ST » 1 5i eee he it a“ ba ee ai at ae a ok ee oe at ” 4 “ a ae dt ot 20 se ve st * ~~ Hy eat ot ot tm a 7 rl Mon % cuenns aes ne in at oe vi ‘vue " na it > im i a a 7 i ay i — f : DT, es | Te ar ug irae 7 Peiunnh i P. i ine hee " Vay | en ve ie i i ut } },, Ter fs a — a a i 7 A Pavel i fi hire Ad ae An ited OAs tel i ’ ae ; i: cae AE ry. \ 7 i) a A yj ' ' 1 j va > : uit} an Sau rh wo nih ‘ee 1, 7 : ‘a a A fo ie 7 », 1 ve ah | : , f ey 4 wih a MCN iOHM/18l Stree, AN eh | ie ta 720) A Poe nay J in + - fl c us a | a we Rest : ’ i] CA ena 6 4 a uy i, “y ni Tea Ie! ie ce 7 Vie: | id Bee af 7 \ ‘ ty Diip/ ee are me i raae x : he } i i ui rs ae at: yee rat _—_ ta os a ue ant et " yey hye ay) Poa 1 Lave % vA : i 9 i i ADVENTURES IN RADIOISOTOPE RESEARCH ny ae is a ve = 7 a : 3 7 _ @ i - 7 an a a 7 ii. wee ‘| iy wos e >, a 59 ae a ee ~ x i : fe 2” ees 7 n ign) ; mo te q a r Se ae re ears fs a me nn 7 a: :. x wil a : eae eae ae 7 _ aa =) en oT : ADVENTURES IN RADIOISOTOPE RESEARCH The Collected Papers of GEORGE HEVESY in Two Volumes VOLUME ONE PERGAMON PRESS NEW YORK - OXFORD - LONDON. PARIS EVG2 PERGAMON PRESS INC. 122 East 55th Street, New York 22, N.Y. 1404 New York Avenue N. W., Washington, 5 D.C. PERGAMON PRESS LTD. Headington Hill Hall, Oxford 4 & 5 Fitsroy Square, London W. 1. PERGAMON PRESS S.A.R.L. 24 Rue des Ecoles, Paris V° PERGAMON PRESS G.m. b.H. Kaiserstrasse 75, Frankfurt am Main Copyright 1962 Pergamon Press J.td. Library of Congress Card No, 60—12557 Printed in Hungary by The Printing House of the Hungarian Academy of Sciences Nw 10 11 12 13 14 15 Clee \ > ‘ NN # CONTENTS aa INORGANIC AND PHYSICAL CHEMISTRY Votume One Analytical Applications Vhe Solubility of Lead Sulphide and Lead Chromate (with F. Paneth) 31 Piatinirmp Slack: (‘withers (SOmiys)! <8 fe ac.o Woks seed asa ctely michtve say isbea sl ont 36 Heads Contentrotehocksn (wath ea Eloi) Sse eeiesacrcie eo siera iets Seine ee 43 Activation Analysis The Action of Neutrons on the Rare Earth Elements (with H. Levi) .... 47 Artificial Activity of Hafnium and Some Other Elements (with H. Levi) 63 Electrochemistry The Problem of the Isotopic Elements (with F. Paneth) ............ 15 Interchange Studies The Velocity of Dissolution of Molecular Layers (with E. Rona) ...... 89 The Exchange of Atoms Between Solid and Liquid Phases .......... 97 Intermolecular Exchange of Atoms of the Same Kind (with L. Zech- ERT STS COTM NG 72a 2 ayo Se MMT OPO PR oust wart 1H naa So ANS DEN e fa hake Oe 103 Selfdiffusion Sel-diffusion. im, Solid’ eadt (with J: Groh) s4.o 0.065. eee. 110 Self-diffusion in Solid Metals (with A. Obrutsheva)................. 114 The Heat of Relaxation of the Lead Lattice (with W. Seith and L. Keil) 116 Dittusionsimis Metals, (wath. Wie, Seuthy)\ Yi ovs:< re /ohey sus cate Saxpis ag Shd oes cles wie 6h 122 Application of Radioactive Recoil in Diffusion Measurements (with W. SSIENUIEY) 3 hic SR Si ES ame ee BN RATT PO ene RU Reem Oo GW GtCN eee 127 Tracers In The Search For Unknown Stable Elements Search for an Inactive Isotope of the Element 84 (Polonium) (with Jaks QGR EN ais ato eset ee 0) ke cd na Anes Pn a ae Ee ord Mc? Oo 140 6 16 ADVENTURES IN RADIOISOTOPE RESEARCH LIFE SCIENCE Application of Radicactive Tracers Occurring in Nature Radiochemical Method of Studying the Circulation of Bismuth in the Body (with J. A. Christiansen and 8. Lomholt) ................ 143 Radiochemical Method of Studying the Circulation of Lead in the Body (with J: A. Christiansen and?’S; Wombholi)y {5.5 ee.e 4... - 145 Skeleton Studies Radioactive Indicators in the Study of Phosphorus Metabolism in Ratss (with: O2 Chiewiatz))i2% rca. 5 seach enae RR IeNe Is oi Sbccgersionere 149 Studies on the Metabolism of Phosphorus in Animals (with O. Chiewitz) 152 Investigations on the Exchange of Phosphorus in Teeth Using Radioactive Phosphorus as Indicator (with J. J. Holst and A. Krogh) .......... 168 Rate of Rejuvenation of the Skeleton (with H. Levi and O. Rebbe) 191 Retention of Atom of Maternal Origin in the Adult White Rat....... 198 Ratetor Renewal of thesiishy Skeleton™ .2.58...c.- aac eee 204 Conservation of Skeletal Calcium Atoms Through Life .............. 21% Path of*AtomsyihroughtGenerations Wuicwiaee sce. eos 22s ee © syne 234 Note on the Chloride Content of the Mineral Constituents of the Skeleton aay cece ee Sane hac aoe aiken spelnep al siraettas othe GhaueWhy Gad Rigi. ch sie: ekeneae eeney omen 241 Phosphatides The Formation of Phosphatides in the Brain Tissue of Adult Animals (Gas) ned Oped s Libel) oie ee Omics oO On cr eemeaereIO oo 6 coral ob oodd 246 Lecithenaemia Following the Administration of Fat (with E. Lundsgaard) 255 Formation of Phosphatides in Liver Perfusion Experiments (with 117205 [1 01 el) eee teas Peete ena nee ROP CAM EN RUPE SSP ces Ss O.B: 6 HORS SO AO AIOVO the 258 Rate of Penetration of Phosphatides Through the Capillary Wall (Aiiatitlo les Dice 3 (1 oh ay ie en rere Pieri Don aio proiniaD Mitno-ard acon IOC 262 Origin of Phosphorus Compounds in Hen’s Eggs (with L. Hahn) .... 