METABOLICEFFECTSOF MARKEDSODIUMRESTRICTION INHYPERTENSIVEPATIENTS. SKIN ELECTROLYTELOSSES1

BY LEWIS K. DAHL, BERNARDG. STALL, III, AND GEORGEC. COTZIAS

(From the Medical Department, Brookhaven Nationl Laboratory, Upton, L. I., N. Y.)

(Submitted for publication August 15, 1954; accepted November 24, 1954)

The increasing use of prolonged Na restrictionin the treatment of cardiovascular diseases in par-ticular, makes study of long-term Na balancepertinent. With diets which contain 2 to 7 mEq.Na per day, combined urine and stool losses aver-age 1 to 5 mEq. Na per day (1, 2). Therefore,over an extended period the loss of electrolytethrough the skin might assume importance. Onlytwo earlier studies were found relative to elec-trolyte losses of the entire skin, both after short-term Na restriction, namely 3 (3) and 14 (4)days.

The data from Benedict's classical study (5) ofan individual who fasted for one month haverelevance to any work on the physiological re-sponse to dietary restriction, but the metaboliceffects of starvation are not directly applicable tothose effected by limitation of only two dietarycomponents-namely sodium and chloride.

For almost seven years the authors of thepresent paper have been using a diet which con-tains about 6 mEq. Na per day in the study andtreatment of patients with essential hypertension.On such a regimen combined urine and stool lossesaverage 2 to 5 mEq. Na per day, and since noevidence of hyponatremia had been found in theabsence of serious renal disease, it was reasonableto suppose that the patients were in Na balance.The current study was undertaken to quantitatethe actual skin losses of several electrolytes onseven subjects who had been on the low Na regi-men for one to five months, during two collectionperiods of 7 and 3 days. Although interest wascentered primarily on Na, Cl and K were meas-ured as well.

MATERIAL

The seven white adult subjects had been admitted tothe metabolic wards of the Brookhaven National Labora-tory Hospital for studies relative to the hypertension

1 This research was sponsored by the Atomic EnergyCommission.

research program. The pertinent clinical and labora-tory data on these patients are summarized in Table I.It will be seen that five of the patients, all females, hadclassical uncomplicated benign hypertension. In two(Di, Jo), the blood pressure had attained near-normallevels by the end of the six-weec control period and be-fore Na restriction had begun. The male with malig-nant hypertension had had a bilateral lumbar sympa-thectomy in 1952 without significant effect on the courseof his disease. This patient on admission had severeheadaches, intermittent gallop rhythm, and impairmet ofrenal function. The seventh subject, a female of forty-five, was a normal volunteer who underwent preciselythe regimen of these hypertensive patients for a fifteen-week period. At the beginning of the first experiment,patient Bl had been on the low Na regimen for 150 days,patient Os for 40, and the remaining five for 26 days.

PROCEDURE

Diet; UrinesAll patients were on a diet accurately calculated and

weighed daily with an intake of approximately 6 mEq.Na per day2 (5.7 + 1.4 mEq.) by analysis of 18 dailydiets.8 Restriction of Na to this amount allowed 45 to55 grams of protein per day in the food; the intake wastherefore supplemented with the low Na protein foodsLesofac®' or Lonalac® in amounts sufficient to give atotal of 1.5 grams protein per Kg. body weight per day.K and Cl were not included in the daily calculations andtherefore their amounts were primarily dependent uponthe limitations imposed by the low Na. Analysis of the

2During the period of control, 10 gm. of entericcoated NaCl had been taken by each patient daily in ad-dition to this basic diet.

8These analyses were carried out at various timesduring the year preceding this experiment Recently,after the completion of this study, analyses were done onsix seven-day collections which were made during theperiod of June 6 to 28. The average value was 4.6 mEq.Na per day (range 4.3 -4.7). The electrolyte contentof vegetables will vary with the content in the soil inwhich grown (6) and the source of supply will of coursevary with the season. Because the previous analyses weredone over a longer period, it is believed that they are morerepresentative of the average intake, and in this study thevalue of 6 mEq. per day has been assumed.

4Obtained through the courtesy of Wyeth & Co., Phila-delphia.

462

SKIN ELECTROLYTELOSSES AFTER NACL RESTRICTION

+ Jmp

+u)0NeT C#) ..

