any period of the series. On the
other hand, there is no doubt that he was happier during the first part
of the fast, rather depressed and silent in the middle, and somewhat
irritable and excitable toward the end, although this irritation was at
no time directed toward the tests. The greatest depression occurred
after a prolonged continuation of bad weather and very much decreased
after he was able to go out in the air. He was also much happier after
having received visitors. He himself remarked that the monotony of
the program was the most difficult thing he had to endure. As to
his physical condition he made few complaints. He felt well through-
out and insisted that he had no sense of hunger even during the first
days.3 The only discomfort of which he spoke was the coated con-
1An idea of his intelligence and interests may be obtained from the association reactions.
See Appendix II, pp. 222-229
2The tests on April 1 1 were tentative and are not included in the curves.
•This is contrary to the experience of most fasters. W. B. Cannon and A. L. Washburn (An
Explanation of Hunger, Am. Journ. Physiol., 1912, 29, p. 441) describe the feeling of hunger aa
follows: "Hunger . . . is a dull ache or gnawing sensation referred to the lower mid-chest
region and the epigastrium. It is the organism's first strong demand for nutriment, and, not
satisfied, is likely to grow into a highly uncomfortable pang, less definitely localized as it becomes
more intense." Further (p. 442): "There is abundant evidence, however, . . . that during
continued fasting hunger wholly disappears after the first few days." Professor Cannon has
recently informed the author that from what certain fasters have told him he believes that sensa-
tions of hunger may be absent from the beginning; that in fact some people may never have the
sensations of hunger as just described. 191
192 A STUDY OF PROLONGED FASTING.
dition of his tongue and the unpleasant taste in his mouth. It was
his idea that the fast should continue until this disappeared and it was
for this reason that he was loath to break his fast on the thirty-first
day.1 Although he seemed more feeble toward the end of the fast and
gave one the impression of a man convalescing from a weakening
illness, yet he was always able to walk without assistance and at no
time was it necessary to omit or alter a test through lack of strength on
his part. On May 15, the day he broke his fast, he suffered severe
colic, induced by the food he ate, and although tests were made the con-
ditions were most unfavorable. It had been planned to continue the
examination for several weeks longer, inasmuch as such tests would
obviously be of inestimable value for comparison with the fasting tests.
Unfortunately that was quite impossible under the circumstances and
an entire year elapsed before further records could be obtained.
Several factors influenced the selection of the tests. In the first
place the time was limited. There was only 1 hour daily available and
it seemed advisable to arrange for as many tests as possible during this
hour in order to obtain a good mental picture. It was therefore neces-
sary to choose short tests and also those requiring the minimum of
effort, as one test had to follow the other without pause for recupera-
tion. For example, prolonged tests for fatigue would have been of
great value, but they could not be considered. In the second place, the
fasting began a few days after L.'s arrival and little time could be
devoted to preliminary trials in order to obtain the best combination,
and the program once arranged could not be fundamentally changed.2
After consultation with Professor Raymond Dodge, a series of tests were
selected. A few days' experience, however, showed the necessity of
several alterations, and the revised program was as follows: (1) Rote
memory for words; (2) tapping test; (3) strength test; (4) tactual-space
threshold; (5) touch threshold; (6) free association and reproduction
reactions; (7) association reactions, genus-species; (8) association reac-
tions, noun- verb; (9) cancellation test; (10) hand- writing;3 (11) visual
acuity; (12) memory for words after 55 minutes. Later the touch
threshold, which was taken on the under part of the lower forearm
with a von Frey hair, was discontinued on account of the impos-
sibility of obtaining reliable results in a short period of time. The
association reaction genus-species was also omitted through difficulty
in finding sufficient reaction words of equal simplicity. In addition to
the tests L. was requested to describe all the dreams he had on the pre-
vious night.4 This was given before the visual acuity test. All the
thirty days were considered sufficient for the physiological tests and he was allowed one day
more to excel Succi's record.
2 A few minor changes were introduced.
3A superficial examination of the daily records revealed no change. A systematic examination
of the data has not yet been made.
4See Appendix I, p. 222.
THE PSYCHO-PHYSIOLOGY OF A FAST.
193
tests with the exception of that of visual acuity were made in a small
room free from disturbing influences.1
The general conditions of the experiments and the nature of the
tests having been described, each test will now be treated separately,
first as to the particular conditions and second as to the results.
MEMORY FOR WORDS.
Ten one-syllable words were chosen and these were read twice to
the subject, who recalled as many as possible immediately after the
the second reading. After 55 minutes the subject again attempted to
recall these words.
tVoof
words
W
A
li-
\
s
\
/
tJS.
/
I
\J
\:
\
1
k
1
:
\
\
i
\
i
\
\
\
1
\
!
\
i
r T ,. , , \^ //rimed/ate ■
J. Memory for »ords\ ,
•
1
'
£
" _, . . . , . .. (A Greatest No be fore f/rsr mistake-
II. Memory for crtgirsX _ . .
\B wttnon/y o/ic mistake -
A/oof
digits
H
9
/
s
A
:
\\
/
\
8
/
V
/
/
*,
\
(
/
\
/
\
/
\
/
//
\
/
\
^
\
\
/
\
/
\
*
\
1
/
\
/
\
IB
I ays 123456
t
fxrsf began
7 9 9 10 II li IS /* 15 IS 17 It 13 to II ti li 14 IS tS 17 U 29 30 31 32 33 3+ I t
Fasf ended Later tests
Fig. 23. — Memory testa.
From the curves (figure 23) it will be seen that there are marked
fluctuations, a circumstance which is always met with in mental tests
and which will be found in all the curves. It will therefore be only
possible to speak of general tendencies throughout. In the curve for
immediate rote memory (A) it will be seen that the poor record made
on the eleventh day (the third day of the test) only occurs once again,
and that on the twenty-fourth day, while a perfect score of the 10 words
was made 3 times and all of them during the last two-thirds of the fast.
It can be said that although the early records reoccur frequently
toward the end, yet the curve as a whole shows a slight general improve-
ment, but so slight that not much significance can be attached to it.
xIt is much to be regretted that time and conditions prevented tests for the threshold of audition
and smell.
194 A STUDY OF PROLONGED FASTING.
The curve B, indicating the amount of retention after 55 minutes, on
the other hand, shows a more or less steady improvement until near
the end of the series, and even when these last trials are included the
general tendency of the curve is decidedly upward. In 4 instances,
and these all in the last two-thirds of the series, the retention curve
crosses the rote memory curve, which means that on these days the
retention after the lapse of almost an hour was better than the im-
mediate memory. L., upon being questioned, was emphatic in his
assurance that he never thought of the words in the interim, so that
this relative improvement in retention was not due to any conscious
repetition during the pause.
TAPPING TESTS.
The instrument used was similar to the tapping-board described by
Whipple.1 It consisted of a board 12 cm. square and covered with
aluminum. This metal is not very well adapted for the tapping-board,
but it was selected for its lightness, it being thought quite probable that
the tests would have to be made toward the end of the experiments
with the subject lying down and the board resting on his chest. The
stylus also had an aluminum point. The records were taken on a
kymograph. The tapping lasted for 30 seconds and periods of 10
seconds were marked off on the records. The subject being left-handed
used that hand. As he was over-sensitive to cold during the fast he
wore, besides a heavy woolen undershirt, a heavy dressing-gown, which
added to the weight he had to lift. Neither the hand nor arm was
allowed to rest on the table during the tapping.
Curve III (figure 24) shows a gradual improvement for the first
6 days, when the maximum of the series — 215 taps or about 7 taps
per second — was reached. The curve then descends for the next
9 days, when the minimum of 170 taps was reached. From this
point to the end of the series there is a rise to a point just below the
maximum. This rise is not, however, gradual, but consists rather of 2
plateaus, one of 9 the other of 7 days, separated by decided jumps and
followed by a gradual but very marked end spurt of 4 days.
The initial improvement can well be due to practice in using those
particular sets of muscles, combined with increasing familiarity with the
work. This same rise also occurred in the dynamometer tests. The
drop, however, begins much sooner than in the dynamometer tests.
In fact, it ends in the former where it begins in the latter. One can
therefore hardly say that it is a matter of muscular fatigue. The first
explanation to suggest itself is a lessening in interest, and this is
strengthened by the fact that the drop occurs at that time when he was
most affected by the monotony of the routine work. In this test less
depends for improvement upon the increase in muscular power than
Whipple's Manual of Mental and Physical Tests, Baltimore, 1910, p. 101.
THE PSYCHO-PHYSIOLOGY OF A FAST.
195
in the dynamometer tests, the main factor being the rapidity of action.
We know that the rate of the reaction time is greatly affected by changes
in attention, and it is probable that the betterment in the muscular
control, which we may assume from the results of the dynamometer
tests did occur, was insufficient to offset this loss of interest. The
results of the last days confirm this assumption, for here we undoubt-
edly have the effect of interest in an end spurt, which, notwithstand-
ing the muscular fatigue undoubtedly present at this time, brings the
curve back to a higher level. In regard to the two plateaus referred
to above, it seems plausible to infer, from what we know of the causes
of plateaus in the learning process in acts of skill, that these sudden
A/o.of
")ays i
OS
\
1
\
It
/
\
-..
»
6
4-
1
e
-l
.
,
i
/
\
.'
1
\
\
/
K
""
^
J
•.
.
'
\
.
-''
\
s
1
.-■
\
i
^
\
;'
\
/
"-.
'
\
/
\
'■•-,
i
\
/
\
/
-
/
1
\
s
'
\
1
s
no
Ho
m
IBS
M
m
•
fast began
It It « 13 14- If IS 17 IB 13 iO U II a 2* tS i6 tl li i3 30 SI 3i 33 3* /
t
Fast ended L afer fesfs
Fig. 24. — Tapping tests.
rises to new levels are due to the learning of some new method or short
cut. Here the most obvious short cut is the lessening of the height
of the stroke.1
An examination of the difference curve (IV, fig. 24), which has been
obtained by subtracting the result of the last 10 seconds from that of
the first 10, still further confirms the assumption of a wavering in in-
terest. There is a gradual increase in the amount of this difference,
which indicates fatigue. This increase is particularly marked toward
the end, when the records are improving, which means that the improve-
ment is caused by a spurt during the first 10 seconds.
'This is an error which is bound to occur with this form of tapping-board. The writer has
therefore, recently constructed a board which regulates the height of the stroke, thus making it
a constant factor.
196
A STUDY OF PROLONGED FASTING.
In general, it may be said that although initial lack of interest1
and later muscular fatigue played a role, both factors being directed
toward a decrease in the amount of work, yet the will impulse toward
the end was sufficiently great to bring the curve back to its initial level
and almost to its maximum.
STRENGTH TESTS.
These tests immediately followed the tapping tests. The subject
stood and received the dynamometer, one of the Collin type, from the
experimenter, and pressing it, returned it to the experimenter. The
record was noted and the instrument returned. The interval between
trials was about a second. Ten trials were made with the left hand,
Days 113 4
fasf began
7 8 9 16 II It 13 I* 15 >t IT 18 19 U tl IZ « Z* 15 U 17 18 19 80 31 it )i 3* I t 3 * 5 6
fast ended Later tests
Fig. 25. — Strength tests.
followed by ten trials with the right. Both in the right-hand (VII,
fig. 26) and left-hand (V, fig. 25) curves there is an initial falling off,
which is more marked with the right hand. The left-hand curve,
however, continues to fall to the tenth day, when it takes a de-
cided drop, while the right-hand curve declines more gradually to the
ninth day, when it reaches its minimum. Both curves then rise to a
maximum, which is reached by the left hand on the sixteenth day and
by the right hand on the twelfth day (the record of the first day not
being considered in speaking of this maximum). The curves then
fall, the left much more than the right, especially in the middle of the
series, the former reaching its minimum on the thirty-first day. Both
'Against this suggestion is the fact that other tests did not show this lack of interest, but it is
quite possible that the interest varied with the different tests.
THE PSYCHO-PHYSIOLOGY OF A FAST.
197
curves show a slight end spurt. This is, as a glance at the curve will
show, merely a rough picture, there being decided rises and falls
throughout.
In interpreting the curve it must be remembered that L.'s left hand
is the practiced hand and it can therefore be assumed that the muscles
of that hand are the stronger. In fact, the results make this more than
an assumption, for the record of this hand is at all times decidedly
better than that of the right hand. The initial falling off is what one
must expect when the subject is not accustomed to the particular
muscular exercise. There is a great exertion at first, and the muscles,
skin, and subcutaneous tissue feel the unusual strain for several days.
The muscles least accustomed to exercise are the most affected, and
Lbs.
95
l
\
\
\
\
i
L.
/
s
/
X
/
\
/
\
/
V
_\
/
V
/
±
/
\
/
^
v
/
\
f
\
/
\
i
\
70
\
\
20
,'
i
,,
"x
12
/
\
;
>
^
,-■
■v
»
s
/
\
i
\
...
! i
/
\
/
\
,'
*«.
/
^„__
X
r.
/
\
i ;
.^
/
'
1
t
■
pt—
, Oays i z j + 5 6 7 j s in ii it a i* is i6 it i« if to ii it 23 24 is it a u » so it si 35 34 i imm
fast ended Later tests
fast JbegfO/?
Fig. 26.— Strength teats.
for this reason the right-hand record drops more than that of the left
hand. Then the muscles gradually recover and the effect of practice
begins to appear. Acting against the practice is the increasing fatigue.
The right hand being the unused hand gives practice more chance for
its influence and although fatigue never allows the curve to reach its
first day's record, yet the drop which soon begins is much more gradual,
as has been pointed out, than it is with the left hand, which shows
more clearly the effect of fatigue.
The difference curves (VI and VIII), which were obtained by sub-
tracting the average of the last three records of each day from the
average of the first three, help to strengthen the conclusions just
drawn. The rise of the difference curve at the same time as the fall
of the main curve means, of course, increasing fatigue, which shows
itself in a greater and greater drop toward the end of the daily series.
198
A STUDY OF PROLONGED FASTING.
This rise in the two difference curves is relatively about the same,
which means that the daily increase in fatigue is relatively the same
for the two hands. Further, if we glance at curves IX and X, fig.
27, we find additional indications in the same direction. This curve
is plotted from the first of the daily series of 10 trials. This trial is
least affected by fatigue and therefore shows the greatest influence of
practice. Here there is a gradual rise for the right hand until next to
the last day, while the curve for the left hand begins to drop where
it should according to our analysis.
In general, we may therefore say that fatigue appears in both hands
early in the series. The curve for the left hand drops far below the
record of the first few days. The curve for the right hand shows less
drop, due to the greater influence of practice, so that the two curves
tend to approach one another.
Lbs.
Days I z 3 * S « 7 t s i» u « u i* is te 17 it 1* to ti ei a « zs i* er it » 30 31 3i 33 3* / 1 3 4- s 6
Fas? began fas/ ended Later tests
Fig. 27. — Strength tests.
TACTUAL-SPACE THRESHOLD.
A pair of dividers with wooden tips were used as an sesthesiometer.
The threshold was found on the volar side of the forearm, about 4
inches from the elbow. The points were applied on either side of a
red-ink dot which was made on the arm on the first day and renewed
when necessary. The method of minimal change, with ascending and
descending series, was employed; 5 trials, excluding one-point "vexier"
trials, were made at each distance; 4 correct out of 5 was considered
the threshold.1
For the first few days the curve (XI, fig. 28) keeps the high level of 7
cm. On the seventh day there is a drop to 5.5 cm., then a slight rise to a
level of 6 cm. and a high threshold of 6.5 cm. on the fourteenth day,
followed by a fall to the minimum of 5 cm. on the twenty-second day,
which minimum is again reached on the twenty-sixth and thirtieth
xIt had been intended to call 3 out of 5 the correct threshold, but this was not found feasible.
The threshold is probably too high, but for the present purpose, where the change and not the
absolute threshold is being investigated, this does not matter. The curve shows no record for the
fourth and fifth days. The experimenter was absent on these days and the physician who kindly
volunteered his services did not deem himself sufficiently skilled in this particular test to undertake it.
THE PSYCHO-PHYSIOLOGY OF A FAST.
199
days. The final days show a rise to 6 cm. The decided drop on the
seventh day may be due to adaptation to the experiment, which in this
instance means the adoption of a definite and clear criterion of dis-
crimination. The drop in the middle of the series, after a more or less
constant level, may be due to a similar cause — that is, a change to a
better criterion. The rises in the latter part of the curve are never as
great as those of the first part, although on the last day the curve
again reaches 6.2 cm. This threshold had to be placed at 5 correct
judgments, as there was a jump from 3 correct judgments. This makes
the threshold probably too high. If we omit the first day and compare
the average of the period from the seventh to the twentieth day with
the average of that from the twenty-first to the thirty-fourth day we
find a difference of 0.4 cm. in favor of the latter period. We may say
Days I S 3 4 3 6 7 8 4p}/0 II IZ 13 !■*■ IS 16 h IS 13 10 Zl ZZ 23 Z4-ZS ZS Z7 18 Z3 30 31 32 35 34- I £ 3 + 5 i'
t
Fasf ended L afer /esfe
fxzsf began
Fig. 28. — Tactual-space threshold and visual acuity.
then, in general, that there is a very slight improvement in the dis-
criminating process, but that there is no end spurt, which latter, from
the very nature of the process under investigation, is not to be expected.
ROTE MEMORY FOR DIGITS.
The usual rote memory test was employed. Increasing series of
digits, beginning with 4 digits, were read aloud once by the experi-
menter to the beat of a metronome with 1-second intervals and were
repeated as far as possible by the subject. The combinations of digits
varied daily.
Curve II B, fig. 23, is obtained by taking the last series that con-
tains only one mistake, curve II A, fig. 23, by taking the number which
immediately precedes the one containing the first mistake. Curve A,
which gives a picture of the rote-memory process, shows two apexes of
maximal value near the middle and another on the thirty-first day.
200 A STUDY OF PROLONGED FASTING.
There is, however, a very low minimum in the second half of the curve
and a decided drop from the maximum of the thirty-first day. One can,
therefore, hardly speak of an improvement. The most that can be
said is that toward the end of the fast the subject was again able to
reach the maximum record of 10 digits obtained near the middle of
the series. From curve B we see that on the third day a mistake was
made at 4 digits, yet the retention is 9 digits; on the eleventh day a
mistake at 4 digits and a retention of 8, etc. It seems fair to assume
from these results that curve B represents in a rough manner the degree
of attention. It is only inattention that can produce results like the
above. Curve B shows a decided rise to the eighteenth day, when it
reaches a maximum, and although it follows a lower level from this day
it never reaches the minimum of the first third of the series. One may
therefore say that there is an improvement in the state of attention, at
least for this experiment, as the fast progressed.
ASSOCIATION TESTS.
The free-association experiments consisted of the daily presentation
of a list of 20 words, which were selected principally from the lists pre-
pared by Woodworth and Wells,1 and with the exception of the list
of May 9, which was a repetition of that of April 11, they were all
different.2 Several days after the tests were begun it was thought
advisable, in order to make the lists as uniform as possible, to have
them composed of an equal number of verbs, concrete nouns, adjectives,
and abstract nouns, in the order given. This arrangement was
adhered to from April 18 to the end of the tests, with the exception of
May 9. The words were read aloud by the experimenter and the time
taken with an ordinary stop-watch. The reproduction experiments
followed these with only a pause of a minute. Although the subject
was told that he need not repeat the same word, if it did not come at
once, yet there is little doubt that his efforts were always directed
toward that end. L. had a good command of the English language,
although it is not his native tongue, but at times he had difficulty in
finding the word he wanted. In such cases he made a gesture as soon
as the idea came to him and the watch was snapped at that time rather
than when the English word was found. This method of procedure
was not often necessary and it seemed a legitimate means of balancing
the slight disadvantage he had as a foreigner. A reserve list was pre-
pared upon which to draw when he did not understand the word of the
main list.
The curve (XIII, fig. 29) is plotted from the daily average. The
average was used in order the better to include the influence of the
'Woodworth and Wells, Association Tests, Psych. Monog., 1911, 3.
2The lists will be found in Appendix n, pp. 222-229. In a few instances the same word appears
in two lists.
THE PSYCHO-PHYSIOLOGY OF A FAST.
201
long times, which might very well be ol importance in these tests.1 The
few exceptionally long times, such as 20 seconds, which may have been
caused by emotional complexes, were not included.
The curve begins with very long reaction times. L. had never
performed such tests before, so that the sudden drop on the third day
■Sees.
Oays
12 3 +56
r
fast began
7 8 5 10 II 12 12 1+ IS 16 17 18 IS 20 21 22 23 24- 25 26 27 28 29 30 51 32 33 3-*
t
fast ended
Fig. 29. — Free association tests.
2 3 •+■ 5 6"
Later teste
Sees.
J
\
l.l
I
\
/
\
r JCIV. Assoc/at/on react/on to
Z.O
i\
\
\
1.3
I
\
1
\
/
\
Jl
IS
\
1
\
/
I
\
1.7
1
\
i
\
j
d
/
\
16
\
i
/
\
1.5
i
1
{
\
r
\
1.4
12
\
\
I
T
/
/
I
\
l
\
\
/
\
/
1
\
j
\
/
J
/
\
LI
1.0
\
\
/
\
Z.I
\
1
|
20
/
\
/
/
\
1.8
1.7
1.6
AS
1.4
a
12
l.l
1.0
/
i
/
\
I
\
/
\
/
\
\
/
s
/
\
\
1
1
I
j
\
\
\
/
/
\
\
\
1
j
\
A
\
/
\
/
\
\
\
/
\
1
1
1
'
\
1
f
\
/
\
/
\
/
\
1
1
/
Days
i
fa
S3 * S 6 7 » 9 10 II IZ 15 14- 15 it 17 18 19 10 11 22 Z3 £•* 25 Z6 27 28 29 30 31 32 33 34- / 2 3 *
T t
sr /began fast ended L ater
Fig. 30. — Association tests. Reactions to verbs and nouns.
+es
-
ts
5
JThe median, which was also calculated, gave the same general curve.
202
A STUDY OF PROLONGED FASTING.
must be attributed to the practice improvement, which at this early-
stage could very well be sudden and of considerable amount, rather
than to the fact that it is the first day of the fast. From this point the
curve descends with a few breaks to the fifteenth day, when it reaches
1.4 seconds; it then rises to the twenty-second day, when it reaches the
maximum (if we do not consider the first few days) and then falls to the
end of the series. On the second from the last day it reaches the min-
imum of 1.3 seconds. Also the record of 1.4 seconds is obtained 3 times
5+cs
QayS I 2 3 ■*■ 5 6 7 8 9 10 U 12 15 <4 15 K5 17 16 19 20 a/ 22 23 24 25 26 27 26 29 30 31 32333+12 3 4-56
t
fa sr ended L ater fesfs
fast began
Fig. 31. — Association tests. Reactions to adjectives.
Sees
26
ts
14
1.3
22
LI
ZJO
1.9
1.8
f.7
1.6
w
IA
1.3
12
II
Days I _
1
,
/
\
/
\
/
\
\
\
Y
\
\
/
/
\
\
\
\
/
\
/
\
/
\
\
\
\
z
\
;
\
\
/
\
\
\
!
\
1
\
/
\
J
\
/
\
7
\
/
\
\
\
f
\
s
/
\
\
\
1 i
J '
f
5
5
r
? i
i
0 1
i i
Z 1
3 /
f 1
S 1
e i
7 I
J /
9 e
oz
/ 2
2 £
32
4-2
5£
6 Z
7 2
8 2
9 I
0 3
1 3
?!
33
4-
i
I
'
y 1
1 (
Fas? began
Fia. 32. — Association tests.
fast ended Later fesfs
Reactions to abstract nouns.
in the second half of the series. If we include the first few days it can
be said in general that there is a very decided betterment in the associa-
tion times; and even if one calculates from the third day there is an
appreciable drop. Especially interesting is the almost steady improve-
ment shown in the last third of the curve.
In order to analyze the curve further, separate curves (XIV, XV,
XVI, and XVII, figs. 30 to 32) have been plotted for each of the four
THE PSYCHO-PHYSIOLOGY OF A FAST.
203
categories of stimulus words. It must be remembered that these
curves begin on the seventh day, when this division into separate
categories was first made. In consideration of the fact that the daily
average is obtained from only 5 reactions, too much importance must
not be attached to sudden daily falls and rises, such as in the abstract
series on the nineteenth and twentieth days and in the adjective series
on the eighteenth day, etc., but rather the convex shape of the verb
curve, the rise in the middle of the noun curve, etc., must be considered.
It is evident that the rise in the main curve about the tenth to thir-
teenth day is caused largely by the noun curve and that the relatively
greatest improvement at the end of the curve as compared with the
Sees.
/
\
/
\
'
/
\
/
\
A
\
/
\
/
\
/
\
/
\
\
/
\
/
\
7
\
/
\
/
\
\
1
\
/
\
/
\
\
\
\
]/
(9
/
1
1.8
/
/
1.6
/
/
/
/
\
I
\
/
\
/
\
\
I
\
/
\
1
/
\
/
v
I
\
/
\
/
'
/
\
/
\
/
\
/
.J
\
/
—
Days I Z 5 4- S 6 7 9 9 10 II IZ /J 14- /S IS 17 16 IS to Zl £z £3 Z-t tS « 17 it 39 30 31 $Z 45 Jf / 2 3 4-5 6
* T T
Fast began Fast ended L ater tests
Fig. 33. — Reproduction tests and mean variations.
beginning is in the abstract curve. On the other hand, the verb and
noun curves have several low averages in the beginning that were not
reached again. In fact, it is hardly possible to say that either of these
curves shows general improvement; certainly not the noun curve. An
examination of the daily fluctuations in the curve shows that the
fluctuation becomes less as the tests progress.
The curve (XVIII, fig. 33) for the mean variation of the main curve
shows a decided improvement as the fast progresses, with a very low
level on the last 3 days.
The reproduction curve (XIX, fig. 33) follows the tendencies of the
association curve. There is the initial drop and many more high peaks
in the first two-thirds of the series. If it were not for the rise on the
last two days the general betterment would be more marked. The reac-
204
A STUDY OF PROLONGED FASTING.
tions were, on the whole, rapid, averaging about 1 second and dropping
as low as 0.8 second. As the number of false reproductions was very
small (see I, table 21), amounting to only 23 in 680 reactions, or 3 per
cent, and never more than 3 in one list, an improvement or the reverse
in this respect would mean little. At least one can say that the quality
of reproduction suffered no deterioration with the progress of the fast,
but that retention was equally as good at the end as at the beginning.
Table 21. — Qualitative analysis.
II. Classification of reaction words in association
III. Mistakes in
experiment.
cancellation test.
No. of test.
I. False
repro-
ductions.
Misun-
Identity.
Persev-
Repeti-
Word com-
Omis-
Incor-
rectly
crossed.
derstood.
eration.
tion.
pounding.
sions.
1st
1
1
0
1
2d
2
1
1
1
0
Fast began.
3d
3
1
0
0
4th
1
3
2
1
5th....
2
1
3
0
6th
0
1
1
2
0
7th
0
3
1
0
8th
0
i
1
0
9th
0
i
2
0
10th
1
i
3
2
11th
1
2
1
12th....
1
l
0
0
13th
0
2
1
14th
0
i
1
0
15th
1
0
0
16th....
0
2
1
1
17th....
0
l
1
1
0
18th
0
4
0
19th
0
l
1
1
1
1
20th
2
2
0
0
21st
1
l
2
0
0
22d
2
0
0
23d
1
1
1
0
24th....
0
l
1
2
25th
0
0
0
26th
1
2
1
1
1
27th
0
1
3
0
28th
0
1
1
1
1
29th
0
1
0
0
30th
0
1
1
31st
0
0
0
32d
0
1
2
0
0
33d
2
2
1
4
1
Fast ended .
34th....
1
2
0
0
Later tests:
1st
0
. .
1
1
4
0
2d
0
3
1
0
3d
0
1
1
0
0
4th
0
1
5th
1
2
1
0
6
6th
0
1
1
1
0
THE PSYCHO-PHYSIOLOGY OF A FAST.
205
The quality of the association reactions was of high grade throughout
the main test (II, table 21). There were no senseless or pure sound
reactions and very few repetitions. Synonyms, word-compoundings,
and misunderstood stimulus words occurred seldom and were scattered
throughout the days. The word "woman" appears a number of times
and "man" slightly less often. There was also evidence of a religious
complex. An examination of the different categories did not show suffi-
cient change to warrant an analysis or tabulation as to quality. It was
thought that the introduction of words designating food might pro-
duce delayed reactions both with the word itself and the words immedi-
ately following. This was not the case. For example, on April 16 we
find egg-white 1.4 seconds; on April 19 omelet-eat, 1.4 seconds; on
April 21 fish-sea, 1.4 seconds; on May 7 candy-sweet, 0.8 second; on
Sees.
1 1 1 J I 1 1 1 1 1
\
c2
-JC2C. Association reaction noun-verb -
i
1
2.1
\
1
10
i
1.9
A 1
\
\
i
IH
"7VT7T
\
\
1.7
/
^
1 A
TT
\
\ ^
1
16
■
X
JlLA
/
\
\
/
\
/
1.5
1
3
/\
/
\
/
\
1-4-
I
\
/
\
/
\
A3
\
/
\
/
^
.
/
12
V
]
\
/
U
/
1.0
Lays I I 3 + S 6 7 9 9 la II II 13 I* IS U 17 IS 19 iO tl 12 Z3 Z* t5 it 17 16 19 30 31 31 33 34- I Z 3 * 5 tf
Fasf began Fast ended Later tests
Fig. 34. — Controlled association tests.
May 9 apple-fruit, 0.8 second; on May 10 roast-meat, 1 second; on
May 13 chocolate-sweet, 1 second. None of these reactions were
followed by unusually long reaction times. It might be of interest to
mention at this point the unusually long reactions which point to com-
plexes. On April 13 we find pulse-hand, 9 seconds; on April 21 death-
eternal, 22.4 seconds; and on April 26 uncertainty-pendulum, 12.6
seconds. These are the only extremely long reaction times. The
next longest is 6 seconds. All of these delayed reactions may be
explained from the same cause. L. had asserted that the chief factor
for a successful fast was faith and confidence and absolute lack of fear.
He thinks it is the fear combined with exposure which causes death
in shipwrecks and other calamities where food is not obtainable, and not
the actual lack of food. It is also claimed that those who fast frequently
cover their mirrors in order that they may not be disturbed by the
evidences of emaciation. One of the supposed dangers in fasting is
heart failure. If L.'s heart had shown alarming symptoms the fast
would have been terminated at once. It does not, therefore, require
206 A STUDY OF PROLONGED FASTING.
a stretch of imagination to suppose that L. would keep his mind from
such subjects as death and uncertainty and that he would even avoid
thought of the condition of his heart and that the mention of these
words would cause hesitation.
The determined association reaction noun-verb was begun on the
eighth day. Curve XX (figure 34) resembles that of the verb curve,
except that the rise continues longer. It starts very low (1 second),
increases with rather large daily fluctuations, and on the last day of the
fast returns to 1.1 seconds. A particularly disturbing factor in this
series was the fact that there was an ever-increasing difficulty to obtain
appropriate words. At first the words had obvious associations.
They were names of common objects, such as dog, gun, eye, etc., but
more unusual words had to be employed in increasing numbers, and
there seems no doubt that this circumstance was at least partly the
cause of the increasing length of the reaction time. It is even more
important in the determined than in the free-association experiments,
to have the quality of the words the same and not more difficult.
For long series of tests the free-association experiments are much to
be preferred.
CANCELLATION TEST.
Special forms were made for this test, consisting of type-written pied
text of 100 a's and 50 of each of the other letters of the alphabet. A
different combination was made each day, so that the subject should
not become accustomed to the order. L. was requested to cancel all
the a's. He used his left hand and the time was taken with a stop-
watch. Special care was observed to have the illumination constant
and the same pencil was employed.
The curve (XXI, fig. 35) represents the time for the completion of the
task. As in some of the other curves, so here we have the initial rise,
which continues to the sixth day, when there is a sudden drop to a
level which slopes slightly to another sudden drop on the twenty-ninth
day and a very low level for the final days. The difference between the
maximum of 3 minutes 48 seconds on the sixth day and the minimum
of 1 minute 53 seconds on the last fast day is very considerable. The
maximum is over double the minimum, and even if we compare the
minimum with the initial time of 3 minutes 7 seconds or with 2 min-
utes 43 seconds of the seventh day, which is the first and largest
practice drop, we still find a very considerable difference. There does
not seem any doubt, therefore, that there is very much of a betterment
in the time as the fast progresses and that this decrease in the time
continues to the end of the series. Nor is this improvement in time
gained at the sacrifice of accuracy. At no time were there many
mistakes made (see III, table 21). In fact, the degree of accuracy was
always so high that we can not place any importance on the slight
THE PSYCHO-PHYSIOLOGY OP A FAST.
207
increase of accuracy in the last half of the series1 nor does the slight
loss of accuracy at the minimum alter the significance of that result.
T/'/neT
—
f
t
/
s
/
/
"^
7
1
\
1
\
/
\
\
\
f
/
\
\
^
\
1
\
\
-~.
\
.—
\
1
\
/
s
/
\
\
/
\
/
^
'
\
/
1 .
Fa
'1
ft
> -
•
be
7<?/
r <
?
t
7 t
i
t 1
1 1
1 1
Z 1
3 /
* /
J- /
S I
7 /
3 1
i i
0 l
i i
Z i
3 2
* Z
5 2
6 11 28 Z9 30 31 32 33 3*
r
Fast ended
Later tests
f
3 5
3* 0
2-40-
est
tit
i o
Czys
Fio. 35. — Cancellation tests.
VISUAL ACUITY.
These tests were made in the large calorimeter room adjoining the
small room in which the previous tests were conducted. The largest
E which had been cut from the Schnellen test-card was used. This
was held by the experimenter at the level of the subject's head when
seated. It was illuminated by an electric lamp held by a second
experimenter in front of the card and moved with it. The shades of
the room were kept drawn during the experiment in order to have
constant illumination as far as possible. The subject suffered from
myopia and wore corrective glasses. A distance was first chosen well
within the threshold at which the subject was asked to judge in what
one of the four possible positions the E was being held. The experi-
menter put the card behind his back when he changed its position.
After a few days of the tests it was thought that the subject might be
using the secondary criterion of the distance of the edge of the E from
the edge of the card, the E not being exactly in the middle. The card
was therefore mounted on a larger cardboard of the same color in order
to obviate this possibility. On account of the surprising results, both
experimenters were at all times keenly attentive to the possibility of
other secondary criteria, but none could be discovered. Ten trials
were made at each distance, the card being moved from the observer
in steps of one foot. That distance was considered the threshold
which preceded the distance at which the subject made two mistakes
^here were 29 mistakes in the first half and 24 in the second half of the series.
208 A STUDY OF PROLONGED FASTING.
out of the ten trials.1 The alteration in the position of the E followed
no definite order, but every means was used in this respect to confuse
the subject in order to remove all possibility of his guessing the position.
Most of the judgments were made without hesitation, both at the very
low and very high thresholds.
The curve (XII, fig. 28) represents the daily threshold in feet.
There is a very rapid rise from the fifth to the fourteenth day, when
the maximum of 37 feet is reached. Then there follows a drop to 24
feet and a rise to 36 feet on the next to the last day of the fast. The
thirty-fourth day shows a drop to 19 feet. The record of the fifth
day is 16 feet, which is the minimum; that of the thirty-second day is
36 feet, which is 1 foot less than the maximum. This difference of 20
feet is very great for visual acuity. He saw twice as far at the end
of the fast as he did at the beginning.
LATER TESTS.
Owing to an attack of colic resulting from the nature of the first food
taken after the fast and the subsequent withdrawal from the labora-
tory, it was impossible to continue the tests during the recuperative
period, as had been planned. Only by later tests for comparison could
a decision be reached as to the efficacy of fasting. One year after the
tests just described L. volunteered as subject for a short series of tests.
These were conducted at the Harvard Psychological Laboratory and
extended over a period of 6 consecutive days. It was not possible to
arrange for them to take place at 5 o'clock as previously and 10 o'clock
in the morning had to be chosen. All the other conditions were
observed as closely as possible. The same tests, with the exception of
the visual-acuity test and the hand-writing test, were performed.
L. seemed in good health. He weighed about 126 pounds, which is
somewhat less than he weighed when he began his fast. His physical
appearance was, however, very much the same as on the day he arrived
at the Nutrition Laboratory. He had remained in America during
the previous year, engaged in medical studies, lecturing, etc., had
not fasted again, and had had no illness during that time. In com-
ing to the laboratory he made a journey of 4 miles and had already
had several hours' work, having risen each day at 5 o'clock, exercised
for half an hour, and made several visits. The conditions previous to
the tests are, therefore, hardly comparable to those of the former series.
It is evident, however, that he was as strenuous, if not more so, than he
had been up to the later hour of 5 o'clock of the previous tests.
The rote memory for digits (II a) was somewhat poorer than it was
during the latter part of the long series. It did not reach the maximum
1Lack of time prevented the threshold being taken in the reverse direction. The tests took
5 to 10 minutes.
THE PSYCHO-PHYSIOLOGY OF A FAST. 209
by two numbers, yet it did not show any poor scores. The curve (II b)
which represents the first mistake, or, as it was supposed above, the
state of attention, shows an improvement over the latter part of the
first series in that it does not drop so far. On the other hand, the rote
memory for words (I a) seemed as good if not better than during the
fast. It reached the former maximum on the fourth day and never
dropped below eight words. The memory after 55 minutes (I b) was
as good as the immediate memory. From these results it may be
concluded that the memory was still, after the year's interval, at about
the level that it was at the end of the fast.
The curve (III) for the tapping begins considerably higher than the
maximum of the fasting tests, and although it drops somewhat, still it
remains above the former maximum. The drop in the difference
curve (IV) is caused principally by a falling off in the initial spurt.
This is concluded from the fact that the results of the last 10 seconds
vary much less than those of the first 10.
The results of the first day of the tactual space threshold can not be
utilized as a comparison (XI). The unusually high threshold was
undoubtedly caused by inattention on the part of L., who admitted
that he had been very much worried over an appointment he had been
forced to miss and upon which his mind had been during these tests.
Apart from this day the curve has the same form it had during the latter
part of the previous trials. The second and third days show the mini-
mum, which was last reached on the thirtieth day of the former trials.
The dynamometer used in the previous tests could not be obtained
until the second day. There are, therefore, only 5 records. The
curves for both the right hand (VII) and the left hand (V) begin with
very high records and drop considerably on the second day, just as they
did in the former series. These first records are very much better than
any made in the previous trials. Even after the drop the right hand
twice surpasses the previous maximum and remains close to it on the
other days. The difference curves (VI and VIII) show that on the first
day the high record for the left hand was made by a sustained effort.
The right-hand spurt caused fatigue toward the end. The large dif-
ferences during the next 3 days for the left hand were caused by spurts
followed by fatigue, that of the right hand by fatigue. It is seen that
the strength of the muscles of the hand had very much increased since
the end of the fast, and judging from the first day's results was much
greater than at the beginning of the fast. One acquires a knack in
gripping the instrument and it may be that this was carried over from
the former tests and made these initial records higher than those of the
year before. In other words, some of the effect of practice was still
present and influenced the results much more than it did when it had
the opposing effect of fatigue.
210 A STUDY OF PROLONGED FASTING.
The free-association reaction time (XIII) begins at the low point
of the last day of the previous series; on the third day it reaches the
shortest time of that series, and again on the fifth day and on the last
day it falls almost one-fifth of a second below this point. That is, the
curve continues the descent it began in the middle of the former series
in as regular a manner as if a year had not intervened. Inasmuch as
some practice was necessary after so long an intermission, it may be
said that the reaction times were better than they were at the end of
the fast. The m. v. (XVIII) was 0.5 second on the first day and 0.15
second on the sixth, with an almost steady decline.
The average reproduction time (XIX) is 0.9 second for all the days;
this is very low, and although 0.8 second was reached 3 times in the
former tests, it is safe to conclude that the reproduction times are at
least as good as they were at the end of the fast. In fact, the average
for these days is better than for any 6 consecutive days of the previous
tests. There was only one false reproduction and that was "wrong"
for "bad." In view of the fewness of the trials little would be gained
by an analysis of the results according to categories XIV, XV, XVI,
XVII. The noun and adjective curves are lower than the verb and
abstract curves. The quality of the reactions is about the same.
Evidence of a religious or mystic complex is as plain here as in the pre-
vious results. "God" was the reaction for "adore," "worship,"
"unseen," "mercy," "Divine," and "Infinite;" "supreme" gave
"Being," "sacred" gave "church," "adorable" gave "saint," "life"
gave "eternal," and "ornament" gave "church." There were no very
long reaction times. In connection with the previous complex it may
be mentioned that "death" was the reaction word for "fear."1
The reaction noun-verb (XX) begins at the average of the thirty-
second day of the former series and on the third and fourth day reaches
the minimum of the next to the last day of the long series. The average
of these days is very much better than that of the last days of the fast
series or even of the first days, so that there is no doubt of an improve-
ment in these reactions.
The cancellation test (XXI) begins at about the point of the twenty-
seventh day and the time gradually decreases, but at the sixth day
has not reached the rapid time of the last fast day; but judging from
the slope of the curve, one would expect it to do so shortly, so that one
can conclude that the mental functions necessary for this test are in
about the same state they were at the end of the fast. There were only
6 mistakes, 4 of them being on the first day.
*It was thought that a year's intermission would make the old lists equivalent to new ones and
as one would then be sure of having the lists of this series of the same quality with those of the
former, the old lists were used on the first day, but 7 of the 20 reactions were the same as those
made a year ago, so that new lists were made.
THE PSYCHO-PHYSIOLOGY OF A FAST. 211
CORRELATIONS.
It would be supposed that there would be very good and very poor
days upon which all the curves would show proportionate increases or
decreases, or that at least similar tests, such as those of the higher
mental processes, would show similar variations. If we compare
some of the crests and valleys, however, we arrive at negative results.
For instance, on the twenty-second day the association time (XIII)
is long and both memory curves (I a and II a) are in a valley, but the
cancellation test (XXI) shows improvement, and the reproduction
times (XIX) are not long. On the sixteenth day the left hand reaches
a maximum in the strength tests (V), but the right hand (VII) shows
no such result. Tapping (III) rises on that day, but it is still compara-
tively low; one memory curve has fallen (II a) and the association time
(XIII) has risen. On the fifteenth and seventeenth days the memory
curve (II a) is at a maximum and the association time (XIII) is also lower;
the cancellation test (XXI) is also low on these days, but the maximum
of the memory tests (I a, II a) on the thirty-first day finds the associa-
tion times (XIII) longer. On the twelfth day the curves for the
strength tests (V, VII) have risen for both hands — it is the maximum
for the right hand — the time for the cancellation tests (XXI) has short-
ened and memory (I a) is better, but the tapping record (III) has fallen
and both association (XIII) and reproduction times (XIX) are at a
peak. The considerable lengthening of the time of the cancellation
test (XXI) on the sixth day finds a betterment in most of the other
tests, the tapping test (III) indeed, having reached its maximum on
that day. The visual acuity curve (XII) rises abruptly to its maxi-
mum on the fourteenth day and, although with a few exceptions the
curves show a slight betterment, the rise is comparatively insignificant.
It must be concluded, therefore, that with the exception of the last
day the daily fluctuations can not be traced to any one cause, such as
a general bodily fatigue and depressed mood or vigorous and cheerful
mental states, but that either there is a change in the one or more
processes essential to the particular test that is showing the exceptional
rise or fall or that there has been a momentary wave of fatigue or dis-
traction or spurt, etc. A diary of the fast was kept in which every
important incident was noted and it is possible that many of the fluctu-
ations in particular curves or changes in general tendencies of several of
the curves could be more or less satisfactorily explained. The follow-
ing considerations, however, make such explanations of doubtful value.
One can not say in advance what the effect of visits or other changes
in the general routine may be. Much depends upon the particular
circumstances. Now, if the results were better after a certain visit,
one could say that the subject was in a pleasant mood after the break
in the monotony of the days and that his mind had been stimulated by
agreeable conversation. If the results were worse on those days, one
212 A STUDY OF PROLONGED FASTING.
could say with equal weight that the fatigue following the unusual
exertion was the cause. Only the most reliable introspection on the
part of the subject before and after each test could have given strength
to such explanations, and both the lack of time and training on the part
of the subject made such a procedure impossible.
It did seem possible, however, to make an exception of the days on
which L. took a drive or was allowed on the roof and that if the curves
showed an agreement in their fluctuations on these days an unequivocal
explanation could be found. The drives were taken on the fourteenth,
seventeenth, twentieth, twenty-second, twenty-fourth, twenty-ninth,
thirty-first, and thirty-second days; the visits to the roof on the tenth,
fifteenth, twenty-first, and thirtieth days. As was stated above, there
was no general agreement even on these days. In regard to the individ-
ual curves, however, the visual acuity curve seemed to show the influence
of the drives. The best result in the visual acuity test was made on the
first drive day and the curve always ascends on the drive days, although
not always to a peak. It falls, however, on all but one day when a visit
was made to the roof; that it rises on the drive days is contrary to what
one would expect and is difficult of explanation, since the subject's eyes
should, if anything, have been fatigued by the increased light. If there
had been a stimulation of the central processes causing a heightened
power of discrimination, this ought to have influenced the other curves
as well.
GENERAL SUMMARY AND CONCLUSIONS.
The fact that a human being could live for a month or longer without
food had already been satisfactorily proved.1 Merlatti is reported
to have fasted for 50 days and Dr. Tanner for 40 days. The fast of
Succi2 is most similar to that of L. in that it was undergone for about
the same length of time and under similarly strict scientific control,
although never before had quite so many precautions been taken as
in the case of L. Succi fasted for 30 days, but took pepton on the
twenty-seventh day. L. continued for one day longer, absolutely
nothing but distilled water passing his lips during that time. Both
men remained in good physical condition throughout and seemed at
no time to suffer any unusual discomfort. It was with difficulty that
L. was persuaded to discontinue his fast on the thirty-first day.
Although Luciani doubted that Succi was mentally normal, general
observations and the tests pointed to a sound mind in the case of L.
Both men were, naturally, men of great determination and above all of
*E. Bardier, in his article "La Faim" (Ch. Richet's Dictionnaire de Physiologie, 1904, 6, p. 3),
remarks in regard to voluntary and involuntary fasts: "On pourra se soumettre volontairement
a un jeune prolonge, comme l'experience en a plusiers fois ete tentee, et endurer assez facilement
les souffrances de la faim. Le besoin de manger sera d'autant moins douloureux, d'autant plus
facile a supporter qu'il suffira d'un signe pour etre mis en face d'un succulent repas. Au contraire,
la faim sera beaucoup plus penible, ses manifestations beaucoup plus douloureuses, si Ton se croit —
dans un naufrage, dans une expedition — voue a une inanition complete sans espoir de salut."
On page 6, in reference to forced fasting, he further says: "La lutte que Ton est oblige de soutenir
contre les causes m6mes de cette inanition augmente la sensation de faim."
2Das Hungern, by Luigi Luciani. Translated into German by Dr. M. O. Fraenkel. 1890.
THE PSYCHO-PHYSIOLOGY OF A FAST. 213
implicit faith and confidence in their idea. L. believed fasting to be a
panacea for all ills and the very fact that he is of that type of man who
can narrow his horizon about an idea and stubbornly resist all inva-
sions gave him the best equipment for the fight against the natural
demands of the flesh. Such a type of mind can not be called abnormal,
although it is unusual. The feeling of hunger was at all times, even
during the first stages of the fast, denied by L. This statement should
not be disbelieved, even though frequently there is extreme discomfort,
which those who fast tell us only disappears after the second or third
day, as in the case of Succi. With L. and perhaps with other fasters
this feeling of hunger may have been suppressed from the beginning
by auto-suggestion. The fact of the deep-ingrained faith in the fast
makes this plausible.1
The condition of Succi's higher mental processes was only ascertained
by general observation. These observations agree with those made
upon L. There was at no time any symptom of hallucination or
lack of clearness in the thought processes. Luciani writes:
"Am 13 Hungertage wollte ich seine Ausdauer bez. geistiger Anstrengungen
auf die Probe stellen, indem ich ihm schwierige oder unlosliche metaphysische
und theosophische Fragen vorlegte und bestandig Einwtirfe gegen seine Ant-
worten erhob, in der Absicht, seinen Verstand zu ermiiden. Ich muss geste-
hen, nicht bemerkt zu haben, dass sein Geist dabei mehr ermudete als der
jedes andern Sterblichen von gleichem Bildungsgrade und gleicher Begabung,
wenn man ihn solchergestalt martert."2
L. is a man of a much higher level of intelligence and intellectual
training than Succi. At all times during the fast he was very eager to
enter into discussions upon abstract subjects such as the value of the
Esperanto language, the political conditions in Malta, the possibility
of mental telepathy, and theories of spiritism, as well as the value of
fasting. It could not be observed that there was any diminution of
his argumentative powers or lack of lucidity of expression. When
aroused to counter argumentation he showed the same energy in reply
at the end as at the beginning of the fast.
Succi's muscular strength as well as his sensory acuity was ascer-
tained in a manner somewhat similar to the method employed for L.,
and the results will be compared in the following summary and inter-
pretation of results:
(1) In the dynamometer tests made upon Succi it is impossible to
tell from the text how many trials were made daily. As the curves for
XE. Bardier, in criticizing Bernheim, writes: " Au sens oil l'entend Bernheim, les jeuneurs qui se
soumettent a l'inanition resistent facilement, tout simplement par le fait d'une auto-suggestion.
Discutant en particulier le jeune de Cetti, il admet que ce dernier—tout en n'etant pas un hysterique—
s'est suggestionne. II demeure convaincu qu'il conservait toute sa force physique, 'cela suffit pour
realiser le ph6nomene; l'idee fait l'acte; il s'exalte, il s'entratne, il se nourrit de son idee, il se
montre avec complaisance a ses visiteurs, il jouit de son triomphe; l'esprit domine le corps;
etc' . . . Le jeftneur, par sa volonte, arrive a resister a l'habitude de manger; il obdit a sa
conscience qui le soumet a l'abstinence, mais certainement sa volont6 doit 6tre incapable de provo-
quer la suppression d'une sensation." La Faim in Ch. Richet's Dictionnaire de Physiologie, 1904,
6, p. 10. See also footnote 3, p. 191 of this publication.
2 Luciani, Das Hungern. German translation by Dr. M. O. Fraenkel, 1890, pp. 68-69.
214 A STUDY OF PROLONGED FASTING.
the 10 trials and for the initial trial for L. are similar, the 10-trial curve
will be considered. It is safe to assume from lack of mention of the
fact and from the nature of the curves that Succi was right-handed.
It will therefore be necessary to compare the curve of the right hand of
Succi with that of the left hand of L.
It will be remembered that the strength of both hands was found to
increase after the drop on the second day until the right hand (VII, fig. 26)
reached its maximum on the twelfth day and the left hand (V, fig. 25) on
the sixteenth day, both curves then dropping steadily from this point,
the right, however, less than the left, for the left reached a minimum
on the thirty-first day, while the right during the fast never dropped as
low as the record of the nineteenth day. There is a very striking
similarity between these and Succi's tests.1 Both of Succi's curves
also drop after the first trials and then rise again, his left reaching a
maximum on the fourteenth, his right hand on the twentieth day, as
compared to the twelfth and sixteenth days of L. Succi's curves then
drop also, but the left drops more than the right, which is the reverse of
L.'s curves. With Succi both maximums are greater than the first
day's records, while with L. this is the case with only the left hand.
This agrees, however, with L.'s records for the initial daily trials (IX
and X, fig. 27). Further, L. was able to make a spurt at the end of the
fast with both hands, this spurt extending through several days. Succi
was only able to spurt with one hand and that on the last day, the curve
for the other hand remaining stationary.
Luciani attributed the rise of the curve alone to auto-suggestion. It
seems quite probable, inasmuch as Succi and possibly L. also believed
that their strength would be increased by the fast, that this idea strength-
ened their determination and that they bettered their results by sheer
"will power."2 There is, however, another possibility which maybe
assumed without denying the influence of auto-suggestion, namely,
that, at least in the case of L., who was unused to such tests, the coor-
dination of the muscles became gradually more perfect, and further, that
these muscles, which were being exercised daily, increased for a time in
strength as they would have done under normal conditions, but in this
case possibly to the detriment of other muscle groups. In both cases,
with both hands, fatigue gained the ascendency over practice effect
and possibly over auto-suggestion about the middle of the fast, causing
the curves to drop. In the case of L.'s unpracticed hand, however, the
effect of practice had more room to work and held the curve up longer
than in the case of the practiced hand.
(2) The tapping test (III, fig. 24) is also influenced by the condition
of the muscular tissue, but there is another factor more essential here
luciani. Das Hungern, 1890, p. 55.
2E. K. Strong, Jr., in his paper entitled " The effect of various types of suggestion upon muscular
activity" (Psych. Rev., 1910, p. 278), says: "The auto-suggestion tends most strongly of all the
types of suggestion to heighten the maxima."
THE PSYCHO-PHYSIOLOGY OF A FAST. 215
than strength, and that is the reaction time. As in the strength tests,
there is a rise at first, but here it is of much shorter duration, the maxi-
mum of 215 taps in 30 seconds being reached on the sixth day.
The following considerable drop until the fifteenth day, at a time when
the strength tests are showing more efficiency, may possibly be caused
by a lessening in the interest for this test.1 About the middle of the
series this interest and increased effort for a good record may have
returned, judging from the results, but fatigue had by that time set in
and the curve, although rising until the last day, is never quite able to
reach the maximum of the sixth day; that is, there was some falling off
in the rapidity of reaction, which, judging from the results of the strength
test, was due rather to a change in the muscle tissue than to a change in
the nervous arc.2 From what we know of the effect of practice in
such tests it is most probable that if it had not been for this increased
muscular fatigue the curve would have reached an appreciable maxi-
mum at the end of the series. From the fact of the very small differ-
ence between the average of the first 10 and last 10 seconds on the sixth
day, when the maximum was reached, as compared with the great
difference in the almost equally good result of the last day, it is evident
that on the first day the good performance of the first 10 seconds prac-
tically continues throughout (in both instances the best record was
made during the first 10 seconds), while on the last day the effect of
practice as shown in the initial performance was counterbalanced
toward the end by fatigue.3 These results seem to cast further doubt
upon Luciani's hypothesis of auto-suggestion in the strength test, for
surely auto-suggestion should play as great, if not a greater, role in
the tapping tests during those days in which according to the strength
tests it would have to be assumed at work. The results of the tapping
tests are indeed directly opposed to such a theory.
To sum up, it may then be said that though initial lack of interest and
later muscular fatigue played a role, both factors being directed toward
a decrease in the amount of work, yet central factors toward the end
brought the curve back to its initial level and almost to its maximum.
JSee pp. 194 and 196.
2As the tapping tests preceded the strength tests, the objection can not be raised that the hand
was being unusually fatigued by these latter tests.
In reference to the tapping test under normal conditions, Wells writes that "The objective
fatigue phenomena which we note in the test are in all probability fatigue phenomena in the refrac-
tory phase or a lowered efficiency of coordination, equally a product of altered synaptic conditions;
the sensations of fatigue, on the other hand, may with equal assurance be ascribed to tissue changes
within the muscles that take place as a result of their continued effort." (F. L. Wells. Normal
performance in the tapping test before and during practice, with special reference to fatigue
phenomena. Am. Journ. Psych., 1908, p. 473.) In the above tests the change in muscular tissue
is due to emaciation, a fact that does not play a role in the test to which Wells refers. At no time
did L. speak of sensations of fatigue, and judging alone from his facial and bodily expressions
there are no data from which to assume that they were greater at the end than at the beginning
of the fast. As to the synaptic conditions, there is nothing in the test to point to a change.
3 Wells writes: "The true practice gain is one mainly in the initial efficiency of performance, as
distinguished from the warming-up gain, which shows itself chiefly in continued efficiency of
performance." Am. Journ. Psych., 1908, p. 478.
216 A STUDY OF PROLONGED FASTING.
(3) The threshold for tactual-space perception (XI, fig. 28) decreased
somewhat as the fast progressed. It was on the average much better
during the last half than the first half of the series. Similar tests were
made upon Succi upon a number of different parts of the body, but only
on 3 days, before the fast, on the fifteenth day, and on the twenty-ninth
day. On some parts of the body there was an increase, on other parts
a decrease. Luciani believed the difference in the 3 days due to differ-
ences in degree of attention. On that part of the body corresponding
most closely to the spot used in these tests, *. e., the lower third of the
volar side of the forearm, there happened to be a rather large decrease
in the threshold, the three thresholds being respectively 16, 11, and
10 mm.1 Authorities differ as to whether practice lowers the threshold
in tests performed under normal conditions. Dresslar,2 for example,
found that practice had a considerable effect. Solomons3 found that
if the subject is not informed of his errors there is no effect of practice.
In the above tests the subject was never told of his mistakes and
"vexier" trials were introduced at frequent intervals and in no special
order, yet there was a lowering of the threshold. This may be and prob-
ably is due to several causes. A physiological cause would be a decrease
in the fat, thus exposing the nerve endings and making them more
sensitive. On the psychological side increased attention, which we
find indicated in other of the tests, would lower the threshold for dis-
crimination. Also, as the tests progress the image of the criterion used
becomes cleared. From what is known of the process of perception,
this is a most important factor in explaining the above effect of practice.
The physiological change is the only one which could be attributed
unequivocally to the fast. The central change occurs in series under
normal conditions.
If, as has often been assumed, the tactual space threshold test is a
measure of mental fatigue, then it must be concluded that there is no
indication of such fatigue during the fast.
(4) The visual acuity (XII, fig. 28) showed an astonishing better-
ment. From 17 feet as the distance of clear vision for the particular
test card employed, the curve ascended rapidly to 37 feet on the four-
teenth day and, although there is a falling off, 36 feet is the record for
the last day of the fast.
If it were not for the maximum of 37 feet midway in the series,
the improvement would be comparatively a steady one. One explana-
tion that suggests itself is that the possible change in intra-ocular
tension caused the eye-ball to change its shape. Unless his glasses
were not the proper ones for him, however, a change in the eye should
cause more rather than less difficulty as long as he wore his glasses.
Further, the suddenness of the rise seems to vitiate such a theory.
'Luciani, Das Hungern, 1890, p. 64.
2F. B. Dresslar. Studies in the psychology of touch. Am. Journ. Psych., pp. 313-368. 1894.
3L. M. Solomons. Discrimination in cutaneous sensations. Psych. Rev., pp. 246-250. 1897.
THE PSYCHO-PHYSIOLOGY OF A FAST. 217
A satisfactory explanation seems difficult to find. It might be said
that the 37-foot record was made by chance. This also seems pre-
cluded by the fact of the number of previous steps in which 10 correct
answers were given and from the evidence of confidence displayed
by the subject.1
Succi's eyes were examined with the ophthalmoscope and his acuity
measured before the fast and on the fifteenth and twenty-eighth days of
the fast, but no change was detected.2 If L. had happened to be
measured on the third, sixteenth, and one of the days toward the end of
the series only, the change would have been thought as negligible as in
the case of Succi. In all such tests where the daily fluctuation is consid-
erable three tests in a month are not sufficient upon which to base a judg-
ment as to the change in sensory acuity or higher mental processes.
(5) The rote memory for digits (II, fig. 23) showed very little change.
There is a slight suggestion of improvement during the first half of
the series. Judging from the curve which indicates the point at which
the first mistake was made (II b), one can say that there was a gradual
improvement in this respect, especially in the first half of the series,
which is probably in part due to a betterment in the perception of the
spoken word, but especially to an increase in attention, it becoming
more sustained as the fast progressed. The rote memory for sense
words (I a) showed a greater improvement than did that for digits.
Here probably the practice effect consisted in the forming of associa-
tions between the words. The most marked improvement of all is
in the retention after a longer period of time, i. e., after 55 minutes (I b).
This is probably also due, in part at least, to the more frequent forming
of associations. Besides, the repetition of the same task through so
many days undoubtedly strengthened the determining tendency, i. e.,
the determination taken at the time of memorizing for the words to ap-
pear in consciousness again, it remaining either in consciousness or
subconsciousness during the interval. According to L.'s statement,
his mind did not revert to the task within the hour. Indeed, the other
tests followed each other so rapidly that this would have been a diffi-
cult thing to do.
Experiments upon memory under normal conditions also show the
effect of practice, as evidenced by an appreciable increase in the mem-
ory span which may continue for a period of 2 months.3
xThe subject did not know whether he was right or wrong or how many correct answers consti-
tuted a threshold, so that the results could not have been prearranged by him; and if they could
have been he would not have allowed such a good record on the fourteenth day. The high thresh-
old on the last day is obviously due to his unusually poor physical condition (when, if at any time,
one might be justified in speaking of a lack of effort).
2Luciani, Das Hungern, 1890, pp. 66-67.
3T. L. Bolton. The growth of memory in school children. Am. Journ. Psych., 1892, pp. 362-380.
G. Milller and F. Schumann. Experimentelle Beitrage zur Untersuchung des Gedachtniss,
Zeitschr. f. Psych., 6, 1894, pp. 81-190, 257-339.
W. H. Winch. The transfer of improvement in memory in school-children. British Journ.
Psych., 1908, pp. 284-293.
218 A STUDY OF PROLONGED FASTING.
(6) The cancellation test (XXI), which employs to a greater degree
the higher functions of perception and attention shows the greatest
improvement of any of the tests used. This improvement continues
from the sixth to the last day of the fast. The accuracy is so high
throughout the series that the slight improvement in the latter part
of the tests is of no significance. Experiments have shown that fatigue
affects the accuracy, so that again we have evidence against an increase
in mental fatigue.1
Besides an improvement in the above-named functions, the increase
in visual acuity may have been a factor in the results. On the other
hand, from the results of the tapping test and strength tests one must
conclude that the betterment is in no degree due either to a betterment
in reaction time or motor ability.
(7) The free association time (XIII, fig. 29) is on the whole shorter
during the latter part of the series. If it were not for a rapid drop in the
middle of the curve after a rise similar to that in the tapping test the
improvement would be comparatively steady. The minimum of 1.3
seconds is reached on the day before the last day of the fast and should
be compared rather with the 1.9 seconds of the third day than with the
2.5 seconds of the first day, when L. was unaccustomed to the manner of
reaction. Even when this comparison is made it is seen that the
improvement is considerable. A separation of the curve into four
curves corresponding to the four categories used made a more minute
analysis possible. The curves XIV, XV, XVI, and XVII, figs. 30 to 32,
show fewer high averages in the second half of the series, but it is only
in the abstract curve and in less degree in the adjective series that there
are more low averages in the second half of the curve. In fact, in neither
of the other two curves is the lowest average of the first half of the series
again equaled. This seems to indicate that the betterment in the
general average of the 20 words is principally due to a betterment in
the reaction to abstract words. It is to be expected that the most
difficult associations would show the greatest practice effect. In the
noun and verb curve there is an almost steady rise in the middle of the
curve corresponding to the rise in the middle of the main curve. I
seems plausible to suppose that there is here, as in the tapping test, a
falling off of interest, and that this would manifest itself more readily in
the easier tasks, in which the reaction is likely to become more nearly
mechanical.
The general improvement is also seen in the decrease in the variations
of the reaction times. In all four curves the daily variation is much
less in the second half of the series. Parallel with this is the decrease in
XB. Bourdon. Observations comparatives sur la reconnaissance, la discrimination et l'associa-
tion. Rev. Phil., 1895, pp. 153-185. A. Binet. Attention et adaptation, Annee Psych., 1900,
6, pp. 248-404. C. Bitter. Ermudungsmessungen, Zeitschr. f. Psych., 1900, pp. 401-444.
THE PSYCHO-PHYSIOLOGY OF A FAST. 219
the variations within each day, as is shown by the decided drop in
the m. v. curve (XVIII, fig. 33) -1
Although the improvement in the reproduction time is not so great
as in the association time, yet it is noticeable, the average of the second
half being lower than that of the first, although the very low time of
0.8 second was made on the second day as well as during the second
half of the series.
The quality of the associations was good throughout (II, table 21)
and showed no striking change.2 The reproductions were so nearly
perfect from the first that nothing can be said in regard to them to
support the results of the memory tests. One might add, however,
that neither do they contradict those results.
The controlled reaction noun-verb (XX, fig. 34) shows an increasing
lengthening of the time until almost the end of the series. It is quite
probable that this was caused by an increasing difficulty in the stimulus
words selected, a factor which could not well be avoided. No other
reason suggests itself why these reactions should have taken a different
course from that of the free association tests.
The present methods of testing mental capacity unfortunately do
not permit one to make dogmatic statements as to the results of any
such tests. In each one a number of functions are involved, any one
of which may have produced the variations which occur. For example,
the cancellation test involves, among other things, attention and inter-
est, apperception and discrimination, nervous impulse and motor dis-
charge. But when, as here, a set of tests are employed in which the
same functions are more or less active and they all show a similar trend,
then a conjecture along general lines seems legitimate. Further, when
there is a very decided difference and it is known that a certain function
is of prime importance, then one is undoubtedly justified in ascribing
the outcome of this test to changes in this function. It is desired to
make it plain that no exact measurement is claimed, but merely that it
has been possible, by means of a number of selected tests, to sketch an
outline picture of the condition of L.'s psycho-physiological organism.
1 Wells conducted long series of association reactions with normal subjects and for all of them
found an improvement in the reaction time. (See Practice effect in free association, Am. Journ.
Psych., 1911, 22, pp. 1-13.)
2 W. Weygandt's results are hardly comparable to those obtained in these tests (Ueber die
Beinfiussung geistiger Leistungen durch Hunger, Psych. Arbeiten, 4, pp. 45-173). His subjects
fasted for periods of only 24 and 48 hours at a time. This intermittent fasting may possibly
cause a much more pronounced disturbance to the organism than a prolonged fast. That there
was greater exhaustion seems to be indicated by the fact that there was an increase in associations
by sound. He also finds that there was an increase in the outer as compared with the inner asso-
ciations. (It is now admitted that such a classification of reaction words can not be made without
introspection.) Weygandt also found memory to be affected. The association time was not
altered. Aschaffenburg studied the effect on association reactions of the exhaustion produced by
a night's work without food or sleep. (Studien ueber Associationen, ii Teil. Die Associationen
in der Erschopfung. Psych. Arbeiten, 2, pp. 1-83) . He too found a similar decrease in the quality
of the reaction words. "Mit der Zunahme der Erschopfung wirkt die zugerufene Vorstellung
immer weniger durch ihre Inhalt; an dessen Stelle bestimmen der Klang und die Tonfarbe die
Reaction."
220 A STUDY OF PROLONGED FASTING.
It will be remembered that the tests range from those involving
principally the muscle groups to those depending in a higher degree
upon central factors. The test depending most on the muscular
reactions, t. e., the strength test, showed a falling off. The tapping
test, which also involved the muscles but in which the rapidity of reac-
tion was a more important factor, showed no improvement. As soon
as one turns, however, to the sensory discriminations one notices an
increased efficiency, which is probably due either to a change in the
peripheral organs or central processes, or both. Finally all the tests
involving the higher processes of attention, perception, and association
show improvement. In a word, there was a loss in muscular strength
due probably to loss of tissue, a possible gain in sensory acuity and a
decided increase in the efficiency of all the central processes. It would
be premature to say that the improvement is the direct result of the
prolonged abstinence from food, as similar improvement has been
observed in such tests under normal conditions, due entirely to the effect
of practice. It can be stated, however, with some degree of certainty,
that the complete abstinence from food for 31 days had little effect
upon the higher mental functions, which were able to develop through
practice very much as they would have done under normal conditions.
This agrees with the observations upon the physiological conditions.
It has been found that during a fast the muscle tissues are the first to
suffer and the nervous tissues the last. From these results it seems
that up to the thirty-first day the nervous tissues have not suffered.
These results also confirm in part the general observations made by
those fasting. It is frequently stated by them that they can do better
mental work. The results show that at least they can do approxi-
mately as well, and it is not at all unlikely that some can do better, for it
must be remembered that there is none of that sluggishness of the
mental processes directly after eating, when the digestive processes are
at their height, and there is also absence of indigestion and the after-
effects of alcohol and tobacco. That, on the other hand, as has been
often claimed, they are able to do more muscular work and that their
power of endurance is greater is in this case at least not true. Probably
the contrast of their actual results compared with what they expected
would happen to a man without food makes the result seem greater
than it is. The claim that the senses are more acute has been verified
as to the visual acuity. It is hardly likely that the slight difference in
the tactual-space threshold would have been noticed by the faster.1
The question remains as to whether prolonged fasting is beneficial
or dangerous to the organism. This can only be satisfactorily answered
*L. stated that the heightened sensitivity for odors made walking on the streets of Malta during
his first fast positively unpleasant. The other senses were examined in the case of Succi and no
appreciable change discovered. Luciani, Das Hungern, 1890.
Whipple (Manual of Mental and Physical Tests, Baltimore, 1910, p. 215), in speaking of the
-effect of practice in the sesthesiometer test, remarks that Dresslar states: "This practice effect ia
. . . rapidly lost, being reduced very definitely within 8 days and completely lost within a month."
THE PSYCHO-PHYSIOLOGY OF A FAST. 221
after an exhaustive physiological examination extending over a long
period of time subsequent to the fast. The tests made after the lapse
of a year permit, however, of some conjecture in this regard concerning
those functions at least which have been discussed in this paper.
The strength test shows a great improvement over the former record.
L. exerted a pressure considerably greater than at any time during
the long series. The record for the tapping test is also above the maxi-
mum of the previous record. The association test shows a marked
improvement and the reproduction is also better, especially in that it
varies less, and the retention of sense words has perhaps also slightly
improved. The tactual-space threshold and the rote memory for sense
words are about the same as at the end of the fast. Only in the case of
the memory for digits and in the cancellation test has the previous
maximum not been reached, but both of these results show consistently
good results. It may be stated, in short, that after an entire year's
intermission the curves continued practically from the point they had
previously reached, if not considerably above that point, without show-
ing that loss of practice which might well have been expected. These
improved conditions are, however, not necessarily traceable directly to
the beneficial effects of the fast. In regard to the association tests L.
has undoubtedly become still better acquainted with the English lan-
guage, and in respect to the strength tests it must be noted that L. has
exercised his muscles daily, according to his report. In general he has
led a careful life, paying especial attention to his diet. The possible
effect of climate and his new surroundings is also to be considered.
Finally, and most important, is the possibility that there was actually
a greater effect of practice in the first series than appeared in the
records, but that it was concealed by certain opposing effects of the
fast, so that the results of the later tests may not be quite what might
be supposed from a comparison of the records.
It remains, however, an indisputable fact that, according to the tests
made, there was no lasting evil effect of the fast, either upon muscular
strength or mental activity.
222
A STUDY OF PROLONGED FASTING.
APPENDIX I.-DREAMS.
As has been already stated, L. was asked to recount the dreams he had had
during the previous night. From these records those dreams are here given
which pertain to food. It will be seen that at one time he ate, at another
refused food, but in neither case was there evi dence of anything but a normal
emotional reaction. According to the Freudian theory this absence of an
intense emotional state (there were no nightmares nor anything else in the
records indicative either of mental or bodily distress) means that the will
("wish") to fast was too strong to allow of any serious conflict of ideas. A
great part of the dreams are of a sexual nature and are not here given.
April 13. I saw a basket covered with a white piece of cloth, which I
imagined full of food. When I tried to uncover it several black rats jumped
out of it and frightened me.
I dreamed I was passing down one of our streets in Malta with a paper bag
under my arm containing cheese-cakes for my daughter. I found myself in
a state of mental excitement and after going a certain distance I found that the
lower end of the bag was opened and the cheese cakes were gone. In their
stead was a white hand.
April 19. I dreamed I was in a shop and on the counter there was a very
big ham, about 10 feet in diameter. The proprietor was riding on the top of it
with a knife in one hand. "It is a very good one," he said. I answered, "I
do not like it. Do you not know I am fasting?" Then a friar came in and said,
"I will take it in his stead, because I like it." He took it and swallowed it.
April 21. I dreamed I had been for a walk in the country. I went to a
country tavern and asked for something to eat. The proprietor gave me a
beefsteak and some fried red fish. I ate them with relish and asked what I
had to pay. He told me $1.50 and asked if that was too much. I said I did
not think so. In coming out of the tavern I saw a river full of these red fish
and people were trying to catch them. I said, " You are fishing out all the
fish and if you continue you will not have any more to eat."
APPENDIX II-COMPLETE SERIES OF ASSOCIATION TESTS.
Average
2.3
April 11, 1912:
April 12:
Stimulus
Reaction
Reaction
Stimulus
Reaction
Reaction
word.
word.
time.
word.
word.
time.
Paper
ink
2.2"
Round
table
2.0"
Bright
light
2.0
Country
green
1.8
Yellow
lemon
1.8
Silver
spoon
2.3
Table
knife
1.2
Rabbit
white
2.0
Spoon
broth
2.8
Chair
cushion
3.0
Apple
stem
2.4
Glass
window
2.0
Sleep
bed
1.6
Flower
odor
2.3
Room
door
1.3
Sun
brightness
3.2
Face
eye
2.0
Bread
white
2.3
Carpet
red
1.8
Wood
hard
3.0
Animal
white
2.6
Well
water
2.4
Rain
noise
5.0
Danger
sea
2.0
Teach
bench
2.0
Tired
bed
2.0
Doctor
knife
4.0
Watch
gold
2.4
Book
no. of pages
3.8
Marble
table
1.6
Store
glass window
1.6
Iron
bar
3.8
Horse
tail
2.2
Bridge
iron
2.8
Island
trees
2.2
Blind
dark
2.4
Journey
ship
2.2
Pencil
wood
3.0
Freedom
banner
2.0
Candy
sweet
3.4
Sweet
sugar
1.2
Average
2.5
THE PSYCHO-PHYSIOLOGY OF A FAST.