273 The Origin of the Phosphorus Compounds in the Embryo of the @hicken (with EH. Levi and O. Rebbe) 2)..9e 4-s04-+-5-2-- eee iNorsaarenineray our WGMke (Ayahelow Ya\g JElG Wao ANI) ssc csaocenengodsbousc6ncn 304 Formation of Lecithin, Cephalin and Sphingomyelin (with L. Hahn) 309 Turmover of Phosphatides (with G. Eliott)) 25. ....22.-e.2--5~- =) 346 40 41 CONTENTS 7 Acid Soluble Phosphorus Compounds Molecular Rejuvenation of Muscle Tissue (with O. Rebbe) ........ 366 Rate of Renewal of the Acid Soluble Phosphorus Compounds of the IRM oihi-(Qyatles IS Isilon) oo bocs gos wenb oucSo010.dd6 UU Ee eee 369 Circulation of Phosphorus in the Frog (with L. Hahn and O. Rebbe) 384 Fatty Acids Turnover Rate of the Fatty Acids of the Liver (with R. Ruyssen and UL, di [gh Yexeo) Fa a oF eH aks) haters CRP Sener ne e SR A rey nen Ue De a 403 Effects of Dinitro-Cyclo-Pentylphenol on the Incorporation of Labelled Acetate Carbon (#4C) Into Tissue Fractions (with L. M. Beeckmans eld CSF Ta CASTS) Brome eter an ca le cules Chen oS atic Mee Ratan Salty Susie, Seeje to eae eoe eRe 408 Determination of the Rate of Renewal From the Rate of Disappearance one tabolled Molecuies. oxi <- casei soos eee totale wie et aheae stare ee hs Be ALT Permeability Studies Rate of Penetration of Ions Through the Capillary Wall (with L. Hahn) 425 Rate of Passage of Water Through Capillary and Cell Walls (with OEE, el BCOWSOIVE ots). Such Riso ahs here al Be He aloo, ARS CSR tA rs et 437 Rate of Penetration of Phosphate into Muscle Cells (with O. Rebbe) 443 The Effect of Excitation on Nerve Permeability (with H. Euler and (Wh, BUI Rey eR RE ee eh eee eee ee 454 Note on the Inorganic Phosphate of Blood Plasma (with G. Elliot and L. Hahn) Fate of the Sulphate Radical in the Animal Body (with A. H. W. Aten) 465 Diplogen and Mish’ (wath) EH Hofer) ...02..0..0.00..e cc css ee wae sence 468 Interaction of Plasma Phosphate with the Phosphorus Compounds Present In The Corpuscles (with A. H. W. Aten)................. 471 Rate of Penetration of Ions Into Erythrocytes (with L. Hahn) ..... 493 Votume Two Labelling of Red Corpuscles A Method of Blood Volume Determination (with L. Hahn) ......... 517 Determination of the Red Corpuscle Content (with K. Zerahn) ..... 523 Thoriume bsLabelled Red “Corpuseles, «.< 2). s.). ssr0c4 gcc ctalocee. sass O81 Clinical Investigations Elimination of Water from the Human Body (with E. Hofer) ...... 536 Excretion of Phosphorus (with L. Hahn and O. Rebbe) ............ 540 On (0 6) 60 67 68 69 74 ADVENTURES IN RADIOISOTOPE RESEARCH Potassium Interchange in the Efuman Body 22:..55:0.......20s00. 553 The Red Corpuscle Content of the Circulating Blood Determined by Labelling the Erythrocytes with MRadio-Phosphorus (with K. H. Koster, G. Sgrensen, E. Warburg and K. Zerahn) .............. 561 Application of 42K Labelled Red Corpuscles in Blood Volume Measure- monte (with "G.. Nylinm) 2.5 sersceepeeeeerieiet «os old. a bes sie ee a okomeeeerte 573 Application of ‘‘Thorium B’’ Labelled Red Corpuscles in Blood Volume Studies. (with:G.. Ny lin) 225 cee eee ence 3.0 na, oa cine ae eee 580 @anecer tANReM1a ;. b. 2 2.000 elk Sy Cho ROPE CID ote bike ois Rice Se eae ee 597 Iron Metabolism Effect of Adrenaline on the Interaction Between Plasma and Tissue Constituents’ (withsG.. Dal sSanto), \aeraretrerie cic ae iene everest 610 Effect of Irradiation on Hemin Formation (with R. Bonnichsen) 624 Haemoglobin Present in the Nuclear Fraction of the Liver (with R. Bonnichsen, G. Ehrenstem and J. Schliack) .........2..57::.... 634 Application of Isotopic Indicators in Haematology ................. 639 Note on the Determination of Radioiron (with K. Agner and R. Bonnich- SEW) eee ee eS ee eae Sole ei eae fase Noe to ve pac pes Sie ewan 651 Embryonal Iron Turnover (with G. v. Ehrenstein) ................ 655 Nucleic Acids Rate of Formation of Nucleic Acid in the Organs of the Rat (with ee OLCESCTa) ee EI, See eee. a Faye ay cia vd eo chien wean oie vet yee eas 663 Rate of Renewal of Ribo- and Desoxyribo Nucleic Acids (with E. Ham- TOY HGS|15101) eR eR Rin eRe AA a eR MR ho OG G.C ba ob.0 a6 ob 673 Turnover of Ribosenucleic Acid in the Jensen-Sarcoma of the Rat Gwithy EL. -Enler and W, Solodkowska)/ i... 7. eee ee 680 Life-Cycle of the Red Corpuscles of the Hen (with J. Ottesen) ..... 688 Studies in Radiation Biology Effect of X-rays on Nucleic Acid Formation in the Jensen- Sarcoma (with: Et. Ailey ey se Loci eee eee ee ec kcls ors: 692 The Effect of X-rays on Nucleic Acid Formation in the Organs Of The Rat (with L.. Ahlstrom’ and’ Hesmuler)-. a2. >. =... ee 721 Turnover of Nucleic Acid in Retrogade Sarcomata (with L. Ahlstrém ro lag] 8 Doel O40 | (eyo) eee oii sls oS atcncueenneroro.t ci0 20 0 7 DIG 13k The Indirect Effect of X-rays on the Jensen-Sarcoma (with L. Ahlstrém 744 Bhavol Tats 1D oa Ae a eC ReE CEOs Cie ono c Silo do Gite ooo ovo Okc 75 SO 81 82 83 84 85 86 87 88 89 90 91 92 93 94 CONTENTS 9 Attempts to Find Products Blocking Nucleic Acid Formation in the Circulation of an Irradiated Organism (with L. Ahlstré6m, H. Euler