Go to National Library of New Zealand Te Puna Mātauranga o Aotearoa
Volume 58, 1928
This text is also available in PDF
(996 KB) Opens in new window
– 264 –

The Vitamin “A” (Anti-Ophthalmic) Content of Butter and Spinach.

[Read before the Otago Institute, 14th September, 1926; received by Editor, 31st December, 1926; issued separately, 8th November, 1927.]

In 1913, McCollum and Davis (1) and Osborne and Mendal (2,3) showed that the presence of vitamin A or fat soluble vitamin was necessary in the diet of albino rats in order to secure normal growth, for the prevention of eye disease and other infections, and for satis-factory reproduction. Without vitamin A normal growth was retard-ed after 60 to 70 days of age, many rats developed an ophthalmia, and unless the deficiency was supplied the weight declined and death resulted at about 3 ½ months of age.

In 1919 E. Mellanby in work on puppies (4) showed that a fat soluble substance seemed necessary for the proper calcification of the bones and called it the “anti-rachitic factor,” because of its value in the prevention and cure of rickets. At that time it was thought a single factor, fat soluble vitamin A, was responsible for all the results when rats suffered from a lack of the fat soluble substances.

Since 1922 the work of Powers, Park and Simmonds (5), McCollum and his associates (6), and Steenbock and associates (7), has shown the existence of at least two substances involved in the fat soluble fraction, one the anti-ophthalmic and the other, the anti-rachitic factor. This is shown by the destruction of the anti-ophthalmic factor in cod liver oil by aeration, (i.e. oxidation) where as the oil after this treatment is still potent in its effect in the cure of rickets, and the securing of a normal deposition of calcium in the bones and of promoting some growth.

As a consequence most of the work hitherto done on vitamin A or fat soluble vitamin will now need to be repeated, or at least reinterpreted in the light of the more recent work, for we must be able to distinguish between the effects resulting from the presence or absence of these two very different substances, the anti-ophthalmic (vitamin A) and the anti-rachitic (vitamin D).

Steenbock, Nelson and Black (8) have revised the technic for determination of vitamin A, so as to differentiate between vitamin A and vitamin D. In order to determine vitamin A, it is necessary to supply vitamin D by the use of aerated cod liver oil or direct radia-tion with ultra violet rays. By means of a basal ration free from both vitamins, addition of aerated cod liver oil or direct radiation supplies vitamin D only, and foods can then be tested for their con-tent of vitamin A. After the vitamin A content is found we can test for content of vitamin D.

– 265 –

We have tested out our basal ration which lacks both vitamins A and D to be certain of its deficiency, we have then added aerated cod liver oil to the basal ration and determined the effects resulting upon this addition of vitamin D. With this ration (basal plus aerated cod liver oil) we have then tested butter and spinach for their content of vitamin A. By the results we can distinguish between the effects produced by the two fat soluble substances in question.

Experimental Procedure.

In these experiments albino rats were used, whose parents were of the same stock and had all been raised on our standard colony diet, with the exceptions noted in the first experiment. This ensured an identical previous pedigree and nutritional history. They were separated from their mothers at 28 or 29 days of age and put on the experimental basal ration. They were weighed once a week and the tables and charts give the average weekly weights. It has been repeatedly determined, that animals of the same age placed on a vitamin -A-free diet, are able to survive the deprivation for varying lengths of time, depending upon the extent to which the previous diet has allowed opportunity for storage of this factor (9). Hence our animals had uniform advantages from the previous diet. Another important condition governing the length of the survival period, is the age of the experimental animals (10). Our animals have been taken at a uniform age, and weighing from 32 to 52 gms. The controls on the basal diet were chosen so as to represent the different litters used in the experiments. No average is taken using only the rats from one litter, for occasionally a whole litter will show peculiar susceptibility or resistance to a lack of vitamin A. The rats of any one litter were therefor put on different amounts of the food being tested.

Large numbers of rats would smooth out the curves, but the general relationships and tendencies from the different diets would probably remain the same, for the deviation of individual rats from the averages is not unreasonable, and the results from all of the rats on any one diet are near enough to the average, and show the same tendencies to justify the conclusions drawn.