10 _doo r KV-4

C4 x k- U).1 U)w) Ul) WU)W -

Hoo-U)9-

H

al =3DI,OF OO

W)

00gU)

-c4

V-~4

c)S..

L

so= FWc ot

co0 0SSlO °

U)z Z

o3t.- Cd

000t)Z

o & %4.b 0 8dZ 8 vw w @Z*:ES °b&'.I

cd

+++0 CCU-)

++W."

-

-

++ coo- F.

0 0

4- z0

++O

+:+ z '

4)(C4 10

o 0

+J a

++00 OE~

0000~~~00eE

z

0000~~00v-z

0

4L)

463

.9%a

.2is.

P.f

N4

0..44-

4-aW04-)

0.

4)4-b

(4

LEWIS K. DAHL, BERNARDG. STALL, III, AND GEORGEC. COTZIAS

diets yielded an average of 70 mEq. K (range 30 to 100)and 12 mEq. Cl (range 9 to 15) per day. Thus the dietin effect provided a low Na, low Cl, normal K intake.All urines were saved, pooled for measurement of 24-hour volume, 200 ml. aliquots removed, and then pre-served with 1 ml. of concentrated H2SO,.

Experimental procedure

Preparation of patients and clothing. On the first dayof both experiments each patient had a shower using anion-free detergent, Brij 35,5 and wash rag to scrub theentire skin and hair. This was followed by repeatedrinses with distilled water after which the patient wasdried with a towel previously washed in the abovedetergent and tap water. Closely fitting one-piece cot-ton underwear was then put on which extended to thewrists and ankles. Both underwear and wash rags hadbeen extracted as follows: an initial wash in an automaticwasher with tap water and Brij 35, rinsed thoroughly in0.1 NH,SO, three rinsings of tap water, left overnightin acidified (with H,SO4) distilled water, then rinsed atleast three times in distilled water and until the washfluid had less than 0.04 mEq. Na per liter as tested bythe direct technique on! a Baird Model DB2 flame photom-eter. No K was found by the same method of testing.Over the underwear, loosely fitting cotton surgical pantsand shirts tied at the wrists, ankles and neck were worn.Short white cotton sox and surgical turbans completedthe outfit except for their individual foot gear. Thepants, gown, sox and head coverings had been washedwith Brij 35 in tap water but had not been chemicallyextracted. The shoes and slippers had not been treatedin any fashion.

Experimental periodsCollections were made during two periods of seven

and three days, the end of the first and start of thesecond being separated by eleven days. Until completionof each of the experimental periods, none of the garmentswere removed except the shoes at night. The patientswere instructed as follows: 1. The hands and face couldbe washed as necessary with tap water and a 1: 500 solu-tion of Brij 35; 2. After each bowel movement, and uri-nation in the females, the perineal urea was to be washedwith Brij 35 and rinsed with distilled water; 3. Nobeauty aids, deodorants, or other toilet preparations couldbe applied; 4. Normal activities could be maintained in-cluding outdoor strolls and occupational therapy projectsproviding no frank sodium contamination was involved(e.g., clay modeling or washing clothes with soap orhigh-sodium detergents was not allowed); 5. Vigorousexercise was to be avoided. The patients were highlycooperative and individually interested in these studiesand it is believed maximal attempts were made to liveup to these stipulations.

Ward temperatures were maintained at 72 to 740 F.during the day and from 65 to 68° at night. After the

5Donated by the Atlas Powder Co., kindness of S. J.Dumovich.

studies were completed, it was found that the averagemaximal daily temperatures outside averaged 50.70 F.and 62.60 F. during the seven and three-day studies,respectively.

