THE DISCOVERY OF THE ACCESSORY FOOD FACTORS, OR VITAMINS
THE accessory food factors were discovered by physio-
logists and biochemists who were investigating the func-
tion in nutrition of the individual groups of food substances
such as " pure " protein, fat, salts, etc. The pure
substances were prepared in the laboratory and their
effect upon the body tested by feeding experiments. A
convenient protein for this purpose is the casein of milk,
for not only can it be easily prepared, but also several
preparations consisting almost entirely of this protein
can be purchased. These commercial preparations are
salt-free, and, after extraction with alcohol and ether to
remove fats and lipoids, consist of practically pure protein.
Pure proteins can also be prepared from egg white, blood
serum, nuts, legumes, cereals, etc. Many " pure "
carbohydrates are available : starch, dextrin, cane sugar,
milk sugar, grape sugar. Lard is generally regarded as
the purest fat ; other pure fats are olive oil and cotton-
seed oil. A mixture of salts has generally been com-
pounded from the purest chemical specimens, this
mixture being made to resemble in quantity and propor-
tion the mineral constituents of a normal diet. Some-
times the ash of the animal's normal diet is used instead
of a salt mixture.
One of the earliest feeding experiments on animals was
made by Lunin (1881) to study the effect of salts in the
food. He fed mice upon a diet of specially prepared milk
fat, casein and cane sugar ; some mice were given salts in
addition and others had no salts. Both sets of mice died
within thirty-six days, but other mice, used as a control,
DISCOVERY OF ACCESSORY FOOD FACTORS 45
fed on fresh milk alone, lived and grew normally. Lunin
concluded from these facts that " other substances indis-
pensable for nutrition must be present in milk besides
casein, lactose, fats and salts."
Many independent workers have since confirmed Lunin 's
results and found it impossible to keep animals for any
length of time on artificial diets intended to contain every-
thing essential to life. The failure was attributed to
confined space, absence of fresh air and the monotony
of the diet, but these suppositions did not explain the fact
that animals lived in good health under exactly similar
conditions of confinement on a monotonous diet of a
natural raw product such as milk or egg.
The discovery of some substance, or substances, in the
food other than protein, carbohydrate and fat, which were
essential for the life of animals, was made by Prof. F. G.
Hopkins; the work was published in 1912, but had been
carried out some years earlier and delayed by his ill health.
Hopkins kept two sets of young rats, approximately i to 2
ounces in weight, on a basal diet consisting of purified
casein, lard, sugar and salts derived from the ash of oats
and dog biscuits, the animals' previous food. The experi-
mental diet satisfied physiological requirements in respect
to protein, fat, carbohydrate, salts and energy value.
One set of rats was given in addition 2 to 4 c.c. of milk
daily, fed by hand, so that it was definitely known to be
consumed. The other set had no milk. The set with milk
grew normally ; the non-milk set declined in weight after
ten to fifteen days.
The same experiment was repeated in an even more
striking fashion. Eight rats were used in each set;
eighteen days after those without milk had begun to
decline in weight they were given milk, and the other set
had their milk discontinued. The original milk set now
declined in weight, but growth began again in the other
set. The results were most graphically illustrated by
their weight charts (Figs. 8 and 9).
In discussing these results Hopkins wrote :
46 VITAMINS AND THE CHOICE OF FOOD
"It is possible that what is absent from artificial diets
and supplied by such addenda as milk and tissue extracts is
of the nature of an organic complex (or complexes) which
the animal body cannot synthesise. But the amount which
seems sufficient to secure growth is so small that a catalytic
or stimulative function seems likely. . . . Stimulation of the
internal secretions of the thyroid and pituitary glands, which
FIG. 8. Lower curve is the average weight of six rats on artificial
diet alone. Upper curve, six similar animals receiving in addition
2 c.c. milk per day.
Reproduced by kind permission from the Journal of Physiology, IQ", 44, 432.
are believed by very suggestive evidence to play an important
part in growth processes, can be legitimately thought of. On
the other hand the influence upon growing tissues may be
direct. If the attachment of such indispensable functions
to specific accessory constituents of diets is foreign to current
views upon nutrition, so also is the experimental fact that
young animals may fail to grow when they are daily absorbing
a sufficient quantity of formative material and energy for the
purpose of growth."