The basal diet, laboratory diet A-2, is the same as that used by Sherman at Columbia University (11) and almost identical with that used by Steenbock at the University of Wisconsin (12) in care-fully conducted quantitative work on vitamin A deficient diets. We have used marmite as a source of vitamin B as used by the Eng-lish investigators, instead of yeast as used by the Americans.

Diet A-2; corn starch 78%, extracted casein 12%, NaCl 1%, salt mixture 4%, marmite 5%.

Steenbock uses dextrinized starch and 2% agar in his basal diet instead of corn starch. The casein was extracted three times with alcohol under a reflux condenser, the boiling continued for one hour at each extraction and the casein filtered by suction while hot. This procedure has been shown by Sherman and Munsell (17) and

– 266 –

also by the results obtained in this laboratory, to be adequate for destruction of vitamin A. The salt mixture was Osborne and Mendel's salt mixture made up to approximate the composition of the ash of milk (13).

The alcoholic extraction of the casein probably removes the anti-rachtic vitamin as well as vitamin A, as shown by the rapid cessation of growth when no additional source of vitamin was added to the diet, and also by the subsequent considerable stimulus to growth, when aerated cod liver oil supplying vitamin D was added. Complete removal of vitamin A is evident from the behaviour of the control rats.

The oxidation of the cod liver oil was conducted as follows; in a boiling water bath a stream of warm air was bubbled through the oil for twenty hours (7). The incidence of ophthalmia when this oxidized cod liver oil was added to the basal diet with the autopsy findings, shows the oxidation and destruction of vitamin A must have been complete; and using this oil at a 3% level in the diet sup-plied sufficient vitamin D (without vitamin A) to stimulate growth. Using the same stock of animals the addition of sufficient butter, dried spinach, or lettuce to this basal diet plus 3% oxidized cod liver, we have obtained normal, or better than normal growth over a period of time longer than used in these experiments. Hence the basal diet is not only free from vitamins A and D, but when supple-mented with these substances is adequate for good growth, so is lacking only in the factor being investigated.

The experimental diet and distilled water were furnished ad libitum through out the experiments.

It is necessary that all other factors except the one being investigated be supplied and all influencing conditions controlled. Rats are now standard reagents, that is, we know their reactions as definitely as we know chemical reactions. The following additional precautions were taken. The temperature of the room was kept at between 13 and 19° C. since it is found that the weights of both normal and experimental rats will be lowered if the temperature of the room drops below about 13° or varies within wider limits. Rats on control or basal diets were kept in separate cages from those receiving additions to the diets because such rats will, we have found, consume sufficient excreta of the other rats to supply an appreciable amount of the missing factor. Also the cages have wire screen bottoms, and no bedding has been used since bedding may collect excreta and also bacteria, a possible source of additional vita-mins. The wire screen bottoms of the cages were changed and washed each day. No direct sunlight ever touched the cages or rats, though the rat room gets some early morning sun. The litera-ture on these matters furnishes experimental evidence of the value of these precautions (11, 12).

Finally, that death was caused by a lack of vitamin A or D or both was confirmed by careful daily observation of the experimental of vitamins A and D as the characteristic abscesses in the glands at the

– 267 –

base of the tongue, the frequency of infections of the sinuses and the middle ear and bronchial tubes, abnormal reproductive organs, abnor-mal intestines and beading of the ribs, or enlarged costochondral junc-tions. The incidence of ophthalmia, respiratory troubles, the condi-tion of the skin and hair, and the humped condition always occurring before death in the control animals were all noted, and left no doubt that the cause of death in the control animals was a lack of the fat soluble substances. Further more, we were able to distinguish between the effects attributable to each vitamin.

Ophthalmia does not evidence itself in all rats supposedly on vitamin A-free diets. Osborne and Mendel (13) found 60% incidence in a group of 493 rats, Stammers (14) had 88% occurrence, Wagner (15) 60 to 70%, Sherman and Munsell (17) 85%.

A week's supply of butter or spinach was weighed up at one time and fed by hand in three allotments on alternate dates.

