6I Slyke, Rhoads, Hiller and Alving [1934a] using urea.

Transcription

1 237 THE RENAL ELIMINATION OF PHENOL RED IN THE DOG 6I2.463 BY H. L. SHEEHAN (From the Department, of Pharmacology, Johns Hopkins Medical School, Baltimore, and the Research Department, Glasgow Royal Maternity Hospital) (Received March 27, 1936) THE introduction of explantation of the kidney by Rhoads [1934] has provided a great advance in methods of investigating renal function by the study of renal-vein blood. The essential step is to bring the left kidney and renal vein to a subcutaneous position so that, for months or even years later, blood can be collected from the renal vein without anawsthesia and without disturbing renal function at all. Previous methods of collecting renal-vein blood have always required anaesthesia and usually the sacrifice of the animal. There is the further advantage that explantation allows the combined study of the renal-extraction ratio and of the urinary clearance of a substance over many successive short periods. The original investigations with this method are by van Slyke, Rhoads, Hiller and Alving [1934a] using urea. The present work is an application of the method to the study of the renal elimination of phenol red in dogs. Investigations with this substance by sacrifice experiments in urethanized rabbits have been reported earlier [Sheehan and Southworth, 1934]. As shown in that paper, there are significant differences between the excretion of phenol red in dogs and in rabbits. The excretion in dogs is also different from that in man [Bernheim, 1926; Shaw, 1925; Goldring, Clarke and Welsh, 1935]; great care must be observed in drawing analogies between renal function even in different types of mammals. In the ptesent work a certain amount of phenol red has been injected into dogs, and at intervals afterwards a series of samples has been collected of arterial blood, renal-vein blood, and total urine. From the data obtained by quantitative analysis of phenol red in these specimens, calculations are then made of (a) the urinary clearance of the dye, (b) the renal extraction ratio, (c) indirectly from (a) and (b), the renal blood flow.

2 238 H. L. SHEEHAN EXPERIMENTAL METHODS Female dogs weighing 1-15 kg. were used. They were carefully trained pets which would lie quietly on the table without restraint during the experiments. They were fed throughout the work on a diet rich in meat and had always unlimited drinking water. Preliminary operation The left kidney was brought to a subcutaneous position in the left loin by Rhoads' method, with the minor modification that the renal vein was enclosed in a skin tube. The formation of this skin tube is not easy, as great care must be taken to avoid any obstruction of the vein: after the operation there must be no swelling of the kidney, and no cedema or haemorrhage around it. The right kidney was not removed, as it was thought undesirable to complicate the experiments by throwing all the work on one kidney [see Rhoads, Alving, Hiller and van Slyke, 1934]. The animals were then left for complete healing of the operation wound; the experiments were conducted at various times from 1 to 8 months after the initial operation. Two dogs with perfect renal-vein preparation were used for numerous experiments and provide most of the data recorded here. The kidney suffers no obvious ill effects from explantation. Two dogs were killed 9 months after explantation; during this time they had both been used for numerous renal-vein experiments. The left kidney was indistinguishable from the right macroscopically or microscopically except for the presence of a rather thick fibrous capsule, and there was no recognizable development of new vessels around it. In one animal the left kidney was 2 g. heavier than the right, in the other it was 2 g. lighter. Renal-vein experiments Phenol red in 6 p.c. solution was injected intravenously in a leg vein in doses of mg. per kg., or subcutaneously in doses of 1 mg. per kg. At definite times afterwards a series of blood samples was collected from the subcutaneous renal vein and from the heart. The venous sample was collected at a moderate rate, in order t6 avoid any danger of drawing blood back from the vena cava into the renal vein. A total of 2 c.c., or for certain purposes 5 c.c., of blood was obtained in each sample. Before the injection a self-retaining rubber catheter was passed into the bladder. All the bladder urine was collected at definite intervals by thorough washing out of the bladder.