DISCOVERY OF ACCESSORY FOOD FACTORS 47
Hopkins termed the unknown substances accessory food
At the time that Hopkins was carrying out these experi-
ments upon rats, Stepp in Germany tried to find out if
animals could live without fat or lipoids in their food. He
found that mice could live for months on a diet of bread
FIG. 9. Lower curve (up to i8th day) shows average weight of 8 rats
upon pure dietary; upper curve 8 similar rats taking 3 c.c. milk each
day. On the i8th day, marked by vertical dotted line, the milk
was transferred from one set to the other.
Reproduced by kind permission from the Journal of Physiology , 1912, 44, 433.
made with milk, but that they only lived for a few weeks
if the same food were first extracted with alcohol and ether
to remove the fat and lipoids. On adding the extract to
the extracted food, the diet again sufficed for normal
nutrition. Stepp at first thought that the substance
essential for growth removed in the extract must be a
lipoid, but on adding in turn all the known lipoids to a
48 VITAMINS AND THE CHOICE OF FOOD
diet of extracted food he failed to maintain the mice.
By continued experiments he proved that the missing
substance was neither fat, lipoid, nor salt, nor any known
chemical substance, and he concluded that " the unknown
substance indispensable for life may go into solution
with the lipoids, and that the latter thereby become what
may be termed ' carriers ' for this substance."
Osborne and Mendel in America, since 1911, have carried
out a large number of feeding experiments on rats to
study the quality of protein in nutrition. Their basal diet
Protein in the form of casein (or gliadin from wheat, or
edestin from hemp, etc.).
Fat in the form of lard.
Carbohydrate in the form of starch or cane sugar.
Salts in the form of a carefully compounded mixture of
pure salts. This mixture was replaced later by a prepara-
tion called " protein-free milk," a kind of whey containing
the salts and sugar of milk but not the fat and casein.
The replacement of the salt mixture by the protein-free
milk was necessitated by the decline of adult rats after a
period of well-being lasting for about 290 days; their
health was restored by replacing the salt mixture and
some of the carbohydrate by the protein-free milk.
On the same diet containing protein-free milk young
rats grew normally for about forty to eighty days, but
then a decline set in and the animals died unless
put back on normal food. The decline could also be
arrested by the substitution of whole-milk powder for
pure protein. Like Hopkins and Stepp, Osborne and
Mendel were forced to conclude that whole milk supplied
some special substance essential for life. Further investi-
gations showed that this growth-promoting substance was
contained in the fat of milk, but not in the other milk
solids nor in the water of milk. Other fats were then
tried ; egg fat or cod-liver oil were as good as butter fat,
but lard or almond oil did not induce growth.
At the same time two other American workers,
DISCOVERY OF ACCESSORY FOOD FACTORS 49
McCollum and Davis, were making independent experi-
ments upon rats to test the value of different salt mixtures
in nutrition. The basal diet of these rats was
Protein in the form of purified casein.
Fat in the form of lard.
Carbohydrate in the form of lactose (milk sugar) and
starch or dextrin.
Salts : the mixture to be tested.
On this artificial food-mixture young rats grew for about
sixty to eighty days and then failed to make further
growth. The rats resumed growth if an ether extract of
egg or butter were given, but extracts of lard, cotton-seed
oil, or of olive oil had no restorative effect.
The combined results of these two separate groups of
workers proved that fats can be divided into two classes,
which may be called good or bad for nutrition :
Good Fats. Bad Fats.
Cod-liver oil. Almond oil.
Egg fat. Olive oil.
Kidney fat. Cotton-seed oil.
From these results it seemed as if the only essential
growth-promoting substance were contained in the good
fats, but neither the diet used by Osborne and Mendel
nor that used by McCollum and Davis could be called
absolutely " pure." Osborne and Mendel used protein-
free milk instead of pure salts, and McCollum and Davis
used a commercial preparation of lactose. McCollum,
whilst carrying out some experiments with birds fed on
white polished rice and suffering from polyneuritis, dis-
covered that the polyneuritis could be cured by the
addition of commercial lactose. This new discovery led
McCollum and Davis to modify their basal diet for rats;
they replaced the lactose by dextrin or starch, and found
that no appreciable growth took place even with the inclu-
sion of 5 per cent, butter fat in the diet. Similarly if the
50 VITAMINS AND THE CHOICE OF FOOD
protein-free milk in Osborne and Mendel's experimental
diet were replaced by artificial salt mixtures the rats did
not grow even if butter or egg fat were used instead of
lard. The commercial lactose in one case and the
protein-free milk in the other were thus supplying another
DISCOVERY OF ACCESSORY FOOD FACTORS 51
essential factor. Two substances, one soluble in water,
the other in fat, were therefore necessary to secure growth
and maintain health.