Collection of samples(a) Bath water. At the end of each experimental pe-

riod, the perineal area was washed as described previouslyand washings discarded. Thereafter the washing pro-cedure was carried out under direct supervision by one ofthe authors or a thoroughly trained nurse. The outergarments including foot gear, sox, turbans, shirts andpants were removed and not used in the subsequentanalyses. Before removing the underwear, the face,neck, hands and feet were washed with Brij solution,rinsed with distilled water, and the washings discarded;it had been decided previously that contamination ofthese areas during the studies could not be preventedand therefore they were eliminated from the calculations.The underwear was then removed and placed in a cleanwax bag. For collection of bath washings, a chemicallyclean stainless steel tub slightly over two feet in alldiameters was provided, with a special drain from whichall wash fluid went directly into collecting bottles. Usingsponge bath technique, each patient washed the entirebody including scalp with the Brij solution and a washrag previously extracted in the manner used for under-wear. Three complete rinses with distilled water fol-lowed after which the patient methodically wiped theinside of the tub with the wash rag which was thenplaced in the bag with the underwear. The empty tubwas thoroughly rinsed with 0.1 N H2SO, followed bydistilled water, and these rinses added to the bath water.After mixing, the total volume was measured and forthe group averaged 4.7 liters. Two hundred ml. aliquotswere preserved with 1 ml. of concentrated H2SO4.

(b) Extraction of underwear and wash rags. Theseitems were treated individually by an initial extractionwith Brij solution, then three extractions with 0.1 NH2SO at which time the washings uniformly had notmore than 0.04 mEq. Na per liter and no K, tested inthe same fashion as before. All of these extracting solu-tions for each patient were pooled, the volume measured,and analyzed separately from the bath water so that ifwidely disparate results in the analysis of bath and ex-tracting solutions had occurred, evidence of contaminationwould have been suggested. The volumes of the extrac-tion fluids averaged 5.6 liters for the group. Aliquotswere removed and preserved in the usual fashion.

MethodsSodium and potassium. Na and K analyses were made

with a Baird Model DB2 flame photometer using an in-ternal lithium standard. Recovery experiments re-peatedly gave maximal errors of 1 per cent. Depend-ing upon the concentration of Na or K in the samples, 1to 40 ml. of solution diluted to 100 ml. were analyzed induplicate, with recalibration of the instrument againststandard solutions at least every second sample.

464

465SKIN ELECTROLYTELOSSES AFTER NACL RESTRICTION

NOsr- mo Ul V- 00\0% 0A0%Mt̂b_f"o V>_aNO0( 0~ _b0co\o %U)

**om .. ... .*** ****6 %6*i C ;C o(~ ;14 . ;In° to NOC4 c ° eo in N-OC°9

P~~~~~~~~~~~~~~~~~~~~-tx~~~~~~~~~~~.40 0 0 C 4C

eq* -wno%qC)no" MN- 4kOeq

. ~ cokO00\ - % e4cN0s e0 0t o 0 4k )mrIvzvNO>n v0 NO0> lqtI-4 Z

in WI W)uz

e~~~00t-. in W- V- No

int ocr rb-e - ;o

reXm o_0 Oiom t ooc-

%0$ oo Oe_O 00C7_oo__cC l) X

_oZ

X. o o o,o v) 00 N _ en~~~~~~~~ NO C4

C4 Mt- O\oot- 0 tool

>r.l * o 4 > ov c^ S e u~~~~~~~~~) 00 m _t-

w_~~~~~~~~~~~~t int (7 0~00 gzv>o ov 0% oo t- NO oD in

bo>\0\0 ~~~~~~Sttoaoe -eV)o0o~~~~~~o NNO 00

+

XSt . ~~~~~~~~~00en_4 t- in In In) )m*t r~~~~-ooo . i tOZ-.+o> W.) M ooo4noe;m0

t~~~~~~~~~~0*i+on W) 0enXmcoC0 -0%k$~~~~~~~~% 00 W) (4 No-M NW-

0 4 U)C

X~~~~~~~~-* -e OIn% 0O W)O-140 t

C1 00 i 0Ib_d40r 00 co NO N0=

rin 0000_ t-e 00uto\ov-- arEeo0 ob - 0 _ oo

o~~~~~~~%*4to oCSt- o oN_ 0@t~~~~~~~~~~~~~~~CMt-o_ooo ONNonOo^ol-ovO~~°+N~~>°>zso ° Ic c°cc

_~ C; CliciC -zC 4. l- ;1 1 l 6

C>coo ;