SMG! IK. VARIN) We 6 AA SMa ce no Cae Boat Gidlqtn Cid co iG Cn 758 Fate of The Nucleic Acid Introduced inte the Circulation (with L. PU SErOMmianiCe ElomENIEeR)) Mies eee gd CEP et nee phasis Gis o's See ec ys oles 761 Formation of Nucleic Acid in Sarcoma Slices. (with L. Ahlstr6m and Ele len waren same RL rer ok oxtreceute ge tee ae waleitec tee cies eee os 770 Application of Labelled Substrates in the Study of Enzymic Processes (wales Ala stromata delete Maler)) . f25 5 oc Bole, oiana sue penser ee ie, Sev che shee « 783 Effect of X-Rays on the Incorporation of Carbon-14 into Desoxy- MOM GleIG eA CIs, aed moh cos Senne! ile eas cians «says hee Sse payee a oats 791 Effect of X-Rays on the Incorporation of Carbon-14 into Animal MISO ae Re ele ES cea ape RG iA ayes of Qioichivete ad aitage altydeie soe) See ater 7193 Effect of X-Rays on the Incorporation of 14-C into Tissue Fractions GlatheeMouse (with G. Dreytus)i 2. 05 cavers were mates) d aecterele 3.2 eueleun seu 795 Effect of Muscular Exercise and of Urethane Administration on the Incorporation of Carbon-14 into Animal Tissue .............. 821 Effect of Irradiation by X-Rays on the Exhalation of Carbon Dioxide bya bhesMonse (wath A Worssberg) — 2.6654. Gie sea in ei a on et 825 Effect of X-Rays and Hormones on Resorption Rate of Injected NETIC Or. A(ayitkt. An, ChOrssbere)) /5:06/s-2q ene e sss oe le Bi ane oe ela cae oe 828 Note on the Effect of X-Rays and Hormones on the Resorption Rate of anjected) NaMHCO; (with A. Forssberg). 2,-2.2... feu e065 6 so dees 838 Effect of Irradiation with X-Rays on the Catabolism of Methyl- Slcon olin here MOMS) ose hee, hots AN a oot Ecslennes hes ete cea eane 9 hae 84] Effect of Irradiation with X-Rays on the Catabolism of Ethylalcohol TIPS UO MOUSO meme aor et yes, G,- a, 5 Acer ti NS eae ae io acl Sasa sia 847 Radioactive Tracers in Radiobiological Studies. The Thirty-Sixth Sulyanus! Chompson Memorial Lecture. 0:20. 62 oss eed eens oss 851 Botanical Studies The Absorption and Translocation of Lead by Plants ............ 876 Atomic Dynamics of Plant Growth (with K. Linderstrg@m-Lang and (Gis MOUS EST ce aoe a ae Cr ae ea dot ones RMN ge beni ee a aah ee 884 Exchange of Phosphorus Atoms in Plants and Seeds (with K. Linder- Spreia Wang andy ©. (OSE). 2s 05. 2s « wlouls hers EAD) Liles ee nc Sun S87 Interaction Between the Phosphorus Atoms of the Wheat Seedling ear Eom UELEIOTIG | SOLUULO Ms Baryl eee Glee cas Vara Ax slam rAgeon L.Sale wo ss 891 Exchange of Nitrogen Atoms in the Leaves of the Sunflower (with K. Linderstrgm-Lang, A. S. Keston and C. Olsen) ............... 905 Zinc Uptake by Neurospora (with I. Andersson-Kotté) ............-. 910 10 ADVENTURES IN RADIOISOTOPE RESEARCH 95 Phosphorus Exchange in Yeast (with K. Linderstrom-Lang and N. Niel- SISTED). 03 aid Se te EN Sh a 5 ic. of AAG a Ae 96 Potassium Interchange in Yeast Cells (with N. Nielsen) ........... 97 Note on the Number of Pollen Grains Identified in the Fruit of the Aspen (with C. Eklundh-Ehrenberg and H. Euler) ............. 92¢ Lectures 98 Some Applications of Isotopic Indicators. Nobel Lecture .......... ‘ 99 The Application of Radioactive Indicators in Biochemistry. Faraday WCEYC bes iy SGP AR Ne von fone en PEPPERS CSS 5 Gute cy Sera) aM 100 Historical Progress of the Isotopic Methodology and its Influences on the Biological Sciences. Read at the Turin Meeting of the Society7oL Nuclear Medicine: ©.’ :::: .} ase ene Gores eres s meee ee Index Sie 210) 0}16! (0/6) :0\),0) 0) 6: 1¢| (e) (0) ,0) 0) ¢.\0)/ 0 6) 6 ja) ‘© (0) 0) 0) ©) 0: 0 16] 0) 6) 0) (6: \e! 0) 0) iei.eNe! (ville! (elle) (exe! (e) @! Ke, (6110) 6170.6) First communicated in Perspectives in Biology and Medicine Vol. I, No. 4, Summer 1958 A SCIENTIFIC CAREER GEORGE HEVESy, Ph. D. (hon.), Ph. nat. D. (hon.), D.Sc. (hon.), Se. D. (hon.) M. D. (hon.) Jur. D. (hon.). I was born in Budapest the 1st of August, 1885. After terminating my studies at the Gymnasium of the Piarist Order in that city, I studied a short time in Budapest and Berlin and Jater in Freiburg, mainly che- mistry and physics, where I took my degree in 1908 an. The subject of my doctoral thesis was the interaction between metallic sodium and molten sodium hydroxide, an interaction responsible for a poor yield often obtained when producing sodium by electrolysis of molten sodi- um hydroxide. Being interested in high-temperature chemistry, I proceeded to Ziirich after obtaining my degree to work under Richard Lorenz, at that time the most eminent representative of that branch of science. The Technische Hochschule of Ziirich was in those days, as it is today, a great place of learning and teaching. The Swiss chemical and pharma- ceutical industry could not have reached its present high standard it represents today without the aid of a great number of able chemists, most of them trained at the Technische Hochschule of Ziirich. When I joined this institution, the permanent head of the chemistry depart- ment was Willstatter. Einstein's First Lecture Shortly after my arrival at Ziirich, Einstein