223
April IS:
April 16:
Stimulus
Reaction
Reaction
Stimulus
Reaction
Reaction
word.
word.
time.
word.
word.
time.
Timid
rabbit
3.0"
Defend
country
1.8"
Pulse
hand
9.0
Deck
ship
1.2
Mystery
religion
5.2
Fresh
air
0.8
Savage
wolf
2.4
Faculty
arts
1.0
Spirit
angel
2.4
Deduct
sum
1.4
Teeth
to eat
2.6
Dinner
good
1.4
Bargain
profit
5.0
Flavor
odor
2.2
Blunder
mistake
3.0
Displease
anyone
3.2
Temper
nervous
2.2
Dog
large
2.0
Abrupt
cascade
2.0
Good
man
0.6
Harp
sing
2.0
Fault
his fault
3.0
Switch
machine
2.4
Egg
white
1.4
Wide
sea
2.2
Green
tree
1.8
Tailor
stuff
3.0
Fright
dog
2.2
Income
money
1.5
Drive
horse
1.2
Splendor
sun
1.8
Fairy
tale
1.4
(Salve) solve
lip
2.6
Hard
stone
0.8
Moon
silver
2.2
Function
ceremony
3.2
Frost
white
1.8
Profess
religion
1.4
License
wine
1.8
Salt
sea
1.4
Average
2.9
Average
1.7
April 14-'
=
April 17:
=
Accept
a reward
2.8"
Crawl
serpent
2.0"
Air
blue
2.6
Clown
buffoon
2.4
Able
sailor
2.0
Dizzy
headache
1.6
Abuse
drink
2.6
Distance
my country
2.0
Address
letter
1.8
Cure
physic
2.6
Blood
red
1.1
Corn
grass
1.8
Bad
man --—
1.4
Easy
chair
1.8
Age
90
1.2
Distress
sorrow
2.0
Agree
wife
1.0
Decorate
church
1.6
Boot
black
1.8
Copper
metal
1.4
(Tall) baU
tree
1.6
Even
ground
2.4
Balance
weight
1.6
Endurance
fasting
1.4
Amuse
theater
1.4
Decline
age
1.0
Bottle
ink
1.4
Cream
sweet
2.0
Band
brass
1.4
Firm
strong
3.4
Climate
mild
0.8
East
west
1.0
Bite
dog
1.5
Degrade
man
1.8
Box
wooden
1.6
Corset
woman
1.0
Contents
book
6.4
Flat
floor
1.8
Boy
small
1.8
End
book
3.0
Average
1.9
Average
1.9
April 15:
=
April 18:
=
Catch
bird
1.6"
Hit
hammer
3.0"
Brain
human
2.6
Swallow
food
1.2
Broad
street
1.4
Suffer
pain
1.2
Courage
man
2.2
Build
house
1.3
Cease
speak
2.2
Rubber
teeth
1.4
Brick
red
1.6
Food
good
1.0
Broken
glass
1.0
Park
large
1.1
Culture
physical culture
1.6
Boat
swim
1.8
Compel
servant
3.4
Smooth
floor
1.1
Cable
iron wire
1.4
Straight
way
1.8
Central
station
1.6
Ugly
man
1.8
Crowd
people
1.2
Gentle
woman
1.4
Confess
priest
1.0
Naughty
man
2.0
Carbon
carbon dioxide
1.8
Power
England
1.6
Common
sense
1.0
Strength
athlete
1.9
Day
night
2.0
Charm
woman
' 3.0
Control
engine
1.0
Cost
money
1.0
Chain
iron
1.0
Kindness
woman
2.2
Course
study
2.2
Break
glass
1.2
Delegate
apostolic
2.0
Jaw
mouth
1.8
Average
1.7
Average
*.6
224
A STUDY OF PROLONGED FASTING.
April 19:
April 22:
Stimulus
Reaction
Reaction
Stimulus
Reaction
Reaction
word.
word.
time.
word.
word.
time.
Produce
field
1.4'
Eat
bread
2.0'
Cry
baby
1.0
Open
door
1.0
Freeze
cold
1.6
Divide
reign
1.8
Follow
soldier
5.8
Fade
flower
1.6
Smoke
pipe
0.8
Travel
ship
2.0
Rope
long
2.0
Umbrella
rain
0.8
Omelet
eat
1.4
Gift
gold
3.0
Cap
head
1.0
Man
long
0.8
Burglar
thief
1.6
Sailor
ship
1.2
Delicate
woman
0.8
School
teacher
1.2
Thick
paper
2.8
Dense
air
2.0
Expensive
money
1.0
Short
man
1.4
Dark
night
1.0
Weary
travel
1.6
Unfair
unjust
2.0
Best
book
5.4
Purpose
Bcope
1.0
Excuse
pardon
1.6
Glory
eternal
1.2
Insult
bad
3.8
Mischief
bad
2.0
Prudence
woman
1.6
Occasion
accident
1.0
Caution
wise man
2.2
Nuisance
wrong
1.6
Conceit
ambition
2.2
Overcoat
dress
1.0
Captain
ship
1.4
Average
1.6
Average
1.9
April 20:
=
April 23:
=
Prefer
office
2.4'
Collapse
sick
2.4'
Crush
crowd
2.0
Excite
nervous
1.6
Allow
pension
1.6
Begin
book
1.8
Drink
water
1.2
Prosper
progress
2.4
(Solution)
salt
2.2
Hat
head
1.2
salute
Sister
brother
1.0
Hip
thigh
1.2
Ham
meat
2.0
Lightning
thunder
2.0
Crime
justice
2.8
Parlor
bedroom
2.4
Tight
shoe
2.0
Snake
serpent
1.0
Solid
stone
1.8
Wicked
man
1.2
Cold
winter
1.6
Rich
millionaire
1.8
Clear
sky
1.4
Clean
body
1.2
Hope
fortune
3.6
Bashful
woman
1.0
Dismay
fear
1.6
True
religion
5.2
Offense
insult
1.4
Exchange
money
1.0
Blunder
mistake
1.0
Style
literature
1.0
Future
time
4.0
Power
gun
1.0
Insist
persist
2.4
Result
good
1.4
Trap
wolf
2.0
Nonsense
foolish
1.6
Oblong
square
1.4
Seed
plant
1.0
Average
2.0
Average
1.7
April 24:
=
April 21:
=
Restore
furniture
1.4'
Pinch
pin
1.4'
Impress
printing
1.8
Satisfy
appetite
0.8
Flirt
woman
1.0
Nourish
food
1.2
Ask
question
1.2
Drift
wind
0.8
Receive
letter
0.8
Abuse
drink
1.2
Baker
bread
1.0
Ditch
deep
1.2
Athlete
strength
1.0
Tiger
fierce
1.0
Cradle
baby
1.0
Music
sweet
1.0
Bundle
hay
1.0
Fish
sea
1.4
Elephant
trunk
1.0
Death
eternal
(22.4)
Cheap
money
3.0
Soft
paste
2.4
Black
dog
0.8
Ugly
man
1.2
Tender
meat
1.4
Watchful
policeman
2.6
Prompt
answer
1.4
Indecent
conduct
3.0
Ignorant
man
1.0
Haste
hurry
1.0
Confidence
familiarity
2.0
Comfort
good
2.0
Jealousy
woman
0.8
Adventure
strange
1.2
Honesty
good
4.2
Practice
long
1.8
Unbelief
atheist
2.4
Untrue
falsehood
1.6
Heroism
warrior
2.0
Merit
high
2.8
Average
1.6
Average
1.5
THE PSYCHO-PHYSIOLOGY OF A FAST.
225
April 25:
April 28:
Stimulus
Reaction
Reaction
Stimulus
Reaction
Reaction
word.
word.
time.
word.
word.
time.
Join
chain
1.8"
Persuade
argument
2.4"
Clasp
hand
1.0
Dig
ditch
1.0
Advance
pretension
2.0
Get
money
1.0
Argue
discussion
2.0
Sting
bee
2.2
Mountain
large
1.0
Preach
priest
1.0
House
beautiful
1.4
Spice
pepper
0.8
Neck
strong
1.0
Star
Venus
1.4
Lamb
quiet
1.2
Ice
cold
1.0
Hero
brave
1.2
Picture
beautiful
1.8
Jealous
woman
1.4
Lip
red
1.4
White
snow
2.0
Easy
chair
1.0
Serious
man
1.0
Unclean
dirty
1.4
Vacant
space
1.0
Red
rose
1.0
Fertile
land
1.0
Rotten
mud
2.0
Reason
mind
1.6
Hard
flint
1.0
Protection
government
1.8
Proposition
geometry
1.6
Solemnity
festivity
1.0
Improvement
progress
1.0
Impudence
woman
3.8
Infamy
calumny
2.2
Convenience
etiquette
3.0
(Competition)
commerce
2.4
Scratch
nail
1.6
competence
Attraction
actress
2.0
Average
1 6
April 26:
=
Average
1.6
Forget
memory
1.2*
April 29:
==
Dislike
people
1.0
Announce
news
1.2"
Prepare
lesson
1.0
Stain
ink
1.0
Admire
virtue
1.8
Finish
lesson
1.4
Protect
children
1.6
Drag
horse
2.0
Starch
white
1.2
Plead
case
2.0
Mutton
meat
14
Cork
bottle
2.0
Ostrich
feather
1.0
Toy
child
1.2
Roof
house
2.0
Key
door
1.2
Little
boy
1.0
Ox
horns
2.2
Funny
buffoon
2.2
River
water
1.6
Gay
sun
1.2
Rusty
iron
1.6
Dead
black
1.2
Ungracious
bear
2.0
Slow
worm
1.6
Irksome
science
2.4
Solemnity
feast
1.6
Equal
balance
4.0
Annoyance
fly
1.0
Late
hour
1.2
Constancy
virtue
3.2
Accusation
importation
2.0
Attention
mind
1.4
Corruption
money
2.0
Uncertainty
pendulum
(12.6)
Poverty
distress
3.2
Imposition
tax
1.0
Average
1.4
Adoration
saint
1.4
April 27:
=====
Accuse
judge
1.8"
Average
1.8
Appear
star
2.0
April 80:
==
Polish
wood
1.2
Adore
saint
2.2"
Repeat
lesson
1.0
Perish
ship
2.2
Condemn
delinquent
2.4
Propose
marriage
1.4
Car
motor
1.8
Uphold
politics
2.8
Knee
leg
1.8
Descend
stairs
1.2
Cloud
white
1.2
Slave
misery
2.8
Fun
joy
1.8
Violin
music
2.0
Violent
wind
1.2
(Path) pot
country
2.4
Sour
acid
1.0
Chapel
church
1.4
Dim
sound
1.0
Trumpet
sound
1.2
Condition
good
1.0
Supreme
being
1.2
Deceit
deceive
3.0
Elegant
woman
1.6
Fraud
wrong
3.0
Impudent
woman
2.0
Brutality
animal
2.0
Blame
offense
2.4
Cup
wine
1.2
Gain
money
1.0
Equality
fraternity
3.0
Idea
noble
1.0
Greasy
pole
1.2
Worship
God
1.0
Violet
odor
1.0
Elevation
spirit
1.4
Noisy
metronome
2.0
Average
1.7
Level
ground
1.0
Average
1.7
226
A STUDY OF PROLONGED FASTING.
May 1:
May 4-
Stimulus
Reaction
Reaction
Stimulus
Reaction
Reaction
word.
word.
time.
word.
word.
time.
Escape
prison
2.0*
Fast
long
1.4*
Admit
argument
2.0
Dream
sleep
2.6
Joke
play
3.0
Taste
food
1.2
Improve
mind
1.6
Cook
food
1.4
Defy
enemy
1.2
Mark
ink
1.0
Lamp
fire
2.0
Sparrow
bird
1.0
Cabbage
green
1.0
Foot
large
1.6
Paste
soft
1.2
Spider
insect
3.2
Poem
beautiful
1.0
Forest
trees
1.0
Spear
piercing
2.6
Stone
heavy
1.0
Harsh
sound
1.2
Purple
color
1.0
Unripe
fruit
1.0
Infamous
calumny
1.2
Unwell
sick
1.0
Refined
art
1.2
Vile
fellow
1.0
Ungracious
bear
1.8
Admission
employment
3.0
Center
circle
1.6
Thankfulness
gratitude
2.0
Awkward
gait
1.8
Dishonor
bad
3.6
Supremacy
authority
2.0
Intimacy
friendship
1.0
Constancy
perseverance
1.6
Revenge
fault
3.4
Time
quick
1.2
Least
thing
2.6
Gin
bad
1.0
Average
1.9
Average
1.5
May 2:
=====
May 6:
==
Deny
favor
2.0'
Invite
guest
1.6*
Burn
fire
1.6
Pin
clothes
1.4
Paint
wall
1.8
(Crumble)
bread
1.4
Betray
faith
1.2
tremble
Dress
clothes
1.4
Attack
enemy
1.2
Mouse
black
2.0
Wood
hard
2.0
Barn
corn
3.0
Dirt
nasty
3.2
Song
beautiful
1.4
Shoe
tight
1.2
Spider
feet
2.6
Camp
large
1.8
Scarlet
fever
1.6
Cannon
big
2.6
Beautiful
woman
1.4
Ashamed
fault
1.4
Yellow
fever
1.8
Unsafe
war
1.6
Modest
girl
2.0
Raw
fruit
2.0
Wealthy
man
2.0
Smooth
ground
1.2
Justice
right
1.4
Fortune
money
1.4
Trouble
bad
2.0
Disdain
angry
2.0
Quantity
large
1.6
Refinement
art
1.8
Reproach
fault
1.2
Activity
work
1.2
Energy
force
2.0
Accident
misfortune
1.6
Crack
nuts
1.0
Scoff
offender
2.4
Noisy
clock
2.0
Average
1.8
May 8:
=====
Average
1.8
Guide
a traveler
6.4"
May 6:
==
Care
a boy
2.3
Dishonor
sin
2.4*
Denounce
principles
3.8
Remove
furniture
1.6
Drop
stone
1.4
Injure
sword
2.4
Suspect
fault
2.2
Plunge
water
1.0
Saddle
horse
1.6
Murder
thief
1.4
Sleep
bed
2.2
Garden
flower
1.0
Fog
fruit
1.0
Nut
crack
2.0
Skin
animal
1.4
Stem
heraldry
2.0
Earth
ground
3.2
Crab
animal
2.0
Rough
weather
1.2
Pickle
burning
2.0
High
mountain
1.2
Noble
gentleman
1.6
Idle
servant
1.4
Nice
fellow
1.2
Humble
man
2.0
Secure
keys
1.0
Active
boy
2.4
Blue
sky
2.0
Health
good
1.4
Swift
sparrow
1.4
Aim
noble
1.8
Disgrace
fault
2.0
Fame
vain
2.8
Security
policeman
2.2
Shame
wrong
2.0
Unhappiness
marriage
2.8
Ability
great
1.2
Rhyme
poetry
1.0
Disaster
Titanic
1.2
Average
2.1
Average
1.7
THE PSYCHO-PHYSIOLOGY OF A PAST.
227
May 7:
May 10:
Stimulus
Reaction
Reaction
Stimulus
Reaction
Reaction
word.
word.
time.
word.
word.
time.
Wash
clothes
1.0"
Roast
meat
1.0"
Elevate
thought
1.4
View
panorama
1.8
Deceive
wrong
2.6
Whistle
a whistle
1.4
Ramble
about
1.6
Alarm
people
2.6
Decay
reign
1.8
Indulge
drinker
1.4
Bible
holy
1.4
Frost
white
1.4
Pencil
lead
1.0
Cask
wine
1.0
Crown
king
1.0
Curtain
silk
1.4
Goat
milk
1.2
Nurse
baby
1.2
Candy
sweet
0.8
Ivy
wall
1.4
Restless
not quiet
2.0
Thankful
grateful
1.0
Simple
countryman
1.6
Steep
stairs
1.2
Reckless
man
1.2
Unwholesome
air
1.0
Eternal
life
1.2
Gentle
woman
1.4
Prosperity
fortune
1.0
Faithful
servant
1.0
Jealousy
woman
1.2
Conflict
nations
1.2
Concealment
to hide
2.4
Anger
bad
2.2
Advancement
progress
0.8
Idleness
vice
2.4
Rancid
butter
1.4
Betrayal
traitor
1.8
Honesty
good
1.0
Denouncement
fault
2.0
Average
1.4
Average
1.5
May 8:
=
May 11:
=
Deserve
merit
1.2"
Plunge
water
1.0"
Wish
fortune
2.4
Guess
enigma
2.4
Boast
glory
3.2
Rescue
wrecked
1.8
Establish
manufactory
1.1
Believe
God
1.4
Barber
razor
1.6
Carve
wood
1.0
Pebble
stone
1.4
Door
house
1.8
Heart
beat
1.2
Barley
corn
1.0
Machine
work
1.4
Eagle
bird
1.0
Statue
marble
1.2
Chin
face
1.6
Certain
thing
2.0
Pulse
beating
1.0
Natural
regime
1.8
Alive
man
1.2
Correct
grammar
2.0
Exquisite
sweet
1.6
Dusty
street
0.8
Empty
barrel
1.2
Enormous
building
1.6
Bitter
quassia
1.8
Commandment
God
1.0
Lazy
fellow
0.8
Excitement
nervous
0.8
Modesty
virtue
1.0
Restoration
food
1.6
Immensity
God
1.6
Density
mercury
1.8
Preservation
alcohol
1.8
Infirmity
sickness
1.8
Prudence
virtue
1.2
Return
voyage
1.6
Indiscretion
vice
1.2
Average
1.6
Average
1.4
May 9:
=
May 12:
=
Paper
write
1.2"
Find
treasure
2.0"
Bright
sun
0.6
Praise
merit
2.0
Yellow
fever
1.4
Pump
water
1.0
Table
mahogany
3.2
Try
lesson
1.8
Spoon
food
1.4
Guard
tower
1.8
Apple
fruit
0.8
Iron
metal
1.8
Sleep
night
2.4
Stomach
empty
1.8
Cut
animal
1.8
Salmon
fish
1.0
Face
beautiful
1.2
Bath
water
1.2
Carpet
ground
1.4
Splinter
wood
1.2
Animal
fierce
1.6
Unfit
unable
2.0
Rain
weather
1.8
Ardent
fire
1.2
Teach
lesson
1.8
North
south
2.2
Doctor
medicine
1.0
Handsome
lady
1.2
Book
interesting
1.4
Price
high
2.2
Store
goods
3.0
Appetite
good
1.2
Horse
animal
1.6
Fable
^Esop
2.0
Island
Malta
1.2
Definition
grammar
1.8
Journey
long
1.0
Queer
sound
2.2
Freedom
liberty
0.8
Ingenuity
simplicity
1.4
Average
1.5
Average
1.7
228
A STUDY OF PROLONGED FASTING.
May 18:
June 2, 1918:
Stimulus
Reaction
Reaction
Stimulus
Reaction
Reaction
word.
word.
time.
word.
word.
time.
Distrust
enemy
1.8"
Adore
God
1.0"
Run
a long way
2.0
Perish
ship
1.0
Agree
friend
1.2
Propose
marriage
1.1
Needle
thread
1.2
Uphold
opinion
1.8
Chocolate
sweet
1.0
Descend
mountain
1.2
Twig
tree
1.2
Slave
poor
2.2
Napkin
white
1.2
Violin
song
1.6
Hill
steep
1.4
Brook
river
1.0
Finger
hand
1.0
Chapel
church
1.0
Pretty
girl
1.2
Trumpet
sound
1.0
Contented
happy
1.0
Supreme
Being
1.2
Absent
minded
1.8
Elegant
lady
1.2
Magical
lantern
1.4
Impudent
boy
2.8
Profane
words
1.2
Blame
fault
2.1
Introduction
to a friend
1.4
Gain
money
0.8
Amusement
theater
1.2
Idea
beautiful
1.4
Remorse
sin
0.8
Worship
God
1.6
Calmness
quietness
1.2
Comfort
pleasure
3.0
Nod
head
1.0
Noisy
room
1.0
Calculate
numbers
1.0
Level
ground
1.0
Average
1.3
Average
1.4
May 14:
=
June 8:
=
Shock
electricity
1.4"
Cover
hat
1.3"
Sweat
heat
1.8
Hasten
pace
1.0
Melt
snow
1.4
Curse
son
3.6
Stun
hit
1.4
Hurt
wound
1.4
Hunt
deer
2.0
Blush
young lady
2.2
Maiden
woman
1.8
Island
Malta
0.8
Bag
sand
2.0
Copper
metal
1.0
Belt
leather
1.2
Water
flowing
1.0
Cake
sweet
1.2
Lettuce
vegetable
1.4
Unhappy
miserable
1.6
Brandy
alcohol
1.0
Pure
blood
1.8
Unseen
God
1.0
Disorderly
irregularity
1.6
Merry
happy
1.6
Unemployed
poor
2.0
Sacred
church
1.4
Wretched
miserable
2.0
Excellent
exam
1.6
Indulgence
vice
1.6
Adorable
saint
1.4
Agreement
friendship
1.2
Life
Eternal
1.2
Advantage
benefit
1.2
Opposition
enemy
1.2
Injury
blow
1.2
Intellect
mind
1.2
Outrage
war
1.6
Sorrow
grief
1.4
Rubber
teeth
1.6
Education
school
1.2
Average
1.6
Average
1.4
May 15:
=
June 4-'
=
Sin
bad
1.4"
Caress
baby
1.4"
Applaud
merit
0.8
Reduce
salary
1.0
Astonish
marvel
1.6
Reward
behavior
1.8
Rejoice
good news
2.0
Talk
English
1.0
Use
tools
1.2
Touch
table
1.0
Spool
loom
1.4
Street
long
1.0
Sheep
fur
1.6
Cane
reed
1.2
Emerald
precious stone
1.8
Soap
soft
1.4
Wagon
coal
1.6
Cheese
English
2.0
Cottage
college
1.6
Drum
sound
1.0
Naughty
boy
1.2
Happy
healthy
2.2
Exacting
demand
2.6
Small
boy
1.0
Thirsty
man
1.2
Difficult
lesson
1.2
Playful
boy
1.2
Painful
wound
1.2
Impulsive
dashing
1.8
Grief
sorrow
1.0
Faithfulness
dog
1.0
Thought
good
1.4
Provocation
insult
1.4
Credit
great
1.6
Contentment
happiness
1.0
Fear
death
1.4
Religion
faith
1.0
Mercy
God
1.2
Profanity
bad word
1.0
Sinful
man
1.0
Average
1.4
Average
1.3
THE PSYCHO-PHYSIOLOGY OF A FAST.
229
June 5:
June 6 — Continued
Stimulus
Reaction
Reaction
Stimulus
Reaction
Reaction
word.
word.
time.
word.
word.
time.
Oppose
enemy
1.2"
Infinite
God
1.0"
Enter
house
1.2
Brave
soldier
1.4
Drive
horse
1.0
Ornamental
church
1.0
Lecture
public
2.2
Dreadful
fight
1.4
Flag
wave
1.0
Chance
good
1.4
Ivory
white
1.0
Quarrel
men
2.0
Bed
sleep
1.2
Conscience
good
1.2
Fountain
water
1.0
Scandal
bad
1.8
Pie
lemon
1.6
Evil
bad
1.6
Awake
morning
1.4
Dull
night
1.4
Average
1.3
Many
friends
1.8
=
Green
leaves
1.2
June 7:
Divine
God
1.0
Irritate
nerves
1.0"
Terror
enemy
1.2
Tame
animal
1.0
Spite
hatred
1.4
Feed
animal
1.2
Advice
council
2.0
Imagine
vision
1.0
Contempt
enemy
1.8
Suffer
pain
1.0
Dispute
question
1.2
Dinner
good
1.2
Telephone
friend
2.6
Raft
sea
1.2
Chart
fever
1.8
Average
1.4
Glove
hand
1.0
=
Bird
sing
1.2
June 6:
Afraid
lion
1.0
Scold
child
1.0"
Blue
sky
0.8
Walk
street
1.0
Anxious
desirous
1.2
Punish
criminal
2.2
Long
street
1.0
Smell
odor
1.2
Audacious
hero
1.2
Send
letter
1.4
Expression
vocal
1.2
MU1
flour
1.0
Mistake
great
1.2
Elbow
hand
1.2
Devotion
church
1.2
Milk
white
1.0
Errand
boy
1.0
Scissors
cut
1.2
Expense
great
1.4
Moon
night
1.2
Quiet
night
1.4
Average
1.14
FECES.
In the days preceding the fasting period, there was more or less regu-
lar defecation, but the special interest in the feces in connection with
this fasting experiment has to do with the defecation immediately pre-
ceding the fast and that on the days following the fast. After the
evening meal on April 13, L. had a large defecation, as was noted in
the history for that day. There was no defecation, however, through-
out the entire fasting period, as no feces were passed from the time of
the defecation on April 13 until 5h30m p. m. on May 15, i. e., about
8 hours after the first food had been taken. It was suggested to the
subject that it would be desirable, especially on the first day or two, to
empty the lower bowel with a warm-water enema, but he preferred not
to do this.
The defecation on May 15 was coincidental with a severe attack of
colic occasioned by the taking of an excessive amount of acid fruits,
which flooded the stomach and the intestinal tract. The defecation,
which was copious, contained a few hard, well-formed lumps of feces
about 1 cm. in diameter and with a total length of 6.5 cm. The rest
of the material was spongy and soft, running like liquid when turned
from the vessel. The feces had a nauseating odor, necessitating
frequent access to the outdoor air in transferring and handling the
material. Another defecation took place about 8 p. m., a third shortly
afterward, and still another during the night. The feces were all of a
very soft and liquid consistency, and of a light yellowish-brown color.
As the hard material was obviously entirely different in nature from the
soft material it was removed, and probably this alone can here be con-
sidered as in any way approximating fasting feces. The second and
third defecations were tested with litmus paper and found to be strongly
acid, probably due in part to the organic acid present in the fruit juices.
The fact that there were no feces throughout the 31 days of this pro-
longed fast is of special significance, as it is commonly stated that fast-
ing men excrete from 2 to 5 grams of dry fecal material each day.
In the earlier experiments at Wesleyan University, no evidence was
found of what might be called strictly fasting feces. In the prolonged
fasting experiment with L., since the last defecation prior to the fast
took place only a half hour after the last meal on April 13, at least a
portion of the feces of May 15 might be expected to result from the food
on April 13, so that we find it difficult to determine what proportion,
if any, of the material defecated should be ascribed to the fasting period.
Unfortunately the exigencies of the situation, especially in view of the
illness of the subject, made it impracticable to preserve and prepare
these feces for a microscopical examination. This is much to be
regretted, as some light might have been thrown upon their source.
The amount was, however, extremely small, as the total air-dried
230
FECES. 231
material from the hard portion of the first defecation amounted to but
20.8 grams.
The amount of fasting feces reported in observations made by other
investigators is somewhat difficult to explain, except by the fact that
they are based in large part upon Mueller's observations on Cetti,
who fasted 10 days. It should be noted, however, that during the
entire fast Cetti smoked cigarettes more or less and unquestionably
shreds of tobacco found their way into the alimentary tract. While
these shreds of tobacco by no means formed the bulk of the fecal
material, they doubtless stimulated peristalsis, which caused a some-
what rapid movement along the alimentary tract of the epithelial
debris and residue of the digestive juices. In view of this probable
stimulation in the experiment with Cetti it appears fortunate that our
subject L. did not use tobacco in any form during his fast.
It is by no means clear whether the weights recorded by Luciani
for the feces in Succi's fast are for dry material — as interpreted by
Mueller1 — or whether the material recovered from the enemata was
dried down to the consistency of normal feces. Luciani's expression,
"un residuo solido di consistenza pastosa,"2 would seem to imply that
the material was by no means anhydrous. With this interpretation,
the amount of dry material found in Succi's 30-day fast would be not
far from 37.5 grams or a little over 1 gram per day, instead of 5 grams
per day, as computed by Mueller.
In the experiment with L. the fecal material which obviously be-
longed either to the fasting period or to the food period prior to the
fast was separated, dried, and analyzed. The results of the analysis are
as follows:
Nitrogen 5 . 26 p. ct.
Fat 21.11
Fat saponified 25 .42
Calcium oxide 2 . 859
Magnesium oxide 1 . 026
Total weight of dried material 20.8 grams.
An inspection of these results shows no noticeable difference from the
composition of ordinary feces, so that we have no chemical indications
of feces which can be specifically ascribed to the fasting period. Con-
sequently the only definite conclusion that can be drawn is that during
the 31-day fast there was no positive evidence of the existence of fast-
ing feces.
In this connection, the following report of Dr. Arthur I. Kendall, of
the Harvard Medical School (now professor of bacteriology in the
Northwestern University Medical School), on the flora of the intestinal
tract of our fasting subject, is of special interest.
Mueller, Archiv f. path. Anat. u. Physiol., 1893, 131, Supp., p. 107.
2Luciani, Fisiologia del digiuno; studi sull' uomo, Florence, 1889, p. 37.
OBSERVATIONS UPON THE BACTERIAL INTESTINAL FLORA OF A
STARVING MAN.
By Arthur I. Kendall.
The question of the composition of the normal bacterial flora of
adult man has never been satisfactorily settled, although the consen-
sus of opinion appears to be that B. coli is the form most commonly
found. The observations recorded below, while not conclusive,
furnish information which tends to show that at least three organisms
may persist in the intestinal tract for a month after all food is withheld,
and in this sense these bacteria are noteworthy. The history of the
case needs no comment here, other than to state that the subject had
no food for 30 days prior to the taking of the sample herein reported.
The material for study was obtained from an enema of sterile physi-
ological salt solution, 300 c.c. in all, which was injected into the rectum,
retained for approximately 5 minutes, and recovered in almost full
volume. The return fluid (collected in a sterile container, with appro-
priate precautions) was turbid, with but little odor, practically color-
less, and, except for a very small amount of cell detritus, free from
particulate matter. No fecal material was recovered.
The fluid was plated in plain agar, in a dilution of ttttwid while a
portion (undiluted, and diluted -n/xnr) was examined for anaerobes
and certain other bacteria. The latter tests were negative.
The total count on agar plates (in duplicate) was 131 and 133 col-
onies, respectively, giving a total of 1,310,000 and 1,330,000 bacteria
per cubic centimeter of washings. Of the 131 colonies, 4 were identical
culturally with B. mesentericus, 17 were found to be Micrococcus ovalis
of Escherich, and the remaining 110 were found to be B. coli. B. coli
was also recovered from fermentation tubes inoculated with 1 c.c. of a
t:ooo,ooo dilution of the washings, thus confirming the count by the
plate method for this organism.
These results, while not striking, are interesting for two reasons :
(1) Certain bacteria appear to be able to live upon the intestinal
secretions, even when all food is withheld for at least a month.
(2) It appears to be impossible to sterilize the intestinal tract by
simple starvation. This latter consideration should be of clinical
interest, since it is customary in certain diseases to try to "starve out"
bacteria from the intestinal tract.
232
EXCRETION THROUGH THE SKIN.
So great is the total excretion from the body, in the respiration, urine,
and feces, that aside from the sensible perspiration, the skin as a path of
excretion is rarely considered in any discussion of the loss of body
material. But leaving the sensible perspiration entirely out of con-
sideration, the skin plays an important part, for there is cutaneous
respiration, including both the absorption of oxygen and the excretion
of carbon dioxide; there is a very considerable insensible perspiration,
which in its strictest meaning refers to the vaporization of water from
the skin surface; and there are the excretions of both nitrogenous
material and chlorides through the skin.
Although the excretion of gaseous and solid material through the
skin of the fasting man would normally be expected to be at a mini-
mum, it seemed desirable, in order to establish sharp balances of the
nitrogen and particularly of the salts, to determine carefully the cuta-
neous excretion of soluble nitrogenous materials as well as the sodium-
chloride excretion. It was not possible to measure the cutaneous
respiration of our subject in any of the forms of respiration apparatus
used in the fasting experiment, for in the calorimeter the cutaneous
respiration is measured with the pulmonary respiration, and with the
respiration apparatus no provision is made for the measurement of
the cutaneous respiration.
The excretion of the nitrogenous material and chlorides through the
skin as the fast progressed was, however, of particular significance, and
arrangements were made for determining these. By nitrogenous
material is meant not the dead cuticle, but the excretion of water-
soluble material, chiefly in the form of urea. In order to determine
this accurately, the body of the subject was given a thorough washing
before the fast. He was then sponged with distilled water and a
freshly extracted and dried cotton union suit was placed upon him.
At the end of the week the union suit was removed, the subject was
again sponged with distilled water, and another freshly extracted and
dried cotton union suit was given him. The union suit which had been
removed was then carefully extracted with distilled water and the
extract water evaporated after the addition of acid. The water in
which the subject had been bathed was also saved and evaporated
after the addition of acid. The entire operation was in the skilled
hands of Mr. T. M. Carpenter. By this procedure it was expected
that the perspiration accumulating during the week would be absorbed
by the cotton union suit and the soluble solids, including salts, urea,
or other material, would be extracted by the distilled water.
The nitrogen was determined by the Kjeldahl method. The chlorine
was determined by titration with silver nitrate and sulphocyanate.
233
234
A STUDY OF PROLONGED FASTING.
The total amount of nitrogen and chlorine found each week is given
in table 22, in which it is seen that the nitrogen ranged from 0.73 gram
in the first week to 0.30 gram in the last week, and the chlorine from
0.39 gram and 0.41 gram for the first two weeks to 0.18 gram in the last
week.
It will be noted that as much as 0.1 gram per day of nitrogen
in water-soluble material may be excreted through the skin during the
first week of fasting and that in all probability this method determines
the minimum rather than the maximum amount, since unquestionably
there is a continual transformation of urea to ammonium carbonate
with a loss of ammonia. On the other hand, it is probably true that
the secretory activity of the skin decreased somewhat as the fast pro-
gressed, as is evidenced by the values for both nitrogen and chlorine.
This loss of nitrogen through the skin has special significance in con-
nection with so-called "nitrogen-balance experiments."
Table 22. — Cutaneous excretion of nitrogen and chlorine in experiment with L.
Date.
Nitrogen.
Chlorine.
1912.
Apr. 13-Apr. 201
Apr. 20-Apr. 27
Apr. 27-May 4
May 4-May 11
gm.
0.73
.39
.31
.30
gm.
0.39
.41
.23
.18
*The subject was bathed on the evening of April
13 and at the end of each week thereafter.
It has previously been shown1 that during severe muscular work as
much as 200 milligrams of nitrogen may be excreted through the
skin per hour. If, therefore, the excretion of nitrogen in a fasting exper-
iment with minimum activity amounts to 0.1 gram or more per day,
it is obvious that nitrogen-balance experiments which do not take into
account this loss through the skin will not give accurate results. I
am unaware of any determinations of this kind made on a fasting man,
although Zuntz and his co-workers on Monte Rosa recorded the loss of
nitrogen and chlorine through the skin in their experiments on the high
Alps.2
The amount of chlorine excreted through the skin of L. was relatively
small, being approximately from 50 to 60 milligrams per day in the
first 2 weeks of the fast. During the fourth week of the fast only
Benedict, Journ. Biol. Chem., 1906, 1, p. 263.
2Schwenkenbecher and Spitta (Arch. f. exp. Path. u. Pharm., 1907, 56, p. 284) found about
0.3 gram each of nitrogen and sodium chloride per 24 hours with a healthy person in bed. Taylor
(Journ. Biol. Chem., 1911, 9, p. 21) found with two men at work but no visible perspiration per
day 0.028 gram sulphur, 0.003 gram phosphorus, and 0.190 gram nitrogen in one case and the
corresponding figures for the other were 0.015, 0.002, and 0.160.
EXCRETION THROUGH THE SKIN. 235
about 25 milligrams per day were thus excreted. While the loss of
nitrogenous material from the surface of the skin by decomposition
might be considerable throughout the week, it is hardly probable that
any large amount of chlorine would be mechanically lost. Thus these
values probably represent very nearly the actual cutaneous excretion
of chlorine during this period. In this connection it is of interest to
note the recent work of Wahlgren,1 indicating that the skin is one of
the principal reservoirs for chlorine in the body. Finally, attention
should be called to the discussion of the excretion of water-vapor
through the skin,2 in which the evidence points towards a decreased
secretory activity of the skin as the fast continued.
Wahlgren, Archiv f. exp. Path. u. Pharm., 1909, 61, p. 97.
2See page 373.
URINE.
Urine analysis has in the past decade undergone a striking revolution
as a result of the development of unique and exceedingly accurate
methods by Folin. Formerly clinical examinations of urine included
urea determinations, usually by the hypobromite method, and quali-
tative or roughly quantitative estimations of phosphates, chlorides,
etc., but to-day the intelligent clinician deals only with the 24-hour
excretion of the various urinary constituents. The introduction of
the Kjeldahl method did much to advance our knowledge of the con-
stituents of the urine by giving us information as to the total organic
nitrogen, but it remained for Folin to show us the methods for the
partition of the nitrogen in the urine and its significance. The ammo-
nia, urea, uric acid, creatinine, and creatine in the urine then began to
be of much greater significance than was the total nitrogen; but in all
these advances in the development of urine analysis, and particularly
in the interpretation of the results, we find stress invariably laid upon
the nitrogenous constituents. To such a degree is this true that we
are inclined for the most part to think of the urine solely as a path for
nitrogen excretion.
Our previous experience with fasting subjects, however, has shown us
that in the urine we have not only indices of the protein katabolism,
but that with acetone, diacetic acid, and /3-oxybutyric acid present,
we have indices regarding the defective fat katabolism; furthermore,
the inorganic constituents, such as chlorine, phosphorus, sulphur, and
the alkaline bases, give us evidence as to the mineral metabolism, the
sulphur excretion also having an importance in interpreting the protein
katabolism. It was therefore essential to study the urine of our fasting
subject not only from the standpoint of protein katabolism, but like-
wise from every other possible standpoint, so that complete analyses
were necessary. In carrying out such a study of the fasting urine, we
have depended more largely upon the results of our former study of
fasting subjects1 than in any other part of the research.2
GENERAL ROUTINE OF COLLECTION AND SAMPLING.
In order to give us as much information as possible about the pre-
vious dietetic habits of this man, particularly for the few weeks prior to
Benedict, Carnegie Inst. Wash. Pub. 77, 1907, pp. 345-419.
2 After the pages of this book were in page proof, my attention was called to the article from
Aoyama's clinic in Tokio, by Watanabe and Sassa, entitled "Die Harnanalyse wahrend des zwei-
wochigen Hungerns eines Mannes" (Zeitschr. f. Biol., 1914, 64, p. 373), issued from Munich en
August 27, 1914, but not received here until late in November. It is thus impossible to make any
comments upon this interesting paper. The authors studied body-weight, measurements of the
body, body-temperature, pulse, respiration, and the blood, but laid special emphasis upon exten-
sive urine analyses. Their findings are, for the most part, in full conformity with those recorded
here.
236
URINE. 237
the fasting experiment, L. was requested to measure and sample the
urine each day from the first of April until he reached the Nutrition
Laboratory on April 10, preserving the samples with chloroform. This
he did most carefully, his training as a pharmacist assisting him materi-
ally in carrying out the routine accurately. When it is considered that
he was traveling rapidly and while on the steamer was obliged to make
all his observations and measurements in the narrow confines of a state-
room having three other occupants, it will be seen that it is much to
his credit that the records were so carefully kept. Although it was im-
possible to keep an accurate record of the amount of food eaten, and
particularly the kind and amount of the various proteins, a study of
these urine samples should give some information as to the normal con-
sumption of protein by this individual.
From the time of his arrival at the Nutrition Laboratory, the col-
lection, measurement, and sampling of the urine were made by members
of the laboratory staff. Particular attention was given to the urine
excreted during the fasting period, as it was especially important to
study the entire output of the body at this time.
When the preliminary arrangements were made for the analyses and
their assignment to the various members of the laboratory staff and its
co-workers, it soon became apparent that the number of determinations
necessary would require a greater volume of urine than would ordinarily
be passed by a fasting man. It was therefore arranged, in accordance
with a suggestion made by Dr. Cathcart, to provide the subject with a
liberal and constant supply of drinking-water. Furthermore, the
smallest volume of sample which would give accurate determinations
was carefully considered in order to obtain the greatest number of
results with the available material. Had it not been for the recent
development of the new Folin methods, it is probable that much
valuable data would have been lost. For example, while formerly
300 c.c. or even more were required for the determination of the uric
acid, with the new Folin method 5 c.c. would suffice. Many of the
determinations of the ammonia as carried out by the new method were
also made with a relatively small amount of urine.
Before the subject came under observation the time of urinating was
more or less irregular. During the three food days preceding the fast,
the subject urinated at irregular times, although ending each day at
approximately 8 a. m. During the fasting period, he was required
to empty the bladder immediately after coming out of the bed calori-
meter in the morning, this being usually not far from 8 o'clock. He
again emptied the bladder shortly before entering the bed calorimeter
at night. We were thus able to divide the urine into two periods, each
approximately 12 hours in length. Use was made of this routine in the
latter part of the fast to study the apportionment of the nitrogen and
ammonia excretion between the day and night periods.
238
A STUDY OF PROLONGED FASTING.
The urine was collected at the laboratory by having the subject
urinate into a previously dried and weighed bottle; the bottle and con-
tents were then carefully weighed and the urine measured in a gradu-
ate and the volume recorded. Shortly after the experiment began,
it was considered advisable, in accordance with a suggestion made by
Dr. Folin, to add sufficient distilled water to bring the urine to a definite
volume each day. Under these circumstances a normal excretion of
urine of 600 to 700 c.c. would be weighed and its specific gravity deter-
mined; it would then be immediately diluted to 1,000 c.c. and division
made for the various analyses. This procedure was very satisfactory
and minimized the calculations.
COMPOSITION OF THE URINE PRIOR TO THE FASTING EXPERIMENT.
As a general indication of the character of the subject's urine prior
to the fasting experiment, we have fragmentary data regarding the
urine passed on the 10 days before he arrived at the laboratory and for
the 3 food days in Boston before the fast-
ing period began. The volume and nitro- ^^tlTS'""''™
gen content of this urine are given in table
23, the nitrogen being determined by the
Kjeldahl method. In addition to the
tabulated data, the ammonia-nitrogen
was determined for the last 3 days by
the old Folin method, the amounts found
being 0.67, 0.65, and 0.59 gram respect-
ively. On the last 2 days the heat of
combustion was 129 and 104 calories
respectively; the total carbon in the urine
for the same days was 11.41 and 9.08
grams respectively. The acidity was
determined on but one day (April 11-12),
this, expressed as cubic centimeters of
N/10 NaOH solution, being 409 c.c.
These data will be used in subsequent discussions and are here recorded
to avoid confusion with the regular examinations of urine in connection
with the fasting experiment.
PHYSICAL CHARACTERISTICS OF THE FASTING URINE.
In considering a subject as complex as is the urinary excretion, it is
advantageous to note first the physical characteristics and then the
chemical composition. The influence of various physical agencies,
particularly the relation between the amount of water drunk and the
volume of urine, may not be without influence upon the chemical
composition, for under certain conditions there may well be a washing
Date.
Volume
Nitrogen
of urine.
in urine.
1912.
c.c.
grams.
Apr. 1-2
1,095
12.07
2-3....
975
10.90
3-4. . . .
1,208
16.03
4-5. . . .
608
8.83
5-6. . . .
581
11.90
6-7. . . .
818
10.45
7-8. . . .
795
11.30
8-9....
1,215
13.36
9-10. . .
1,151
12.25
10-11...
1,485
17.02
11-12. . .
1,521
15.92
12-13...
1,528
14.48
13-14. . .
1,441
11.54
URINE.
239
out of the end-products of protein katabolism by the excess water.
Furthermore, the specific gravity (when accurately determined) and
also the total solid matter have an interest second only to the chemical
constituents of the urine. Accordingly in table 24 a record is given
of the water consumed, the volume of urine, the amount of urine in
grams, the water in the urine, the ratio of water in the urine to the
water consumed, the specific gravity, the total solids, either computed
or determined, and the ratio of the total solids to the specific gravity.
Table 24.
-Relations between water consumed, water in urine, specific gravity, and total solid
matter in experiment with L.
a hi •
a
Day of
73
«
s
■
a
o
6
a
1
"o
6
a
1
O
6
a
°E
3
a
Ratio of water i
urine to wate
consumed (d-*-a'
T3 .
o a
CO -g
o '3
a,
B
°3 .•
Date.
fast.
6
fa
8
■*»
a
j3
>
43
'3
e
«3
"3
9
do
3
O
H
Ratio
solids
gravit}
A
B
C
D
E
F
G
H
1912
gm.
ex.
gm.
gm.
gm.
Apr. 14-15
1st
720
660
673.7
630.2
0.875
1.0206
243.51
15-16
2d
750
468
482.0
436.6
.582
1.0303
245.38
16-17
3d
750
565
581.2
530.6
.707
1.028
250.62
17-18
4th
750
713
730.5
674.4
.899
1.0246
256.13
18-19
5th
750
667
682.6
633.5
.845
1.023
249.09
19-20
6th
750
610
623.9
577.8
.770
1.0236
246.07
20-21
7th ... .
750
524
536.5
495.9
.661
1.0242
240.58
21-22
8th ... .
750
587
601.0
556.9
.743
1.0235
244.14
22-23
9th....
750
607
622.1
575.3
.767
1.0241
246.81
23-24
10th....
750
565
577.8
535.3
.714
1.0235
242.49
24-25
11th....
900
564
577.2
535.1
.595
1.0233
242.05
25-26
12th....
900
517
529.1
489.9
.544
1.0237
239.21
26-27
13th....
900
561
574.3
532.3
.591
1.0234
242.01
27-28
14th....
900
647
659.9
619.3
.688
1.0196
240.58
28-29
15th
900
758
768.3
735.8
.818
1.0134
232.50
29-30
16th
900
889
902.3
861.2
.957
1.0154
41.12
3.0
Apr. 30-May I...
17th....
900
848
860.9
821.4
.913
1.0153
39.51
3.0
May 1-2
18th
900
657
668.8
633.3
.704
1.0177
35.47
3.1
2-3
19th
900
728
739.6
705.0
.783
1.0153
34.59
3.1
3-4
20th
900
699
708.7
678.6
.754
1.0143
30.06
3.0
4-5
21st
900
708
717.2
685.3
.761
1.013
31.88
3.5
5-6
22d
900
785
794.6
763.4
.848
1.0127
31.18
3.1
6- 7
23d
900
556
565.8
536.5
.596
1.0176
29.30
3.0
7-8
24th
900
750
759.5
727.5
.808
1.013
32.01
3.3
8-9
25th
900
713
722.1
691.8
.769
1.0135
30.32
3.1
9-10
26th
900
728
737.1
706.1
.785
1.0129
31.04
3.3
10-11
27th
900
653
662.7
631.2
.701
1.0147
31.52
3.3
11-12
28th
900
655
663.4
634.3
.705
1.0134
29.06
3.3
12-13
29th
900
697
705.8
676.2
.751
1.0129
29.64
3.3
13-14
30th
900
771
780.3
750.7
.834
1.0119
29.58
3.2
14-15
31st
900
566
574.5
547.4
.608
1.0150
27.07
3.2
lfThe amounts of water in urine from Apr. 14-15 to Apr. 28-29 have been obtained by means
of computed amounts of total solids.
2Calculated by means of the average ratio (3.2) of total solids to specific gravity, determined
in the last 16 days of the fast. For the formula U3ed in the computation, see Benedict, Carnegie
Inst. Wash. Pub. No. 77, 1907. page 354.
240 A STUDY OF PROLONGED FASTING.
VOLUME OF URINE.
In the publication giving the results of the earlier fasting research,
it was clearly brought out that no one factor affects the volume of
urine as does the volume of the water ingested, particularly when the
volume of drinking-water is over 1,000 c.c. In the experiment with
L., the water consumed was always under 1 liter and hence the influence
of the amount of the drinking-water on the volume of the urine was not
so obvious.
The volume of urine varied from 468 c.c. on April 15-16 to 889 c.c.
on April 29-30. The average volume was 659 c.c. On April 24-25,
the volume of the water taken was changed from 750 c.c. to 900 c.c,
and it was expected that this change would materially affect the volume
of the urine. An inspection of the data shows, however, that it pro-
duced no marked effect upon the water excreted in the urine, at least
during the first 3 days. Thus, on the 9 days from the second to the
tenth days of the fast, inclusive, when the subject drank 750 c.c. of
water daily, the average volume of urine was 590 c.c. per day and on the
3 days from the eleventh to the thirteenth days of the fast, inclusive,
when the daily amount had been increased to 900 c.c, the average
volume of urine per day was 547 c.c. Subsequently there was a dis-
tinct tendency for the urine volume to increase and on the next 10 days
the average volume was 728 c.c, an increase of 138 c.c. over the period
when 750 c.c. of water was taken, closely approximating the increase in
the amount of drinking-water. It is to be noted, however, that at this
stage of the fast, 900 c.c. of drinking-water was proportionately large
for this man's needs, since he had decreased materially in weight. In
the earlier fasting experiments, when several liters of water were taken
daily, the amount of water drunk unquestionably influenced the
amount of water in the urine excreted, but with the comparatively
small amount taken by the subject L., the effect was evidently at a
minimum, and the absence of any flushing-out of the end-products of
protein katabolism simplifies the subsequent discussion.
In comparing the water drunk with the volume of urine, the discus-
sion may be based more advantageously upon the water in the urine.
A determination of the solids in the urine was made only on the last
16 days of the fast, but amounts for the earlier days of the fast have
been computed and these are sufficiently accurate to use in this con-
nection for obtaining the amount of water in the urine.
From the values given in column e of table 24, showing the relation-
ship between the water in the urine and the water consumed, it is seen
that in the first days of fasting, when the amount of water taken by
the subject was only 750 c.c, about 74 per cent of the water consumed
appeared in the urine. When the amount of drinking-water was
increased to 900 c.c, there were marked disturbances in the ratio for
the next 7 days. The widest variations in the entire series appeared in
URINE. 241
these 7 days, namely, from 54 per cent on April 25-26 to as high as 96
per cent on April 29-30. On the other hand, from May 1-2 to the end
of the fast, the ratio remained very constant, with but minor variations
above or below the average figure for the whole series of 0.744, essen-
tially that obtaining on the first 10 days of the fast. The disturbance
in the ratio was found, therefore, only during the 7 days immediately
following the change in the amount of water consumed from 750 c.c.
to 900 c.c.
This surprising constancy in the ratio between the water of urine and
the water consumed, aside from the 7 days mentioned, is difficult to
explain, particularly since at least two factors might have been ex-
pected to disturb this relationship. During the first 10 days of the
fast, there was a considerable loss of preformed water from the body,1
ranging from 769 grams to 183 grams. This loss of water might be
expected to increase the volume of urine during these days. Indeed, if
the volume of urine were not increased, in the absence of other evidence,
this might be taken as an argument against such an excretion of pre-
formed water.
Another point which should be taken into consideration in this
connection is that one would expect that with the greatly diminished
body substance, the amount of drinking-water consumed might exceed
the physiological need and hence would disturb the relationship
between the volume of urine and the volume of water consumed. On
the other hand, it is well known that during fasting there is a tendency
for all the tissues to become water-rich, this retention of water possibly
compensating for the decrease in the physiological need following the
decrease in the size of the organism.
That the relationship between the water of urine and the water
consumed is reasonably constant, even when the quantity of water is
but 1 liter or less, is likewise substantiated by calculations from data
published regarding Cathcart's experiment with Beauts, in which the
volume of drinking-water was also constant, i. e., 1,000 c.c. per day. By
using the data for the volume of urine, the specific gravity, and the
factor 3.2 (see page 244), we have computed the total solids and also
the water in the urine for the sixth, seventh, eighth, tenth, eleventh,
twelfth, and fourteenth days of Beaute's fast. The ratio of water in
the urine to the water consumed was for the several days as follows:
0.813; 0.649; 0.630; 0.940; 0.612; 0.576; 0.662. The average value
was 0.697, which is not materially different from 0.750, the average
obtained for the values for L. when the 7 days referred to have been
omitted. The high value of 0.940 found on the tenth day with Cath-
cart's subject exceeds any found with our subject L.
In general, then, the volume of water in the urine is approximately
75 per cent of the amount of drinking-water taken, even when but
1See discussion of preformed water loss on page 408.
242 A STUDY OF PROLONGED FASTING.
1 liter or less is taken, provided the intake of water is constant. It
is obvious, however, that this would hold true only when the factors
influencing the loss of water, such as environmental temperature and
exercise, also remain constant.
The evidence is clear, therefore, that even with the small amounts of
water taken in this fast, there was a reasonably constant relationship
between the water consumed and the water in the urine, the absolute
fluctuations in the volume of urine noted being so small that there
could have been no disproportionate washing out from the tissues of
the crystalline end-products of protein katabolism. There is, to be
sure, a distinct increase in the average volume of the urine after the
twelfth day, but in this experiment we deal with an average increase
of approximately 100 c.c, and hence these absolute variations in volume
are not to be compared with the very large variations noted in the
earlier fasts at Wesleyan University, when the amount of drinking-
water varied within wide limits.
SPECIFIC GRAVITY.
The specific gravity of the urine was carefully determined by Miss
Alice Johnson, at a constant temperature of 20°C, on a Westphal bal-
ance. The position of the scale when the weight was suspended in
distilled water was accurately checked, the temperature of the sur-
rounding liquid being invariably artificially maintained at 20° C.
Thus the specific gravity of the urine for each day of the fast was
readily obtained to the fifth significant figure.
These values, which are given in table 24, ranged from 1.0303 on
April 15-16 to 1.0119 on May 13-14, and are therefore well within
normal limits. As the fast progressed, there was a distinct tendency
for the specific gravity to decrease, although for the first 10 or 12 days
it was approximately constant at about 1.024, falling thereafter some-
what sharply and remaining at about 1.015 for the remainder of the fast.
This approximation to constancy may be in part accounted for by the
approximately constant volumes of urine passed. In the short fasting
experiments at Wesleyan University, the specific gravity ranged be-
tween 1.0338 and 1.0032, but the lowest specific gravity was accom-
panied by a very large volume of urine and the high specific gravity
by a very small volume of urine.
It is obvious that the nature of the solids dissolved in the urine has
a noticeable influence upon the specific gravity. For instance, a solu-
tion containing 100 grams of sodium chloride in 1 liter has a density at
15° C. of 1.073, while a solution of urea 1 to 10 has a density of but
1.028; consequently a large amount of sodium chloride in the urine
would have considerable effect upon the specific gravity. From other
fasting studies it is known that a large amount of sodium chloride is
excreted in the first days of fasting, which would thus increase the
URINE. 243
specific gravity. After the first few days, the excretion would be in
large part of an organic nature, accompanied by the usual salts other
than sodium chloride; hence we should normally expect to find the
specific gravity somewhat lower in the latter part of the fast.
So close is the relationship between the total solids and the specific
gravity that a formula has been in use for many years for computing
the total solids by means of this index. Thus, the approximate weight
in grams of total solids in 1 liter of normal urine may be calculated by
multiplying the last two figures of the specific gravity (as ordinarily
expressed in 3 decimal places) by the factor 2.33. The values for the
specific gravity were so used for computing the total solids in the urine
for a part of the fasting period, substituting 3.2 as the factor.
TOTAL SOLIDS.
In the pressure upon the laboratory staff necessary for carrying out
the many details of this elaborate research, the determination of the
total solids in the urine was unfortunately overlooked until the latter
part of the fast. The data secured in previous researches as to the
total solids for the first days of fasting are, however, fairly complete,
and we are thus able to supplement these by the important data which
were obtained in the last part of this fasting experiment. The proced-
ure followed in determining the total solids was primarily developed
for the determination of carbon in urine, and the description of the
method applies likewise to the method used for securing the carbon
excretion. Three samples of each specimen of urine were prepared in
the following manner :
A small soft-metal bottle-cap was first weighed and in this were
placed 50 milligrams of pure salicylic acid. With a carefully calibrated
pipette, 20 c.c. of urine were next added. The bottle-caps containing
the acid and the urine were then placed on the laboratory table in such
a position that a current of air from an electric fan would blow over
them. This drying was continued over night, usually for a period of
about 24 hours. The samples were next dried for 24 hours in a high
vacuum in a desiccator. Subsequently the lead capsules with their
dried contents were quickly weighed, the 50 milligrams of salicylic
acid being subtracted from the final weight of dry matter, thus giving
the weight of the total solids in 20 c.c. of urine.
In determining the carbon and the heat of combustion by this
method, it is unnecessary to dry the substance in the capsules com-
pletely in a high vacuum, but at the end of the 24-hour drying in the
current of air, the thick pasty material may at once be transferred
to the small capsules used in the combustion bomb. It is thus seen
that had we only delayed the weighing of the soft-metal bottle-caps for
24 hours, it would also have been possible for us to determine the total
solids in the urine for the first part of the fast.
244 A STUDY OF PROLONGED FASTING.
When the total amount of dry matter was determined, the contents
were afterwards carefully transferred from the metal bottle-cap to the
nickel capsule by means of a swab of ignited asbestos, the last traces
of solid material adhering to the bottle-cap being removed by a bit of
asbestos wool moistened with water. Finally, the material was dried
in an air-current to a pasty consistency and then placed in a high
vacuum until dry enough to burn.
This method has been previously described1 and need only be referred
to here. The 3 samples always gave perfectly agreeing results, indica-
ting that the drying was essentially complete, and testifying to the
skilful technique of Mr. Arthur W. Cornell, who carried out the deter-
minations of the total solids, carbon, and heat of combustion. The
results of the determinations of the total solids for the last 16 days of the
fast are given in table 24, and range from 41.12 grams on April 29-30
to 27.07 grams on the last day of the fast. From these absolute
determinations of the total solids, together with the volumes of urine
and the specific gravity, it was possible to compute a factor indicating
the ratio between the total solids and the specific gravity. This factor,
although higher than the value for normal individuals (2.33), remained
very constant in these later days of the fast, ranging from 3.0 to 3.5,
with an average of 3.2. The value 3.2, which represents the average
ratio between the total solids and the specific gravity in the last part
of the fast, agrees very closely with those found in three of the fasting
experiments with the subject S. A. B. in the Middletown research,2
namely, experiments Nos. 71, 73, and 75, the ratios being 3.0, 3.4, and
3.3 respectively. In experiment No. 77 with the same subject, much
larger amounts of sodium chloride were excreted, and this doubtless
was the cause, at least in part, for the lower ratio of 2.5 for this 4-day
fasting experiment. In the later days of the fasting experiment with
our subject L. there was undoubtedly a minimum sodium-chloride
excretion, and the constancy in the ratio between the total solids and
the specific gravity points towards an approximately constant relation-
ship between the organic and inorganic solids of the urine. Unfor-
tunately, we are unable to apportion the total solids between the min-
eral and organic constituents, since it was impossible to determine
the ash content of the fasting urine, owing to the deficiency in material.
It can only be pointed out here, therefore, that while the factor 3.2
is considerably larger than that accepted for normal people, namely,
2.33, it is probably explained in part by the fact that there were pro-
ducts of defective fat katabolism in the urine.
This average ratio, i. e., 3.2, was used for computing the total solids
for the first 15 days of the fast. The results of these computations are
also given in table 24. In comparing the values we find that the largest
1Higgins and Benedict, Am. Journ. Physiol., 1911, 28, p. 291.
Benedict, Carnegie Inst. Wash. Pub. 77, 1907, p. 355.
URINE.
245
amount was 56.13 grams on April 17-18. It is probable that, owing
to the increased sodium-chloride excretion in the first days of fasting,
the values are somewhat too high and that the factor used should have
been less than 3.2.
DAY AND NIGHT URINES.
While with a fast as prolonged as this the main interest lies in a
comparison of the urinary excretion from day to day and there is but
little interest in a subdivision of the day into 12-hour periods, yet we
have certain fragmentary data regarding the diurnal excretion of urine
which are of sufficient importance to record here. Usually the bladder
was emptied at 8 a. m. and at 8 p. m., the 24-hour day being thus divided
into two periods, i.e., from 8 a. m. to 8 p. m. and from 8 p. m. to 8 a. m.
While this subdivision was not made exactly each day, nevertheless
the variations were generally well within one-half hour. On one day
Table 25. — Periodic distribution of volume and nitrogen of urine in experiment with L.
Day period.
Night period.
Nitrogen.
Nitrogen.
Date.
Day
of
fast.
Duration.
Vol.
Propor-
Duration.
Vol.
Propor-
A
mt.
tion of
A
mt.
tion of
il
total for
total for
24 hours.
24 hours.
1912.
a.m.
p.m.
c.c. g
m.
p. ct.
p.m.
a.m.
c.c. g
m.
p. ct.
Apr. 18-19....
5th.
8h05m
to 7h 27m
307 .
yb 27m
to 7h 57m
360 .
19-20
6th.
7 57
8 06
312 .
8 06
8 05
298 .
20-21
7th.
8 05
8 02
256 .
8 02
8 04
268 .
21-22
8th.
8 04
8 05
298 .
8 05
7 59
289 .
22-23
9th.
7 59
8 05
315 .
8 05
8 05
292 .
23-24
10th.
8 05
8 03
305 .
8 03
7 57
260
24-25
11th.
7 57
8 08
284 .
8 08
8 04
280 .
25-26
12th.
8 04
8 05
267 .
8 05
7 55
250 .
26-27
13th.
7 55
7 22
288 .
7 22
8 14
273 .
27-28....
14th.
8 14
7 25
252 .
7 25
8 07
395 .
28-29
15th.
8 07
7 00
263 .
7 00
7 55
495 .
29-30
16th.
7 55
6 55
407 .
6 55
7 54
482 .
Apr. 30-May 1
17th.
7 54
6 53
328 .
6 53
8 06
520 .
May 1-2
18th.
8 06
6 48
284 .
6 48
8 00
373 .
2-3....
19th.
8 00
6 45
250 .
6 45
8 07
478 .
3-4....
20th
8 07
6 45
251
6 45
8 01
448 .
4-5....
21st .
8 01
8 02
'300 3
.71
i
to
8
8 02
8 10
x418 4
.22
53 .2
5-6....
22d..
8 10
8 02
335
8 02
7 57
450 .
6-7....
23d..
7 57
8 11
266 .
8 11
8 00
290 .
7-8....
24th.
8 00
8 00
»316 3
.94
47
8
8 00
7 55
!443 4
!30
52.2
8-9....
25th.
7 55
8 07
323 3
.84
48
5
8 07
8 05
390 4
.07
51.6
9-10....
26th.
8 05
8 12
282 3
.96
49
5
8 12
8 10
446 4
.04
50.5
11-12
28th.
8 08
8 08
242 3
.34
44
8
8 08
7 39
413 4
.12
55.2
12-13
29th.
7 39
8 05
300 3
.82
50
0
8 05
7 41
396 3
.82
50.0
13-14
30th.
7 41
8 00
311 3
.74
49
2
8 00
7 44
460 3
.86
50.8
14-15
31st .
7 44
8 09
266 3
.30
48
3
8 09
8 02
300 3
.53
51.7
KJra
246 A STUDY OF PROLONGED FASTING.
(May 10-11) the day period was 14 hours and 5 minutes and the night
period only 9 hours and 55 minutes, a discrepancy which resulted in the
omission of the day from the record of the periodic distribution of the
urine. Beginning with April 18-19, the volumes of the day and night
urines were separately recorded. Furthermore, toward the end of the
fast, periodic determinations of the nitrogen for the day and night
were made. The data thus obtained are given in table 25. Without
laying emphasis at this time on the absolute values of total nitrogen
excreted — a discussion which belongs later in this report — we may
properly consider the data in this table as indicating the division of
the urinary excretion between night and day when no food was taken.
The volume of urine during the day ranged from a minimum of
242 c.c. on May 11-12 to a maximum of 407 c.c. on April 29-30. The
average volume for the day urine was 293 c.c. The average volume for
the night urine was 376 c.c, an increase of 83 c.c. On several occasions
there were large differences between the day and night urines, which
are not easily explained. Thus, on April 28-29 there were but 263
c.c. in the daytime and 495 c.c. during the night. Although the day
period was but 11 hours and the night period 13 hours, this difference
in volume is very large. While in general the volumes for the day and
the night were not far apart, the average difference, as we have seen,
being but 83 c.c, it is indeed surprising that a larger volume was not
excreted during the day, for the subject drank his entire allotment
of 900 c.c. of water before 8 p. m., taking it in fairly regular portions
throughout the day. Usually but a small amount was left after 6 p. m.
The drinking of water was thus distributed to obviate the necessity for
urinating inside the calorimeter chamber during the night; and indeed,
throughout the fast, the subject retained the urine in the bladder
the entire night period.
On the days for which we have the data for both the volume and
the nitrogen of the urine, we find that on an average 42 per cent of the
total volume of urine and 48.1 per cent of the total nitrogen were
excreted during the daytime.1 If the volume of urine had had a
material effect upon the total nitrogen, one would expect that a some-
what greater proportion of nitrogen would have been excreted during
the night than was actually found, and it is reasonable to suppose that,
with the relatively small total volume of urine here involved there could
have been but little washing out of the nitrogenous products as a result
of the differences in volume. But in view of the well-known fact that
large quantities of water assist in washing out nitrogenous material, no
other explanation than the increase in the volume of urine seems possi-
ble for this small but positive increase in the nitrogen output during
the night.
1Obviously slight corrections for variations in the relative length of the day and night periods
should be made, but an inspection of the table shows that the percentage figures would not be
materially altered.
URINE. 247
CHEMICAL CONSTITUENTS OF FASTING URINE.
In modern urinary analysis, as carried out in connection with
a metabolism experiment, we have several distinct classifications or
subdivisions : First, the total nitrogen and the partition of nitrogen in
accordance with the analytical scheme of Folin; second, the acid
radicles, which would include the chlorine, phosphorus, sulphur, total
acidity, and /3-oxybutyric acid; third, the bases — calcium, magnesium,
potassium, and sodium oxides —whose excretion might perhaps be
discussed in connection with the ammonia, itself a base; and finally,
attention should be given to the determination of the reducing power,
total carbon, and the heat of combustion of urine.
With these various determinations, several ratios can be intelligently
discussed, of which the most important may be the ratio of nitrogen
to sulphur, nitrogen to phosphorus, carbon and energy to nitrogen,
and the heat of combustion to carbon. The presentation of the data
secured in our fasting experiment with L. will follow essentially the
analytical scheme thus outlined.
TOTAL NITROGEN.
From the early days of the Liebig titration method for determining
total nitrogen down through the various modifications to the present-day
development of the Kjeldahl technique, one of the first and most impor-
tant determinations of the constituents of the urine in metabolism experi-
ments has been that of the total nitrogen, the importance increasing in
proportion as the technique has been developed. After the accuracy of
the Kjeldahl method had been demonstrated and an exact method
was thus available, we were informed by Folin, as a result of his beauti-
ful systematic analyses of the urine in which the partition of the nitro-
gen has been made, that the value of the total nitrogen in the urine
did not have the importance which had formerly been attributed to it.
For instance, the determination of the carbon-dioxide excretion in a
metabolism experiment has great value in itself, but the apportionment
of this carbon dioxide to fat, carbohydrate, and protein katabolism
has a much greater value; similarly, although the determination of the
total nitrogen in the urine is not without value, yet the apportionment
of the nitrogen among the various constituents of the urine is much
more illuminating and scientifically intelligible than the amount of
the total nitrogen. The determinations of the total nitrogen in the
fasting urine were therefore made primarily as a preliminary to con-
sidering the partition of the nitrogen.
Comparison of Total Nitrogen Excretion of L. with that of Other Fasting Subjects.
While the total nitrogen has been determined in the greater number
of prolonged fasting experiments, in relatively few of these studies has
the determination been made by the modern Kjeldahl method, the
248 A STUDY OF PROLONGED FASTING.
results in many experiments having been recorded in terms of "urea."
Nevertheless, the nitrogen values found in several fasts of 7 or more
days are considered of sufficient importance in connection with the
study of the fasting urine of our subject to be reproduced here and
are accordingly recorded in table 26.
In this table the body-weight at the beginning of the fast is given
for nearly every subject, and frequently for comparison the nitrogen
excretion is included for the day prior to the fasting. The values for
L. are first shown, these being followed by the nitrogen found in Succi's
fasts. Unfortunately these latter values are not strictly comparable
with the others, owing to the differences in methods of determination.
Those found for Cetti can be relied upon, as can those for Beauty,
Schenk, Tosca, and S. A. B. The values reported for Succi for the
London and Naples fasts are undoubtedly somewhat low, but those
for the Florence fast have been corrected by Munk. Even when these
points are taken into consideration, the most striking feature in this
whole group of results is the fact that the nitrogen excretion of our
subject L. continues to be extraordinarily high to the fifteenth day of
the fast, and, indeed, throughout the remainder of the fast the values
are noticeably higher than those found in any other study of prolonged
fasting.1 Values as high, and even higher, are shown for Cetti for
the 10 days of his fast, and also for the 7-day experiment of S. A. B.,
but in none of the longer fasts are such high values so continuously
shown. In Succi's 30-day fast in Hamburg the value found for the
last day (8.42 grams) was higher than that for the thirtieth day of
the fasting experiment with L., but the earlier values were measurably
lower. Another point of interest is that the general tendency is toward
a low nitrogen output on the first day of the fast, with a higher nitrogen
excretion on the subsequent one or two days. This characteristic is
shown in the fasts with L., S. A. B., Tosca, and Beauts, and may
easily be attributed to the protecting action of the body-storage of
glycogen during the first few days.
One striking fact in connection with the high nitrogen output in
L.'s fast is that on the fifteenth day there was a sudden fall of nearly
2 grams. An inspection of the values for the other subjects shows
that in all of the fasts this sudden fall in the nitrogen excretion occurred
at some point. Thus, in Succi's fast at Florence, there was a fall of
1 gram on the eighth day; in the London fast there was a fall of 1.2
xNone of these subjects shows as absolute a minimum value for nitrogen excretion as was found
on one day with Grafe's insane subject (Grafe, Zeitschr, f. physiol. Chemie, 1910, 65, p. 21), when
the very low excretion of 1.057 grams of nitrogen was found. Since the body-weight of this
subject was at the time 49.25 kilograms, this excretion would correspond to approximately 0.0215
gram of nitrogen per kilogram of body-weight. This surprisingly low value is difficult to explain,
for while Grafe states that during the latter part of the experiment the urine was frequently spon-
taneously passed and hence the 24-hour periods could not be accurately determined, yet his dis-
cussion of this low value of 1.057 grams indicates that he believed it represented a 24-hour excre-
tion of nitrogen. This remains the lowest value that we have as yet seen reported in any fasting
observation on men or women.
URINE.
249
grams on the seventh day; in the Naples fast, the nitrogen output
fell 1.8 grams on the eighth day; in the Rome fast, it fell 1.8 grams on
the ninth day; while in the Vienna fast, it fell 2.9 grams on the tenth
day. With Cetti there was a fall in the nitrogen excretion of 2 grams
on the eighth day; with Beaute* the decrease was 1.1 grams on the
seventh day; with Schenk it was 1 gram on the tenth day; with Tosca,
Table 26 -
-Nitrogen eliminated in urine
daily by fasting subjects
Day of fast.
co
5
M
CO
d
CO
Succi.
o5
o
r3
M
©
'43
o
O
CO
O
%
CO
ui
CO
WU
3
03
8
PQ
m
o
M
eo
CO
"5
^5
a
<u
A
a
GO
03
8
CO
o
GO
o
IQ
OJ
»o
PQ
<
03
■ CO
<u o
5 --
r1 CO
fo CO
00
a oo
A
o
co rS
a ja
Pi CO
i< CD
CO
O
aj '£
c3
a
a
CO
>
a in
Last food day .
1st
gm.
11.54
7.10
8.40
11.34
11.87
10.41
10.18
9.79
10.27
10.74
10.05
10.25
10.13
10.35
10.43
8.46
9.58
8.81
8.27
8.37
7.69
7.93
7.75
7.31
8.15
7.81
7.88
8.07
7.62
7.54
7.83
6.94
gm.
117.85
15.19
12.13
15.25
14.08
14.12
11.13
10.31
9.37
8.56
7.43
8.67
7.88
3.86
5.87
5.66
6.05
6.78
6.00
5.54
4.82
4.27
3.52
5.23
6.11
6.65
5.57
5.90
6.16
4.49
7.28
gm.
311.41
12.62
12.00
10.46
9.80
8.57
7.86
7.50
7.10
6.51
6.86
6.14
5.95
5.36
5.35
5.40
4.13
3.98
4.64
4.00
5.16
4.66
4.70
4.32
4.12
3.81
3.40
4.13
4.56
4.36
4.79
4.75
4.80
4.80
3.95
5.00
4.77
4.99
5.56
5.82
gm.
28.99
8.72
8.45
9.05
8.51
9.87
8.62
7.62
5.84
6.90
5.37
5.10
6.19
4.83
3.83
4.14
3.24
5.01
4.06
3.49
4.77
5.37
gm.
9.13
8.91
9.17
8.68
8.46
10.01
9.42
8.58
8.14
6.35
5.71
4.94
5.11
4.78
4.41
2.83
3.15
3.32
4.06
3.82
3.45
gm.
gm.
gm.