3 ELIMINATION OF PHENOL RED Venous blood collection. In earlier experiments, with kidneys explanted by the original Rhoads' method, great difficulty was often experienced in inserting a needle into the renal vein. Several attempts were frequently required, as the vein lay obliquely at some depth beneath the skin, and it was usually found necessary to fix the vein by stretching the skin along the line of the pedicle of the kidney. There was in addition the unlikely but worrying possibility that the needle might penetrate the renal arteries or the ureter. These points do not normally detract from the efficiency of the method; it is astonishing how much the renal pedicle may be traumatized by multiple punctures which miss the vein, often without any apparent interference with renal function. Nevertheless, the skin tube method was found very much more satisfactory. Using an ordinary 5 c.c. aspirating syringe with piston, the collection of blood from the renal vein is as easy and rapid and almost as certain as an ordinary subcutaneous injection. The vein is immediately subcutaneous and accurately localized, and there is no need to touch or pull the skin over the flank in any way. Thus the kidney and renal pedicle are never subjected to any indirect pressure. After the first puncture, further samples are obtained by inserting the needle through the original hole in the skin; in this way several samples can easily be collected at minute intervals. Sometimes, after a few months, the skin tube becomes adherent below and flattens out, but the vein itself does not sink and thus the aspiration of blood is not interfered with. Heart blood collection. The heart blood was collected in a syringe by a long needle passed through the chest wall directly into the ventricle. It was usually obtained about half a minute after the renal-vein blood. The interval was always measured so that the dye content of the heart blood could be corrected to the time of the renal-vein blood collection. This method of collecting arterial blood has one serious cause of possible error. A good heart puncture has no obvious effect on the dog or on its renal function. A bad puncture may cause pain or even collapse, and renal function is then interfered with; in one dog, in which marked collapse occurred at the height of phenol-red excretion, not a trace of dye could be found in the bladder washings for 2 min. afterwards. The condition is, however, uncommon and is easily recognized. Chemical estimations Phenol red was estimated in blood and tissues by the method used in previous work [Sheehan, 1931; Sheehan and Southworth, 1934]. Only one alcohol extraction of the blood was made; the amounts of dye 239

4 24 H. L. SHEEHAN found in the supernatant were corrected according to the recovery of known amounts from blood over the whole range. Recovery is less complete at low concentrations than at high ones, but under standard conditions it is very constant. Small amounts of dye were read colorimetrically in very long thin Nessler tubes. The general accuracy of estimations is to about 2 p.c. with a downward limit of satisfactory readings at about -2 mg. per 1 c.c. plasma. All estimations of phenol red in blood were made on whole blood. The figures thus obtained have been multiplied by a constant factor of 1-6 to bring them to plasma concentrations; the dye is carried only in plasma. THEORY OF CALCULATION OF RESULTS After the intravenous injection of phenol red: (1) The dye content of the blood peaks during about the first 2 sec., and then falls, at first rapidly and then more slowly. (2) The dye content of the kidneys (including the urine in the tubules) rises rapidly for about 2 min. and more slowly for about 5 min.; then it falls gradually during the excretion of the dye. (3) The dye first appears in the bladder urine in about 21 min. and reaches its peak in about 7 min. It is therefore clear that the dye content of the urine at any particular moment cannot be directly compared with the amount of dye in the blood at that time. The two factors of importance which must be taken into consideration are: (a) the initial accumulation of dye in the kidney, (b) the time required for any individual molecule of dye to go from the renal artery to the bladder urine. (a) It is clearly impossible to find the amount of dye which is accumulated in the kidneys of a single dog at various intervals during the experiment. Indirect information can, however, be obtained by killing a number of dogs at various intervals after injection and analysing the kidneys; the results are somewhat variable, but give a general idea of the state of affairs. (The dye content of the kidneys includes, of course, the dye in the lumen of the tubules.) The circles in Fig. 1 are from dogs which were given an intravenous injection of 24 mg. phenol red per kg. in 1 sec. After a prearranged period they were then given 5 c.c. of 2 p.c. KCN intravenously in 5 sec., the abdomen was opened, and the kidneys were removed within 16 sec. after beginning the injection of KCN. The points in Fig. 1 are from similar experiments by Bernheim [1926] and Olivet and Prufer