was appointed associate professor of theoretical physics on the University. I was one of the audience of about twenty who attended his inaugural lecture on the determination of the ratio of charge and mass of the electron. (Einstein left after a few years for Prague and returned later to Ziirich to fill the chair of theoretical physics on the Technische Hochschule.) When he visited our laboratory, I had the privilege to show him around. I remember vividly his astonishment when shown a hydrogen electrode. He thought such an electrode to be only a theoretical concept. Twenty-three years later, after terminating my Baker lectureship on the Cornell University at Ithaca, I met Einstein in Pasadena. I visited a iby ADVENTURES IN RADIOISOTOPE RESEARCH barber shop whose owner, a son of the City of Constance, praised the beauties of life in California, mentioning that his only wish in life was to be permitted once to cut Einstein’s hair. I told him that this wish would not be easy to fulfill as, according to rumors, Mrs. Einstein used to perform this work. When I told Einstein about the barber’s wish, he remarked : “Da er sich auf Ihrem Kopfe nicht austoben konnte, wollte er meinen Kopf haben” (“As he could not sufficiently exercise himself on your head [I had poor hair] he wants to have mine’’). Einstein talked repeatedly to me on the problem of causality. He dis- agreed with Bohr’s views on this topic and asked me to convey his objections to Bohr. He wished an explanation on a classical basis. When Lorenz left Zirich for the University of Frankfurt (I was later, after his death, asked to fill his chair). Willstitter called on me to make the short statement: “In Germany the assistant belongs to the professor, in Switzerland to the laboratory—you stay here.”’ I did not, as I got much interested in the catalytic synthesis of ammonia by Haber, a discovery which at that date rightly impressed all those interested in chemistry. My monthly salary in Ziirich corresponding to $36 was entirely ade- quate, as 1 was charged $15 a month for a very nice room and two good meals a day. When I was promoted to a “first assistant’’, I was told that my salary would be raised to $60 a month, the highest pay ever allotted to an assistant. When I was leaving the laboratory one evening together with Will- statter, he told me that he was moving to Berlin to take over one of the Kaiser Wilhelm Institutes. I asked him, much astonished, why he was leaving. He was the permanent head of the chemistry faculty and had a very fine laboratory, and postgraduate students from all over the world were anxious to work under his guidance. His answer was: “If the fatherland calls, it is my duty to go.”’ Thirty-two years later I was present at the meeting of the Danish Academy of Sciences when the president, S. P. L. Sérensen, death written on his face, read a letter from Willstatter requesting that the Proceedings of the Academy should no more be sent to him. Willstatter went on, saying: “I have no home any more. I have lost all my belongings, which I do not mind much. What chagrins me is that I lost my fatherland.”’ Haber wished me to work in another field than that of catalytic synthe- sis. I was to investigate whether or not oxidation of molten zine is accompanied by emission of electrons. No one in Haber’s institute had experience in the field of the conductivity of electricity in gases. I pro- posed therefore to Haber that I proceed to England to acquire some knowledge in this new field of physics and return later to his laboratory. Haber entirely shared my view, and I left in the first days of January, 1911, for Manchester to work under Rutherford. A SCIENTIFIC CAREER 13 Years with Rutherford The physics laboratory of the University of Manchester was housed in a spacious building. The chief equipment of the institute was electro- scopes built with cocoa cans, sealing wax, sulphur rods, gold leaves, and reading microscopes. Once adjusted, the electroscopes were not permitted to be cleaned, and the smokey atmosphere of Manchester left its visible marks all over the laboratory. The years I had the privilege to spend in Rutherford’s laboratory in Manchester, between 1911 and 1914, witnessed some of the greatest discoveries in the history of physics. I could follow from close quarters the discovery of the atomic nucleus and how Rutherford devised, carried out, and interpreted the results of ex- periments. All this was done with the greatest ease, without visible effort. Niels Bohr came to Manchester in 1912. He recently remarked in an after-dinner speech that I was the first-one he met when he entered Rutherford’s institute. Rutherford—and not he alone—soon realized Bohr’s genius. When I was enjoying Rutherford’s hospitality one Sun- day afternoon, soon after the discovery of the atomic nucleus, I happened to ask him about the origin of B-rays. The a-rays clearly originated from the nucleus, but what about the origin of B-rays? Rutherford answered promptly, “Ask Bohr’’, and the answer was at once given by the