13.49
13.55
12.59
13.12
12.39
10.70
10.10
10.89
8.90
10.83
9.47
gm.
16.45
10.51
14.38
13.72
13.72
(11.30)
10.77
9.67
9.52
(9.39)
8.38
8.49
8.77
(8.97)
7.78
gm.
8.41
6.50
7.78
7.86
7.82
7.13
6.20
5.40
4.38
5.17
5.38
8.11
5.96
5.10
4.07
gm.
13.99
8.76
8.38
10.73
9.40
7.87
7.73
6.11
7.70
7.35
6.80
6.14
6.97
5.62
4.08
gm.
19.50
12.24
12.45
13.02
11.63
10.87
10.74
10.13
17.0
11.2
10.55
10.8
11.19
11.01
8.79
9.74
10.05
7.12
6.32
6.84
5.14
4.66
5.05
4.23
5.4
3.6
5.7
3.3
2.82
2d
3d
4th
5th
6th
7th
8th
9th
10th
11th
12th
13th
14th
15th
16th
17th
18th
19th
20th
21st
22d
23d
5.84
6.41
6.27
6.18
6.30
4.44
4.19
8.42
24th
25th
26th
27th
28th
29th
30th
31st
32d
33d
34th
34th
35th
36th
37th
*8th
59th
K)th
JThe figures in this column are given for the first 10 days of the fast as corrected by Munk. The results for
t e remaining days have been increased in like proportion.
*Given by Ajello and Solaro as urea and here converted to nitrogen for purposes of comparison. Since the
a thors do not give the method employed, no attempt is here made to correct the figures.
3The results in this column were reported by the investigators as grains of urea, but are here converted to
g ams of nitrogen in urea for purposes of comparison.
250 A STUDY OF PROLONGED FASTING.
1.5 grams on the fourteenth day; and with S. A. B., 1.4 grams on the
fourth day. These sudden drops were almost invariably permanent
and were sometimes followed by a day on which even lower values were
found. It is difficult to predict at what point this break in the nitrogen
curve is likely to appear, and the irregularity of certain curves does not
justify giving serious attention at present to this feature of the general
course ; nevertheless the fact that it is characteristic of all long fasting
experiments is worthy of note.
The most accurate nitrogen determinations for the prolonged fasts
shown in table 26 are unquestionably those made by Brugsch for Succi's
fast at Hamburg. These values are somewhat lower than those found
for L., although Succi's body-weight was 18 kilograms greater than
that of our subject. In none of the fasting experiments do we find,
save perhaps in the Hamburg fast, any indication of an increase in the
nitrogen excretion near the end of the fast which may be considered
as corresponding to the so-called "pre-mortal" rise which has been
observed with many fasting animals, particularly with rabbits. It may
be said, therefore, that the values found for L. follow much the same
general course as the values found with the subjects of earlier fasting
experiments, except that the level of the nitrogen excretion after the
first 7 days was distinctly higher than with the other subjects.
Daily Excretion or Nitrogen.
Since L. had carefully preserved specimens of the urine from April 1
until the time of his arrival at the Nutrition Laboratory, we were able
to obtain information as to the nitrogen excretion of this subject for
13 days preceding the fasting experiment. By reference to the results
of these determinations (see table 23, page 238), it will be seen that in
general the nitrogen excretion was on a moderately high level, aver-
aging not far from 13 grams per day, and even exceeding this when the
low value of 8.83 grams is excluded.
On his first day in Boston (April 10-11), the nitrogen excretion was
17.02 grams. This was the highest value found and doubtless resulted
in part from the large beefsteak eaten by the subject on the night of
his arrival. The nitrogen excretion subsequently decreased until on
the last day before the fast it was but 11.54 grams. From April 10
until the beginning of the fast, therefore, the total nitrogen in the urine
averaged over 14 grams per day. This is significant as indicating that
L. was subsisting on a nitrogenous diet, which was quite inconsistent
with his claim that he was a "low-proteid vegetarian."
The values for the nitrogen excretion for the whole experiment,
including not only those for the fasting period, but for the food days
prior to and following the fast, are given in table 27.
As noted in the comparison with other fasting subjects, two striking
features of these values for the total nitrogen excretion are the immedi-
URINE.
251
ate decrease with the beginning of the fast and the continuance of
the high values until after the fourteenth day. This decrease in the
nitrogen excretion in the first few days of the fasting period has already
been explained as being due to the relatively large katabolism of glyco-
gen on those days. The average nitrogen excretion for the first 10
days of the fast was over 10 grams. The highest value (11.87 grams)
in the whole of the series was found on the fourth day, and the lowest
value (6.94 grams) on the thirty-first day. That the lowest value was
Table 27. — Nitrogen excreted in urine, per day and per kilogram of body-weight,
in experiment with L.
Date.
Day of
fast.
Nitrogen excreted.
Date.
Day of
fast.
Nitrogen excreted.
Per day.
Per
kilogram
of body-
weight
per day.
Per day.
Per
kilogram
of body-
weight
per day.
1912.
Apr. 11-12. .
gm.
15.92
14.48
11.64
7.10
8.40
11.34
11.87
10.41
10.18
9.79
10.27
10.74
10.05
10.25
10.13
10.35
10.43
8.46
9.58
gm.
0.264
.238
.190
.118
.142
.195
.207
.184
.181
.176
.186
.196
.185
.190
.189
.193
.196
.160
.182
1912.
Apr. 30-May 1..
May 1-2
2-3
3-4
4-5
5-6
6-7
7-8
8-9
9-10 ,
10-11
11-12
12-13
13-14
14-15 , ,
15-16
17th...
18th...
19th...
20th . . .
21st . . .
22d. . . .
23d
24th...
25th...
26th...
27th...
28th...
29th...
30th. ..
31st. . .
gm.
8.81
8.27
8.37
7.69
7.93
7.75
7.31
8.15
7.81
7.88
8.07
7.62
7.54
7.83
6.94
4.83
3.81
x2.75
gm.
0.169
.160
.163
.151
.156
.154
.146
.164
.158
.160
.165
.157
.156
.163
.146
.102
.081
1.058
12-13 . .
13-14. .
14-15..
15-16..
16-17..
17-18..
18-19 . .
19-20. .
20-21 . .
21-22 . .
22-23..
23-24..
24-25..
25-26 . .
26-27..
27-28..
28-29 . .
29-30..
1st. . .
2d. . . .
3d....
4th...
5th .. .
6th...
7th...
8th...
9th...
10th...
11th...
12th...
13th...
14th...
15th . . .
16th...
16-17
17-18
determined in urine for about 22 hours.
but 0.16 gram lower than that found on the first day may be explained
by the fact that on the first day L.'s energy requirement was in part
met by the combustion of about 70 grams of glycogen (see table
63, page 412). On the last day the katabolism was essentially a
protein-fat katabolism, unassisted by the combustion of any measurable
amount of carbohydrate. During the 31-day fast this subject actually
excreted 277.32 grams of nitrogen in the urine, thus averaging 8.95
grams of nitrogen per day. This would correspond to 1,664 grams of
protein, or 8,319.60 grams of flesh. Since the entire loss in body-weight
of this subject was 13.25 kilograms, it can be seen that 63 per cent of
the total loss may be accounted for in flesh katabolized.
252
A STUDY OF PROLONGED FASTING.
Although this is primarily a study of the excretion of nitrogen during
fasting, the values found for the 3 days subsequent to the fast have a
certain interest. During these days the subject took an almost
protein-free1 diet, consisting of fruit juices and honey. The large
amount of carbohydrate contained in this diet immediately protected
the protein in the body and in consequence there was a continually
decreasing nitrogen excretion, until on the last day we have the lowest
amount found with this subject, namely, 2.75 grams. This 22-hour
value is actually somewhat lower than that found with Beaute" by
Cathcart in a 3-day experiment with a starch-cream diet of Folin —
Beaut6, with a body-weight of not far from 58 kilograms, showing a
minimum nitrogen output of 2.84 grams. Since L. had a body-weight
at this time of only 47.5 kilograms, it would perhaps be expected that
his nitrogen excretion would be much lower than that of Cathcart's
subject; it should be noted, however, that his total nitrogen level was
considerably higher than that shown by Beaute\
The values found for our subject L. on these days of food following
the fast have a special interest, in that they show that the excess of
carbohydrate in the diet acted as a great protection of the body protein,
and hence we have here probably the nearest to the minimum protein
requirement of this man, corresponding to the "Abnutzungsquote"
of Rubner.
Nitrogen Excretion per Kilogram or Body-weight.
We have no information as to the fluctuations in the body-weight
prior to the arrival of the subject at the laboratory, but accurate obser-
vations were made from April 11 to the end of the experiment, and the
nitrogen per kilogram of body-weight may thus be computed for that
period. These values are also given in table 27. On the first day of the
fast the nitrogen output per kilogram of body- weight was very low, being
only 0. 1 18 gram. It then rose regularly until it reached a maximum on
the fourth day of 0.207 gram. Thereafter there was, in general, a steady
fall, with two minima of 0.146 on the twenty- third and the thirty-first
xTo aid in indicating the kinds and amounts of food eaten on the first two days of food fol-
lowing the fast, the estimated amounts and composition of food eaten are tabulated herewith:
Date.
Kind of food.
Amount
eaten.
Protein.
Fat.
Carbo-
hydrates.
Nitrogen.
1912.
May 15-16
May 16-17
Lemons
Oranges
Honey
Grape juice. . .
Total
Lemon juice . .
Honey
Orange juice . .
Total
gm.
100
450
311
1072
gm.
1.00
3.60
1.24
gm.
0.70
.90
gm.
8.5
52.2
252.5
178.6
gm.
0.16
.59
.19
5.84
1.60
491.8
0.94
80
139
1128
6.55
6.05
7.8
112.8
128.1
6.08
1.01
6.60
248.7
1.09
URINE.
253
days respectively. After food was again taken, the nitrogen excretion
decreased to the surprisingly low value of 0.058 gram per kilogram of
body-weight.
Comparison of Methods fob Determining Total Nitrogen and Ammonia-nitrogen.
The microchemical methods had been developed just previous to
this fasting experiment and were therefore used by Mr. H. L. Higgins
for determining the total nitrogen and the ammonia-nitrogen. The
total nitrogen was also determined by the Kjeldahl method and the
ammonia-nitrogen by the old Folin method, both determinations being
made by Miss E. B. Babcock. These analyses of the fasting urines
were therefore the first control analyses which had been made outside of
the Folin laboratory. Both the Kjeldahl method and the Folin micro-
chemical method were frequently tested by determining the nitrogen of
Table 28. — Comparison of the determinations of nitrogen and ammonia-nitrogen by former
methods and the new microchemical methods of Folin.
Date.
Day of
fast.
Total nitrogen.
Ammonia-N.
Kjeldahl
method.
Folin
Folin
Folin
micro
method
micro
method.
(original).
method.
1912.
gm.
gm.
gm.
gm.
Apr. 16-16
2d
0.60
0.60
16-17
3d
11.34
10.26
.95
.95
17-18
4th....
11.87
11.46
1.40
1.40
18-19
5th....
10.41
9.94
1.60
1.63
19-20
6th....
10.18
9.91
1.67
1.69
20-21
7th....
9.79
9.30
1.52
1.56
21-22
8th....
10.27
9.91
1.62
1.68
22-23
9th
10.74
10.74
1.70
1.68
23-24
10th....
10.05
10.02
1.57
1.60
24-25
11th....
10.25
10.45
1.56
1.60
25-26
12th....
10.13
10.11
1.47
1.51
26-27
13th
J9.91
10.00
*1.45
11.62
27-28
14th
10.43
10.25
1.57
1.61
28-29
15th....
8.46
8.58
1.43
1.46
29-30
16th
9.58
9.47
1.91
1.97
Apr. 30-May 1
17th
8.81
8.77
1.90
1.93
May 1-2
18th
19th....
21st
24th
25th
26th
8.27
8.37
7.93
8.15
7.81
7.88
8.45
8.11
7.90
8.24
7.91
8.00
1.80
1.76
1.52
1.51
1.42
1.80
1.81
1.55
1.52
1.43
2-3
4-5
7-8
8-9
9-10
10-11
27th....
8.07
7.99
1.36
1.39
11-12
28th....
7.62
7.46
1.28
1.29
12-13
29th....
7.54
7.64
1.32
1.32
13-14
30th
7.83
7.60
1.32
1.32
14-15
31st
6.94
6.83
1.25
1.22
16-16
3.72
3.81
3.64
3.98
.52
.36
.53
.36
16-17
1For total amounts of nitrogen and ammonia-N on this day, see table 29.
254 A STUDY OF PROLONGED FASTING.
known substances, such as ammonium-sulphate, ammonium ferrous
sulphate, urea, and uric acid. We were thus assured of the accuracy
of the methods. Since the Folin microchemical methods played such
an important role in these analyses, particularly in an economical
distribution of the available urine, it seems desirable to publish the
results of the control tests. Accordingly, in table 28, the values
obtained for the total nitrogen by the Kjeldahl method are compared
with those secured by the microchemical methods; the values for the
ammonia-nitrogen obtained with the new and old Folin methods are
also compared. As will be seen, the results of such comparison are
most satisfactory. We wish again to emphasize the great value of
these methods, particularly when there is urgent necessity for the use of
small samples.
THE PARTITION OF THE NITROGEN EXCRETION.
While the total nitrogen excretion in the urine of a fasting man has
a general interest, more particularly in the apportionment of the total
energy requirement and the energy output among the various factors,
protein, carbohydrate, and fat, a clear understanding of the nature of
the disintegration of the nitrogenous material is obtained only when a
partition of the nitrogen excretion is made according to the analytical
scheme of Folin. Fortunately, with all of the samples of urine col-
lected for the 31 days of the fast we were able to secure a complete
partition of the nitrogen, with the single exception of the determination
of the total purines; we were, however, able to determine the uric
acid-nitrogen. This partition included the determination of the total
nitrogen, and the nitrogen from urea, ammonia, uric acid, creatinine
preformed, and total creatinine. The nitrogen undetermined is given
as "rest nitrogen." Furthermore, since Folin has shown the great
significance of the proportionate distribution of the nitrogen derived
from these various sources, we have computed the percentage of the
total nitrogen in these nitrogenous constituents of the urine. The
values for each day, expressed in grams and in percentages of total
nitrogen, are given in table 29.
Urea.
With the microchemical method of Folin, the urea-nitrogen in the
fasting urines could be determined with great accuracy. The deter-
minations were accordingly made by Mr. Higgins for the 31 days of the
fasting period and for the 3 days following the fast. The results are
given in table 29, together with the percentage of urea-nitrogen in
terms of total nitrogen.
In general the course of the excretion of the urea-nitrogen follows
quite closely that of the total nitrogen. The regular increase shown
in the first 3 days for the total nitrogen is also apparent here; the
URINE.
;hv
255
smallest amount of urea-nitrogen (4.84 grams) is likewise found on the
last day of the fast, but unlike the total nitrogen is considerably smaller
than the excretion for the first day of the fast. The percentage of urea-
nitrogen shown for the first few days, i. e., approximately 80 per cent,
is distinctly lower than that found with normal urine, Folin1 giving as
Table 29. — Partition of nitrogen excreted in urine in experiment with L.
Date.
Day of
fast.
Excretion of nitrogen.
Total
nitrogen.
Urea-N.
Ammo-
nia-N.
Uric
acid-N.
Creati-
nine-N
(pre-
formed).
Total
creati-
nine-N.
Rest-N.
1912.
gm.
gm.
gm.
gm.
gm.
gm.
gm.
Apr. 11-12. . .
15.92
14.48
11.54
7.10
5.68
0.67
.65
.59
.41
0.112
6!51
0.48
0.42
12-13 . . .
13-14 . . .
14-15 . . .
1st
15-16...
2d
8.40
6.69
.60
.049
.46
.46
.60
16-17...
3d
11.34
9.11
.95
.042
.46
.55
.69
17-18...
4th....
11.87
9.03
1.40
.044
.42
.54
.86
18-19...
6th....
10.41
7.58
1.62
.059
.41
.51
.64
19-20. ..
6th....
10.18
7.36
1.68
.097
.39
.52
.52
20-21 . . .
7th....
9.79
7.02
1.54
.112
.38
.49
.63
21-22 . . .
8th....
10.27
7.45
1.65
.108
.38
.50
.56
22-23 . . .
9th....
10.74
7.83
1.69
.099
.37
.50
.62
23-24...
10th....
10.05
7.44
1.59
.118
.37
.49
.41
24-25. ..
11th....
10.25
7.66
1.58
.116
.37
.49
.40
25-26.. .
12th
10.13
7.43
1.49
.154
.37
.49
.57
26-27...
13th
10.35
7.69
1.55
.093
.35
.48
.54
27-28...
14th....
10.43
7.69
1.59
.125
.33
.44
.59
28-29 . . .
15th....
8.46
6.18
1.45
.071
.30
.38
.38
29-30...
16th
9.58
6.71
1.94
.099
.32
.42
.41
Apr. 30-May 1
17th....
8.81
5.95
1.92
.100
.31
.40
.44
May 1- 2 . . .
18th
8.27
5.70
1.80
.122
.34
.41
.24
2- 3 . . .
19th....
8.37
5.58
1.79
.130
.30
.38
.49
3-4...
20th....
7.69
5.36
1.58
.115
.31
.38
.26
4-6...
21st
7.93
5.54
1.57
.112
.31
.38
.33
5- 6. ..
22d. , .
7.75
5.60
1.51
.110
.31
.36
.17
6-7...
23d,
7.31
5.01
1.49
.097
.34
.36
.35
7-8...
24th....
8.15
5.92
1.54
.114
.30
.34
.24
8-9...
25th....
7.81
5.43
1.52
.098
.28
.35
.41
9-10...
26th
7.88
5.62
1.43
.063
.29
.36
.41
10-11. . .
27th
8.07
5.90
1.38
.089
.29
.35
.35
11-12. . .
28th
7.62
5.46
1.29
.095
.28
.34
.44
12-13.. .
29th
7.54
5.55
1.32
.101
.29
.35
.22
13-14...
30th....
7.83
5.53
1.32
.106
.29
.33
.54
14-15...
31st
6.94
4.84
1.24
.122
.30
.32
.42
15-16. . .
4.83
3.81
2.75
3.21
2.69
1.54
.69
.36
.35
.140
.144
.111
.35
.34
.35
.37
.34
.33
.42
.28
.42
16-17. . .
117-18 . . .
xThe amounts for this day were determined in the urine for about 22 hours.
an average for his subjects 87.5 per cent of the total nitrogen in the form
of urea. The percentage of the urea-nitrogen then falls rapidly until
on the fifth day it is but 72.82 per cent. The lowest percentage found
^olin, Am. Journ. Physiol., 1905, 13, p. 62.
256 A STUDY OF PROLONGED FASTING.
Table 29. — Partition of nitrogen excreted in urine in experiment tvith L. — Continued.
Date.
Day of
Proportion of total nitrogen in —
fast.
Urea.
Ammo-
nia.
Uric
acid.
Creatinine
(pre-
formed).
Total
creatinine.
Rest-N.
1912.
p. ct.
p. ct.
p. ct.
p. ct.
p. ct.
p. ct.
Apr. 11-12
4.21
12-13
80.00
4.49
6.11
5.77
1.58
7.18
6.76
5.89
13-14
14-16
1st
16-16
2d
79.64
7.14
.58
5.48
5.48
7.16
16-17
3d
80.33
8.38
.37
4.06
4.85
6.07
17-18
4th....
76.07
11.79
.37
3.54
4.55
7.22
18-19
6th ... .
72.82
15.66
.67
3.94
4.90
6.15
19-20
6th
72.30
16.50
.95
3.83
5.11
5.14
20-21
7th....
71.71
15.73
1.14
3.88
5.01
6.41
21-22
8th....
72.54
16.07
1.05
3.70
4.87
5.47
22-23
9th
72.90
15.73
.92
3.44
4.66
5.79
23-24
10th
74.03
15.82
1.17
3.68
4.88
4.10
24-25
11th....
74.73
15.41
1.13
3.61
4.78
3.95
25-26
12th
73.35
14.71
1.52
3.65
4.84
5.58
26-27
13th....
74.30
14.98
.90
3.38
4.64
5.18
27-28
14th
73.73
15.24
1.20
3.16
4.22
5.61
28-29
15th
73.05
17.14
.84
3.55
4.49
4.48
29-30
16th
70.04
20.25
1.03
3.34
4.38
4.30
Apr. 30-May 1
17th....
67.54
21.79
1.14
3.52
4.54
4.99
May 1-2
18th....
68.92
21.77
1.48
4.11
4.96
2.87
2-3
19th
66.67
21.39
1.55
3.58
4.54
5.85
3-4
20th
69.70
20.55
1.50
4.03
4.94
3.31
4- 5
21st
69.86
19.80
1.41
3.91
4.79
4.14
5-6
22d
72.26
19.49
1.42
4.00
4.65
2.18
6-7
23d
68.54
20.38
1.33
4.65
4.92
4.83
7-8
24th....
72.64
18.90
1.40
3.68
4.17
2.89
8- 9
25th
69.52
19.46
1.25
3.69
4.48
5.29
9-10
26th
71.32
18.15
.80
3.68
4.57
5.16
10-11
27th
73.12
17.10
1.10
3.59
4.34
4.34
11-12
28th
71.66
16.93
1.25
3.67
4.46
5.70
12-13
29th
73.61
17.51
1.34
3.85
4.64
2.90
13-14
30th
70.63
16.86
1.35
3.70
4.21
6.95
14-15
31st
69.74
17.87
1.76
4.32
4.61
6.02
15-16
66.46
70.60
56.00
14.28
9.45
12.73
2.90
3.78
4.04
7.25
8.92
12.73
7.66
8.92
12.00
8.70
7.25
15.23
16-17
U7-18
*The amounts for this day were determined in the urine for about 22 hours.
during the last 26 days was 66.67 per cent on the nineteenth day and
the highest on the eleventh day of 74.73 per cent. The average value
for these days was 71.5 per cent, with a distinct tendency towards
constancy. Upon the resumption of food, there was at first no marked
disturbance in this ratio, but on the third day the percentage of urea-
nitrogen decreased to the low value of 56 per cent, this value being
found at the time that only 2.75 grams of nitrogen were excreted in
the urine. These low values in the percentage of urea-nitrogen are
perfectly comparable with those found by Folin on subjects subsisting
URINE. 257
on a starch-cream diet, with an excretion of nitrogen corresponding
to not far from 4 to 5 grams per day.
These figures are also substantiated by the observations of Cathcart.
While he found in the first 3 days of Beauty's fast that the urea-nitro-
gen averaged not far from 87 per cent instead of the 80 per cent found
with our subject L., and that the values also averaged somewhat higher
for the remainder of the fast, nevertheless the percentage fell as low as
71 per cent on the eighth day of the fast. On the food days following
the fast, the urea-nitrogen fell to 61.97 per cent on the day when the
minimum nitrogen excretion was observed.
E. and 0. Freund found in their observations on Succi that the urea-
nitrogen was 82 per cent or more of the total nitrogen excretion for
the first two weeks of the fast. There was then a rapid fall in the
percentage until but 56 and 58 per cent of urea-nitrogen were found
on the last 2 days. Brugsch's observations on Succi in Hamburg show
that for the last 8 days of the 30-day fast, the urea-nitrogen was not
far from 60 per cent of the total nitrogen.
Van Hoogenhuyze and Verploegh, in their observations on Tosca,
note most irregular proportions of urea-nitrogen. The percentages
of urea-nitrogen computed by us from their data are as follows, the
dayof the fast being given in parentheses: (1)68.84; (2)79.11; (3)93.29;
(4) 84.90; (5) 66.07; (6) 59.00?; (7)81.02; (8)84.42; (9)88.43; (10)87.65;
(11) 85.84; (12) 86.80; (13) 86.84; (14) 88.73. On the sixth day the
low value of 59 per cent is from data questioned by the authors. In
the light of other fasting studies, there is no obvious explanation
for the unusually high average value, especially for the last 8 days of
the fast.
From all the evidence it can be seen that in general during fasting
the urea output approximately parallels that of the total nitrogen,
there being a decided increase on the first few days of fasting, followed
by a decrease. In practically every instance when there is a fluctua-
tion in the total nitrogen output, this is paralleled by the urea-nitrogen.
It would thus appear that the determining factor in the fluctuations
of the total nitrogen is probably the proportion of urea-nitrogen, and
not the gross alterations in the other factors. The variation in the
percentage distribution can, however, be intelligently treated only after
a consideration of the changes in the output of ammonia-nitrogen.
Ammonia.
The ammonia-nitrogen, on account of its great significance in con-
sidering the products of defective fat katabolism, was determined by
both the old and the new Folin methods. The results of these two
series of determinations are given in table 28, page 253. An average of
these two series of values is also given in table 29.
Normal urine always contains a relatively small amount of ammonia,
and the amounts found for L. on the 3 days prior to the fast were
258 A STUDY OF PROLONGED FASTING.
approximately those which would be noted for normal individuals
subsisting on a diet containing about 15 grams of nitrogen. The
ammonia-nitrogen formed not far from 4.5 per cent of the total nitrogen
excretion per day. At the beginning of the fasting period the amount
of ammonia-nitrogen excreted fell somewhat, and not until the third
day do we find values exceeding those obtained before the fast. On
the fourth day it rose quite sharply to 1.40 grams and then continued
to rise steadily, with slight fluctuations, until the maximum value of
1.94 grams was reached on the sixteenth day. Thereafter it slowly
and quite regularly decreased until the end of the experiment, the
excretion of ammonia-nitrogen on the last day being 1.24 grams. This
gradual increase and decrease was exactly that observed by Cathcart,
although the maximum value with his subject was observed on the
eighth day, while with L. it did not appear until the sixteenth day.
Brugsch found the excretion of ammonia-nitrogen quite regular, rang-
ing between 1.26 grams and 1.72 grams in the last 8 days of Succi's
fast in Hamburg.
Since the amount of ammonia-nitrogen would normally be expected
to fluctuate somewhat with the fluctuations in the total nitrogen
excretion, the percentage of ammonia-nitrogen in the total nitrogen
must be considered. On this basis the minimum percentage (5.77 per
cent) is found on the first day of the fast, with the maximum (21.79
per cent) on the seventeenth day. Even on the percentage basis the
ammonia-nitrogen tends to increase until the middle of the fast and
then slowly to decrease, although towards the end it was still 10 per
cent or more larger than it was prior to the fasting period.
It is clear, therefore, that there was some factor which stimulated
the excretion of ammonia-nitrogen. From previous experience with
fasting subjects, it is obvious that this increase in the excretion was
due to the organic acids, chiefly the /3-oxybutyric acid resulting from
defective fat katabolism. The large excretion of ammonia therefore
undoubtedly corresponds to an increasing acidosis, the production of
the ammonia being a protective action on the part of the body to over-
come the effect of the acids. The amounts of ammonia-nitrogen found
for L. during fasting were measurably greater than those observed by
Cathcart for his subject and the percentage of the total nitrogen was
also considerably greater. These values would therefore imply that L.
had a somewhat greater acidosis than had Beaute" in Cathcart's research.
The results of the observations of E. and O. Freund on Succi are
diametrically opposed to the values found by Cathcart on Beauts,
Brugsch on Succi, and by us with L., for both the absolute amount and
the percentage of the total nitrogen found, and one is inclined to ques-
tion somewhat the technique of the Freunds. While the observations of
Bonniger and Mohr1 on the fasting woman Schenk are complicated by the
1Bonniger and Mohr, Zeitschr. f. exp. Path. u. Therapie, 1906, 3, p. 675.
URINE. 259
introduction of amino-acids, their results shoul d also be cited. The abso-
lute amounts of ammonia-nitrogen excreted varied from 0.40 gram on
the second fasting day to 1 .84 gram on the seventh day of fasting. The
percentage of ammonia-nitrogen in the total nitrogen varied from 5.28
per cent on the second day to the high value of 30.57 per cent on the
sixteenth day of the fast. In general, these values are not unlike those
observed by us on our subject L.
If we compare the values obtained with L. for the urea-nitrogen and
the ammonia-nitrogen, it is evident that the ammonia-nitrogen was
formed at the expense of urea-nitrogen, for whenever the values for
the urea-nitrogen decrease, those for the ammonia-nitrogen increase,
and vice versa. A fact of particular interest is the rapid return to a
small excretion of ammonia-nitrogen with the taking of food by the
subject. Even on the first day with food, the ammonia-nitrogen fell
to 0.69 gram. On subsequent days the absolute values decreased to
0.36 gram and 0.35 gram respectively, though still forming 10 to 12 per
cent of the total nitrogen. It is obvious, therefore, that, with the inges-
tion of a large amount of carbohydrate and a decrease in the acidosis,
there was no necessity for an excessive ammonia-nitrogen excretion.
Uric Acid.
In the earlier fasting experiments at Wesleyan University, material
was available for only a few determinations of the uric acid, which were
made possible through the courtesy of Professor Lafayette B. Mendel,
of Yale University. In this experiment with L. the determinations
of uric acid were personally made by Professor Otto Folin, by his new
colorimetric method,1 small specimens of the 24-hour urine being sent
to the Harvard Medical School daily for analysis. The results of his
determinations, expressed as uric-acid nitrogen, are given in table 29.
No determinations of the uric-acid nitrogen were made prior to the
fasting period. During the fast the minimum amount of 0.042 gram
Was obtained on the third day and the maximum of 0.154 gram on the
twelfth day, the amounts varying considerably throughout the entire
fasting period.
With the origin of uric acid still the subject of considerable critical
debate, particularly between Siven2 on the one hand and Mares3 and
his colleagues on the other, it can be seen that although the general
tendency is for physiologists to uphold the views of Mares,4 it is difficult
to interpret the values for the uric-acid excretion obtained in this fast
on any of the present hypotheses as to its origin. That it is a deriva-
^olin and Denis, Journ. Biol. Chem., 1913, 14, p. 95.
2Siven, Archiv f. d. ges. Physiol., 1912, 146, p. 499.
3Mare§, Archiv f. d. ges. Physiol., 1912-1913, 149, p. 275. See also Smetdnka, Archiv f. d.
ges. Physiol., 1911, 138, p. 217; ibid., 1912-1913, 149, p. 287.
4Mares, Archiv f. d. ges. Physiol., 1910, 134, pp. 59-102
260 A STUDY OF PROLONGED FASTING.
tion of nuclein katabolism is undoubtedly true. Beyond this univer-
sally accepted fact, the evidence as to the influence of intestinal,
glandular, kidney, and muscular activity is much debated.
In previous fasting experiments it has been shown that the glandular
activity, at least so far as the digestive organs are concerned, is at a
minimum,1 and hence we should expect to find that if any considerable
proportion of the uric acid were obtained as the result of glandular
activity, there would be a minimum uric-acid nitrogen excretion during
the fast. Aside from the admittedly low values on the second to the
fifth days of the fasting experiment with L., this was not the case, for
there are many days during the fast on which there was fully as much
uric-acid nitrogen, if not indeed more, than one would expect to find
with a normal individual subsisting on a purine-free diet.2
After L. had taken food, there was a slight increase in the total
uric-acid nitrogen excretion from 0.122 gram, which was obtained on
the last day of the fast, to 0.140 gram on the first day with food. Ordi-
narily this would not be considered as a substantial increase in the
total uric-acid nitrogen excretion, and yet in recent controversial
papers, Mares has pointed out that a 10 per cent increase is not to be
ignored and that it does represent an actual increase. While accord-
ing to Mares, the ingestion of food which stimulates the digestive
glands to their greatest activity is most productive of an increase in
the excretion of uric acid, nevertheless, Smetanka has found that such
an excretion took place even with honey. Inasmuch as the food taken
by L. was in large part carbohydrate, a portion of the diet being
honey, it is not impossible to believe that there may have been
a positive increase in the uric-acid nitrogen as a result of the ingestion
of this food. This is the more credible if we give heed to Smetanka's
criticism that 24-hour periods are not best suited to the study of this
problem, as the influence of the activity of the digestive glands dis-
appears rapidly after the ingestion of food.
During the fasting period, the percentage of the total nitrogen in
the form of uric-acid nitrogen likewise underwent considerable fluctua-
tion, approximately paralleling the absolute amount of this constituent.
The percentage value shows a distinctly increased uric-acid nitrogen
excretion in the post-fasting period, rising to 4.04 per cent, and while
the evidence is by no means complete, this may well be interpreted
as an index of an increased glandular (digestive) activity.
Cathcart also determined the uric-acid nitrogen excretion for his
subject, using the method of Hopkins-Folin. During the first half
of the fast he found a distinctly lower value for the uric-acid nitrogen
than in the preceding food period, but there was a continual tendency
for this form of nitrogen to increase as the fast progressed. With the
taking of food there was an increase in the uric-acid nitrogen — at
'Luciani, Das Hungem, Hamburg and Leipsic, 1890, p. 44.
2Folin, Am. Journ. Physiol., 1905, 13, p. 62.
URINE. 261
least on the first day — followed by 4 days with the same excretion as
in the last period of the fast. Thus, in Beauty's fast there was a
definite tendency towards regularity1 in the excretion of the uric-acid
nitrogen, inasmuch as there was from the beginning to the end a
regular, though slight, increase — a regularity that was by no means
paralleled by the observations on L.
While the values found for L. were decidedly variable, yet in several
points they strikingly confirm the earlier observations. Thus the
marked fall on the second fasting day, followed by a low excretion for
several days and a subsequent rise, with a higher value on the first
food day than that shown on the last fasting day, are all in agreement
with the observations of Cathcart with Beaute, and of Van Hoogen-
huyze and Verploegh in their 14-day experiment with Tosca.
The rapid fall in the uric-acid nitrogen excretion on the second day
has also been observed in shorter fasts by Schreiber and Waldvogel,2
by Hirschstein,3 and by Feldmann.4 Scaffidi,5 in experimenting on
the purine metabolism in fasting, concluded that with those animals
with which there was a formation of oxidative uric acid, the uric-acid
nitrogen decreased during starvation and there was no regularity in
the relation of total uric-acid nitrogen to the total nitrogen.
This decrease in the uric-acid excretion of L. was coincident with
that period of the fast when the supply of glycogen was rapidly being
depleted and the subsequent increase followed sharply the incidence
of the protein-fat katabolism characteristic of the remainder of the
fast. The cessation of the glandular activity of digestion on the first
few days may explain the fall in the uric-acid nitrogen, but the sub-
sequent increase can only be explained by an increase in the katabolism
of the active protoplasmic tissue. The fact that this excretion does not
remain constant, or at least does not regularly increase or decrease,
is indeed difficult to explain, since there are no obvious reasons for
assuming that the changes in the amount of the excretion of uric-acid
nitrogen are due to corresponding changes in the rate of destruction
of the active protoplasmic tissue. Indeed, if we observe the total
creatinine excretion, the regular decrease of that urinary constituent
would imply a proportional regularity in the rate of destruction of the
active protoplasmic tissue. It is evident that the values obtained for
the excretion of the uric-acid nitrogen in this fasting experiment offer
no proof of the validity of any of the present-day conceptions as to
the origin of endogenous uric-acid excretion.
Regularity in the uric-acid excretion as the fast progressed was also noted for the greater part
of the experiment with Tosca by Van Hoogenhuyze and Verploegh (loc. cit.) and with Succi in
Vienna by E. and O. Freund (Wiener klin. Rundsch., 1901, 15, p. 69). Brugsch (Zeitschr, f.
exp^Path. u. Therapie, 1905, 1, p. 419) found very constant values for the purine-nitrogen during
the last eight days of Succi's Hamburg fast.
2Schreiber and Waldvogel, Arch. f. exp. Path. u. Pharm., 1899, 42, p. 69.
'Hirschstein, Arch. f. exp. Path. u. Pharm., 1907, 57, p. 229.
4Feldmann, cited by Siven, Archiv. f. d. ges. Physiol., 1912, 146, p. 499.
'Scaffidi, Biochem. Zeitschr., 1911, 33, p. 153.
262 A STUDY OF PROLONGED FASTING.
Creatinine.
As a result of Folin's fundamental observations on the partition of
the nitrogen of normal urines, special stress was laid upon the deter-
mination of the creatinine existing in the urine of this fasting man,
Folin having emphasized the fact that the total creatinine may be
looked upon as an index of the total tissue metabolism.1 Immediately
after Folin's papers had appeared, a large number of researches on the
creatinine in urine were reported, and observations were simultane-
ously made in several laboratories which implied the presence of
creatine in urine under certain conditions, particularly in pathological
cases and during fasting.
Folin's analytical scheme enabled the direct determination of crea-
tine, the creatinine being first determined, and subsequently any creatine
in the urine was converted to creatinine by heating with acid for 3
hours. These results were reported by him as creatinine preformed
and total creatinine, the difference in the two values being estimated
to be a measure of the creatine expressed in terms of creatinine.
There is little of positive value to be said regarding the determinations
of creatine or creatinine in fasting experiments prior to the introduction
of the Folin method, hence Baldi's2 observations as well as those of the
Freunds3 on Succi can have but little quantitative interest. One of the
first studies made after the Folin method was put forth is that of Van
Hoogenhuyze and Verploegh on Tosca.4 The considerable decrease in
the creatinine obtained by them at the beginning of the fast, followed
by an increase after food, shows (in part at least) that only preformed
creatinine was determined, and hence their values correspond more
nearly to the values reported for our subject as preformed creatinine.
In the experiments at Wesleyan University, the total creatinine
was invariably found to be somewhat higher in the fasting periods
than the preformed creatinine, the difference disappearing when food
was taken, and thus the conclusion was drawn that creatine was to be
found regularly in the urine of fasting subjects. This observation was
simultaneously made and published by Cathcart.5 Laying special
emphasis upon the difference between preformed and total creatinine,
Cathcart discussed at considerable length the appearance of creatine
in the urine. Although our interpretation of the presence of creatine
in the urine differs considerably from that of Cathcart, the fact remains
that since that time considerable research has been carried out on both
creatinine and creatine excretion, and hence an observation of the
amount excreted by our fasting subject was particularly desirable.
^olin, Am. Journ. Physiol., 1905, 13, pp. 66 and 117.
2Baldi, Sperimentale, March 1889; Centrlb. f. klin. Med., 1889, 10, p. 651.
3E. and O. Freund, Wiener klin. Rundsch., 190 1, 15, pp. 69 and 91.
4Van Hoogenhuyze and Verploegh, Zeitschr. f. physiol. Chemie, 1905, 46, p. 415.
'Cathcart, Biochem. Zeitschr., 1907, 6, p. 109.
URINE. 263
The only other fasts on human subjects in which the Folin methods
have been used are the two 7-day fasts reported by Howe, Mattill,
and Hawk.1 Here, as in our experiments, preformed creatinine and
total creatinine were determined without prior treatment to remove
aceto-acetic acid. The large amount of creatine-nitrogen found by
them on the first fasting day with their subject E., namely, 0.269 gram,
is wholly inexplicable, being larger than any observation with which we
are familiar. After the first 3 days there was a regular decrease in
the creatine-nitrogen excretion, until on the seventh day it was prac-
tically nothing. With subject H. they found an increasing creatine-
nitrogen excretion for the first 4 days, followed by a decrease. It
is thus seen that, as the result of determinations by the Folin method,
several observers have noted that the total creatinine is greater than
the preformed creatinine.
In interpreting the increase in the creatinine excretion in the fasting
experiments at Wesleyan University as being due to the presence of
creatine, attention was especially directed2 to the possible influence
upon the Jaffe" reaction of substances other than creatine and creati-
nine. Immediately after this publication had appeared, and in sub-
sequent visits to foreign laboratories where work upon creatine and
creatinine was being carried out, I was assured that no substance
existing in either fasting or diabetic urine appreciably affected the
reaction and that this interpretation of the difference between pre-
formed creatinine and total creatinine was undoubtedly justifiable.
Accordingly, in presenting the results obtained with L., the assumption
has been made that a difference between preformed creatinine and
total creatinine, as ordinarily determined, could be taken as an index
of the presence of creatine.
The creatinine determinations in these fasting urines were made by
Miss Alice Johnson, under the immediate supervision of Dr. A. W.
Peters, numerous control tests being made with samples of pure creati-
nine kindly furnished by Professor Otto Folin. The preformed creati-
nine was determined according to the usual Folin method and in no
instance was abnormal fading or alteration in color noted. Further-
more, in the light of recent investigations, it is important to note that
no special treatment was given for the removal of acetone bodies which
would possibly have affected the color. After the determination of the
preformed creatinine, the total creatinine was determined by heating
another specimen of the urine with hydrochloric acid for 3 hours on
an electric plate to convert the creatine to creatinine. The readings
were carefully controlled and the illumination of the colorimeter was
given special attention. The results as found by these two processes
are expressed as nitrogen of creatinine preformed and total creatinine
'Howe, Mattill, and Hawk, Journ. Am. Chem. Soc, 1911, 33, p. 568.
2Benedict, Carnegie Inst. Wash. Pub. 77, 1907, p. 395.
264 A STUDY OF PROLONGED FASTING.
and are given in table 29. As with the earlier fasting experiments,
there was a difference between the creatinine preformed and the total
creatinine as thus measured.
At the time these analyses were made, we were firmly of the opinion
that the results obtained by such analysis would represent quanti-
tatively the amount of creatine in urine, expressed as creatinine.
Since these investigations were carried out, two papers have appeared
which lead us to question the quantitative relations exhibited in the
values here given. Greenwald,1 in his work on diabetic urines, con-
siders the effect of the acetone bodies upon the Jaffe reaction and con-
cludes that urine containing aceto-acetic acid and acetone will give
correct results for creatinine by the Folin method only after the
removal of these substances. More recently Graham and Poulton2
have studied the influence of aceto-acetic acid on the estimation of
creatinine and have called into question the interpretation of the dif-
ference between preformed and total creatinine as reported in fasting
experiments and in experiments with carbohydrate starvation. They
conclude that " acetone and /3-oxybutyric acid, if present in amounts
comparable to those which usually occur in urine, produce practically no
error in the estimation of creatinine." They furthermore state that
aceto-acetic acid causes an error in the preformed creatinine determina-
tion, but does not affect the determination of total creatinine. On
the basis of their experiments, they conclude that the difference
between creatinine preformed and total creatinine does not represent
creatine, since in their experiments with carbohydrate-free diets, they
found no excretion of creatine.
It is obvious, therefore, that these two researches throw considerable
doubt upon the value of the figures reported under the head of the
preformed creatinine, but affect in no wise the values for the total
creatinine. We may therefore without further reservation proceed to
a consideration of the total creatinine-nitrogen as reported in table 29.
One of the most striking features of these results is their great regu-
larity. After the first 2 days there was an almost continuous fall in
the total creatinine-nitrogen from the maximum on the third day of
0.55 to a minimum on the last day of the fast of 0.32. The minor
fluctuations from this regular fall are so few as to be negligible.
In the fasting experiments with S. A. B. at Wesleyan University,
the almost absolute constancy in the total creatinine-nitrogen was a
matter of special comment and in the report of the results it was
pointed out that it was probably more than a mere coincidence that
the sum of the creatine-nitrogen and the preformed creatinine-nitrogen
remained constant each day as the fast progressed. It will be seen
that for the first 7 days of the fast with L., while the results were not
Greenwald, Journ. Biol. Chem., 1913, 14, p. 87.
Graham and Poulton, Proc. Royal Soc, ser. B., 1914, 87, p. 205.
URINE. 265
constant, nevertheless the amount excreted did not vary greatly,
although the variation was wholly in the line of a decrease toward the
end. It was only after the thirteenth day that any considerable de-
crease in the total creatinine-nitrogen appeared; from that time it
remained at a lower level, gradually decreasing until the end of the
fast, the lowest value being found on the last day of the fast.
Of particular interest is the striking regularity in the percentage of the
total nitrogen excreted in the form of total creatinine. Omitting the
first 2 days, the percentage ranged from 5.11 per cent on the sixth day
to 4.17 per cent on the twenty-fourth day.
On the first 2 days of the fast, the values for preformed creatinine-
nitrogen and for total creatinine-nitrogen are essentially the same,
and it is only with the third day that we begin to find the measurable
differences which have been ascribed to creatine. In an attempt to
explain the presence of creatine in the fasting urine, two hypotheses
have been presented: one, that as a result of inanition the body loses
its power of converting creatine to creatinine before excreting it, and
the other that creatine was representative of the flesh katabolized, as
in the disappearance of the body material a certain amount of creatine
normally existing in the flesh was liberated and excreted in the urine.
Against the validity of this latter assumption was the obviously sig-
nificant fact that the sum of the creatinine-nitrogen and the creatine-
nitrogen remained constant. While without doubt the observations
of Greenwald and of Graham and Poulton affect the quantitative value
of the difference between preformed creatinine and total creatinine, the
accumulative evidence of the past eight years is such as to make it
reasonably certain that creatine is excreted unchanged in fasting and
pathological urines and our uncertainty lies, therefore, only in our
knowledge as to the quantities thus excreted.
In a recent paper Folin1 has reiterated his belief regarding the
interpretation of the creatinine output in the following terms: "The
creatinine elimination becomes more clearly than ever the most clear-
cut index or measure of the total normal tissue metabolism." In this
paper Folin explains the appearance of creatine as an abnormal split-
ting off of a cleavage product which is normally excreted as creatinine
and disclaims the presence of isolated, uncombined creatine in flesh,
admitting that in fasting and in various pathological conditions the
normal breaking down into creatinine is accompanied more or less by
an abnormal breaking down into creatine. It is clear, therefore, that
according to all of Folin's recent interpretations the exact quantitative
knowledge regarding the creatine in urine has no longer the significance
formerly attributed to it, and hence the criticisms of Greenwald and of
Graham and Poulton, while admittedly justifiable so far as technique
is concerned, apply to a determination which bids fair to have but
^olin and Denis, Journ. Biol. Chem., 1914, 17, p. 501.
266 A STUDY OF PROLONGED FASTING.
little physiological significance. The pathological significance must,
however, be measurably increased in value as a result of their researches.
Of fundamental importance is the possibility of using Folin's inter-
pretation of the total creatinine excretion as an index of the metabolism
of normal tissue. Throughout this entire monograph stress has been
laid upon the importance of knowing, if possible, the active mass of
protoplasmic tissue in the body,1 and the question arises, "Have we in
the total creatinine excretion an interpretable index of the changes
in the active mass of protoplasmic tissue?" If, as is pointed out in
the section on energy transformation, the active mass of protoplasmic
tissue is the fundamental factor in the determination of the total
energy requirement, we should normally expect to find that the total
energy output decreased with the loss of creatinine, indicating a con-
tinual decrease in the tissue metabolism. As a matter of fact, as the
fast progresses, such a decrease in the total heat-production is clearly
shown by various methods and yet, as a reference to the section on the
pulse-rate will show, our whole conception of the relationship between
creatinine and energy transformation is seriously affected by the
increase in the pulse-rate found with this subject during the last week
of the fast. The total energy transformation is clearly due to the
active mass of protoplasmic tissue and the stimulus to cellular activity,
but at no part of the fast were conditions so sharply differentiated that
we may say with accuracy that the loss in the heat-production was
directly comparable to the loss in the normal katabolized tissue, as
indicated by the decrease in the total creatinine excretion. Similarly,
we find no definite relationship between the creatinine excretion and
the total basal heat production, utilizing the creatinine elimination as
an index of the total mass of active protoplasmic tissue remaining in
the body. It is, however, of great significance that the total creatinine
excretion decreased regularly as the fast progressed, thus indicating
not an approximation to depletion but a distinct tendency on the
part of the body to a conservation of its active protoplasmic tissue.
The decreasing differences between the preformed creatinine and
the total creatinine observable toward the end of the fast are not
easily explained upon the ground of the influence of acetone bodies
in the urine, since from the determinations of the ammonia and the
/3-oxybutyric acid we have no reason to believe that the acidosis was
materially less in the last week than at any other time. These values
would therefore indicate that a sufficient amount of acetone bodies
was not present in these urines to affect materially the quantitative
determinations of preformed creatinine.
1During the reading of this proof, we received the admirable article, "Basal Metabolism and
Creatinine Elimination," by W. W. Palmer, J. H. Means, and J. L. Gamble (Journ. Biol. Chem.,
1914, 19, p. 239).
URINE.
267
With the taking of food, there was a slight increase in the total
creatinine-nitrogen . on the first day, and the difference between the
nitrogen of the creatinine preformed and of the total creatinine dis-
appears after this day.
Table 30. — Total creatinine excreted in urine, per day and per kilogram
of body-weight, in experiment with L.
Date.
Day of fast.
Creatinine excreted.
Per kilogram
of body-weight
per day. ,.
Per day.
/
1912.
Apr. 14-15
1st
•
gm. ~
1.29
1.23
1.47
1.45
1.37
1.40
1.31
1.35
1.35
1.31
1.31
1.31
1.28
1.19
1.03
1.14
1.07
1.09
1.03
1.01
1.01
.96
.98
.92
.94
.96
.95
.91
.94
.89
.86
1.00
.91
1.88
mg.
21.5
20.8
25.2
25.3
24.2
24.9
23.5
24.4
24.6
24.1
24.3
24.4
23.9
22.3
19.4
21.7
20.6
21.1
20.1
19.8
19.9
19.1
19.6
18.5
19.0
19.5
19.4
18.7
19.5
18.6
18.1
21.2
19.3
18.5
15-16
2d
16-17
3d
17-18
4th
18-19
5th
19-20
6th
20-21
7th
21-22
8th
22-23
9th
23-24
10th
24-25
11th
25-26
12th
26-27
13th
27-28
14th
28-29
15th
29-30
16th
Apr. 30-May 1
17th
May 1-2
18th
2- 3
19th
3-4
20th
4- 5
21st
6- 6
22d
6-7
23d
7- 8
24th
8- 9
25th
9-10
26th
10-11
27th
11-12
28th
12-13
29th
13-14
30th
14-15
31st
15-16
16-17
17-18
determined in urine for about 22 hours.
As an indication of the relationship between the metabolism of
active tissue and the body-weight — a relationship that must at best
be somewhat approximate, depending in large part upon the composi-
tion of the body — it has been customary to obtain a so-called creatinine
coefficient by dividing the total creatinine excretion each day by the
body-weight. This coefficient is normally found to be not far from
268 A STUDY OF PROLONGED FASTING.
20 to 30 milligrams. We have computed the creatinine coefficient for
our subject L. for each day of the 31-day fast and give the values in
table 30. After the first 2 days of fasting, the coefficient remains
practically constant until the fourteenth day; it then shows a tendency
to fall for a week, and the last 10 days it remains nearly constant at
19 milligrams.
Rest Nitrogen.
When the urea-nitrogen, the ammonia-nitrogen, the uric-acid nitro-
gen, and the total creatinine-nitrogen are combined, we find that the
total amount is somewhat less than the total nitrogen found by th
Kjeldahl method. This remainder, or the so-called "rest-nitrogen,"
amounts usually to not far from 0.5 gram of nitrogen in the observa-
tions with our subject L. The values for this undetermined nitrogen
are recorded in table 29, in which it is seen that the largest amount
(0.86 gram) was found on the fourth day and the smallest (0.17 gram)
on the twenty-second day. Since these values include all the errors in
the analyses, the amounts thus recorded are not unexpected. Usually
they form about 5 per cent of the total nitrogen and in these records
vary from 7.22 per cent on the fourth day of the fast to 2.18 per cent
on the twenty-second day. In the food period following the fast, we
find a large increase on the percentage basis in this rest nitrogen,
which reaches 15.23 per cent of the total nitrogen excretion. It thus
appears that the chief factor affecting the nitrogen excretion on the last
day was the urea, for the absolute amounts of ammonia-nitrogen, uric-
acid nitrogen, creatinine-nitrogen, and rest-nitrogen remained essen-
tially constant on all three days of food.
ACID RADICLES.
Fasting urine contains a large number of acid radicles which
may be either organic or inorganic. Thus there are always present
considerable amounts of chlorine, phosphorus pentoxide, sulphur
trioxide, and (particularly in fasting experiments) /3-oxybutyric and
other fatty acids. By using Folin's titration method, it has been possi-
ble to determine the "total acidity." For a clear understanding of
the quantitative relationships of these various acid radicles, direct
determinations were made of the total chlorine, phosphorus pentoxide,
total sulphur, and /3-oxybutyric acid.
Chlorine.
Owing to the general satisfaction with the standard Volhard method,
chlorine has perhaps received more attention than any other of the
inorganic constituents of the urine. Recognizing the importance of
the determination of this constituent, especially in view of the emphasis
laid upon the supposed excess storage of chlorine in the body under
URINE.
269
normal conditions, we made duplicate determinations of the chlorine in
the urine for each day of the fast. The chlorine excretion in fasting
experiments has always had a particular interest, as the chlorine curve
almost invariably follows a fairly regular course. It is furthermore
important as presumably demonstrating whether or not the fast is
a true one, for it is commonly supposed that unless the food sur-
reptitiously taken is pure fat or pure carbohydrate, it is practically
impossible for a subject to break his fast without almost immediately
affecting the chlorine excretion. On the other hand, it was found,
Table 31. — Chlorine (CI) excreted in urine
daily by fasting
r subjects.
Day of fast.
L.
Succi.
5
O
i
«
B
m
O
<
GO
6
6
a
I
E
J
%
a
Z
1
>
g
a
3
w
Last food day . . .
gm.
gm.
26.322
1.350
.539
1.155
.848
.817
.840
.800
.736
.550
.513
.332
.405
.230
.119
.137
.113
.130
.258
.298
.311
.234
.216
.219
.235
.204
.118
.139
.239
.428
.688
gm.
4.51
4.65
2.86
2.00
.47
.55
.38
.37
.25
.44
.47
.64
.86
.62
.50
.58
.47
.80
.67
.40
.71
.56
gm.
4.792
3.908
2.212
1.799
1.198
1.092
1.044
.973
.700
.702
.412
.434
.567
.532
.497
.436
.403
.322
.306
.217
.233
gm.
gm.
gm.
5.432
1.606
2.303
1.7
1.548
1.396
1.088
.95
.814
1.104
.62
gm.
6.7
3.2
2.0
1.5
1.3
1.0
.84
.69
.39
.30
.18
gm.
7.51
2.99
1.73
3.66
1.90
.38
.30
.32
1.15
1.32
1.07
.98
1.29
.85
.68
gm.
4.11
1.45
1.34
.62
.25
!39
.42
1st
3.77
1.02
.79
.59
.41
.40
.55
.32
.31
.2.8
.36
.31
.32
.26
.16
.14
.12
.15
.16
.15
.18
.21
.18
.10
.18
.16
.16
.14
.12
.14
.13
.23
.26
8.18
9.03
3.21
1.55
1.48
1.18
1.29
1.11
1.12
1.21
.87
.92
.82
.58
.65
.51
.45
.58
.44
.67
.42
.42
2d
3d
4th
6th
6th
7th
8th
9th
10th
11th
12th
13th
14th
.24
16th
16th
17th
18th
19th
20th
21st
22d
23d
0.30
.29
.20
.21
.15
.19
.21
.33
24th
25th
26th
27th
28th
29th
30th
31st
1st food day ....
2d food day
3d food day
Reported by the investigators as NaCl, but converted to chlorine for purposes of comparison .
2 Average of 6 days before fast began.
'Determined in the urine for about 22 hours.
270 A STUDY OF PROLONGED FASTING.
in the experiment with L., that when the subject took considerable
amounts of food on the 3 days following the fast, the excretion of chlor-
ine was but slightly affected, certainly not enough to be considered as
proof that food had been taken. The character of the diet on these
food days easily explains this absence of influence upon the excretion
of chlorine.
The determinations of the chlorine excretion were made under Dr.
Peters's supervision by Mr. W. F. O'Hara according to the Volhard
method, the excess of silver nitrate added being determined in a filtered
portion of the urine. These values are compared in table 31 with the
values obtained for several other subjects in long fasting experiments.
With L. there was a large excretion of chlorine on the first day of the
fast, doubtless from the food previously taken. This was followed by
a marked fall, even on the second day, this decrease continuing almost
regularly until the fifteenth day, when the excretion reached a new
minimum level. It subsequently fluctuated slightly until the end of
the fast. On the 3 food days there was a slight increase over the latter
part of the fasting period.
An intelligent comparison of the values found for L. with those found
for other fasting subjects is somewhat difficult, owing to the facts that
frequently the basis upon which the chlorine is reported is somewhat
obscure and that some of the subjects, Succi in particular, were accus-
tomed to drink water containing more or less chlorine. Apparently
the chlorine excretion in complete inanition varies widely with different
individuals, for it will be seen at once, by inspection of the values in
table 31, that the excretion found for L., particularly in the first part
of the fast, was lower than that found for any other subject except for
a few days with Succi1 in the Naples fast, for 3 days with Tosca, and for
4 days with S. A. B. In the latter part of the long fasts, however,
there is more of a tendency toward uniformity, although the values
for Succi at Naples and at Vienna, and those for the latter portion
of Tosca's fast, are much higher than those found with L. There is
a general tendency shown with all of the subjects for the excretion to
decrease gradually until the fifteenth day, but not so rapidly as was
found for our subject. The observations of Brugsch on Succi at Ham-
burg give values that agree well with those found for L.
Relationship between chlorine excretion and preformed water lost. — A
critical examination of the tables in the report of the earlier fasting
experiments2 shows a rather interesting relationship between the excre-
tion of chlorine and the loss of preformed water from the body. For
a long time we have been at a loss to explain the marked variations in
the absolute amounts of chlorine excreted on different days of a fast
*I have been unable to obtain the original chlorine data in the study made of Succi's urine by
Koranyi, published in Orvosi hetilap, 1894, Nos. 39-40. See autoreferat, Maly, Jahrb. d. Tier-
Chemie, 1894, 24, p. 268. Koranyi also studied the depression of the freezing-point.
Benedict, Carnegie Inst. Wash. Pub. 77, 1907, table 216, page 415, and table 229, page 469.
URINE.
271
by different subjects, there being almost no uniformity; some subjects,
as Tosca, Cetti, and Succi in Vienna, excreting much larger amounts
than others. In a relatively few fasting experiments, both the loss of
preformed water from the body and the chlorine excretion have been
determined, thus supplying data which permit comparison. Such a
comparison is made in table 32, which gives the amount of water taken,
Table 32. — Preformed water eliminated from the body and accompanying excretion of chlorine
(CI) in experiments with fasting subjects.
Subject.
Day of
fast.
Water
con-
sumed.
A
Chlorine
(CI) in
urine.
B
Preformed water
lost.1
Water in urine.
Total.1
C
Per
gram of
chlorine.
(C-S-B)
D
Total.
E
Per
gram of
chlorine.
(b-hb)
F
L
1st
2d
3d
4th
1st
2d
1st
2d
1st
2d
1st
2d
1st
2d
1st
2d
1st
2d
1st
2d
1st
2d
1st
2d
gm.
720
750
750
750
2082
2747
1973
1729
2048
1593
783
340
133
206
291
194
858
1093
1467
884
705
708
115
357
gm.
3.77
1.02
.79
.59
1.63
.47
1.45
1.34
5.29
1.67
2.92
3.62
8.90
4.03
3.88
2.79
3.45
6.71
.52
.63
4.58
1.46
5.86
1.81
gm.
585
448
350
225
377
341
-334
273
736
713
471
623
1500
1132
540
777
47
991
-142
286
744
231
822
356
gm.
155
439
443
381
231
726
-230
204
139
427
161
172
169
281
139
278
14
148
-273
454
162
158
140
197
gm.
630.2
436.6
530.6
674.4
2225.8
2928.2
1469.6
1839.9
2528.2
2122.9
996.1
810.0
1105.8
734.8
621.3
783.2
538.6
1740.1
1159.2
1012.4
1145.1
642.4
599.6
492.2
gm.
167
428
672
1143
1366
6230
1014
1373
478
1271
341
224
124
182
160
281
156
259
2229
1607
250
440
102
272
S.A.B., Exp. 73...
Exp. 75 . . .
Exp. 77. ..
H.E. S., Exp. 79. ..
C. R. Y., Exp. 80. . .
A. H. M., Exp. 81 . .
H. C. K., Exp. 82 . .
H. R. D., Exp. 83 . .
N.M.P.,Exp.85..
D. W., Exp. 89. . . .
1Preformed water other than that resulting from the disintegration of flesh and fat. (Column
j, table 62, and Carnegie Institution of Washington Pub. 77, table 229, column c). Since the
chlorine in flesh is not a large proportion of the total chlorine, the water of flesh and fat is purposely
omitted in this discussion.
the chlorine excreted in the urine, the preformed water lost from or stored
in the body,1 the preformed water lost per gram of chlorine excreted,
the water in the urine, and the water in the urine per gram of chlorine
excreted during fasting. These factors are given not only for the
first 4 days of the experiment with L., but also for the first 2 days for
1For discussion of this factor, see page 408.
272 A STUDY OF PROLONGED FASTING.
a number of short fasting experiments made with eight subjects at
Wesleyan University.
It is conceivable that variations in the chlorine excretion might be
due, in part at least, to a flushing out of the body, and hence we should
expect to find the chlorine excretion varying to a certain degree with the
variations in the volume of urine or of the water taken. An examina-
tion of the figures in the last column of table 32 (column f) shows,
however, that the water in the urine per gram of chlorine excreted
undergoes wide variations, ranging from 102 grams to 6,230 grams, with
no obvious average value.
If we consider the body as losing regularly not only carbohydrate, fat,
and protein from its original store of substance at the beginning of
the fast, but also losing regular amounts of preformed water, i. e., water
of flesh and fatty tissue (see note to table 32), and water existing in
the fluids of the body, we can see that the importance of the determina-
tion of the preformed water lost is much greater than would at first
appear. The values for the preformed water lost by these fasting sub-
jects are given in column c of table 32, and for the preformed water
lost per gram of chlorine in column d. An examination of the values
for the preformed water lost per gram of chlorine shows that there is
at least a semblance to regularity. This is more clearly seen if the first
day of the experiment is omitted, as may properly be done, since it is
natural to suppose that the chlorine excretion on the first day may
have been influenced by the previous diet. Excluding the values for
the first day, then, we find that the remaining values range from 726 on
the second day of experiment 73 with S. A. B. to as low as 148 on the
second day of experiment 82 with H. C. K. Aside from the very high
value found in experiment 73, it will be seen that there is a fairly close
agreement between the chlorine excreted and the preformed water lost
from the body. It would thus appear that, in the discharge of this
water, there is excreted simultaneously an amount of chlorine approx-
imately proportional to the total preformed water lost. It is obvious
that all of the factors involved in the determination of the preformed
water lost from the body are such as to make the absolute values of
some of the determinations problematical, and yet we believe that as
a whole there is sufficient agreement here to indicate some approximate
relationship between the chlorine and the preformed water lost from
the body.
Source of chlorine excreted. — The exact source of the chlorine excreted
in the first days of fasting is by no means certain, but it is clear that
the small amounts excreted, in the urine after the first few days of
fasting, correspond to the usual percentage of chlorine commonly con-
sidered as belonging to human flesh. Thus Katz,1 whose analyses
have been considered as remarkably accurate, maintains that human
xKatz, Archiv f. d. ges. Physiol., 1896, 63, p. 1.
URINE. 273
muscle contains 0.07 per cent chlorine. Furthermore, Magnus-Levy's1
analysis of the flesh of a suicide agrees remarkably well with the values of
Katz. If we use this factor 0.07 for computing the chlorine in the flesh
katabolized by the subject L. as recorded in column q, table 61 (page
403), the values found would be approximately the amounts of chlorine
actually excreted. For instance, on the twenty-fifth day, there were
235 grams of flesh katabolized, the excretion of chlorine being somewhat
larger than the average for this part of the fast. Applying the factor
of Katz, we find that 0.07 per cent of 235 grams would give 0.165 gram
of chlorine, while the amount actually excreted on that day as shown
by analysis was 0.18 gram. It is clear, therefore, that, at least in the
later stages of inanition, chlorine is derived for the most part from dis-
integrated muscle substance. The large storage of salt in the skin,
which was noted by Wahlgren2 and subsequently further studied by
Padtberg,3 and Scholz and Hinkel,4 must therefore have been rapidly
depleted during the first days of the fast. In any event, the total
amount of chlorine involved throughout the whole 31 days of our
fasting experiment was but 12.27 grams, an amount so small as to cast
a doubt upon the theory that there is an excess5 of chlorine stored in
the body.
Phosphorus.
Since phosphorus has so intimate a relationship with both the mineral
and the organic constituents of the body, observations have been made
of the amounts present in the fasting urine of a large number of subjects.
The phosphorus in the urine of L. was determined by Mr. W. F.
O'Hara under the supervision of Dr. Peters, for each day of the fast
and for the following food days, by titration with uranium acetate.
Usage is followed here in expressing the values as phosphorus pentoxide
instead of as phosphorus, although the inconsistency of expressing
the elements in terms of their compounds is obvious. The absolute
amounts of phosphorus pentoxide determined in these urines, together
with those found in other long fasting experiments, are given in table 33.
The values obtained for L. show an increasing amount for the first
4 days and thereafter a very regular decrease for the remainder of the
fast. The maximum amount, 2.90 grams of phosphorus pentoxide,
was observed on the fourth day and the minimum amount, 1.32 grams,
on the last day. While in the first part of the fast the values for L.
are exceeded by those for Cetti and Beaute* and approximately equaled
by those for S. A. B., in the latter part of the fast they are measurably
Magnus-Levy, Biochem. Zeitschr., 1910, 24, p. 363.
*Wahlgren, Archiv f. exp. Path. u. Pharm., 1909, 61, p. 97.
'Padtberg, Archiv f. exp. Path. u. Pharm., 1910, 63, p. 60.
<Scholz and Hinkel, Deutsch. Archiv f. klin. Med., 1913, 112, p. 334.
*For a discussion of this point, see Magnus-Levy, Physiologie des Stoffwechsels, von Noorden's
Handbuch der Path, des Stoffwechsels, Berlin, 1906, 1, p. 451; Munk, Archiv f. Path. Anat. u.
Physiol., 1893, Supp. 131, p. 146; and Morawitz, Oppenheimer's Handbuch der Biochemie, Jena,
1910, 4 (2), p. 282.
274
A STUDY OF PROLONGED FASTING.
higher than those recorded for any other subject. The only other fast
in which the phosphorus was determined, and which extended over
so long a period as that for our subject, was Succi's fast in Florence,
but the amount excreted by this subject was considerably less than
that found for L. The values obtained by Brugsch on Succi in the
Table 33. — Phosphorus (P2O6) eliminated in urine daily by fasting subjects.
Day of fast.
L.
Succi.
"■+3
■♦a
6
<3
1
S
c3
8
O
PQ
<
00
i
a
%
!
6
1
03
t*
1
gm.
gm.
gm.
1.90
1.78
1.82
1.95
1.46
2 64
2.47
2.32
1.48
1.49
1.23
1.22
1.98
1.11
1.14
1.33
1.50
1.02
1.36
1.02
1.19
1.11
gm.
1.792
2.499
1.559
1.528
1.662
2.100
1.561
1.678
1.158
.841
.662
.518
.769
.879
.428
.465
1.162
1.079
.725
.610
gm.
gm.
gm.
2.76
2.597
2.925
3.289
2.974
2,871
2.667
2.663
1.722
2.065
.948
gm.
4.14
2.26
2.93
2.98
2.91
2.37
1.84
1.89
1.60
1.54
1.55
gm.
2.670
1.550
1.830
2 654
2.934
1.749
1.069
.713
1.658
1.702
1.461
1.097
1.312
1.114
.869
gm.
2.318
1.431
2.256
2.055
2.406
2.078
2.071
2.081
1st
1.66
2.48
2.51
2.90
2.64
2.33
1.84
1.84
2.13
1.97
1.95
1.70
1.95
1.86
1.47
2.04
1.99
1.86
1.75
1.47
1.60
1.57
1.62
1.55
1.53
1.49
1.41
1.35
1.46
1.39
1.32
.74
.31
*.21
1.9W
2.051
2.090
2.120
2.394
2.150
1.865
1.601
1.360
1.246
1.420
1.012
.363
.996
1.029
1.077
1.218
1.005
.953
.875
.747
.718
1.049
.790
.592
.783
.861
.945
.789
1.019
2.98
2.75
2. 52
2.54
2.51
2.27
2.13
2.31
2.40
1.68
1.41
1.35
1.04
.99
1.32
.876
1.34
.86
1.14
.67
.64
2d
3d
4th
6th
6th
7th
8th
9th
10th
11th
12th. .
13th. . .
14th
1.25
15th. .
16th.
17th. . .
18th. . .
19th. .
20th. .
21st
22d . .
23d
0.96
1.062
.980
.900
1.056
.901
.754
1.545
24th
25th
26th
27th. .
28th. .
29th. .
30th.
31st
1st food day. .
2d food day . .
3d food day. .
♦Determined in urine for about 22 hours.
Hamburg fast are also measurably less than those reported for our
fasting subject.
Relationship between phosphorus and total nitrogen. — Owing to the
intimate relationship between phosphorus and the organic tissues of the
body, particularly muscle, the ratio between phosphorus and total
URINE.
275
nitrogen has frequently been computed for fasting experiments. The
well-known determinations of phosphorus pentoxide and nitrogen in
muscle show that for each gram of phosphorus pentoxide there should
be 6.6 grams of nitrogen. The ratios between nitrogen and phos-
phorus pentoxide have been computed, not only for the fasting experi-
ment with L., but likewise, in so far as the data permit, for those of
Table 34.-
—Ratio
of nitrogen to phosphorus (^r^r-)
in urine of fasting subjects.
Day of fast.
L.
Succi.
Cetti.
Beaute.
Tosca.
S. A. B.
Flor-
ence.
Rome.
Vienna.
Ham-
burg.
5.10
3.57
5.88
5.68
5.09
4.77
6.03
5.11
7.03
7.55
8.63
9.54
6.65
5.44
10.30
6.77
2.86
3.76
5.27
5.66
4.89
6.22
4.30
3.99
4.17
3.73
3.79
4.09
5.17
5.24
10.00
3.97
4.65
4.90
4.60
4.71
4.54
5.25
5.03
5.23
5.51
5.65
5.24
5.65
4.58
4.04
3.20
4.50
7.23
8.57
4.64
4.32
4.65
5.60
5.31
5.04
4.70
8.41
8.55
5.52
6.34
4.83
5.23
5.19
4.87
1st
4.28
3.39
4.52
4.09
3.94
4.37
6.32
6.58
5.04
5.10
5.26
5.96
5.31
5.61
5.76
4.70
4.43
4.45
4.78
5.23
4.96
4.94
4.51
5.26
5.10
6.29
5.72
5.64
5.16
5.63
5.26
6.53
12.29
13.10
*7.87
5.91
7.30
6.64
5.90
5.18
5.53
5.85
6.29
5.96
6.11
7.79
10.63
5.89
6.50
5.62
5.57
6.97
5.81
5.51
6.72
4.90
4.99
7.73
11.23
7.11
6.85
6.52
5.69
7.14
5.70
4.07
4.19
4.25
4.46
4.85
4.13
4.22
4.19
4.24
4.48
5.07
4.94
4.71
3.83
4.83
4.03
4.19
5.00
4.93
4.41
2d
3d
4th
5th
6th
7th
8th
9th
10th
11th .
12th. .
13th .
14th
6.22
15th .
16th. .
17th
18th
19th
20th
21st. .
22d
23d
6.08
6.04
6.40
6.87
5.97
4.93
5.56
5.45
24th .
25 th
26th
27th .
28th. .
29th. .
30th
31st .
1st food day. . .
2d food day. . .
3d food day ....
♦The ratios shown in this column have been obtained by means of the nitrogen as corrected
by Munk. (See table 26, p. 249.)
i
the earlier fasting experiments reported in table 33. These ratios are
given in table 34.
Considering the fasting values for L., we find that in no case do they
reach the theoretical relationship found with muscle, namely, 6.6. The
highest value was 5.96 on the twelfth day and the lowest value was
276 A STUDY OF PKOLONGED FASTING.
3.39 on the second day of the fast. Furthermore, the figures show no
definite increment in the ratio as the fast progressed. While it is true
that values less than 5 are not found on the last 8 days of the fast,
nevertheless there is a period — that between the seventh and the fif-
teenth days — when the values again all lie above 5, while between the
fifteenth and the twenty-fourth days values as low as 4.4 are found.
In examining the values for the other fasts, we find several which
show distinctly abnormal values for the first day, i. e., 7.87 for the
first day of Succi's Florence fast, and 8.55 for the first day of the fast
of S. A. B., but this would naturally be expected. On the other hand,
after the first day the values should lie well below 6.6 and any above
this are open to suspicion. The ratios established by the Freunds for
the Vienna fast of Succi are remarkably constant, as are those of
Brugsch for a later fast of Succi in Hamburg. Perhaps the greatest
uniformity in ratios is shown by Cathcart's subject Beauts, these
ratios gradually and steadily increasing as the fast progressed. The
values for Succi in the Florence fast are somewhat vitiated by the
uncertainty in the nitrogen determinations, although the values were
computed on the basis of nitrogen as corrected by Munk.