McCollum named these
two accessories " fat
soluble A " and " water
soluble B," as their
chemical nature was un-
known. They objected
to the term " vitamine "
introduced some years
previously by Funk in
connection with beri-
beri. They disliked it
partly because it was
not proved that the un-
known substances were
organic bases as the term
" amine " implied, and
partly because other
organic substances, such
as tryptophan (p. 105),
are also essential for life.
The term " accessory
food factor " used by
Hopkins was not con-
sidered satisfactory, as
it suggested a subordin-
ate role and might lead
to a confusion with
condiments. The word
" vitamine " has, how-
ever, crept into general
use and is now spelled
" vitamin," the terminal
e having been dropped, as it implied that the substances
had a known chemical constitution.
The absence of A-vitamin in a diet with the conse-
52 VITAMINS AND THE CHOICE OF FOOD
quent cessation of growth is shown in curves i, 2 and 3
of Fig. 10. At the point marked o, A-factor (butter fat)
was added to complete the diet, and growth was resumed.
Curve 4 in this figure shows the normal growth of young
rats nursed by mothers on an artificial diet containing
butter fat. In curve 5 is seen the cessation of growth
of young rats nursed by mothers on a diet deficient in
The need of B-vitamin for growth is illustrated in
Fig. ii. The dotted curves 3, 4, 5, and the dotted parts
of curves i and 2 show loss of weight in absence of this
vitamin. Resumption of growth immediately followed
the addition of the vitamin as shown in the continuous
line of the curve.
A young rat suffering from B-vitamin deficiency is
seen in Fig. 12. The animal is unable to use its hind
legs, which are stretched out in a helpless manner. After
a dose of yeast extract, the animal was able to walk
with ease at the end of twenty-four hours.
THE NEED FOR A THIRD ACCESSORY FACTOR.
Rats, on account of their small size, are a convenient
animal for laboratory experiments. It is possible to prepare
sufficient artificial food for their use for long periods, and
they soon show the effect of any inadequacy in the diet.
Further, their short span of life, about three years, enables
records to be kept for more than one generation.
Rats are able to live, grow, reproduce and rear healthy
offspring on a diet containing purified protein, fat, carbo-
hydrate and salts, together with some source of A and
B factors. Upon such a diet man would get scurvy
unless some fresh fruit or vegetables were also given.
This peculiarity is not confined to man but shared with
several other species of animals. Guinea-pigs, monkeys
and also pigs suffer from scurvy, and Captain Cook
recorded symptoms of scurvy in goats and sheep after
several months on board ship. Rats, cats, rabbits and
birds show no visible signs of scurvy.
DISCOVERY OF ACCESSORY FOOD FACTORS 53
The antiscorbutic substance is now included amongst
the vitamins, and is called " water soluble C." The three
FIG. 13. Weight charts of two male rats on diets containing
A-, B-, and C- Vitamins and A- and B- Vitamins respectively.
8 12 16 20 24 28 32
FIG. 14. Weight charts of two female rats on similar diets.
(Harden and Zilva.)
Reproduced by kind permission from the Biochemical Journal. 1018, 12, 414
(Cambridge University Press).
vitamins are often called A, B, and C factors. All three
factors are really essential for perfect growth and main-
54 VITAMINS AND THE CHOICE OF FOOD
tenance ; even rats grow better if the G factor is added to
their basal diet as well as A and B factors ; this is shown
in the weight charts reproduced in Figs. 13 and 14.
PROBABLE IDENTITY OF THE WATER-SOLUBLE B FACTOR
AND THE ANTIBERI-BERI VITAMIN.
The water-soluble substance contained in commercial
lactose was found by McCollum and Davis to cure poly-
neuritis in birds, and also to supply the deficiency in the
diet of rats fed on purified protein, butter fat, carbo-
hydrate and salts. They therefore concluded that poly-
neuritis in birds was the specific result of a lack of the
water-soluble B factor in the food. It is, of course,
impossible to prove the identity of the antiberi-beri vita-
min with the B factor until the substance has been
isolated and compared. Their probable identity is in-
dicated by their very similar distribution as well as their
relative proportion in the various food-stuffs (which is
seen in the following table), and by their similar
B Factor tested
tested on Birds.
Value for Antiberi-
+ -f +
+ + +
+ + +
+ + + +
+ + +
+ + +
+ + +
Commercial Meat Extract
White Bread .
Polished Rice. .
Both Factors missing.
The identity of these two factors is now generally