latter, emphasizing the difference between nuclear and non-nuclear /-particles. Bohr was however not always easy to understand. When he briefly stated, “Argon is not the right argon’, he made a statement that was at that date not easy to interpret. It was then, in 1912, already clear to him that it is not the mass number but the atomic number that is decisive for the place of an element in the periodic system. Soon after, this fact was decisively brought out by Moseley’s work. I consider myself lucky to have had the opportunity to help Moseley set up the first X-ray spectrograph. We turned to the steward of the chemistry department, Mr. Edwards, who handed us a beautiful, very large potassium ferri- cyanide crystal which found application in Moseley’s spectrograph. A mag- netic device served to bring small metal disks covered with the element to be investigated into the electron beam, which had to excite the X-rays. Moseley’s fundamental work brought out, among other things, that, while the atomic weight of argon is higher than that of potassium, its atomic number is not—that “argon is not the right argon’, as stated by Bohr previously. Moseley has also shown that the anomaly of the posi- tions of tellurium and iodine in the periodic system disappears if we consider the atomic number instead of the atomic weight. When I arrived at Manchester, Rutherford wished me to study the solubility of actinium emanation in various liquids. It was not an easy task in view of the short life of this emanation, now called actinon, the half-life of which is four seconds only. This was, however, a very good 14 ADVENTURES IN RADIOISOTOPE RESEARCH school to learn the handling of short-lived substances. I later became engaged with the study of the electrochemical properties of radioele- ments of unknown chemical character and the measurement of their valency from diffusion data. The early origin of the famous Geiger-counter goes back to those Manchester days as well. Rutherford and Geiger counted a-particles by making use of a galvanometer which registered the arrival of each a-particle. The ionization produced was magnified by using the principle of production of ions by collision. The much more difficult task of counting f-particles was solved later, after the first World War, by Geiger, then at Kiel. When I was in Manchester, Rutherford was much interested to come into the possession of a strong radium D sample. Large amounts of radium D were stored in the laboratory, but imbedded in huge amounts of lead. The great German chemist Haber intended to pay Germany’s war debts after the first World War by extracting gold from the ocean. First he undertook to check the correctness of the available gold analyses of sea water. He found the gold content of the ocean to be very much lower than previously found. He summarized the depressing results of his expedition by stating : “Dilution is the death of all value’. Ruther- ford could have made the same remark when glancing at the hundreds of kilograms of lead chloride extracted from pitchblende and presented to him by the owner of the Joachimsthal mines, the Austrian govern- ment. Radioactive Tracers One day I met Rutherford in the basement of the laboratory where the lead chloride was stored. He addressed me by saying: “If you are worth your salt, you separate radium D from all that nuisance of lead.”’ Being a young man, I was an optimist and felt sure that I should succeed in my task. Trying during a year all sorts of separation methods and making the greatest efforts, it looked sometimes as if I succeeded, but I soon found out that it was radium E, the disintegration product of radium D, a bismuth isotope, which I separated. The result of my efforts was entire failure. To make the best of this depressing situation, I thought to avail myself of the fact that radium D is inseparable from lead, and to label small amounts of lead by addition of radium D of known acti- vity obtained from tubes in which radium emanation decayed. From such tubes pure radium D can be obtained. It was the Vienna Institute for Radium Research which owned in those days by far the greatest amount of radium and, correspondingly, of radium emanation. This fact induced me to interrupt my stay in Manchester and to proceed to Vienna. In the Vienna Institute there were A SCIENTIFIC CAREER 15 very large amounts of lead chloride, obtained from pitchblende as well, and Paneth, assistant at the Institute, unaware of my efforrs at Man- chester, made very extensive studies to achieve separation. His great efforts were as abortive aS mine. At my suggestion we associated in the application of labelled lead.The first use of this method, early in 1913, was the determination of the solubility in water of sparingly soluble salts such as lead sulphide and lead chromate. We then proceeded to study the electrochemistry of bismuth and lead by making use of the method of radioactive indicators. We could show, among other things, that Nernst’s law of the dependence of the electrode potential on the ionic