Source of phosphorus excreted. — Comparing these ratios, particularly
those for L., with the theoretical relationship with flesh, we find in
all of the experiments a tendency toward a much larger excretion of
phosphorus pentoxide in its relation to nitrogen than occurs in the
ordinary composition of flesh. The possible sources of phosphorus in
the body other than the flesh are of course the nucleins and, above all,
the mineral matter of the bony structure. It is now the consensus
of opinion that the disturbance in the relationship between the phos-
phorus pentoxide and nitrogen in fasting experiments is due exclusively
to the material draft upon the bony structure as the fast progresses.1
This was clearly set forth by Munk in his discussion of the experi-
ments with Cetti and Breithaupt, but as these were short experiments
Munk frankly stated that he expected to find that the ratio would
become smaller and smaller as the fast progressed. The ratios in the
31-day fast with L. do not, however, become smaller as the fast con-
tinued, but on the other hand tend to become higher in the last week
than at any other time. Brugsch has already commented upon the
very high values found in the last week of his study with Succi.
During the fast with L., 277.32 grams of nitrogen2 were excreted,
corresponding to 8,319.6 grams of flesh katabolized. If we assume
that this flesh had normally combined with it 0.5 per cent of phos-
phorus pentoxide, the total amount combined with the katabolized
flesh would be equal to 41.6 grams. Since 56.63 grams of phosphorus
xWellman (Archiv f. d. ges. Physiol., 1907-1908, 121, p. 508) believes that the calcium and
phosphorus losses found by him conform with Munk's conclusion that there is considerable
loss of calcium and phosphorus from the bones during starvation.
2This amount does not include the nitrogen excreted through the skin (see table 22).
URINE. 277
pentoxide were excreted during the fast, it will be seen that there was
distinctly an excess excretion, amounting to 15 grams of phosphorus
pentoxide for the whole fast. This was undoubtedly derived — in large
part, at least — from the bones. It was hoped that the present-day
technique of Roentgen photography would show any material attacks
upon the bony structure and possible depletion of calcium, but the
excellent series of X-ray photographs taken by Dr. Francis H. Williams,
of the Boston City Hospital, did not indicate this. On the other hand,
when one considers the large storage of calcium in the body and the
relatively small draft, an ocular indication of such a draft which could
be measured could hardly be expected.
The exact apportionment of the phosphorus pentoxide between
muscle and bone is not permissible with the experimental data at
present in our hands. In all probability the amount of phosphorus
pentoxide actually drawn from the skeleton was considerably more
than 15 grams.
Since after the first few days there was no material increase in the
uric-acid excretion as the fast progressed, there was doubtless no direct
attack upon the nucleins, and thus the increase in the phosphorus
pentoxide could not have been derived from that source.
Sulphur.
Sulphur as an integral component of protoplasm is oxidized and ex-
creted in the urine in several forms: first, as sulphates; second, in the
form of conjugated sulphates, or sulphuric acid combined with organic
radicles; and finally, as so-called neutral or unoxidized sulphur. The
apportionment of the total sulphur output among these various com-
ponents has been studied during fasting in considerable detail, both
in the laboratory of Wesleyan University and by Cathcart. In the
series of experiments carried out on L., it was impossible to separate
the sulphur, owing to the lack of experimental material, and only the
total sulphur was determined. These determinations were personally
made by Dr. A. W. Peters, the constancy in the results testifying to
his technical skill.
It has been the custom of many writers to report the sulphur excre-
tion as sulphuric acid or sulphur trioxide, but the values given for L.
in table 35 represent the daily excretion of total sulphur. For com-
parison, the sulphur excretion has been computed from the results
given by other observers for fasting subjects, and these values are
included in the table. Two difficulties arise in comparing our results
with those obtained with other subjects. In the first place, frequently
only the total sulphuric acid was determined and the organic sul-
phur was not included. Secondly (in the case of Succi, at least), the
subject often drank mineral water containing measurable amounts of
sulphates.
278
A STUDY OF PROLONGED FASTING.
With our subject L., the sulphur excretion followed a course not unlike
that of the nitrogen excretion, there being a slightly increasing amount
excreted for the first 3 days and thereafter an almost continuous
decrease until the end of the fast. The maximum amount (0.68 gram)
Table 35. — Total sulphur (S) excreted in urine daily by fasting subjects.
Day of fast.
L.
Succi.
tt
Florence. Na
pies.
Rome. Vie
nn
a.
L. 1H auu'.
Last food day. .
1st
gm.
0.46
.61
.68
.67
.65
.65
.62
.64
.66
.61
.62
.62
.62
.60
.60
.69
.53
.54
.55
.61
.51
.50
.61
.49
.49
.64
.52
.53
.52
.52
.49
.39
.22
».36
gm.1 g
1
0.75
.72
.75
.71
.68
.59
.65
.49
.43
.42
.45
.43
.30
.33
.28
.26
.32
.21
.22
.24
.21
.19
.29
.35
.28
.19
.17
.19
.14
.24
.04
.76
.48
.72
.60
.58
.40
.49
.48
.44
.68
.30
.79
.38
.29
.20
.31
.70
.66
.35
.46
.61
gm. g
0.632
.690 1
.277
.346
.249
.312
.187
.276
.258
.272 !
.120
.130
.251
.207
.117
m.
21
5:
6-
r>(
4C
k
3f
2i
3(
2;
gm.
t 0.99(
1
! .921
t
.71]
\ .78]
.831
.61(
>
(
1.
)
.
!
L
I
im.
33
614
934
801
856
712
644
615
556
570
577
536
f
1
im.
36
62
67
75
72
67
66
62
2d
3d
4th
6th
6th
7th
8th
9th
10th
11th
12th
i
) '.
13th
14th
15th
16th
17th
18th
19th
20th
21st
22d
23d
24th
25th
26th
27th
28th
29th
30th
31st
1st food day. . .
2d food day. . . .
3d food day
xProbably inorganic and ethereal. Given by the investigator as H2SO4, but converted to S
for purposes of comparison.
2Probably inorganic and ethereal.
3Determined in urine for about 22 hours.
was excreted on the third day of the fast, and the minimum amount
(0.49 gram) was found on the twenty-fourth, twenty-fifth, and thirty-
first days.
The most carefully determined results for the other subjects are
undoubtedly those for Cathcart's subject Beaute\ Here again we find
a decrease in the sulphur excretion as the fast progresses, the values for
URINE.
279
the entire fast being not unlike those found for L. This is likewise true
for the results of the observations on S. A. B.
Relationship between total nitrogen and total sulphur. — The total
sulphur excretion has a special significance in that it is so intimately
combined with the protoplasm that it is frequently suggested as an
index of the total muscle katabolized. Since there is a relatively
constant relation between the nitrogen and sulphur in muscle, i. e.,
Table 36. — Ratio of nitrogen to total sulphur (— ) in urine of fasting subjects.
Day of fast.
L.
Succi at ,
Vienna.
3etti. B
eau
te. S. A. B.
Last food day
1st
15.4
13.8
16.7
17.7
16.0
15.7
15.8
16.0
16.3
16.5
16.6
16.3
16.7
17.4
16.9
16.2
16.6
15.3
15.2
15.1
15.5
15.5
14.3
16.6
15.9
14.6
15.5
14.4
14.5
15.1
14.2
12.4
17.3
7.6
13 .3
19.0
20.3
16.9
19.7
15.7
21.2
]
]
19.7
18.0
14.6
15.0
14.3
13.7
L4.*2
L5.1
L3.9
...
13. 0
L5.3
;
...
]
L2.
L7.
L5.
L7.
LG.
L5.
15. (
L5.<
15. (
14. <
15. J
L4.\
2 14.3
I 19.6
J 18.6
L 17.4
) 16.1
16.3
I 16.3
) 16.3
I
2d
3d
4th
5th
6th
7th
8th
9th
)
I
10th
11th
12th
13th
14th
15th
16th
17th
18th
19th
20th
21st
22d
23d
24th
25th
26th
27th
28th
29th
30th
31st
3d food day
about 13.3 grams to 1 gram of sulphur, the relationship between the
excretions of total nitrogen and total sulphur is worthy of note. The
nitrogen-sulphur ratio has been computed not only for the subject L.,
but for a number of the other fasting subjects; these computed ratios
are given in table 36. But one ratio for Succi is included in this table,
that for the Vienna fast, in which the observations were made by the
Freunds, for while it was permissible to include the observations of the
280 A STUDY OF PROLONGED FASTING.
sulphur excretion in the other fasts in table 35, it was not permissible
to compute the ratio between the nitrogen and the sulphur, since there
were undoubtedly errors in the nitrogen determinations and probably
likewise in those for the sulphur excretion.
The ratios found for L. show at first a slight increase, rising on the
fourth day to a maximum of 17.7. They then decrease with consider-
able regularity, the lowest value, 14.2, being on the last day of the fast.
In general, the ratios remain within very narrow limits. These values
are again not unlike those found by Cathcart with Beauts, which
showed a tendency to decrease as the fast progressed. The same ten-
dency is shown by the values for S. A. B., although at no time during
the 7-day fasting experiment did the ratio fall below 16.1.
All of the values found for L. show a somewhat higher excretion of
nitrogen than would normally accompany the amount of sulphur ex-
creted; it would appear, therefore, that there was a disintegration of
sulphur-poor and nitrogen-rich substance other than muscle. Through-
out all of the observations, it has been noted that on certain days there
was always a marked lowering in the excretion of the urinary compo-
nents. This lowering may be due either to an actual decrease in the
katabolism on that day or to a possible loss of urine. The handling,
sampling, and preservation of the urine were so strictly controlled that
it would seem impossible for such a loss to occur. It is conceivable,
of course, that the subject may not have emptied the bladder com-
pletely in the morning, but a compensating increase in the urine of
the next day would then be expected, which was not observed. If we
examine the values for sulphur, total creatinine, nitrogen, chlorine,
and phosphorus, we find that there was a distinct lowering in the
amount excreted on the fifteenth day of the fast, with occasionally a
slight indication of a compensating increase on the sixteenth day.
This points toward a loss of urine which we are as yet unable to
account for. As a matter of fact, no disturbance in the nitrogen-
sulphur ratio was found on the fifteenth day and, indeed, such dis-
turbance was not expected.
On the other hand, we note a very definite increase in the sulphur-
nitrogen ratio from 14.3 on the twenty-third day to 16.6 on the twenty-
fourth day. An examination of the values for the total nitrogen
excretion shows that here also there was a marked increase, thus indi-
cating the disintegration of a nitrogen-rich and a sulphur-poor sub-
stance other than muscle. The general course of the values for sulphur
and total creatinine show a striking similarity, indicating that the
katabolism resulting in the excretion of total creatinine is accompanied
by an excretion of sulphur. If we are to accept Folin's view that the
total creatinine is an admirable index of the total katabolism of tissue,
we may then conversely state that the sulphur is likewise an index.
URINE.
281
Total Acidity.
When the conditions are such that acidosis may be expected to
develop, as in fasting, a determination of the total acidity of the urine
is of special value. Accordingly the total acidity of the urine in this
experiment was determined for each day of the fasting period and also
for the 3 days following when food was taken. Under the direction
of Dr. A. W. Peters, the determinations were made by W. F. O'Hara,
according to the method of Folin,1 in which potassium oxalate was used
and the titration was carried out with 25 c.c. of urine. The values are
given in table 37, expressed in terms of cubic centimeters of deci-normal
sodium hydroxide solution.
Table 37. — Total acidity (— NaOH) of urine of fasting subjects.
Day of fast.
L.
Beaut6.
Day of fast.
L.
Succi at
Hamburg.
Last food day . .
1st
c.c.
*409
285
499
658
655
570
467
337
399
415
376
365
297
362
1 332
313
505
485
c.c.
582
378
640
687
604
454 *
358
344
280
256
212
228
18th
c.c.
383
347
264
296
301
287
328
303
289
284
253
268
263
227
139
56
359
c.c.
2384
295
263
300
283
265
156
103
19th
2d
20th
3d
21st
4th
22d
5th :
6th
23d
24th
7th
25th
8th
26th
9th
27th
10th
28th
11th
29th
12th
30th
13th
31st
14th
1st food day. .
2d food day. .
3d food day . .
15th
16th
17th
1 Acidity of urine on third food day before the fast.
2The figures in this column were reported by the investigator as normal sulphuric acid, but
are here proportionately increased for purposes of comparison.
"Determined on urine for 22 hours.
At the beginning of the fast, the acidity increased rapidly until the
maximum of 655 c.c. was reached on the fourth day. It then showed a
general tendency to decrease as the fast progressed, although occa-
sionally unusually high values were found, as on the sixteenth and
seventeenth days of the fast. In the 3 food days following the fast, the
acidity dropped almost immediately to about 60 c.c.
The fasting values best suited for comparison with the values for
acidity found for L. are those determined by Cathcart on Beaute* and
by Brugsch on Succi in Hamburg. Cathcart's figures for the total
acidity in the 14-day fast of Beaute* show values ranging from a maxi-
^olin, Am. Journ. Physiol., 1903, 9, p. 265.
282 A STUDY OF PKOLONGED FASTING.
mum of 687 on the third day of the fast to a minimum of 212 on the
twelfth day of the fast. Following the fasting period the subject was
given a starch-cream diet, and the acidity immediately dropped to
approximately 100 c.c.
From the twenty-third to the thirtieth days of the fast in Hamburg,
Brugsch found with Succi values ranging from 384 c.c. on the twenty-
third day to 103 c.c. on the thirtieth day — values which are not incom-
parable with those found by us with L.
It was impossible, owing to the insufficient supply of urine, to deter-
mine the mineral acidity in the urine of our subject.
jS-OXTBTJTTBIC AdD.
In the earlier fasting experiments at Wesleyan University, there was
strong evidence that a nitrogen-poor, carbon-rich substance was present
in the urine in large amounts. This was shown by determinations of
the carbon in the urine and the relationship between the carbon and
nitrogen and calories. At that time the opinion was expressed that in
all probability the material was jS-oxybutyric acid, but we were then
unable to make the determinations in addition to the other analyses.
Special effort was therefore made in this fasting experiment to deter-
mine the /3-oxybutyric acid in the urine as accurately as possible,
although the interesting paper of Brugsch, reporting the Hamburg
fast with Succi, and more recently the paper by Grafe,1 have left no
doubt as to the nature of this excess non-nitrogenous material in the
urine of a fasting man.
Results of determinations. — Of the methods for the determination of
/3-oxybutyric acid which were available at the time of this experiment,
that of Black2 was best fitted for our purpose. By this method, plaster
of paris is first mixed with the acidulated, dried urine and the mixture
is then extracted with ether, the /3-oxybutyric acid removed being
determined with the polariscope.3 The determinations were carried
out by Miss Alice Johnson under the supervision of Dr. A. W. Peters.
Since these determinations were made, a large amount of research
on /3-oxybutyric acid has been carried out; in the light of the technique
existing at the time of the experiment, however, the determinations,
while admittedly having a relatively large error, are nevertheless suffi-
ciently accurate to indicate that there was a material excretion of
jS-oxybutyric acid throughout the fast. The results are recorded in
table 38 (column f).
On the second day of the fast only 0.5 gram of /3-oxybutyric acid was
found; this was wholly in fine with what would be expected. Subse-
quently no values less than 1.4 grams were found until the fast had been
xGrafe, Zeitschr. f. phyaiol. Chemie, 1910, 65, p. 21.
2Black, Journ. Biol. Chem., 1908, 5, p. 207.
'For a more explicit statement of this procedure, see Benedict and Joslin, Carnegie Inst. Wash.
Pub. 136, 1910, p. 25.
URINE.
283
concluded. On the first day with food, the /3-oxybutyric acid dropped
at once to 0.8 gram. During the last week of the fast, the results
obtained by the optical method showed a considerable amount of
/3-oxybutyric acid present in the urine.
Table 38. — p-oxybutyric acid excretea
in urine in experiment with L.
Date.
Day of
fast.
Total
nitrogen.
Carbon.
/3-oxybutyric acid.
Computed
normal
Deter-
Differ-
Calcu-
lated.
Deter-
excretion.
mined.
(c-b)
/dX100\
^ 46.11 '
mined.
(aXO.79).
A
B
C
D
£
F
1912.
gm.
gm.
gm.
gm.
gm.
gm.
Apr. 14-15
1st
7.10
5.61
6.82
0.21
0.46
15-16
2d
8.40
6.64
7.99
1.35
2.93
6.5
16-17
3d
11.34
8.96
10.35
1.39
3.01
2.1
17-18
4th
11.87
9.38
11.88
2.50
5.42
3.5
18-19
5th....
10.41
8.22
10.69
2.47
6.36
2.1
19-20
6th ... .
10.18
8.04
10.42
2.38
6.16
3.6
20-21
7th
9.79
7.73
9.06
1.33
2.88
2.8
21-22
8th....
10.27
8.11
10.30
2.19
4.75
1.6
22-23
9th
10.74
8.48
10.92
2.44
6.29
3.5
23-24
10th
10.05
7.94
9.92
1.98
4.29
3.5
24-25
11th
10.25
8.10
9.59
1.49
3.23
1.4
25-26
12th
10.13
8.00
9.05
1.05
2.28
2.4
26-27
13th
10.35
8.18
10.15
1.97
4.27
4.2
27-28
14th
10.43
8.24
9.95
1.71
3.71
4.7
28-29
15th
8.46
6.68
8.71
2.03
4.40
1.6
29-30
16th
9.58
7.57
11.39
3.82
8.28
5.2
Apr. 30-May 1 . .
17th....
8.81
6.96
10.91
3.95
8.57
3.6
May 1-2
18th
8.27
6.53
9.65
3.12
6.77
4.4
2- 3
19th....
8.37
6.61
9.56
2.95
6.40
7.0
3-4
20th....
7.69
6.08
8.07
1.99
4.32
4.4
4-5
21st
7.93
6.26
8.69
2.33
5.05
5.0
5- 6
22d
7.75
6.12
8.40
2.28
4.95
3.1
6- 7
23d
7.31
5.77
7.25
1.48
3.21
6.0
7-8
24th
8.15
6.44
8.68
2.24
4.86
6.9
8- 9
25th....
7.81
6.17
8.58
2.41
5.23
4.4
9-10
26th....
7.88
6.23
8.56
2.33
5.05
6.1
10-11
27th ....
8.07
6.38
8.23
1.85
4.01
4.0
11-12
28th
7.62
6.02
7.73
1.71
3.71
4.9
12-13
29th
7.54
5.96
7.94
1.98
4.29
5.6
13-14
30th ....
7.83
6.19
7.95
1.76
3.82
5.4
14-15
31st
6.94
5.48
7.37
1.89
4.10
4.5
15-16
4.83
3.81
»2.75
3.82
3.01
2.17
7.13
4.28
3.31
1.27
7.18
2.75
.8
.5
1.5
16-17
17-18
determined in urine for about 22 hours.
The determination of /3-oxybutyric acid by this method was so
defective that it seems unwise to compute the amount of ammonia
which would theoretically combine with the acid and to discuss any
relationship arising therefrom.
Indirect computation of amounts of p-oxybutyric acid excreted. — In
addition to the determinations made by the Black method, the amounts
284 A STUDY OF PROLONGED FASTING.
of /8-oxybutyric acid excreted were also computed by an indirect
method, using the relationship between the nitrogen and the carbon.
The carbon and the nitrogen in the urine were determined for every-
day of the fast and for the 3 food days following the fasting period.
Normally there exists a relatively definite relationship between the
carbon and the nitrogen of urine, a relationship which was determined
by Munk on Breithaupt and Cetti as being not far from 1 part of
nitrogen to 0.82 of carbon. This relationship was determined for L.
for 2 days before the fast, and a ratio found, which may be termed
"normal," of 1 to 0.79. We therefore computed the amount of carbon
that would normally be excreted from the amount of nitrogen actually
excreted, and deducted the result found from the total amount of
carbon found in the urine. The excess carbon would be due, in all
probability, to /3-oxybutyric acid or to acetone bodies. The results
of this computation are considered of sufficient importance to be
included in table 38, which gives the total nitrogen for each day, the
values for the normal excretion of carbon as computed with the ratio
C : N = 0.79, the total carbon as actually determined, the difference
which would be ascribed to /3-oxybutyric acid, and finally the values for
/3-oxybutyric acid obtained by using 46.11 as the percentage of carbon
in /3-oxybutyric acid. As the values for /3-oxybutyric acid actually
determined are also given in this table, a comparison may readily be
made.
In general the determined values are somewhat lower than those
obtained by computation. Occasionally, especially toward the end of
the fast, the determined values are higher than the calculated values.
On the whole, however, there is sufficient agreement to give an approxi-
mate estimate as to the amount of /3-oxybutyric acid present. Indeed,
taking everything into consideration and the regularity of the figures,
it would appear that the computed values are probably more nearly
accurate than those which were actually determined. In any event,
the amounts here found are noticeably less than those computed from the
laevo-rotation of the urine observed by Brugsch with Succi and by Grafe
with his fasting woman. When the computed values are compared, it
is found that the greatest excretion of /3-oxybutyric acid was from the
sixteenth to the nineteenth days, when for 4 days an average amount
of over 7 grams was excreted.
The amounts of /8-oxybutyric acid present in these urines are not
unlike those found in short fasting experiments or in experiments with
a carbohydrate-free diet, and the stimulating effect of these acids upon
metabolism is certainly not to be ignored. On the other hand, with the
acid present in the body for so long a time, the subject might easily
have become accustomed to its presence and therefore the reaction be
less, as has been found in cases of severe diabetes.
It is perfectly clear, however, that the amounts of /3-oxybutyric acid
involved in these determinations are not sufficient to affect the respira-
URINE. 285
tory quotients to an appreciable degree, since, as Magnus-Levy1 has
pointed out, the excretion of 20 grams of /3-oxybutyric acid per day
results in a lowering of the respiratory quotient only 0.006. The pres-
ence of this acid is, however, sufficient to account for the increase in
the ammonia excretion as the fast continued. The effort of the body
to correct this undue acidity by combining ammonia with the acid is
thus clearly shown.
Bonniger and Mohr2 found with Schenk much greater amounts of
acid than were found or indeed calculated for the urine of L. Brugsch3
likewise found large amounts with Succi, and the excretion for Grafe's4
insane patient was also large. The uncertainty in the method for the
quantitative determination of /3-oxybutyric acid does not permit use
of these values for comparison, and it is sufficient to state that lsevo-
rotatory /3-oxybutyric acid in appreciable amounts is excreted in the
urine during fasting.
Beginning on the afternoon of April 21, 1912, and continuing at
least every other afternoon during the fast, a qualitative test for acetone
in the breath was made, using the reagent of Scott- Wilson.5 Several
tests were also made each day after the fast, the latest being on May
17 at 7 p. m. In every case the test showed acetone present, but it is
impossible to draw quantitative deductions from the results.
MINERAL METABOLISM.
With normal man the mineral metabolism has two main paths for
excretion — through the solid salts of the urine and through the feces.
With L. the entire mineral excretion took place through the urine, if
one excepts the small amount of sodium chloride excreted through the
skin. Usually much stress is laid upon the determination of the acid
radicles in the urine, namely, chlorine, sulphur trioxide, and phosphorus
pentoxide, and but little attention is given to the calcium, magnesium,
potassium, and sodium metabolism. The important relationship
between phosphorus and calcium and the possible draft upon the
skeletal tissue of the body made the determination of the mineral
constituents of the urine desirable, and although a very large number of
determinations were made of the various components of the urine of
our subject, it was possible, by combining and apportioning the mate-
rial, to secure a sufficient sample of urine for the determination of the
mineral constituents.
The analyses were made by Mr. J. C. Bock, who was at that time a
member of the Laboratory staff and who had had previous experience
in mineral analysis. Since special training was necessary for this inves-
^agnus-Levy, Zeitschr. f. klin. Med., 1905. 56, p. 83.
2Bonniger and Mohr, Zeitschr. f. exp. Path. u. Therapie, 1906, 3, p. 675.
3Brugsch. Zeitschr. f. exp. Pathol, u. Therapie, 1905, 1, p. 419.
4Grafe, Zeitschr. f. physiol. Chemie, 1910, 65, p. 27.
8Scott-Wilson, Journ. Physiol., 1911, 42, p. 444.
286
A STUDY OF PROLONGED FASTING.
tigation of the mineral metabolism during fasting, Mr. Bock was allowed
the privilege, through the kindness of Dr. Rufus S. Cole of the Rocke-
feller Hospital, of working with Dr. Francis H. McCrudden, at that
time of the Rockefeller Hospital, and whose researches in mineral
metabolism are too well known to need special mention here. Having
acquired certain of Dr. McCrudden's methods and technique, Mr.
Bock made a most careful analysis of the urine of L., determining the
calcium, magnesium, sodium, and potassium, so that we have a fairly
Table 39. — Mineral metabolism (urine excretion) in experiment with L.
Date.
Total.
Calcium
(Ca).
Magnesium
(Mg).
I
Aver-
age per Total,
day-
Aver-
age per
day.
Potassium
(K).
Total.
Aver-
age per
day.
Sodium
(Na).
Total.
Aver-
age per
day.
gm.
gm.
gm.
gm.
1.630
368
368
368
445
445
.883
.883
.883
1.006
1.006
.070
.779
.552
.463
.199
2.070
.926
.926
.926
I .276
i .276
f .154
I .154
I .154
f .100
I .100
1.627
.814
.814
.217
.109
.109
i:
{:
:
:
:
.;
:
676
676
644
644
643
643
787
787
656
656
585
585
606
606
430
430
.116
.102
.131
.166
.129
.109
.071
.105
.107
.046
.051
.051
.066
.066
.083
.083
.065
.065
.055
.055
.036
.036
.053
.053
.054
.054
.046
determinations in urine for about 22 hours.
URINE. 287
complete picture of the mineral metabolism of this subject throughout
the entire fast. As a rule, the urines for 2 days were combined, the
determinations thus representing the amounts for 2-day periods. On
two occasions it was necessary to combine the urines of 3 days, while
the mineral metabolism was determined for the days preceding the
fast on a sample representing the total urine for those 4 days. The
results are given in table 39, the values per day for convenience being
interpolated.
Throughout the entire fast there was a material excretion of all four
of these elements. The calcium excretion remained relatively constant,
with a slight tendency to fall off as the fast progressed, particularly after
the third week. The highest amount, 0.274 gram per day, was observed
on the fifth and sixth days of fasting; the smallest amount, 0.131 gram,
was found on the twenty-eighth and twenty-ninth days of the fast.
With magnesium there was a very considerable increase in the first
portion of the fast, and even for the last days of the fasting period the
average was 0.052 gram, which was considerably more than the amount
excreted per day in the four days prior to the fast, i. e., 0.034 gram.
With the potassium there was a very great decrease as the fast pro-
gressed. The largest amount, 1.63 grams, was found on the first day
of the fast, the excretion steadily falling until on the twenty-eighth and
twenty-ninth days of the fast it reached 0.585 gram. It is interesting
to note that at all times there was measurably over 0.5 gram of potas-
sium excreted per day.
With sodium we find perhaps greater variations than for any other
of these four substances, there being a rapid decrease for the first few
days, followed by a more moderate but steady decrease for the remain-
der of the fast. The values range from 2.07 grams on the first day of
the fast to 0.036 gram on the twenty-eighth and twenty-ninth days.
It is thus seen that the decreases in the minerals excreted were by no
means parallel.
Relationships of the Mineral Constituents.
In studying the mineral metabolism of the subject L. as indicated
by the results of the urinary analyses given in table 39 we see in-
stantly the great advantage in having the subject drink only distilled
water, for the entire mineral output is then derived solely from the
body tissue; we can therefore consider the relationships between
calcium and magnesium and between sodium and potassium without
the discussion being complicated by the possibility of varying amounts
of these elements in the drinking-water.
While a continuous decrease is evident in the excretion of all four
elements, it is likewise clear that the diminution in the excretion is
more marked with sodium than with any other element. Fortunately
the relations between potassium and sodium and calcium and magne-
288 A STUDY OF PROLONGED FASTING.
shim in the animal body have been recently carefully studied,1 and we
may more advantageously study the excretion of these four elements by
noting the ratios of their oxides to each other than in any other way.
Consequently the percentages of calcium oxide and magnesium oxide
excreted have been computed, using the total of the calcium oxide and
magnesium oxide as 100. Similarly the potassium and sodium oxides
have been added together and the percentages of potassium oxide and
sodium oxide computed. These percentages, which are given in table
40, were computed for each sample of urine analyzed.
Prior to the fast there was about eight times as much calcium oxide
excreted in the urine as magnesium oxide, but this relationship became
4 to 1 on the first fasting day and subsequently the percentages were
fairly constant at approximately 75 per cent for the calcium oxide and
25 per cent for the magnesium oxide. It was not until the third day
with food following the fast that this relationship was disturbed, when
the original percentages prior to the fast were again approximated.
According to these ratios, then, there were relatively much larger
amounts of magnesium excreted during fasting than on the food days.
In considering the ratios for the potassium oxide and sodium oxide,
we find that the problem is considerably complicated by the fact that
a certain amount of sodium chloride exists in the body which is very
loosely, if at all, combined in the tissues. Consequently, on the first
day of the fast, only 41.3 per cent of the total alkali was found in the
form of potassium oxide and 58.7 per cent in the form of sodium oxide.
On the next day the proportion was reversed, and from that time until
the eighteenth day there was a tendency toward a gradually increasing
percentage of potassium oxide. For the remaining days of the fast,
about 90 per cent of the alkali was in the form of potassium oxide and
10 per cent or less in the form of sodium oxide. One specimen of urine —
that for the twenty-eighth and twenty-ninth days — shows the high
percentage of 93.6 for potassium oxide as compared with 6.4 per cent
for sodium oxide.
It is obvious from the values for the potassium and sodium that
shortly after the excretion of the uncombined sodium chloride on the
first few days, the body excreted the potassium and sodium from muscle
substance and we then have the large differences between these two
elements; in the last part of the fast, approximately nine or ten times
more potassium was excreted than sodium. This far exceeded the
ordinary proportion between potassium and sodium in muscle given
as a result of the analyses of Bunge.2
The relationship between calcium and magnesium in animal tissues
has been extensively studied, particularly with dogs and horses.
Gerard, Ann. de l'Inst. Pasteur, 1912, 26, p. 12.
sSee Aron, Oppenheimer's Handbuch der Biochemie des Menschen u. der Tiere, 1909, 1, p. 84,
where the relationship is given as 5 to 6 times as much potassium as sodium.
URINE. 289
Table 40. — Distribution of mineral metabolism (urine excretion) in experiment with L.
Day
of fast.
Calcium and magnesium.
Potassium and sodium.
Total,
(grams)
CaO.
MgO.
Total.
(grams)
K20.
Na20.
Date.
Grams.
Per cent.
/bX100\
Grams.
Per cent.
/dX100\
Grams.
Per cent.
Grams.
Per cent.
(xXJOO)
A
B
c
D
£
F
G
H
I
J
1912.
\pr. 10-11
1
11-12
>1.942
0.381
[l.545
1.718
0.304
1.020
88.5
79.8
66.0
0.224
0.077
0.525
11.5
20.2
34.0
12-13
4.752
8.688
1.964
4.944
41.3
56.9
2.788
3.744
58.7
43.1
13-14
14-15
15-16
16-17
17-18
1st .
2d..
3d..
4th.
18-19......
19-20
5th.
6th.
|l.090
il.409
0.765
70.2
0.325
29.8
4.224
3.480
82.4
0.744
17.6
20-21
21-22
22-23
7th.
8th.
9th.
1.062
75.4
0.347
24.6
3.816
3.192
83.6
0.624
16.4
23-24
24-25
25-26
26-27
10th.
11th.
12th.
13th.
)o.854
Jo. 819
0.616
0.604
72.1
73.7
0.238
0.215
27.9
26.3
2.692
2.424
90.0
0.268
10.0
27-28
28-29
14th.
15th.
jo. 893
]o.856
0.659
73.8
0.234
26.2
2.253
1.960
87.0
0.293
13.0
29-30
\.pr. 30-May 1..
16th.
17th.
0.597
69.7
0.259
30.3
Vlay 1-2
2-3
18th.
19th.
Jo. 897
0.701
78.1
0.196
21.9
1.766
1.628
92.2
0.138
7.8
3-4
4-5
20th.
21st..
Jo. 838
0.664
79.2
0.174
20.8
1.728
1.552
89.8
0.176
10.2
5-6
6-7
22d..
23d..
Jo. 664
0.500
75.3
0.164
24.7
1.772
1.548
87.4
0.224
12.6
7- 8
8- 9
24th.
25th.
Jo. 652
0.468
71.8
0.184
28.2
2.070
1.896
91.6
0.174
8.4
9-10
10-11
26th.
27th.
Jo. 596
0.428
71.8
0.168
28.2
1.726
1.580
91.5
0.146
8.5
11-12
12-13
28th.
29th.
Jo. 520
0.366
70.4
0.154
29.6
1.504
1.408
93.6
0.096
6.4
13-14
14-15
15-16
30th.
31st. .
Jo. 558
Jo. 263
0.147
0.385
69.0
0.173
31.0
1.602
1.460
91.1
0.142
8.9
16-17
0.202
0.134
76.8
91.2
0.061
0.013
23.2
8.8
1.179
0.203
1.035
0.140
87.8
69.0
0.144
0.063
12.2
31.0
17-18*
1 Determinations in urine for about 22 hours.
Toyonaga1 in Tokio found that in the muscles there is always less
calcium than magnesium, thus confirming the analyses of Katz,2 but
in the glands there is always more calcium than magnesium.
Aloy,3 studying the calcium and magnesium content of muscle, found
about twice as much magnesium as calcium. Perhaps of more interest
toyonaga, Bui. Coll. of Agr., Tokio, 1902-1903, 5, p. 143, and p. 455; also 1904-1905, 6, pp.
89 and 357.
2Katz, Archiv f. d. ges. Physiol., 1896, 63, p. 1.
3Aloy, Compt. rend. Soc. Biol., 1902, 54, p. 604.
290 A STUDY OF PROLONGED FASTING.
in this particular study are the observations of Magnus-Levy,1 who ana-
lyzed the body of a man who had committed suicide. His results
showed about three times as much magnesium as calcium in the muscles.
On the basis of all the analyses prior to those of Magnus-Levy, Aron2
gives the average figure for the relationship of magnesium to calcium
in the muscle of dogs as 1 to 0.54-0.60 and in the muscle of horses as
1 to 0.34. In general, then, we may assume that there is approxi-
mately three times as much magnesium as calcium in the muscle of man.
We may consider the magnesium excretion as more nearly an index
of the muscle disintegration than calcium, for while there is, to be sure,
a small percentage of magnesium oxide in bone, there is a much larger
available supply of calcium oxide in the form of bone which is unques-
tionably drawn upon. The calcium-oxide excretion is therefore the
resultant of two factors, i. e., muscle disintegration and bone disin-
tegration, while magnesium oxide is derived almost exclusively from
the non-osseous tissue.
It is a matter of regret that the small amount of magnesium present
in the urine and the necessity for combining the samples for several
days renders it very difficult to make an exact comparison between
the magnesium excretion and the other urinary constituents. In
general, however, the magnesium excretion follows approximately
the nitrogen excretion. Of striking interest is the fact that, on the
first day of the fast, an extraordinarily low amount of both magnesium
and calcium were excreted. The increment in the magnesium excre-
tion on the second day (i. e., 100 per cent) was not approximated by
the excretion of any other element in the body.
The results of the analyses reported by Cathcart for Beaute, while
relatively few in number, are in full conformity with our findings,
save that the magnesium excretion on the days prior to the fast is
much larger than that found with L. Furthermore, L. excreted con-
siderably more calcium per day than did Beaute" during the fasting
period.
In the food period following the fast, it is interesting to note the
striking fall in the excretion of all of the minerals save sodium. For
the first 2 food days there was but half as much calcium excreted as
on the last day of the fast, about one-third as much magnesium, seven-
tenths as much potassium, and about the same amount of sodium.
On the last day of the food period after the fast, the calcium was seven-
tenths that of the last fasting day, the magnesium one-seventh, the
potassium one-fifth, and the sodium about the same as on the last fast-
ing day.
The amounts of calcium, magnesium, potassium, and sodium intake
are unknown for these days, yet these elements must have been present
in the food taken. It is obvious that the effect of the ingestion of a large
Magnus-Levy, Biochem. Zeitschr., 1910, 24, p. 363.
2Aron, Oppenheimer's Handbuch der Biochemie des Menschen u. der Tiere, 1909, 1, p. 88.
URINE. 291
amount of carbohydrates upon the mineral metabolism was consider-
able, resulting in a marked retention of the inorganic salts introduced,
with a noticeable lessening in the attack upon the storage of mineral
matter in the body. On the other hand, it should be remembered that
on these days fecal matter was passed in considerable amounts and we
may have here to deal only with the disturbance in the paths of excre-
tion of mineral matter. Thus, in the total amount of fecal material
excreted between 5 p. m. and 8 a. m. on the first food day following the
fast, there were excreted 1.78 grams of calcium oxide and 0.748 gram
of magnesium oxide, as determined by Mr. Bock on the dry matter of
feces. On this basis 30 per cent of the earthy alkali was magnesium.
REDUCING POWER.
The presence, even in fasting urines, of reducing substances other
than dextrose, has frequently been noted. Munk1 especially has
studied this subject and made extensive observations of the reducing
power of the urines in Breithaupt's experiment, using a reduction
method developed by himself2 and further elaborated and tested by
Hagemann.3 Munk's method gives as the reducing power for normal
urines from 0.16 to 0.47 per cent, with an average of 0.3 per cent. In
the fasting experiment with Breithaupt he found in the urine of the
last 2 food days as high as 7.7 grams of reducing substance (calculated
as dextrose). Even in fasting periods amounts were obtained ranging
from 3 to 7 grams per day.4 Furthermore, he noted very considerable
fluctuations from day to day. Munk considers the reducing action to
be due in large part to the formation of glykuronic acid. The reducing
power of the urine bore no relationship to the amount of carbohydrate
ingested, but there was a tendency to parallel the protein disintegrated.
In connection with the experiment with L., Dr. A. W. Peters suggested
that it would be desirable to test the reducing power of the urine. As
Dr. Peters had previously developed an accurate method5 for testing
the amounts of reducing substances in urine, this could be done to
advantage, and accordingly the determinations were made for each day
of the fast by W. F. O'Hara under Dr. Peters's supervision. The results,
expressed in terms of dextrose, are given in table 41.
The Peters method gives considerably less reducing substance in the
urine than the method of Munk. Thus, for normal urines, Peters has
found in this laboratory from 0.03 to 0.12 per cent as compared with
the values of 0.16 to 0.47 per cent found by Munk. Since the observa-
tions of the urine in our fasting experiment were to be wholly compara-
tive, either method was suitable for studying the variations in the
^unk, Archiv f. path. Anat. u. Physiol., 1893, 131, Supp., p. 138.
2Munk, Archiv f. path. Anat. u. Physiol., 1886, 105, p. 73.
3Hagemann, Archiv f. d. ges. Physiol., 1888, 43, p. 501.
4Munk, Archiv f. p*th. Anat. u. Physiol., 1893, 131, Supp., p. 68, table 7.
6See description of method by Dr. Peters in Benedict and Joslin, Carnegie Inst. Wash. Pub.
176, 1912, p. 8.
292
A STUDY OF PROLONGED FASTING.
reducing power as the fast progressed, regardless of any inherent differ-
ences which might exist in the two methods.
The reducing power of the fasting urines in the experiment with L.
was at all times well within normal limits, averaging not far from the
values observed on control normals in this laboratory. The largest
amount on the fasting days was 498 milligrams for the fourth and six-
teenth days, and the smallest amount was 296 milligrams for the third
and sixth days.
A large amount of reducing power was found on the first day of food
following the fast. On this day the subject ate in a relatively short
time about 500 grams of carbohydrate, chiefly in the form of soluble
Table 41.-
— Total reducing power
of urine in experiment with L.
Date.
Day of
fast.
Reducing
power (as
dextrose).
Date.
Day of
fast.
Reducing
power (as
dextrose) .
1912.
Apr. 14-15
15-16
16-17
17-18
18-19
19-20
20-21
21-22
22-23
23-24
24-25
25-26
26-27
27-28
28-29
29-30
Apr. 30-May 1 . . .
1st
2d
3d
4th ... .
5th....
6th....
7th....
8th....
9th....
10th....
11th
12th
13th
14th....
15th....
16th....
17th....
mg.
450
414
296
498
328
296
376
376
414
356
342
396
343
384
328
498
434
1912.
May 1- 2. . .
2- 3 . . .
3- 4 . . .
4- 5 . . .
5- 6 . . .
6-7...
7-8...
8-9...
9-10...
10-11.. .
11-12...
12-13 . . .
13-14...
14-15. ..
15-16 . . .
18th . . .
19th . . .
20th...
21st. . .
22d
23d
24th . . .
25th...
26th . . .
27th . . .
28th. . .
29th...
30th . . .
31st. . .
mg.
396
376
376
312
342
328
376
356
376
328
342
376
396
342
4441
267
x246
16-17. . .
17-18
determined in urine for about 22 hours.
dextrose, the diet consisting of honey and the juice of grapes, oranges,
and lemons. Unquestionably this amount exceeded his carbohydrate
tolerance on that day and 4.44 grams of dextrose were therefore
excreted in the urine. It is possible that we find here with man the
condition with dogs described by Hofmeister1 as "hunger diabetes,"
and it may be an interesting confirmation of his theory.
Recently, also, Rietschel2 has noted that the fasting of infants has
resulted in a marked lowering of their tolerance for carbohydrate sub-
stances. This lowering has been so noticeable as to lead Rietschel to
warn clinicians in the following words against the undue use of starving
in treating pathological cases:
"Dass der Hunger spez. die absolute Nahrungsentziehung fiirdengesunden,
wie besonders fur den ernahrungsgestorten Saugling auch schwere Gefahren
nach sich Ziehen kann, ist heute allgemein anerkannt."
hofmeister, Archiv f. exp. Path. u. Pharm., 1889-1890, 26, p. 355.
2Rietschel, Heubner's Festschrift, Berlin, 1913, p. 516.
URINE. 293
To attempt an analysis of the reducing substances in the urine during
the fasting period would be somewhat difficult. A certain portion
has already been ascribed to creatinine and to uric acid, but there is no
definite relationship between the amounts of uric acid, creatinine, and
the reducing power to be noted in the results obtained in this fasting
experiment.
CARBON IN URINE.
While the excretion of carbon in the form of carbon dioxide is of great
significance as indicating the total amount of energy transformed into
heat in the body, nevertheless when a study of the total loss of body
material is of importance, as it is during complete fasting, the carbon
in the urine must be taken into consideration. Since we were also
making a study of the energy output of the urine by burning samples
in a calorimetric bomb, it was relatively simple to combine the deter-
minations of the energy output and the carbon content. Conse-
quently, after the dried urine had been burned inside the bomb, the
carbon dioxide produced in the combustion was allowed to escape from
the vessel through weighed soda-lime tubes, in accordance with the
method of Fries.1 These determinations were skilfully carried out by
Mr. A. W. Cornell, of the Laboratory staff.
The description of the method of preparation and drying of the urine
samples for the bomb calorimeter has already been given in the dis-
cussion of the total solids in the urine.2 From the weight of the carbon
dioxide in the soda-lime tubes and the weight of the urine, the amount
of carbon per day excreted in the urine is readily computed. These
values are given in table 42.
The determination of carbon in urine has been for many years a
subject of research in this laboratory and in the chemical laboratory j
of Wesleyan University. Various methods have been tried, including
the moist combustion process, drying with and without the addition
of salicylic acid, and with and without the use of the cellulose filter
blocks recommended by Kellner.3 The method which gives the largest
percentage of carbon is presumably the best one, and this has been our
criterion. No method that we have thus far used approaches the large
percentage of carbon which is obtained by the method previously de-
scribed,4 namely, drying first with 50 milligrams of salicylic acid, then
transferring to a nickel capsule, drying in a desiccator until ready to burn,
and finally burning in compressed oxygen in a bomb calorimeter, and
allowing the carbon dioxide to escape into soda-lime. When the heat
of combustion is desired, the determination of carbon occupies but
a few moments additional, thus providing the simplest and best method
for obtaining the required values. The preliminary operations of drying
require but little attention from the assistant.
^ries, Journ. Am. Chem. Soc, 1909, 31, p. 272.
2See p. 243.
3Kellner, Landw. Jahrb., 1896, 47, p. 297.
4Higgins and Benedict, Am. Journ. Physiol, 1911, 28, p. 291.
294
A STUDY OF PROLONGED FASTING.
The total amount of carbon excreted in the urine ranged, according
to the values in table 42, from 5.82 grams on the first day of the fast
to 11.88 grams on the fourth day of the fast. In the first half of the
fasting period the carbon excretion averaged somewhat above 10 grams
per day, but in the latter part the excretion was not far from 8 grams.
Table 42. — Nitrogen, carbon, and energy of urine in experiment with L.
Date.
Day of
fast.
Nitrogen.
Carbon.
Carbon
per gram
of nitro-
Energy of urine.
Per gram of
Per gram
(C:N)
Total.
nitrogen.
of carbon.
(Cals. : N)
(Cals. : C)
1912.
gm.
gm.
gm.
cals.
cals.
cals.
Apr. 12-13
14.48
11.41
0.788
129
8.91
11.30
13-14
11.54
9.08
.787
104
9.01
11.45
14-15
1st
7.10
5.82
.820
65
9.15
11.17
15-16
2d
8.40
7.99
.951
89
10.60
11.14
16-17
3d
11.34
10.35
.913
118
10.40
11.40
17-18
4th....
11.87
11.88
1.001
134
11.29
11.28
18-19
5th
10.41
10.69
1.027
123
11.82
11.51
19-20
6th
10.18
10.42
1.024
116
11.40
11.13
20-21
7th....
9.79
9.06
.925
104
10.62
11.48
21-22
8th
10.27
10.30
1.003
116
11.30
11.26
22-23
9th ... .
10.74
10.92
1.017
124
11.54
11.36
23-24
10th
10.05
9.92
.987
111
11.04
11.19
24-25
11th....
10.25
9.59
.936
110
10.73
11.47
25-26
12th
10.13
9.05
.893
105
10.36
11.60
26-27
13th
10.35
10.15
.981
114
11.01
11.23
27-28
14th
10.43
9.95
.954
111
10.64
11.16
28-29
15th....
8.46
8.71
1.030
95
11.23
10.91
29-30
16th
9.58
11.39
1.189
123
12.84
10.80
Apr. 30-May 1 .
17th....
8.81
10.91
1.238
117
13.28
10.72
May 1-2
18th
8.27
9.65
1.167
104
12.58
10.78
2-3
19th
8.37
9.56
1.142
105
12.54
10.98
3- 4
20th....
7.69
8.07
1.049
91
11.83
11.28
4-5
21st
7.93
8.59
1.083
95
11.98
11.06
5- 6
22d
7.75
8.40
1.084
93
12.00
11.07
6- 7
23d
7.31
7.25
.992
88
12.04
12.14
7-8
24th....
8.15
8.68
1.065
95
11.66
10.94
8-9
25th....
7.81
8.68
1.099
91
11.65
10.61
9-10
26th
7.88
8.56
1.086
90
11.42
10.51
10-11
27th
8.07
8.23
1.020
90
11.15
10.94
11-12
28th
7.62
7.73
1.014
85
11.15
11.00
12-13
29th
7.54
7.94
1.053
87
11.54
10.96
13-14
30th....
7.83
7.95
1.015
87
11.11
10.94
14-15
31st
6.94
7.37
1.062
80
11.53
10.85
15-16
4.83
3.81
7.13
4.28
1.476
1.123
74
45
15.32
11.81
10.38
10.51
16-17
In the fasting experiments at Wesleyan University the carbon
content of the urine increased noticeably on a number of days as the
fast continued. This increase in carbon was accompanied by an
increase in the energy content, which was attributed at the time to the
presence of a large amount of /3-oxybutyric acid. A similar increment
in the carbon content of fasting urine was noted with an insane patient
by Benedict and Diefendorf.1 This excess of carbonaceous material
Benedict and Diefendorf, Am. Journ. Physiol., 1907, 18, p. 362.
URINE. 295
became apparent when the ratio between the carbon and nitrogen
of normal urine was compared with that obtained in these fasting
experiments.
Cakbon-Nitrogen Ratio.
Since the carbon content of the urine naturally fluctuates to a certain
degree with the nitrogen content, it is obvious that the determinations
of the carbon alone would not have the significance of the ratio between
the carbon and nitrogen, for we are interested not so much in the carbon
normally accompanying nitrogen in the urine as in the carbon other
than that in nitrogenous material. With an ordinary diet, essentially
constant ratios have been found for normal urines by various investi-
gators, averaging not far from 0.8 gram of carbon for each gram of
nitrogen excreted in the urine. Thus Benedict and Milner1 found an
average carbon-nitrogen ratio of 0.73 for 58 metabolism experiments
upon normal individuals with rest and work in the respiration calori-
meter at Wesleyan University. The variations in these experiments
were, in general, very small, the carbon-nitrogen ratio ranging from
0.67 to 0.89. Richardson,2 working with fewer urines, obtained a
carbon-nitrogen ratio varying from 0.74 to 1.01, with an average of 0.88.
Magnus- Alsleben3 concludes that with healthy individuals the carbon-
nitrogen ratio will not pass beyond the limits of 0.7 and 1.0, regardless
of diet. Loewy,4 Pregl,6 Reale,6 and others reported values with normal
individuals essentially within these limits.
The composition of the diet seems to have but little effect upon this
ratio. Benedict and Milner noted no appreciable variation due to
change in diet containing a preponderance of either carbohydrate or
fat, but Tangl7 found an average ratio of 0.96 on days when the diet
was rich in carbohydrates and poor in fat, which was considerably
higher than the average carbon-nitrogen ratio of 0.75 which was found
when the diet was poor in carbohydrate and rich in fat. Notwith-
standing the fact that the diet on these days was extraordinary, it may
be noted that Tangl's figures fell within the limits set by Magnus-
Alsleben.
Moderate muscular work has been shown both by Tangl and by
Benedict and Milner to increase the ratios little, if any, over those for
rest, although the severe muscular exercise incidental to the strenuous
work of a Marathon race was found by Higgins and Benedict8 to result
in distinctly abnormal carbon-nitrogen ratios in a number of cases, a
ratio as high as 1.517 being found in one instance.
Benedict and Milner, U. S. Dept. Agr., Office Expt. Sta. Bui. 175, 1907, p. 145.
2Richardson, Bulletin Mt. Hope Retreat Laboratory, 1900. Cited in Maly's Jahrsb. d. Tier-
Chemie, 1901, 31, p. 703.
3Magnus-Alsleben, Zeitschr. f. klin. Med., 1909, 68, p. 358.
4Loewy, Verhndl. der physiol. Gesellsch. zu Berlin, 1905-1906, p. 11.
6Pregl, Archiv f. d. ges. Physiol., 1899, 75, p. 87.
6Reale, Biochem. Zeitschr., 1912, 47, p. 355.
'Tangl, Archiv f. Anat. u. Physiol., 1899, Physiol. Abth. Supp., p. 251.
8Higgins and Benedict, Am. Journ. Physiol., 1911, 28, p. 291.
296 A STUDY OF PROLONGED FASTING.
With normal urines, therefore, one may conclude that the carbon-
nitrogen ratio may vary from 0.67 to 1.0. Many interesting cases
are recorded in which the disturbance of the carbon-nitrogen ratio has
been found. In the fasting experiments at Wesleyan University
carbon-nitrogen ratios during fasting ranged from a minimum of 0.660
to a maximum of 1.293. On 7 days out of 43 the ratio was over 1,
these 7 days being the last 4 days of a 7-day fast and the last 3 of a
4-day fast. In certain types of fever, also, Magnus-Alsleben1 found
an increase of the carbon-nitrogen ratio, while in others there was a
marked decrease. He also reports three cases in which extremely
high ratios were obtained after severe muscular work, these ratios
being 3.262, 1.926, and 1.038 respectively. It is thus clear that under
conditions which result in an abnormal katabolism, disturbances in
the carbon-nitrogen ratio are found, and conversely a disturbance of
this ratio may be taken as prima facie evidence of a distinctly disturbed
katabolism.
The carbon-nitrogen ratio has been computed for each day of the
fasting experiment with L. and likewise for the food days prior and sub-
sequent to the fast. These values are included in table 42, together
with the values for the total nitrogen. On the 2 days before the fast
the ratio was very constant, averaging 0.79. It then rose rapidly until
the fourth day, when it was slightly over 1.0 and remained at approxi-
mately 1 until the maximum level was reached between the sixteenth
and nineteenth days of fasting. The very high value of 1.476 on the
first day with food after the fast is in part explained by the excretion
of 4.44 grams of dextrose in the urine. It is seen from these ratios,
therefore, that the urine excretion after the first day or two regularly
contained some nitrogen-poor and carbon-rich substance which, from
all evidence, appears to be /3-oxybutyric acid. Since the 2 days with
food before the fast agree so perfectly, we have felt justified in using the
average value of 0.79 for computing the normal amount of carbon
accompanying nitrogen in the indirect computation of the amount of
jS-oxybutyric acid present in the urine. (See column B,table 38, page283.)
The values found with L. during fasting are materially higher
throughout the entire fast than those reported by Munk for Breithaupt,
for on 6 days of the fast Munk found no difference in the carbon-
nitrogen ratio between the fasting days and the 2 days with food
following the fast. On the last 6 days of a 3-weeks fast Grafe2 found
extraordinarily high carbon-nitrogen ratios as follows: 1.714, 1.642,
2.016, 1.873, 1.63, 1.53. On the first food day the ratio fell to 0.746.
The three observations of Pettenkofer and Voit3 may also be cited,
these investigators finding on the first fasting day an average of 0.7
as the carbon-nitrogen ratio.
1Magnus-Alsleben, Zeitschr. f. klin. Med., 1909, 68, p. 358.
2Grafe, Zeitschr. f. Physiol. Chem., 1910, 65, p. 21.
3Pettenkofer and Voit, Zeitschr. f. Biol., 1866, 2, p. 459.
URINE. 297
ENERGY OF URINE.
In studying the metabolism of a fasting man, although we are par-
ticularly interested in the energy transformed in the body and leaving
the body as heat, a complete picture of the total breaking-down of
tissue and loss of body material can not be had without a knowledge of
the potential energy of unoxidized material in the urine throughout
the fasting period. Determinations of the heat of combustion were
made by Mr. A. W. Cornell, the results given in table 42 being always
the average of two or three well-agreeing analyses.
During the fasting period the total amount of energy lost in the
urine ranged from 65 calories on the first day to 134 calories on the
fourth day. There was a general tendency after the fourth day for
the values to fall off gradually as the fast continued; excluding the first
day, the smallest amount (80 calories) was found on the last day of the
fast. The energy was also determined for the 2 days immediately
preceding the fast, the values being 129 and 104 calories respectively.
On the days with food following the fast, very small amounts of energy
were found, these being 74 calories on the first and 45 calories on the
second food day.
Calorie-Nitrogen Ratio.
Since the total amount of energy lost per day may vary with the
amount of nitrogen excreted, and since there will always be a certain
amount of potential energy normally accompanying each gram of
nitrogen in the urine, it is important to compute the number of calories
per gram of nitrogen. The results of such computation are also given
in table 42.
On the 2 days before the fast, there were about 9 calories daily per
gram of nitrogen. The ratio remained practically unchanged on the
first day of the fast, but for the next 4 days it showed a distinct ten-
dency to increase. The highest ratio found during the fasting period
was 13.28 calories per gram of nitrogen on the seventeenth day, and the
lowest was 9. 15 calories on the first day of the fast. It is thus seen that,
except on the first day, the ratios throughout the fast were compara-
tively high, exceeding those found for the preceding food days. The
ratios for the 2 days with food following the fast are influenced by the
fact that on the first day, when the diet contained an excessive amount
of carbohydrate, there was a measurable amount of sugar excreted
(4.44 grams), which would obviously increase the energy but have no
effect upon the nitrogen; furthermore, on the second day a very small
amount of nitrogen was excreted. Neither of these values can of course
be looked upon as obtained under normal conditions.
With normal subjects and an ordinary diet, the ratio of calories
to nitrogen is essentially constant. Thus Benedict and Milner1 found
Benedict and Milner, U. S. Dept. Agr., Office Exp. Sta. Bull. 175, 1907, p. 145.
298 A STUDY OF PROLONGED FASTING.
an average calorie-nitrogen ratio of 8.09 for 58 rest and work experi-
ments with normal individuals in the respiration calorimeter at Wes-
leyan University, the variations being extremely small, ranging from
7.3 to 8.94. When unbalanced diets are taken, this ratio may be some-
what altered. Tangl1 reports the ratio considerably higher on the
days when the diet was rich in carbohydrates and poor in fat, the ratio
becoming as high as 11.67 on the carbohydrate-rich days and falling
to 9.63 on the carbohydrate-poor days. On the other hand, Benedict
and Milner noted no appreciable change due to diet.
In the earlier fasting experiments the energy in the urine has been
rarely determined, the most extensive investigation being that in Wes-
leyan University;2 in one experiment a calorie-nitrogen ratio was found
ranging from 8.0 to 19.75.
The large ratios found during fasting experiments are unquestionably
to be explained by an excretion of nitrogen-poor, carbon-rich material,
which is chiefly /3-oxybutyric acid. Unfortunately, in the earlier
observation of Benedict and Diefendorf, in which the very high ratio
of 19.75 was found, direct evidence of the excretion of j8-oxybutyric
acid was not obtained, as the determinations were not then feasible.
Assuming that the high calorie-nitrogen ratio is due to the presence of
acetone bodies, it can be seen that during the fasting experiment of
the subject L. the highest acidosis as measured by this means occurred
between the fifteenth and twenty-fifth days, high ratios prevailing for
this entire period.
Calorie-Carbon Ratio.
While it is perfectly possible to have a carbon-rich, nitrogen-free
substance in the urine which would profoundly affect the calorie-
nitrogen ratio, the presence of carbonaceous material in all energy-
producing material found in urine would lead one to suppose that the
relationship between calories and carbon would be much more regular
than that between the calories and the nitrogen. The calorie-carbon
ratios have been computed for this experiment and are included in
table 42. The striking irregularities in the other ratios given in this
table are entirely absent in the calorie-carbon ratio, for they remain
remarkably constant under all conditions. The values for the entire
series, including both the first and second food periods, range only
between 12.14 on the twenty-third day of the fast and 10.38 on the first
day with food after the fast. In the fasting period itself the minimum
value was 10.51 on the twenty-sixth day. It is thus seen that during
the total 31 days of the fast there were, on the average, 11.12 calories
for each gram of carbon.
The average calorie-carbon ratio in these fasting urines, namely,
11.12 calories, is almost exactly the same as that found by Higgins and
^angl, Archiv f. Anat. u. Physiol., 1899, Physiol. Abth. Supp., p. 251.
2Benedict, Carnegie Inst. Wash. Pub. 77, 1907, p. 493, and Benedict and Diefendorf, Am
Journ. Physiol., 1907, 18, p. 362.
URINE. 299
Benedict,1 their average calorie-carbon ratio for 18 specimens of urine
excreted after severe muscular exercise incidental to a Marathon race
being 11.02. This is also very close to the average calorie-carbon ratio
obtained by Benedict and Milner, namely, 10.96. The striking uni-
formity in the ratio between calories and carbon again emphasizes the
importance of the development of some simple, rapid method of
determining the carbon in urine which will not require the employment
of a complicated bomb calorimeter.
Still another relationship may be studied by comparing the total
potential energy in the urine with the total estimated heat output.
This comparison, however, is made in another section of the report.
(See table 64, page 414.) The values given in this table show clearly
that the total amount of energy excreted in the urine by the fasting
subject L. equals approximately 8 to 10 per cent of the daily quota,
and hence may not be neglected in any consideration of the energy
lost by this man as the fast continued.
Wiggins and Benedict, Am. Journ. Physiol., 1911, 28, p. 291.
MICROSCOPY OF URINE AND TESTS FOR ALBUMIN.
By Harry W. Goodall, M. D.
The heat test was used in making the albumin determinations. For
the sake of uniformity the microscopic examination was made as fol-
lows : Two 15 mm. X 15 mm. cover-glass fields were examined with each
specimen, 20 minutes being given to searching for and counting casts.
The urine was centrifuged at a uniform rate for 5 minutes. The results
of the tests are given herewith.
DETAILED RESULTS.
April llf-15 (first day of fast). — Albumin, absent. Sediment, no casts,
blood, or pus; rare round cell, occasional squamous cell; little mucus.
April 15-16 (second day of fast). — Albumin, absent. Sediment, no casts
or pus; rare normal red blood corpuscle, rare small round cell,
occasional squamous cell; little mucus.
April 16-17 (third day of fast). — Albumin, absent. Sediment, one hyaline
cast, small diameter, no blood or pus; few small round cells, rare large
caudate cell, numerous squamous cells; rare spermatozoa, normal
in appearance; little mucus.
April 17-18 (fourth day of fast). — Albumin, absent. Sediment, no casts,
blood, or pus; few small round and squamous cells, little mucus.
April 18-19 (fifth day of fast). — Albumin, slightest possible trace. Sedi-
ment, 13 hyaline casts, 3 coarse granular casts, all of large diameter;
a few of the casts with cells adherent; small round cells more numer-
ous; few squamous cells.
April 19-20 (sixth day of fast). — Albumin, least possible trace. Sediment,
15 hyaline casts, 2 coarse granular casts, nearly all of large diameter,
some with cells adherent; few small round and squamous cells.
April 20-21 (seventh day of fast) . — Albumin, slightest possible trace. Sedi-
ment, 5 hyaline casts, 9 coarse granular cells, all of large diameter,
some with cells adherent; rare normal red blood globule, numerous
small round cells, few squamous cells; rare spermatozoa, normal in
appearance.
April 21-22 (eighth day of fast). — Albumin, slightest possible trace. Sedi-
ment, 4 hyaline casts, 5 coarse granular casts, nearly all of large
diameter, some with cells adherent; numerous medium and small
round cells; rare small caudate cell; little mucus.
April 22-23 (ninth day of fast). — Albumin, slightest possible trace. Sedi-
ment, 5 hyaline casts, three of which had a few cells and fat drops
adherent; 5 coarse granular casts with cells adherent, all casts of
large diameter; no blood or pus; few small and medium round and
squamous cells; rare spermatozoa, normal in appearance.
April 23-24 (tenth day of fast). — Albumin, slightest possible trace (albumin
cloud more marked than at previous examinations). Sediment,
8 hyaline casts, some with cells and fat drops adherent; 4 coarse
granular casts; all casts of large diameter; no blood or pus; few small
and medium round and squamous cells; little mucus.
April 24~25 (eleventh day of fast). — Albumin, slightest possible trace (same
as last examination). Sediment, 4 hyaline casts; 3 coarse granular
casts; general tendency to diminution in diameter of casts; adherent
cells and fat drops less numerous; few small and medium round and
squamous cells.
300
MICROSCOPY OF URINE AND TESTS FOR ALBUMIN. 301
April 25-26 (twelfth day of fast). — Albumin, least possible trace (same as
last examination) . Sediment, 7 hyaline casts, 6 coarse granular casts,
chiefly small diameter ; few fat drops and cells adherent ; few leucocytes
and small round cells; few acid sodium-urate crystals; little mucus.
April 26-27 (thirteenth day of fast). — Albumin, slightest possible trace
(reaction less marked). Sediment, 15 hyaline casts; 2 coarse gran-
ular casts of medium size and with a few cells adherent; few leuco-
cytes and small round cells.
April 27-28 (fourteenth day of fast). — Albumin, slightest possible trace
(same as last examination). Sediment, 6 hyaline casts, 2 coarse
granular casts, all casts of medium size and with a few cells adherent;
few leucocytes and small round cells; numerous spermatozoa.
April 28-29 (fifteenth day of fast). — Albumin, slightest possible trace (very
faint reaction). Sediment, 15 hyaline casts; 10 coarse granular casts,
few cells and fat drops adherent; general tendency to decrease in
diameter of casts; few small and medium round cells; rare small
caudate and neck-of-bladder cells.
April 29-30 (sixteenth day of fast). — Albumin, slightest possible trace
(same as last examination). Sediment, 14 hyaline casts; 12 coarse
granular casts, chiefly of small diameter; few small and medium
round cells; rare neck-of-bladder cells.
April 30-May 1 (seventeenth day of fast). — Albumin, slightest possible
trace (very faint reaction) . — Sediment, 7 hyaline casts, 2 fine granular
casts, some with few fat drops and cells adherent; casts of small
diameter; few small and medium round cells; rare neck-of-bladder
cells; few squamous cells.
May 1-2 (eighteenth day of fast). — Albumin, slightest possible trace (very
faint reaction). Sediment, 8 hyaline casts; 4 coarse granular casts,
small diameter, few fat drops and cells adherent; few small and
medium round cells; few squamous cells.
May 2-3 (nineteenth day of fast). — Albumin, slightest possible trace (very
faint reaction). Sediment, 8 hyaline casts; 4 coarse granular casts,
small diameter, few fat drops and cells adherent; few small and
medium round and squamous cells; little mucus.
May 3-4 (twentieth day of fast). — Albumin, slightest possible trace (very
faint reaction). Sediment, 6 hyaline casts; 2 fine granular casts, all
casts of small diameter, a few containing fat drops; few squamous
cells; little mucus.
May 4~5 (twenty-first day of fast). — Albumin, least possible trace (very faint
reaction). Sediment, 4 hyaline casts; 2 fine granular casts, of small
diameter; few small and medium round cells; few squamous cells.
May 5-6 (twenty-second day of fast). — Albumin, slightest possible trace
(very faint reaction). Sediment, 7 hyaline casts; 4 coarse granular
casts, all casts of small diameter; few small round and squamous cells.
May 6-7 (twenty-third day of fast). — Albumin, slightest possible trace
(very faint reaction). Sediment, 7 hyaline casts; 3 coarse granular
casts, all casts of small diameter; few small round and squamous cells.
May 7-8 (twenty-fourth day of fast). — Albumin, least possible trace (very
faint reaction). Sediment, 6 hyaline casts; 4 coarse granular casts, all
of small diameter; few small and medium round and squamous cells.
May 8-9 (twenty-fifth day of fast). — Albumin, slightest possible trace
(very faint reaction). Sediment, 6 hyaline casts; 2 coarse granular
casts; all casts of small diameter, occasional fat drops and cells
adherent; few small and medium round and squamous cells; some
cells slightly fatty; rare spermatozoa, normal in appearance.
302 A STUDY OF PROLONGED FASTING.
May 9-10 (twenty-sixth day of fast). — Albumin, slightest possible trace
(very faint reaction). Sediment, 8 hyaline casts; 1 coarse granular
cast; all casts of small diameter; occasional fat drops and cells
adherent; few small and medium round and squamous cells, some
cells slightly fatty; rare spermatozoa, normal in appearance.
May 10-11 (twenty-seventh day of fast). — Albumin, slightest possible trace
(very faint reaction). Sediment, 8 hyaline casts; 1 coarse granular
cast; all casts of small diameter, with a few fat drops and cells adher-
ent; rare spermatozoa, normal in appearance; few medium round
cells, slightly fatty; few squamous cells.
May 11-12 (twenty-eighth day of fast). — Albumin, slightest possible trace
(very faint reaction). Sediment, 6 hyaline casts; 2 coarse granular
casts; all casts of small diameter, a few fat drops and cells adherent;
numerous spermatozoa, normal in appearance.
May 12-13 (twenty-ninth day of fast). — Albumin, slightest possible trace
(very faint reaction). Sediment, 10 hyaline casts, 3 coarse granular
casts; all casts of small diameter, few fat drops and numerous cells
adherent; few small and medium round cells; few squamous cells;
few spermatozoa, normal in appearance.
May 18-14 (thirtieth day of fast). — Albumin, slightest possible trace (very
faint reaction). Sediment, 12 hyaline casts; 1 coarse granular cast;
all casts of small diameter, few fat drops and rather numerous cells
adherent; few small and medium round cells; few squamous cells;
few spermatozoa, normal in appearance.
May 14~15 (thirty-first day of fast). — Albumin, slightest possible trace
(distinctly more than last examination). Sediment, 36 hyaline casts;
2 coarse granular casts; casts of small and large diameter about equal
in number; few fat drops and epithelial cells adherent; few small
and medium round cells; numerous squamous cells.
May 15-16 (first day after breaking fast). — Albumin, slightest possible
trace. Sediment, 18 hyaline casts; chiefly of small diameter, with a
few fat drops and rare epithelial cells adherent; few small and medium
round cells; few squamous cells; very many spermatozoa, normal in
appearance; many acid sodium-urate crystals.
May 16-17 (second day after breaking fast). — Albumin, slightest possible
trace. Sediment, 2 hyaline casts of small diameter; few small and
medium round cells; few squamous cells; little mucus.
May 17-18 (third day after breaking fast). — Albumin, slightest possible
trace. Sediment, 2 hyaline casts; 1 epithelial cast of small diameter;
few small and medium round cells; few squamous cells; many sperm-
atozoa, normal in appearance; many calcium-oxalate crystals.
October 19 (five months after breaking fast). — Albumin, least possible trace.
Sediment, 2 hyaline casts of small diameter; few small and medium
round cells; occasional neck-of-bladder cells; few squamous cells.
SUMMARY.
The most remarkable change in character of the urine noted during
the fast was the appearance of albumin and casts on the fifth day, which
persisted throughout. A summary of the results is given in table 43.
During the first 24 hours that food was taken, the urine contained sugar1
and the sediment showed numerous calcium-oxalate crystals.
The sexual system of the subject remained active throughout the
entire period of observation. According to his own statements, a
seminal emission occurred two nights before the fast began, he had
^ee p. 292.
MICROSCOPY OF URINE AND TESTS FOR ALBUMIN.
303
Table 43. — Summary of results.
Day of fast.
Albumin.
Hyaline
casts.
Granular
casts.
Size of casts.
3d
0
1
0
13
15
5
4
5
8
4
7
15
6
15
14
7
8
8
6
4
7
7
6
6
8
8
6
10
12
36
18
2
2
2
0
0
3
2
9
5
5
4
3
6
2
2
10
12
2
4
4
2
2
4
3
4
2
1
1
2
3
1
2
0
0
1
0
Predominating casts large diam.
Do.
Do.
Do.
Do.
Do.
Size diminishing.
Do.
Do.
Do.
Small diameter.
Do.
Do.
Do.
Do.
Do.
Do.
Do.
Do.
Do.
Do.
Do.
Do.
Do.
Do.
Do.
Large and small about equal.
Small diameter.
Do.
Do.
4th
0
5th
SI. possible tr . . .
do
6th
7th
do
8th
do
9th
do
10th
do
11th
do
12th
do
13th
do
14th
do
15th
do
16th
...do
17th
do
18th
do
19th
do
20th
do
21st
do
22d
do
23d
do
24th
do
25th
do
26th
do
27th
do
28th
do
29th
do
30th
do
31st
. .. .do
do
do
do
do
voluptuous dreams without ejaculation on the night of the thirteenth
fasting day, and a seminal emission on the night of the fifteenth fasting
day. The urinary sediment contained rare spermatozoa on the third,
seventh, and ninth fasting days, very numerous spermatozoa on the
fifteenth fasting day, rare spermatozoa on the twenty-fourth, twenty-
fifth, and twenty-sixth fasting days, numerous spermatozoa on the
twenty-seventh, and a few on the twenty-eighth and twenty-ninth
days. On the first and third days after breaking the fast, during the
period of extreme mental disturbance, numerous spermatozoa were seen.
Microscopically the spermatozoa at all times appeared to be normal in
size and shape. There was no motility, but in every instance the urine
had been standing for some time before the examination was made.1
1These observations are fully in line with the observations made by Albitsky on fasting rabbits,
which exhibited sexual excitement after 28 to 30 days of fasting. (Cited by Pashutin, in a Course
of general and experimental pathology, St. Petersburg, 1902, 2, part 1 .) He also notes that sperma-
tozoa were frequently found in the urine of starving rabbits. Pashutin likewise cites Manassein,
who observed spermatozoa motile in the urine of rabbits even 33 hours after the animals had per-
ished from the want of food, thus indicating the intense persistency of sexual activity.
In striking contrast to these observations are those of Pojarkov (Compt. rend. Soc. Biol., 1913,
74, p. 141), who with two dogs which had fasted three months noted profound influence upon
the sexual activity. Indeed, the depression of the sexual activity was so great and so lasting that
Pojarkov even suggested that prolonged fasting may result in a bloodless castration. — F. G. B.
THE RESPIRATORY EXCHANGE.
Any study of the respiratory exchange in a living animal is of dual
value, for if properly conducted it throws light upon the character of
the katabolism and also supplies data for computing by the indirect
method the heat-production of the body, thus serving as a control
upon the direct calorimetric measurements of the heat-production.
In the fasting experiments at Wesleyan University, the measurements of
the respiratory exchange and of the heat-production were coincidental
throughout the entire period of the fast. This simultaneous deter-
mination had certain great advantages, particularly in establishing
the fundamental laws of metabolism obtaining in the early stages of
inanition. On the other hand, it gave very little opportunity for
securing evidence regarding the minimum or basal metabolism of the
fasting subject, since a prerequisite of a study of the basal metabolism
is a period of absolute muscular repose.
In these earlier fasting experiments, such a period of absolute mus-
cular repose was best secured during the night, when the subject was
in bed and supposedly sound asleep, i. e., quietly at rest. The subject
went to bed at 11 o'clock and as the experimental periods were all of
2 hours' duration, the period from 1 a. m. to 7 a. m. could reasonably
be taken as the time when the subject had a minimum activity. Unfor-
tunately no direct evidence regarding the degree of muscular activity
could be obtained. Furthermore, there was no evidence as to whether
the subject was asleep the entire time or more or less awake and some-
what restless.