0 NOC4 co co C4 C4 NO(4 C4-4X-

c:; um>FE. co tft *5 -H40' ;C; C 6C40

V-4eq1 V-4 " V-4 V-4 V-1 V-4 " V-4 V-4 " " " C4 C4 " C4 C4 " " (N m

LEWIS K. DAHL, BERNARDG. STALL, III, AND GEORGEC. COTZIAS

Chloride. The Van Slyke and Hiller modification (7)of Sendroy's method (8) with the nomogram of VanSlyke (9) for solubility corrections of Ag IO, was used;because of the low pH of the solutions (1.5 to 2.0) fromthe HISO4 preservative, upon the advice of Dr. D. D.Van Slyke, no H,PO, was added. Furthermore, becauseof the low chloride in most solutions, 25 ml. aliquotswere precipitated. All analyses were done in a constanttemperature room main-tained at 20° C. Experimentswith standard added to these solutions gave 98.3 to 100.5per cent recovery (99.4 per cent average) but as a furthercheck on reagents and method, standard solutions wereanalyzed at the beginning and end of each day's titrations.

OBSERVATIONS

The primary data on skin and urine losses ofNa, K, and Cl are shown in Table II.

Sodium-SkinIt may be seen that despite variation within the

group, daily loss of Na from the skin in all indi-viduals was low, ranging from 0.09 to 2.59 mEq.(line 12) but with the majority about 0.5 to 1.5mEq. per day. The highest daily losses in bothseven and three-day experiments (1.41 and 2.59mEq., respectively) were in the smallest individual(Ke) and the lowest values (0.09 and 0.38 mEq.,respectively) in the one patient (Os) with malig-nant hypertension. Skin losses in the normal con-trol patient (De) were roughly half way (0.28and 0.99 mEq. per day) between the lowest valuesand the mean (0.80 and 1.70 mEq. per day) ofthe five patients with essential hypertension, aswell as being less than losses of any individualamong those five. That significant contaminationdid not occur in these experiments is suggested bythe generally low total values for Na as well as bythe lack of striking disparity between Na valuesfor bath and underwear (lines 9 and 10).

Sodium-Urine

Urinary excretion of Na is summarized in lines13 and 14. With the exception of patient Os withmalignant hypertension, the values found werewithin the range observed in hypertensive patientsduring the past six years on this same regimen,namely about 1 to 4 mEq. per day. The value of7 mEq. per day shown by patient Di during thethree-day experiment lies within the range of day-to-day fluctuation in the urinary excretion of Nain patients maintained on a constant low Na in-

take. The normal control did not differ from thefive patients with essential hypertension in herability to conserve Na on a restricted intake. Pa-tient Os, with serious renal disease, excreted anaverage of 10 and 22 rnEq. per day in these twoexperiments. It may be recalled that this pa-tient was shown to lose phenomenally smallamounts of Na from the skin-considerably lessthan the six subjects with good renal function.By contrast, patient Ke, who had the largest skinloss of Na, had the smallest urinary loss (1.07 and0.77 mEq. per day).

Potassium-Skin

In general, the findings paralleled closely thosedescribed above for Na. Potassium loss was onlyslightly greater than that of Na with a range of0.08 to 2.69 mEq. per day (line 20) in the group.The five patients with essential hypertension av-eraged 1.1 and 1.9 mEq. per day, the normal con-trol 0.42 and 0.79 mEq. per day, and patient Os0.08 and 0.23 mEq. per day with the seven andthree-day experiments, respectively. Thus, rankedby absolute daily losses, the seven members weredistributed in the same order found for Na loss inskin. And again, patient Ke had the greatest lossin both experiments, as patient Os had the least.

Potassium-Urine

The K intake of 30 to 100 mEq. per day is re-flected in the large urinary excretions shown inlines 21-22 and did not appear to be correlatedwith the condition of individual patients.

Chloride-SkinThe previously described low values for Na and

K skin loss were imitated by those for Cl in thatonly 0.29 to 1.71 mEq. per day were measured(line 28). The five patients with essential hyper-tension were grouped closely about an average of0.8 and 1.4 mEq. per day with the control slightlylower (0.47 and 0.91 mEq.) and patient Os, low-est, with 0.29 and 0.63 mEq. per day.

Chloride-UrineExcept for the three-day period with patient

Os, excretion was from 6.20 to 12.57 mEq. perday. No definite correlations with clinical statuswere evident.