concentration is valid even at exceedingly low concentrations. Paneth then directed his interest toward the interaction of the lead ions present in the surface layer of lead sulphate and the labelled lead ions of the surrounding solution. I studied the interaction of the lead atoms of a lead foil and also of lead peroxide with the lead ions of a solution, employing labelled lead foils and non-radioactive lead salt solution, or vice versa. In the last of the numerous joint investigations with Paneth, we succeeded in preparing visible amounts of radium D from radium emanation. By comparing the electrode potential of radium D peroxide with that of lead peroxide, we were able to show that these cannot be distinguished from each other. During my stay in Vienna, I undertook balloon ascents in the company of Hess and Paneth. On one of his trips Hess took an electrometer with him to follow the change in the ionization of the air with height. He assumed this ionization to be due to terrestrial radiation and corres- pondingly expected it to decrease with height. The opposite, however, was found to be the case. With such simple means and without much effort this observation led to the discovery of cosmic radiation. Madame Marie Curie When passing through Paris on the way to Manchester, I never failed to call on Marie Curie and I was always sure to find her amidst experi- mental work. She was usually surrounded by several girl assistants precipitating or crystallizing preparations. The only protection that she used was finger caps of rubber. When engaged with the concentration of actinium from rare-earth samples, she generously presented me with an actinium preparation. I consider this specimen one of my most pre- cious belongings. As the years pass by, the bottle containing the radio- active sample is getting more and more coloured, indicating the many years which have elapsed since I met this most remarkable personality and great pioneer. At a later visit to the Institut de Radium, I met Joliot, who was then a young assistant engaged in the study of the electrochemistry of polo- 16 ADVENTURES IN RADIOISOTOPE RESEARCH nium, which many years earlier was in the center of interest of Paneth and myself. Also, Irene Curie worked in the laboratory of her mother. When I saw her in 1938, she mentioned that by neutron bombardment of thorium she had obtained a lanthanum-like radioactive body. I asked her if she was sure that this substance was not actinium. She answered that she was pretty sure she was dealing with an element much lighter than one of the radioactive disintegration series. A few months later Otto Hahn and Strassman made their fundamental discovery of nuclear fission. I first met Hahn in Vienna in 1913. Already at that date he had made such important discoveries as the existence of radiothorium and mesothorium and the separation of radioelements by making use of the recoil phenomenon. The years to come, brought new discoveries of great importance, many of them in collaboration with Lise Meitner. When I asked Rutherford in 1912 whom of his students he considered to be the most merited one, he answered without hesitation **Otto Hahn’. On my way to Manchester I usually stopped in London. On such an occasion I had the opportunity of being present in the House of Commons at the introduction of the much discussed budget by Lloyd George, then Chancellor of the Exchequer, who characterized his introduction of heavy death duties and other taxes as “bringing rare and refreshing fruit”’! I was also present when J. J. Thomson in April, 1913, delivered his Bakerian Lecture in the Royal Society on the two neon parabolas ob- tained in his positive ray studies. He did not make any allusion to the analogy between the two neons and the isotopes in the field of radioacti- vity. This omission induced me to write to him drawing his attention to the analogy between the two kinds of neon, on one hand, and radium D and lead, on the other. He stated in his answer that he did not share my view. While not adopting the view that the heavier constituent of neon was a compound NeH,, which could have given the observed ato- mic weight within the limits of experimental error, Thomson was not convinced that this explanation was absolutely excluded. As Lord Ray- leigh remarks in The Life of Sir J. J. Thomson, he had always been haunted by this suspicion about hydrogen compounds and, for that rea- son, hesitated for a time to accept Aston’s later results about isotopes of other elements. When we were on a ski-trip at Finse in Norway, Aston related that when he first succeeded in getting two lines on a mass spectrum photograph — one indicating *Cl, the other °7Cl — Thomson refused to look at the photograph, which, Aston added, was the most beautiful one he ever obtained. Aston was an ingenious and most merited experimenter, who was the first one to prove the complexity of the com- mon elements. In 1914 Moseley moved to Oxford and, being much interested in X-ray spectroscopy, I intended to work with him. We wanted