It seemed advisable, therefore, in planning the study of the long fast
at the Nutrition Laboratory, so to arrange the experimental technique
and routine as to include a series of respiration experiments which
would throw definite light upon the character of the katabolism,
measured in both long and short periods, and to be able to isolate
certain periods in which the subject was perfectly quiet and under the
same conditions of muscular activity throughout the entire series of
fasting days. The respiratory exchange of the fasting subject was
therefore studied in two ways, i. e., by using the bed calorimeter
and the so-called "universal respiration apparatus." With the bed
calorimeter both the heat-production and the gaseous exchange could
be studied throughout the period that the subject remained in the
chamber. In the experiments at Wesleyan University the subject
remained in the chamber throughout the whole fasting period and the
respiratory exchange could be studied in 24-hour periods. In the pro-
longed fasting experiment at the Nutrition Laboratory, however, the
calorimeter periods usually began at 9 or 10 p. m. and continued until
8 a. m. the following day. Thus for 10 or 11 and but rarely for 9 con-
304
THE RESPIRATORY EXCHANGE. 305
secutive hours, the gaseous metabolism of the subject inside the
chamber could be studied. Graphic records of the degree of muscular
activity were also secured by a special form of bed.
The possible differences in the degree of sleep and the degree of
restlessness and the impossibility of determining the actual periods
of wakefulness and sleep made it seem undesirable to rely wholly upon
the determinations of the respiratory exchange made by this method
for a comparison of the metabolism as the fast progressed. Conse-
quently each morning, at the conclusion of the night experiment, the
respiratory exchange alone was studied for two or three experimental
periods by means of the universal respiration apparatus. With this
apparatus it was possible to obtain the gaseous metabolism for several
15-minute periods in which the subject lay perfectly quiet and awake,
thus giving material for comparison for each day of the fasting period.
The element of uncertainty as to the degree of muscular activity and the
degree of wakefulness or sleep was by this method entirely eliminated.
With both forms of apparatus the respiratory quotient could be accu-
rately determined, so that a mutual control was obtained on the accuracy
of the two methods of determining the respiratory exchange.
APPARATUS AND METHODS USED IN THE CALORIMETER EXPERIMENTS.
It is unnecessary to enter into the details of the construction and the
technique of using the respiration calorimeter or of the methods of cal-
culating the results obtained with it, as these have been fully explained
elsewhere.1 Since the publication of this description, however, a number
of minor though important changes have been made in the apparatus,
and hence it seems desirable to discuss them here somewhat at length, so
that the complete technique used in these experiments may be available.
Although the fundamental principle involved in the determination of the
respiratory exchange has not been altered in the slightest degree, our
accumulated experience enabled us to develop a technique to meet the
special conditions of the fasting experiment which not only facilitated
the manipulation of the apparatus but also gave greater accuracy.
ABSORPTION OF WATER- VAPOR AND CARBON DIOXIDE.
According to the usual arrangement of the absorbing system of
this apparatus, large porcelain vessels of special form, made by the
Royal Berlin Porcelain Works of Berlin, Germany, are used to hold the
sulphuric acid for absorbing the water from the air-current, and silver-
plated brass cans for holding the soda-lime for absorbing the carbon
dioxide. Both of these containers weigh considerable and when the
amount of water or carbon dioxide absorbed is 20 to 30 grams, the
balance on which they are weighed (which is accurate to about 0.05
gram) is sufficiently exact. On the other hand, when small amounts of
1Benedict and Carpenter, Carnegie Inst. Wash. Pub. 123, 1910.
306 A STUDY OF PROLONGED FASTING.
water or carbon dioxide are to be determined, as for example 10 grams
or less, these vessels are too large.
In the earlier part of the fasting experiment, the calorimeter experi-
ments were usually subdivided into three periods, so that the amount of
carbon dioxide to be weighed represented that produced in about 3
hours, or approximately 45 to 60 grams. Toward the end of the fasting
period it seemed desirable to obtain more definite information regarding
the progress of the metabolism throughout the night and hence an
attempt was made to secure shorter periods. On one night (May 4-5)
the periods were but 1 hour long; under these circumstances only about
11 grams of carbon dioxide were absorbed in each period. It was
necessary, therefore, to have some form of absorbing vessel which would
weigh considerably less than the usual containers, as the error in weigh-
ing might make a measurable difference in the results. We accordingly
replaced the large vessels with soda-lime bottles and glass sulphuric-
acid containers, i. e., "Williams bottles," similar to those used in the
absorbing circuit of the respiration apparatus, and from that time
divided the calorimeter experiments into relatively short periods
throughout the night, weighing the water and carbon dioxide in these
smaller and more accurately weighed containers. A complete descrip-
tion of the glass soda-lime bottle and the Williams bottle is given else-
where.1 Soda-lime was used as the absorbent for carbon dioxide
throughout the whole fasting experiment, for although at the time the
description of the respiration calorimeter was published experiments were
being made with potash-lime, subsequent experience has convinced us
that, as yet, the original form of soda-lime has not been improved upon.
ANALYSIS OF CHAMBER AIR AT THE END OF PERIODS.
While it was desirable to obtain as short and as many periods as
possible in the night calorimeter experiments, it naturally became diffi-
cult to arrange the routine so as to secure the largest number of periods
without decreasing the accuracy and overtaxing the strength of the
assistants, especially as a continuous metabolism experiment of 35 to
40 days was quite outside of our experience. With three periods in
each experiment, as was at first planned, it was possible to arrange the
program so that a trained observer with two responsible assistants
could readily carry out the routine of a calorimeter experiment. One
of the difficult parts of the program was to make provision for the
analysis of the air residual in the chamber at the end of each period.
It had been our custom to do this by deflecting a certain volume of the
air from the outgoing air-current through a series of U-tubes containing
soda-lime, and pumice stone and sulphuric acid respectively, which
absorbed the carbon dioxide and water-vapor from the air. The air
Benedict, Deutsch. Archiv f. klin. Med., 1912, 107, p. 166. See, also, figure 39, p. 316, of
this publication, for a diagrammatic representation of these bottles.
THE RESPIRATORY EXCHANGE. 307
was then passed through a calibrated Bohr gas-meter and returned to
the system. By this method the amount of carbon dioxide and water-
vapor in the air residual in the chamber at the end of the period could
be readily computed. This manipulation of U -tubes, reading of gas-
meter, barometer, and temperature, all of which was checked by an
independent observer according to our usual procedure, made an added
complex at the end of each experimental period, which of itself formed a
considerable part of the routine. Some simpler method for obtaining
data regarding the composition of the air inside the chamber was there-
fore sought.
Since the publication of the description of the calorimeter, it has been
our good fortune to become thoroughly familiar with the very ingenious
and accurate gas-analysis apparatus of Sonden of Stockholm, which
was used in a research on the composition of outdoor air carried out at
the Nutrition Laboratory.1 This apparatus was employed in the fast-
ing experiment for determining the carbon-dioxide content of the air
in the calorimeter chamber. By means of a system of previously
dried gas-sampling tubes, samples of the air could readily be taken in
a few seconds at the end of each experimental period. The carbon
dioxide in these samples of air could then be determined on the Sonden
apparatus the next morning by a skilled assistant.
Unfortunately, while ideal determinations of the carbon dioxide
could be obtained by this method, it is no less important to know the
volume of water- vapor residual in the chamber at the end of the period,
and the accurate determination of this factor has been one of our most
perplexing problems. With the U-tube system previously described
very satisfactory results could be obtained, but with the substitution
of the gas sampling and the subsequent determination of the carbon
dioxide on a Sonden apparatus, it was necessary to find some method
of determining the exact amounts of the water-vapor in the air. It is
obviously impossible to take a sample of air, even over mercury, and
retain the water-content for any length of time, as it would either be
deposited upon the glass tube or so affected that no accurate analysis
could be made. Furthermore, for analysis by the ordinary gravimetric
method a sufficient volume of air could not be so stored. We therefore
began experimenting with a delicate wet- and dry-bulb psychrometer.
Two mercury thermometers, graduated in 0.01° C, were placed in
the air-current, one of the thermometers having a moistened strip of
linen wrapped around the bulb. The depression of the wet bulb was care-
fully noted and by comparing the observations thus obtained with
results secured by the U-tube method and particularly with the unique
and extremely accurate hygrometer of Sonden,2 it was found that
Benedict, Carnegie Inst. Wash. Pub. 166, 1912, p. 75.
2Sonden, Bihang till K. Svenska Vet.-Akad. Handlingar, 1891, 17, p. 3; see also Meteorolo-
gische Zeitschr., 1892, p. 81.
308 A STUDY OF PROLONGED FASTING.
concordant results could be obtained. Hence, to secure records of
the water-vapor in the residual air, it was only necessary to place this
psychrometer in the air-current inside the respiration chamber. It
was so arranged that the air leaving the chamber came through a pipe
opening at the rear and extending along the bottom to the front of
the chamber near the glass window. The air passed over the dry
bulb of the psychrometer and from there over the wet bulb, and then
directly to the blower outside. It was therefore possible for the obser-
ver on the outside to read both mercury thermometers through the
glass front of the calorimeter chamber; readings to 0.01° C. could
ordinarily be relied upon. By means of psychrometric tables, the
amount of water-vapor residual in the chamber could be very readily
computed. It is thus apparent that by reading the psychrometer and
taking a single sample of the air in the chamber and subsequently ana-
lyzing it, it was possible to obtain information regarding the content
of water-vapor and carbon dioxide in the chamber air at the end of each
period with a minimum utilization of the assistant's time during the
night.
The psychrometer is at present used in this laboratory for short
experiments with babies and small animals in which respiration
chambers are employed. Having tested this method of determining
the water-vapor by two other methods, we felt justified in employing
it in the long calorimeter experiments, especially as it requires only
the reading of the two thermometers by the assistant at the end of
each period. To use the psychrometer successfully, it is necessary
that the air should pass rapidly over the bulbs of the thermometers.
Care should also be taken that the fabric about the wet-bulb ther-
mometer is kept moist, as in long experiments of 10 or 12 hours it
occasionally becomes dry, so that false readings are obtained.
It is obvious that no one of these methods, i. e., the sulphuric acid-
pumice stone, the Sond6n hygrometer, or the psychrometer, gives the
true value for the water-vapor inside the chamber, as they measure
only the water- vapor in the outgoing air, and there is certainly an area
about the ingoing air pipe (where the air is entering absolutely free
from water-vapor) which is of a much lower water-content, notwith-
standing the fact that the air is fairly well circulated inside the chamber
by means of an electric fan. It should be considered, however, that
the values obtained at the beginning and end of each period are for
comparison only and we deal here with differences rather than with
absolute amounts.
This change in methods was particularly advantageous for the deter-
mination of the residual carbon dioxide. When U -tubes are used, it
is necessary to pass 10 liters of air through them in order to secure a
weighable amount of carbon dioxide. For doing this in a relatively
short time, such as 3 minutes, a ventilation through the U -tubes of
THE RESPIRATORY EXCHANGE. 309
about 3? liters per minute is required and the use of fairly large U-tubes,
each weighing about 80 to 90 grams. A combination of one soda-lime
U-tube followed by a pumice-stone sulphuric acid U-tube weighs not
far from 160 to 180 or even 190 grams, while the amount of carbon
dioxide to be weighed is sometimes no more than 60 milligrams. When
experimenting with a man awake or doing severe muscular work, the
method is perfectly satisfactory, but when experimenting with a man as
emaciated as our fasting subject, with a minimum metabolism, and
producing only a small amount of carbon dioxide per hour, it is obvious
that the residual carbon dioxide in the chamber would be low and it
would be difficult to obtain very accurate determinations under these
conditions. With the Sonden gas-analysis apparatus, on the contrary,
it was possible to determine 0.5 per cent of carbon dioxide to the third
significant figure with great accuracy. The advantage of thus obtain-
ing a more delicate determination of the carbon dioxide and at the same
time decreasing the work required of an assistant during the long night
period made it desirable to introduce the method by which the sam-
pling pipette with subsequent analyses could be used to determine the
carbon dioxide and the wet- and dry-bulb psychrometer for determining
the water- vapor. The sampling tubes were collected each morning after
the night experiment was over, and the analyses made on the Sonden
apparatus by Miss Alice Johnson, whose technical skill in the use of
this apparatus was well attested in the research on the composition of
outdoor air previously referred to.
TENSION EQUALIZER.
The rubber bathing cap used as a tension equalizer in the earlier
form of respiration calorimeter has been replaced by a spirometer which
was first designed for the universal respiration apparatus. A brief
description of this spirometer with diagram (figure 40) is given on
page 318, but a more detailed description may be found in an earlier
publication.1
The spirometer is attached directly to the side of the respiration
chamber and thereby becomes a part of the chamber volume, thus
providing for fluctuations in the volume of the air inside the apparatus.
As the carbon dioxide is absorbed by the soda-lime and the oxygen is
used by the man inside the calorimeter, the total volume of the air
inside the chamber gradually decreases. Accordingly the spirometer
bell slowly falls until a certain point is reached where an electric contact
(not shown in figure 40) is made and oxygen thereby automatically
admitted by means of an electric valve attached to the oxygen cylinder.
When sufficient oxygen is admitted to raise the bell and break the
contact, the flow of oxygen is automatically stopped; in this way the
supply of oxygen is under continuous control.
Benedict, Deutsch. Archiv f. klin. Med., 1912, 107, p. 172. See also Benedict and Talbot.
Carnegie Inst. Wash. Pub. 201, 1914, p. 43.
310 A STUDY IN PROLONGED FASTING.
In the fasting experiment, use was made of this spirometer to indi-
cate the constancy of conditions. For example, if the oxygen supply
was completely shut off for 5 or 6 minutes before the probable end of
an experimental period and the pointer on the spirometer bell was
allowed to rest against the smoked-paper drum, a regular rising curve
would be drawn on the rotating drum, thus indicating the slow, steady
fall of the spirometer bell. If the subject made a muscular movement,
as in turning over, or for any reason there was an irregularity in the
curve, it was obvious that there was a sudden expansion or contraction
of the air in the chamber which could not be corrected for by the meas-
urement of the temperature and barometer. Consequently, if such an
irregularity in the line occurred during the last 5 minutes of the experi-
mental period, the length of the period was extended until a regular
curve could be secured. This was most helpful in many instances.
Specimen curves are given in figure 36 showing this method of utilizing
the spirometer. It is of course necessary to note the exact height of the
spirometer at the moment the period is ended. This is done by read-
ing the position of the pointer attached to the counterpoise of the spi-
rometer as it travels over a millimeter scale.
MAY 1 1. 1912
8.47 A.M.
Fig. 36. — Specimen records of change in volume of the spirometer on the bed
calorimeter during last 5 minutes of periods in experiment with L.
ARGON IN OXYGEN FROM LIQUID AIR.
In recent years we have used the nearly pure oxygen obtained from
liquid air by the Linde Air Products Company. At first we were
unaware of the fact that the residual gas was not, as commonly con-
sidered, all nitrogen with an atomic weight of 14, but consisted in large
part of argon with an atomic weight of 40. Hence it has been necessary
to emphasize the fact that in computing either the volume of oxygen
admitted from a cylinder or in calibrating a gas-meter by the method of
weighing the oxygen,1 the composition of this residual gas should be
taken into consideration, as otherwise an appreciable error in the per-
centage of oxygen may easily occur. Thus, in a series of observations
carried out on diabetics2 and likewise another series on muscular work,3
Benedict, Physical Review, 1906, 22, p. 294.
2Benedict and Joslin, Deutsch. Archiv f. klin. Med., 1913, 111, p 350.
'Benedict and Cathcart, Carnegie Inst. Wash. Pub. 187, 1913, p. 74.
THE RESPIRATOEY EXCHANGE. 311
it was found that this correction for argon in place of nitrogen altered
the oxygen consumption about 1 per cent and consequently altered the
values for the respiratory quotient by a like amount.
GRAPHIC REGISTRATION OF DEGREE OF MUSCULAR REPOSE OF SUBJECT
INSIDE THE RESPIRATION CALORIMETER.
The intimate relationship exhibited between the degree of muscular
repose and the total metabolism compelled us to make sure, at the
beginning of the fasting experiment, that the measurements of the
metabolism made with this subject from time to time were comparable
so far as muscular repose was concerned. Great care was taken to
secure experimental periods when the subject was perfectly quiet and
awake, and likewise when he was asleep. All this care would have
been of no avail, however, if we had not been able to secure periods of
like muscular repose or activity. If, for example, the subject had been
noticeably restless on the first few nights of the period of prolonged
fasting and on the last nights was especially quiet, the decrease in
the metabolism could not be shown to have been due to the influence of
the fast, but might have been due to the differences in the degree of
muscular repose. While statements could be secured from the subject
or from observers as to how well the subject slept or how quietly he lay,
no reliance could be placed upon them, as our experience has been that
such observations are usually untrustworthy. Hence we made use of
a method of graphic registration, which succeeded the ocular observa-
tions of the muscular activity used in connection with the experiments
in Wesleyan University, and likewise had its own development later
in this laboratory. In the earlier experiments we placed about the
body of the subject either one or two tube pneumographs in such a
position that not only was the respiration-rate recorded, but likewise
any muscular movement of the body.1 When the subject was lying
down, it was found that these pneumographs became irksome if worn
for several hours. In the experiments with diabetics,2 the use of the
pneumograph was found to be satisfactory, as the apparatus was
rarely worn continuously for more than 3 hours. In the fasting experi-
ment, however, it would be necessary for the subject to wear the pneu-
mograph for some 12 or 13 hours each night, and as the degree of emacia-
tion became greater it was quite possible that the discomfort might
be such as to disturb his sleep if not, indeed, cause pain; also, that the
subject might turn over during the night and cramp the transmission
tube in such a way as to prevent proper registration.
Previous experiences in this laboratory with a suspended cage or
crib for dogs or infants led us to apply the same principle for devising
a special form of bed for use in experiments with the universal respira-
^enedict, Carnegie Inst. Wash. Pub. 77, 1907, p. 10.
2Benedict and Joslin, Carnegie Inst. Wash. Pub. 136, 1910, p. 22.
312
A STUDY OP PROLONGED FASTING.
tion apparatus in which the subject lies upon a couch. This bed was
so suspended that the slightest change in the center of gravity of the
body, such as moving the hand or the foot, would alter the tension on
the spring inside a pneumograph and thus transmit the movement to
a tambour and kymograph. By this means the least muscular activity
would be recorded. With the suspended crib used in experiments with
infants, experience has shown that the best point of support was at the
Fig. 37. — Method for obtaining graphic record of activity in bed calorimeter.
The subject lies on the bed on the framework inside the calorimeter. One side of the frame
rests on a knife-edge, K ; the other side is supported by two stout spiral springs, S and S'. Any
change in the tension on the springs likewise affects the tension on the pneumograph, P, thus
altering the tension of the air in the pneumograph. By means of a rubber tube and a metal pipe
passing through the copper wall, C, zinc wall, Z, and asbestos wall, A, of the calorimeter, the lower
end of the pneumograph communicates with a tambour which writes on the kymograph placed
above the calorimeter. Any lateral change in the center of gravity of the body instantly produces
a movement of the pointer on the kymograph.
THE RESPIRATORY EXCHANGE. 313
foot of the crib, the spring being placed at the head. With adults, how-
ever, we soon found that the major movements were lateral rather than
lengthwise of the body and the supports and springs were accordingly
placed at the sides of the bed instead of at the head and foot.
This bed, when used in the calorimeter chamber, was supported at
one side on two frictionless steel points and at the other by two stout
spiral springs which could be adjusted by turnbuckles to bring the bed
to a level position. Obviously any change in the center of gravity of
the body altered the tension upon the two supporting springs, which
were therefore elongated or shortened. When the pneumograph was
attached to the bed, the same force producing the elongation or
contraction of the springs affected the pneumograph. The change
in the tension of the air inside the pneumograph was transmitted by the
usual method, i. e., by means of a metal tube passing through the walls
of the chamber and subsequently by a rubber tube connecting with the
tambour, writing-point, and kymograph. The method of obtaining
this graphic registration of the muscular activity is shown in figure 37.
In this figure the open end of the bed calorimeter is shown in per-
spective and in a somewhat schematic way. C, Z, and A represent
respectively the inner copper wall of the chamber, the zinc middle wall,
and the outer asbestos wall. The framework of the bed is seen at the
bottom of the calorimeter chamber with the left-hand edge resting
on the steel support, K. The two spiral springs, S and S', each pro-
vided with a turnbuckle, are attached at the upper end to the wall of the
calorimeter chamber and at the lower end to the right-hand edge of the
bed. Midway between the springs is attached a pneumograph, P, the
upper end of which is attached to the wall of the calorimeter chamber.
The subject, lying upon an air mattress which is in turn resting upon
a long plate of galvanized iron, is slid on to the bed framework feet first.
As the weight of the body falls upon the framework, the springs, S and
S', become extended, the adjustment necessary to secure perfect level-
ing of the bed being made by means of the turnbuckles. If the subject
turns during the night, a greater tension is put upon the springs, S
and S', and the pneumograph, P, is elongated.1 The air in the tube
connecting the pneumograph with the outside of the chamber is thus
somewhat rarefied and the tambour pointer sinks, thus producing a
depression in the line on the kymograph drum.
*For the benefit of other workers in this field, it is of interest to record here the recent experi-
ence of Dr. Paul Roth, of Battle Creek, Michigan. In recording the body-movements of men or
women lying on beds, he replaced the pneumograph with a small Politzer bulb, so adjusted
as to be somewhat compressed by the bed-frame. The bulb was connected to the tambour and
kymograph. Preliminary tests made in the Nutrition Laboratory with the Politzer bulb arrange-
ment have shown that the results of the variation in pressure on the bulb by variation in muscular
activity are most satisfactory, not only with adults but also with small animals — a fact of special
interest in connection with the research on infants. Two serious objections to the pneumograph,
i. e., the danger of leaks through the rubber and the difficulty of renewing the rubber, are thus
obviated by the use of this bulb. A flexible rubber bulb of small size is best used.
314
A STUDY OF PROLONGED FASTING.
I. APRIL 10-11. 1912
t
M*
1*0* A.M.
aS
TMS A.M.
4J*
«.!•
ill
M.OC A.M.
U» ' -
l.»» * ...
H*» AJ*
ha
•!.•» KH.
I. APRIL
14-15.
TlLM P.M.
1912
7*9
,
1 4*0 AH
6.IS —
KB
4J0
rs.00 aj«.
S3i
*—"
£«4
LSI
T&M AM
iijR
— '
ILM AJ4.
11.13
IO20
-TM7AK HI. APRIL 29-30. 1912
SA7
Ul
t&OOAM.
~~ *
xso t — ■
f
U>« " '
1.10
11.17 AM. '
IU»
IOJ5
r —
EZ. MAY 13-14. 1912
T7.3I AM.
J~
127
rJ*.
•XT
tZ/JI AJ*.
fx.
ILS2
•uUr
+- *"
Fig. 38. — Specimen pneumograph records of movements of bed calorimeter lever mattress
support in night experiments with L.
THE RESPIRATORY EXCHANGE. 315
The apparatus is extremely sensitive and shows plainly such minor
muscular motions as movements of the hand or arm to one side or,
indeed, the twisting of the feet. Throughout his whole stay in the
laboratory, L. slept on this bed inside the respiration chamber each
night and a kymograph record was therefore obtained for the whole
period. The drum of the kymograph was usually rotated at such a speed
as to give one revolution of 500 mm. per hour.
Although these kymograph records were not secured primarily for
publication, four typical records have been arbitrarily selected for
reproduction in figure 38. The curve for April 10-11, 1912 (Curve I),
which represents the record obtained on the first night which the sub-
ject spent inside the chamber, begins at 11 p. m., April 10, and ends at
8h 02m a. m., April 11. For the most part the subject was remarkably
quiet, showing no considerable degree of restlessness until about 5 a. m.
on April 11.
The curve for April 14-15, 1912 (Curve II) represents the record
obtained on the first night of the fasting period. The line here is
remarkably regular, showing relatively few major muscular movements.
When there is a distinct change in the level of the mark made by the
pointer, as is seen at approximately 5h 12m a. m., this is an indication
that the subject changed his position, probably turning on his side.
Another curve (Curve III) has been selected from those obtained
about the middle of the fasting period, which represents the record for
April 29-30, 1912. This also shows a general regularity of line, with
occasional indications of changes in position.
The last curve (Curve IV) was obtained on the next to the last night
of the fasting period, May 13-14, 1912. Between 12h 45m a. m. and
lh 45 m a. m. there was considerable movement for a few moments at
two or three different times, but otherwise the record has much the
same characteristics as the other curves shown.
From these curves and also from other curves which it is imprac-
ticable to reproduce here, we may logically infer that this subject
was particularly quiet inside the respiration chamber. While he did
not sleep the entire time, yet the kymograph records show that he
was for the most part very comfortable inside the chamber — indeed,
he repeatedly made the statement that he was very comfortable
throughout the night. It is of interest to note that at the hospital
he said that the bed there was not so comfortable as the bed inside the
calorimeter chamber.1
METHODS USED IN EXPERIMENTS WITH THE RESPIRATION APPARATUS.
The universal respiration apparatus, which was used for the respira-
tion experiments throughout the day, is based upon the same principle
as the large respiration calorimeters. A great variety of experiments
xSee page 51.
316
A STUDY OF PROLONGED FASTING.
have been made with it in this laboratory, and its accuracy has been
thoroughly tested. Not only men and women have been used as sub-
jects, but, by adding a small chamber, experiments have also been
made with infants and small animals. The apparatus has been
described in detail elsewhere.1
With this apparatus the subject lay quietly on the same bed upon
which he slept during the night, the bed being withdrawn from the
respiration chamber and placed upon a small framework in the calo-
rimeter laboratory. He was covered with bed clothing and two soft-
rubber nose-pieces were inserted in the nostrils, the subject being cau-
tioned to keep his mouth closed. After he had breathed a few minutes
through a two-way valve opening into the room, the valve was turned
and he began to breathe into a closed volume of air — some 8 or 10
liters — which was kept in motion by a ventilator or blower. As the
air left the nostrils of the man, it was carried by the blower to suitable
Fig. 39. — Schematic outline of universal respiration apparatus.
The subject, lying upon a couch or bed, breathes either through the two nose-pieces or a mouth-
piece into a ventilating current of air, kept in motion by a rotary pump. The moisture in the
air is absorbed in two glass vessels containing sulphuric acid, an empty glass vessel serving as a
trap to prevent accidental back suction of acid. The dried air then passes through soda-lime and
again through sulphuric acid in a special form of bottle and finally through a can containing
sodium bicarbonate to free the air of any traces of acid vapor. Oxygen is introduced as desired.
The air is then ready to be inhaled by the lungs. As the air leaves the lungs, the changes in the
volume of the confined air are recorded on the spirometer, which moves freely up and down with
each inspiration and expiration. The change in weight of the soda-lime vessel and its accompany-
ing sulphuric-acid bottle gives the weight of the carbon dioxide produced. The weight of the
oxygen is obtained either by noting the loss in weight of the cylinder of the gas or measuring the
gas carefully admitted through a meter.
Benedict, Deutsch. Archiv f. klin. Med., 1912, 107, p. 156.
THE RESPIRATORY EXCHANGE. 317
containers in which the water and carbon dioxide were absorbed;
oxygen was next added from a cylinder of weighed gas or through a
calibrated meter to replace that used by the man; the air was then
returned to the subject. The amount of carbon dioxide excreted was
obtained from the changes in weight of the absorbers and the amount
of oxygen consumed from the record of the oxygen admitted to the air-
current. Experiments could be made with this apparatus with periods
as short as 15 minutes. The general scheme of the respiration appa-
ratus is shown in figure 39.
In this apparatus provision has been made for fluctuations in the
volume by attaching a tension equalizer. In the earlier forms of the
respiration apparatus, a rubber bathing cap was used as a tension
equalizer, but more recently this has been replaced by a spirometer.
This spirometer not only provides for the fluctuations in the volume of
air, but has been utilized for recording the character of the respiration,
as has already been noted in a previous section of this publication.
(See page 158.) It has also been used with the bed calorimeter for indi-
cating the constancy of conditions. (See page 310.) The details of
the spirometer are shown in figure 40.
With each inspiration and expiration, the thin copper bell, c, of the
spirometer falls and rises in the annular space between the copper walls,
a and b, this space being filled with water. To the counterweight rod,
g, g, g, is attached a pointer, h, which writes on the smoked-paper
surface of the kymograph drum. A wheel, r, with a milled edge, is
rotated by each upward movement of the cord, t, which rests in a groove
in the edge of the wheel, the pawl, u, preventing any backward move-
ment of the wheel. By means of a platinum contact on the periphery
of r, each complete revolution of the wheel may be recorded. An expla-
nation of the use made of the records obtained with this spirometer
and a series of typical kymograph curves are found on pages 158 to 160.
The respiration apparatus was used regularly each morning of the
fast for an experiment immediately following the night's sojourn in the
bed calorimeter. The subject, lying upon his bed, was transferred
directly from the calorimeter chamber to the respiration apparatus.
He then turned upon his side and urinated without rising and the
respiration experiment was begun shortly afterwards. In the latter
part of the fast, the apparatus was used for a respiration experiment
each evening about an hour before the subject entered the calorimeter.
At irregular intervals throughout the fast the respiratory exchange was
also studied with the respiration apparatus, while the subject was
sitting quietly in a chair or writing steadily, as he did much of the time.
The apparatus was likewise used for experiments in which the subject
breathed an oxygen-rich atmosphere while lying upon the couch.
All of these experiments included two or three periods of approxi-
mately 15 minutes each, in which records of the degree of muscular
318
A STUDY OF PROLONGED FASTING.
Air leaving the lungs enters the pipe, m n,
and passes into the spirometer bell, c. This
bell moves up and down in the annular space
filled with water between the inner copper
shell, a, and the outer copper shell, b. The
bell is counterpoised by a rod, g g g, and a
small supplementary weight, I, attached to a
cord, t, the counterpoise being «upported by
an aluminium wheel, e, over which the thread
d passes. The height of the spirometer can
be read by the pointer, h, on a millimeter
scale, p, or the pointer h can be made to
write directly upon a kymograph drum. A
work adder wheel, r, is so attached that the
cord t passes through a groove in its peri-
phery, and the wheel conse-
quently rotates with each down-
ward motion of the bell c. The
pawl, m, prevents back motion.
A projection, w, in the periphery
of the wheel makes an electric
contact and permits graphic re-
cord of each complete revolu-
tion of the work adder wheel.
For counting the respirations
a new attachment, s, has been
employed. A steel wire is loosely
fastened around the hub of the
wheel, r. By means of a light
hair spring, lateral tension is
brought against
the movable rod, g.
As the rod ascends,
the steel wire ro-
tates out of the
mercury cup,
breaking the elec-
tric contact; as the
rod moves down,
the contact is
made. A small
stop above the
mercury cup pre-
vents the wire from
rising too far.
Fig. 40. —Spirometer for studying thejnechanics of ventilation.
THE RESPIRATORY EXCHANGE. 319
repose were obtained by means of the special form of suspended bed,
and the pulse-rate was regularly observed.1 The subject L. adjusted
himself very readily to this apparatus, finding it not at all uncomfort-
able. Indeed, on one or two occasions he expressed his enjoyment of
the soothing sensation produced by the slight sound of the blower.
He seemed to be in no wise affected by the apparatus and to him appar-
ently the respiration experiment was but a slight incident in the day's
program. These experiments were all personally supervised by Mr.
Carpenter, who found this subject more nearly ideal than any subject
he had ever studied, as the man could be relied upon to keep absolutely
quiet throughout the whole period. The record of the degree of mus-
cular repose also shows this fact,2 and it is especially advantageous to
have this assurance that, as the fast progressed, whatever disturbance
in the total metabolism is observed as a result of varying body position
or the inhalation of oxygen-rich atmospheres, it certainly was not
complicated by extraneous muscular activity.
Aside from the value of being able to study the respiratory exchange
under the different conditions of waking, sitting, and writing, the
universal respiration apparatus offered special advantages for likewise
studying the mechanics of respiration, including the respiration-rate,
the character and volume of each respiration, and the ventilation of
the lungs per minute, and for indicating in general any abnormality in
the mechanics of respiration. It was also possible to determine the
alveolar air of the subject in these respiration experiments.3
As both the morning and evening series of respiration experiments
were made under the same conditions of muscular repose and with the
subject awake, the results obtained are perfectly comparable, so that
they give excellent data for drawing sharp conclusions as to the influ-
ence of prolonged fasting upon the general metabolism. Furthermore,
since two wholly independent series of respiration experiments were
obtained with different apparatus and at a different time of day, the
individual periods of both the experiments with the bed calorimeter
and the respiration apparatus are made doubly valuable by this check.
It is important to bear in mind, however, that the experiments with the
universal respiration apparatus gave no evidence regarding either the
water-vapor exhaled from the lungs and skin or the cutaneous respiration.
As Magnus-Levy has pointed out,4 there is a greater amount of carbon
dioxide excreted through the skin than of oxygen absorbed, so that
there is a tendency for the respiratory quotient to be affected by about
0.01. When, therefore, a respiratory quotient of 0.73 is obtained with
a subject in the respiration chamber of the calorimeter, a respiratory
^ee pp. 311, 99, and 110.
2See figure 22 on page 159.
3See discussion of these results on pages 168 to 181.
4Magnus-Levy, von Noorden's Handbuch der Pathologie des Stoffwechsels, 1896, 1, p. 218.
320 A STUDY OF PROLONGED FASTING.
quotient of but 0.72 would, under like conditions, be obtained with the
respiration apparatus. On the other hand, the experiments made
with the universal respiration apparatus are extremely helpful as a
general index of the respiratory exchange from which the calorimetry
can be computed by the indirect method. We considered it of im-
portance to make a special effort to secure experiments of short
duration, as the technique of the experiments made by Luciani on Succi
have been adversely criticized by Zuntz and the experiments made by
Zuntz and his co-workers on the fasters Breithaupt and Cetti in Berlin
were certainly complicated by colic and a head cold. Moreover, it is
not unreasonable to suppose that the technique in thirty years has
been materially improved.
STUDIES WITH THE BED CALORIMETER.
ATMOSPHERIC CONDITIONS INSIDE THE CHAMBER.
Prior to a consideration of the results of the study of the gaseous
exchange inside the bed calorimeter, it is advisable to note the exact
conditions of ventilation, temperature, and particularly relative
humidity under which this subject was living in the chamber. The
respiration calorimeter was ventilated at a rate of approximately 40
liters per minute, or roughly speaking, 2,400 liters per hour. Since
the volume of the chamber was not far from 800 liters, theoretically
the air would be replaced inside the chamber three times each hour.
The cooling arrangement prevented an abnormal rise in the tempera-
ture, and a study of the relative humidity under these conditions pre-
sents certain features of interest. Since the air is dried over sulphuric
acid before it is returned to the calorimeter, it enters the respiration
chamber absolutely water-free and consequently the only sources
of water-vapor inside the chamber are the lungs and the skin of the
man. The ventilation of the chamber per hour, the amount of water
vaporized per hour, the average temperature of the calorimeter cham-
ber, and the relative humidity of the air are given in table 44.
Daily tests, in which the number of revolutions of the blower are
recorded by an automatic counter, have shown that in general 210 revo-
lutions of the blower correspond to a ventilation of 1 cubic foot or 28.315
liters of air. The total ventilation may therefore be obtained by divid-
ing the number of revolutions by this factor and multiplying by 28.315.
As a matter of fact, the number of revolutions per cubic foot of air
was determined each day and this variable used in the calculation.
From the length of period as given in table 44, the ventilation per
hour was readily found.
All of the water-vapor removed from the chamber was absorbed by
sulphuric acid as the ventilating current passed through the absorbing
system. The total amount was corrected for the small amount of
THE RESPIRATORY EXCHANGE.
321
water vaporized from the wet bulb of the psychrometer, and the amount
per hour found in the usual way.
The average calorimeter temperature was secured by means of a
series of resistance thermometers. The relative humidity was calcu-
lated from the amount of water vaporized per liter of ventilation and
the number of milligrams of water-vapor in one liter of air saturated
at the calorimeter temperature.
The rate of ventilation and the rate of carbon-dioxide production
were such that the usual proportion of carbon dioxide residual in the
Table 44. — Ventilation of chamber and relative humidity during experiments with L. in bed
calorimeter at night.
Date.
Day of
fast.
Period.
Ventila-
tion per
hour.
Water
vapor-
ized per
hour.
B
Average
temper-
ature
of the
chamber.
C
Relative
humid-
ity.
1912.
Apr. 10-11
11-12
12-13
13-14
14-15
15-16. . . .
16-17. . . .
17-18
18-19
19-20
20-21
21-22
22-23
23-24
24-25
25-26
26-27. . . .
27-28
28-29
29-30
Apr. 30-May 1
May 1- 2
2-3....
3-4....
4-5
5- 6.
6- 7.
7- 8.
8- 9.
9-10.
10-11.
11-12.
12-13.
13-14.
14-15.
16-17.
17-18.
1st.
2d..
3d..
4th.
5th.
6th.
7th.
8th.
9th.
10th.
11th.
12th.
13th.
14th.
15th.
16th.
17th.
18th.
19th.
20th.
21st.
22d..
23d..
24th.
25th.
26th.
27th.
28th.
29th.
30th.
31st.
llh
10
11
11
9
9
9
8
9
38m p.m.
13 p.m.
13 p.m.
09 p.m.
30 p.m.
41 p.m.
22 p.m.
58 p.m.
57 p.m.
30 p.m.
23 p.m
47 p.m,
50 p.m,
12 p.m,
18 p.m,
40 p.m,
53 p.m,
13 p.m,
27 p.m.
46 p.m.
29 p.m.
15 p.m,
51 p.m.
35 p.m,
31 p.m,
14 p.m.
35 p.m.
34 p.m,
41 p.m.
08 p.m,
47 p.m.
55 p.m.
30 p.m.
01 p.m.
38 p.m.
54 p.m.
57 p.m.
to8h
02m a.m.
00 a.m.
00 a.m.
00 a.m.
00 a.m.
00 a.m.
55 a.m.
00 a.m.
50 a.m.
00 a.m.
58 a.m.
55 a.m.
00 a.m.
52 a.m.
00 a.m.
52 a.m.
00 a.m.
00 a.m.
51 a.m.
47 a.m.
00 a.m.
50 a.m.
00 a.m.
55 a.m.
01 a.m.
50 a.m.
46 a.m.
50 a.m.
59 a.m.
03 a.m.
00 a.m.
30 a.m.
30 a.m.
31 a.m.
56 a.m.
54 a.m.
54 a.m.
liters.
2253
2472
2473
2454
2474
2468
2469
2369
2328
2410
2467
2432
2395
2204
2347
2254
2376
2313
2134
2321
2309
2334
2307
2356
2268
2212
2137
2195
2205
2309
2243
2312
2318
2257
2395
2318
2455
gm.
25.3
26.3
26.6
27.1
22.8
25.6
28.7
22.8
21.1
19.4
18.9
19.2
21.1
17.0
18.3
17.9
18.4
17.6
13.6
15.9
15.6
16.0
15.5
15.8
14.6
15.7
16.7
16.5
17.5
18.0
18.7
18.9
19.3
19.7
17.9
20.0
23.7
°C.
21.32
20.17
21.28
20.72
20.77
20.65
21.41
20.77
21.19
20.09
20.76
20.37
21.28
20.82
20.91
20.75
21.16
20.58
20.08
20.08
20.19
20.65
20.23
20.57
20.73
20.68
20.77
20.54
20.64
20.53
20.71
21.02
21.25
21.12
21.10
20.65
20.46
p. ct.
62
62
59
63
52
59
63
54
50
47
43
46
48
44
44
45
43
43
38
40
39
39
39
38
38
40
45
44
46
46
48
46
46
49
42
50
56
322 A STUDY OF PROLONGED FASTING.
chamber at the end of each experimental period throughout the night
was not far from 0.4 per cent by volume or approximately 13 times that
of normal air. This percentage of carbon dioxide in the air, and indeed
a very much higher percentage, has been shown to be entirely without
effect upon persons breathing such an atmosphere,1 so that it may be
stated with perfect confidence that the excess amount of carbon dioxide
present in the chamber could in no way have influenced either the
respiratory exchange or the heat-production of the subject.
As will be seen from table 44, the ventilation of the chamber aver-
aged not far from 2,200 to 2,300 liters per hour throughout the 31 days
of the experiment, the range being from 2,134 liters to 2,474 liters per
hour. The hourly vaporization of water had a tendency to decrease as
the fast progressed, the largest amount being on the night of the third
day of fasting and the smallest on the night of the fifteenth day of
fasting. The average temperature of the calorimeter remained for the
most part within a few tenths of a degree of the average figure, 20.6° C.
The relative humidity shows an interesting course. Beginning with
approximately 60 per cent on the nights following food, it decreased
to a minimum level of approximately 39 per cent from the fifteenth to
the twenty-first day and thereafter rose gradually to the end of the
fast. The variations in the excretion of water-vapor and the cause of
the fluctuations in the relative humidity will be discussed in a subse-
quent section of this publication.2 The results secured in the measure-
ments of the respiratory exchange inside the bed calorimeter may,
therefore, now be considered.
MEASUREMENT OF THE RESPIRATORY EXCHANGE INSIDE THE
BED CALORIMETER.
With the bed calorimeter it is possible to determine simultaneously
the water vaporized inside the chamber, the carbon dioxide produced,
and the oxygen consumed. These determinations were made directly
on four nights prior to the fast, on the 31 nights of the fast, and for two
nights after the fast. While the greatest emphasis must be laid upon
the total amounts measured, the absorbing vessels were changed several
times during the night, so that the experiment was usually subdivided
into three periods. In the latter part of the fast, the measurements
were made in five or six periods, and on two nights, seven and nine
periods respectively. It was accordingly possible to compute the car-
bon-dioxide output, the oxygen intake, and the respiratory quotient for
the whole experiment and also for the individual periods, thus giving a
control on the measurement of the respiratory exchange.
Benedict and Milner, U. S. Dept. Agr., Office Exp. Stas. Bui. No. 175, 1907, p. 237.
2See page 373.
THE RESPIRATORY EXCHANGE.
323
Periodic Changes in the Metabolism.
The average results for the experimental periods have been plotted
in the form of curves for each night that the subject spent inside the
respiration chamber. (See figures 41 to 44.) These results are given
in cubic centimeters per minute, the scale values on the outside indicat-
ing the values for the oxygen consumed, and those on the inside the
values for the carbon dioxide produced. The respiratory quotient for
each individual period is placed between the oxygen and the carbon-
dioxide curves. Thus, on the night of April 10-11, the oxygen intake
for the first period averaged 284 c.c. per minute, for the second period
! :00P.M. 10.00 I2;00 2:00A.M. 4:00 6:00 8:00
230
.220
tre
26C
2JC.
24C .230
.220
.210
280 .
270 .
260 .
2S0 .
240 .
230 230
220
210
200
190
180
0.84. 0.81
APR. 13-14
CO,
00 P.M. 10.00 12-.0Q 2:00A.M. 4:00 6:00 8:00
Oj
830
coa
-180 .
o«
APR. I4--I5
0
220
0.73
210
0.78
190
_170
.160
C02
0.77
220
°i
APR. 15-16
<7\
210
0.78
.170
.160
C02
0.74-
.ISO
210
°i
APR. 16-17
(5)
200
.155
eog
0.75
190
0.71
.145
210
• °!
APR. 17-18
. 0
200
190
C02
0.73
0.7 S
180
0.76
.140
Fig. 41. — Curves showing oxygen consumption,
carbon-dioxide production, and respiratory
quotient during night periods in the bed
calorimeter for the four days preceding the
fast and the first to the fourth days of the
fast.
324
A STUDY OF PROLONGED FASTING.
265 c.c. per minute, and for the third period 270 c.c. per minute. The
carbon-dioxide production for the corresponding periods was respec-
tively 227, 221, and 218 c.c. per minute and the respiratory quotients
0.80, 0.84, and 0.81 respectively.
The subdivision into experimental periods was made in an attempt
to secure information regarding the periodic changes throughout the
night. But from fundamental factors in the technique of the calori-
meter experiments, the longer the experimental periods, the more accu-
rate are the measurements of the carbon-dioxide production and es-
pecially of the oxygen consumption; hence by subdividing these total
values, a certain degree of accuracy is sacrificed in the measurements,
although the average values for the night are unaffected. This may
explain certain discrepancies in the respiratory quotient and in the
general conformity of the curves for the oxygen and the carbon dioxide.
5 It will be seen that as a rule the curves for the carbon dioxide and
oxygen are essentially parallel, although they are by no means straight
8:00 P.M. ifcQO IfcQO 2O0A.M 4 00 8:00 6:00 BjOOP.M I0:0Q 12:00 2:OOA.M *:O0 6.00 8:00
°,
o«
0
180
170
180
170
160
180
170
I7S
165
175
165
155
170
16C
CO, APR. 23-24.
140 °l 1
0.75
%
0.°* co«
0.7 Z
APR. 18-19
0.7 e
130 CO,
071
190 .
0.78
°»
G
vii
170
0.77
0.70
.130 CO,
APR. 24-25
I3S
0.74
0.74
200
o,
©
■20
0.70
APR. 19-
-1ZO
190
_140
co»
0.67
ISO
« APR. 25-26
<S
0.74
CO, 07*
H
0,
APR. 20-
G\
-125
200
21
0.70
190
°z
APR. 26-27
<3
-140
CO,
0.71
L.30 CO, •»*
ieo
0.73
0.71
.120
.130
°»
<5
°*
©
0.69
.130
.120 C0»
APR. 27-28
.140 CO,
0.70
190
APR.
21-22
0.75
180
0.73
1
0.73
1
0.77
.110
.130
J
°2
APR. 22-23
©
APR.
°2
28-29
' 0.71
C
.145
-135 CO,
0.72
1
1
I7C
0.81
.125
CO, °-72
1
r
113
—
FIO- 42. Curves showing oxygen consumption, carbon-dioxide production, and respiratory quotient durir
night periods in the bed calorimeter for the fifth to the fifteenth days of the fast.
THE RESPIRATORY EXCHANGE.
325
lines for the whole experiment. Occasionally, discrepancies are found,
as on April 10-11 (the first night of the experiment) when the carbon-
dioxide production is higher in the second period than in the last period,
while as a matter of fact the oxygen consumption is apparently some-
what lower. This is especially noticeable on April 22-23, when the
carbon-dioxide production in the last period increased and the oxygen
consumption decreased. As would be expected, the possibilities for a
discrepancy between the curves increase as the period is shortened and
consequently we find on the night of May 4-5, when the experiment was
divided into nine periods, that while the values as a whole are approxi-
mately parallel, in the sixth and seventh periods there is a great in-
crease in the oxygen consumed which is unaccompanied by a corre-
sponding increase in the carbon-dioxide production. Similar irregu-
larities are to be noted on the night of May 13-14. In general, however,
there is a striking tendency toward parallelism in the two curves.
After the first three nights of fasting, the minimum values for carbon
dioxide and oxygen are usually found in the middle period of the night,
i. e., from 2 to 4 a. m., or thereabouts, the morning period almost inva-
J:00P.M.
10:00
12:00. 2:00A.M. 4:00
6:00
8:00
1 t
0 .
CO,
.0 .125
.115
0*
APR. 29-30
Q
C02
0.69
0.72
0.72
1 i0.
1 0 .120
.110
°l
*PR.30-MAY 1
to
co2
0.70
0.73
0.73
1 5 .
Ii5 .125
.115
o,
MAY 1-2
1 0.75
O
co2
0.70
0.7Z
1
|
1 ».
1 <0 .115
.I0S
o.
MAY 2-3
L_
(3)
COj
0.70
0.70
0.7*
! ».
>0 .115
-!<35
o,
MAY 3-4.
©
CO,
0.70
0.71
0.73
r~
6.00 P.M. 10.00 12.00 2:00A.M. 4:00 6:00
I :o. 43. — Curves showing oxygen consumption, carbon-
dioxide production, and respiratory quotient during ,50
night periods in the bed calorimeter for the sixteenth
to the twenty-fourth days of the fast.
326
A STUDY OF PROLONGED FASTING.
riably showing a tendency to rise. This may be seen with great regu-
larity throughout most of the curves, although there are enough
exceptions (for instance, on May 4-5 and May 6-7) to make it inappli-
cable in all cases.
After the first few days of the fast, one would not expect a great
change in the respiratory quotient, since there would be no material
alteration in the character of the material oxidized in the body. An
examination of the respiratory quotients given with the curves shows
that they run not far from a constant value throughout the night with
the different conditions of food and fasting. Thus, on the first few
nights with food, the values are considerably above 0.80, but with the
beginning of the fast they drop rapidly to about 0.74, remaining not
far from 0.72 throughout the remainder of the fast. Occasionally cer-
tain fluctuations above or below the average figure may be observed,
8:00P.M. 10:00 12:00 f.QQAM. 4:00 6:00
CO,
115
105
125
115
105
iL.
MAY 9-10
©
may io-ii
MAY 11-12
CO, Q.75
MAY 12-13
©
>0 P.M. 10:00 12:00 2:00 A.M. 4:00 6jOJJ 9:00
200
190
180
CO,
120
110
200
190
180
135 coz
125
IIS
"MAY 16-17
0.80 0.84
0 MAY 13-14
—1 .
itt 1 r^
C02
— 1 ,. 0.74 n«7
L
o,
MAY 14-15
049
~L
0.73
C02
0.70
0.72 074
©
Fig. 44. — Curves showing oxygen consumption, carbon-dioxide production, and respiratory quotient during nigfc
periods in the bed calorimeter for the twenty-6fth to the thirty-first days of the fast and the secon
and third food days.
THE RESPIRATORY EXCHANGE. 327
but these may easily be attributed to the fact that the shortness of the
period affected the determinations of the oxygen consumption. It may
be considered an established fact that respiratory quotients more than
0.02 above or below the average value for the night are due to acci-
dental variations in the determinations. A low respiratory quotient
following a high quotient may frequently be noted, showing that there
is a compensation in the measurement of the oxygen consumption as
the experiment continues.
Inasmuch as the determination of the respiratory quotient requires
extremely accurate determinations of both the oxygen consumption
and the carbon-dioxide production, it is obviously very much more
difficult to secure accurate respiratory quotients than accurate measure-
ments of either the carbon dioxide or the oxygen. Accordingly, while
we feel that it is legitimate to accept the values for the carbon dioxide
and even for the oxygen for short periods, we are by no means certain
that we are justified in laying considerable stress upon the respiratory
quotients in periods so short as those in the calorimeter experiments.
A careful scrutiny of all of the kymograph records shows that the
extraneous muscular movements, although not absolutely constant in
every period, are so slight that they may be practically neglected.
Such movement as there was did not correspond closely to the general
trend of the katabolism, for although the subject was more active
during the morning period, the activity was not sufficient to account
for the great difference in the katabolism. On the other hand, it is per-
fectly clear from his own records that the subject was usually in deep
sleep in the middle of the night, as he often reported in the morning
that he awoke about 4 a. m. and lay awake until the end of the experi-
ment. A relationship between deep sleep and the metabolism is there-
fore indicated, a relationship which will be discussed in a subsequent
section.
Perhaps the most striking fact shown by the whole series of curves
is that the subject had by no means a constant metabolism. This man
was living on a low metabolic plane, had a remarkable degree of mus-
cular repose as shown by the kymograph records, and was without
food in the alimentary tract ; and yet, as has already been pointed out,
the curves show a distinct tendency to fall off in the first part of the
night until a minimum is reached from 2 to 4 a. m., and then to rise
again in the later morning hours.
Total Metabolism.
The difficulties incidental to comparing the short-period values for
the gaseous exchange are eliminated when one uses as a unit the
results obtained during the entire sojourn of the subject inside the
respiration chamber during any given night — that is, for a period of
328
A STUDY OF PROLONGED FASTING.
10 or 12 hours. Consequently a comparison may be made of the
results obtained for the individual nights as the fast progressed. We
may, indeed, go further and compare not only the average values found
throughout the night, but likewise the average values for the minimum
periods in the experiments. While obviously there is an opportunity
for possible error in thus selecting minimum periods, particularly in
the measurements of the oxygen consumption, nevertheless it is be-
lieved that such errors will be equalized throughout a 31 -day fast.
Accordingly, in table 45 we give both the average and the minimum
Table 45. — Gaseous exchange of subject L. during experiments in the bed calorimeter at night.
Carbon dioxide
Oxygen per
Date.
Day of
fast.
per minute.1
minute.1
Respira-
tory
quotient.
(A-C)
Average
pulse-
rate.
Average
body-
tempera-
ture.
Aver-
age.
Mini-
mum.2
Aver-
age.
Mini-
mum.2
A
B
C
D
E
F
G
1912.
ex.
c.c.
c.c.
c.c.
°C.
Apr. 10-11
224
228
218
180
165
218
217
196
173
152
276
258
265
24fi
0.81
.88
.86
.81
.78
82
76
78
70
68
11-12. .
12-13
252 235
13-14
221
212
208
196
14-15
1st
15-16
2d
159
154
211
208
.75
66
36.41
16-17
3d
151
148
206 198
.73
62
17-18
4th ... .
150
140
202 1 187
.74
65
36.55
18-19
5th ... .
143
137
192
176
.75
63
36.58
19-20
6th ... .
134
131
194
185
.68
60
36.44
20-21
7th ... .
135
132
190
185
.71
59
36.42
21-22
8th....
137
135
187
177
.73
61
36.55
22-23
9th....
134
131
178
173
.75
59
36.50
23-24
10th
130
127
180
179
.72
57
36.64
24-25
11th....
128
124
176
166
.72
57
36.80
25-26
12th
129
126
175
173
.73
68
36.85
26-27
13th
126
125
171
167
.74
56
36.62
27-28
14th
120
116
167
160
.72
53
36.30
28-29
15th
117 114
163
162
.71
53
36.43
29-30
16th
117
114
165
158
.71
53
36.40
Apr. 30-Mav 1
17th....
115
113
160
154
.72
52
36.42
Mav 1-2
18th....
115
112
159
154
.72
52
36.30
2-3
19th....
113
111
158
153
.71
52
36.21
3-4
20th....
114
112
160
159
.71
52
36.51
4-5
21st
112
103
154
137
.73
54
36.12
5-6
22d
111
109
154
153
.72
53
36.10
6-7
23d
112
106
156
144
.72
66
35.98
7-8
24th
109
106
158
152
.69
55
35.88
8-9
25th ....
111
108
153
147
.72
55
36.31
9-10
26th ....
111
106
159
151
.70
56
10-11
27th
111
107
153
145
.72
57
36.03
11-12
28th
115
109
162
145
.71
59
36.37
12-13
29th
112
104
158
152
.72
58
36.23
13-14
30th....
110
103
151
147
.72
58
36.06
14-15
31st. . . .
115
109
160
148
.72
57
36.14
16-17
124
188
117
176
154
194
143
182
.80
.97
64
90
36.79
37.53
17-18. .
1For the duration of the period during which the metabolism was measured see table 44.
2The duration of the periods in which these minimum values were observed varies in general
from 3 hours to 1 hour.
THE RESPIRATORY EXCHANGE. 329
periodic values for the total gaseous exchange and the respiratory
quotient for each night of the experiment.
The average for the carbon-dioxide production varied from 228 c.c.
per minute on April 11-12 (the second night of the preliminary food
period) to 109 c.c. on May 7-8 (the twenty-fourth night of the fast).
The carbon-dioxide excretion during the fasting period was much less
than during the days when food was taken, ranging from 165 c.c. on
the first night to 109 c.c. on the twenty-fourth night. This increase
with the taking of food is also shown in the two nights following the
fasting period, when the carbon-dioxide production increased from 115
c.c. to 124 c.c. on the second night after the food was taken, and to
188 c.c. on the third night.
The minimum periods have a special interest, as they indicate the
lowest plane of metabolism during the experiment. These values range
from 218 c.c. on the first night with food to 103 c.c. on both the twenty-
first and thirtieth nights of the fast. As with the average values, an
increase after taking food is noted in the minimum periods for the two
nights following the fast.
While the carbon dioxide of itself is a well-known index of the metab-
olism, particularly in fasting, when the character of the material
burned remains relatively constant, nevertheless the values for oxygen
likewise have significance. The average values for the oxygen consump-
tion range from 276 c.c. on the first night with food to 151 c.c. on the
thirtieth night of the fast. Considering only the fasting values, the
oxygen consumed ranges from 212 c.c. on the first night of the fast to
the minimum noted above. While no increase in the oxygen con-
sumption is apparent in the first value obtained after the taking of
food, a considerable increase is shown on May 17-18, the oxygen con-
sumed being 194 c.c.
The minimum periodic values for the oxygen consumption can not be
considered to have the same degree of accuracy as the minimum
periodic values for the carbon-dioxide production, but a comparison is
of interest. These values range from 265 c.c. on the first night with
food to 137 c.c. on the twenty-first night of the fast. During the fast-
ing period the range is from 208 c.c. on the second night of the fast to
137 c.c. on the twenty-first night. As with the average values, the
minimum periodic values for the oxygen consumption do not show an
increase after taking food until the third night.
From an examination of all of the data given in table 45 for the car-
bon-dioxide production and the oxygen consumption, it will be seen
that there was a striking tendency for the total metabolism to decrease
as the fast progressed. Both the average and the minimum periodic
values show that the metabolism reached a low point about the
twentieth day of the fast, and from that time to the end of the fast there
was but little alteration.
330 A STUDY OF PROLONGED FASTING.
The oxygen consumption showed the same general course as the
carbon-dioxide production, there being a steady decrease until about
the twentieth day. While the values for the oxygen between the
twenty-first and the thirty-first days do not show the regularity that
was observed with the carbon dioxide during the same period, they still
do not fluctuate widely from the low value, the average values remaining
between 151 c.c. and 162 c.c. and the values for the minimum periods
between 137 c.c. and 153 c.c. It will be seen, therefore, that the metab-
olism as indicated by the carbon-dioxide production and the oxygen
consumption decreased regularly until the twentieth day and from that
time until the end of the fast remained essentially constant.
This observation is strikingly significant, inasmuch as one would nat-
urally expect that, as the organism wasted away as a result of fasting,
the total metabolism would decrease and likewise the intensity of the
metabolism. The fact that the decrease in the metabolism did not con-
tinue beyond the twentieth day is the more surprising, since the loss
in weight continued regularly throughout the fast. The absence of a
continued decrease in the metabolism will subsequently be given special
discussion.
Respiratory Quotient.
At present the best index we have of the character of the material
burned in the body is the relationship between the volume of the carbon
dioxide excreted and the oxygen consumed, i. e., the so-called respira-
tory quotient. When carbohydrates are burned, the volume of carbon
dioxide produced is equal to that of the oxygen consumed, the respira-
tory quotient being 1.0. On the contrary, when fat is burned, there is
a much less volume of carbon dioxide produced per liter of oxygen and
the respiratory quotient is not far from 0.7.
In the study of short fasts previously made at Wesleyan University,
it was shown that the carbohydrates stored in the body (chiefly in the
form of glycogen) were heavily drawn upon in the first few days of the
fast and thereafter the body subsisted substantially upon fat and pro-
tein, but chiefly fat. In this fasting experiment, therefore, a rapid
fall in the respiratory quotient would be expected during the first days
of the fast, and a possible constancy during the remaining days, show-
ing a combustion of fat.
By reference to the values for the respiratory quotient given in table
45, it will be seen that on the 4 nights prior to the fast the values
ranged from 0.81 to 0.88, averaging not far from 0.84. This quotient
is approximately that which would be expected with individuals sub-
sisting upon a mixed diet. On the first day of fasting, the respiratory
quotient fell to 0.78 and remained for the next few days not far from
0.74 to 0.75. On the sixth day a low value was found of 0.68, but for
the remainder of the fast the respiratory quotient ranged above or
THE RESPIRATORY EXCHANGE. 331
below the average of 0.72. No average value lower than 0.68 was found
in any of the experiments. These respiratory quotients would indicate
that the combustion in the body after the first few days of fasting was
principally of fat. As will be seen later, there was the formation of a
small amount of /3-oxybutyric acid, which would have a tendency to
lower slightly the respiratory quotient, but this would be in part com-
pensated by the consumption of protein and a possible steady, though
very small, oxidation of carbohydrate, both of which would tend to
increase the respiratory quotient. Such an increase is indicated by the
slightly higher average value of 0.72.
The two nights after the fast, when food had been taken, show a
marked increase in the respiratory quotient, the quotient for May 16-
17 being 0.80. On the night of May 17-18, when the whole alimentary
tract of the subject was filled with carbohydrate material, due to the
excessive amount of fruit juices and honey which he had taken, the
extraordinarily high value of 0.97 was obtained.
The main conclusions to be drawn from the average respiratory
quotients found in the experiments with the bed calorimeter as the fast
progressed may be summed up as follows :
First, no very low values were found, such as have been observed and
reported by other investigators. Quotients below 0.68 were very
rarely found for the individual periods, and the average value for the
31 nights of the fast was 0.72.
Second, from the course of the respiratory quotients, it is clear that
carbohydrate was burned on the first few days of fasting, which is in
full conformity with the results found in the experiments carried out at
Wesleyan University.
Finally, after the subject had fasted for 6 or 7 days, the respiratory
quotients reached a point which indicated essentially a fat katabolism
and continued at this point until the end of the fast, the formation of a
small amount of /3-oxybutyric acid tending to lower the respiratory
quotient and the combustion of a small amount of protein, with possibly
a small amount of glycogen, tending to increase the quotient above that
which would be obtained with the combustion of pure fat.
Relationships of Pulse-Rate, Bodt-Tempebatube, and Metabolism.
In considering both the curves of the respiratory exchange and the
average values shown in table 45, it should be noted that two of the
factors affecting the total metabolism were absent, i. e., muscular
exercise and the digestion of food. Considering that the subject is
living on a low metabolic plane, we might expect that the metabolism
would be represented by a straight line, were it not for the influence of
a third important factor — the internal muscular activity. The best
index of the internal activity is the pulse-rate. Johansson1 has also
Johansson, Skand. Archiv f. Physiol., 1898, 8, p. 85.
332 A STUDY OF PROLONGED FASTING.
pointed out that there is an intimate relationship between the body-
temperature and the metabolism. It is important, therefore, to con-
sider the relationship between the metabolism as shown by the gaseous
exchange and these two indices of the internal condition.
It has seemed impracticable to complicate the curves in figures 41 to
44 by superimposing others, but a comparison can be made by referring
to the curves for the pulse-rate and the body-temperature1 given in
previous sections of this publication. Such a comparison shows that
the curves for the carbon-dioxide excretion, the oxygen consumption,
the pulse-rate, and the body-temperature have in general the same
course for each experiment, with a distinct tendency to fall off during
the evening until a minimum is reached about the middle of the night,
and then to rise in the morning. This parallelism with the metabolism
is shown more clearly in the curves for the body-temperature, as there
are more variations in the curves for the pulse-rate, but the general
rhythm of the latter is much like that exhibited by the curves for the
metabolism. Furthermore, there does not appear to be a material
difference in these two relationships at the beginning and end of the
fast, so that it would seem that fasting per se does not affect them.
The body acts as a unit, therefore, irrespective of the state of nutrition.
The intimate relationship between the pulse-rate and the metabolism
(which has been emphasized in this laboratory for a number of years)
and the relationship between the body-temperature and the metabolism
are thus not only demonstrated in a remarkable manner, but are also
shown to be unaffected by a prolonged fasting period.
The relationship between the pulse-rate and the total metabolism
and the body-temperature and the metabolism as the fast progressed
may be discussed more in detail in connection with the values given
in table 45, using the average values rather than those for the minimum
periods. In comparing these factors with the total metabolism on
successive nights, it should be borne in mind that the relationships
would not logically be expected to remain constant, for we have on the
one hand the pulse-rate and the body-temperature governed by certain
laws and on the other an organism producing heat, the heat-producing
mechanism of which is constantly diminishing in size.
The pulse-records for the nights preceding the fast are somewhat
irregular, but the technique for making the observations was not then
so perfectly developed as it was later in the experiment and the assis-
tant had not the time to make such frequent records. It will be seen,
however, that the high pulse-rates were obtained with the high values
for the carbon dioxide and the oxygen during the 4 nights prior to the
fast and throughout the first 2 weeks of the fasting period. In the
latter portion of the fast there was a distinct tendency for the average
pulse-rate to increase without a corresponding increase in the total
1See figures 4 to 8, pages 90 to 94, and figures 12 to 18, pages 104 to 110.
THE RESPIRATORY EXCHANGE. 333
carbon-dioxide output and oxygen intake. On May 17-18, however,
the greatly increased pulse-rate was accompanied by a corresponding
increase in both the carbon-dioxide output and the oxygen intake.
In general, then, one may infer that even with an organism whose heat-
producing mechanism is constantly decreasing in size, there is still
an intimate relationship between the pulse-rate per minute and the
total heat production. It should also be recognized that this relation-
ship was somewhat disturbed during the latter portion of the fast,
but not sufficiently disturbed as not to be again apparent on the third
day with food.
The body-temperature was not recorded on the nights preceding
the fast, but observations were made nearly every night of the fast and
for two nights following. The values given in table 45 for the fasting
period have a distinct tendency to remain not far from an average
value of 36.36° C, and range from 36.85° C. on the twelfth night of
the fast to 35.88° C. on the twenty-fourth night. From the twenty-first
night of the fast, the values for the most part lie distinctly below the
average of the first 3 weeks of the fasting period; but little if any rela-
tionship is shown between the average body-temperature and the total
metabolism. On the other hand, on the last night of observation
after the fast (May 17-18), the increased metabolism and increased
pulse-rate were accompanied by the highest average temperature
found on any night with this subject.
It is evident, therefore, that while there is a tendency towards a
parallelism of the body-temperature and the metabolism throughout
any given night, there is no distinct tendency towards parallelism
between the average temperatures of successive nights and the total
metabolism as measured. Evidently the heat-regulating mechanism
of the body is in large part independent of the total heat-production
or of the condition of nutrition of the subject.
STUDIES WITH THE UNIVERSAL RESPIRATION APPARATUS.
The facility with which experiments could be carried out with the
universal respiration apparatus made it specially adapted for measuring
the metabolism of the fasting subject under various conditions, such as
lying awake, sitting up either quietly or writing, or lying awake breath-
ing an oxygen-rich atmosphere. It was also possible to obtain accu-
rate determinations of the respiratory quotient with this apparatus.
This was of special importance, since it was desired to establish as
sharply as possible the respiratory quotient obtaining during prolonged
fasting, particularly as the low quotients found by Luciani and by
Zuntz and his co-workers have been the subject of much discussion.
Consequently it was decided that, throughout the entire fast, respira-
tion experiments would be made as frequently as practicable in which
the respiratory exchange would be determined under various condi-
334
A STUDY OF PROLONGED FASTING.
tions. A summary of the data obtained in these experiments is given
in table 46.
For purposes of comparison the oxygen absorbed and the carbon
dioxide produced were calculated on the basis of cubic centimeters
per minute. The respiratory quotient for each experiment and the
average pulse-rate are also given in this table. The morning respira-
tion experiments were made immediately following the calorimeter
Table 46. — Gaseous exchange of subject L. at different times
activity. (Respiration apparatus.)
of the day and with varying
Date.
Day of
fast.
Lying (usually 8h 30° a.m. to
9h 30™ a.m.).
Lying (usually 7 p.m. to
7h45mp.m.).
Carbon
dioxide
per
minute.
Oxygen
per
minute.
Respi-
ratory
quo-
tient.
Aver-
age
pulse-
rate.
Carb
dioxi
per
minu
on --.
de 0xygen
per
, minute.
;e.
R
r:
c
ti
espi- Aver-
tory age
[uo- pulse-
ent. rate.
1912.
c.c.
c.c.
c.c.
c.c.
Apr. 11
186
196
200
182
185
231
220
225
223
237
0.81
.89
.89
.82
.78
72
73
72
73
74
12
13
14
15
1st
16
2d
180
227
.79
73
17
3d
169
226
.75
70
18
4th
159
212
.75
68
19
5th ... .
158
205
.77
67
20. . ..
6th
148
200
.74
64
21
7th ... .
153
204
.75
64
22
8th
151
203
.74
65
23
9th....
143
190
.75
63
24
10th....
143
187
.76
63
25
11th....
140
187
.75
61
26 ... .
12th
140
187
.75
61
13<
) 193
0
72 62
27
13th....
140
192
.73
59
li3(
\ !195
i
70 »59
28
14th
134
181
.74
58
134
I 190
71 59
29
15th
132
179
.74
57
137
189
72 61
30
16th
133
182
.73
58
134
[ 190
71 59
May 1 . . . .
17th
130
182
.71
57
13C
1 188
69 61
2
18th....
123
174
.71
56
128
1 189
68 62
3
19th
127
177
.72
57
12e
182
69 60
4
20th
124
173
.72
58
5
21st
126
174
.73
59
12£
182
69 57
6....
22d
124
170
.73
59
121
176
71 63
7
23d
121
165
.73
58
12e
175
72 60
8
24th ....
122
167
.73
59
12£
177
71 63
9
25th
125
166
.75
60
124
176
70 63
10
26th
123
168
.73
61
128
180
71 66
11
27th....
129
172
.75
62
126
181
70 66
12
28th
124
166
.75
61
123
178
69 66
13
29th....
124
171
.73
63
123
178
69 67
14
30th ....
119
166
.72
59
127
183
69 71
15....
31st
120
166
.72
60
17
133
172
170
183
.78
.94
72
84
18
1During a period from 3h 16m p.m. to 3h 51m p. m. on this day, with the subject in the lying
position, the observations were: Carbon dioxide, 140 c.c; oxygen, 189 c.c; respiratory quotient,
0.74; pulse-rate, 61 per minute.
THE RESPIRATORY EXCHANGE.
335
Table 46. — Gaseous exchange of subject L. at different times of the day and with varying
activity. {Respiration apparatus.) — Continued.
Date.
Day of
fast.
Sitting.1
Period.
Carbon
dioxide
per
minute.
Oxygen
per
minute.
Respira-
tory
quotient.
Average
pulse-
rate.
1912.
Apr. 16. . .
19...
23...
24...
26...
27...
29...
May 1 . . .
4...
7...
14...
2d
5th....
9th....
10th
12th
13th....
15th....
17th....
20th....
23d
30th
4h 00m p.m. to 4b 35m p.m.
4 10 p.m. 4 43 p.m.*
3 52 p.m. 4 28 p.m.
3 58 p.m. 4 57 p.m.
3 13 p. n. 4 11 p.m.
12 14 p.m. 12 48 p.m.
3 23 p.m. 3 56 p.m.*
9 31 a.m. 10 04 a.m.*
9 35 a.m. 10 10 a.m.*
3 43 p.m. 4 14 p.m.*
6 32 p.m. 7 02 p.m.*
c.c.
179
198
129
144
124
141
164
153
141
159
156
c.c.
244
269
187
194
183
200
233
215
208
222
221
0.73
.74
.69
.74
.68
.71
.70
.71
.68
.72
.71
82
80
62
69
60
68
68
69
65
69
75
1Periods indicated by an asterisk were obtained with the subject sitting, writing.
experiment, with the subject still lying upon the couch in essentially
the same position as when he left the calorimeter chamber. The
evening respiration experiments were made in the latter days of the
fast just before the subject entered the calorimeter chamber. The
experiments when the subject was sitting were of two kinds. In certain
experiments he sat quietly in his chair, but in others he was writing,
exactly as is shown in plate 1, figure b. It is thus seen that observa-
tions were made with this fasting subject on every day of the fast and
on certain days experiments were made in several body positions.
These data also give an indication of the diurnal variations in the katab-
olism, since observations were made under identical body conditions, so
far as muscular activity and absence of food are concerned, both in
the morning after the subject left the calorimeter and in the evening
on the same day just prior to entering the chamber for the night.
VARIATIONS IN THE METABOLISM AS THE FAST PROGRESSED.
The carbon-dioxide production of the subject while lying on the
couch in the morning respiration experiments ranged from 200 c.c.
on the morning of April 13 (one of the days preceding the fast) to 119 c.c
on May 14 (the thirtieth day of fasting). During the fasting period
the values ranged from 185 c.c. to 119 c.c. When the subject again
took food the rise in the carbon-dioxide production was very notice-
able, particularly on the last day of observation.
In general, the data obtained for the oxygen consumption nearly
paralleled those for the carbon-dioxide production. The maximum
336 A STUDY OF PROLONGED FASTING.
value of 237 c.c. was obtained on the first day of the fast, while the
minimum value of 165 c.c. was found on the twenty-third day of the
fast. The striking constancy shown in the values for the oxygen con-
sumption on the first four mornings prior to the fast and on the first
day of the fasting period is worthy of special notice, as it gives evidence
in the first place of the remarkable constancy in the katabolism of this
man and likewise of the regularity of his muscular repose. Both the
carbon-dioxide production and the oxygen consumption fell off as the
fast progressed, and although the minimum value for the oxj'gen con-
sumption was reached on the twenty-third day, yet the remaining days
of the fasting period indicate a katabolism not far from the minimum
value of 165 c.c. It is of particular interest to note that on the 17th
of May, i. e., the second day of taking food, the metabolism had not
materially increased, as shown by the oxygen consumption, but on the
last morning (May 18) there was a marked increase from 170 c.c. to
183 c.c.
The trend of the respiratory quotient is likewise significant. On the
first 4 days with food, the respiratory quotient varied from 0.81 to 0.89,
this being not far from the average respiratory quotient found with
normal individuals subsisting on a mixed diet. At the beginning of the
fast, the respiratory quotient was a little lower on the first few days
and then steadily decreased until a minimum value of 0.71 was found
on the seventeenth and eighteenth days. During the remainder of
the fast, the value for the respiratory quotient remained at about 0.73.