466

SKIN ELECTROLYTELOSSES AFTER NACL RESTRICTION

o _ WCS) oad N-d -4 f

'0 t e.Vc

vo '0bUo U

C'4 o'-'O o

o oEnu>^ o

0* * *_4 ~+o 1I)Uf U)

* * --4 -

+

-q C'4 +~o. do-Um o)oo U . +"4 +

o co-In)U)s4 %O

O tr#)O- - c4 4 -4 +

o o-'po o.

O U>) _e

N us0~0 bv

+ o+ +O _ U) _*q *4 . * .60 0 4> 0

cr

8com4J4)z

0 OCEOU)O0 00%6 - 1-4 d

0 00,,'- -4 C'0 :, 64

0

0

0- _ UI)N 00

16t0 C; qWI.4No

o. i . . -

CE) U.) Cy.4

4ot -Ousc6 0-I

o 0Ou)u) +0. E)0 -

_0 C

+

CE) CE) d

'-60t-~

+00-0 I

'0 c'Oluv dOE - I>N -

_IE

O '0O-t +C

* . . . . .4a0"0*U)

(4 (_

4.

(4C3

467

- t

"..

CE)I

fE

0 O

-.9

.-4r

N.'

'0s. ',,00a_4)._

00-_4)4 )

4) 4)-";0 04

bOg

._

* 4-++ S.

LEWIS K. DAHL, BERNARDG. STALL, III, AND GEORGEC. COTZIAS

%AlQ2 C3 040Q5 0.7 Q9 2 3 4 5 6 7.891C

K (meq/d)A

10.

0Q2 Q3 Q.4Q5 Q7C19CJ(meq/d)

C

2 3 4 5 6 78910

B

IC98765-4

3

2K

(meq/d) 00 *

0

0L7Q15 _

0.4 o 7 DAYSO *3 DAYS

_ AC._

02-

0Z 03 Q4C45 Q709CG(meq/d)

D

2 3 4 5 6 78910

FIG. 1. RATIOS OF CONCENTRATIONSOF DAILY ELECTROLYTELoss FROMSKINSince the data are plotted logarithmically, constant ratios will be on parallel straight lines. Points below any line

will have lower ratios than that indicated by the line, and points above a line, higher ratios. Graph A = Na/K,Graph B = Na/Cl, Graph C-.Na+ K and Graph D = K

The two numerical ratios, shown in Graph B, are the ranges quoted by Robinson (14) for normal sweat values.

Balance dataIn Table III, balance calculations for Na and

Cl on this group have been made using stool ex-

cretion data from previous similar experiments(1, 2). Because of the wide variations possiblein intake, balance calculations were not made forK. It will be seen that all of the patients were

in approximate Na balance with the exception ofpatient Os, and the three-day period in patient

Di, whose long-term balance was positive as in-dicated by urinary excretion and plasma Na. Thenegative balance of Os will be noted under COR-

RELATIONS and requires no amplification. Theintake data for chloride are less reliable thanthose for Na since Cl intake was dependent solelyupon the limitation imposed by the Na restriction.Although several patients show slight negativebalances for these short-term studies, with the ex-

468

9-716-NK- LO

4 K

3 X

1.0 _18

08~~~~~~).6 / 0

0 7 DAYSD2 - o3D*3DAYS -

12

Natmeq/d

No+K(meq/c

3 _ I l/876-5-40

3 0 °.

d) o 5(

-'z0-9 Na+K g0.80Q7C26 _ / g _

050A o 7 DAYS

0 3 DAYSQ3

_2

0.1

ii

c

SKIN ELECTROLYTELOSSES AFTER NACL RESTRICTION

ception of Os, there was no long-term evidence ofsuch as indicated by urine excretion or significantdecrease in plasma Cl.

CORRELATIONS

In the accompanying graphs the electrolyte ra-Na Na Na + K and KK ' Cl ' Cl and C- for the skn col-

lections are plotted, and considering the small ab-solute amounts involved, the ratios in the two ex-

periments were in surprisingly good agreement for

any single individual. The Na and N-l ratiosmuch more closely approximate those found insweat after Na restriction (1), and the admini-stration of adrenal control hormones (10-12) thanthose which might be expected to come throughthe skin by diffusion of extracellular fluid with in-sensible perspiration.