5 ELIMINATION OF PHENOL RED [1928] on normal or lightly urethanized dogs which had been given phenol red intravenously in doses of about 5 mg. per kg. (One observation of the latter workers was made on a very emaciated abnormal animal, and is not included here.) In order to find the actual amount of dye extracted from the blood by the kidneys, the curve of kidney dye must be added to the curve of excretion of dye in the bladder urine. From the combined curve it is possible to calculate the amount of dye extracted from the blood by the kidneys during any particular minute. It is, of course, clear that the curve of kidney dye is only approximate, in particular for any individual animal, and thus the calculations from it are open to some error. However, except in the first few minutes after injection, the dye content Xto 1~~~~~~~~~~~~~~ Minutes after intravenous injection Fig. 1. Amounts of phenol red in kidneys at various times after intravenous injection of 5-24 mg. per kg. o =Present data. * =Data of Bernheim, and Olivet and Prufer. of the kidneys is small relative to the dye content of the urine, so that minor errors in the former amount are of little significance in the sum. The urinary output of dye in any minute, with this correction for change in kidney content of dye during the minute, gives the amount of dye extracted from the plasma during that minute. The true clearance is found by dividing this figure by the amount of dye in 1 c.c. of plasma during the same minute. (b) The other approach to the problem is to find how long the dye takes to go from the renal artery to the bladder urine, allowing for any temporary accumulation in the kidney cells or delay in the tubules, and for the dead space to the lower end of the ureters. Fig. 2 shows the early excretion of the dye after intravenous injection of 24 mg. per kg. during 1 sec.; the bladder being washed out at intervals of 1 or 2 min. Two of the dogs were normal; two had the left kidney explanted. There is no

6 242 H. L. SHEEHAN significant difference between them. It will be seen that the dye appears in the urine at 2 min. but does not reach its maximum until 7 min., though as mentioned earlier the peak of the dye in the blood is before 2 sec. after injection. Mathematical examination of the curves representing the dye content of heart blood, urine and kidneys indicates that nearly all the dye extracted from the blood by the kidneys in any particular minute reached the bladder between the end of the third minute and the end of the ninth minute, about half reaching the bladder by the end of the sixth minute. Fig. 2. o lo Minutes after intravenous injection Rate of excretion of phenol red in urine after intravenous injection of 24 mg. phenol red per kg. The time of this delayed exrcretion iis of course somewhat approxrimate, as it is computed by the dissection of large curves into a number of small curves. A further criticism is that the calculations are made with most certainty during the first 1 min. after injection, and at this time the bladder exrcretion of the dye may be more delayed than later. Nevertheless, the estimated delaymay be accepted as sufficientlyaccurate for the present purpose. In particular it is presumed here that the dye content of the bladder urine represents on the average the dye exrtracted from the blood about 6 min. earlier by the kidneys. All clearances are therefore calculated from the dye content of the blood 6 min. before the mid-point of the period of urine collection; the clearance being of course