On the second day with food it rose to 0.78, and on the third day with
food it reached the extraordinarily high value of 0.94, indicating that
the subject was surcharged with carbohydrate material. The absence
of very low quotients during the fast was noticeable. It should be
borne in mind that the values for the oxygen consumption represent
more nearly the true index of the total metabolism than do the values
for the carbon-dioxide production, particularly in the 4 days with food
preceding the fast and the first few days of fasting. After the third or
fourth day of fasting, however, the values for the carbon-dioxide pro-
duction and the oxygen consumption were essentially parallel, so that
either may be looked upon as a true measure of the total metabolism.
RELATIONSHIP BETWEEN THE PULSE-RATE AND THE METABOLISM.
The pulse-rate remained remarkably constant for the first 4 days
with food before the fast and likewise on the first few days of fasting,
ranging from 72 to 74, which is in general conformity with the measure-
ments of the oxygen consumption. It then fell with a considerable
degree of regularity until a minimum value of 56 was reached on the
eighteenth day of the fast. Subsequently the values show a slight,
though definite, tendency to rise gradually to the end of the fasting
period. The increase on the second day after food was taken was
THE RESPIRATORY EXCHANGE. 337
considerable, with a still further increase on the last day on which the
observations were made.
A careful examination of the fluctuations in the values for the oxygen
consumption and the pulse-rate shows a remarkable regularity in the
relationship between them, although the absolute minimum values were
not found on the days that the minimum pulse-rate was found. On
the second day with food after the fast, the pulse-rate rose to 72 and
the oxygen consumption likewise rose, reaching 170 c.c. The values
taken as a whole, however, show that in the earlier days of this long
fast the relationship between the oxygen consumption and the pulse-
rate was reasonably close, but in the latter part of the fasting period
there was a slight divergence, as a somewhat increased pulse-rate was
occasionally accompanied by an actual decrease in the oxygen con-
sumption. It should be considered, however, that the organism was
changing from day to day, and while the total tissue available for
metabolism was slowly decreasing the pulse-rate may still have a
definite relationship to the total active tissue remaining. Thus a
decrease in the amount of tissue may in part be compensated for by an
increase in the pulse-rate, although this latter factor may still have too
small an effect to prevent a lowering of the total metabolism. Further
discussion along this line must be deferred until the metabolism per
unit of body-weight and per unit of body-surface are considered.
DIURNAL VARIATIONS IN METABOLISM.
The determination of the respiratory exchange at various times
during the day gives an excellent opportunity for studying the diurnal
variations in the metabolism of the same individual during fasting.
The data given in table 46 show that the metabolism during the evening
experiments was invariably higher than in the morning experiments,
regardless of whether the carbon-dioxide production or the oxygen
consumption is used as an index.
The pulse-rate was also a few beats higher in the evening hours, thus
indicating a close relationship between the pulse-rate and the metab-
olism. The slight tendency for the pulse-rate to rise in the morning
experiments beginning with the eighteenth day of the fast and con-
tinuing to the end was even more marked in the records of the pulse-
rate for the series of evening experiments, in which the minimum value
of 57 was found on the twenty-first day of the fast and the maximum
of 71 on the thirtieth day. A general relationship between the oxygen
consumption and the pulse-rate is shown throughout all of the series
of experiments, although as the fast progressed this relationship was
not so pronounced as at the beginning. It should be considered here
again, however, that the active mass of protoplasmic tissue was gradu-
ally decreasing, so that the relationship can not be expected to hold
constant.
338 A STUDY OF PROLONGED FASTING.
The respiratory quotients obtained in the evening were not unlike
those obtained in the morning experiments, with a slight tendency for
the early evening quotients to be somewhat lower than those obtained
in the morning experiments. This lowering of the quotient in the
evening experiments might be taken as an indication that there may
have been a formation of carbohydrate from fat by a storage of oxygen,
or a greater formation of /S-oxybutyric acid, and that the next morning
either the formation of the /3-oxybutyric acid was less or that the slight
supply of glycogen formed during the early evening was being burned.
Unfortunately, although these respiratory quotients were determined
with the best technique that we know of at present, we do not feel
justified in laying much stress upon a change of one or two units in the
quotients.
EXTERNAL INFLUENCES UPON METABOLISM.
While the values for the carbon-dioxide output, the oxygen intake,
the respiratory quotient, and the pulse-rate are given in table 46 for the
experiments in which the subject was lying upon a couch and sitting
up in a chair, either writing or quietly at rest, a better understanding
of the influence of a change in conditions may perhaps be secured by
studying each change by itself. To this end several small tables have
been prepared which show the influence of the change in condition
upon the total metabolism and also upon the mechanics of respiration.
Effect of Changes in Body Position.
On the second, tenth, twelfth, and thirteenth days of the fast, the
metabolism was studied while the subject was sitting in a chair. It was
thus possible to compare the metabolism and the mechanics of respi-
ration for the two positions. This comparison is made in table 47, in
which is given the increase or decrease in the values due to the change
to the sitting position. The figures show that in general there was
practically no increase in the carbon-dioxide production — and, indeed,
in two instances a considerable decrease. The oxygen consumption
was increased in 3 out of the 5 experiments, with a slight decrease in
the other 2, and there was a perceptible though probably not significant
change in the respiratory quotient, which may have been caused by
the absence of change in the carbon-dioxide production. There was
an average increase in the pulse-rate and respiration-rate and an
increase in the lung ventilation, but varying results in the volume per
respiration.
Since the lying experiments were made in the early morning and the
sitting experiments late in the afternoon, they are not, strictly speaking,
comparable. On the other hand, one would expect that there would be
a higher metabolism normally in the late afternoon than in the morning
after the subject came out of the calorimeter, and it is accordingly very
THE RESPIRATORY EXCHANGE.
339
difficult to explain the results obtained on the ninth and twelfth days of
the fast, when there was an actual decrease in the oxygen consumption,
with a slight falling off of the pulse-rate.
The general course of the metabolism noted on the second, tenth,
and thirteenth days is essentially that which is found with normal
individuals — namely, a small increase due to the position of sitting.
This increase was also accompanied by an increased pulse-rate. The
parallelism shown here between the increase of the oxygen consumption
and the pulse-rate is worthy of special attention.
The figures also show that the change to the position of sitting
invariably results in an increased ventilation of the lungs, although
the respiration-rate changes so that the actual volume per inspiration
may be above or below that when the subject was lying. No positive
deductions can be drawn as to the influence of the change in position
upon the volume per inspiration. If an average of these five experi-
ments were permissible, it would be seen that there was an increase in
metabolism of not far from 5 c.c. per minute, or about 2 to 2.5 per cent
Table 47. — Comparison of the gaseous exchange and lung ventilation of subject L., lying
on couch and sitting in chair. (Respiration apparatus.)
Date.
Day
of
fast.
Position.
No.
of
peri-
ods.
Time.
Car-
bon
diox-
ide
per
min-
ute.
Oxy-
gen
per
min-
ute.
Respi-
ratory
quo-
tient.
Respi-
ration
rate.
Lung
venti-
lation
per
min-
ute.1
Vol-
ume
per
inspi-
ration.2
Pulse-
rate.
1912.
L .pr. 16
. .pr. 23
. .pr. 24
. v.pr. 26
'..pr.27
2d
9th
10th
12th
13th
Lying
Sitting ....
3
2
8h 34m a.m. to 9h 37m a.m.
4h 00™ p.m. to 4h 35m p.m.
C.C.
180
179
c.c.
227
244
0.79
.73
10.9
10.3
liters.
5.18
5.58
C.C.
576
660
73
82
-1
17
-.6
.40
84
9
Lying
Sitting ....
3
2
8h25ma.m. to 9h18ma.m.
3h 52m p.m. to 4h 28m p.m.
143
129
190
187
0.75
.69
12.1
16.7
4.65
5.48
476
402
63
62
-14
-3
4.6
.83
-74
-1
Lying
Sitting ....
3
3
8h 19m a.m. to 9h 37m a.m.
3h58mp.m. to 4h57mp.m.
143
144
187
194
0.76
.74
10.9
14.6
4.55
5.83
504
484
63
69
1
7
3.7
1.28
-20
6
Lying
Sitting ....
3
3
8h 21m a.m. to 9h 26m a.m .
3h 13m p.m. to 4h llm p.m.
140
124
187
183
0.75
.68
12.8
15.8
4.64
5.37
429
404
61
60
-16
-4
3.0
.73
-25
-1
Lying
Sitting . . . .
3
2
8h 37m a.m. to 9h 33m a.m .
12h 14mp.m.to 12h48mp.m.
140
141
192
200
0.73
.71
12.8
12.8
4.63
5.55
437
525
59
68
1
8
0.0
.92
88
9
^he lung ventilation observed is here reduced to 0° C. and 760 mm. pressure.
Calculated to the pressure existing in the lungs and to 37° C.
340
A STUDY OF PROLONGED FASTING.
of the oxygen consumption. In the light of these varying results, it is
to be regretted that further observations were not made with the sub-
ject sitting quietly. However, a number of observations made when
the man was sitting up and writing actively may also be compared.
Influence op the Work op Writing.
On the fifth, fifteenth, seventeenth, twentieth, twenty-third, and
thirtieth days of the fast, the metabolism was studied while the subject
was sitting up writing, an employment that occupied much of his spare
time during the entire fast. These six experiments are compared in
table 48 with data obtained on the same day when the subject was
lying upon the couch in the morning experiment. Two of these experi-
ments— those on the seventeenth and twentieth days — immediately
Table 48. — Comparison of the gaseous exchange and lung ventilation of subject L., lying
on couch and sitting writing. (Respiration apparatus.)
Date.
Day
of
fast.
Position.
No.
of
peri-
ods.
Time.
Car-
bon
diox-
ide
per
min-
ute.
Oxy-
gen
per
min-
ute.
Respi-
ratory
quo-
tient.
Respi-
ration-
rate.
Lung
venti-
lation
per
min-
ute.1
Volumt
per
inspi-
ration.'
Pulse
rate.
1912.
Apr. 19
Apr. 29
May 1
May 4
May 7
May 14
5th
15th
17th
20th
23d
30th
Lying
Writing. . .
Increase
Lying
Writing . . .
Increase
Lying
Writing. . .
Increase.
4
2
8h21ma.m.to 9h32ma.m.
4h 10™ p.m. to 4h43mp.m.
ex.
158
198
c.c.
205
269
0.77
.74
11.8
17.9
liter 8.
4.88
7.54
c.c.
507
517
67
80
40
64
6.1
2.66
10
13
3
2
8h19ma.m.to 9h 19"1 a.m.
3h23mp.m.to 3h56mp.m.
132
164
179
233
0.74
.70
12.3
18.7
4.55
7.88
446
510
57
68
32
54
6.4
3.33
64
11
3
2
8h22ma.m.to 9h 1&» a.m.
9h 31m a.m. to 10h 04m a.m.
130
153
182
215
0.71
.71
12.3
14.6
4.81
6.57
471
542
57
69
23
33
2.3
1.76
71
12
Lying
Writing . . .
Increase.
3
2
8h22ma.m. to 9h 15™ a.m.
9h 35m a.m. to 10h 10" a.m.
124
141
173
208
0.72
.68
14.3
15.3
4.90
6.22
413
490
58
65
17
35
1.0
1.32
77
7
Lying
Writing. . .
Increase.
3
2
8h15ma.m.to 9h 17m a.m.
3h43mp.m. to 4h14mp.m.
121
159
165
222
0.73
.72
14.0
16.1
4.76
7.62
410
573
58
69
38
57
2.1
2.86
163
11
Lying
Writing . . .
3
2
8h03ma.m. to 8h56ma.m.
6h32mp.m.to 7h02mp.m.
119
156
166
221
0.72
.71
14.8
17.8
4.80
8.05
391
546
59
75
37
55
3.0
3.25
155
10
xThe lung v
Calculated
sntilat
to the
ion observed is here reduced
pressure existing in the lung
to0°<
i and 1
3. and
,o37°
760 m]
C.
n. pres
sure.
THE RESPIRATORY EXCHANGE. 341
followed the morning experiments, and it is again to be regretted that
this routine was not carried out in all cases.
There was a noticeable increase in the metabolism in all of the writing
experiments, which is shown by both the carbon-dioxide excretion and
the oxygen consumption, but the respiratory quotient tended to become
a few points lower during the writing period. The pulse- and respira-
tion-rates were both invariably increased, the increase in the pulse-rate
ranging from 7 to 16 beats per minute. The ventilation of the lungs
per minute likewise increased very considerably during the writing
experiment, this increase at times amounting to 3£ liters. The volume
per inspiration did not increase materially, except on the twenty-third
and thirtieth days of the fast.
Under ordinary conditions one would normally expect an increased
metabolism in the afternoon over the morning, but the comparison
previously made between the metabolism for the lying and sitting posi-
tions showed no increase for the sitting position in some instances and
in others there was an actual decrease. Consequently a sharp compar-
ison is difficult to make for the writing experiments.
We may assume from these experiments, however, that when the
subject was writing there was invariably an increased metabolism, but
that on the 2 days when the writing experiment immediately followed
the lying experiment the increase was only 50 to 75 per cent of that
obtained on the days when the writing experiment was in the afternoon.
On the seventeenth and twentieth days of fasting, when the writing
experiments were in the morning, the increase in the metabolism due to
writing was represented by an increased consumption of about 35 c.c.
of oxygen, or not far from 20 per cent. The absolute maximum increase
of 64 c.c. above the lying position on the fifth day of the fast amounted
to approximately 30 per cent, but while the absolute increase on the
twenty-third and thirtieth days was a few cubic centimeters less, the
percentage increase was greatest, i. e., 34 and 33 per cent respectively.
The work of writing, therefore, produced a distinct increase in the
metabolism, which is shown not only by the increase in the carbon-
dioxide production and the oxygen consumption, but also by an
increase in the respiration-rate, the ventilation of the lungs, and the
volume per inspiration. Finally, there was a regular and distinct increase
in the pulse-rate. These values are used subsequently in computing
the probable metabolism of this subject during several hours in the
day when he sat in the balcony and wrote.
Influence of Breathing an Oxygen-rich Atmosphere.
On the twenty-eighth, twenty-ninth, and thirtieth days of the fast
a series of experiments was made in which the subject breathed an
oxygen-rich atmosphere varying from 95 to 75 per cent of oxygen.
At the beginning of each experiment the percentage of oxygen in the
342
A STUDY OP PROLONGED FASTING.
atmosphere was probably not far from 95 per cent. This fell off quite
rapidly through the experimental period, so that at the end of the
experiment the proportion of oxygen in the atmosphere was between
75 and 80 per cent. The results of these experiments are not included
in table 46, but are compared in table 49 with the morning experiments
in which the subject breathed a normal atmosphere. In each case
the oxygen experiment immediately followed the regular morning
experiment.
Table 49. — Comparison of the gaseous exchange and lung ventilation of subject L., breath-
ing different air mixtures. (Respiration apparatus, subject lying, in the morning.)
Date.
Day
of
fast.
Air mixture.
No.
of
peri-
ods.
Car-
bon
diox-
ide
per
min-
ute.
Oxy-
gen
per
min-
ute.
Respi-
ratory
quo-
tient.
Alveolar
carbon
dioxide.
Respi-
ration-
rate.
Lung
venti-
lation
per
min-
ute.1
Volume
per
inspi-
ration.2
Pulse-
rate.
1912.
May 12
May 13
May 14
28th
29th
30th
Oxygen-rich. . .
Increase. . . .
Oxygen-rich. . .
Increase. . . .
Oxygen-rich. . .
Increase. . . .
3
2
c.c.
124
137
c.c.
166
179
0.75
.77
p. cl.
3.66
3.43
14.8
13.6
liters.
5.04
5.50
c.c.
410
487
61
62
13
13
-.23
-1.2
.46
77
1
3
2
124
130
171
176
0.73
.74
3.67
3.34
14.1
14.0
4.96
5.44
426
471
63
61
6
5
-.33
-.1
.48
45
-2
3
2
119
130
166
169
0.72
.77
3.80
3.44
14.8
14.2
4.80
5.34
391
454
59
58
11
3
-.36
-.6
.54
63
-1
*The lung ventilation observed is here reduced to 0° C. and 760 mm. pressure.
Calculated to the pressure existing in the lungs and to 37° C.
This series of experiments maybe compared with an extensive research
made in this laboratory by Mr. H. L. Higgins on the influence upon
the metabolism of normal individuals of breathing oxygen-rich atmos-
pheres. In the previous experimenting it was found that there was no
evidence of the increased percentage of oxygen affecting the respiratory
exchange, but there was a slight tendency for the pulse-rate to decrease
with an increase in the percentage of oxygen. In the experiments with
L., observations were made of the effect on the carbon-dioxide excre-
tion, oxygen consumption, respiratory quotient, the respiration-rate,
lung ventilation, volume per inspiration, alveolar carbon dioxide, and
the pulse-rate.
The data given in table 49 show a slight increase in the metabolism
indicated by both the carbon-dioxide production and the oxygen con-
sumption. The respiration-rate had a tendency to fall off somewhat;
THE RESPIRATORY EXCHANGE. 343
the lung ventilation was considerably increased — the increase ranging
from 0.46 liter to 0.54 liter — and the volume per inspiration was like-
wise increased from 45 c.c. to 77 c.c. per inspiration. The pulse-rate
was but little affected.
It is clear, therefore, that, even with an emaciated subject after a
prolonged fast, the inhalation of an oxygen-rich atmosphere did not
materially affect the carbon-dioxide production and oxygen con-
sumption, the slight increase being not far from 7 c.c. or approximately
5 per cent. The increase in the lung ventilation was, however, con-
siderable, amounting to about 10 per cent. This has special signifi-
cance with reference to the clinical use of oxygen, since it has been
contended that one of the advantages in the clinical administration
of oxygen was the fact that the patient was able to secure a sufficient
amount of oxygen for the aeration of the blood with less effort than
would be required to pump the lungs full of ordinary room air. The
larger ventilation of the lungs in the oxygen experiments would imply
that the increased amount of oxygen from inhaling an oxygen-rich
atmosphere was obtained with a greater mechanical effort, which would
thus offset the supposed advantage of this method of securing a suitable
percentage of oxygen for the oxygenation of the blood.
Influence of Sleep.
The measurements of the metabolism in the bed-calorimeter experi-
ments were made for the most part when the subject was resting very
quietly. During a portion of the time he was asleep; it is equally
certain that during certain periods he was awake. The results obtained,
therefore, show the metabolism of a subject who was a part of the
time awake and a part of the time asleep.
The kymograph curves also show that there was but little muscular
activity, so this can be eliminated in studying the factors influencing
the metabolism. As the subject was fasting, the influence of the inges-
tion of food may likewise be left out of consideration. Hence we have
an ideal condition for studying the basal metabolism of the subject.
According to the prevailing belief, with a complete absence of mus-
cular activity and of food in the alimentary tract, the metabolism
should have a constant value. But, as was pointed out in discussing
the metabolism curves, the course is irregular, the minimum values
apparently being secured in the periods of the night when the subject
was in the deepest sleep. Notwithstanding the fact that it is believed
by many investigators that sleep per se has no influence upon the metab-
olism, such an influence is indicated in the curves shown.
In the morning experiments with the respiration apparatus, the
measurements of the metabolism were made under exactly the same
conditions as to the absence of muscular activity and the ingestion of
food as were the calorimeter experiments, the only difference being
344 A STUDY OF PKOLONGED FASTING.
that the measurements were made with another apparatus and the
subject was awake throughout the whole experiment. We have,
therefore, as a result of the experiments with the respiration apparatus,
the basal minimum metabolism of the subject when he was awake.
It is accordingly of great interest to compare the metabolism as
measured in the bed calorimeter during the night with the metabolism
as measured with the respiration apparatus in the morning immediately
after the subject has been taken out of the calorimeter chamber.
Such a comparison has been made in table 50, in which the average
carbon-dioxide production and oxygen consumption are given for the
bed-calorimeter experiments when the subject was a part of the time
asleep and a part of the time awake, and also for the minimum periods
when the subject was presumably in deep sleep. The average values
for the carbon-dioxide production and the oxygen consumption are
likewise given for the experiments with the respiration apparatus, in
which the subject was awake throughout the whole period. The com-
parison of these values is also made by noting the increase when the
subject was awake. In the same table the minimum level of the pulse-
rate observed in the bed-calorimeter experiments1 is compared with
the average pulse-rate for the experiments with the respiration appa-
ratus. The increase in the pulse-rate when the subject was awake in
the experiments with the respiration apparatus is also shown. Finally,
the respiratory quotients obtained in the bed-calorimeter experiments
and in the experiments with the respiration apparatus are compared.
In comparing the results obtained during the first 4 days of the
experiment, when the subject was taking food, we find that the carbon-
dioxide production was 32 c.c. less per minute on the first day than dur-
ing the minimum period in the bed calorimeter when the subject was
asleep. On the night of April 12-13, the carbon-dioxide production
was lower by 4 c.c. and on the following night by 9 c.c. per minute when
the subject was asleep than when awake. These values, however, pre-
sent no abnormalities, since they are easily explained by the previous
ingestion of carbohydrate-rich food in the evening meal. This resulted
in a greater production of carbon dioxide in the bed-calorimeter exper-
iment than was obtained in the experiment with the respiration appa-
ratus, when the subject was without breakfast and therefore in the
post-absorptive condition. The values for the oxygen consumption
followed approximately the same course. The pulse-rate, also, in two
instances was less in the morning experiment than it was the night
previous and in two others it was 3 and 9 beats higher.
However, the results obtained on days when food was taken are not
of such great interest as those obtained during the fasting period.
Beginning with the first night of the fast, it will be seen that in the
xSee curves showing the pulse-rate in the bed-calorimeter experiments in figures 12 to 18, on
pages 104 to 110.
THE RESPIRATORY EXCHANGE.
345
Table 50.
-Comparison of the gaseous exchange of subject L., in the bed calorimeter at night and
awake on the respiration apparatus in the morning.
Carbon dioxide
per minute.
Oxygen per minute.
Pulse-rate.
Respiratory
quotient.
i
« .
oj
0J
(4
03
a>
u
03
Date.
Day
of
fast.
Bed calori-
meter.
P.
ft
B
d tn
•2 5
■§ fi
u u
'ft
■
0J
•5 l
*o
03 o3
£ 1
o c3
a
i-i
Bed calori-
meter.
ft
ft
03
0 'A
O 3
11
ft
m
«
<0 •
If
•f.9
3 B
TS &
n
ft
ft
03
.2 2
03 u
.a
o.
03
V
H
o «
2 M
03 03
o o]
a
M
<0
£ +
T3
0J
«
ft—s
ft o
03 .|.
a o
o ^
h
ft g
0)
H
Aver-
age.
Mini-
mum.1
Aver-
age.
Mini-
mum.1
A
B
c
D
E
F
G
H
I
J
K
L
M
1912.
c.c.
c.c.
c.c.
c.c.
c.c.
c.c.
c.c.
c.c.
Apr. 10-1 13
224
228
218
180
165
218
217
196
173
152
186
196
200
182
185
-32
-21
4
9
33
276
258
252
221
212
265
246
235
208
196
231
220
225
223
237
-34
-26
-10
15
41
76
70
73
64
64
72
73
72
73
74
-4
3
-1
9
10
0.81
.88
.86
.81
.78
0.81
.89
.89
.82
.78
11-123
12-133
13-143
14-15
1st
15-16
2d
159
154
180
26
211
208
227
19
63
73
10
.75
.79
16-17
3d
151
148
169
21
206
198
226
28
60
70
10
.73
.75
17-18
4th
150
140
159
19
202
187
212
25
58
68
10
.74
.75
18-19
5th
143
137
158
21
192
176
205
29
59
67
8
.75
.77
19-20
6th
134
131
148
17
194
185
200
15
57
64
7
.68
.74
20-21
7th
135
132
153
21
190
185
204
19
56
64
8
.71
.75
21-22
8th
137
135
151
16
187
177
203
26
58
65
7
.73
.74
22-23
9th
134
131
143
12
178
173
190
17
57
63
6
.75
.75
23-24
10th
130
127
143
16
180
179
187
8
55
63
8
.72
.76
24-25
11th
128
124
140
16
176
166
187
21
54
61
7
.72
.75
25-26
12th
129
126
140
14
175
173
187
14
56
61
5
.73
.75
26-27
13th
126
125
140
15
171
167
192
25
54
69
5
.74
.73
27-28
14th
120
116
134
18
167
160
181
21
51
58
7
.72
.74
28-29
15th
117
114
132
18
163
162
179
17
51
57
6
.71
.74
29-30
16th
117
114
133
19
165
158
182
24
52
58
6
.71
.73
Apr. 30-May 1
17th
115
113
130
17
160
154
182
28
49
57
8
.72
.71
May 1-2
18th
115
112
123
11
159
154
174
20
51
56
5
.72
.71
2-3
19th
113
111
127
16
158
153
177
24
50
57
7
.71
.72
3-4
20th
114
112
124
12
160
159
173
14
51
58
7
.71
.72
4-5
21st
112
103
126
23
154
137
174
37
51
59
8
.73
.73
5-6
22d
111
109
124
15
154
153
170
17
51
59
8
.72
.73
6-7
23d
112
106
121
15
156
144
165
21
53
58
5
.72
.73
7-8
24th
109
106
122
16
158
152
167
15
53
59
6
.69
.73
8-9
25th
111
108
125
17
153
147
166
19
53
60
7
.72
.75
9-10
26th
111
106
123
17
159
151
168
17
54
61
7
.70
.73
10-11
27th
111
107
129
22
153
145
172
27
55
62
7
.72
.75
11-12
28th
115
109
124
15
162
145
166
21
57
61
4
.71
.75
12-13
29th
112
104
124
20
158
152
171
19
55
63
8
.72
.73
13-14
30th
110
103
119
16
151
147
166
19
55
59
4
.72
.72
14-15
31st
115
109
120
11
160
148
166
18
54
60
6
.72
.72
16-173
124
188
117
176
133
172
16
-4
154
194
143
182
170
183
27
1
60
84
72
84
12
0
.80
.97
.78
.94
17-183
xThe duration of the periods in which these minimum values were observed varies in general from 3
hours to 1 hour.
Represents period of lowest pulse-rate observed. See page 112.
8On the days preceding and following the fast the night experiments were made after the ingestion
}f food. The subject was without breakfast during the morning respiration experiments.
346 A STUDY OF PROLONGED FASTING.
morning experiment 33 c.c. more of carbon dioxide were produced than
during the period of deep sleep in the night. There was likewise an
excess of 41 c.c. of oxygen consumed and an increase in pulse-rate of
10 beats. This increase in the gaseous metabolism and the pulse-rate
persists throughout the whole 31 days of the fast, for on no day was the
metabolism during the morning experiment, when the subject was
awake, less than during the minimum period of the night experiment,
when the subject was in deep sleep. Furthermore, in no instance was
an average value secured in the morning experiment with the respira-
tion apparatus which was less than the average value for the whole
experiment with the bed calorimeter, when the subject was asleep a
part of the time and awake a part of the time.
It will be observed that this increase in the metabolism and in the
pulse-rate underwent extreme variations, the range for the carbon-
dioxide production being from 33 c.c. on the first night to 11 c.c. on
the eighteenth and thirty-first nights, and for the oxygen consumption
from 41 c.c. on the first night to 8 c.c. on the tenth night. Never-
theless there was a fairly average value of 17 c.c. for the increase in
the carbon-dioxide production, the variation as the fast progressed
being not far from this value, except on the twenty-first, twenty-
seventh, and twenty-ninth nights. With the oxygen consumption,
the variations are much more irregular; these may partly be explained
by the difficulties of determining the oxygen consumption exactly,
especially when the values for the minimum periods are selected. It
is likewise important to note that the oxygen consumption for the
morning is invariably higher than not only the minimum during the
night but also the average for the whole night.
The increase in the pulse-rate undergoes a much more regular change.
On the first 4 mornings of the fast, the pulse-rate is 10 beats higher when
the subject is awake than when he is asleep during the night. This
increase throughout the fast remains reasonably constant at 8 beats.
On but two occasions, i. e., on the twenty-eighth and the thirtieth
nights, was the increase only 4 beats.
It is thus clear that, so far as the absolute metabolism is concerned,
there is an increase in the period when the subject lay awake on the
respiration apparatus over that of the average for the night, when the
subject slept for a part or the whole of the time, and particularly over
the minimum for the night.
In the post-fasting period, the metabolism on May 16-17 continued
to be greater in the morning experiment than in the minimum period
for the night experiment. On May 17-18, however, when the subject
was in much distress and had consumed a large amount of carbohydrate
material, a considerable proportion of which had been retained, the
carbon-dioxide production was less in the morning than at night and
the oxygen consumption practically the same. There was likewise no
THE RESPIRATORY EXCHANGE. 347
change in the pulse-rate of 84, which was for this subject a very high
pulse-rate.
If we consider the respiratory quotients obtained in these experi-
ments, we find that the differences shown in the gaseous metabolism
and the pulse-rate practically disappear. This would naturally be
expected, since the respiratory quotient indicates the character of the
katabolism and there is no reason why this should change essentially
during the fasting experiment, regardless of whether the subject was
inside the bed calorimeter or lying on the couch with the respiration
apparatus. On the first four nights with food we find essentially the
same respiratory quotients with both apparatus, while during the
fasting nights there is a decrease in the respiratory quotients. The
respiratory quotients for the bed-calorimeter experiments are on the
average a little lower than those obtained with the respiration appa-
ratus, particularly in the first half of the fast, but this may be due to
a slight error in the determination of the residual amount of oxygen
in the chamber as a result of faulty measurements of the water-vapor
in the chamber air.1 In any event, we believe that the two series of
respiratory quotients are comparable, and it will be seen that they follow
essentially the same course throughout the fast. An excellent demon-
stration of the validity of these comparisons is the fact that, on the two
nights with food after the fast, the increased respiratory quotient of 0.80
in the bed-calorimeter experiment was accompanied by a quotient of
0.78 in the experiment with the respiration apparatus the following
morning, while on the next night the very high quotient of 0.97 in the
bed-calorimeter experiment was followed by a quotient of 0.94 in the
experiment with the respiration apparatus. It is thus seen that the
respiratory quotient shows essentially no difference in the character
of the katabolism as determined by either the bed calorimeter or the
respiration apparatus. Furthermore, the fact that the respiratory
quotient has the same general value in both instances is an excellent
demonstration of the probable trend of this factor, since these quotients
were obtained by two methods, within a few hours of each other, so
that each figure is an excellent control upon the other.
There are two possible explanations of the differences in the meta-
bolism as measured in the bed calorimeter and with the respiration
apparatus. It may be reasonably questioned (1) whether or not the
two apparatus measure the respiratory exchange with the same degree
of exactness, and (2) whether or not the subject was more active when
awake on the respiration apparatus than when asleep in the bed calori-
meter.
The measurement of the metabolism by these two apparatus has
been carefully compared in a long series of experiments carried out by
Mr. T. M. Carpenter. Certain of these comparison tests have already
^ee page 308.
348 A STUDY OF PKOLONGED FASTING.
been published,1 but a still larger number are being prepared by Mr.
Carpenter for early publication. His results give a complete answer
to the first criticism, as they show that the measurement of the metabo-
lism by the bed calorimeter is essentially the same as that with the
respiration apparatus.
Furthermore, since it has already been shown that the character of
the katabolism as indicated by the respiratory quotient was accurately
determined by both of these apparatus, it is unreasonable to suppose
that there would be any material difference in the measurements of
the total metabolism, especially as the two apparatus are constructed
on the same principle, i. e.,the Regnault-Reiset closed-circuit principle.
If one apparatus had been built on the closed-circuit principle and the
other upon the open-circuit principle, an error in the measurement of
the total ventilation or of the total volume of air passing through the
chamber would affect the measurement of the total metabolism without
affecting the respiratory quotient. Since the two apparatus were
built upon the same principle, however, the possibility of such an error
would be very small and any question of a fundamental difference
between the results obtained with these two forms of apparatus may
therefore be excluded from this discussion.
In considering the second criticism, namely, that there might be a
difference in activity when the subject was awake and asleep, it should
be pointed out that each experiment in both series was accompanied
by a graphic record of the degree of muscular repose of the subject
by means of the movable bed previously described.2 If the records
are compared, it will be seen that during the night there were slight
movements, as would be natural, since no individual could lie in
absolutely the same position for a period of 10 or 12 hours. On
the other hand, the record for the experiment with the respiration
apparatus in the morning was almost invariably a straight line, showing
that the subject had not moved. Mr. Carpenter, who was in charge
of the respiration experiments, was of the opinion that the subject was
entirely indifferent to the mechanical part of the respiration apparatus
and showed not the slightest evidence of muscle tension or appre-
hension, which might increase the pulse-rate or general muscle tonus.
In other words, this subject unquestionably approximated, as nearly as
any subject that we have ever experimented with, the conditions
described by Johansson3 as "Vorsatzliche Muskelruhe." It is, further-
more, clear that the average results obtained in the two series of experi-
ments show that as the fast progressed there was a tendency for the
values to decrease and essentially at the same rate. This would hardly
be expected if the subject were under any great mental strain or anxiety
Benedict and Joslin, Carnegie Inst. Wash. Pub. 136, 1910, p. 173.
2See page 311.
'Johansson, Skand. Archiv f. Physiol., 1898, 8, p. 85.
THE RESPIRATORY EXCHANGE. 349
and it would probably not continue for 31 days. L. frequently said
that the experiments were very simple and occasionally remarked that
they were quieting and restful. It would appear, therefore, that he
was as nearly unaffected by the experimental routine as any person
could be, was without anxiety or apprehension, and that there was no
evidence of tenseness of muscles.
It thus appears that the only evidence we have of muscular activity
during the two series of experiments is the slight activity shown by the
kymograph records for the experiments with the bed calorimeter. But
such activity would tend to increase the metabolism during the night,
when the subject was a part of the time asleep and a part of the time
awake, over that when the subject was fully awake on the respiration
apparatus. The difference between the results would thus be even
greater if the subject were absolutely quiet throughout the night in
the bed calorimeter.
Since there was no difference in the methods of measurement and
such activity as existed would tend to lessen rather than increase the
difference between the results, it seems evident that the difference in the
metabolism during the sleeping and waking conditions must be due to
the influence of sleep, an influence which has hitherto been disregarded
by experimenters. No series of experiments with which we are familiar
shows so completely and in such a controlled manner this striking
difference in the metabolism between a subject asleep and a subject
awake. The pulse-rate is likewise increased during the waking period,
being fairly uniform with the increases in the carbon-dioxide production
and oxygen consumption. This completely substantiates our view that
the pulse-rate gives an admirable index of the internal activity of the
body, which largely determines the basal metabolism.
To show more clearly the increase in the metabolism when the sub-
ject is awake over the metabolism when he is asleep, the percentage
increases in the carbon-dioxide production, oxygen consumption, and
pulse-rate are given in table 51 for each day of the experiment. Dis-
regarding the results for the first 4 nights, which are complicated by the
influence of food upon the metabolism during the sleeping period, we
find on examining the values for the fasting period that there is invari-
ably a considerable percentage increase for the carbon dioxide, oxygen,
and pulse-rate. For instance, the percentage increase in the carbon-
dioxide production rises as high as 21.7 percent on the morning following
the first night and falls as low as 9.2 per cent on the morning after the
ninth night. In general, in the latter part of the fast there was not far
from 15 per cent increase in the carbon-dioxide production when the
subject was awake over that when he was asleep in the minimum period
of the calorimeter experiment. Similar fluctuations were observed in
the oxygen consumption, but with hardly the regularity noted in the
values for the carbon-dioxide production. During the latter part of
350
A STUDY OF PROLONGED FASTING.
the fast, the average oxygen consumption of the subject awake was
not far from 13 to 14 per cent greater than during sleep.
While there is no particular reason to give the increase in the pulse-
rate on a percentage basis and expect that it would have a value cor-
responding to those for the carbon-dioxide production and oxygen con-
sumption, for want of a better comparison at the moment it seems
Table 51. — Increase in metabolism of subject awake as compared with
metabolism of subject asleep.
Date.
Day of
fast.
Carbon
dioxide.
Oxygen.
Pulse-
rate.
1912
Apr. 10-11
p. ct.
—14.7
—9.7
2.0
5.2
21.7
16.9
14.2
13.6
15.3
13.0
15.9
11.9
9.2
12.6
12.9
11.1
12.0
15.6
15.8
16.7
15.0
9.8
14.4
10.7
22.3
13.8
14.2
15.1
15.7
16.0
20.6
13.8
19.2
15.5
10.1
13.7
—2.3
p. ct.
—12.8
—10.6
--4.3
7.2
20.9
9.1
14.1
13.4
16.5
8.1
10.3
14.7
9.8
4.5
12.7
8.1
15.0
13.1
10.6
15.2
18.2
13.0
15.7
8.8
27.0
11.1
14.6
9.9
12.9
11.3
18.6
14.5
12.5
12.9
12.2
18.9
0.5
p. ct.
—5.3
4.3
—1.4
14.1
15.6
15.9
16.7
17.2
13.6
12.3
14.3
12.1
10.5
14.5
13.0
8.9
9.3
13.7
11.8
11.6
16.3
9.8
14.0
13.7
15.7
15.7
9.4
11.3
13.2
13.0
12.7
7.0
14.6
7.3
11.1
20.0
0.0
11-12
12-13
13-14
14-15
15-16
16-17
17-18
18-19
19-20
20-21
21-22
22-23
23-24
24-25
25-26
26-27
27-28
28-29
29-30
Apr. 30-May 1
May 1-2
2-3
3-4
4- 5
5- 6
6- 7
7-8
8- 9
9-10
10-11
11-12
12-13
13-14
14-15
16-17
1st
2d
3d
4th
5th....
6th
7th....
8th....
9th....
10th....
11th....
12th
13th
14th....
15th....
16th....
17th....
18th....
19th
20th....
21st
22d
23d , ,
24th....
25th....
26th
27th
28th....
29th
30th
31st
17-18
advisable to present it in this way. Thus we note here an increase in
the pulse-rate ranging from 17.2 per cent on the morning following the
fourth night of fasting to as low as 7 per cent after the twenty-eighth
night. In the latter part of the fast the pulse-rate is not far from 12
per cent above that found during the sleeping period. There was
practically no increase in the carbon-dioxide production, the oxygen
THE RESPIRATORY EXCHANGE. 351
consumption, or the pulse-rate on the last morning of observation fol-
lowing the taking of food, i. e., May 17-18.
An average of the increases for the 31 days shows that the carbon-
dioxide production increased 14.7 per cent, the oxygen consumption
13.2 per cent, and the pulse-rate 12.8 per cent. It is perhaps somewhat
surprising to note that although there are individual variations on the
different days of the fasting experiment and that a given increase
in the pulse-rate is not always accompanied by the same percentage
increase in the carbon-dioxide production and oxygen consumption,
nevertheless the average percentage increase for the carbon-dioxide
production, oxygen consumption, and pulse-rate shows a most striking
uniformity. While this is not the first time in this laboratory that an
attempt has been made to establish a percentage relationship between
the pulse-rate and the metabolism, it would appear that in these experi-
ments the increase in metabolism was directly proportional to the
increase in the pulse-rate. Under ordinary conditions of laboratory
experimentation, in which factors other than those governing the basal
minimum metabolism enter, it is hardly probable that this sharp mathe-
matical regularity would obtain, but these closely agreeing results are
of considerable significance.
These figures are strikingly at variance with those found by other
observers, particularly Johansson1 and Loewy.2 The experiments of
Johansson are wholly unique, in that he possesses, as probably no other
living man does, the power to relax completely his own muscles, thereby
lowering his metabolism to a minimum value. His experiments are,
however, distinctly vitiated by the fact that his entire series of measure-
ments is based upon the carbon-dioxide output alone, and while con-
siderable care was given to secure regularity in the ingestion of food, it
is a fact that the carbon dioxide is not an ideal index of the total katabo-
lism. From the experience in this laboratory with various subjects,
in which we have had more or less definite information as to the metabo-
lism of a subject when asleep and when awake, we are perfectly con-
vinced that the metabolism during deep sleep is profoundly affected
by the sleeping condition and is much lower than the metabolism when
the subject is awake.
METABOLISM PER UNIT OF WEIGHT AND SURFACE.
We have seen in the previous discussion that the most common
factors affecting the metabolism — external muscular activity and the
ingestion of food — were lacking when the metabolism was studied in
this fast, and that the fluctuations which were observed must have
been due to other factors; also, that the metabolism was not constant
throughout the day. Even under conditions when the subject was
Johansson, Skand. Archiv f. Physiol., 1898, 8, p. 116.
2Loewy, Berlin klin. Woch., 1891, p. 434.
352 A STUDY OP PROLONGED FASTING.
lying perfectly quiet, there was still an absence of constancy, for it
was found that the metabolism was much lower during the hours of
deep sleep at night than in the morning, when the subject was lying
upon a couch and connected with the respiration apparatus; also that
the metabolism during the evening just prior to the night calorimeter
experiment was higher than in the morning. We thus find a definitely
established daily rhythm, with the minimum metabolism in the early
morning not far from 2 to 4 a. m., a somewhat increased metabolism
between 8 a. m. and 9h 30m a. m., and a still higher metabolism
between 7 p. m. and 8 p. m. As the measurements in all cases were
made under conditions when there was muscular repose and no food
in the alimentary tract, it is clear that a certain factor not ordinarily
considered influences the daily rhythm. This is unquestionably the
factor which we may, for want of a better term, designate as " internal
muscular work" or cellular activity. When the stimulus to this cel-
lular activity is increased, the subject shows a correspondingly higher
metabolism.
As we have already found, the pulse-rate is an admirable index to
this tonicity or cellular activity, since the closest correlation has been
shown to exist between the pulse-rate and the metabolism. This is
perhaps no more strikingly brought out than in comparing the metabo-
lism when the subject was asleep with the metabolism when the subject
was awake. While, therefore, the three series of experiments in which
the metabolism was determined were not primarily designed to throw
light upon the daily rhythm, inasmuch as the main purpose was the
study of the alterations in the metabolism as the fast progressed, yet
since the data were obtained within the 24 hours, they offer a good
demonstration of this diurnal variation in the metabolism and a hint
as to its nature. Furthermore, they show the intimate relationship
between the pulse-rate and the diurnal variations.
During the 24 hours there was of course no material alteration in
body-weight or in body-surface, and hence the data need not be com-
pared upon any special basis other than that of the body as a whole.
On the other hand, as the fast progressed there were certain fundamental
changes taking place which should be considered in any attempt to
interpret the variations in the metabolism noted during the fast. In
the first place, the man was obviously losing weight every day. As was
pointed out in the discussion of the losses in body-weight of L. during
the fast, it would have been possible with a fasting animal to arrange
the conditions of the fast so that the loss in weight would have been
regular; with a fasting man, however, it was impossible so to control
the daily activity, the ingestion of water, the collection of the urine,
and the environmental temperature that the loss in weight would
follow a mathematical curve; and yet, as has already been shown, it
proceeded with a considerable degree of regularity.
THE RESPIRATORY EXCHANGE. 353
As the weight decreases during a fast, there will unquestionably be
a change in the body-surface. With certain individuals such loss of
weight, either through old age or illness, is not accompanied by a cor-
responding shrinkage of the skin, and the surface of the skin is conse-
quently wrinkled and hangs in large folds. In general, however, this
is not the case; that such a condition certainly did not exist with our
fasting subject is clearly shown by the photographs given in plates
4 and 5. Consequently with L. there was undoubtedly a decrease
in the radiating body-surface. These factors of decreasing body-weight
and body-surface may reasonably be expected to play a role in the
metabolism during a long fast, while in a fast of but 24 hours they would
be negligible.
Using as indices the changes in the pulse-rate, the blood-pressure,
and the pulse-pressure, we find that there was also a considerable
variation in the internal muscular activity. Not only do we find
variations in the pulse-rate between the conditions of lying asleep and
lying awake, but as the fast progressed we find that there were likewise
changes from day to day in these values. This is clearly brought
out in table 50. On the other hand, the pulse-rate did not continually
decrease, for, as was pointed out in a previous discussion of the changes
in the pulse-rate, there was a period in which the pulse-rate fell rather
rapidly, followed by a period when it remained approximately constant,
while toward the end of the fast there was a tendency for the pulse-
rate to increase. These changes in the pulse-rate indicate a consider-
able alteration in the internal muscular activity of the body as the fast
continued, thus clearly establishing a factor in metabolism which has
heretofore been almost neglected. It is the purpose of this section to
examine more closely some of the various factors which affected the
metabolism during this prolonged fast.
METABOLISM PER KILOGRAM OF BODY-WEIGHT.
In the attempt to find some unit of comparison, it has long been the
custom of many writers to use the kilogram of body-weight, presumably
on the ground that with animals differing in size, the larger animal
would normally have the greater metabolism. In other words, it has
been the custom to assume that the metabolism per kilogram of body-
weight is essentially the same for most animals of the same species.
As a result of using this unit certain discrepancies have appeared which
have been recognized by many writers, but nevertheless this method
of comparison is still adhered to and, indeed, individuals of widely
varying body-weight have been compared in this way.
If we could determine the composition of the human body, we should
certainly find great differences in individuals with different body-
weights. With this fasting man we have an ideal condition for study-
ing the metabolism per kilogram of body-weight in that we have an
354 A STUDY OF PROLONGED FASTING.
organism continually losing weight from the beginning to the end of
the period of fasting. Furthermore, we can tell with a reasonable
degree of accuracy the character of these losses and thus secure some
indication as to the probable composition of the human body at the
beginning and the end of the fast.
As the fast continues, the changes in the body-weight show a loss
of body material. It has been demonstrated, in previous fasting
studies,1 that at the beginning of a fast this loss consists in large part
of water, much of which is preformed water. During the first two
days of the fast there is unquestionably a further loss of several hundred
grams of carbohydrates in the form of glycogen. Subsequently, the
loss is of fat ; there is also a fairly regular loss of protein from day to
day, but after the first few days of the fast the loss is chiefly fat and
water. Thus the first 5 kilograms lost from the body in a 31-day fast
would certainly be of greatly different composition from the last 5
kilograms lost. The composition of the organism is therefore not the
same on the tenth day of fasting, for instance, as on the first day, and
varies considerably as to the absolute amounts of fat, carbohydrate,
and protein. Consequently, to compare the metabolism on the basis
of body-weight is wholly illogical, and although this method of compari-
son is habitually used by many writers, it is certainly inconsistent with
their knowledge of the character of the body losses.
The character of the body material lost may be determined with
considerable accuracy. The loss of protein may be computed from the
nitrogen found in the urine; the loss of carbohydrate and fat may be
computed from the respiratory quotient and the carbon dioxide pro-
duced or the oxygen consumed, making due correction for the carbon
dioxide produced and oxygen consumed in the combustion of the pro-
tein; the amount of water lost may also be found by modern technique.
But in considering changes in the metabolism, we are dealing not with
the material lost from the body, but with the body material remaining
in the organism, and to determine the composition of the body at any
given period has been found very difficult. While it may be reasonable
to attribute any difference in the total metabolism for the first and
thirty-first days of the fast to the metabolism that would normally
belong to the material lost, this will be true only when we are
assured that the living tissue in each case had precisely the same effi-
ciency as to the production of heat and the maintenance of the vital
processes.
Certain evidence that has been brought forward in discussing the
pulse-rate, and particularly the comparison of the metabolism for a
subject lying awake and lying asleep, leads us to believe that there are
influences affecting the total heat-production, entirely aside from the
organized mass of heat-producing material. Thus we may not say
Benedict, Carnegie Inst. Wash. Pub. 77, 1907, p. 468.
THE RESPIRATORY EXCHANGE. 355
that the subject on the thirty-first day of fasting, with a weight of
2 kilograms less, has the same metabolism as on the twenty-fourth day,
for the organism at the end of the fast is living on a higher metabolic
plane, as is evidenced by the higher pulse-rate. Consequently a strict
comparison of the results on the basis of the metabolism per kilogram
of body-weight is precluded.
It is commonly considered that the active heat-forming mass of the
body is not found in the fatty tissue nor in the water, but in the organ-
ized protoplasmic tissue. If we could assume, for example, that a
fasting man when losing weight could lose only fat and water and no
organized nitrogenous material, one would expect that as the fast
progressed the metabolism per kilogram of body-weight would increase,
for while the original mechanism for the production of heat would not
alter in any way, the inert material (fat and water) which hitherto
contributed to the body-weight, and thus reduced the heat output per
kilogram of body-weight, would be removed and the heat output per
kilogram should accordingly be increased.
Two important factors militate against this assumption. In the
first place, it is impossible for a man to fast for any number of days
without a considerable loss of nitrogenous tissue. This may or may
not be derived from active protoplasmic tissue, but it certainly is in
part a loss of heat-producing tissue. On the other hand, there are
known instances when very large amounts of nitrogenous material
have been fed to individuals and a considerable proportion of the nitro-
gen has been retained by the body in some form without apparently
changing the value of the heat-producing mechanism, since the heat-
production per kilogram of body-weight did not alter. The most
notable instance of this is the experiment reported by Mueller Hn Vienna,
who increased the nitrogen content of the body of his subject by 210
grams in 28 days, and yet was unable to obtain the slightest increase
in the metabolism per kilogram of body-weight. Apparently the nitro-
gen added to the body did not enter into the active protoplasmic tissue
or contribute to the heat-producing qualities of the body as a whole.
Reference has already been made to several remarkable experiments
with dogs carried out by Awrorow, in 1898, in the Imperial Medical
Academy in St. Petersburg.2 These dogs fasted for periods ranging
from 16 to 66 days, without water, and remained for 22 hours out of
each day inside the Pashutin respiration calorimeter, being catheterized
daily. The carbon-dioxide production was measured by absorption
in potassium hydroxide and the heat-production by the Pashutin
calorimeter. These observations of Awrorow are of such importance
in this connection that it seems advisable to reproduce the charts for
two of the experiments,3 i. e., those for dogs No. 2 and No. 3, in which
the fast continued for 44 days and 60 days respectively.
Mueller, Zentrlb. f. d. ges. Physiol, u. Path, dea Stoff., 1911, 6, p. 617.
2See page 78. 3See footnote on page 79.
356
A STUDY OF PROLONGED FASTING.
The curves for the body-weight will be recognized as comparable to
those given in figure 3 (page 77), although it will be noted that in these
two charts the percentage of loss in body-weight is plotted, while in fig-
ure 3 the actual body-weights are plotted. The total heat-production
and the total carbon-dioxide production for 24 hours fell rapidly for
the first 3 days with dog No. 2 and for 7 days with dog No. 3. There-
OAYS OF FAST 0
AWROROW'S EXPERI
5 10 15 20
Fig. 45. — Complete metabolism chart of fasting dog (Awrorow No. 2).
The calories from fat may be found by deducting the calories from protein from the total
calories as indicated by the scale.
THE RESPIRATORY EXCHANGE.
AWROROW'S EXPERIMENT N0.3.
357
DAYS OF FAST 5
^^^^^^^^^Mj&^gk™21£Ba«Ui22
j. ., ^v^^^^^SSS^^SSSsSs"
Fig. 46. — Complete metabolism chart of fasting dog (Awrorow No. 3).
The calories from fat may be found by deducting the calories from protein from the total
calories as indicated by the scale.
358 A STUDY OF PROLONGED FASTING.
after there was a period of 4 or 5 days of equal production and sub-
sequently a more or less regular fall until the end of the fasting period.
It is thus seen that the organisms of these dogs acted not unlike that of
our fasting man, that is, the total heat-production and the total carbon-
dioxide output decreased with general regularity as the fast progressed.
On the other hand, in both experiments, the curves for the carbon-
dioxide and heat-production per kilogram of body-weight rise with
great regularity throughout the entire fast, falling only on the last 2
or 3 days. Unfortunately, Awrorow does not give the pulse-rate, but
from the curves for the temperature given in the upper part of the
charts it will be seen that the sharp fall in the last few days of the fast,
in both the total heat-production and the heat-production per kilogram
of body-weight, was coincidental with a rapidly falling temperature.
Since both dogs died, it is probable that they were in a moribund con-
dition in the last day or two of the experiments. On the other hand,
it is of interest to note that there was no appearance of a premortal
rise in the nitrogen excretion.
Since these dogs had a uniform external muscular activity during
their stay in the respiration chamber and probably a uniform internal
muscular activity, we deal here only with the relationship between
cellular activity and the total body-weight. The increase in the car-
bon-dioxide and the heat-production per kilogram of body-weight
found in these experiments indicates most strongly a resistance to
destruction of the heat-producing mechanism in the body which was
wholly disproportionate to the losses in body-weight. It is thus seen
that in the experiments with these dogs, in which the metabolism was
unaffected by muscular work or by the ingestion of food, this distinct
conservation of organized material had a marked influence; accordingly,
the heat-production per kilogram of body-weight was not constant,
but, as a matter of fact, increased as the inert water and fat were lost.
Such ideal conditions for experimenting are obviously impossible with
men, and even with Awrorow's dogs there was the disturbing element of
falling body-temperature on the last few days, which affected pro-
foundly both the total heat-output and the heat-output per kilogram
of body-weight. Since, however, it is the custom of writers to give
the heat-production per kilogram of body-weight, it has been necessary
to use this basis in computing the metabolism for the experiment with
L. in order that the values may be comparable with those of other
investigators. The results of these computations are given in tables
52, 53, and 54, which show the metabolism per kilogram of body-weight
per minute for the experiments with the bed calorimeter and with the
respiration apparatus. In the belief that a question of such funda-
mental importance should be considered from every standpoint, not
only the average results are given for the bed-calorimeter experiments
(see table 52), but also the results for the minimum periods obtained
with that apparatus (see table 53). The values for the morning experi-
ments with the respiration apparatus are given in table 54.
THE RESPIRATORY EXCHANGE.
359
Metabolism per Kilogram of Body- Weight in the Calorimeter Experiments.
Considering first the metabolism as indicated by the average values
for the carbon-dioxide production and oxygen consumption during the
night in the bed calorimeter (table 52), we find that the carbon-dioxide
Table 52. — Metabolism per kilogram of body-weight and per square meter of body-surface
(Meeh) in experiments with L. (Bed calorimeter at night.)
Average carbon
Average
dioxide.
oxygen.
Date.
Day
of
fast.
Weight
without
cloth-
Body-
sur-
face.
Aver-
age
pulse-
Aver-
age
body-
tem-
Per
Per
kilo-
Per
square
Per
Per
kilo-
Per
square
mg.1
min-
gram
meter
min-
gram
meter
rate.
per-
ute.
per
min-
ute.
per
hour.
ute.
per
min-
ute.
per
hour.
ature.
1912
kilos.
8q. m.
ex.
c.c.
gin.
c.c.
c.c.
am.
°C.
Apr. 10-112
60.37
1.90
224
3.71
13.89
276
4.57
12.45
82
11-122 .
60.82
1.90
228
3.75
14.14
258
4.24
11.64
76
12-132
61.19
1.91
218
3.56
13.45
252
4.12
11.31
78
13-142
60.87
1.91
180
2.96
11.11
221
3.63
9.92
70
14-15.. . .
1st..
69.86
1.88
165
2.76
10.34
212
3.54
9.67
68
15-16....
2d..
58.91
1.86
159
2.70
10.07
211
3.68
9.72
66
36.41
16-17....
3d..
58.01
1.84
151
2.60
9.67
206
3.55
9.60
62
17-18.. . .
4th.
57.22
1.83
150
2.62
9.66
202
3.53
9.46
65
36.55
18-19.. . .
5th.
56.53
1.81
143
2.53
9.31
192
3.40
9.09
63
36.58
19-20....
6th.
56.01
1.80
134
2.39
8.77
194
3.46
9.24
60
36.44
20-21.. . .
7th.
55.60
1.79
135
2.43
8.89
190
3.42
9.10
59
36.42
21-22.. . .
8th.
55.18
1.78
137
2.48
9.07
187
3.39
9.00
61
36.55
22-23....
9th.
54.74
1.77
134
2.45
8.92
178
3.25
8.62
59
36.50
23-24.. . .
10th.
54.25
1.77
130
2.40
8.66
180
3.32
8.72
57
36.64
24-25.. . .
11th.
53.94
1.76
128
2.37
8.57
176
3.26
8.57
57
36.80
25-26.. . .
12th.
53.64
1.75
129
2.40
8.69
175
3.26
8.57
58
36.85
26-27.. . .
13th.
53.48
1.75
126
2.36
8.49
171
3.20
8.38
56
36.62
27-28.. . .
14th.
53.22
1.74
120
2.25
8.13
167
3.14
8.23
53
36.30
28-29.. . .
15th.
52.92
1.74
117
2.21
7.92
163
3.08
8.03
53
36.43
29-30.. . .
16th.
52.40
1.73
117
2.23
7.97
165
3.15
8.17
53
36.40
Apr. 30-May 1
17th.
51.91
1.71
115
2.22
7.93
160
3.08
8.02
52
36.42
May 1- 2.. . .
18th.
51.57
1.71
115
2.23
7.93
159
3.08
7.97
62
36.30
2- 3....
19th.
51.21
1.70
113
2.21
7.83
158
3.09
7.97
52
36.21
3- 4....
20th.
50.97
1.69
114
2.24
7.95
160
3.14
8.11
52
36.51
4- 5....
21st..
50.60
1.69
112
2.21
7.81
154
3.04
7.81
54
36.12
5- 6... .
22d..
50.22
1.68
111
2.21
7.79
154
3.07
7.86
53
36.10
6- 7....
23d..
50.00
1.67
112
2.24
7.90
156
3.12
8.01
56
35.98
7- 8.. . .
24th.
49.70
1.67
109
2.19
7.69
158
3.18
8.11
55
35.88
8-9....
25th.
49.40
1.66
111
2.25
7.88
153
3.10
7.90
55
36.31
9-10.. . .
26th.
49.10
1.65
111
2.26
7.93
159
3.24
8.26
56
10-11....
27th.
48.78
1.64
111
2.28
7.98
153
3.14
8.00
57
36.03
11-12....
28th.
48.52
1.64
115
2.37
8.26
162
3.34
8.47
59
36.37
12-13....
29th.
48.19
1.63
112
2.32
8.10
158
3.28
8.31
58
36.23
13-14....
30th.
47.79
1.62
110
2.30
8.00
151
3.16
7.99
58
36.06
14-15....
31st..
47.47
1.61
115
2.42
8.42
160
3.37
8.52
57
36.14
16-17*
47.37
1.61
124
2.62
9.08
154
3.25
8.20
64
36.79
17-183
48.40
1.64
188
3.88
13.51
194
4.01
10.14
90
37.53
'Calculated from the weights observed each morning (see table 2) by adding one-quarter of
the loss on each day to the weight obtained on the morning following the night experiment.
2The subject had eaten as usual during the day.
3The fast was ended with the taking of food on the morning of May 15. The subject ate at
intervals throughout the subsequent days.
360 A STUDY OF PROLONGED FASTING.
production per kilogram of body-weight ranged from 3.88 c.c. per min-
ute (the extraordinarily high value found on May 17-18) to 2.19 c.c.
per minute on the twenty-fourth day of fasting. Excluding the days
with food, the highest value found was on the first night of the fast,
namely, 2.76 c.c. per minute. The values show a distinct tendency to
fall until the fourteenth day of the fast, with a subsequent essentially
constant production of carbon dioxide until the twenty-sixth day.
From that time until the end of the fast the carbon-dioxide produc-
tion tends to be somewhat higher per kilogram of body- weight.
The highest value for the oxygen consumption was found on April
10-11, the first night which the subject spent inside the respiration
chamber, i. e., 4.57 c.c. per minute, and the lowest value was 3.04 c.c.
on the twenty-first night of the fast. Excluding the periods when food
was taken, the highest value, 3.58 c.c, was obtained on the second
night of the fast. The oxygen consumption per kilogram of body-
weight followed a course which was not unlike that of the carbon-
dioxide production, that is, a persistent fall until the fourteenth day
of the fast, and thereafter an approximate constancy, with a tendency
toward a rise from the twenty-sixth day to the end of the fast.
In the previous comparison of the pulse-rate and the total metabo-
lism, it was found, in the latter part of the fast, that there was a ten-
dency for the pulse-rate to rise which was unaccompanied by an increase
in the total carbon-dioxide output and oxygen intake. When the
values are computed on the basis of per kilogram of body-weight, the
course of the carbon-dioxide and the oxygen is found to be strikingly
similar to that of the pulse-rate. At this point, one noticeable anomaly
in the otherwise nearly constant relationship between the pulse-rate
and the metabolism should be emphasized. On the last night of the
fast, the oxygen consumption was 3.37 c.c. per kilogram per minute,
while the pulse-rate was 57. On the second night thereafter (namely,
on May 16-17) the oxygen consumption was somewhat less, being only
3.25 c.c. per minute, while there was an increase of 7 beats per minute
in the pulse-rate. This discrepancy is in part explained by taking
into consideration the differences in the calorific value of oxygen with
the different respiratory quotients; but we still have a discrepancy
which is striking, the only one of any magnitude noted in the long series
of observations. Its explanation is not simple.
The average values found throughout the night are complicated by
varying conditions of sleeping, waking, and some muscular activity;
hence, for the most accurate comparison, use should be made of the
values given in table 53, which represent the metabolism per kilogram
of body-weight in the minimum periods of the night. With this basis
of comparison we find that in certain instances the carbon-dioxide
production does not follow closely the oxygen consumption, but the
general picture of the progress of the metabolism during the fast is not
THE RESPIRATORY EXCHANGE.
361
altered by these few discrepancies. The pulse-rates given in table 53
should be used with considerable reserve, since it is difficult to select
them intelligently on this basis, as the minimum period for the pulse-
rate may not have been coincident with the minimum period for either
the carbon-dioxide production or the oxygen consumption. It is
safe to conclude, however, that the general trend of the pulse-rate is
such that its use in this general comparison is not illogical.
Table 53. — Minimum metabolism per kilogram of body-weight and per square meter of body-
surface (Meeh) in experiments with L. (Bed calorimeter at night.)
Minimum carbon dioxide.
Minimum oxygen.
Mini-
mum
Date.
Day
of fast.
Per
Per
kilogram
Per
square
Per
Per
kilogram
Per
square
pulse-
rate
(subject
asleep) .*
minute.
per
meter
minute.
per
meter
minute.
per hour.
minute.
per hour.
1912
c.c.
c.c.
am.
c.c.
c.c.
firm.
Apr. 10-112
218
217
196
173
152
3.61
3.57
3.20
2.84
2.54
13.52
13.46
12.09
10.67
9.53
265
246
235
208
196
4.39
4.04
3.84
3.42
3.27
11.95
11.10
10.55
9.33
8.94
76
70
73
64
64
11-122
12-132
13-142
14-15
1st
15-16
2d
154
2.61
9.76
208
3.53
9.58
63
16-17
3d
148
2.55
9.48
198
3.41
9.22
60
17-18
4th
140
2.45
9.02
187
3.27
8.76
58
18-19
5th ... .
137
2.42
8.92
176
3.11
8.33
59
19-20
6th ... .
131
2.34
8.58
185
3.30
8.81
57
20-21
7th ... .
132
2.37
8.69
185
3.33
8.86
56
21-22
8th....
135
2.45
8.94
177
3.21
8.52
58
22-23
9th
131
2.39
8.72
173
3.16
8.38
57
23-24
10th ....
127
2.34
8.46
179
3.30
8.67
55
24-25
11th....
124
2.30
8.30
166
3.08
8.08
54
25-26
12th
126
2.35
8.49
173
3.23
8.47
56
26-27
13th ....
125
2.34
8.42
167
3.12
8.18
54
27-28
14th
116
2.18
7.86
160
3.01
7.88
51
28-29
15th....
114
2.15
7.72
162
3.06
7.98
51
29-30
16th ....
114
2.18
7.77
158
3.02
7.83
52
Apr. 30-May 1 . . . .
17th ....
113
2.18
7.79
154
2.97
7.72
49
May 1- 2
18th ....
112
2.17
7.72
154
2.99
7.72
51
2-3
19th ....
111
2.17
7.69
153
2.99
7.71
50
3-4
20th ....
112
2.20
7.81
159
3.12
8.06
51
4-5
21st
103
2.04
7.18
137
2.71
6.95
51
5-6
22d
109
2.17
7.65
153
3.05
7.81
51
6-7
23d
106
2.12
7.48
144
2.88
7.39
53
7-8
24th
106
2.13
7.48
152
3.06
7.80
53
8-9
25th
108
2.19
7.67
147
2.98
7.59
53
9-10
26th....
106
2.16
7.57
151
3.08
7.84
54
10-11
27th....
107
2.19
7.69
145
2.97
7.58
55
11-12
28th ....
109
2.25
7.83
145
2.99
7.58
57
12-13
29th....
104
2.16
7.52
152
3.15
7.99
55
13-14
30th
103
2.16
7.49
147
3.08
7.78
55
14-15
31st
109
2.30
7.98
148
3.12
7.88 .
54
16-173
117
176
2.47
3.64
8.56
12.65
143
182
3.02
3.76
7.61 '
9.51
60
84
17-183
'See page 112.
2The subject had eaten as usual during the day.
8The fast was ended with the taking of food on the morning of May 15.
intervals throughout the subsequent days.
The subject ate at
362 A STUDY OF PROLONGED FASTING.
The highest value for the carbon-dioxide production during the
minimum periods is found, as with the average values, on the last day
of the experiment, May 17-18, being 3.64 c.c. per minute. The lowest
value is 2.04 c.c. on the twenty-first day of fasting. Disregarding the
days with food, we then find the highest value to be 2.61 c.c. on the
second day of the fast. There is a general tendency for the carbon-
dioxide production per kilogram of body-weight to decrease until the
fourteenth day, when there is a marked fall ; from that time till the end
of the fast the carbon-dioxide production remains essentially constant,
with the exception of the low value found on the twenty-first day of
fasting and the high values on the twenty-eighth and thirty-first days.
Owing to the many difficulties in determining the oxygen consump-
tion in short periods — difficulties which have already been empha-
sized1— the selection of the minimum periods for the oxygen consump-
tion is not easy and hence we must not expect to find the regularity
in these values that we should find in the values for the carbon-dioxide
output. The figures given in table 53 have, however, been selected
with due care for accuracy. The values ranged from 4.39 c.c. for the
first night the subject spent in the chamber to 2.71 c.c. for the twenty-
first night. The highest value found for the fasting period was 3.53
c.c, on the second night of the fast. While there is a general tendency
for the oxygen values to decrease for the first 15 days of the fast and
thereafter to remain essentially constant, one may hardly generalize
from such irregular figures.
The pulse-rates, which are taken from table 6, show a regular fall
until the fourteenth day, when there is a drop of 3 points. The pulse
then stays at the low value of 50 or 51 until the twenty-third day and
subsequently shows a definite tendency to increase slowly. The pulse
curve is therefore not strictly parallel to that for the carbon-dioxide
production per kilogram per minute. The discrepancy previously
found in the relationship between the pulse-rate and the average oxy-
gen consumption also appears in the values given here for the minimum
periods. Thus, on the last night, the oxygen consumption was 3.12 c.c.
per minute and the pulse-rate 54 ; on the second night with food, the
oxygen value was only 0.1 c.c. less or 3.02 c.c, whereas the pulse-rate
increased 6 beats per minute. Since it was noted that the minimum
values for the body-temperature were not strictly comparable with the
metabolism on this basis,we have not included these values in this table.
Metabolism per Kilogram of Body-Weight in Respiration-Apparatus Experiments.
The difficulties in comparing the values obtained in the bed calori-
meter experiments, due to the varying conditions of waking, sleeping,
and some muscular activity, disappear when we consider the values
obtained in the experiments with the respiration apparatus. In these
^ee page 324.
THE RESPIRATORY EXCHANGE.
363
experiments we have complete muscular repose and a relatively regular
pulse-rate in each group of experimental periods ; hence we may properly
average the values for the metabolism per kilogram of body-weight
from day to day. These values have been computed and are given in
table 54, together with the average pulse-rate and body-temperature for
each experiment. Of the three series of values compared, this set is
probably the most comparable, since we have the greatest muscular
repose, controlled graphically by the movable bed, and the subject is
always awake. Furthermore the carbon-dioxide production, the oxy-
gen consumption, and the pulse-rate are simultaneously determined.
Table 54. — Metabolism per kilogram of body-weight and per square meter of body-surface
(Meeh) in experiments with L. (Respiration apparatus, subject lying in the morning.)
Carbon dioxide.
Oxygen.
Aver-
age
body-
tem-
Date.
Day
of
fast.
Weight
without
cloth-
Body-
sur-
face.
Aver-
age
per
Per
kilo-
gram
Per
square
meter
Aver-
age
per
Per
kilo-
gram
Per
square
meter
Aver-
age
pulse-
rate.
ing.
min-
per
per
min-
per
per
per-
ute.
min-
ute.
hour.
ute.
min-
ute.
hour.
ature.
1912
kilos.
sq. m.
c.c.
c.c.
gm.
c.c.
c c.
gm.
°C.
Apr. 10-11. . .
60.13
1.89
186
3.09
11.60
231
3.84
10.48
72
11-12. . .
60.53
1.90
196
3.24
12.16
220
3.63
9.92
73
12-13. . .
60.95
60.64
59.60
1.91
1.90
1.88
200
182
185
3.28
3.00
3.10
12.34
11.29
11.60
225
223
237
3.69
3.68
3.98
10.10
10.06
10.80
72
73
74
13-14. . .
14-15. . .
1st..
15-16. . .
2d..
68.68
1.86
180
3.07
11.40
227
3.87
10.46
73
16-17. . .
3d..
57.79
1.84
169
2.92
10.82
226
3.91
10.53
70
17-18. . .
4th.
57.03
1.82
159
2.79
10.30
212
3.72
9.98
68
18-19. . .
5th.
56.37
1.81
158
2.80
10.29
205
3.64
9.71
67
19-20. . .
6th.
55.89
1.80
148
2.65
9.69
200
3.58
9.52
64
20-21. . .
7th.
55.50
1.79
153
2.76
10.07
204
3.68
9.77
64
21-22. . .
8th.
55.08
1.78
151
2.74
10.00
203
3.69
9.77
65
22-23. . .
9th.
54.63
1.77
143
2.62
9.52
190
3.48
9.20
63
23-24. . .
10th.
54.13
1.76
143
2.64
9.58
187
3.45
9.11
63
24-25. . .
11th.
53.88
1.76
140
2.60
9.37
187
3.47
9.11
61
25-26. . .
12th.
53.56
1.75
140
2.61
9.43
187
3.49
9.16
61
26-27. . .
13th.
53.45
1.75
140
2.62
9.43
192
3.59
9.40
59
27-28. . .
14th.
53.15
1.74
134
2.52
9.08
181
3.41
8.92
58
28-29. . .
15th.
62.84
1.73
132
2.50
8.99
179
3.39
8.87
67
29-30. . .
16th.
52.26
1.72
133
2.54
9.11
182
3.48
9.07
58
Apr. 30-Mayl
17th.
51.79
1.71
130
2.51
8.96
182
3.51
9.12
57
36.73
May 1- 2. . .
18th.
51.50
1.70
123
2.39
8.53
174
3.38
8.77
66
36.54
2- 3. . .
19th.
51.11
1.70
127
2.48
8.80
177
3.46
8.92
67
36.55
3- 4. . .
20th.
50.93
1.69
124
2.43
8.65
173
3.40
8.77
58
36.85
4-5...
21st..
50.49
1.68
126
2.50
8.84
174
3.45
8.88
59
36.39
5- 6. . .
22d..
50.13
1.67
124
2.47
8.75
170
3.39
8.72
59
36.63
6-7...
23d..
49.96
1.67
121
2.42
8.54
165
3.30
8.47
58
36.24
7- 8. . .
24th.
49.62
1.66
122
2.46
8.66
167
3.37
8.62
59
36.11
8-9...
25th.
49.33
1.66
125
2.53
8.87
166
3.37
8.57
60
36.59
9-10. . .
26th.
49.02
1.65
123
2.51
8.79
168
3.43
8.73
61
10-11...
27th.
48.70
1.64
129
2.65
9.27
172
3.53
8.99
62
36.04
11-12...
28th.
48.46
1.64
124
2.56
8.91
166
3.43
8.68
61
36.59
12-13. . .
29th.
48.10
1.63
124
2.58
8.97
171
3.55
8.99
63
36.42
13-14. . .
30th.
47.69
1.62
119
2.50
8.66
166
3.48
8.78
59
36.18
14-15. . .
31st..
47.39
1.61
120
2.53
8.78
166
3.50
8.84
60
36.42
16-17. . .
47.12
48.17
1.61
1.63
133
172
2.82
3.57
9.74
12.44
170
183
3.61
3.80
9.05
9.62
72
84
37.12
37.64
17-18. . .
364 A STUDY OF PROLONGED FASTING.
As in the calorimeter experiments, we again find the highest record
for the carbon-dioxide production in the last observation of the series,
namely, 3.57 c.c. per minute on May 18. The lowest record was 2.39
c.c. per minute on May 2. Disregarding the food days, we find the
highest value of 3.10 c.c per minute on April 15, at the end of the first
day of fasting. The course of the carbon-dioxide excretion per minute
was remarkably regular in the first part of the fast, there being an
almost steady fall until the minimum >value was reached on May 2;
subsequently there was some fluctuation, with a general tendency
toward a slight rise.