to study A SCIENTIFIC CAREER ily the X-ray spectrum of the elements 68 through 72. I was already in Holland on the way to Oxford when the first World War broke out, soon followed by the tragic death of Moseley. While talking on a field telephone at Gallipoli, a bullet struck the head of this ingenious and most remarkable man. By a curious coincidence, I was to oceupy myself extensively with the X-ray spectrum of the element 72 eight years later. Measurement of Self-diffusion AsI was at that time a Hungarian subject (I am now a Swedish citizen) I was drafted into the Austro—Hungarian army. I spent much of that time as technical supervisor of electrolytic copper works. While located in Carpathian plants, I fitted up a laboratory on a very modest scale and studied the difference in the chemical behaviour of the active deposit of thorium when present in ionic and colloidal state. For several months after the end of the war it was not possible to leave Hungary. During these months I started with my friend Groh to study self-diffusion in molten and in solid lead, using radium D as an indicator. We fused a radiolead-rod on to the top of an inactive lead-rod, heated this solid system to 200°—300°, and determined the dislocation of the radium D atoms. From the extent of dislocation, the rate of self-diffusion of lead was calculated. This early, rough method was improved later during my stay in Copenhagen. Together with the Russian scientist Mrs. Ob- rutsheva, we condensed the lead isotope thorium B on top of a lead foil and counted the number of scintillations produced by the a-rays emitted by the disintegration products of thorium B. Upon heating of the sample, thorium B diffused into the lead foil, resulting in a reduction of the number of scintillations observed. From this reduction, the diffusion rate of lead in lead could be calculated. Heisenberg, then lecturer in Copenhagen, very kindly at that time helped us with these calculations. Later on in Freiburg, Seith and myself made use of the recoil pheno- menon to measure self-diffusion in lead. This is an exceedingly sensitive method, which permitted measurement of diffusion rates as slow as 104* cm?/day. The Rockefeller Foundation started to support my investigations in 1930 and continued most generously to do so for the following twenty- five years. Niels Bohr’s Institute In the first days of May, 1919, I left for Copenhagen to spend some time with Niels Bohr at the charming summer house in Tibirke. At that time his premises were at the Technological Institute of Copenhagen, from which he directed the construction of his new institute. When he 2 Hevesy 18 ADVENTURES IN RADIOISOTOPE RESEARCH had to decide on a name for the new institute, he hesitated between “Theoretical Physics” and “Atomic Physics’. His choice fell on the first one ; he felt that the latter might be too exacting and possibly too special as well. In front of the Technological Institute there is a statue of Olaus Romer, the first physicist to measure the velocity of light. I once pointed out when passing this monument that space is available for a future monument of Niels Bohr. My companion smiled at this remark. Today he would not smile any more. It was settled with Bohr that I should be back in Copenhagen in the spring 1920, to start activities at the new institute which was to be opened by that date. I spent the remaining six months with my friend Zech- meister in Budapest carrying out exchange studies by the application of radioactive indicators. When dissolving in water both 1 mol of labelled lead nitrate and 1 mol of non-radioactive lead chloride, or labelled lead chloride and non-radioactive lead nitrate, after separation of the two compounds, we found the radioactivity equally distributed between PbCl, and PbNO,. When dissolving non-radioactive tetraphenyl lead and radioactive lead nitrate, after separation all radioactivity was con- served in the nitrate sample, as the lead atoms of tetraphenyl lead are not exchangeable. When I met Svante Arrhenius in 1922 he told me about his interest in the above-mentioned work. The experiments with labelled lead chloride and non-labelled lead nitrate, or vice versa, are the most direct proof of the correctness of the theory of electrolytic dissociation. After the war, I was anxious to go to England as soon as possible. The atmosphere at that time, however, radically differed from the one that prevailed after the second World War. When in 1921 I wrote from Copen- hagen to Rutherford, a very liberal man, that I wished to visit England, he answered that it was still too’early for a former enemy to come to England. In 1923, however, when he was elected president of the British Association meeting which was to take place in Liverpool, he invited me to address that meeting on the discovery of hafnium. I recall a lunch party at Liverpool in which Lord and Lady Rutherford, Niels Bohr, Millikan, Aston, Coster, and myself took part. Lady Rutherford remarked that this party included four Nobel Laureates. Rutherford added, ‘‘And