THE USE OF FERROUS GLUCONATE IN THETREATMENT OF HYPOCHROMIC ANEMIA

Paul Reznikoff, Walther F. Goebel

J Clin Invest. 1937;16(4):547-554. https://doi.org/10.1172/JCI100881.

Research Article

Find the latest version:

http://jci.me/100881/pdf

Pdf

THE USEOF FERROUSGLUCONATEIN THE TREATMENTOFHYPOCHROMICANEMIA

By PAULREZNIKOFFANDWALTHERF. GOEBEL(From the New York Hospital and the Department of Medicine, Cornell University Medical

College, and the Hospital of the Rockefeller Institute for Medical Research,New York City)

(Received for publication January 23, 1937)

Although hypochromic anemia due to iron defi-ciency is universally treated with iron medication,there is not as much unanimity with respect to theparticular type of iron compound to be used. Thefollowing postulates, however, are accepted: Thedosage must be adequate to insure a reasonablyrapid increase of hemoglobin, the iron compound,in the dosage given,'must be tolerated by the pa-tient without undue distress, and the cost of medi-cation must be within the financial means of thepatient.

Many individuals receiving iron complain ofsuch symptoms as nausea, epigastric discomfort,diarrhea or constipation. Some physicians hesi-tate to subject their patients to possible upset incases of gastric or duodenal ulcer, colitis, anddiarrhea or constipation. It was thought desir-able, therefore, to prepare an iron compoundwhich might have minimum irritating effects.One of the characteristics of ferric compounds istheir ability to precipitate proteins. Most ferrouscompounds are oxidized readily to the ferric state.This precipitating effect on proteins might explainthe irritating action of iron compounds on thegastro-intestinal tract of some patients. Ferrousgluconate, prepared anaerobically, was found tohave no precipitating action on proteins even whenconverted into the ferric state. It was decided,therefore, to treat patients with this compound todetermine not only its efficacy but also its toxicitycompared to medications now in common use.As a further test of its lack of toxic effects, solu-tions of ferrous gluconate were administered in-tramuscularly in quantities containing as much as50 mgm. of iron, with no systemic disturbancesand rarely with local discomfort. The method ofpreparing this substance and its use in treatinghypochromic anemia in rats have been describedpreviously (1).

Recently we have been able to simplify the methodbased upon a report of Neiger and Neuschul (2). These

investigators, studying the photochemical reactions offerrous gluconate, prepared dilute solutions of this saltby boiling aqueous gluconic acid with iron filings. Al-though these workers do not describe the isolation of thecrystalline salt, it occurred to us that crystalline ferrousgluconate might be prepared in quantity and in a highstate of purity by employing this simple reaction. Con-sequently 100 grams of crystalline calcium gluconate 1were dissolved in 700 cc. of boiling water. A solution of29.3 grams of oxalic acid dissolved in 150 cc. of warmwater was added. The precipitated 'calcium oxalate wasseparated by filtration, and the clear filtrate containinggluconic acid was concentrated to 350 cc. in vacuo. Thesolution of gluconic acid was placed in a one liter, threenecked, round bottomed flask bearing a mercury sealedstirrer. One outlet of the flask was fitted with a smallwater trap to permit the'escape of evolved hydrogen gaswhereas the third outlet was closed with a rubber stopper.The flask was now heated in a water bath, and the con-tents rapidly stirred. Twenty-six grams (2 equivalents)of pure powdered iron (Merck's " Iron by Hydrogen ")were added. A rapid evolution of hydrogen took place.At the end of two hours the solution in the flask wasneutral to litmus paper. The hot solution of ferrousgluconate, colored a pale green, was carefully filteredthrough a sintered 'glass funnel of fine porosity. Thefiltration was conducted in such a manner that at no timedid the solution of ferrous gluconate come in contact withatmospheric oxygen. This was accomplished by conduct-ing the solution from the reaction flask by suction into anenclosed filtering system in which all air had been dis-placed by carbon dioxide.

The solution was allowed to cool in an atmosphere ofcarbon dioxide and after crystallization of ferrous glu-conate was complete, the product was filtered rapidly ina Buchner funnel, washed with a small amount of 50 percent alcohol, and finally with pure acetone. The sub-stance was'placed in a vacuum desiccator to remove alltraces of acetone. Eighty-eight grams of ferrous glu-conate were recovered.

The product thus obtained is a fluffy whitepowder with a very slight greenish tint. Thesubstance crystallizes with one molecule of waterand contains no detectable ferric iron. The fer-

1 Wehave been able to simplify the preparation of fer-rous gluconate still further recently by making it fromtechnical gluconic acid.