Measurements of both plasma Na and Cl madefive and eight days after the completion of theseven and three-day periods, respectively, showedno statistically significant changes from the con-

trol values, except in patient Os with impairedrenal function: in him plasma Na and Cl levelsimmediately before Na restriction were 142.5 and100.2 mEq. per liter, respectively. Ten days laterthey were 135.8 and 86.2 mEq. per liter. Thevalues after the two studies were: 7 days-Na123.5, Cl 84.3; 3 days-Na 129.7, Cl 85.2 mEq.per liter. The patient had received a total of 288mEq. added NaCl orally by the time of the lastplasma chloride study. With this single excep-

tion it may be said that the low skin losses of Naand Cl were not associated with a chronic decreasein extracellular concentrations.

DISCUSSION

These observations have shown that quite smallamounts of Na, Cl, and K we're lost through theskin under the conditions of this experiment.From the data of DuBois and DuBois (13), itwas estimated that the areas not included by thepresent study amounted to approximately 15 per

cent of the total body surface. It is not possible tostate this fraction's contribution to the total elec-trolyte loss, but addition of 15 to 25 per cent over

the listed values would not affect the interpre-tations, nor would calculations based on loss per

unit of surface area. These data indicate that the

minimal dermal loss of about 0.2 mEq. Cl per hourquoted by Robinson and Robinson (14) for nor-mal individuals was diminished to levels ofroughly 0.01 to 0.1 mEq. per hour in these pa-tients. There are very few previous studies withwhich the present data may be compared, sinceinterest has centered largely on the electrolytelosses associated with sweating after exercise orexposure to heat (e.g., 12, 15, 16, 17). In thisgroup, neither of these factors was importantsince gross sweating did not occur. Indeed, itis questionable whether the quantities measuredhere represent the electrolytes of sweat or thosesalts which came through the skin in associationwith insensible perspiration as well as from des-quamated epithelial cells and sebaceous secretion

Na(18). As noted under CORRELATIONS, the -K and

N ratios, at least, are more suggestive of sweat

ratios than diffusion products. But, whether theloss of electrolytes came chiefly from unnoticedperspiration, or via so-called insensible perspira-tion, it is evident that such loss is low, and exceptfor Benedict's fasting patient (5), generally lowerthan previously reported (3, 4, 18). It seems fairto conclude, then, that at least in the absence ofmarked sweating, skin losses of these electrolyteson this regimen may usually be disregarded, evenafter considerable periods of restriction.

It is of interest that the absolute amounts ofelectrolytes washed from the skin in both experi-ments were very similar in any single individual;when calculated on a per diem basis, the losses ofthe three-day experiment of course exceeded thosein the seven-day study. There is evidence thatwith rise in skin temperature more chloride is ex-creted in the sweat at least (19); since skin tem-perature is partly a function of the ambient tem-perature, it is possible that the mean increase of120 in maximal outside temperatures accountedfor the larger daily excretion of electrolytes in theshorter experiment, either as the result of un-noticed sweat in small amounts or by diffusionwith insensible perspiration (18). Another in-triguing possibility, however, is that the electro-lytes were reabsorbed through the intact skin bya process the reverse of that which occurs whenthe electrolytes diffuse outwards during insensibleperspiration.

Several preliminary experiments which were

469

LEWIS K. DAHL, BERNARDG. STALL, III, AND GEORGEC. COTZIAS

done with Na2' to test this conjecture indicatedquite clearly that sodium could be absorbedthrough intact skin. If this is true for Na, it maybe true for other electrolytes as well. No evi-dence is available to indicate whether such absorp-tion occurred via the sweat glands or directlythrough the skin. From this, however, the thesisemerges that in the absence of frank sweating atsome dermal concentration of electrolyte, whichconcentration may vary perhaps with both bodylocation and individual, reabsorption would beginand continue until the excess electrolyte wasremoved.

SUMMARY

1. The loss of Na, K, and Cl from the skin hasbeen measured in seven patients-six hyperten-sives and one normal-all of whomhad been on alow Na, low Cl, normal K diet for periods rangingfrom 26 to 150 days when the experiments began.

2. The measured losses amounted to 0.09 to2.59 mEq. per day, 0.08 to 2.69 mEq. per day,and 0.29 to 1.71 mEq. per day for Na, K, and Cl,respectively.