The highest value for the oxygen consumption during the fasting
period (3.98 c.c.) was also found at the end of the first fasting day, and
the lowest (3.30 c.c.) was reached on the twenty-third day. The course
of the oxygen consumption was in a large measure parallel with that
of the carbon-dioxide production, the values showing a slight tendency
toward a rise during the last week.
It is of special interest to note that the curve for the pulse-rate pre-
pared on this basis follows with remarkable fidelity the curves for the
carbon-dioxide output and oxygen consumption, although the positive
differences in the pulse-rate after the tenth day of fasting are very small,
the minimum value being 56 and the highest 63. Our recent experi-
ence in this laboratory with the photographic registration of the pulse-
rate, and particularly with the duration of the pulse-cycle, leads us to
believe that our inability to secure registration of the pulse-rate on
account of the breaking of the string galvanometer was a loss much
greater than was realized at the time of experimentation. The pulse-
rate records were made with as great accuracy as would ordinarily
be expected with the stethoscope, but the differences are so small that
photographic registration would have been most desirable.
Conclusions Regarding the Metabolism per Kilogram of Body-Weight.
In these three series of comparisons it will be seen that the metabo-
lism per kilogram of body-weight fell rapidly during the first half of the
fast, then remained constant for a time, and during the last week
usually increased slightly. During the entire fasting period the body
lost 13.25 kilograms of body material, of which (according to our best
estimates) there were 7.33 kilograms of water, 201 grams of carbo-
hydrate, 3,650 grams of fat, and 1,664 grams of protein.1 It is clear,
therefore, that a considerable proportion of the loss, or 87.4 per cent,
was made up of material other than protein. It has commonly been
believed that the heat-producing organism of the body consists in large
part of protein. Under these circumstances, we should expect the
heat-production per kilogram of body-weight to increase as the fast
progressed, for the total loss of material was much greater in proportion
than the loss of active protoplasmic tissue could have been.
iSee tables 2, 61, 62, and 63.
THE RESPIRATORY EXCHANGE. 365
An examination of the records of the pulse-rate shows, however, that
another factor influenced the metabolism, namely, the stimulus to the
internal muscular activity or the muscle tonus. The exact nature of
this stimulus is difficult at this time to state. Short periods of fasting
have shown the presence of acidosis, and an acidosis experimentally
induced by giving a carbohydrate-free diet has been found to increase
the metabolism.1 On the other hand, it is known that diabetics tolerate
a considerable acidosis without great increase in the metabolism, and
the theory has been advanced that the diabetic organism becomes
immune to or reacts less vigorously to the stimulus of the acidosis.2
While there was unquestionably an acidosis in the 31-day fast, it
is highly probable that its influence upon the metabolism must have
been practically constant or continually decreasing. Nevertheless,
during the last week there was a distinct increase in the pulse-rate,
accompanied by a constancy in the metabolism, thus giving striking
evidence that the organism was being influenced by some factor of an
internal nature. This constancy in the total metabolism may possibly
account for the increase in the metabolism per kilogram of body-weight
during the last week of the fast, which is similar to the striking increase
observed throughout the entire fasting experiments with the dogs
studied by Professor Awrorow.
During the first 14 days of the fast there was certainly no constancy
between the carbon-dioxide output and the body-weight, for the
carbon-dioxide output per kilogram of body-weight was continually
decreasing. The fact that for two weeks or more the carbon-dioxide
output and the oxygen intake per kilogram of body-weight remained
constant should not be taken as a demonstration of an intimate rela-
tionship between body-weight and metabolism, since the metabolism
had a tendency to increase during the last week. A more intelligent
interpretation of the phenomena would be that the total metabolism
decreased by virtue of the fact that there was a decrease not only in the
size of the heat-producing organism, but in the intensity of the cellular
activity, as was indicated by the pulse-rate. When this activity was
increased, then the relationship between body-weight and metabolism
was entirely obscured. If it were possible to have an experiment in
which the pulse-rate remained constant and the change in body-weight
was wholly due to inert material, then the values for the heat-production
per kilogram of body-weight could logically be compared.
The decrease in the total heat-production as the fast progressed was
therefore only in part due to the fact that there was a decrease in the
size of the heat-producing organism. With a constant size of organism,
as would obtain on any given experimental day during the fast, there
still were material differences in the heat production and the intensity
of cellular activity which were evidenced by the change in the pulse-
rate and by the accompanying noticeable increase in the metabolism.
Benedict and Joslin, Carnegie Inst. Wash. Pub. 176, 1912, p. 125. 2Ibid., p. 134.
366 A STUDY OF PROLONGED FASTING.
While, therefore, the computation of the metabolism on the basis
of per kilogram of body-weight is justifiable as a gross index of the
probable metabolism of the organism, for the proper interpretation of
the values found in this research this method of computation is wholly
without avail. It is furthermore of great importance to note that
the level of intensity of metabolism in any given organism may have
equal, if not indeed greater, influence upon the total metabolism than
may the body-weight. It is illogical, therefore, to compare subj ect A with
a pulse-rate of 70 with subject B with a pulse-rate of 70 and assume
that the metabolism per kilogram of body-weight should be the same.
On the other hand, subject A, with a pulse-rate of 70, will almost
invariably have a higher metabolism than with a pulse-rate of 60.
Evidence with regard to the influence of long-continued changes in
the body-weight, either as a result of fasting, as a result of diminished
diet in obesity cures, or as a result of the increased difference of body
material following periods of fasting, are thus far too few to enable us
to make general deductions. The enormous addition of nitrogen
to the body in the experiments of Mueller1 without a corresponding
increase in the metabolism per kilogram of body-weight suggests a
line of research that should prove most profitable. This fundamentally
important observation should certainly be repeated and with every
control. It is sufficient here only to point out that while custom has
sanctioned the general usage of the method of computing the metabo-
lism on the basis of per kilogram of body-weight, such computation
can have but little significance when a careful scientific analysis of the
results of metabolism experiments is desired.
METABOLISM PER SQUARE METER OF BODY-SURFACE.
Since the heat-production of an animal body varies as to its size,
physiologists have long sought to establish some relationship between
them. One of the earliest views was that primarily based upon Newton's
law of cooling; this is set forth in the simplest form by Bergmann2
to the effect that the surface of the body is a much more logical factor
for considering the needs for heat-production than any other.
Bergmann's original discussion is of interest here :
' "Die Oberflache ist ein einfacher und genau zu ermittelnder Factor fur die
Warmeverluste, dessen Werth, zusammengenommen mit der Beschaffenheit
dieser Oberflache (Bedeckung mit Haaren u. s. w.), der Differenz zwischen
Temperatur des Thieres und des umgebenden Mediums und Beschaffenheit
dieses Mediums (ob es Luft oder Wasser ist) die Warmeverluste bestimmt.
"Das Volumen des Thieres dagegen wird als ein Maass fur die mogliche
Warmebildung betrachtet werden konnen. Gewiss ist in gleichem Volumen
sowohl verschiedener Thiere als auch desselben Thieres zu verschiedener Zeit
die Warmebildung sehr verschieden. Aber man wird es nicht gewagt finden,
Mueller, Zentrlb. f. d. ges. Physiol, u. Path, des Stoff., 1911, 6, p. 617.
2Bergmann, Ueber die Verhaltnisse der Warmeokonomie der Thiere zu ihrer Grdsse., Got-
tingen, 1848, p. 9.
THE RESPIRATORY EXCHANGE. 367
wenn wir annehmen, dass es fiir die Warmebildung ein Maximum gebe, in der
Art, dass ein gewisses Quantum animalischer Substanz im lebenden Korper
nicht im Stande ist, mehr als ein gewisses Quantum Warme in einer gegebenen
Zeit zu liefern.
"Nun vergrossern oder vermindern sich ja der cubische Inhalt von Korpern
und die Ausdehnung ihrer Oberflache nicht nach demselben Verhaltnisse,
sondern, wenn wir die einzelnen Dimensionen eines Korpers z. B. sammtlich
im Verhaltnisse von 1 zu 2 vergrossern, so wachst die Oberflache von 1 zu 4
und der cubische Inhalt von 1 zu 8.
"Es ist also entschieden, dass die Thiere, je grosser sie sind, um so weniger
Warme im Verhaltniss zu ihrer Grosse zu bilden brauchen, um eine gewisse
Erhohung ihrer Temperatur iiber die der Umgebung zu gewinnen."
Three decades later, Rubner1 amplified this idea and presented
evidence, secured in researches with his calorimeter, to show that the
heat-production was proportional to the body-surface and, indeed,
that this law held true with practically all warm-blooded animals.
An ingenious method of explaining the apparent relationship between
the metabolism and the size of the body-surface was put forth by von
Hosslin.2 Nevertheless Rubner's view has obtained and has been the
basis of much discussion since it was first put forth.
The true significance of the apparent relationship between metabo-
lism and body-surface has recently been extensively discussed in con-
nection with the report of a long series of observations on infants
recently published from this laboratory,3 in which it was shown that
with infants of approximately normal average weight, height, and age
not only was the heat-production per square meter of body-surface by
no means constant, but that with atrophic infants the disturbance of
the relationship between body-surface and metabolism was far too great
to be accounted for by the usual method of explanation, namely, that
there was a disturbance in the formula for computing the body-surface.
It was pointed out that in all probability the active mass of proto-
plasmic tissue played a dominant role and that the body-surface was
simply a normal function of growth, and with normal conditions of
nourishment bore a simple mathematical relationship to the body-
weight. According to the recent observations of Dreyer and his
associates,4 the same mathematical relationship obtains between body-
weight and various other physiological constants of the body, par-
ticularly with the cross-section of the trachea, the aorta, and more
particularly the blood volume. While, therefore, the general law of
the relationship between the body-surface and metabolism may hold
approximately for animals of normal growth and weight, the cause for
this relationship is not, we believe, due to Newton's law of cooling, but
Rubner, Zeitschr. f. Biol., 1883, 19, p. 545.
2Von Hosslin, Archiv f. Anat. u. Physiol., 1888, p. 323.
'Benedict and Talbot, Carnegie Inst. Wash. Pub. 201, 1914, p. 157.
4Dreyer and Ray, Phil. Trans., 1909-1910, 201, ser. B., p. 133; Dreyer, Ray, and Walker,
Proc. Roy. Soc, 1912-1913, 86, Ser. B, pp. 39 and 56.
368 A STUDY OF PROLONGED FASTING.
to the fact that the heat-producing organism of the body, of which
the blood volume is probably a fair index, bears a similar relation to
the law of growth as does the body-surface.
It was believed that further information might be secured on this
important point in the experiment with the fasting" man, and since the
heat-production and likewise the oxygen consumption and carbon-
dioxide production are discussed by many physiologists on the basis
of per square meter of body-surface, such a form of computation has
been used here, the values being given in tables 52, 53, and 54.
With the fasting man we have a heat-producing organism which is
continually decreasing in weight and consequently should decrease in
surface. In our calculations it has been assumed that the usual
formula for computing the body-surface, namely, that of Meeh, in
which the surface is equal to 12.312 y/W2, holds for each weight as
the fast progresses. In former publications from this laboratory, in dis-
cussing the metabolism of greatly emaciated patients, particularly
diabetics, the contention has been made that there may be a lack of
proportion between body-weight and body-surface which is not cor-
rectly considered by this formula. But with this fasting man, as has
been pointed out previously, and indeed with diabetics studied in this
laboratory, there has been no evidence that the skin did not shrink
in proportion to the loss in weight, inasmuch as there was no folding of
the skin which indicated such a disturbance in the relationship.
Rubner,1 in his experiments on a fat boy and a thin boy, found that
by computing the surface of the thin boy by Bouchard's formula he
obtained 10,480 sq. cm. instead of 10,730 sq. cm. as computed by the
Meeh formula. With the fat boy the discrepancy was much greater, the
Bouchard formula giving 13,522 sq. cm. and the Meeh formula 14,554
sq. cm. In any event the maximum error was only about 7 per cent.
For the sake of argument, it may be considered that on the first day
of the fast the body-surface of the subject L. was properly indicated
by the Meeh formula, but judging from Rubners experience with his
fat boy, the Meeh formula would give too high a value. As the fast
progressed, the error with the Meeh formula would become less and
less, until toward the end of the fast it would give the true body-
surface. It might then be considered that at the beginning of the
experiment the computed value for the body-surface was actually
somewhat large and that the computed value for the metabolism per
square meter of body-surface would consequently be slightly too small.
Nevertheless, the fact that the error can not be more than 7 to 10 per
cent should be borne in mind. Furthermore, any correction of this
nature will tend to raise the values for the metabolism per square meter
of body-surface in the early part of the fast, but will not affect those
found for the latter part of the fast.
Rubner, Beitrage zur Ernahrung im Knabenalter mit besonderer Beriicksichtigung der Fett-
sucht. Berlin, 1902, p. 40.
THE EESPIRATORY EXCHANGE. 369
With this correction in mind, a comparison may be made of the
metabolism found with the bed calorimeter and the respiration appa-
ratus on the basis of per square meter of body-surface.
Metabolism per Square Meter of Body-Surface in Calorimeter Experiments.
Table 52 gives the average metabolism per square meter of body-
surface as computed from the values found with the bed calorimeter.
On this basis of computation, the carbon-dioxide production ranged
from 13.89 grams on the first night inside the respiration chamber to
7.69 grams on the twenty-fourth night. Disregarding the food days,
the highest value was 10.34 grams on the first night of the fast. On the
last food night of the observation, the carbon-dioxide production was
essentially as high as it was on the first 3 nights prior to the fast,
namely, 13.51 grams per square meter per hour.
Considering the values for the oxygen consumption per square meter
of body-surface, we find that it varied from 12.45 grams on the first
night inside the respiration chamber to a minimum of 7.81 grams on
the twenty-first night. Again disregarding the food days, the highest
value observed is 9.72 grams on the second night of the fast.
If the law holds true that the metabolism per square meter of body-
surface remains constant, it is surprising that the variation in the car-
bon-dioxide production during the fast is so great, i. e., 10.34 grams to
7.69 grams — a decrease of 26 per cent. An approximately similar
decrease is noted in the oxygen consumption per square meter of body-
surface. Considering the values for both the oxygen consumption and
the carbon-dioxide production, we find that the gaseous exchange per
square meter of body-surface remained essentially constant from the
fifteenth day to the twenty-seventh day of the fast, but in the last 4
days there was a distinct tendency for the gaseous exchange to increase
slightly, even on the basis of per square meter of body-surface. While
the values were not in complete uniformity with the curve for the pulse-
rate, they showed a distinct tendency to follow the same course. No
clearly established relationship, however, can be found between the
body-temperature and the metabolism per square meter of body-surface.
Since the objections previously cited in reference to comparisons of
the average values for the gaseous exchange may apply with equal force
in this connection, the gaseous exchange per square meter of body-surface
in the selected minimum periods may also be compared on this
basis. These values are given in table 53. The carbon-dioxide pro-
duction ranged from 13.52 grams on the first night inside the chamber
to 7.18 grams on the twenty-first night; but if the food days are dis-
regarded, the maximum value was 9.76 grams on the second night of
the fast. We find here, however, that there is still a range of 26 per
cent in the metabolism per square meter of body-surface, essentially
the same as that shown with the average values. The difficulties in
370 A STUDY OF PROLONGED FASTING.
making proper selections for the periods for the oxygen consumption
have already been outlined, but we find, from such selections as we have
been able to make, that the values per square meter of body-surface
ranged from 9.58 grams on the second night of the fast to 6.95 grams
on the twenty-first night.
A comparison of the values for the gaseous exchange with the records
for the pulse-rate, also given in table 53, must necessarily be made with
reserve, owing to the character of the selection as pointed out in the
comparison of the values per kilogram of body-weight, but it will be
noted that, at least during the first part of the fast, there was a distinct
tendency for the metabolism per square meter of body-surface to follow
approximately the curve of the pulse-rate. Indeed, even in these
minimum selected periods, there was a tendency for the metabolism to
rise in the last week of the fast, although this tendency was by no
means as pronounced here as in the values given in table 52.
Metabolism per Square Meter op Body-Surpace in the Respiration- Apparatus
Experiments.
Perhaps the most satisfactory comparison is that of the values
obtained with the respiration apparatus in the morning, just after
the subject came from the calorimeter chamber, since the conditions
were constant in all of the experiments, i. e., simultaneous measure-
ments of the carbon-dioxide production, oxygen consumption, and
pulse-rate, when the subject was without food in the alimentary tract
and there was complete absence of muscular activity. Under these
conditions the values given in table 54 show a range in the carbon-
dioxide production per square meter of body-surface from 12.44 grams
on May 18 (the last day of observation) to 8.53 grams on the eighteenth
day of the fast. Disregarding the values for the food days before and
after the fast, the maximum value of 11.6 grams is found at the end of
the first fasting day. With the oxygen consumption per square meter
of body-surface essentially the same results are found, i. e., a maximum
of 10.80 grams at the end of the first fasting day and a minimum of
8.47 grams at the end of the twenty-third day.
These figures indicate that the metabolism per square meter of body-
surface tends distinctly to decrease in unison with the pulse-rate during
the first half of the fast. In the last 15 days the fluctuation was not so
sharply marked, though there was a tendency toward high values in the
last week coincidental with the slight tendency to a rise in pulse-rate.
Conclusions reqardino the Metabolism per Square Meter op Body-surface.
From a consideration of the fundamental theory regarding the metab-
olism per square meter of body-surface, it will be seen that the values
should be constant, assuming that there is no error in the formula for
computing the body-surface from the body-weight. But even if a cor-
THE RESPIRATORY EXCHANGE. 371
rection of 6 or 7 per cent is made according to Rubner's observations,
this serves only to increase the high values found in the early part
of the fast and to leave unaltered the low values in the latter part, thus
making the discrepancy even more striking than before. It is perhaps
of more than passing significance that there was a distinct tendency
for the metabolism per square meter of body-surface to increase slightly
in the last week of the fast, an increase that was accompanied by a
slight, though persistent, increase in the pulse-rate. It would appear
that this measurable increase in the pulse-rate must be an index of in-
creased cellular activity, which of itself is sufficient to account for the
increased total metabolism, irrespective of body-surface or body-weight.
The distinct decrease in the metabolism as the fast progressed, par-
ticularly in the first part of the fast, is wholly in conformity with the
experience with the fasting subject in Middletown, Connecticut. In
the two fasts of 5 and 7 days, respectively, the oxygen consumption and
the carbon-dioxide production on the basis of both per kilogram of
body-weight and per square meter of body-surface decreased as the
fast progressed. Owing to the fact that there were unavoidable
differences from day to day in the muscular activity of the continuous
sojourn inside the chamber, selected periods, namely, from 1 a. m.
to 7 a. m., when the subject was, if not sound asleep, resting quietly
in bed, were used for comparison.1 In the report of these experiments
it is clearly shown that in fasting experiments Nos. 71, 73, and 75 there
was a pronounced tendency for the metabolism to decrease both per
kilogram of body-weight and per square meter of body-surface as the
fast progressed. This was noticeably true in the first 15 days of the
fast with the subject L., but in the latter part of the fast the metabo-
lism was essentially constant.
With the Middletown subject S. A. B., the pulse-rate showed a
tendency to fall regularly throughout the entire fast. With both the
subjects S. A. B. and L. there was for the most part a striking uni-
formity between the pulse-rate and the metabolism measurements.
The observations on L. have shown that on any given day the pulse-
rate is a strikingly constant index of the total metabolism and to a
certain extent gives us a distinct idea of the metabolism as the fast
progresses. At the same time it must be borne in mind that the rela-
tionship between the pulse-rate and the metabolism is not mathemati-
cally so firmly established that it can be said to apply to different
individuals. If we consider that L. was continually changing as the
fast progressed, and therefore a new organism was being studied each
day, it is especially surprising that the pulse-rate is so close an index
of the metabolism. This is of particular significance in considering the
values for the last week of the fast, for while the rise in the metabolism
which accompanied the rise in the pulse-rate was but slight, yet the
lSee Benedict, Carnegie Inst. Wash. Pub. 77, page 501, table 241.
372 A STUDY OF PROLONGED FASTING.
metabolism stopped falling and the superimposed factor of an altered
cellular activity compensated in part for the natural fall in the metabo-
lism which would otherwise have been expected.
SUMMARY OF RESULTS REGARDING THE METABOLISM PER KILOGRAM OF
BODY-WEIGHT AND PER SQUARE METER OF BODY-SURFACE.
If we attempt to analyze the prevailing views in regard to the metab-
olism with the progress of a fast, we find that in the first place it is
believed that the metabolism per kilogram of body-weight should rise
when the greater part of the loss in weight is made up of inert fat and
water. Second, according to the theory of the relationship between
the metabolism and the body-surface, the metabolism per square meter
of body-surface should remain constant.
So far as the metabolism per kilogram of body-weight is concerned,
it is illogical to attribute the same heat-producing value to the body-
substance which remains as the body loses weight during the progress
of the fast, and this method of computation may not be used advan-
tageously in considering the metabolism of a fasting man. Further-
more, the values for the metabolism per square meter of body-surface
show differences of approximately 25 per cent, a result which can not
be accounted for in any way by a possible discrepancy in the formula
used for computing the body-surface each day. Finally, the evidence
is strikingly in favor of the belief that the pulse-rate is an admirable
index of the intensity of cellular activity, an activity that plays a very
important role in interpreting the total metabolism, entirely irrespec-
tive of body-size or body-weight.
An examination of the values appearing in tables 52, 53, and 54
shows that there is a tendency for the metabolism to divide into three
periods during the fast. The first period, which extends nearly to the
middle of the fast, is characterized by a rapidly falling metabolism and
a rapidly falling pulse-rate, the fall in the metabolism being shown not
only in the total values, but in the values calculated on the basis of
per kilogram of body-weight and of per square meter of body-surface.
The second period of approximately 10 days shows a comparatively
level metabolism per kilogram and per square meter with an approxi-
mately level pulse-rate. The third and last period, or the last week of
the fast, shows a general tendency toward an increase in the metabolism,
although this was not so apparent in the values for the metabolism
per kilogram of body-weight in the minimum periods of the calorimeter
experiments. There was also a distinct tendency toward an increase
in the pulse-rate in the last week. It thus appears that the striking
factor of pulse-rate must continually be reckoned with, even with these
conditions of rapidly changing body-weight and body composition.
A more exact statement would perhaps be that the intensity of cellular
activity plays an important part which is not seriously affected either
by changes in the body-weight or the body-surface.
THE RESPIRATORY EXCHANGE. 373
ELIMINATION OF WATER THROUGH LUNGS AND SKIN.
Since the vaporization of water from the lungs and skin is one of the
important methods of heat regulation, its determination was formerly
given great attention, particularly in the series of 24-hour respiration
experiments carried out by Atwater and his associates in Middletown,
Connecticut. In the course of these investigations, it became apparent
that the water thus vaporized was subject to considerable variation
which was, for the most part, irregular in character. Consequently, a
knowledge of the water vaporized from the lungs and skin, per se, has
relatively little value in ordinary metabolism studies, especially with
normal subjects, and as such determinations are difficult and expensive,
they do not seem justifiable, except so far as it is necessary to note the
amount of water vaporized in the calorimeter chamber in order to
correct for the heat absorbed in the vaporization. But with the unu-
sual conditions obtaining in the experiments with L., when there was
a constantly shrinking skin and at times distinct evidence of a physio-
logical need of water, it seemed desirable to devote the additional time
and expense to securing accurate measurements of the water vaporized
from the lungs and skin of this man during the time he was inclosed in
the calorimeter chamber.
In the discussion of the atmospheric conditions inside the chamber of
the bed calorimeter, the values were given in table 44 * for the amount of
water vaporized per hour, with due correction for the water vaporized
from the wet-bulb thermometer. This hourly vaporization of water
varied from 28.7 grams on the third night of the fast to 13.6 grams on
the fifteenth night of the fast. An inspection of all the data in table 44
shows that there was apparently no constant relationship between
the amount of water vaporized and the ventilation of the chamber; nor
was there a constant relationship between the water vaporized and the
chamber temperature, for although the highest average temperature was
observed on the night of the highest vaporization of water per hour,
yet the first night the subject was inside the chamber the temperature
was within 0.1 degree of the maximum, while the vaporization of water
was only 25.3 grams. In making a further comparison with the relative
humidity, it should be noted that the values for the relative humidity
were calculated from the total amount of water vaporized. These
computed values, however, agree remarkably well with a number of
values obtained by means of a psychrometer in the outgoing air-cur-
rent, although usually they are a little lower. They may thus be relied
upon as indicating the general course of the relative humidity during
the fast. An inspection of the values given in table 44 shows that the
relationship between the relative humidity and the total vaporization
of water was approximately constant. The only scientific use of these
figures is in making corrections for the heat required for the vaporization.
^ee page 321.
374 A STUDY OF PROLONGED FASTING.
It is possible, however, to apportion the vaporization of water between
the lungs and the skin by using a method based onthe work of Zuntz, by
means of which the water vaporized from the lungs may be estimated.
During the observations of Zuntz and his co-workers, it was estab-
lished that, under the experimental conditions obtaining, each cubic
centimeter of oxygen absorbed from the air was accompanied by a lung
ventilation of 21 c.c. In the earlier publication1 giving the results of
the fasting experiments carried out in Middletown, Connecticut, this
figure was used for computing indirectly the amount of water-vapor
given off from the lungs. In the series of experiments with L. in which
the respiration apparatus was used, accurate measurements of the lung
ventilation accompanied the determinations of the oxygen consump-
tion, and consequently the relationship between the lung ventilation and
the oxygen consumption may be computed for these experiments.
From the results of these computations, which are given in table 55,
it will be seen that in the extended series of morning experiments the
lung ventilation per cubic centimeter of oxygen consumed varied from
34.0 c.c. to 22.6 c.c, these values being materially greater than those
obtained in the observations of Zuntz. As the fast progressed, there
was a distinct tendency for the lung ventilation per cubic centimeter of
oxygen to increase. Observations were also made of the influence of
conditions other than that of lying quietly in the morning, such as
change of position, time of day, the muscular activity of writing, and the
inhalation of oxygen-rich atmospheres, all of which increased the lung
ventilation per cubic centimeter of oxygen consumed. This increase
was usually not far from 4 to 5 c.c, except that in the last 3 of the 6
experiments with writing, the increase was approximately 9 c.c. over the
lung ventilation when the subject was lying quietly in the morning.
Since the relationship between the lung ventilation and the oxygen
consumption in the morning respiration experiments with the subject
lying quietly is thus well established, it is not unreasonable to assume
that the lung ventilation may in turn be obtained for the calorimeter
experiments from the oxygen consumed when the subject was inside the
respiration chamber. Accordingly, using the oxygen consumption for
the night periods and the lung ventilation per cubic centimeter of
oxygen consumed for the experiment with the respiration apparatus
on the following morning, the values for the ventilation of the lungs
have been computed for each of the bed-calorimeter experiments,
the results being recorded in table 56.
The air which the subject inhaled while inside the calorimeter
chamber obviously contained water-vapor. The amount inhaled may
readily be computed from the ventilation of the lungs and the percent-
age of water-vapor in the air inside the chamber. For instance, it will
be seen by reference to the values given in table 44 for the ventilation
Benedict, Carnegie Inst. Wash. Pub. 77, 1907, p. 435.
THE RESPIRATORY EXCHANGE.
375
of the chamber and the water vaporized per hour that the water-vapor
in each liter of air on April 10-11 was approximately 11.44 milligrams.
Consequently, the water in the air inspired per hour would be equal
to 11.44 milligrams multiplied by 415.7, or 4.76 grams. (See columns
a and b, table 56). The air expired is assumed to be saturated at
Table 55. — Ventilation of lungs per volume of oxygen in experiments with subject L.
{Respiration apparatus.)
Date.
Day of
fast.
Lying (usually 8h 30™ a.m. to
9h30ma.m.).
Lying (usually 7 p.m. to
7h45mp.m.).
Oxygen
per.
minute.
Lung ventilation
(observed) .
Lung ventilation
(observed) .
UX
Per c.c. of J
ygen
oer
nute.
Total per
minute.
Per c.c. of
Total per
minute.
oxygen. mi
/bX1000\
oxygen.
/EXIOCKT)
^ D '
A
B
C
D E
F
1912.
c.c.
liters.
c.c. (
:.c. h
ters.
c.c.
Apr. 11
231
220
225
223
237
5.80
5.66
5.61
5.14
5.35
25.1
25.7
24.9
23.0
22.6
12
13
14
15
1st
16
2d
227
5.65
24.9
17
3d
226
5.71
25.3
18
4th
212
5.21
24.6
19
5th
205
5.38
26.2
20
6th
200
5.11
25.6
21
7th
204
5.19
25.4
22
8th
203
5.10
25.1
23
9th
190
5.17
27.2
24
10th
187
4.94
26.4
25
11th
187
4.75
25.4
26
12th
187
4.94
26.4 ]
L93 5.64
29.2
27
13th
192
5.03
26.2 J]
L95 '5.87
^O.l
28
14th
181
5.00
27.6 ]
L90 5.79
30.5
29
15th
179
4.94
27.6 :
L89 6.36
33.7
30
16th
182
5.44
29.9 ]
L90 6.55
34.5
17th
182
5.21
28.6 ]
L88 6.30
33.5
2
18th
174
5.01
28.8 ]
189 6.33
33.5
3
19th
177
5.18
29.3 ]
L82 6.16
33.8
4
20th
173
5.31
30.7
5
21st
174
4.78
27.5 ]
L82 6.24
34.3
6
22d
170
5.30
31.2 ]
L76 6.18
35.1
7
23d
165
5.17
31.3 ]
L75 6.59
37.7
8
24th
167
5.13
30.7 ]
L77 6.37
36.0
9
25th
166
5.46
32.9 ]
176 6.36
36.1
10
26th
168
5.22
31.1 ]
L80 6.11
33.9
11
27th
172
5.30
30.8 ]
181 6.32
34.9
12
28th
166
5.46
32.9 ]
L78 6.28
35.3
13
29th
171
5.41
31.6 ]
178 6.28
35.3
14
30th
166
5.21
31.4 ]
L83 6.44
35.2
15
31st
166
5.24
31.6
17
170
183
4.32
6.22
25.4"
34.0
18
during the period 3h 16m p.m. to 3h 51m p.m., with the subject in the lying position, the
observations were: oxygen, 189 c.c; ventilation per minute, 5.63 liters; ventilation per c.c. of
oxygen, 29.8 c.c.
376
A STUDY OF PROLONGED FASTING.
Table 55. — Ventilation of lungs per volume of oxygen in experiments with subject L.
{Respiration apparatus)— Continued.
Date.
Day of
fast.
Sitting.1
Period.
Oxygen
per
minute.
G
Lung ventilation
(observed).
Total per
minute.
H
Per c.c. of
oxygen.
(hXIOOOX
I
1912.
Apr. 16 ,
19
23
24
26
27
29
May 1
4
7
14
2d
5th , ,
9th
10th
12th ,
13th . . ,
15th
17th
20th
23d
30th, ,
4h W* p.m. to 4h 35m p.m. . . .
4 10 p.m. 4 43 p.m.*. .
3 52 p.m. 4 28 p.m
3 58 p.m. 4 57 p.m
3 13 p.m. 4 11 p.m
12 14 p.m. 12 48 p.m. . . .
3 23 p.m. 3 56 p.m.* . .
9 31 a.m. 10 04 a.m.* . .
9 35 a.m. 10 10 a.m.* . .
3 43 p.m. 4 14 p.m.*. .
6 32 p.m. 7 02 p.m.*. .
c.c.
244
269
187
194
183
200
233
215
208
222
221
liters.
6.13
8.33
6.06
6.37
5.77
6.05
8.60
7.12
6.74
8.32
8.77
c.c.
25.1
31.0
32.4
32.8
31.5
30.3
36.9
33.1
32.4
37.5
39.7
'Periods indicated by an asterisk were obtained with the subject sitting, writing.
a temperature of 37° C. and, knowing the ventilation of the lungs, it
is easy to compute, by means of well-known tables, the amount of water
in the air expired from the lungs. For April 10-11 this was found to
be 18.07 grams. The amount of water-vapor eliminated from the lungs
is therefore the difference between the water in the air inspired and
that in the air expired, this being 13.3 grams per hour for April 10-11.
The amount of water-vapor eliminated per hour from the lungs and
skin, i. e., vaporized inside the calorimeter, on this date was 25.3 grams
(see column d), so that the amount of water-vapor given off from the
skin would be represented by the difference between columns d and e
or 12.0 grams (column f) for April 10-11. The percentage distribution
of the water-vapor from the lungs and skin has also been calculated
and recorded in this table in columns G and h, and finally (for reference),
the relative humidity is given in the last column of the table.
The assumption that the expired air is saturated at 37° C. will un-
undoubtedly be questioned in the light of the recent work of Loewy1
and Galeotti.2 Neither of these researches has, however, stood the
test of severe criticism, and while undoubtedly the water-vapor is
probably not quite so great as that represented by assuming the air
to be saturated at 37° C, nevertheless, for want of firmly established
values for this factor, we adhere to the older assumption. It is, further-
more, important to note that any error in this assumption affects only
the apportionment of the water vaporization between the lungs and
xLoewy and Gerhartz, Biochem. Zeitschr., 1912, 47, p. 343.
2Galeotti, Biochem. Zeitschr., 1912, 46, p. 173.
THE RESPIRATORY EXCHANGE.
377
the skin for each individual night. The most important comparisons
are those made between different days of the fast and with these a
considerable error in the assumption might be made without affecting
the general deduction.
Table 56. — Water eliminated from the lungs and skin during experiments with L in the bed
calorimeter at night. {Amounts per hour.1)
Date.
Day
of fast.
Ventila-
tion of
lungs.2
Water
in air
in-
spired.3
Water
in air
ex-
pired.4
Water eliminated.
1
>
'•+3
c3
From
lungs
and
skin.8
From
lungs.
(c-b)
From
skin.
(D — E)
Proportion.
From
lungs.
From
skin.
/exioox
(FT)
A
B
C
D
E
F
G
H
I
1912.
liters.
gm.
gm.
gm.
gm.
gm.
p. ct.
p. ct.
p.ct.
Apr. 10-11
415.7
4.76
18.07
25.3
13.3
12.0
52.6
47.4
62
62
11-12
397.8
4.30
17.29
26.3
13.0
13.3
49.4
50.6
12-13
376.5
4.12
16.36
26.6
12.2
14.4
45.9
54.1
59
63
13-14
305.0
3.43
13.26
27.1
9.8
17.3
36.2
63.8
14-15 . . .
1st...
287.5
2.69
12.50
22.8
9.8
13.0
43.0
57.0
52
15-16...
2d...
315.2
3.32
13.70
25.6
10.4
15.2
40.6
59.4
59
16-17...
3d...
312.7
3.69
13.59
28.7
9.9
18.8
34.5
65.5
63
17-18. . .
4th..
298.2
2.91
12.96
22.8
10.1
12.7
44.3
55.7
54
18-19...
5th..
301.8
2.78
13.12
21.1
10.3
10.8
48.8
51.2
50
19-20. . .
6th..
298.0
2.44
12.95
19.4
10.5
8.9
54.1
45.9
47
20-21 . . .
7th..
289.6
2.26
12.59
18.9
10.3
8.6
54.5
45.5
43
21-22 . . .
8th..
281.6
2.25
12.24
19.2
10.0
9.2
52.1
47.9
46
22-23 . . .
9th..
290.5
2.60
12.63
21.1
10.0
11.1
47.4
52.6
48
23-24. . .
10th..
285.1
2.24
12.39
17.0
10.2
6.8
60.0
40.0
44
24-25. . .
11th..
268.2
2.12
11.66
18.3
9.5
8.8
51.9
48.1
44
25-26. . .
12th..
277.2
2.23
12.05
17.9
9.8
8.1
54.7
45.3
45
26-27...
13th..
268.8
2.12
11.68
18.4
9.6
8.8
52.2
47.8
43
27-28. . .
14th..
276.6
2.13
12.02
17.6
9.9
7.7
56.3
43.7
43
28-29 . . .
15th..
269.9
1.74
11.73
13.6
10.0
3.6
73.5
26.5
38
29-30...
16th..
296.0
2.06
12.87
15.9
10.8
5.1
67.9
32.1
40
Apr. 30-May 1
17th..
274.6
1.88
11.94
15.6
10.1
5.5
64.7
35.3
39
May 1- 2 . . .
18th..
274.8
1.92
11.94
16.0
10.0
6.0
62.5
37.5
39
2- 3 . . .
19th..
277.8
1.89
12.07
15.5
10.2
5.3
65.8
34.2
39
3- 4 . . .
20th..
294.7
2.00
12.81
15.8
10.8
5.0
68.4
31.6
38
4- 5 . . .
21st...
254.1
1.71
11.04
14.6
9.3
5.3
63.7
36.3
38
5- 6 . . .
22d. . .
288.3
2.08
12.53
15.7
10.5
5.2
66.9
33.1
40
6-7...
23d. . .
293.0
2.37
12.74
16.7
10.4
6.3
62.3
37.7
45
7- 8 . . .
24th..
291.0
2.24
12.65
16.5
10.4
6.1
63.0
37.0
44
8-9...
25th..
302.0
2.47
13.13
17.5
10.7
6.8
61.1
38.9
46
9-10.. .
26th..
296.7
2.42
12.90
18.0
10.5
7.5
58.3
41.7
46
10-11 . . .
27th . .
282.7
2.42
12.29
18.7
9.9
8.8
52.9
47.1
48
11-12...
28th..
319.8
2.68
13.90
18.9
11.2
7.7
59.3
40.7
46
12-13 . . .
29th..
299.6
2.56
13.02
19.3
10.5
8.8
54.4
45.6
46
13-14. . .
30th..
284.5
2.55
12.37
19.7
9.8
9.9
49.7
50.3
49
14-15 . . .
31st...
303.4
2.32
13.19
17.9
10.9
7.0
60.9
39.1
42
16-17
234.7
2.08
10.20
20.0
8.1
11.9
40.5
59.5
50
17-18
395.8
3.91
17.20
23.7
13.3
10.4
56.1
43.9
56
'For the duration of periods see table 44.
Calculated from the oxygen consumption during the night and the ventilation per volume of
oxygen (table 55) observed on the following morning.
3Computed by means of the ventilation of the lungs (column a) and the water absorbed per
liter of the ventilating air-current, as computed'from columns a and b, table 44.
*It is assumed that the air expired was saturated and at a temperature of 37° C.
BSee "Water vaporized per hour," table 44.
378 A STUDY OF PROLONGED FASTING.
The apportionment of the water-vapor between the two paths of
excretion presents certain particularly interesting features. On the
4 nights prior to the fast there was a continually decreasing percentage
of water excreted from the lungs until the low value of 36.2 per cent was
noted on the last night. During the fasting period there was a distinct
tendency for the proportion given off from the lungs to increase grad-
ually, and although there were marked irregularities in this increase,
particularly on the tenth and fifteenth nights of fasting, yet it may be
said that in general the proportion of water vaporized from the lungs
became greater and that from the skin less as the fast progressed. The
apportionment between the lungs and the skin does not, however,
follow any definite law.1
Another interesting comparison may be made between the relative
humidity and the excretion of water- vapor from the skin. We have
existing inside the chamber an essentially constant ventilation and an
essentially constant temperature. With a low relative humidity, a
high vaporization of water from the skin would be looked for, since one
would naturally expect water to vaporize more rapidly from the moist
skin of the subject as the air surrounding the body became drier.2 We
find, however, that the opposite is true, since the general course of both
the total amounts and the percentage values follows with approximate
constancy the values for the relative humidity, the lowest percentage
of water excreted from the skin, i. e., 26. 5 per cent on the fifteenth night,
being coincidental with the lowest relative humidity. As has been
pointed out in a previous section, there was a slight tendency for the
relative humidity to increase during the latter part of the fast ; this was
accompanied in general by an increase in the water vaporized from the
skin. It may be inferred, however, that, in general, the body-surface
of the man became smaller and the skin less and less moist as the fast
progressed and perhaps, in consequence, less capable of losing water.
Nevertheless the wide variations noted by Langlois and Boussaguet (Compt. rend. Soc. Biol.,
1912, 72, p. 967) do not appear in this fasting experiment; the method here employed is, however,
fundamentally different from that employed in the French research.
2 Wolpert. Archiv f. Hygiene, 1902, 41, p. 301.
CALORIMETRY.
DIRECT CALORIMETRY.
A unique feature of the experiments carried out at Wesleyan Uni-
versity on fasting men was the direct determination of the heat, not
only eliminated but produced, by means of a specially designed respira-
tion calorimeter. In these earlier experiments the subject remained
the entire fasting period inside the respiration chamber and thus con-
tinuous records of the heat output could advantageously be obtained.
Knowing that it would be impracticable (and, indeed, unwise) to con-
fine the subject L. inside the respiration chamber for the 31 days of
the fast, inasmuch as such confinement would prevent a large number
of other important observations, it was decided to make the heat
measurements only during the night period.
The significance of the heat measurements in this fasting experiment
is altogether different from that of the measurements obtained in the
earlier researches at Wesleyan University. In the earlier experiments
the primary object was to secure a complete balance of income and
outgo and study the transformations of both matter and energy
throughout the fast. Consequently, special stress was laid upon these
direct heat measurements, the elementary analyses of the urine, and
the computations of the amounts of protein, fat, and carbohydrate
katabolized; finally a comparison was made of the heat actually
measured with that theoretically resulting from the katabolism of the
protein, fat, and carbohydrates. Owing to the shortness of the fasts,
but little evidence could be obtained regarding the effect of prolonged
fasting upon the transformations of matter and energy. In this 31-
day fast, however, special emphasis was laid upon the effect of inanition
upon the katabolism as the fast progressed. For this purpose, as we
have seen, a study of the gaseous exchange at any given hour of each
day is amply sufficient, as the influence of the fast upon the gaseous
exchange has been found to be essentially the same, regardless of
whether it was studied in the morning or in the late afternoon on the
respiration apparatus or during the night in the bed calorimeter.
While a continuous measurement of the heat throughout the fast
was not practicable, it seemed desirable to have the subject sleep in
a chamber fitted with calorimetric appliances1 and thus simultane-
ously determine the gaseous exchange and the heat output. Since the
whole period of sojourn inside the chamber during the night might vary
as to muscular activity and particularly as to wakefulness, any attempt
to make a sharp comparison of the results obtained for the various
nights of the fast was not to be expected, although subsequently it was
found that the remarkable degree of quiet shown by this man enabled
xBenedict and Carpenter, Carnegie Inst. Wash. Pub. 123, 1910.
379
380 A STUDY OF PROLONGED FASTING.
us to make fairly satisfactory comparisons. It is clear, however, that
the direct measurements of the heat output in this particular series of
calorimeter experiments were only incidental and not of prime impor-
tance in this study. This is particularly fortunate, since, as will be
pointed out later, certain unavoidable conditions, particularly with
reference to the temperature environment in the calorimeter room,
vitiated to a considerable extent the direct measurements of the heat
output.
Prior to the arrival of the subject in Boston, the accuracy of the
apparatus as to heat measurement had been repeatedly tested1 by
electrical check tests, in which a definite amount of heat was developed
with an electric current of known amperage and voltage, and also as
to the accuracy of the measurement of the carbon dioxide, water-vapor,
and heat produced, as well as oxygen consumed, by burning known
amounts of alcohol inside the chamber. It was possible, therefore,
to put the subject inside the calorimeter immediately on his arrival
at the laboratory and determine the heat output on the several nights
preceding the fast.
The calorimeter laboratory is so constructed that the temperature of
the room can be kept constant, and during calorimeter experiments it
is ordinarily kept at the temperature of the calorimeter. Furthermore,
the calorimeter is so made that it normally measures the heat-produc-
tion with great accuracy, irrespective of whether the environmental
temperature is 1° above or below the temperature of the calorimeter.
In this fasting experiment, however, a great difficulty was encoun-
tered at the outset, in that the subject (who spent the day on a bal-
cony in the calorimeter room) complained so much of cold that it
became necessary to increase the temperature of the room throughout
the day and to cool it somewhat during the night. Accordingly the
calorimeters in the calorimeter laboratory became very much over-
heated and the measurements of the heat output were inevitably
somewhat too large. After the conclusion of the observations on L.,
a series of check tests was carried out, with conditions as nearly
comparable as possible with those existing while the man was in the
laboratory, and corrections for the heat output were computed from
the results of the tests. It is unnecessary here to discuss the theo-
retical difficulties of measuring heat with the increased temperature
necessary to make the subject comfortable, but we believe the heat
measurements as finally recorded here are within 2 or 3 per cent of the
actual values. It may be safely said that the measurement is always
too great, and the correction would therefore be a minus correction, the
probable error being about 2.5 per cent.
It should be borne in mind that the heat production as measured by
a respiration calorimeter is made up of several factors. First, a certain
1Benedict, Riche, and Emmes, Am. Journ. Physiol., 1910, 26, p. 1.
CALORIMETRY. 381
amount of heat is given off by radiation and conduction and is absorbed
by heat-absorbing pipes through which cold water is passed inside the
calorimeter. While this includes the greater portion of the heat meas-
ured, there is likewise a considerable elimination of heat due to the
vaporization of water from the lungs and skin of the subject. This is
measured by weighing the water-vapor in the air-current and making
due allowance for the heat of vaporization per gram of water, namely,
0.586 calorie. Furthermore, there should be corrections for any loss
of heat through the excretion of urine or feces, and for changes in
body-weight. This latter correction is of somewhat more importance
in fasting experiments than in ordinary respiration-calorimeter experi-
ments, since the losses in weight are naturally greater. The corrected
value thus obtained may properly be called the heat eliminated by
the body during the experimental period.
By far the most important correction, however, is that for the
changes in the body-temperature, for, as was pointed out in the section
in which this factor was considered, a complete comparison of direct
and indirect calorimetry, which is always attempted in respiration-
calorimeter experiments, necessitates a knowledge of the fluctuations
in the body-temperature.1 While it has been shown that the tempera-
ture fluctuations in the various parts of the body follow very closely
the temperature fluctuations in the rectum,2 it is still true that one of
the most important and difficult problems in computing the heat pro-
duction from the heat measured directly by the calorimeter is the possi-
bility of the storage of a considerable amount of heat inside the human
body. For example, the body-temperature of a man weighing 60
kilograms falls 0.5° C. This would correspond to a loss of nearly 25
calories, or the total heat production in 20 to 30 minutes of a man
asleep; thus, by taking into consideration the temperature changes of
the body and the body-weight multiplied by the specific heat (assumed
to be 0.83), the amount of heat lost from or stored in the body may be
computed and in turn may be subtracted from or added to the heat
eliminated. The heat measurement thus corrected is termed the heat
produced. In the tabular presentation of our results, we deal usually
with the heat-production and not with the heat elimination.
The intelligent use of corrections in the direct measurement of the
heat output of man is so difficult that the computation of the heat
produced in short periods is highly questionable. An inspection of the
heat measurements obtained with our subject during the night shows a
lack of uniformity which makes it impracticable to present the heat pro-
duction in the form of curves, as was done for the carbon dioxide pro-
duced and oxygen consumed in figures 41 to 44, but there is distinct
evidence of the lowest heat production occurring during the midnight
Benedict and Joslin, Carnegie Inst. Wash. Pub. 136, 1910, p. 19.
2Benedict and Slack, Carnegie Inst. Wash. Pub. 155, 1911, p. 72.
382
A STUDY OF PROLONGED FASTING.
hours, this being coincident with the minimum respiratory exchange.
Of far greater significance is the change in the average heat production
throughout the night as the fast progressed.
The subject spent varying lengths of time inside the calorimeter
each night, averaging not far from 10 hours, but to make the values
comparable the computations are all based either upon the amount of
heat produced per hour or per 24 hours. The results of these computa-
tions are recorded in table 57, in which the heat produced by the
subject in the bed calorimeter at night is given on the basis of the total
Table 57. — Heat produced by subject L. during experiments in the bed calorimeter at night.
Computed to basis
Date.
Day of
fast.
of 24 hours.
Per
Per sq.
meter of
Total.
kilogram
of body-
weight.
Total.
body-
surface
(Meeh).
1912.
cats.
cats.
cats.
cats.
Apr. 10-111
83.4
1.38
2002
1054
H-121
81.0
74.9
68.3
64.0
1.33
1.22
1.12
1.07
1944
1798
1639
1536
1023
941
858
817
12-131
13-141
14-151
1st
16-16
2d
64.3
1.09
1543
830
16-171
3d
64.1
1.10
1538
836
17-18
4th
63.1
1.10
1514
827
18-19
5th
57.6
1.02
1382
764
19-20
6th
58.1
1.04
1394
774
20-21
7th
66.7
1.02
1361
760
21-22
8th
58.6
1.06
1406
790
22-23
9th
53.1
.97
1274
720
23-24
10th
53.5
.99
1284
725
24-25
11th
52.4
.97
1258
715
25-26
12th
51.9
.97
1246
712
26-27
13th
51.7
.97
1241
709
27-28
14th
50.6
.95
1214
698
28-29
15th
47.1
.89
1130
649
29-30
16th
46.1
.88
1106
639
Apr. 30-May 1 . . . .
17th
45.7
.88
1097
642
May 1-2
18th
46.5
.90
1116
653
2-3 w .
3-4
19th
47.9
.94
1150
676
20th
46.9
.92
1126
666
4-5
21st
44.0
.87
1056
625
5-6
22d
45.7
.91
1097
653
6-7
23d
45.6
.91
1094
655
7-8
24th
45.3
.91
1087
651
8-9
25th
44.1
.89
1058
637
9-101
26th
47.8
.97
1147
695
10-11
27th
46.0
.94
1104
673
11-12
28th
46.2
.95
1109
676
12-13
29th
46.9
.97
1126
691
13-14
30th
47.1
.99
1130
698
14-15
31st
47.0
.99
1128
701
16-17
46.1
62.6
.97
1.29
1106
1502
687
916
17-18
xThe heat measured during the night experiments on the days noted has been corrected only
for the change in body-weight and not for change in body-temperature.
CALORIMETRY. 383
amount per hour, the amount per kilogram per hour, the total amount
per 24 hours, and the amount per square meter of body-surface per 24
hours. With the exception of the first few days, the values given are
all for heat produced, i. e., the heat measured by the apparatus cor-
rected for both change in body-weight and change in body-temperature.
For those nights on which the body-temperature was lacking, the cor-
rection is made only for changes in body-weight, i. e., only the heat
eliminated is given.
Before discussing the values in table 57, it should be noted that the
figures are not strictly comparable, since, as has already been stated,
the muscular activity was not absolutely uniform. The kymograph
records show that the subject was remarkably quiet, so that no great
correction would be necessary for differences in the muscular activity;
but unquestionably the amount of sleep varied and, as has already been
shown, the influence of sleep per se upon the metabolism should be
taken into consideration. Nevertheless, the values are determined
under sufficiently uniform conditions to justify their comparison if it
be clearly understood that there is a reasonably constant error of tech-
nique due to the extreme heat of the calorimeter laboratory, a certain
lack of uniformity in the muscular activity, and, finally, material differ-
ences in the amount of time spent by the subject awake or asleep.
The total heat produced varied from a maximum of 83.4 calories per
hour on the first night inside the chamber to a minimum of 44.0 calories
per hour on the twenty-first night of fasting. When the heat pro-
duction is computed on a 24-hour basis, a maximum is obtained of
2,002 calories and a minimum of 1,056 calories. The steady fall in
the total heat production per hour during the first 2 weeks of fasting,
followed by a period of approximately constant output of heat with a
slight tendency towards an increase in the last week, is consistent with
our observations on metabolism as recorded in the earlier chapters of
this book.
As a concession to those writers who are wont to consider the heat
production on the basis of per kilogram of body-weight and per square
meter of body-surface, both values have been computed and included
in table 57. The heat per kilogram of body-weight per hour varies
from a maximum of 1.38 calories on the first night inside the chamber to
a minimum of 0.87 on the twenty-first night of the fast. Even on this
basis there is a distinct fall in the heat production per kilogram of
body-weight during the first two weeks, with an approximately con-
stant level for a short period and then a distinct tendency to increase
during the last week of the fast.
On the basis of per square meter of body-surface in 24 hours, the
heat production varied from a maximum of 1,054 calories on the first
night inside the chamber to a minimum of 625 calories on the twenty-
first night of the fast. On the first 4 nights the heat production is
clearly affected by the previous ingestion of food on the evening before,
384 A STUDY OF PROLONGED FASTING.
but during the strictly fasting period we have a variation ranging
from 836 calories per 24 hours on the third night of the fast to a mini-
mum of 625 calories per 24 hours on the twenty-first night, a variation
of somewhat more than 25 per cent. Even on this basis, when pre-
sumably the values should all be constant, there is a distinct falling
off in value during the first 2 weeks of fasting, followed by a period of
approximately constant heat production per square meter, with a
distinct rise in the values in the last week of fasting.
It is thus clear that the deductions drawn from an inspection of the
data for the respiratory exchange in the foregoing chapters apply
equally well to the total heat production as measured in the respiration
calorimeter, even admitting the discrepancies outlined previously.
The general picture is by no means obscured by the assumptions made,
this showing that the influence of fasting per se upon not only the total
heat production but the heat production per kilogram of body-weight
and per square meter of body-surface was fully in accord with the influ-
ence of fasting upon the respiratory exchange. It seems unwise,
however, to dwell further upon these results of the heat measurements
and thereby possibly ascribe to them too much importance. The
errors cited bring them clearly out of the sphere of accurate physio-
logical experimentation and make them of value only for subsidiary
evidence. For a more careful and certainly more nearly exact con-
sideration of the heat production under different conditions during the
fast, we must resort to the method of indirect calorimetry.
INDIRECT CALORIMETRY.
Second only in value to accurate direct measurements of the heat
output from the body is the method of so-called "indirect calorimetry/ '
this being an excellent substitute for the difficult and costly direct
calorimetry. By considering the nitrogenous material excreted in the
urine and the amount of carbon dioxide produced and oxygen absorbed,
it is possible to apportion the katabolism among the various body con-
stituents, these being chiefly protein, fat, and carbohydrate, and to
compute (from the well-known heats of combustion of these constitu-
ents) the amount of heat liberated in the process of their disintegration.
Several methods of computing the heat production thus indirectly
have been employed by various writers. Perhaps the most elaborate
and fundamentally exact is that based upon 24-hour studies of the
gaseous exchange, such as were made with the respiration calorimeter
formerly used at Wesleyan University. The elementary analyses of
food, feces, and urine, and the direct determination of the oxygen
absorbed and the carbon dioxide produced gave data for computing,
by simultaneous equations,1 the amounts of protein, carbohydrate (gly-
cogen), fat, and water participating in the metabolism. From these
data and the heats of combustion of the various body constituents and
of the urine and feces, the total energy production could be accurately
Benedict, Carnegie Inst. Wash. Pub. 77, 1907, pp. 36 and 452.
CALOKIMETRY. 385
computed. Such computed energy transformations agreed remarkably
well with the direct measurements of the heat produced per 24 hours.
When, as is frequently the case, an exact knowledge of the various
amounts of carbohydrate, fat, and protein disintegrated is not of par-
ticular significance, it is possible to secure the heat-production indi-
rectly from the measurements of the carbon dioxide produced and the
oxygen consumed without computing the different constituents oxi-
dized. Zuntz has computed most carefully the calorific equivalents of
both oxygen and carbon dioxide and assumes that for every liter of
oxygen absorbed in normal metabolism a definite number of calories
must have been developed. The calorific value of carbon dioxide
varies greatly, the variations depending upon whether the carbon diox-
ide is produced by the combustion of carbohydrate or of fat. Thus, with
every gram of carbon dioxide resulting from the oxidation of carbohy-
drate 2.57 calories of heat are produced, and from the oxidation of fat
3.41 calories. On the other hand, the calorific equivalent of oxygen
does not fluctuate so widely, for when carbohydrate is burned, 3.53
calories of heat are developed for every gram of oxygen consumed, and
when pure fat is burned, 3.28 calories are produced. It is therefore
only necessary to determine the carbon-dioxide production and oxygen
consumption very exactly, and from the respiratory quotient and
the absolute amount of either oxygen or carbon dioxide it is possible
to compute the heat production by using the calorific value for the
oxygen or the carbon dioxide.
With the Zuntz-Geppert apparatus, oxygen is determined as accu-
rately as is carbon dioxide, and since the calorific value of oxygen
remains essentially constant, Zuntz has regularly used the oxygen
measurements and multiplied by the calorific equivalent of oxygen.
In connection with his researches on the physiology of marching, pub-
lished with Schumburg, Zuntz1 worked out a table which gave the
respiratory quotients and calorific equivalents of oxygen with varying
proportions of fat and carbohydrate in the material oxidized.
A more difficult matter is the determination of the calorific equiva-
lent of oxygen when protein is burned. Loewy2 has correctly pointed
out that in experiments of short duration it is justifiable to compute
the heat production indirectly by using the carbon-dioxide output and
oxygen intake without taking into account the protein disintegration,
since the protein disintegration may or may not be simultaneous with
the nitrogen excretion. As a matter of fact, the calorific equivalents of
oxygen and carbon dioxide when protein is burned are not greatly
different from those for either fat or carbohydrate; and furthermore,
since the protein usually furnishes only about 15 per cent of the total
energy, the error in thus neglecting the protein is extremely small,
especially for short experiments.
TZuntz and Schumburg, Physiologie des Marsches, Berlin, 1901, p. 361.
2Loewy, Oppenheimer's Handbuch der Biochemie, Jena, 1911, 4 (1), p. 281.
386 A STUDY OF PROLONGED FASTING.
On the other hand, in considering the results of long experiments, it
is not justifiable to neglect the protein entirely. As the result of a large
number of experiments made by Rubner and by Zuntz and his associ-
ates, a certain number of standard figures regarding the combustion
of protein may be used with considerable confidence. Thus, they have
definitely established that 1 gram of nitrogen in the urine corresponds
to a heat production of 26.51 calories as a result of the oxidation of
protein. Furthermore, Zuntz has computed that for each gram of
nitrogen in the urine 5.91 liters of oxygen are absorbed and 4.75 liters
of carbon dioxide are produced. These values were obtained from ex-
periments on animals which were fed a diet of meat only. The method
of computing the energy from protein by making a special computation
for the energy accompanying the nitrogen in the urine was employed
by Zuntz1 in computing the energy output from the values found with
the fasting subjects Cetti and Breithaupt.
Williams, Riche, and Lusk2 have also made use of these values for
computing the heat production even in short periods, and in accordance
with the computations of Loewy have calculated the non-protein
respiratory quotient by deducting the carbon dioxide and oxygen
apportioned to the protein oxidation from the total measured amounts
of these gases. Using this non-protein respiratory quotient and
Zuntz and Schumburg's table, they have computed the energy of fat
and carbohydrate. The total urinary nitrogen multiplied by 26.51
gave the energy value for protein.
Magnus-Levy3 and Loewy,4 in discussing the indirect method of
calculating the heat output, emphasize the fact that the computation
gives accurate results only when the oxidation in the body proceeds
along normal lines. Recognizing the fact that there is a not inconsider-
able disturbance of the intermediary metabolism during prolonged
fasting, due in part to the excretion of large amounts of ammonia
and the formation of /3-oxybutyric acid, Grafe5 has attempted to com-
pute the heat production by making numerous corrections, some of
which are certainly based upon premises not beyond question.
After a careful consideration of the various methods for computing
indirectly the energy production from the gaseous exchange and the
urinary excretion of nitrogen, and taking into account all the possible
errors and the weight of the errors in the different determinations,
it seemed most advantageous to use in our calculations the method
recommended by Magnus-Levy, in which the heat output can be
directly computed by using the calorific equivalent of oxygen for
^ehmann, Mueller, Munk, Senator, and Zuntz, Archiv f. path. Anat. u. Physiol, u. f. klin.
Med., 1893, 131, Supp., p. 208.
2Williams, Riche, and Lusk, Journ. Biol. Chem., 1912, 12, p. 357.
3Magnus-Levy, von Noorden's Handbuch der Pathologie des Stoffwechsels, Berlin, 1896,
1, p. 217.
4Loewy, Oppenheimer's Handbuch der Biochemie, Jena, 1911, 4 (1), p. 281.
'Grafe, Zeitschr. f. physiol. Chem., 1910, 65, p. 21.
CALORIMETRY. 387
the various respiratory quotients. Magnus-Levy1 computed that if
0.031 calorie per liter is deducted from the calorific equivalent of oxy-
gen, the correction for the energy requirement of protein is thereby
made, assuming that only about 15 per cent of the total energy of the
day is obtained from protein. If the energy from protein is 30 per cent
of the daily quota, the correction would be 0.064 calorie. Thus,
according to the table of Zuntz and Schumburg, a non-protein respira-
tory quotient of 0.738 would correspond to a calorific equivalent of
oxygen equal to 4.724 calories; but with a respiratory quotient of 0.738
and a division of the energy between 30 per cent of protein and 70 per
cent of fat, the calorific equivalent would be, according to Magnus-
Levy, 4.660 calories or 0.064 calorie less.
A computation of the energy which should be apportioned to protein
during the calorimeter period showed that in general about 20 per cent
of the total energy was derived from protein. Making use of this value
and computing the correction so that it may be used with carbon
dioxide instead of oxygen, we have employed a correction of 0.057
calorie per liter of carbon dioxide. Hence, by deducting 0.057 calorie
from the calorific equivalent of carbon dioxide as presented in the table
given by Benedict and Talbot,2 the heat output is rapidly calculated
and the correction for the difference in the factor due to the energy
from protein is simultaneously made. Since the computations for
energy are chiefly for purposes of comparison as the fast progresses,
and since we do not have the nitrogen output during the first nights
of the fast, the value 0.057 has been used for computing the entire
series. The results are given in table 58.
As was pointed out in discussing the values found for the gaseous
exchange, at least three bases for comparison may be used:
First, a comparison of the heat output during the bed-calorimeter
experiments has a distinct advantage, since the experimental periods
were so long, being approximately 10 hours.
Second, as there was irregularity in the amount of activity inside the
bed calorimeter, and also a difference in the time spent in sleep, it is
distinctly advantageous to select periods throughout the night in which
there was a minimum amount of activity and compare the minimum
metabolism for the various nights as the fast progressed.
Finally, it is probable that no periods throughout the day are so
comparable as the morning respiration experiments, in which the con-
ditions were ideal, as the subject lay perfectly quiet upon the couch
and always awake; consequently, comparisons can be made between
the results obtained with the respiration apparatus and with the bed
calorimeter.
'See Loewy, Oppenheimer's Handbuch der Biochemie, Jena, 1911, 4 (1), p. 281; also, Magnus-
Levy, von Noorden's Handbuch der Pathologie des Stoffwechsels, Berlin, 1896, 1, p. 207.
Benedict and Talbot, Carnegie Inst. Wash. Pub. 201, 1914, p. 29.
388
A STUDY OF PROLONGED FASTING.
Table 58. — Heat production (indirect) computed from the gaseous exchange in experiment with
subject L. in the lying position.
Date.
Day
of
fast.1
Bed calorimeter (night).1
Average basis.
Carbon
Respi-
dioxide
ratory
per
quo-
minute.
tient.
A
B
liter.
0.224
0.81
.228
.88
.218
.86
.180
.81
.165
.78
.159
.75
.151
.73
.150
.74
.143
.75
.134
.68
.135
.71
.137
.73
.134
.75
.130
.72
.128
.72
.129
.73
.126
.74
.120
.72
.117
.71
.117
.71
.115
.72
.115
.72
.113
.71
.114
.71
.112
.73
.111
.72
.112
.72
.109
.69
.111
.72
.111
.70
.111
.72
.115
.71
.112
.72
.110
.72
.115
.72
.124
.80
.188
.97
Calorific
equivalent
of carbon
dioxide.
C
Heat per
24 hours.
(aXcX1440)
Minimum basis.
Carbon
dioxide
per
minute.
£
Heat per
24 hours.
(eXcX1440)
F
1912.
Apr. 10-1 12
11-12*
12-132
13-14*
14-15
15-16
16-17
17-18
18-19
19-20
20-21
21-22
22-23
23-24
24-25
25-26
26-27
27-28
28-29
29-30
Apr. 30-May 1 .
May 1- 2
2- 3
3-
4-
5-
6-
7-
8-
9-10.
10-11.
11-12.
12-13.
13-14.
14-15.
16-172
17-182
1st .
2d..
3d..
4th.
5th.
6th.
7th.
8th.
9th.
10th.
11th.
12th.
13th.
14th.
15th.
16th.
17th.
18th.
19th.
20th.
21st.
22d..
23d..
24th.
25th.
26th.
27th.
28th.
29th.
30th.
31st .
cats.
5.885
5.511
5.612
5.885
6.066
6.262
6.401
6.331
6.262
6.637
6.549
6.401
6.262
6.474
6.474
6.401
6.331
6.474
6.549
6.549
6.474
6.474
6.549
6.549
6.401
6.474
6.474
6.637
6.474
6.637
6.474
6.549
6.474
6.474
6.474
6.001
5.165
cals.
1898
1809
1762
1525
1441
1434
1392
1367
1289
1281
1273
1263
1208
1212
1193
1189
1149
1119
1103
1103
1072
1072
1066
1075
1032
1035
1044
1042
1035
1061
1035
1085
1044
1025
1072
1072
1398
liter.
0.218
.217
.196
.173
.152
.154
.148
.140
.137
.131
.132
.135
.131
.127
.124
.126
.125
.116
.114
.114
.113
.112
.111
.112
.103
.109
.106
.106
.108
.106
.107
.109
.104
.103
.109
.117
.176
cals.
1847
1722
1584
1466
1328
1389
1364
1276
1235
1252
1245
1244
1181
1184
1156
1161
1140
1081
1075
1075
1053
1044
1047
1056
949
1016
988
1013
1007
1013
998
1028
970
960
1016
1011
1309
1For duration of periods, see table 44.
2On the days preceding and following the fast the night experiments were made after the
ingestiouof'food.
CALORIMETRY.
389
Table 58. — Heat production (indirect) computed from the gaseous exchange in experiment
with subject L. in the lying position — Continued.
Respiration apparatus (morning).1
Calo-
Heat.
Date.
Day
Carbon
Respi-
rific
Per hour.
Per 24 hours.
Increase
of fast.
dioxide
ratory
equiva-
lent of
carbon
diox-
(awake)
per
quo-
above night
minute.
tient.
Per kilo-
Per
minimum
ide.
Total
(gXiX60).
gram of
body-
weight.
Total
(JX24).
square
meter oJ
body-
surface.
(asleep)
on 24-
hour basis
(L-F).
G
H
I
J
K
L
M
N
1912.
liter.
cats.
cats.
cals.
cals.
cals.
cals.
Apr. 10-1 12
0.186
0.81
5.885
65.7
1.09
1577
834
-270
11-122
.196
.200
.182
.185
.89
.89
.82
.78
5.462
5.462
5.827
6.066
64.2
65.5
63.6
67.3
1.06
1.07
1.05
1.13
1541
1572
1526
1615
811
823
803
859
-181
- 12
60
287
12-132
13-142
14-15. ..
1st..
15-16. ..
2d..
.180
.79
6.005
64.9
1.11
1558
838
169
16-17. ..
3d..
.169
.75
6.262
63.5
1.10
1524
828
160
17-18...
4th.
.159
.75
6.262
59.7
1.05
1433
787
157
18-19. ..
5th.
.158
.77
6.130
58.1
1.03
1394
770
159
19-20. ..
6th.
.148
.74
6.331
56.2
1.01
1349
749
97
20-21 . . .
7th.
.153
.75
6.262
57.5
1.04
1380
771
135
21-22. ..
8th.
.151
.74
6.331
57.4
1.04
1378
774
134
22-23 . . .
9th.
.143
.75
6.262
53.7
.98
1289
728
108
23-24. . .
10th.
.143
.76
6.196
53.2
.98
1277
726
93
24-25. ..
11th.
.140
.75
6.262
52.6
.98
1262
717
106
25-26 . . .
12th.
.140
.75
6.262
52.6
.98
1262
721
101
26-27. . .
13th.
.140
.73
6.401
53.8
1.01
1291
738
151
27-28. ..
14th.
.134
.74
6.331
50.9
.96
1222
702
141
28-29 . . .
15th.
.132
.74
6.331
50.1
.95
1202
695
127
29-30. ..
16th.
.133
.73
6.401
51.1
.98
1226
713
151
Apr. 30-May 1
17th.
.130
.71
6.549
51.1
.99
1226
717
173
May 1- 2 . . .
18th.
.123
.71
6.549
48.3
.94
1159
682
115
2- 3 . . .
19th.
.127
.72
6.474
49.3
.96
1183
696
136
3-4...
20th.
.124
.72
6.474
48.2
.95
1157
685
101
4- 5 . . .
21st .
.126
.73
6.401
48.4
.96
1162
692
213
5- 6 . . .
22d..
.124
.73
6.401
47.6
.95
1142
684
126
6-7...
23d..
.121
.73
6.401
46.5
.93
1116
668
128
7-8...
24th.
.122
.73
6.401
46.9
.95
1126
678
113
8-9
25th.
.125
.75
6.262
47.0
.95
1128
680
121
9-10. ..
26th.
.123
.73
6.401
47.2
.96
1133
687
120
10-11. . .
27th.
.129
.75
6.262
48.5
1.00
1164
710
166
11-12. . .
28th.
.124
.75
6.262
46.6
.96
1118
682
90
12-13 . . .
29th.
.124
.73
6.401
47.6
.99
1142
701
172
13-14 . . .
30th.
.119
.72
6.474
46.2
.97
1109
685
149
14-15 . . .
31st .
.120
.72
6.474
46.6
.98
1118
694
102
16-172
.133
.78
6.123
48.9
1.04
1174
729
163
17-182
.172
.94
5.290
54.6
1.13
1310
804
1
*Duration of period was usually 8h 30m a. m. to 9h 30m a. m.
2The subject was without breakfast.
390 A STUDY OF PROLONGED FASTING.
The results of the computation of the heat output on these three
bases are given in table 58, in which are recorded first the heat output
per 24 hours, computed on the average values obtained with the bed
calorimeter throughout the night; second, the heat per 24 hours com-
puted from results obtained in the selected minimum periods during
the night; and third, the heat per 24 hours computed from the results
obtained with the respiration apparatus; and finally, the heat per kilo-
gram of body-weight per hour and per square meter of surface per 24
hours computed from the observations with the respiration apparatus.
These values were computed from the total carbon-dioxide output and
the calorific equivalent of carbon dioxide, corrected for the difference
due to the energy from protein, as stated previously, by deducting 0.057
calorie per liter.
The average heat output on the 24-hour basis ranged from 1,898
calories on the first night inside the chamber to a minimum of 1,025
calories on the thirtieth night. Since the results obtained during the
first 4 nights in the chamber were complicated by the previous ingestion
of food, particularly the night of April 10-11, when a large amount of
protein was taken in the evening meal, the results obtained for the
fasting period are more properly compared. The highest value for the
fasting period is 1,441 calories on the first night and the lowest value is
1,025 calories on the thirtieth night. It is thus seen that there was a
steady decrease in the total heat output as the fast progressed, although
from the twenty-first to the thirty-first day the heat production
showed a tendency towards constancy.
Using only the minimum periods during the night, we find a maxi-
mum value of 1,847 calories on the first night in the chamber and a
minimum of 949 calories on the twenty-first night. For the fasting
period, the values ranged from 1,389 calories on the second night of
the fast to 949 calories on the twenty-first night. Here, again, it will
be observed there was a steady fall in the heat production to about the
twenty-first day of the fast, and from that time the heat output showed
a tendency to remain approximately constant for the last 10 nights.
Any possible criticisms of the comparisons of the minimum bed-
calorimeter periods disappear in comparing the values found with the
respiration apparatus. Here we find that the heat production ranged
from 1,615 calories on the morning following the first night of the fast
to 1,109 calories on the morning following the thirtieth night of the
fast. There is a distinct tendency for the heat production to fall pro-
gressively until the twenty-third day, and from that time on it remains
more or less a constant.
Of special importance in connection with this comparison is the fact
that on the first 3 nights the heat production in the bed-calorimeter
experiments was greater than in the experiments with the respiration
apparatus the following morning. This was undoubtedly due to the
CALORIMETRY. 391
heavy evening meal, which, except on the night of April 13-14, con-
tained a not inconsiderable amount of protein. The heat production
on the morning following the fourth night with food was 1 calorie
greater than the average heat output during the night, and the heat
production, from that time on, was invariably larger in the morning
experiments than during the night. Naturally, when we use the values
computed for the minimum periods during the night, this increase in
the heat production in the morning experiment is even greater, the
increase ranging from 287 calories on the morning following the first
night of the fast to 90 calories following the twenty-eighth night of the
fast. These increments, which are given in the last column of the table,
indicate admirably the increase in the total heat production due to the
difference between lying asleep and lying awake.
Comparing the values for the heat production obtained with the
respiration apparatus on the basis of per kilogram of body-weight, we
find that a maximum heat production of 1.13 calories was found on the
morning following the first night of the fast and also on the last morn-
ing of the whole series of experiments, with a minimum heat production
of 0.93 on the morning following the twenty-third night. These
values also show a general tendency to decrease until about the twenti-
eth day of the fast, remaining approximately constant at a low level
for a number of days and rising toward the end of the fast, thus indi-
cating a distinct tendency for the heat production per kilogram of
body-weight to be somewhat higher in the last week than it is in the
third week of fasting.