547

PAUL REZNIKOFF AND WALTHERF. GOEBEL

.

4-..C0

kQco >1

ft

* 0

m...

0

U)

>.

~OD

4)04

I-.U)

0'

0

ol-

2.0ci)

to.*0

0

4)

3

.00

0

b..0

4)

0

u

0C:U)

98

.0

1.04)

44.

04-.

4)

>1m

0U)0.

4 A4-)

o D

0

04.b 4J

aU)

10.

8.0: ; g a

4) =b: &4 3.

.C

.A

0

-0

4-b -

80+

4J00-4U

0-8..U)

4J4.

19u

t49300 O 4 0 00 09.0.!: ~~0%A-. 0%'0C--g- qf)o o In) 14I

3 - c - -0 eq

4~~~~~- 1414sol: lo 11 s

wa0 %O'0% o000 0% o0 00 m 0 00 e

,d%4).9 -U)00 t0 ( 31)20

A M 1-- 00~~~~~~~~~~~~~0t.

0%i1 00 00

'0'-4 '0% ~ 0 I) 0

"'" .000 . .. 0

0~ 0

*4- I)'0__4_ ___ __Id.-0 Ul) 'IO ~~~~~~~~~~~~~~' i) 0 00 00*4 --

eq" ) -) tJ4 coo00e '00('-4 C4Oeq 0- q R000 C)4 NO

4-.0

0.

4-J~~~~4~~~~ o~~~~~~~c

q0 00J0 0~.~~ ~4)4-'4) .ed... 0; .

.

Cd~~ ~ .-.o0

'5

.444

C's

X

I)

548

*11

.X4i

I'

el)

.0i?I%.44

.',¢ u

aa

FERROUSGLUCONATEIN HYPOCHROMICANEMIA

it

¢V

549

0

N-

'0

Ca

o

oM._..E8

c*4)

PAUL REZNIKOFF AND WALTHERF. GOEBEL

rous gluconate prepared in this manner is in allrespects identical with that previously described.The above method, however, is considerably sim-pler than that originally described by us for thepreparation of crystalline ferrous gluconate (1),and has the additional advantage of being moreeconomical and more easily carried out.

Thirteen female patients ranging in age from24 to 49 years and averaging 40, and sufferingfrom hypochromic anemia were treated with fer-rous gluconate, two by intramuscular injection,eight by oral administration and three by bothmethods (Table I). In addition, two patientswho had demonstrated marked intolerance to otheriron compounds were given ferrous gluconate.In most instances, daily reticulocyte, red blood celland hemoglobin determinations were made untilnormal values were obtained, and subsequentlythe blood was studied as frequently as seemedindicated. In this study 14.5 grams of hemo-globin per 100 cc. was equivalent to 100 per cent.

The diagnosis in four of the patients was " idio-pathic " hypochromic anemia without complicatingfactors. In the others, the following conditionswere found, in some cases more than one beingpresent: intestinal adhesions, 1; post-thyroidec-tomy, 2; uterine fibroids, 3; previous miscarriages,1; rheumatic heart disease, 3; ulcerative colitisand ileostomy, 1; duodenal ulcer, 1.

The initial hemoglobin was less than 7.25 gramsper 100 cc., or 50 per cent, in 3 of the patients.In the remaining, the hemoglobin before medica-tion was greater than 7.25 grams. This observa-tion is important since Heath (3) states that a 1per cent rise in hemoglobin per day is the low limitof a satisfactory response to treatment when theinitial hemoglobin is below 50 per cent.

The volume index of one patient who had beenbleeding from a duodenal ulcer was 0.97. An-other had a volume index of 0.8; in two the deter-mination was not made; and in the other nine, thevalues varied from 0.58 to 0.76.

Gastric analysis was not performed in 3 cases.Of the rest, 1 had normal hydrochloric acid con-tent after alcohol and histamine; 4 had hypo-chlorhydria; and 5, achlorhydria.

The anemia in all of these patients respondedwell to therapy. In only two did the hemoglobinlevel fail to reach 11.6 grams or 80 per cent and

in these the toxicity of ferrous gluconate whengiven intramuscularly was tested and no attemptwas made to complete the treatment with this ironsalt. Six of the 13 patients attained hemoglobinvalues ranging from 13.1 grams (90 per cent) to14.2 grams (98 per cent); and 5, values rangingfrom 11.7 grams (81 per cent) to 12.6 (87 percent) .