Experiments on Butter as a Source of Vitamin A.

Rats used in the first series of experiments were of our new colony, and inferior in weight to those used later, but the growth curves show the same general tendencies and are included for comparison. By using a good stock diet we have since improved our colony animlas.

[The section below cannot be correctly rendered as it contains complex formatting. See the image of the page for a more accurate rendering.]

Table 1.
Additions of Butter to the Basal Diet.
Average weekly weights from 35 days after weaning.
Butter gms. daily No rats Cessation of growth in days Survival in days Weeks
5 6 7 8 9 10 11 12
0 4 75 63 67 67 70 68 67 61 54 D
.025 3 35 73 75 74 73 71 71 71 66 K
.05 7 35 61 63 67 68 68 69 72 74 K
.10 6 36 58 62 64 65 67 67 71 69 K
.20 3 37 59 63 64 71 74 73 73 73 K

Different amounts of butter added after active growth had ceased, i.e. after about 35 days on the basal diet.

The effect of additions of butter to the basal diet are shown in Table 1 and Chart 1. Except for those included in Chart 3 the weights and growth curves are not given for the preliminary period on the basal diet (from weaning at 28 days of age until 35 days later when active growth had ceased). The charts show the average weights of the rats on any one diet after the preliminary “running down” period and show the effects produced by the additions of butter.

Apparently with the first rats of somewhat inferior quality .025 gm. butter daily added after the preliminary period on the basal

– 268 –

diet was inadequate to supplement the basal diet, since the curve shows no growth, and but a slightly longer survival than the control rats. .05 gm. butter daily gave a slow but evident impetus to growth. The lack of the anti-rachitic vitamin was not supplied by the larger amounts of butter (.1 and .2 gm. daily), as shown by the very slightly better growth made on those amounts.

Picture icon

Chart 1.
Gains in weight after additions of Butter, added when growth had ceased, at about 35 days after weaning.

This experiment shows butter to the extent of .2 gm. daily to be inadequate in its supply of anti-rachitic vitamin to induce normal growth when added to a basal diet lacking both vitamins A and D. Drummond and Coward (9a) found .2 gm. butter daily, gave slow but steady growth to a non-growing rat. Sherman and Munsell (23)

– 269 –

found .25 gm. butter daily added to this same basal diet adequate to induce a gain of 25 gms. in weight in eight weeks. In the case of no single rat did we get any gain on this amount; they all lost from 2 to 13 gms. weight in seven weeks. Macroscopic examination and autopsy findings, show an absence of eye trouble and pus at the base of the tongue in all rats where butter was added, so the supply of vitamin A was adequate to prevent infections when added in as small amounts as .025 gm. per day.

The butter purchased by Sherman and Munsell on the New York City market, and that used by Drummond and Coward secured in London appear to have contained more vitamin D than did our supply. The cows in England and in New York State are kept in stalls and fed on dry food for a part of the year, whereas the cows in this vicinity are left out doors the entire year, and we expected more anti-rachitic factor to be present in the butter on our markets. This point will need to be investigated further by the use of butter of more than one brand secured at different seasons of the year and used with a better stock of animals.

The four control rats with no butter all developed ophthalmia, the average occurrence being at 54 days after weaning (82 days of age), all showed pus at the base of the tongue and very little or no fat in the abdominal cavity. Practically all of the twenty-two rats used in this experiment showed some evidence of haemorrhagic ribs and enlarged costochondral junctions, which were not found in rats supplied with oxidized cod liver oil (supplying vitamin D), although the condition of those receiving butter appeared superior to that of the controls.

[The section below cannot be correctly rendered as it contains complex formatting. See the image of the page for a more accurate rendering.]

Table 2.
Additions of Cod Liver Oil and Butter to the Basal Diet.
Weaning or days after weaning when cod liver oil and butter were added Gm. Butter daily Cessa tion of growth in days Survival in days No. rats Weeks Average weekly weights in gms.
3% c l o Butter 5 6 7 8 9 10 11 12 13
42 42 95 2 80 86 91 103 110 115 106 89 66 D
W 88 4 92 96 96 95 92 88 85 75 D
44 44 .025 44 92 4 93 99 109 121 126 129 121 109 88 D
W 46 .05 46 85 3 106 108 101 101 92 87 84 76 D
W 40 .10 40 5 89 94 97 96 104 105 110 116 117 K
W W 10 3 88 90 97 102 114 118 123 130 137 K

Additions of 3% oxidized cod liver oil and varying amounts of butter to the basal diet at weaning and after cessation of growth.