7 ELIMINATION OF PHENOL RED 243 at the earlier time. In the calculations of true clearances this timeallowance method gives very similar figures to the kidney-content method (a). The method of calculating the results may best be illustrated by part of the protocol of an experiment; this includes only the few data relevant to the particular calculations. The dye content of the heart blood at any particular time is obtained by interpolation along the curve of the series of heart bloods in the individual experiment. In the calculation of renal extraction ratios, the correction to renal-vein time which is applied to the heart blood dye content is in most cases negligible. Dog 5. 1 kg. Surface area 43 sq. m. Weight of kidneys 56-2 g mg. phenol red, 6 p.c. intravenously Heart blood 6 2*77 mg. dye per 1 c.c. plasma Urine Renal-vein blood mg. dye per 1 c.c. plasma Heart blood mg. dye per 1 c.c. plasma Urine mg. dye total. At heart blood=248 mg. dye per 1 c.c. plasma. So renal extraction ratio = 1 (281.48) =29 p.c. By comparison with the previous and subsequent renal-extraction ratios it appears that the ratios at 32.3 and 38.3 are probably also 29 p.c. (a) Mean urinary excretion of dye at 38.3 is 1-65 mg. per mm. From graph of kidney dye contents, estimated loss of dye from kidneys at 38.3 is -13 mg. per min. So mean dye removed from blood by kidneys at 38.3 is 1-52 mg. per min. At 38.3 heart blood=223 mg. dye per 1 c.c. plasma. So plasma clearance at 38.3 is 1 52 x 1 = 75 c.c. per min. So renal blood flow at 38.3 is 1 29 x 75 x 1-6=43 c.c. per min. (b) Mean urinary excretion of dye at 38.3 is 1-65 mg. per min. At 32.3 heart blood=2-42 mg. dye per 1 c.c. plasma. So plasma clearance at 32.3 is 1-65 x l=68 c.c. per min. So renal blood 1 flow at 32.3 is f9 x 68 x 1-6 =39 c.c. per min. EXPERIMENTAL RESULTS Two complete experiments are graphed in Figs. 3 and 4; one after intravenous and one after subcutanous injection of phenol red. These show the large number of observations that can be made in a single experiment and the general regularity of the results. There are two striking differences between the two experiments, the renal-extraction

8 244 H. L. SHEEHAN. '4 4-a p4 C; Minutes after intravenous injection Fig. 3. Dog. Wt. 12*5 kg. Surface area X52 sq. m. given 3 g. phenol red intravenously at -2 min.

9 ELIMINATION OF PHENOL RED 245 C) C) C) C; C.) P 5 1 Minutes after subcutaneous injection Fig. 4. Dog. Wt kg. Surface area -52 sq. m. given 125 mg. phenol red subcutaneously at min.

10 246 H. L. SHEEHAN ratios and the plasma clearances. This appears to be due not to the different route of injection but to the different levels of the heart blood dye contents. In experiments which were continued long enough after intravenous injection to allow the dye content of the heart blood to fall to the level of those in Fig. 4, the extraction ratios and clearances increased to the level of those in Fig. 4 also. The matter is illustrated in S 6 Q o!. 4 o\ 2 l ci ~~ 2.. ~~~~~~~~~~~~~~~. o~~~~~~~~~o 2~~~~~~~~ mg. phenol red per 1 c.c. plasma Fig. 5. Relationship of extraction ratios to amount of phenol red in plasma. O = Present data. * = Data of Marshall. Figs. 5 and 6, which show the relationships of extraction ratios and clearances to the dye content of the blood; the data are combined from a number of experiments. It will be seen from Fig. 5 that the extraction ratios fall from about 5 or 6 p.c. at 3 mg. to about 1 p.c. at 2 mg. dye per loo c.c. plasma, and that they stay relatively constant above this concentration. In this figure are included also the data of Marshall [1931] on dogs under paraldehyde aneesthesia; it will be seen that these are in good

11 ELIMINATION OF PHENOL RED 247 correspondence with the present results. The dye is carried only in the plasma, so that there is no back diffusion from corpuscles to plasma in the renal-vein blood after passing through the kidney. Marshall's measurements of extraction ratios, using plasma, are thus quite comparable to the present ones, where whole blood was used for the chemical estimations. Apart from this relationship of extraction ratios to plasma concentrations, there is no obvious relationship of extraction ratios to time xooo t I \ OC a S1~~~~I O-S 1 I-S 1 S S 12 rep h a m ~~~~~~~~~ o. oo Fig plasma mgc phenol red per1 c.c. Relationship of plasma clearances to amount of phenol red in plasma. after intravenous injection after the first minute. (No studies have been made of the extraction ratios in dogs during the first minute, but they are probably high at that time.) In Fig. 6 the clearances show a similar curve, becoming relatively constant at above 2 mg. per 1 c.c. plasma. They are comparable among each other, as they are from dogs of roughly equal size. Most of them were measured on a falling blood dye concentration, but, as Shannon [1935] points out, this is not of significance with regard to phenol red. The shape of this curve has been noted previously by Marshall [1931] and Shannon [1935]. The following table gives a combination of their published figures, obtained on rather larger dogs than were used in the present work. PH. LXXXVII. I11