Of the 13 patients, one had an initial erythro-cyte count between 2,000,000 and 2,500,000; one,between 2,500,000 and 3,000,000; four, between3,000,000 and 3,500,000; one, between 3,500,000and 4,000,000; two, between 4,000,000 and 4,500,-000; and four, between 4,500,000 and 5,000,000.After treatment, three patients had erythrocytecounts between 4,000,000 and 4,500,000; five, be-tween 4,500,000 and 5,000,000; and five, above5,000,000. Since this response of the red bloodcells to treatment showed no abnormalities, thisphase of the subject will not be considered furtherin this report.

Symptomatically, 7 patients were apparentlycured with the attainment of a normal bloodcount. In 4, profuse menstrual bleeding per-sisted even after a normal blood count wasreached. In 2 of these cases hysterectomy wasperformed (Cases 4 and 9); in the other twoCases 1 and 7), the blood count has been normalfor five months in spite of the fact that they havecontinued to menstruate profusely and have re-ceived no medication. One patient (Case 5) whohad an ileostomy for ulcerative colitis, becameanemic again two months after the administrationof ferrous gluconate was stopped although shereceived large doses of ferrous sulphate and intra-muscular liver extract. Another patient (Case10) returned three months after the cessation offerrous gluconate therapy with evidences of hy-perchromic anemia and a history of severe mal-nutrition. She is responding well to liver ther-apy.

The effect of ferrous gluconate therapy is sum-marized in Table I. An analysis of these results,without a critical consideration of each case, showsthat with respect to the reticulocyte count, gain inhemoglobin per day, and percentage utilization ofiron, the ferrous gluconate was strikingly effec-tive in the 13 patients treated. Since in this studysmall oral doses of iron (108 mgm. daily) were

550

FERROUSGLUCONATEIN HYPOCHROMICANEMIA

FIG. 1. RETICULOCYTEPEAKS ATrrMND BY PATIENTS PLOTTED ON HEATH'S (3) CURVEOF ADEQUATERESPONSE

usually administered to determine the response tominimum dosage, the time elapsing before a nor-mal blood count was obtained was not remarkable.However, when for purposes of expediency largeroral doses (216 mgm. daily) were given to Pa-tients 2 and 13, normal blood counts were attainedwithin three weeks. Patients 1, 6 and 8 showeda normal count with only 108 mgm. of iron dailyin 20, 16 and 23 days respectively. Case 6, how-ever, represents a posthemorrhagic patient, andthe marked response may not be due chiefly to thetherapy.

Figure 1 shows the reticulocyte peaks attainedby the patients in this study. These are repre-sented by a dot for each case when compared toHeath's curve (3) for adequate reticulocyte re-sponse. Patient 5 suffering from ulcerative coli-tis attained 22 per cent reticulocytes, but her ini-tial count was 10 per cent. Patient 6 who suf-fered from hematemesis reached a peak of 15.2per cent from an initial count of 12.6 per cent.

Obviously, these results cannot be considered tobe entirely due to the treatment. Patient 12 at-tained a reticulocyte peak of only 4.6 per cent withan initial hemoglobin of 7.9 grams. But prelim-inary counts were as high as 2.4 per cent, and shehad a red blood cell count of 4,800,000 beforetherapy was started which would tend to lowerthe peak.

The average daily increase of hemoglobin inthe patients receiving ferrous gluconate intra-muscularly, exclusive of Patient 5 in whom theresults of intramuscular and oral administrationoverlapped, was 0.189 gram of 1.30 per cent. IfPatient 3 is excluded because of her high initialcount of 10.2 grams, the results with the remain-ing three patients, two of whomhad initial hemo-globin values of less than 7.25 grams, give anaverage daily increase of hemoglobin for intra-muscular therapy of 0.215 gram or 1.48 per cent.The average daily gain for all patients given fer-rous gluconate orally was 0.181 gram or 1.25 per

551

PAUL REZNIKOFF AND WALTHERF. GOEBEL

cent. It is obviously unfair to include Patient 6in this'series as the rapid rise in her hemoglobinfollowing the acute hemorrhage is certainly notdue to medication entirely. Patient 2 whose ini-tial hemoglobin before oral therapy was startedwas 10.2 grams and Patient 10 who subsequentlydemonstrated a maturative deficiency probablyshould not be considered in this calculation. Ifthe last two patients are included, the averagedaily gain of hemoglobin for the group is 0.173gram or 1.19 per cent; if these two cases are ex-cluded, the average daily gain of hemoglobin is0.191 gram or 1.32 per cent. Patient 7, the onlyone in the series whose initial hemoglobin wasless than 7.25 grams before oral therapy wasstarted, had a daily hemoglobin gain of 0.225gram or 1.55 per cent.2