In the second series of experiments on butter as a source of vita-min A (Table 2, Chart 2) oxidized cod liver oil, 3% of the diet, was added to the basal diet at weaning, death occured in an average of 88 days. When oxidized cod liver oil was added at the time active growth ceased, the resulting stimulus to growth was pronounced, but

– 270 –
Picture icon

Chart 2.
1. Plus cod liver oil from weaning.
2. Plus cod liver oil after cessation of growth.
3. Plus cod liver oil and .025 gm. butter daily after cessation of growth.
Plus cod liver oil at weaning—
4. and .05 gm. butter daily after cessation growth.
5. and .10 gm. butter daily after cessation growth.
6. and .10 gm. butter daily after cessation weaning.

– 271 –

disater resulted in 95 days. When oxidized cod liver oil, 3% of the diet, and .025 gm. butter daily were both added when growth had ceased, death resulted in 92 days after weaning, and the curve is almost identical to that when cod liver oil, but no butter was added. The stimulus to growth was evidence of the presence of vitamin D in the oil. The stores of vitamin A were apparently exhausted and .025 gm. butter daily did not supplement for the deficiency, a support to the failure obtained in the first experiment when .25 gm. butter without cod liver oil gave no better result than the basal diet alone.

The second set of curves in Chart 2, shows the changes in weight after 35 days from weaning when 3% oxidized cod liver oil was added at weaning, i.e., a basal diet including the antirachitic factor, and varying amounts of butter added when active growth had ceased. Even with the addition of .1 gm. butter daily, the curve is considerably below a normal curve.

It is of interest to compare the effect of addition of .05 gm. butter daily after growth had ceased when no cod liver oil was added, with the result when cod liver oil had been added at weaning. The increased growth due to the active principle furnished in the cod liver oil, vitamin D, apparently means a greater requirement for vitamin A on the part of the more rapidly-growing organism. In Chart 1.05 gm. butter was added to diets of rats of an average weight of 61 gms. and in Chart 2 the same amount of butter was added for rats of an average of 106 gms. weight. This amount of butter was inadequate after the growth-stimulation caused by the vitamin D in the oil.

The use of .1 gm. butter daily from weaning gives a nearer approach to the normal curve bearing out the evidence of Osborne and Mendel (18) that in their experience smaller amounts of vitamin containing foods are required, when such foods are supplied from the start as a preventative measure than when given as a supplement after severe nutritive decline has set in. This gives experimental evidence of the superior value of preventative dietary measures over curative treatment.

Another comparison of these data is given in Chart 3, in which the weight curves from weaning of rats on the basal diet alone are compared with the impetus to growth given by additions after growth had practically ceased, on .025 gm. butter daily (curve 2), 3% oxidized cod liver oil (curve 3), and the same amounts of butter and oil added together (curve 4). The rats used when the cod liver oil was added (curves 3 and 4) were of later stock and superior to those used at first (curves 1 and 2); however, death resulted in about the same length of time. Evidently death was caused by a lack of vitamin A, and the addition of oxidized cod liver oil supply-ing factor D, and even of small amounts of butter, did not adequate-ly supply the missing essential substance, although better growth resulted for a time. The rats included in curves 3 and 4 from better stock, were of larger weights at weaning, and presumably would have better stores of both vitamin A and D, from the previous diet of the parents and what they ate of the stock diet before weaning.

– 272 –

This merely emphasizes the inevitable failure resulting from a lack of vitamin A.

Experiments on Spinach as a Source of Vitamin A.