12 248 H. L. SHEEHAN Plasma clearance, c.c. per min. Plasma phenol red Marshall mg. per 1 c.c. (mean figures) Shannon *2-* * * S han no n, with considerable justification, attaches more significance to the clearances relative to those of inulin. As in his work the inulin clearances were fairly constant, his dye clearances in absolute figures give a similar curve to that relative to inulin. Similar curves relative to creatinine clearances in the dog have been found by Elsom, Bott and Landis [1934] using neoskiodan and hippuran, but not with skiodan. An analogy may be tentatively suggested between the excretion of the two former substances and that of phenol red. Renal circulation rates The renal blood flow is calculated by dividing the blood clearance per minute by one-hundredth of the extraction ratio. The measurement of clearances is somewhat uncertain during the first quarter of an hour after injection, as the blood content is falling very rapidly; no blood flows are calculated for this time. The table below shows the renal blood flows calculated per square metre surface area; they are higher than the blood flows found by van Slyke, Rhoads, Hiller and Alving [1934a] in dogs on a diet poor in meat, but in the same region as the blood flows found by them [1934 b] in dogs on a diet rich in meat. As a renal blood flow of 5 c.c. per sq. m. corresponds in the dog to a renal circulation rate of about 5 c.c. per min. per g. kidney, it appears that the renal blood flow in the dog is very much higher than in the rabbit. Renal blood flow Number of c.c. per mim. per sq. m. surface area observations In any individual experiment the renal blood flow was usually fairly constant, but it was noticed to vary somewhat in any dog from day to day.

13 ELIMINATION OF PHENOL RED 249 INTERPRETATIONS The primary matter in the discussion of the results in Figs. 5 and 6 is the "binding" of most of the phenol red in the plasma by plasma proteins, so that only a part of the phenol red is filterable through a collodion membrane [Marshall and Vickers, 1923]. The amount of binding depends on the concentration of dye in the plasma. No measurements have been made in the present work; the following are approximate figures for the dog taken from Grollman [1925], Marshall [1931], and particularly Shannon [1935]. The exact amounts vary somewhat in individual animals. mg. phenol red per 1 c.c. plasma Percentage bound E k e horn [1935] reviews the evidence that the glomerular membrane is practically impermeable to normal plasma proteins and that its approximate limit of permeability is for proteins of molecular weight between 34, and 6,. If the phenol red is bound to normal plasma proteins, as would appear from the work of Grollman [1925 and 1926], the bound phenol red would thus not be filterable at the glomerulus. If it is bound in the plasma to smaller molecules which cannot pass through a collodion membrane but which can pass through the glomerular membrane, then the bound dye can be filtered at the glomerulus. Richards and Walker [193] have shown that, in the perfusion of frog's kidney, the phenol red that is "bound" in either frog plasma or half-strength horse serum is not filtered at the glomerulus; the percentage of dye filtered at the glomerulus is in each case the same as that filtered through a collodion membrane. This evidence cannot, however, be applied directly to the mammalian kidney, and the conclusion that the glomerular membrane is impermeable to bound phenol red is not universally accepted [E keho rn, 1935]. The extraction of phenol red will therefore be considered here from the two aspects of the bound dye being unfilterable or filterable at the glomerulus. (a) Assuming that the bound dye cannot be filtered at the glomerulus. In the table below, the average figures of the experiments are correlated with the amounts of free dye from the data given above. It will be seen that with less than about 2 mg. total phenol red per 1 c.c. plasma the 17-2