The percentage utilization of iron was calcu-lated by multiplying the total gain in grams ofhemoglobin per cubic centimeter for each patientby 5000, the approximate adult blood volume, andby 0.0033, the approximate percentage of iron inhemoglobin, and the result obtained was dividedby the total amount of iron given the patient.The percentage utilization in the 4 patients whoreceived intramuscular ferrous gluconate calcu-lated in this manner was well over 100 per cent,actually averaging 148 per cent. This is in keep-ing with the findings of Heath, Strauss and Castle(4) and of Whipple and Robscheit-Robbins (5).Whether this is due to the erroneous assumptionof blood volume as the former authors suggest orto a salt effect on iron stored in the body as thelatter workers state cannot be determined by ourexperiments. The percentage utilization of ironin the patients receiving ferrous gluconate bymouth averaged 27.2. If the three patients ex-cluded from the final calculations of the dailyhemoglobin increase are likewise omitted fromthis determination the utilization for oral admin-istration of ferrous gluconate in doses rangingfrom 0.108 to 0.216 mgm. of iron daily is 28.5per cent.

The problem of toxicity remains to be consid-

2 After these cases were compiled, a patient was stud-ied who had an initial hemoglobin of 5.7 grams or 39 percent and was given daily doses of ferrous gluconate con-

taining 324 mgm. Fe. In 11 days her hemoglobin rose to11.6 grams or 80 per cent, a daily gain of 0.54 gram or

3.7 per cent.

ered. It is difficult to obtain objective evidence ofintolerance to iron therapy. The following case,for example, which Dr. William Murphy of Bos-ton kindly permits us to cite, illustrates an expe-rience which is occasionally encountered. Thepatient had suffered from gastric and intestinaldisturbances characterized by nausea, constipationand resulting hemorrhoids following the admin-istration of various forms of iron. She also suf-fered from rectal irritation associated with noc-turia, frequency and burning on urination. Aftertaking 0.3 gram of ferrous gluconate three timesa day, these symptoms were all produced but to aless degree than with other forms of iron whichshe tried. Dr. Murphy writes, " I am sure thatshe could take short courses with this form ofiron with less difficulty than any of'those which Ihave previously tried,-and I think it would bewell for her to have these capsules for periodic useif it is possible to obtain them."

While this case is not very striking, a secondpatient offered a better means of studying the rela-tive toxicity of ferrous gluconate. The patient inquestion was recovering from a Caesarian deliveryand her obstetrician hesitated to give her iron fora slight anemia because she had suffered fromgastro-intestinal distress and urticaria when shehad received iron previously. At one time an in-jection of some iron compound had produced in-tense urticaria. When seen 17 days after heroperation she had a red blood cell count of 3,600,-000 and a hemoglobin of 11.7 grams or 80 percent. She was not treated at the time but one andone-half months after discharge from the hospitalshe was given ferrous gluconate in increasingdoses until she received 0.9 gram daily containing108 mgm. of iron with no ill effects. She wasthen given 35 mgm. of iron in the form of ferroussulphate and within a few hours suffered a violentgastro-intestinal upset. To determine the effectof the intramuscular injection of ferrous glu-conate, she was given 0.45 cc. of a solution con-taining 25 mgm. of iron per cc. in her left glutealmuscle. For a few hours she had some swellingand soreness but no general reaction and the nextmorning the local region was practically normal.

At least two patients in this series who sufferedno ill effects when taking ferrous gluconate inquantities containing 216 mgm. of iron per day,

552

FERROUSGLUCONATEIN HYPOCHROMICANEMIA

complained of indigestion characterized by ano-rexia, gas and epigastric distress or of constipa-tion when receiving ferrous sulphate in quantitiescontaining from 180 to 249 mgm. of iron per day.