3. Among the seven patients the one with malig-nant hypertension lost the smallest amounts ofthese electrolytes from the skin; the normal con-trol had values roughly mid-way between the ex-tremely low values of that patient and the greaterlosses of the five patients with essential hyper-tension.

4. The data indicate that despite the rigiddietary restrictions, electrolyte balance was pos-sible among the six members with good renalfunction.

5. The skin washings for the seven and three-day periods contained very similar amounts ofelectrolytes. Therefore, it was proposed that atsome unknown concentration on the skin surface,reabsorption of electrolyte might begin and con-tinue until the usual situation had been rees-tablished.

ACKNOWLEDGMENT

Wewish to thank Mrs. Miriam Brown for her tech-nical assistance.

REFERENCES

1. Dole, V. P., Dahl, L. K., Cotzias, G. C., Eder, H.A., and Krebs, M. E., Dietary treatment of hy-pertension. Clinical and metabolic studies of pa-

tients on the rice-fruit diet. J. Clin. Invest., 1950,29, 1189.

2. Dole, V. P., Dahl, L. K., Cotzias, G. C., Dziewiat-kowski, D. D., and Harris, C., Dietary treatmentof hypertension. II. Sodium depletion as relatedto the therapeutic effect. J. Clin. Invest., 1951, 30,584.

3. Freyberg, R H., and Grant, R. L., Loss of mineralsthrough the skin of normal humans when sweatingis avoided. J. Clini. Invest., 1937, 16, 729.

4. Arn, K D., and Reimer, A., Minimal sodium lossesthrough the skin. J. Cin. Invest., 1950, 29,1342.

5. Benedict, F. G., A study of prolonged fasting. Car-negie Inst. of Washington, 1915, Publication No.203, p. 233.

6. Miller, E. C., Plant Physiology. 2d ed., New York,McGraw-Hill Book Co., Inc., 1938, p. 288.

7. Van Slyke, D. D., and Hiller, A., Application ofSendroy's iodometric chloride titration to protein-containing fluids. J. Biol. Chem., 1947, 167, 107.

8. Sendroy, J., Jr., Microdetermination of chloride inbiological fluids, with solid silver iodate. II. Titri-metric analysis. J. Biol. Chem., 1937, 120, 405.

9. Van Slyke, D. D., Nomogram for correction of lowurine chloride values determined by the silveriodate reaction. J. Biol. Chem., 1947, 171, 467.

10. Conm, J. W., Louis, L H., Johnston, M. W., andJohnson, B. J., The electrolyte content of thermalsweat as an index of adrenal cortical function. J.ClirL Invest., 1948, 27, 529.

11. Conn, J. W., Electrolyte comnposition of sweat. Clini-cal implications as an index of adrenal corticalfunction. Arch. Int. Med., 1949, 83, 416.

12. Conn, J. W., The mechanism of acclimatization toheat in Advances in Internal Medicine, Dock, W.,and Snapper, I., Editors. New York, IntersciencePublishers, Inc., 1949, vol. 3, p. 373.

13. DuBois, D., and DuBois, E. F., The measurement ofthe surface area of man. Arch. Int. Med., 1915,15, 868.

14. Robinson, S., and Robinson, A. H., Chemical com-position of sweat. Physiol. Rev., 1954, 34, 202.

15. Moss, K. N., Some effects of high air temperaturesand muscular exertion upon colliers. Proc. Roy.Soc., London, s.B., 1924, 95, 181.

16. Dill, D. B., Life, Heat and Altitude. Cambridge,Harvard University Press, 1938, p. 39.

17. McCance, R A., Experimental sodium chloride de-ficiency in man. Proc. Roy. Soc., London, s.B.,1936, 119B, 245.

18. Hancock, W., Witehouse, A. G. R., and Haldane, J. S.,The loss of water and salts through the skin, andthe corresponding physiological adjustments.Proc. Roy. Soc., London, s.B., 1929, 105, 43.

19. Robinson, S., Gerking, S. D., Turrell, E. S., andKincaid, R. K, Effect of sldn temperature on saltconcentration of sweat. J. Applied Physiol., 1950,2, 654.

470