14 25 H. L. SHEEHAN kidney extracts more than all the free dye from the plasma, but at high concentrations it extracts only about a quarter of the free dye from it. On the other hand, the kidney extracts an absolute amount (averaging about.4 mg. dye per 1 c.c. plasma) in excess of 2 p.c. of the free dye in the plasma, when the plasma concentration is above 3 mg. per 1 c.c. At lower plasma concentrations the absolute amount falls. Extracted mg. per looc.c. plasma in Mg. phenol red per 1 c.c. plasma Extracted as excess of _ percentage 2 p.c. of Total Free 2 p.c. of free Extracted of free dye free dye * * * This relationship can be most easily explained by a combination of the interpretation of phenol-red excretion by Shannon [1935] with the filtration figures put forward by van Slyke, Hiller and Miller [1935]. The latter observers worked with creatinine, ferrocyanide and inulin, and conclude that about 2 p.c. of the fluid of the plasma is filtered at the glomeruli. If this figure be accepted and applied to the present data, it would appear that the kidney extracts by filtration about 2 p.c. of the free dye reaching it in the plasma, and that in addition it extracts a small but fairly constant absolute amount of the total dye by direct absorption into tubular epithelium. This is probably about 4 mg. dye per 1 c.c. plasma passing through the kidney at plasma concentrations above 3 mg. per 1 c.c. With fairly steady blood flows of about 5 c.c. per min. the absolute amount would also be related to time (in these dogs about 1-2 mg. per min.); it is not clear which relationship is to be regarded as the more significant, time or blood flow. Such a passage of phenol red directly from the blood into the tubular epithelium requires short discussion. During the first minute after intravenous injection certain dyes are very rapidly taken up by tubular epithelium directly from the blood. Ekehorn [1935] makes the reasonable suggestion that this is a process of simple diffusion, the dye in the tubular cytoplasm being almost immediately bound on microscopic surfaces so that the "free" dye in the cytoplasm is always at a very low level. The gross impairment of the process by tubular damage [Sheehan, 1932] and the relatively low diffusibility of the dyes are, however, serious difficulties in the way of this view. But for the present

15 ELIMINATION OF PHENOL RED 251 purpose it is not of great significance whether the process be regarded as one of vital absorption or as one of diffusion dependent on the vital activity of the epithelium. The essential point is that under given conditions it can be proved that certain dyes pass directly from blood into tubular epithelium. This potential absorptive activity of the capillary surface of tubular epithelium to phenol red cannot be neglected, though absolute proof is lacking in the case of this dye. There is, of course, no theoretical difficulty in the kidney extracting in this way more than the total amount of free dye; direct absorption of free dye would disturb the equilibrium in the capillary plasma, and bound dye there would then become free to attempt to maintain the equilibrium. (b) Assuming that the bound dye can be filtered at the glomerulus. If a little over half of the plasma is filtered at the glomeruli, it is unnecessary to postulate any direct absorption of the dye. It is, however, necessary that, at high concentrations of phenol red in the plasma, either a great reduction of filtration occurs or about two-thirds of the filtered dye is reabsorbed and returned secondarily to the blood. If less than half of the plasma is filtered, some of the dye must be directly absorbed by tubular epithelium, increasing in amount as the blood concentration falls. The varying co binations are too numerous to warrant more precise speculation he. Accepting for pur oses of discussion the view that 2 p.c. of the free phenol red is filtered and that, in addition, about 4 mg. of phenol red per 1 c.c. plasma is directly absorbed by tubular epithelium, the fate of the directly absorbed dye must be considered. The dye is excreted almost quantitatively and, its destruction in tubular epithelium thus cannot be accepted. It is equally impossible to assume that the dye remains stored in the epithelium. The direct absorption of 1-2 mg. per min. would give in the course of an hour 8 mg. of dye in tubular epithelium: an hour after intravenous injection of 25 mg. of phenol red, the kidneys (including all the dye in tubular urine) contain only about 7 mg. of dye. Finally, if it be suggested that the dye which is being directly absorbed throughout the whole period is also being returned to the blood in the same amount and at the same time, it is necessary to go back to the view that over half the total dye in the plasma, free and bound, is filtered at the glomeruli. The simplest explanation is that the dye absorbed directly by the tubular epithelium is then secreted into the tubular lumen. This accounts satisfactorily for the very high clearances of the dye at low plasma