DISCUSSION

To evaluate the efficacy of ferrous gluconate, acomparison with results obtained with various ironcompounds by other workers is necessary. How-ever, it must be remembered that the experimentalconditions may not be the same in any two studies.For example, Reimann and Fritsch (6) using fer-rous chloride in doses which contained 100 mgm.of iron a day demonstrated remarkable utilizationof the iron, ranging from 17 to 45 per cent.However, all their patients had other forms of ironuntil a few days before the ferrous chloride wasstarted and some demonstrated excellent reticulo-cyte responses with these other iron compoundswhich were supposed to give no appreciable hemo-globin increase. It is obvious that the action ofthe alleged inactive iron salts cannot be disre-garded in the final computations. In fact, theseauthors concluded that all ferrous compoundshave approximately the same effect. It is alsocontrary to the experience of all other workersthat large doses of ferric salts are ineffective inhypochromic anemia as Reimann and Fritsch as-sert. Schulten (7) found that ferrous chloridehad to be given in much larger doses and saw nodistinct advantage in this iron compound. David-son (8) reported excellent results with ferrouschloride in doses containing 122 mgm. of iron aday but only 2 of his 7 patients attained 80 percent hemoglobin. Witts (9) gives as the mini-mumeffective daily dose of ferrous carbonate, anamount containing 300 mgm. of iron. Probablythe best comparison of the efficacy of various ironcompounds has been made by Fullerton (10).Table II represents a summary of his results com-pared with those obtained by oral administrationof ferrous gluconate. Only those cases are in-cluded in which it seems reasonably certain thatthere are no factors which either interfere withor accentuate the iron effect. This control neces-sarily makes the available cases few. It is alsoimportant to compare separately the results inpatients whose initial hemoglobin values were be-low and above 50 per cent. In our series, pa-

TABLE II

Relative efficacy of various iron compounds *

Aver- Average Uti-Compound Daily ironInitial Nu-age tim

liza-Compound Daily iron I lemo- her daily before tionused dosage globin cases g hloobinglob ofrise rise io

grams per per" days percent cent cent

Ferrous sulfate .... 0.180 <50 12 1.175 30 15.70Ferrous sulfate .... 0.120 >50 3 0.650 33 13.00

Iron ammoniumcitrate ........ 1.215 <50 30 1.270 2.50

Iron ammoniumcitrate ........ 1.215 >50 3 1.030 2.03

Ferrous carbonate 0.110 <50 6 0.955 20.80Ferrous carbonate 0.110 >50 3 0.520 11.30Ferrous carbonate 0.220 <50 8 0.803 8.80Ferrous carbonate 0.220 >50 1 0.180 1.96Ferrous carbonate 0.330 <50 10 1.125 8.18Ferrous carbonate 0.330 >50 3 0.940 6.84

Ferrous chloride. . 0.132-0.198 <50 4 1.420 20.70Ferrous chloride. . 0.132-0.198 >50 3 1.000 14.50

Ferrous gluconate 0.108 <50 1 1.550 38 37.50Ferrous gluconate 0.108 >50 5 1.230 29 29.00Ferrous gluconate 0.216 >50 2 1.240 18 17.25

* 14.5 grams = 100 per cent hemoglobin.

tients whose initi-a hemoglobin readings werebelow 7.25 grams are rare; in Fullerton's studyinitial hemoglobin values below 50 per cent wereusual. In comparing the effect of the variousiron salts it is important to note that all the pa-tients receiving ferrous gluconate orally attainedhemoglobin values greater than 11.6 grams or 80per cent, while in Fullerton's series 9 of the 15patients treated with ferrous sulphate and 2 of the3 taking ferrous carbonate for whomdata is givenfailed to reach such a level.

Since Barkan's (11) and Meulengracht's (12)reports, most clinicians feel that large doses ofiron are 'essential in treating hypochromic anemia(13, 14, 15, 3, 16). Whipple and Robscheit-Robbins (17), working with standard anemicdogs, emphasize the fact that the particular typeof iron is unimportant as long as it is given inlarge doses. However, Furth and Scholl (18)found that ferrous salts are much more easily ab-sorbed from intestinal loops of rabbits than ferriccompounds and most workers who advocate largedoses of iron admit that ferrous salts are moreefficiently utilized in patients (14, 3). Bethell,Goldhamer, Isaacs and Sturgis (19) feel the sameabout soluble iron salts. The results of moststudies show, however, that large doses of ironmay not produce a normal blood count for a con-siderable time (15) and occasionally rather small

553

PAUL REZNIKOFF AND WALTHERF. GOEBEL

doses of iron will cause surprising improvement(20).

The real problem of iron therapy is not the the-oretical utilization of iron, or the reticulocyte re-sponse, or even the daily increase of hemoglobinfor any particular period of treatment. Theseare important only as they indicate the return ofthe patients' blood to normal in a reasonably shorttime without undue inconvenience. Most patientssuffering from hypochromic anemia respond wellto most forms of iron when administered in ade-quate dosage. For the patients who cannot tol-erate the usual iron compounds, it is important tohave a medication which is effective and whichcauses minimum disturbance. For all patients inneed of iron it is desirable to use a compoundwhich gives good results with the least discomfort.Ferrous gluconate seems to be such a medicament.