The basal diet already described was used, including 2% oxi-dized cod liver oil to supply vitamin D. Varying amounts of dried spinach were furnished either at weaning as a protective measure or

Picture icon

Chart 3.
Growth from weaning, showing the effects of additions to the Basal Diet, of cod liver oil and butter after cessation of growth.
1. Basal Deit only.
2. Plus .025 gm. butter at A.
3. Plus cod liver oil at A.
4. Plus cod liver oil and .025 gm butter at A.

as a curative supplement after active growth had ceased. The spinach was dried in a large gas oven (gas stove) at a very low tem-perature with the door wide open, so there was no strong current of

– 273 –

air going through the oven. It was stored in an air-tight tin, taking care throughout that oxidation of vitamin A was avoided as much as possible.

[The section below cannot be correctly rendered as it contains complex formatting. See the image of the page for a more accurate rendering.]

Table 3.
Additions of Dried Spinach to the Basal Diet Plus Cod
Liver Oil.
Changes in weight when spinach added after cessation of growth.
Spinach gms. weekly No. rats Cessation of growth in days Survival in days Weeks Average weekly weights in gms.
5 6 7 8 9 10 11 12 13 14
0 4 88 92 96 96 95 92 88 85 75 68 D
.015 4 43 88 93 95 99 99 105 105 105 109 109
.025 3 43 91 91 98 103 105 108 111 113 119 120
.05 4 40 84 85 91 98 101 107 113 115 121 121
.10 3 37 85 85 95 105 107 112 116 120 125 128
.20 2 36 84 92 99 110 114 126 132 139 145 147

Additions of varying amounts of dried spinach to the Basal Diet supplemented by 2% oxidized cod liver oil. The spinach was added after cessation of active growth. The weights given are the averages after 35 days on the Basal Diet with cod liver oil.

In Table 3 and Chart 4 the weights are not given for the pre-liminary period from weaning to 35 days later. Active growth ceased in from 36 to 43 days on the basal diet supplemented with the oxidized cod liver oil. The weights are given from 35 days on, so that effects due to additions of spinach are evident. .015 gm. dried spinach per week or .0021 gm. daily, was sufficient to prevent any evidence of ophthalmia, and in no case was pus found under the tongue. Evidently adequate vitamin A, was furnished by this small amount of spinach to protect against infections, but not enough for much more than maintenance of weight. As other investigators have found in similar work, there is a wide middle ground between main-tenance and good growth.

Sherman and Munsell (17) found in feeding fresh spinach, that .017–.018 gm. fed daily (.119) to .126 gm. fresh per week) gave a gain of about 25 gms. in weight in eight weeks. We secured an equal gain on dried spinach with .025 gm. a week or the equivalent of .312 gm., fresh spinach per week. The experiments are not comparable because they were relying on the spinach to furnish vitamin D, which means that furnishing vitamin D, we expected to secure the same growth on smaller quantities of spinach. We have in all probability lost some of the original content of vitamin A by the drying process, but our spinach could not have had the original content of vitamin A, which was contained by that used by Sherman and Munsell, obtained on the New York City market.

McClendon and Shuck fed dried spinach as .1% of the food eaten, and secured a distinct retardation of ophthalmia. Rats of

– 274 –

this age and weight, eat about 45 gms. of food per week, which means .045 gm. dried spinach per week.

Further experiments are in progress using spinach from wean-ing instead of after a nutritive decline has set in; and while not suffi-ciently complete for results, the indications are that as with butter, a slightly better uniform growth is secured on corresponding amounts

Picture icon

Chart 4.
Changes in weight after 35 days on Basal Diet and Cod Liver Oil upon additions of dried spinach. Specified amounts fed weekly.

of spinach, and the weights of the rats at the end of 14 weeks from weaning, are equal to or better than those of this experiment.

Practical and Social Significance.

Both experiments with additions of butter and of spinach, show that very small amounts protect against infections, but very much

– 275 –

larger amounts are necessary before normal growth can be obtained. Furthermore, what we call normal growth is not necessarily optimum. There is a wide field where increasing quantities of vita-min A, induce increasingly more favourable results, but for optimum physical well-being much more than a minimum is necessary. It is recognized, that rats grow at a much faster rate than human beings; and as Mendel has said, we do not know the factor to enable us to transfer findings on animals, such as the rat to human beings; but we do know from the dietary experiment unwittingly conducted during the war, that much human experience can now be interpreted in the light of animal experimentation. We now know that in human beings it is quite possible that general physical debility, lack of stamina and vitality, and quite possibly size of stature could be greatly improved by an optimum instead of a minimum supply of essentials such as the vitamins have proved themselves to be.