16 252 H. L. SHEEHAN concentrations. By taking inulin clearance as a measure of filtration, Shannon found the "secreted moiety" to be as high in certain cases as 5 mg. per min., though, as he points out, the absolute figures are somewhat variable. 5 mg. per min. is definitely more than is calculated in the present experiments, but the dogs he used were much larger; the relative amounts of the "filtered and secreted moieties" in his experiments are not greatly different from those found in the present work. Secreted moiety as percentage of total mg. per 1 c.c.,a_ plasma Shannon Present experiments * Reference may also be made to certain recent studies on the tubular excretion of phenol red. These are on different lines from those recorded here and therefore will not be discussed in detail [Gersh, 1934; Chambers and Kempton, 1933; Chambers, 1935; Marshall, 1934]. SUMMARY In the dog the extraction of phenol red from the plasma by the kidney has been measured during the excretion of the dye at various times up to 3 hours after injection. The extraction ratio is about 5-6 p.c. at low plasma concentrations, and about 1 p.c. at high plasma concentrations. The clearances of the dye also show an inverse relationship to the plasma concentrations. These facts are in accordance with the view that about 2 p.c. of the plasma is filtered at the glomeruli, and that in addition there is secretion of a relatively constant amount of the dye by the tubular epithelium. Other explanations are,* however, possible, dependent on the question of the filterability of bound phenol red by the glomeruli. My thanks are due to Prof. E. K. Marshall for his hospitality in permitting me to do this work in his laboratory, and for his constant encouragement and critical advice.

17 ELIMINATION OF PHENOL RED 253 REFERENCES Bernheim, E. (1926). Z. klin. Med. 14, 24. Chambers, R. (1935). Proc. Soc. exp. Biol., N.Y., 32, Chambers, R. and Kempton, J. (1933). J. cell. comp. Physiol. 3, 131. Ekehorn, G. (1935). Virchow8 Arch. path. Anat. 295, 256. Elsom, K. A., Bott, P. A. and Landis, E. M. (1934). Proc. Soc. exp. Biol., N.Y., 32, 77. Gersh, I. (1934). Amer. J. Physiol. 18, 355. Goldring, W., Clarke, R. W. and Welsh, C. (1935). Proc. Soc. exp. Biol., N.Y., 32,979. Grollman, A. (1925). J. biol. Chem. 64, 141. Grollman, A. (1926). J. gen. Physiol. 9, 813. Grollman, A. (1926). Amer. J. Physiol. 75, 287. Marshall, E. K. (1931). Ibid. 99, 77. Marshall, E. K. (1934). Physiol. Rev. 14, 133. Marshall, E. K. and Vickers, J. L. (1923). Johns Hopk. Ho8p. Bull. 34, 1. Olivet, J. and Prufer, J. (1928). Z. klin. Med. 18, 653. Rhoads, C. P. (1934). Amer. J. Phy8iol. 19, 324. Rhoads, C. P., Alving, A. S., Hiller, A. and van Slyke, D. D. (1934). Ibid. 19, 329. Richards, A. N. and Walker, A. M. (193). J. biol. Chem. 87, 479. Shannon, J. A. (1935). Ibid. 113, 62. Shaw, E. C. (1925). J. Urol. 13, 575. Sheehan, H. L. (1931). J. Phy8iol. 72, 21. Sheehan, H. L. (1932). J. Path. Bact., Lond., 35, 589. Sheehan, H. L. and Southworth, H. (1934). J. Physiol. 82, 438. van Slyke, D. D., Hiller, A. and Miller, B. F. (1935). Amer. J. Physiol. 113, 611. van Slyke, D. D., Rhoads, C. P., Hiller, A. and Alving, A. S. (1934 a). Ibid. 19, 336. van Slyke, D. D., Rhoads, C. P., Hiller, A. and Alving, A. S. (1934 b). Ibid. 11, 387.