When the values for the heat production are considered on the basis
of per square meter of body-surface, a maximum value is found of 859
calories on the morning following the first night of the fast, and a
minimum value of 668 calories on the morning following the twenty-
third night of the fast. There is a distinct tendency for this value to
decrease progressively until the eighteenth day, and from that time to
show a general tendency toward equalization, but there is no definite
indication of the increase during the last week of the fast observed in
values found by other methods of computation. It is interesting to
note that the values for the heat production per square meter of body-
surface, as computed from the data obtained with the respiration appa-
ratus, are the only values which do not indicate the tendency to an
increased metabolism in the last week of fasting.
All of these computations of the heat production by the indirect
method therefore substantiate almost completely the inferences
already drawn as to the effect of a prolonged fast upon the respiratory
exchange, i. e., there is here a progressive falling off in the total amount
of heat produced during the first 21 days of fasting, with a tendency
thereafter for the heat to remain constant or to increase somewhat.
BALANCE OF INCOME AND OUTGO.
While it was impracticable to have this fasting subject remain in the
respiration calorimeter for the entire time, as in the fasting experiments
at Wesleyan University, and thus secure ideal conditions for studying
the total metabolism and energy transformation during the fast, yet,
recognizing the great value of a knowledge of the entire 24-hour energy
transformation, we attempted to secure as frequent respiration experi-
ments as possible to supplement the data obtained inside the respiration
calorimeter during the night. This supplied a logical basis for a subse-
quent computation of the total energy transformation.
The intake of this subject consisted of pure distilled water and oxy-
gen from the air. The output consisted of the water of respiration and
perspiration, the carbon-dioxide excretion from the lungs, and the water
and solids of the urine. Theoretically, one should not overlook the
small quantities of solid matter in the perspiration, these being removed
by the baths in distilled water.
The oxygen in the intake appeared in the output combined with
either carbon or hydrogen, since the oxygen entered into the combustion
of body material. In this combustion heat was liberated, which left
the body through various paths — by simple radiation and conduction;
as the sensible heat of excreta, i. e., the heat of the urine; and as heat
required to vaporize water and to warm the inspired air to the body-
temperature.
For a complete understanding of the various components of the out-
put, particularly when considered upon the basis of the total energy
transformation in 24 hours, it is necessary to know the elementary
composition of the excreta. This requires a knowledge of the total
amount of water given off, either as liquid water or as water-vapor, the
carbon, hydrogen, and oxygen in the solid matter of urine, and the ni-
trogen in the urine and the perspiration. Certain of these factors were
directly determined. Thus we have direct determinations of the
nitrogen and carbon in the urine and of the nitrogen of perspiration.
The others can only be computed indirectly, for while we have accurate
evidence regarding the carbon-dioxide excretion throughout the night
period, when the subject was inside the bed calorimeter, we have no
complete evidence of the carbon-dioxide excretion for the daytime,
when he was outside of the calorimeter. Our first problem, then, is
to estimate the probable katabolism of the subject throughout each 24
hours.
TOTAL KATABOLISM PER 24 HOURS.
Knowing, as we do, the influence upon the katabolism of the slightest
muscular activity, it is obvious that the carbon-dioxide output during
the night represents a minimum amount for this subject and that during
392
BALANCE OF INCOME AND OUTGO. 393
the day this will be materially increased. Fortunately the experiments
made with the respiration apparatus supply evidence which enables
us to compute this excretion of carbon dioxide with reasonable accuracy.
Thus we have respiration experiments in the morning, with the subject
lying upon a couch, which give values higher than those obtained during
the night in the respiration chamber. We have also, in the latter part
of the fast at least, a series of observations with the subject lying on
a couch just prior to his entering the calorimeter, which give values
still higher than those obtained with the same apparatus in the morn-
ing. On certain days we have values obtained while the subject was
sitting quietly in a chair or sitting writing. The increase in katabolism
due to these changes in body position and activity has already been
noted in considering the respiratory exchange.
Roughly speaking, the subject was inside the respiration calorimeter
for approximately 11 or 12 hours each night; during 10 hours of this
time the katabolism was directly determined; from the measured value
we may estimate the probable katabolism during the entire sojourn of
11 or 12 hours in the respiration calorimeter. The katabolism was
determined on the respiration apparatus for a period of approximately
2 hours in the morning and an hour at night, making a sum total of
time during which the respiratory exchange was actually measured,
at least in the last part of the fast, of from 14 to 15 hours. The remain-
der of the time, i. e., some 9 or 10 hours, the subject was variously
occupied in the laboratory, for the most part sitting, either writing,
or talking with the various men examining him. A number of respira-
tion experiments were made which were specially designed to secure
information regarding the probable katabolism during such periods of
minor physical activity.
DAILY ACTIVITY.
From the notes made by the observers on the various experimental
record sheets and by the watchers at other times, it was possible to
compute very carefully the number of minutes spent by the subject
on each day of the fast in different occupations. His daily routine
consisted of practically five degrees of activity — lying, sitting resting,
sitting active, standing, and walking. Since there was a distinct
diurnal variation in the katabolism, lying in the evening, which was
based upon the evening respiration experiments, was accorded a dif-
ferent value from that given to lying in the morning.
To show the subdivision of the day into these various activities, the
number of hours and minutes occupied in the various positions is recorded
in table 59. The first column shows the period when the subject was
lying in the morning, this including the time from the end of the calori-
meter experiment to the time that he was weighed. The lying in the
evening includes the entire time from the moment he lay down on the
394
A STUDY OF PROLONGED FASTING.
couch up to the beginning of the first period of the calorimeter experi-
ment, although a part of the time was spent lying inside the calorimeter
chamber, i. e., the preliminary period of the calorimeter experiment.
The third column gives the time spent by the subject lying in the
apparatus during the calorimeter experimental period, which usually
ended about 8 a. m.
Table 59. — Summary of daily activities in experiment with L. (8 a. m. to 8 a. m.).1
Lying.
Sitting.
Date.
Day of
fast.
Stand-
ing.
Walk-
ing.
Morning.
Evening.
Night.
Active.
Resting.
1912.
hr.
min.
hr. min.
hr.
min.
hr
min.
hr
min.
min.
min.
Apr. 14-15
1st. . .
1
49
0 50
10
30
9
2
1
30
4
15
15-16
2d...
1
43
1 29
10
19
8
40
1
30
4
15
16-17
3d...
1
45
1 2
10
38
7
31
2
25
24
15
17-18
4th..
2
0
0 39
11
2
7
8
2
52
4
15
18-19
6th..
1
45
1 13
10
3
8
45
1
30
29
15
19-20
6th..
1
40
1 12
10
30
8
49
1
30
4
15
20-21
7th..
1
35
0 59
10
37
8
19
2
1
14
15
21-22
8th..
1
25
1 27
10
13
8
31
2
0
4
20
22-23
9th..
1
60
1 30
10
10
8
41
1
30
4
15
23-24
10th..
1
24
0 56
10
48
7
57
2
16
24
15
24-25
11th..
1
46
0 59
10
42
7
23
2
51
4
15
25-26
12th..
1
40
1 15
10
20
7
56
2
0
29
20
26-27
13th . .
1
36
2 21
10
7
6
8
3
24
4
20
27-28
14th..
1
40
3 27
9
47
6
4
2
13
14
35
28-29
15th..
1
32
2 27
10
33
7
29
1
30
14
15
29-30
16th. .
1
25
2 37
10
14
7
55
1
30
4
15
Apr. 30-May 1 .
17th . .
1
30
2 30
10
31
7
20
1
30
24
15
May 1-2
18th . .
1
25
2 21
10
45
7
40
1
30
4
15
2-3
19th . .
1
27
3 0
10
9
6
40
2
0
29
15
3-4
20th . .
1
39
2 35
10
25
7
32
1
30
4
15
4-6
21st...
1
25
1 6
10
29
9
1
1
30
14
15
5-6
22d. . .
1
28
3 35
9
46
7
22
1
30
4
15
6-7....
23d. . .
1
45
2 44
10
25
7
17
1
30
4
15
7-8
24th . .
1
28
2 38
10
26
7
19
1
30
24
15
8-9
25th . .
1
30
3 54
9
19
7
3
1
30
29
15
9-10
26th..
1
40
3 9
9
52
7
30
1
30
4
15
10-11
27th..
1
30
3 62
9
13
7
36
1
30
4
15
11-12
28th . .
1
16
2 53
10
5
7
47
1
30
14
15
12-13
29th..
1
55
2 45
10
30
7
31
1
0
4
15
13-14
30th . .
2
0
3 11
9
69
7
26
1
0
4
20
14-15
31st...
2
14
2 23
10
22
6
57
1
0
49
15
xTo the activity here given should be added the work of dressing and undressing, bathing
and raising and lowering the body on the stairs. For explanatio n of estimates in certain portions
of this summary, see text.
For a large part of the day the subject was busy writing, handling his
papers, gesticulating, and arguing, and hence was on a distinctly
higher metabolic level than when he sat quietly with complete mus-
cular repose. The sitting periods have therefore been classified under
two heads, designated respectively as " sitting resting" and "sitting
active." All of the time not otherwise accounted for in the table
is classified as "sitting active," including the periods of the writing
BALANCE OF INCOME AND OUTGO. 395
respiration experiments and the psychological and other tests, except
for the time occupied by Professor Anderson's physical measurements
and Dr. Langfeld's dynamometer tests.
The designation " standing" includes the time spent in the dyna-
mometer tests, in being photographed, in the physical measurements,
and in bathing. The dynamometer tests required 4 minutes, the photo-
graphing 20 minutes, the physical measurements 25 minutes, and the
standing for the bath 10 minutes. Practically 15 minutes each day
were spent in walking, although on a few days somewhat more time than
this was so occupied. Besides the regular daily routine, there were
varying amounts of work done by lifting the body in going up and down
the balcony stairs and occasionally to other portions of the building.
This work has also been taken account of in the computations.
TOTAL CARBON-DIOXIDE PRODUCTION AND OXYGEN CONSUMPTION
PER 24 HOURS.
The results of the computation of the total amount of carbon dioxide
probably produced by the subject and the oxygen consumed in 24
hours are given in table 60. Using the values given in table 59 for the
amount of time spent in the various activities, and the carbon dioxide
and oxygen measured in the morning respiration experiments, we may
easily compute the carbon-dioxide output and oxygen intake of the
subject while lying in the morning. The values given in table 60 for
lying in the evening were obtained by using the data secured in the
evening respiration experiments. These were available only after the
eleventh day of the fast, but since an increment in the metabolism
was regularly noted in the evening experiments over that of the morn-
ing experiments, a correction of 5 per cent has been applied to the morn-
ing values to secure the values for lying in the evening when actual
measurements were not available.
For the calorimeter experiments we have, of course, the direct meas-
urements of the carbon-dioxide output and oxygen intake without
further computation. Moreover, for computing the values for "sitting
active" and "sitting resting," we have on various days direct measure-
ments of both the carbon dioxide produced and the oxygen consumed
during such periods. These absolute values tended to decrease regularly
as the fast continued and with such a degree of regularity that we may
very properly interpolate and thus obtain a highly probable figure for
each day of the fast to supplement the results of actual determinations.
For a relatively short time each day the subject was standing.
Information regarding the increase in the metabolism over lying due to
standing is as yet very incomplete. During the winter of 1913-1914
Dr. Hans Murschhauser, Research Associate of the Carnegie Institution
of Washington, conducted a series of experiments at the Nutrition
396
A STUDY OF PROLONGED FASTING.
Table 60. — Carbon-dioxide production and oxygen consumption in liters per 24 hours in
experiment with L. (8 a. m. to 8 a. m.)
Date.
Day
of fast.
Carbon dioxide.
Lying.
Sitting.
Stand-
Walk-
Dress-
ing and
Going
up and
Total.
Morn-
Even-
Rest-
ing.
ing.
undress-
down
ing.
ing.
Night.
Active.
ing.
ing.
stairs.
1912.
Apr. 14-15
1st.
19.84
9.55
104.05
125.74
16.65
0.80
5.46
3.33
0.89
286.3
15-16
2d..
19.06
17.27
98.55
120.64
16.65
.82
5.55
1.67
.85
281.1
16-17
3d..
18.90
11.72
96.30
101.48
25.96
4.75
5.40
6.44
.82
271.8
17-18
4th.
20.28
6.90
99.15
90.74
29.93
.74
5.07
1.57
.80
255.2
18-19
5th.
16.70
12.19
86.53
105.00
15.21
5.08
4.77
6.08
.79
252.4
19-20
6th.
15.80
11.95
82.99
104.74
14.76
.70
4.74
1.48
.77
237.9
20-21
7th.
14.06
9.15
86.11
94.31
19.36
2.28
4.44
5.76
.76
*237.9
21-22
8th.
13.01
14.01
83.86
96.58
18.72
.67
6.12
1.41
.76
235.1
22-23
9th.
16.61
14.31
81.55
98.47
13.68
.66
4.53
2.74
.77
233.3
23-24
10th.
12.01
8.40
84.19
85.38
20.13
3.77
4.29
3.99
.75
222.9
24-25
11th.
15.16
8.85
81.95
79.30
24.62
.63
4.29
2.59
.75
218.1
25-26
12th.
14.00
11.03
79.92
83.30
17.16
4.47
5.60
3.86
1.48
220.8
26-27
13th.
13.44
19.60
76.16
64.40
28.97
.62
5.60
2.56
.74
212.1
27-28
14th.
14.00
28.34
70.55
63.70
18.75
2.16
9.80
3.81
1.82
*214.5
28-29
15th.
12.33
19.70
74.29
75.43
12.06
2.06
4.02
2.41
1.90
204.2
29-30
16th.
11.22
21.51
72.23
77.90
11.88
.58
3.96
1.19
1.07
201.5
Apr. 30-May 1 .
17th.
11.97
20.10
72.45
69.52
11.97
3.50
3.99
4.78
.70
199.0
May 1- 2
18th.
11.05
18.33
74.03
70.38
11.70
.57
3.90
1.17
1.38
192.5
2-3
19th.
10.70
23.04
68.65
61.20
14.76
3.92
3.69
4.42
.68
191.1
3-4
20th.
12.57
19.53
71.57
69.16
11.43
.56
3.81
1.15
1.36
191.1
4- 5
21st.
10.54
8.58
70.62
82.77
11.16
1.90
3.72
4.46
1.39
*196.5
5- 6
22d..
11.09
26.88
65.29
67.63
11.34
.56
3.78
1.14
1.35
189.1
6- 7
23d..
13.02
20.50
70.39
66.86
11.16
.54
3.72
2.23
.68
189.1
7-8
24th.
10.65
19.91
68.52
69.80
10.89
3.19
3.63
3.27
1.01
190.9
8- 9
25th.
10.98
29.25
62.08
67.26
10.98
3.89
3.66
4.40
.67
193.2
9-10
26th.
12.50
23.44
65.93
71.10
11.25
.55
3.75
1.13
.66
190.3
10-11
27th.
11.07
29.70
61.65
72.05
11.07
.54
3.69
2.21
1.32
193.3
11-12
28th.
9.80
21.80
69.72
73.32
11.61
1.99
3.87
3.48
.98
*198.0
12-13
29th.
14.84
20.30
70.82
70.81
7.44
.54
3.72
2.23
1.31
192.0
13-14
30th.
15.12
23.49
66.05
69.58
7.44
.54
4.96
1.12
1.29
189.6
14-15
31st.
16.48
18.16
71.33
65.05
7.14
6.42
3.57
6.42
.95
195.5
♦The carbon-dioxide production assumed for the work of bathing was on April 20-21, 1.63
liters; April 27-28, 1.54 liters; May 4-5, 1.36 liters; May 11-12, 1.42 liters.
Laboratory with a professional athlete, in which the subject walked on
a treadmill of special construction. Basal figures for these studies were
supplied by the results of a large number of determinations (made by
Dr. Cathcart)1 of the metabolism of this man while in a lying position.
A control experiment which was made within a few months by Mr. L.
E. Emmes, of the Laboratory staff, shows a close agreement with the
earlier observations of Cathcart. Dr. Mursehhauser's observations
other than walking have dealt entirely with determinations of the
metabolism while the subject was standing in various positions, such as
'Benedict and Cathcart, Carnegie Inst. Wash. Pub. 187, 1913, p. 77.
BALANCE OF
INCOME AND OUTGO.
397
Table 60. — Carbon-dioxide production and oxygen consumption in liters
experiment with L. (8 a. m. to 8 a m.) — Continued.
per 24 hours in
Date.
Day
of fast.
Oxygen.
Lying.
Sitting.
Stand-
Walk-
Dress-
ing and
Going
up and
Total.
Morn-
ing.
Even-
ing.
Night.
Active.
Rest-
ing.
ing.
ing.
undress-
ing.
down
stairs.
1912.
Apr. 14-15
1st.
24.31
11.70
133.33
169.10
22.86
0.98
6.69
4.57
1.11
374.7
15-16
2d..
24.41
22.16
130.59
162.24
22.86
1.04
7.11
2.29
1.10
373.8
16-17
3d..
23.84
14.76
131.38
135.30
35.38
6.00
6.81
8.78
1.09
363.3
17-18
4th.
27.12
9.24
133.73
127.97
40.76
1.00
6.78
2.14
1.07
349.8
18-19
5th.
22.26
16.28
115.68
147.53
20.70
6.76
6.36
8.28
1.05
344.9
19-20
6th.
20.50
15.48
122.04
142.30
20.16
.90
6.15
2.02
1.05
330.6
20-21
7th.
19.00
12.39
121.40
131.74
26.38
3.08
6.00
7.84
1.04
*331 . 1
21-22
8th.
17.34
18.62
114.85
134.90
25.44
.90
8.16
1.91
1.03
323.2
22-23
9th.
22.33
19.17
108.45
137.54
18.54
.89
6.09
3.71
1.03
317.8
23-24
10th.
15.96
11.20
116.52
118.77
27.20
5.02
5.70
5.40
1.01
306.8
24-25
11th.
19.82
11.56
112.88
109.42
33.17
.82
5.61
3.49
1.01
297.8
25-26
12th.
18.70
14.70
108.76
117.10
23.28
5.97
7.48
5.24
2.00
303.2
26-27
13th.
17.95
27.21
102.47
90.53
39.58
.82
7.48
3.49
1.00
290.5
27-28
14th.
19.20
40.08
98.06
89.54
26.60
2.95
13.44
5.40
2.53
*299.9
28-29
15th.
16.65
27.93
103.70
106.86
16.92
2.79
5.43
3.38
2.64
286.3
29-30
16th.
15.22
29.67
101.92
110.68
16.92
.79
5.37
1.69
1.49
283.8
A pr. 30-May 1 .
17th.
16.38
28.50
101.14
98.56
16.92
4.80
5.46
6.76
.97
279.5
May 1-2
18th.
15.47
26.51
102.78
98.90
16.92
.80
5.46
1.69
1.94
270.5
2-3
19th.
15.14
34.02
96.22
86.00
21.60
5.54
5.22
6.48
.97
271.2
3-4
20th.
17.52
28.21
100.32
97.18
16.20
.78
5.31
1.62
1.92
269.1
4-5
21st.
14.71
12.01
97.04
116.32
16.20
2.66
5.19
6.48
1.90
*274 . 4
5-6
22d..
15.31
39.13
90.30
95.03
16.20
.76
5.22
1.62
1.90
265.5
6-7
23d..
17.85
28.86
97.73
93.96
15.84
.75
5.10
3.17
.94
264.2
7- 8
24th.
14.52
27.65
99.18
97.46
15.48
4.37
4.95
4.65
1.40
269.7
8- 9
25th.
15.03
41.42
85.69
93.91
15.48
5.34
5.01
6.20
.93
269.0
9-10
26th.
16.60
33.26
94.37
99.90
15.48
.73
4.98
1.55
.92
267.8
10-11
27th.
15.12
41.76
85.63
101.23
15.48
.74
5.04
3.10
1.84
269.9
11-12
28th.
13.07
31.31
98.11
103.21
15.48
2.65
5.16
4.65
1.36
*276.9
12-13
29th.
19.67
29.37
98.32
99.67
10.32
.73
4.98
3.10
1.82
268.0
13-14
30th.
20.76
34.00
91.75
98.57
10.32
.75
6.84
1.55
1.79
266.3
14-15
31st.
22.38
26.17
99.41
92.16
10.32
8.97
4.98
9.30
1.34
275.0
*The oxygen consumption assumed for the work of bathing was on April 20-21, 2.20 liters;
April 27-28, 2.11 liters; May 4-5, 1.90 liters; May 11-12, 1.89 liters.
" attention/' "relaxed," and similar positions. Using as a basis Cath-
cart's and Emmes' figures for lying and Dr. Murschhauser's values for
standing, we find an increase over lying of about 10 per cent. Conse-
quently this 10 per cent increment has been used to compute the value
for standing for our fasting subject.
For the values for walking we have relied implicitly upon the exten-
sive series of observations made by Dr. Murschhauser with the
professional athlete. This subject showed an increment of 190 per
cent in the metabolism while walking at a slow pace — 60.6 meters
per minute — over that while lying. Recognizing the fact that Dr.
398 A STUDY OF PROLONGED FASTING.
Murschhauser's subject was a trained athlete and represented the
highest degree of muscular vigor and that our subject was a non-
athlete, flabby, and, as the fast progressed, still more emaciated and
less inclined to physical exercise, we have employed but 100 per cent
for the computation of the value for L., since there was here the possi-
bility of a very considerable error. Fortunately, as will be seen by
reference to table 60, the total carbon-dioxide production and oxygen
consumption while the subject was walking was hardly 2 per cent of
the total value for the day; we may therefore consider any error present
as entirely negligible.
For the numerous physical tests and the photographing, the subject
had to dress and undress at least once or twice each day and on some
days several times. The metabolism involved in dressing and undress-
ing was studied in a large number of experiments in the respiration
calorimeter at Wesley an University,1 and it was found that the resting
metabolism was increased 30 per cent during one hour. Accordingly
the values obtained while the subject was sitting resting were increased
by 30 per cent of the resting metabolism for one hour to allow for this
muscular work of dressing and undressing. Here, again, the absolute
values involved are extremely small when compared with the total for
the day. On 4 days an additional allowance was made for the work of
bathing, this being considered as equivalent to the metabolism for 10
minutes of standing. The value for this on the seventh day of the fast
was 1.63 liters of carbon dioxide and 2.20 liters of oxygen; on the four-
teenth day it was 1.54 liters of carbon dioxide and 2.11 liters of oxygen;
on the twenty-first day it was 1.36 liters of carbon dioxide and 1.90
liters of oxygen; and on the twenty-eighth day it was 1.42 liters of
carbon dioxide and 1.89 liters of oxygen.
A further computation was made of the metabolism involved in the
work of going up and down stairs. For lack of better evidence, it was
estimated that the amount of work performed in going down stairs
was equal to half of that going up. The number of kilogrammeters
were computed by multiplying the weight of the man by the height
through which he lifted his body, increasing this by one-half when
the work of going down stairs was included. The number of calories
were then obtained by dividing the number of kilogrammeters by the
factor 426.6. Finally, by using the calorific equivalent for the aver-
age respiratory quotient for the day, computed from the results of
the experiments with the calorimeter and the respiration apparatus,
we were able to calculate the amount of carbon dioxide excreted and
oxygen consumed in the muscular work of lifting the body. The total
probable carbon-dioxide production and oxygen consumption for the
24 hours is therefore the sum of the values computed for the varying
degrees of activity; these totals are given in the last column of table 60.
Benedict and Carpenter, Carnegie Inst. Wash. Pub. 126, 1910, p. 247.
BALANCE OF INCOME AND OUTGO. 399
In estimating the probable accuracy of the figures, it is evident that
the greatest error is likely to be found in the values for the metabolism
with the subject sitting active, since this is a not inconsiderable part
of the metabolic activity for the day. An increment of approximately
25 per cent was actually measured during the writing experiments and
it has been assumed that when the subject was talking, gesticulating,
handling his papers, writing, and not otherwise at complete rest, he
was on a metabolic plane comparable to that during the actual period
of measurement while he was writing. The values on certain days are
undoubtedly a little high and on others a little low, but it is highly
improbable that there is an appreciable percentage error when the total
24-hour amounts are considered. The results of these computations
for the carbon-dioxide production and oxygen consumption per 24
hours are therefore probably within 3 per cent of the values which
would have been obtained had it been possible to make complete and
continuous determinations of the respiratory exchange.
The total 24-hour quantity of carbon dioxide produced steadily
decreased as the fast progressed, with slightly higher values in the last
week. The maximum, 286.3 liters, was found on the first day and the
minimum, 189.1 liters, on the twenty-second and twenty-third days.
The oxygen consumption followed essentially a parallel course, with
a maximum of 374.7 liters on the first day and a minimum of 264.2
liters on the twenty-third day.
CHARACTER OF THE KATABOLISM.
In the discussion of the metabolism of the fasting subject in previous
sections of this book, emphasis has been laid upon the minimum
amounts or comparable values obtained under like external conditions,
in an attempt to study exclusively the influence of the internal pro-
cesses upon the metabolism, particularly the effect of the prolonged
fasting. Hence we find emphasis laid upon the daily alteration in
rhythm and the influence of inanition as the fast progressed.
Since Rubner's law of the isodynamic value of the nutrients makes it
of relatively little importance as to whether the diet of a normal indi-
vidual contains 25 or 40 per cent of its total energy in carbohydrates,
interest in normal katabolism is centered chiefly upon the total heat-
production. On the other hand, it is clear that with a subject living
upon body material, any evidence regarding the character of the com-
position of the katabolized material and the source of energy, particu-
larly in regard to the effect of prolonged fasting upon them, will also
have much more than ordinary interest.
We have here a subject living substantially the same daily routine
for 31 days of fasting and the 3 days following with food. In the days
prior to the fast the subject was distinctly more active. Utilizing the
results of the numerous respiration experiments, the measurements
400 A STUDY OF PROLONGED FASTING.
of the metabolism made during the long sojourn of the subject in the
respiration calorimeter at night, the daily records of the occupations of
the subject during the fast, and standard figures regarding the effect
upon the katabolism of variations in activity, we are able to make for
practically every hour of the day a reasonably satisfactory estimate of
the probable activity and thus obtain the total katabolism.
When complete 24-hour determinations of the various factors enter-
ing into the computation of the character of the katabolism can be
obtained, namely, the direct measurement of the total carbon-dioxide
output, oxygen intake, and elementary composition of the solid matter
of urine, the most exact method for computing the character of the
katabolism and the kinds and the amounts of material burned is that
outlined in the previous publication on inanition, in which use was
made of simultaneous equations.1 When, as is the case with this
fasting subject, an element of uncertainty enters into the fundamental
figures which would be used in such computations and it is necessary
to make our estimates of the total carbon-dioxide output and oxygen
intake from the results of fragmentary, periodic determinations, it
does not seem justifiable, or indeed logical, to employ the extremely
exact mathematical computation used in the earlier publication.
It is true that the protein katabolism may be as readily obtained for
this experiment as for the experiments previously reported, since we
have data as to the total nitrogen excretion, but for the apportionment
of the katabolism between fat and carbohydrate a simpler method is
probably more justifiable. We have thus first computed the protein
katabolism, and then apportioned the katabolism of carbohydrate
and fat by a method to be subsequently described.
PROTEIN KATABOLISM.
In accordance with the current conception of the relationship
between the nitrogen in the urine and protein katabolism, the total
amount of protein katabolized may be computed from the total amount
of nitrogen in the urine by using the percentage of nitrogen in the tissue
broken down. In ordinary metabolism experiments, when the food
ingested is of varying composition and contains proteins of various
kinds, it is commonly assumed that the nitrogen in the tissue broken
down is represented by the factor 6.25, so that the amount of nitrogen
in the urine may be multiplied by this factor to obtain the amount of
protein in the tissue broken down. In a fasting experiment, however,
only body-material is katabolized and the percentage of nitrogen in
dry, fat-free, and ash-free flesh has been shown to be 16.62; hence the
factor 6.0 should be used instead of 6.25 for computing the protein in
the flesh katabolized from the amount of nitrogen in the urine. Accord-
ingly this factor was used for our computations of the protein kata-
bolism in this experiment.
Benedict, Carnegie Inst. Wash. Pub. 77, 1907, pp. 36 and 452.
BALANCE OF INCOME AND OUTGO. 401
In a previous section of this publication (see table 27, p. 251), it has
been shown that the nitrogen in the urine gradually decreased as the
fast progressed, but the nitrogen excretion per kilogram of body-weight
is also of interest in this connection as showing whether or not the
decrease was proportional to the loss in weight. From the values given
in table 27 it will be seen that after the first 4 days of the fast, during
which there was a marked rise in the nitrogen excretion, the values
per kilogram of body-weight had a tendency to fall as the fast pro-
gressed, the highest value (0.207 gram) being found on the fourth day
of fasting and the lowest (0.146 gram) on the twenty-third and
thirty-first days of the fast. It is thus unquestionable that there
was a distinct tendency in the latter part of the fast for the nitrogen
excretion per kilogram of body-weight to be lower than in the first
part of the fast. The values for the total nitrogen excreted and
the computed amounts of protein katabolized are given in columns
a and p, table 61. The protein katabolized ranged from a maxi-
mum of 71.2 grams on the fourth day of the fast to a minimum of
41.6 grams on the last day, the course of the protein curve naturally
following that of the nitrogen excretion. Finally, on the assump-
tion that flesh contains 20 per cent protein and 80 per cent water,
the flesh equivalent of the protein katabolized has been computed by
multiplying the protein by 5, the results being given in column q of
table 61. These values obviously follow those for the nitrogen and
the protein, the maximum value being 356 grams on the fourth day of
fasting and the minimum 208 grams on the last day of the fast.
APPORTIONMENT OF NON-PROTEIN KATABOLISM BETWEEN CARBOHYDRATE
AND FAT.
The method used in this publication for computing the katabolism of
fat and carbohydrate for our fasting subject was, in brief, as follows:
The amounts of carbon dioxide produced and oxygen consumed in
the combustion of carbohydrate and fat were first obtained by deducting
the computed amounts for the combustion of protein from the total
amounts of carbon dioxide produced and oxygen consumed. From
these values the non-protein respiratory quotient for each day of the
fast was readily calculated. Using these non-protein respiratory
quotients, we next found the total heat given off in the combustion of
the carbohydrate and fat, the apportionment of the heat output
between them being made subsequently by the use of certain factors.
From the heat output due to the combustion of the carbohydrate
and the fat, the amounts of these two body materials katabolized were
finally calculated by means of their heats of combustion.
402
A STUDY OF PROLONGED FASTING.
Carbon Dioxide Produced and Oxygen Consumed in the Katabolism of
Carbohydrate and Fat.
Since we know the amount of nitrogen excreted in the urine, the car-
bon dioxide produced and the oxygen consumed in the katabolism of
protein may be computed by using factors established by Zuntz and
his co-workers. Thus, for every gram of nitrogen in the urine, Zuntz
has found that 4.75 liters of carbon dioxide are produced and 5.91 liters
of oxygen are consumed. The total nitrogen excretion for each day
of the fast is given in column a of table 61, and the carbon dioxide
produced and the oxygen consumed in the combustion of protein are
given in columns c and f respectively. Deducting these values from
Table 61.
—Body materials katabolized and total heat production
[computed)
per 24 hours
in
experiment with L.
Carbon dioxide.
Oxygen.
Non-
Date.
Day
of fast.
Nitrogen
in urine.
Total.1
From
protein
burned
(AX4.75).
From fat
and
carbo-
hydrate
(b-c).
Total.1
For com-
bustion
of
protein
(AX5.91).
For com-
bustion
of fat and
carbo-
hydrate
(e — f).
protein
respi-
ratory
quotient
(D-rG).
A
B
C
D
E
F
G
H
1912.
gm.
liter 8.
liters.
liters.
liter 8.
liters.
liters.
Apr. 14-15 ....
1st.
7.10
286.3
33.7
252.6
374.7
42.0
332 . 7
0.76
15-16
2d..
8.40
281.1
39.9
241.2
373.8
49.6
324.2
.74
16-17
3d..
11.34
271.8
53.9
217.9
363.3
67.0
296.3
.74
17-18
4th.
11.87
255.2
56.4
198.8
349.8
70.2
279.6
.71
18-19
5th.
10.41
252.4
49.4
203.0
344.9
61.5
283.4
.72
19-20
6th.
10.18
237.9
48.4
189.5
330.6
60.2
270.4
.70
20-21
7th.
9.79
237.9
46.5
191.4
331.1
57.9
273.2
.70
21-22
8th.
10.27
235.1
48.8
186.3
323.2
60.7
262.5
.71
22-23
9th.
10.74
233.3
51.0
182.3
317.8
63.5
254.3
.72
23-24
10th .
10.05
222.9
47.7
175.2
306.8
59.4
247.4
.71
24-25
11th.
10.25
218.1
48.7
169.4
297.8
60.6
237.2
.71
25-26
12th.
10.13
220.8
48.1
172.7
303.2
59.9
243.3
.71
26-27
13th.
10.35
212.1
49.2
162.9
290.5
61.2
229.3
.71
27-28
14th.
10.43
214.5
49.5
165.0
299.9
61.6
238.3
.69
28-29
15th.
8.46
204.2
40.2
164.0
286.3
50.0
236.3
.69
29-30
16th.
9.58
201.5
45.5
156.0
283.8
56.6
227.2
.69
Apr. 30-May 1 .
17th.
8.81
199.0
41.8
157.2
279.5
52.1
227.4
.69
Mav 1- 2
18th.
8.27
192.5
39.3
153.2
270.5
48.9
221.6
.69
2-3
19th.
8.37
191.1
39.8
151.3
271.2
49.5
221.7
.68
3-4
20th.
7.69
191.1
36.5
154.6
269.1
45.5
223.6
.69
4-5
21st .
7.93
196.5
37.7
158.8
274.4
46.9
227.5
.70
5-6....
22d..
7.75
189.1
36.8
152.3
265.5
45.8
219.7
.69
6-7....
23d..
7.31
189.1
34.7
154.4
264.2
43.2
221.0
.70
7-8....
24th.
8.15
190.9
38.7
152.2
269.7
48.2
221.5
.69
8-9
25th.
7.81
193.2
37.1
156.1
269.0
46.2
222.8
.70
9-10
26th.
7.88
190.3
37.4
152.9
267.8
46.6
221.2
.69
10-11
27th.
8.07
193.3
38.3
155.0
269.9
47.7
222.2
.70
11-12
28th.
7.62
198.0
36.2
161.8
276.9
45.0
231.9
.70
12-13
29th.
7.54
192.0
35.8
156.2
268.0
44.6
223.4
.70
13-14
30th
7.83
189.6
37.2
152.4
266.3
46.3
220.0
.69
14-15
31st
6.94
195.5
33.0
162.5
275.0
41.0
234.0
.69
'See table 60.
BALANCE OF INCOME AND OUTGO.
403
the computed 24-hour values for the total carbon dioxide given off
and oxygen consumed, we obtain the carbon dioxide and oxygen for
the combined combustion of the fat and carbohydrate, these amounts
being given in columns d and g of the table. The ratio of these two
values gives the so-called non-protein respiratory quotient, which is
recorded in column h.
Significance of the Non-Protein Respiratory Quotients.
Since the non-protein respiratory quotient has a special significance
in this connection, a discussion of the values for the quotients given in
table 61 may be made here before going further with the computation
of the amounts of body materials katabolized.
Table 61.
-Body materials katabolized and total heat production (computed) per 2/t hours in experiment
with L. — Continued.
Heat computed.
Body materials katabolized.
Flesh
equi-
Date.
Day
of fast.
From
fat and
car-
bohy-
drate.1
From
car-
bohy-
drate.1
From
fat.1
From
protein
(AX26.51).
Total
(j+k+l).
Carbo-
hydrate
(j+4.23).
Fat
(k 4-9.54).
Protein
(AX6.0).
valent
of pro-
tein
(pX5).
I
J
K
L
M
N
o
P
Q
1912.
cats.
cats.
cats.
cats.
cats.
gm.
gm.
gm.
gm.
Apr. 14-15
1st.
1581
291
1290
188
1769
68
.8
135
42.6
213
15-16
2d..
1533
178
1355
223
1756
42
.1
142
50.4
252
16-17
3d..
1401
163
1238
301
1702
38
.5
130
68.0
340
17-18
4th.
1311
18
1293
315
1626
4
.3
136
71.2
356
18-19
5th.
1333
64
1269
276
1609
15
.1
133
62.5
312
19-20
6th.
1267
1267
270
1537
133
61.1
306
20-21
7th.
1280
1280
260
1540
134
58.7
294
21-22
8th.
1231
17
1214
272
1503
4
.0
127
61.6
308
22-23
9th.
1196
57
1139
285
1481
13
.5
119
64.4
322
23-24
10th.
1160
16
1144
266
1426
3
.8
120
60.3
302
24-25
11th.
1113
16
1097
272
1385
3
.8
115
61.5
308
25-26
12th.
1141
16
1125
269
1410
3
.8
118
60.8
304
26-27
13th.
1075
15
1060
274
1349
3
.5
111
62.1
311
27-28
14th.
1117
1117
277
1394
117
62.6
313
28-29
15th.
1107
1107
224
1331
116
50.8
254
29-30
16th.
1065
1065
254
1319
112
57.5
287
Apr. 30-May 1.
17th.
1066
1066
234
1300
112
52.9
265
May 1-2
18th.
1038
1038
219
1257
109
49.6
248
2-3
19th.
1039
1039
222
1261
109
50.2
251
3-4
20th.
1048
1048
204
1252
110
46.1
231
4-5
21st.
1066
1066
210
1276
112
47.6
238
5-6
22d..
1030
1030
205
1235
108
46.5
233
6-7
23d..
1036
1036
194
1230
109
43.9
220
7-8
24th.
1038
1038
216
1254
109
48.9
245
8-9
25th.
1044
1044
207
1251
109
46.9
235
9-10
26th.
1037
1037
209
1246
109
47.3
237
10-11
27th.
1041
1041
214
1255
109
48.4
242
11-12
28th.
1087
1087
202
1289
114
45.7
229
12-13... .
29th.
1047
1047
200
1247
110
45.2
226
13-14. . . .
30th.
1031
1031
208
1239
108
47.0
235
14-15
31st.
1097
1097
184
1281
115
41.6
208
xFor the factors used in this computation see Williams, Riche and Lusk, Journ. Biol. Chem., 1912, 12, p. 357.
404 A STUDY OP PROLONGED FASTING.
We found in this experiment none of the extraordinarily low values
reported by other observers with fasting man, the lowest quotient
obtained with our subject being 0.68, which was found for but one day.
When the difficulties in securing accurate determinations of the
respiratory quotients are considered — all of the errors in the determina-
tion of both the oxygen consumption and the carbon-dioxide production
affecting the value of the respiratory quotient — it is perhaps surprising
that values more abnormal than those shown in column h are not
found. It is of course not impossible that abnormal katabolism result-
ing from the acidosis of this man may in part account for this slightly
lowered respiratory quotient.
In fasting there is unquestionably a larger excretion of ammonia in
the urine and likewise an excretion of /3-oxybutyric acid. Indeed,
an attempt has been made by Grafe1 to correct for the alteration in the
values for the respiratory quotient when ammonia is excreted in the
urine instead of the ordinary urea. It may be questioned, however,
whether data obtainable in fasting experiments are sufficiently accurate
and the oxygen measurements are sufficiently exact to warrant such
an attempt. Furthermore, since the alteration from urea to ammonia
is merely a matter of hydrolysis, the computation may prove to be
somewhat problematical. In all probability the protein katabolism
is not greatly affected, for while there is a distinct disturbance in the
relationship between the ammonia and the total nitrogen, there is not
an excessive amount of amino-acids formed, as the rest nitrogen is not
unduly large. On the other hand, Magnus-Levy2 has clearly shown
that the formation of acetone bodies due to the partial oxidation of
fat results in the absorption of oxygen and production of carbon diox-
ide, thus influencing slightly the respiratory quotient, but concludes
that a production of 40 grams of /3-oxybutyric acid alters the respira-
tory quotient not more than 0.012.
A not inconsiderable proportion of the total energy transformation,
as here computed, is based upon experiments made with the respiration
apparatus, in which the cutaneous respiration was not considered.
The respiratory quotient might thus have been slightly higher had the
cutaneous respiration also been determined, Magnus-Levy mentioning
0.015 as the probable correction to be applied to the respiratory quo-
tient determined by nose or mouth appliances.
While the determinations of the respiratory quotient by means of
the bed calorimeter and the respiration apparatus check each other
and show that the results are within a few units of being accurate,
nevertheless when the protein katabolism correction amounts to but
a few thousandths and the correction for the formation of acetone
bodies also amounts to only a few thousandths, it can be seen that the
iGrafe, Zeitschr. f. physiol. Chemie, 1910, 65, p. 21.
2Magnus-Levy, Zeitschr. f. klin. Med., 1905, 56, p. 83.
BALANCE OF INCOME AND OUTGO. 405
degree of accuracy in the determination of the respiratory quotient
must be extraordinarily high to admit of an intelligent discussion
of these points. Furthermore, it is reasonable to question whether
or not one can logically use the determination of the respiratory
quotient for analyzing in any way the organic processes which are
affected during fasting. It is certain, of course, that the combustion
of fat forms the greatest part of the total combustion. On the other
hand, there may be a slight formation of carbohydrate on some days
which would alter the quotient ; there may also be an abnormal excre-
tion of materials in the urine which would likewise alter the respiratory
quotient; and there is always a possibility of the production of /3-oxy-
butyric acid. All of these factors, however, affect the quotient but
slightly, and therefore no great stress need be laid upon them. In fact,
it is doubtful, as was pointed out in the earlier fasting publication,
whether the computation of the total glycogen output was sufficiently
accurate to justify the assumption that glycogen was actually produced
on 2 days of fasting. I believe that the computations of the 24-hour
gaseous exchange in the experiment with L. are not sufficiently accu-
rate to throw definite light upon the question as to whether or not
glycogen was produced after the first few days of the fast.
The general course of the respiratory quotients found with L. is
wholly in accord with those observed with the Middletown subject
S. A. B., who remained in the respiration calorimeter the entire period
of the experiment, except that the quotients obtained for the latter
subject were almost invariably somewhat higher. Thus, the average
respiratory quotients for the 7 days of the fasting experiment with
S. A. B. were 0.78, 0.75, 0.74, 0.75, 0.74, 0.75, and 0.74, these values
being substantially higher than those observed for our subject L. on
the fourth to seventh days of his fast. The respiratory quotients for
L. given in column h of table 61 are, however, non-protein respiratory
quotients, while those here given for S. A. B. are the total respiratory
quotients, i. e., they include the protein katabolism. Hence the dis-
parity is not so great, as the respiratory quotient for protein is not far
from 0.81. In any event, it is clear that we do not have here sufficient
evidence of so great a disturbance in either fat, carbohydrate, or protein
katabolism as to lead us to believe that there was a fundamental altera-
tion in the character of the katabolism, other than that accompany-
ing the acidosis during the 31-day fast.
Energy Derived from Katabolism of Carbohydrate and Fat.
Using these values for the non-protein respiratory quotient, the oxy-
gen consumption required for the combustion of the fat and carbohydrate
(column g of table 61), and the calorific equivalents of oxygen given in
Zuntz and Schumburg's table,1 we are able to compute the energy
^untz and Schumburg, Physiologie des Marsche3, Berlin, 1901, p. 361.
406 A STUDY OF PROLONGED FASTING.
derived from the carbohydrate and fat katabolized. For instance,
on April 14-15, the first day of the fast, the non-protein respiratory
quotient was 0.76. The calorific equivalent of oxygen with this re-
spiratory quotient is 4.752 calories per liter; hence the total heat given
off from the combustion of the fat and carbohydrate would be equal
to the oxygen consumed in the combustion (332.7 liters) multiplied
by the calorific equivalent for this non-protein respiratory quotient,
i. e., 4.752, the calculation being 332.7X4.752 = 1,581 calories. The
values for the total heat resulting from the combustion of the fat and
carbohydrate for each day of the fast are given in column i of table 61.
The table of Zuntz and Schumburg has been elaborated in a most
helpful and ingenious way by Williams, Riche, and Lusk,1 who have
apportioned (in terms of percentage of total energy) the fat-carbo-
hydrate katabolism, using the non-protein respiratory quotient. With
a non-protein respiratory quotient of 0.76, these authors compute that
18.4 per cent of the energy is derived from carbohydrate and 81.6 per
cent from fat; consequently, on April 14-15 about 18.4 per cent of the
1,581 calories as calculated previously, or 291 calories, are derived from
the katabolism of carbohydrate, and the remaining 1,290 calories from
fat. Obviously, as the non-protein respiratory quotient falls, the per-
centage of energy derived from carbohydrate becomes less and less,
until, at 0.70 or less, it is assumed that no carbohydrate is burned. In
fact, on several days when the non-protein respiratory quotient was
0.69 and on one day 0.68, we have followed Lusk's usage in employing
the value for the quotient of 0.70.
Amounts of Carbohydrate and of Fat Katabolized.
Since Williams, Riche, and Lusk were particularly interested in the
energy transformation, their apportionment of the non-protein katab-
olism between fat and carbohydrate was expressed in percentage of
energy from carbohydrate and percentage of energy from fat. For our
purpose, since we desire the weight of the carbohydrate and fat burned,
it is very simple to compute the heat production and then by dividing
by the appropriate calorific values for carbohydrate and fat, obtain
directly the number of grams of carbohydrate and fat entering into the
katabolism.
The heat of combustion of the carbohydrate burned in the body is
taken as that of glycogen, 4.23 calories per gram,2 while the heat of
combustion for the fat is taken as that of body fat, 9.54 calories per
gram.3 By dividing the respective amounts of energy by these heats
of combustion, the number of grams of glycogen and body fat partici-
pating in the katabolism are computed. These are recorded in columns
n and o in table 61. While this method of computing the probable
Williams, Riche, and Lusk, Journ. Biol. Chem., 1912, 12, p. 357.
2Emery and Benedict, Am. Journ. Physiol., 1911, 28, p. 301.
3Benedict and Osterberg, Am. Journ. Physiol., 1900, 4, p. 69.
BALANCE OF INCOME AND OUTGO. 407
24-hour amounts of body materials katabolized is by no means com-
parable with that resulting from exact continuous measurements with
the respiration calorimeter, it nevertheless has a value.
Katabolism of carbohydrate. — On the basis of the non-protein respi-
ratory quotients as computed for 24 hours, it is seen that there is a
combustion of body carbohydrate or glycogen for the first 5 days and a
probable combustion for 6 days later in the fast. The maximum
amount of carbohydrate katabolized was 68.8 grams on the first day
of fasting and the minimum was practically 4 grams on the 4 days
from the tenth to the thirteenth days inclusive. These amounts are
somewhat smaller than those found for the average of all of the short
fasting experiments at Wesleyan University. In these experiments
the average katabolism for the 7 days was 110, 40.3, 21.8, 23.3, 8.2, 21.7,
and 18.7 grams, respectively; these values are not dissimilar from those
found in the 7-day experiment with S. A. B., in which but 64.9 grams of
glycogen were katabolized on the first day. The findings with L. are
wholly in line with those observed on S. A. B. and show that there is a
material katabolism of body carbohydrate or glycogen continuing
throughout the first 7 to 10 days of the fast.
Katabolism of fat. — The katabolism of fat decreased regularly as the
fast progressed, the maximum amount of 142 grams being observed on
the second day of fasting and the minimum amount of 108 grams on
the twenty-second and thirtieth days of fasting respectively.
LOSS OF WATER FROM THE BODY.
The loss of organized tissue has been calculated in the previous
section with great care and thus light has been thrown upon the trans-
formations of matter in the body during a prolonged fast. Water does
not enter into energy changes and yet the data regarding the loss to
the body by the excretion of water are of much interest.
An important observation was made in connection with the report
of the 7-day fasting experiment in the earlier book1 to the effect that
there was distinct evidence that a large amount of preformed water,
other than that of flesh, was discharged from the body in the first 4
or 5 days of the fast and thereafter the loss of water was approximately
proportional to the flesh disintegrated and the fat burned.
Water, in so far as it enters into the katabolism of the fasting man,
may be considered as coming from two sources, first, preformed water,
and second, the water of oxidation of organic matter. From the
determined amounts of carbohydrate, fat, and protein katabolized, it is
possible, by means of well-known chemical formulae, to compute the
amount of organic hydrogen oxidized during the fast, but the draft
upon the store of preformed water as the fast progressed is of much
more significance.
Benedict, Carnegie Inst. Wash. Pub. 77, 1907, p. 467.
408 A STUDY OF PROLONGED FASTING.
Although L. went without food for 31 days, and hence the compli-
cated factor of determining the intake of water in the food was elimi-
nated, nevertheless he drank a definite amount of water each day, and
water was lost from the body through various paths. We know that
in the urine a certain amount of water was mixed with the solids. This
was determined very carefully in the latter part of the fast, so that we
had accurate information as to the relationship between the total
amount of water and total solids, and were accordingly able to compute
the water excreted in the urine during the earlier days of the fast.
Water was also lost by vaporization from the lungs and skin. In the
method formerly used for obtaining the loss of water from the body
it was likewise necessary to determine carefully the amount of water
thus vaporized in the 24 hours. In this fasting experiment, while such
determinations were made for the period when the subject was inside
the respiration calorimeter, we have no direct evidence regarding the
probable water vaporized from the lungs and skin during the time that
L. was outside of the chamber. Even the respiration experiments,
while supplying information regarding the carbon-dioxide output and
oxygen intake, give no evidence as to the excretion of water. Accord-
ingly, the total outgo of water must be determined by indirect compu-
tation. Fortunately the data regarding the katabolism of L. are so
full and exact that we are justified in employing such a method.
LOSS OF PREFORMED WATER.
To determine the actual drafts upon body material in the form of
preformed water, the computation proceeds in the following manner,
the results being given in table 62. We have first the " insensible
perspiration" (column a), which is likewise recorded in table 4 (page
84). From the discussion in the preceding section we have informa-
tion as to the amounts of carbohydrate, fat, and protein which were
katabolized in the body. The carbohydrate was completely oxidized,
leaving the body in the form of carbon dioxide and water, and hence
was a complete loss. The fat was similarly burned to carbon dioxide
and water. On the other hand, only a portion of the protein was
actually oxidized and, so to speak, volatilized and lost from the body.
Using standard figures, Loewy1 has computed that for each 100
grams of combustible anhydrous flesh which is katabolized there are
available for oxidation and conversion to carbon dioxide and water
41.50 grams of carbon, 4.40 grams of hydrogen, and 7.69 grams of
oxygen. Thus, of the total protein molecule, 53.6 per cent may be
burned and 46.4 per cent may be excreted in urine or feces. In com-
puting the total weights of carbohydrate, fat, and protein oxidized and
volatilized in the body we must therefore allow for 46.4 per cent of the
protein excreted in the urine. The values given in column b of table 62
1Loewy, Oppenheimer's Handbuch der Biochemie, Jena, 1911, 4 (1), p. 156.
BALANCE OF INCOME AND OUTGO.
409
for the anhydrous material burned therefore represent the total weight
of carbohydrate burned (column n of table 61), the total weight of fat
burned (column o), and 53.6 per cent of the total weight given in table
61 (column p) for the amount of protein burned.
Since the insensible perspiration represents the actual insensible
loss from the body, the amount of preformed water vaporized may be
Table 62 —
Preformed water vaporized or excreted from the body and drafts
upon the
original supply in experiment with L.
Body
material
(anhy-
drous)
burned.1
Preformed water.
Vaporized or excreted.
Loss of preformed water.
Date.
Day
Insen-
sible
Vapor-
Water
From
of fast.
perspi-
ized
Ex-
con-
From
sources
ration.
from
creted
Total
sumed.
Total
From
flesh
other
body
in
(C+D).
(e— f).
fatty
(JVX6.0
than fat
(A-B).
urine.
tissue.2
X4).
and flesh
g-(h+i).
A
B
C
D
£
F
G
H
I
J
1912.
gm.
gm.
gm.
gm.
gm.
gm.
gm.
gm.
gm.
gm.
Apr. 14-15
1st.
1086
227
859
630
1489
720
769
14
170
585
15-16
2d..
1188
211
977
437
1414
750
664
14
202
448
16-17
3d..
1059
205
854
531
1385
750
635
13
272
350
17-18
4th.
779
179
600
674
1274
750
524
14
285
225
18-19
5th.
727
182
545
634
1179
750
429
13
250
166
19-20
6th.
606
166
440
578
1018
750
268
13
244
11
20-21
7th.
603
166
437
496
933
750
183
13
235
- 65
21-22
8th.
569
164
405
557
962
750
212
13
246
- 47
22-23
9th.
578
167
411
575
986
750
236
12
258
- 34
23-24
10th.
672
156
516
535
1051
750
301
12
241
48
24-25
11th.
573
152
421
535
956
900
56
12
246
-202
25-26
12th.
691
154
537
490
1027
900
127
12
243
-128
26-27
13th.
436
148
288
532
820
900
- 80
11
248
-339
27-28
14th.
540
151
389
619
1008
900
108
12
250
-154
28-29
15th.
442
143
299
736
1035
900
135
12
203
- 80
29-30
16th.
578
143
435
861
1296
900
396
11
230
155
Apr. 30-May 1 .
17th.
509
140
369
821
1190
900
290
11
212
67
May 1- 2
18th.
521
136
385
633
1018
900
118
11
198
- 91
2- 3
19th.
550
136
414
705
1119
900
219
11
201
7
3- 4
20th.
371
135
236
679
915
900
15
11
184
-180
4-5
21st.
623
138
485
685
1170
900
270
11
190
69
5-6
22d..
465
133
332
763
1095
900
195
11
186
- 2
6- 7
23d..
504
133
371
537
908
900
8
11
176
-179
7-8
24th.
480
135
345
728
1073
900
173
11
196
- 34
8- 9
25th.
468
134
334
692
1026
900
126
11
188
- 73
9-10
26th.
473
134
339
706
1045
900
145
11
189
- 55
10-11
27th.
557
135
422
631
1053
900
153
11
194
- 52
11-12
28th.
477
139
338
634
972
900
72
11
183
-122
12-13
29th.
554
134
420
676
1096
900
196
11
181
4
13-14
30th.
530
133
397
751
1148
900
248
11
188
49
14-15.. ..
31st.
625
137
488
547
1035
900
135
12
166
- 43
1AUowance has been made for the proportion (46.4 per cent) of the protein katabolized assumed to have
been excreted in the urine.
2It is assumed that the amount of water from fatty tissue is equal to 10 per cent of the fat katabolized.
410 A STUDY OF PROLONGED FASTING.
obtained by deducting the total weight of anhydrous material burned
from the total amount of insensible perspiration. The results are
recorded in column c of table 62, these values representing the total
amount of water actually vaporized from the lungs and skin of the
subject. We know from former experience1 that the preformed water
lost in this way is affected largely by muscular activity and that it is
made up in part of water which has been combined with flesh and with
fatty tissue.
In addition to the loss of water through the lungs and skin, there was
also a loss of preformed water through the urine, the amounts being
given in column d, while the total amount of preformed water, either
vaporized or excreted, is shown in column e. There was, however,
not only an outgo but an income of water, as the subject drank a
measured amount of water each day. To find the actual loss of pre-
formed water from the store in the body, this income of water (see
column f) should be deducted, the values in column g showing the
actual loss of preformed water from the store in the body.
The total loss of preformed water varied from 769 grams on the first
day to 8 grams on the twenty-third day; as a matter of fact the body
actually stored 80 grams of water on the thirteenth day. The heaviest
losses occurred on the first few days of the fast. In this discussion it
is necessary to consider that the body loses fairly regularly a quantity
of protein and fat each day. This loss is accompanied by a loss of the
water normally combined in the flesh and fatty tissue, the amount
being relatively constant and averaging not far from 200 grams of water
per day. The portion of the body remaining (which may, for conven-
ience, be called the "residue") likewise has a water-content which under-
goes fluctuation. The total loss of preformed water in column g may
therefore be classified under three heads : (1) water of flesh broken down ;
(2) water of fatty tissue broken down ; (3) water from other sources,
%. e., from the residue.
The amounts of water from the flesh disintegrated and fatty tissue
broken down may be computed by the use of certain factors. Thus it
is assumed that there are 4 grams of water combined with every gram
of protein to form flesh; furthermore, that body fat is combined with
10 per cent2 of its weight in water to form fatty tissue. The draft
upon the store of preformed water due to the katabolism of protein
may therefore be found by multiplying the amount of protein kata-
bolized by 4, and the draft upon the preformed water due to the
katabolism of fatty tissue by finding 10 per cent of the amount of fat
Benedict, Carnegie Inst. Wash. Pub. 77, p. 429.
2Munk (Lehmann, Mueller, Munk, Senator, and Zuntz, Archiv f. path. Anat. u. Physiol.,
1893, 131, Supp., p. 216) in computing the body-losses uses 10 per cent; Albu and Neuberg
(Physiol, u. Path, des Mineralstoffwechsels, Berlin, 1906, p. 9) also assume 10 per cent. On the
other hand, Bozenraad (Deutsch. Archiv f. klin. Med., 1911, 103, p. 120) states that the water-
content of human fat may vary from 7 to 46 per cent. The average water-content of the fat in
fat people was 13.2 per cent, while in emaciated persons it was 28.2 per cent.
BALANCE OF INCOME AND OUTGO. 411
katabolized. These values have been computed and the results are
given in columns i and h of table 62. Deducting the amount of pre-
formed water lost through the breaking down of flesh and fatty tissue
from the total amount of preformed water lost from the body, we have
the preformed water lost from sources other than fat and flesh (col-
umn j). It is obvious that the minus signs here indicate an actual
addition of water to the residue rather than a loss.
From the figures given in column g, it is seen that the total loss of
preformed water from the body in the experiment with L. became less
as the fast progressed — that is, there was a tendency to conserve the
store of water. This is the more apparent if we consider the figures
in column J, which represent the loss or gain of preformed water in the
residue. It should be noted, however, that in the computation of these
values it was assumed that the proportion of water in the flesh and
fatty tissue katabolized was not altered during the course of the fast —
an assumption which may very properly be made, since the fluctuations
in the water-content of the flesh and fat katabolized, as compared with
those in the water-content of the whole body, would be so small as to
be negligible.
From the figures in column j it will be seen that on the first 5 days
of the fast there was a very considerable loss of water from the residue
other than that belonging to the flesh and fat katabolized. A period
of 5 days followed in which there was an approximate water equilibrium.
When the amount of water taken by the subject was increased at the
suggestion of Dr. Goodall, there was a retention of water by the residue
for some 4 days, thus implying an actual physiological water-need.
From that time until the end of the fast, with the exception of 4 days
when approximately 50 to 100 grams of water were lost, the store
of water showed a distinct tendency towards an addition of water to
the residual body-material. From the eleventh to the thirty-first day
of the fast there was a positive addition to the storage of water in the
residue amounting to 1,383 grams.
This evidence of a retention of water in the body of the fasting man
is of especial interest, inasmuch as it is in conformity with other experi-
mental evidence showing a tendency, during fasting,1 for the flesh to
increase in water-content. It has been demonstrated by Abderhalden,
Bergell, and Dorpinghaus,2 however, that there is no change in the
composition of the proteid in the body, for, according to the Fischer
esterification method, the composition of the fractions does not alter.
^ichtenfelt, Archiv f. d. ges. Physiol., 1904, p. 353; and Sedlmair, Zeitschr. f. Biol., 1899,37,
p. 25. That the greatest storage is probably in flesh is shown by Engels, Archiv exp. Path. u.
Pharm., 1903, 51, p. 346; other depots are noted by Mayer, Compt. rend. Soc. Biol., 1906, 60,
p. 588, and Paladino, Biochem. Zeitschr., 1912, 38, p. 443.
'Abderhalden, Bergell, and Dorpinghaus, Zeitschr. f. physiol. Chem., 1904, 41, p. 153.
412
A STUDY OF PROLONGED FASTING.
TOTAL LOSS OF ORIGINAL BODY SUBSTANCE.
From an analysis of the foregoing data we are in a position to state,
with a considerable degree of accuracy, the loss of carbohydrate, fat,
protein, and preformed water, and solids excreted in urine and from the
skin, and thus present a general picture of the total loss of original body
substance during fasting. This is shown in table 63. The combustion
of carbohydrate ceased after the first 13 days of the fast. The combus-
Table 63. — Distribution of loss
, in grams of original body-substance per 2U hours,
in experiment with L.
Date.
Day
of fast
Body materials katabolized.
Preformed water lost.
Solid excreta.
Protein.
Fat.
C
Car-
bohy-
drate.
D
From
flesh
(aX4).
E
From
fatty
tissue.2
F
From
sources
other
than
flesh
and
fat.2
G
Total
solids
in
urine.8
H
From
skin.4
I
Feces.
J
Total
(NX6.0).
A
Burned
(aXO.536).1
B
1912.
Apr. 14-15 ....
15-16.. ..
16-17
17-18.
18-19 ....
19-20
20-21
21-22
22-23
23-24
24-25
25-26
26-27
27-28
28-29
29-30
Apr. 30-May 1 .
Mav 1- 2
2- 3
3- 4
4- 5
5- 6
6- 7
7- 8
S- 9
9-10
10-11
11-12
12-13
13-14
14-15
1st.
2d..
3d..
4th.
5th.
6th.
7th.
8th.
9th.
10th.
11th.
12th.
13th.
14th.
15th.
16th.
17th.
18th.
19th.
20th.
21st.
22d..
23d..
24th.
25th.
26th.
27th.
28th.
29th.
30th.
31st.
42.6
50.4
68.0
71.2
62.5
61.1
68.7
61.6
64.4
60.3
61.5
60.8
62.1
62.6
60.8
57.5
52.9
49.6
50.2
46.1
47.6
46.5
43.9
48.9
46.9
47.3
48.4
45.7
45.2
47.0
41.6
22.8
27.0
36.4
38.2
33.5
32.7
31.5
33.0
34.5
32.3
33.0
32.6
33.3
33.6
27.2
30.8
28.4
26.6
26.9
24.7
25.5
24.9
23.5
26.2
25.1
25.4
25.9
24.5
24.2
25.2
22.3
135
142
130
136
133
133
134
127
119
120
115
118
111
117
116
112
112
109
109
110
112
108
109
109
109
109
109
114
110
108
115
68.8
42.1
38.5
4.3
15.1
4.0
13.5
3.8
3.8
3.8
3.5
170
202
272
285
250
244
235
246
258
241
246
243
248
250
203
230
212
198
201
184
190
186
176
196
188
189
194
183
181
188
166
14
14
13
14
13
13
13
13
12
12
12
12
11
12
12
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
12
585
448
350
225
166
11
- 65
- 47
- 34
48
-202
-128
-339
-154
- 80
155
67
- 91
7
-180
69
- 2
-179
- 34
- 73
- 55
- 52
-122
4
49
- 43
43.51
45.38
50.62
56.13
49.09
46.07
40.58
44.14
46.81
42.49
42.05
39.21
42.01
40.58
32.50
41.12
39.51
35.47
34.59
30.06
31.88
31.18
29.30
32.01
30.32
31.04
31.52
29.06
29.64
29.58
27.07
•2.20
1.52
1.04
■ .94
?
B
O
^ee page 408.
2See table 62.
'Amounts calculated for first 15 days of fast. (See table 24.)
4Urea plus sodium chloride calculated from values given in table 22 (p. 234.) Excreta from the skin were
determined each week, e. g., from the time of bathing on evening of April 13 to the time of bathing on the
evening of April 20. For the total nitrogen excreted in the urine each day, see table 29.
BALANCE OF INCOME AND OUTGO. 413
tion of fat and protein continued without material alteration in relative
proportions until the end of the fast. The preformed water lost from
the body came from three sources, first, that combined with the protein
to form flesh; second, that combined with fat to form fatty tissue;
third, water from other sources in the body. Finally, we have the
solid excreta from the body made up of the solids of the urine1 and the
solids from the skin. There was no fecal loss.
TOTAL ENERGY LOSS.
The loss in energy as the fast progressed was in two forms, i. e., the
kinetic energy in the form of heat, which includes the sensible heat of
excreta, the heat lost by radiation and conduction, and the heat of
vaporization of water; and the potential energy of the solid excreta,
chiefly that of the organic material in the urine.
Considering first the kinetic energy or the heat liberated, it is clear
that the heat results from the actual combustion of organic material,
such as protein, carbohydrate, and fat. From our former computations
(see table 61), we can apportion the heat-production among these three
compounds. This apportionment is shown in table 64, which, for sub-
sequent discussion, also gives the total amount of energy lost from the
body.
The greatest loss of heat falls upon the fat, with the protein supply-
ing the greater part of the remainder. As a matter of fact, although
the carbohydrate (glycogen) is rapidly drawn upon and apparently the
available supply is quickly depleted, sufficient glycogen is burned the
first day of the fast to furnish somewhat more heat than was supplied by
the protein on that day. After the first 3 days, the heat was derived
almost wholly from a fat-protein katabolism, with a minimum supply of
glycogen. The combustion of glycogen ceases after the thirteenth day.
The actual quota each compound supplies to the daily energy require-
ment is best seen by placing the heat output subdivision upon the per-
centage basis. This is done in table 65, which shows that on the first
3 days from 10 to 16 per cent of the heat was derived from glycogen.
Subsequently, the glycogen combustion furnished hardly more than
1 to 3 per cent of the heat. After the thirteenth day, approximately
17.5 per cent of the heat was derived from the oxidation of the protein
and the remainder from the combustion of fat.
In the fasting experiment with L., the body lost not only the kinetic
energy of material burned, but also the potential energy of solid unoxi-
dized material contained in the urine. Since there were no feces during
the fast, no potential energy was lost in this way. While both glyco-
gen and fat burn completely to form carbon dioxide and water in
*It is obvious that the unoxidized material of protein is a constituent in two columns of the
table. In column a, the total protein includes the unoxidized material which appears again in
the solids in the urine.
414
A STUDY OF PROLONGED FASTING.
normal katabolism, under certain pathological conditions and in
inanition, the fat combustion is defective and certain partially oxidized
bodies — the so-called "acetone bodies" — are produced and excreted
chiefly in the urine.
In the earlier report of fasting experiments1 direct evidence of the
presence of these bodies was not available, and in computing the heat
Table 64. — Distribution of heat produced per 24 hours to body-materials katabolized, and total
energy loss of the body in experiment with L.
Heat produced (calculated).
Date.
Day of
fast.
Energy
(potential)
of urine.
Total
energy
loss
From car-
From
From
Total
bohydrate.
fat.
protein.
(A + B + C).
(D + E).
A
B
C
D
E
F
1912.
cats.
cats.
cols.
cols.
cals.
cals.
Apr. 14-15
1st
291
1290
188
1769
65
1834
15-16
2d
178
1355
223
1756
89
1845
16-17
3d
163
1238
301
1702
118
1820
17-18
4th....
18
1293
315
1626
134
1760
18-19
5th....
64
1269
276
1609
123
1732
19-20
6th....
1267
270
1537
116
1653
20-21
7th....
1280
260
1540
104
1644
21-22
8th....
17
1214
272
1503
116
1619
22-23
9th ... .
57
1139
285
1481
124
1605
23-24
10th....
16
1144
266
1426
111
1537
24-25
11th....
16
1097
272
1385
110
1495
25-26
12th
16
1125
269
1410
105
1515
26-27
13th....
15
1060
274
1349
114
1463
27-28
14th....
1117
277
1394
111
1505
28-29
15th.. .
1107
224
1331
95
1426
29-30
16th....
1065
254
1319
123
1442
Apr. 30-May 1
17th....
1066
234
1300
117
1417
May 1- 2
18th
1038
219
1257
104
1361
2-3
19th....
1039
222
1261
105
1366
3-4
20th....
1048
204
1252
91
1343
4- 5
21st
1066
210
1276
95
1371
5- 6
22d
1030
205
1235
93
1328
6- 7
23d
1036
194
1230
88
1318
7-8
24th....
1038
216
1254
95
1349
8- 9
25th....
1044
207
1251
91
1342
9-10
26th....
1037
209
1246
90
1336
10-11
27th ....
1041
214
1255
90
1345
11-12
28th ....
1087
202
1289
85
1374
12-13
29th....
1047
200
1247
87
1334
13-14
30th
1031
208
1239
87
1326
14-15
31st
1097
184
1281
80
1361
from protein we assumed erroneously that all of the unoxidized material
in the urine was derived from the cleavage products of protein. The
method of simultaneous equations there employed is fundamentally
based upon the assumption that the cleavages all proceed in a normal
manner. By this method, the total carbon excreted, including that
of urine and carbon dioxide, was apportioned among the protein, car-
Benedict, Carnegie Inst. Pub. 77, 1907, p. 495.
BALANCE OF INCOME AND OUTGO.
415
bohydrate, and fat on the assumption that the katabolism proceeded
in a normal manner. As a matter of fact, inspection will show that
while the amounts of carbon, hydrogen, and oxygen employed in the
various equations would not be greatly affected in their distribution by
this method of computation, the error in apportionment has a theoreti-
cal importance which should not be overlooked. If due consideration
be shown to these conditions, it will be seen that the method of appor-
tioning the protein, fat, and carbohydrate combustion employed in the
present research may be distinctly preferable to that in the earlier
publication, although the problem of defective-fat katabolism is only
in part theoretically solved by this method.
Table 65. — Proportions
of total heat
production derived from different body-materials in
experiment with L. {Computed per 2 4 hours.)
From
From
From
Date.
Day of
car-
From
protein
Date.
Day of
From
protein
fast.
bohy- .
fat.
(net
fast.
fat.
(net
drate.
energy).
energy).
1912.
p. ct.
p. ct.
p. cl.
1912.
p. ct.
p. ct.
Apr. 14-15...
1st...
16.5
72.9
10.6
Apr. 30-May 1 .
17th . .
82.0
18.0
15-16...
2d...
10.1
77.2
12.7
May 1- 2
18th . .
82.6
17.4
16-17.. .
3d...
9.6
72.7
17.7
2-3
19th . .
82.4
17.6
17-18...
4th..
1.1
79.5
19.4
3- 4
20th . .
83.7
16.3
18-19. . .
5th. .
4.0
78.9
17.1
4-5
21st.. .
83.5
16.5
19-20. . .
6th..
82.4
17.6
5-6
22d...
83.4
16.6
20-21 . . .
7th..
83.1
16.9
6-7
23d...
84.2
15.8
21-22. . .
8th..
1.1
80.8
18.1
7- 8
24th . .
82.8
17.2
22-23.. .
9th. .
3.8
76.9
19.3
8-9
25th. .
83.5
16.5
23-24 . . .
10th..
1.1
80.2
18.7
9-10
26th . .
83.2
16.8
24-25...
11th..
1.2
79.2
19.6
10-11
27th..
82.9
17.1
25-26. . .
12th . .
1.1
79.8
19.1
11-12
28th . .
84.3
15.7
26-27...
13th..
1.1
78.6
20.3
12-13
29th . .
84.0
16.0
27-28. . .
14th..
80.1
19.9
13-14
30th..
83.2
16.8
28-29. . .
15th . .
83.2
16.8
14-15
31st . .
85.6
14.4
29-30...
16th..
80.7
19.3
It is important, therefore, in considering the total energy loss from
the body, to include not only the energy given out as heat and appor-
tioned among the carbohydrate, fat, and protein, but also the amount
of potential energy lost in the urine; this potential energy, from the
cleavage products of protein and fat, is given in table 64, column e,
as derived from the heat of combustion. The entire loss from the body
of this fasting man thus represents the kinetic energy of the materials
burned plus the potential energy of the solid materials in the urine,
this total loss being shown in column f of table 64 for each day of the
fast. In general, both the loss in kinetic and total energy steadily
decreased as the fast continued and this man lost from his body, in the
form of kinetic and potential energy, an average of 1,489 calories per
day for 31 consecutive days.
416 A STUDY OF PROLONGED FASTING.
Table 66. — Complete metabolism results with subject L. during 4 days pre-fasting
period, 31 days fasting, and 3 days post-fasting period.
In no series of derived tables is it possible to draw complete comparisons of
the various factors of metabolism, and this can be done satisfactorily only
when all of the data are grouped together. I have been requested by a number
of physiologists to summarize all of the data obtained with this subject in
one large table at the end of the book. Since so much stress has been laid
upon the desirability of such a presentation, the data are given here in full.
In order to bring together comparative data for each day of the fast, it
has been necessary to place in each column results which represent in the aggre-
gate a total of 36 hours, each single result, however, being either taken at the
end of 24 hours or representing a total for 24 hours. The results recorded on
the thirty-first day to which an asterisk has been affixed were, as a matter of
fact, obtained a short time after the first food had been taken.
Fio. 47. — Metabolism chart of the most important factors measured on subject L.
throughout the fast.
Although table 66 gives the exact mathematical expression of the quantities
of materials involved in the different measurements of the metabolism of this
subject, a visualization of the relationships between the various factors of
metabolism is best secured by a series of curves. A chart containing curves
for all of the factors measured on this subject would be of such great size as
to preclude convenient inspection. Hence we have collected here only the
most important or more generally observed factors. Although no specific
reference is made to these curves for any comparisons in the text, it is obvious
that a constant reference to this chart is presupposed in a careful reading of
the report.
H
n*tl ■
T
f
it.-
BINDING SECT AUG 1 * 1972
>
PLEASE DO NOT REMOVE
CARDS OR SLIPS FROM THIS POCKET
UNIVERSITY OF TORONTO LIBRARY
QP Benedict, Francis Gano
J^1, A study of prolonged
B396 fasting
BioMed
Live Music Archive
Librivox Free Audio
Metropolitan Museum
Cleveland Museum of Art
Internet Arcade
Console Living Room
Books to Borrow
Open Library
TV News
Understanding 9/11