Another practical aspect is evident in the better condition of rats whose dietary included the factor vitamin A from weaning over those who were allowed to cease active growth before being supplied the supplement of vitamin A. The general social and economic importance of prevention over cure is self evident.


This work was undertaken to ascertain the vitamin A content of butter and spinach obtainable locally. Before any quantitative work on food content of vitamin D can be carried out it has been necessary to determine vitamin A and vitamin D as separate entities, the results upon deprivation of each, and find a source of vitamin A which can be supplied with the basal diet without vitamin D. Knowing now the growth to be expected when vitamin D is furnished in adequate amount and vitamin A is supplied by .1 gm. of butter daily and .025 gm. spinach, we are starting on the vitamin D content of butter and spinach. Last year we did a few calcium determina-tions on the animals used, but this year, because of inadequate laboratory space and time, no such work has been possible. However we have found that our normal animals on our stock diet have the same amount of calcium in their bodies as do those of other labora-tories as given in the literature. By feeding low or free from vita-min D rations the calcium content will need to be determined also.

The thanks of the author are due to Miss E. N. Todhunter for techinal aid.


1. Mccollum, E. V. and Davis, M., 1913. Journal of Biological Chemistry, vol. 15, page 167.

2. Osborne, T. B., and Mendel, L. B., 1913. Journal of Biological Chemistry, vol. 15, page 311.

Osborne, T. B., and Mendel, L. B., 1913. Journal of Biological Chemistry, vol. 16, page 423.

4. Mellanby, E., 1919. The Lancet, vol. 1. page 407, and Journal of Phy-siology, vol. 52, page 53.

– 276 –

5. Powers, G. F., Park, E. A., and Simmonds, N., 1922. Proceedings of Society of Experimental Biology and Medicine, vol. 20, page 81.

6. McCollum, E. V., Simmonds, N., Becker, J. E., and Shipley, P. G., 1922. Journal of Biological Chemistry, vol. 53, page 293.

7. Steenbock, H., and Nelson, E. M., 1923. Journal of Biological Chemistry, vol. 56, page 355.

8. Steenbock, H. Nelson, E. M., and Black, A., 1924. Journal of Biological Chemistry, vol. 62 page 275.

9 (a.) Drummond, J. C., and Coward, K. H., 1920. Biochemical Journal, vol. 14, page 661 and 734.

(b.) Sherman, H. C., and Kramer, M. M., 1924. Journal of American Chemical Society, vol. 46, page 1055.

10. (a.) Sherman, H. C., and Storms, L. B., 1925. Journal of American Chemical Society, vol. 47, page 1653.

(b.) Drummond, J. C., 1919. Biochemical Journal, vol. 13, page 81 and 95.

11. Sherman, H. C., and Cammack, A. M., 1926. Journal of Biological Chem-istry, vol. 68, page 69.

12. Nelson, E. M., and Steenbock, H., 1925, Journal of Biological Chemistry, vol. 62, page 575.

13. Osborne, T. B., and Mendel, L. B., 1924. American Journal of Physiology, vol. 69, page 543.

14. Stammers, A. D., 1924. Biochemical Journal, vol. 18, page 11.

15. Wagner, E., 1926. Chemical Abstracts, vol. 17, page 2908.

16. Osborne, T. B., and Mendel, L. B., 1919. Journal of Biological Chemistry, vol. 37, page 572.

17. Sherman, H. C., and Munsell, H., 1925. Journal of Biological Chemistry, Society, vol. 67, page 1639.

18. Osborne, T. B., and Mendel, L. B., 1920. Journal of Biological Chemistry, vol. 41, page 459.

19 McClendon, J. F., and Shuck, C., 1923. Proceedings of Society of Experi-mental Biology and Medicine, vol. 20, page 288.