CONCLUSIONS

1. Ferrous gluconate prepared in the absence ofoxygen has been used in the treatment of 13 pa-tients suffering from hypochromic anemia.

2. The use of ferrous gluconate compared withother iron preparations results in satisfactoryreticulocyte responses, a high percentage utiliza-tion of iron, and such daily increase in hemoglobinthat a normal level occurs in a reasonably shorttime.

3. Four patients, who showed toxic reactions toother iron compounds, were able to take ferrousgluconate without any undue distress.

4. In the patients who received ferrous glu-conate intramuscularly up to the present no sys-temic and only rare and mild local reactions oc-curred. However, in view of the efficacy of theoral administration of ferrous gluconate and itslack of toxicity there is seldom any reason for itsparenteral administration.

BIBLIOGRAPHY1. Reznikoff, P., and Goebel, W. F., The preparation of

ferrous gluconate and its use in the treatment ofhypochromic anemia in rats. J. Pharmacol. andExper. Therap., 1937, 59, 182.

2. Neiger, R., and Neuschul, P., Uber photochemischeReaktionen der Eisengluconate. Ztschr. f. phys-ikal. Chem., 1936, 177, Alt A., 355.

3. Heath, C. W., Oral administration of iron in hypo-chromic anemia. Arch. Int. Med., 1933, 51, 459.

4. Heath, C. W., Strauss, M. B., and Castle, W. B.,Quantitative aspects of iron deficiency in hypo-chromic anemia (The parenteral administration ofiron). J. Clin. Invest., 1932, 11, 1293.

5. Whipple, G. H., and Robscheit-Robbins, F. S., Bloodregeneration in severe anemia. XVI. Optimumiron therapy and salt effect. Am. J. Physiol.,1930, 92, 362.

6. Reimann, F., and Fritsch, F., Vergleichende Unter-suchungen zur therapeutischen Wirksamkeit derEisenverbindungen bei den sekundaren Anamien.Ztschr. f. klin. Med., 1930, 115, 13.

7. Schulten, H., Zur Klinik der essentiellen hypochromenAnamie. Munch. med. Wchnschr., 1935, 82, 697.

8. Davidson, L. S. P., Discussion on the treatment ofthe anemias before Section of Therapeutics andSection of Medicine. Proc. Roy. Soc. Med., 1933,26, 616.

9. Witts, L. J., Discussion on the treatment of theanaemias. Proc. Roy. Soc. Med., 1933, 26, 607.

10. Fullerton, H. W., The treatment of hypochromicanaemia with soluble ferrous salts. Edinburgh M.J., 1934, 41, 99.

11. Barkan, G., Therapie der Anamien mit grossen Eisen-gaben. Klin. Wchnschr., 1923, 2, 1748.

12. Meulengracht, E., Large doses of iron in the differentkinds of anemia in a medical department. Actamed. Scandinav., 1923, 58, 594.

13. Keefer, C. S., Huang, K. K., and Yang, C. S., Liverextract, liver ash and iron in the treatment ofanemia. J. Clin. Invest., 1931, 9, 533.

14. Minot, G. R., Idiopathic hypochromic anemia. Eman-uel Libman Anniversary Volume, InternationalPress, New York, 1932, 2, 831.

15. Wintrobe, M. M., and Beebe, R. T., Idiopathic hypo-chromic anemia. Medicine, 1933, 12, 187.

16. Bethell, F. H., The application of diagnostic criteriato the treatment of the anemias. New York StateJ. Med., 1935, 35, 799.

17. Whipple, G. H., and Robscheit-Robbins, F. S., I. Ironand its utilization in experimental anemia. Am. J.M. Sc., 1936, 191, 11.

18. Furth, O., and Scholl, R., The absorption of ferrousand ferric compounds from the intestines of rab-bits. J. Pharmacol. and Exper. Therap., 1936, 58,14.

19. Bethell, F. H., Goldhamer, S. M., Isaacs, R., andSturgis, C. C., The diagnosis and treatment of theiron-deficiency anemias. J. A. M. A., 1934, 103,797.

20. Vaughan, J. M., The haematopoietic response to ther-apy in non-tropical anaemias. Tr. Roy. Soc. Trop.Med. and Hyg., 1934, 27, 533.

554