some embryos’’. Rudolf Schoenheimer During my stay in Liverpool I was told about the work of Blair- Bell who claimed a successful cancer therapy through administration of lead compounds. These results induced me, when I worked at the University of Freiburg some time later, to study the distribution of labelled lead compounds between cancerous and normal tissue. A study A SCIENTIFIC CAREER 19 of the distribution of labelled lead and bismuth in healthy rabbits was carried out earlier, in 1924, in Copenhagen. I approached the great pathologist Aschoff to delegate one of his collaborators to help us in our work. He first delegated the director of a hospital on the island of Formosa and later, to help him, his chief chemist, Rudolf Schoenheimer. This was Schoenheimer’s first experience with tracer work, a field to which he later, jointly with his eminent colleague Rittenberg, made unsurpassed contributions. Schoenheimer was already at that date a very nervous man. He moved his limbs incessantly, smoked cigarettes, and consumed coffee on a much too liberal scale. When our work was finished, he left Freiburg and I never saw this most merited man again. mee | Separation of Isotopes When I went to Copenhagen in the spring of 1920, Bohr’s institute was not yet ready. I associated with the eminent physicochemist Brgn- sted to investigate a problem of great interest to both of us, namely, the partial separation of isotopes on a preparative scale. We based our pro- cedure on the more rapid rate of evaporation of the lighter isotope from a liquid. We distilled mercury in high vacuum at 40° and prevented the more rapidly evaporating lighter isotopes from being reflected back into the liquid mercury by freezing them on a glass surface cooled with liquid air. By repeating this process some hundred times, we obtained a light and a heavy mercury fraction. The results were controlled by both density measurements and atomic weight determinations, the latter being carried out by Hénigschmid in Munich. When partially separating the isotopes of chlorine, we made use of the above-mentioned method again. We distilled concentrated solutions of hydrochloric acid in water and obtained several liters of water con- taining hydrochloric acid with different isotopic chloride composition. I suggested to Brgnsted that he have a look at the density of the water ob- tained. He objected to my suggestion, as shortly before two distinguished German chemists, Vollmer and Stern, had searched without success for other isotopes of hydrogen and oxygen than !H and 16O. These workers carried out diffusion experiments through porous membranes. When I, after Urey’s discovery of deuterium, reminded Brgnsted of my suggestion, he answered: “A discovery like this should not be made fortuitously ; it should be based on careful considerations like Urey’s.” Bohr was highly interested in our separation experiments and keenly followed our progress. Bohr’s greatness is due not only to his ingenuity but to the unique catholicity of his interests, his sagacity, and his immense conscientiousness. When as a young man he intended to publish his first ‘‘letter’’ to the editor of Nature, he wrote the note over and over again. Finally his brother, who later achieved fame as a great mathema- 2* 20 ADVENTURES IN RADIOISOTOPE RESEARCH tician, suggested he should now mail the “letter’’, Niels Bohr was shocked by this suggestion, since, he said, this was the first trial of the first con- cept of the “‘letter.”’ In this spirit all his papers were written. How fabulously far-sighted Bohr was, is seen from a letter which the present writer addressed to Rutherford after the Birmingham meeting of the British Association for Advancement of Science from Budapest the 14th October 1914. “The meetings on Monday and Tuesday have been very interesting. It is a most remarkable fact that Aston succeeded to separate the two Neons by diffusion and gave a definite proof that elements of different atomic weights can have the same chemical properties. Thomson came in his paper on X, to the conclusion that the latter is a polymerized hydrogen, a kind of H, (like O3). In the following discussion Bohr —in his usual modest way — suggested the possibility that X, being an H atom with one central charge, but having a three-times heavier nucleus than hydrogen. He suggested to let a mixture of H and X, diffuse through palladium and try if it is possible to separate them, as the heavier X, atom has to diffuse much slower. “Bohr had not been understood properly and Thomson gave a rather quick answer, saying — after a brief consultation with Ramsey — that Bohr’s suggestion is useless, for not molecules, but the atoms of H diffuse through Palladium. Certainly, but this was just Bohr’s point. “The general appearance was, that he told something highly ingenious and Bohr something very stupid. Just the contrary was the case. So I felt bound to stick up for Bohr and explained the meaning of Bohr’s, suggestion in more concrete terms, saying that Bohr’s suggestion is that Xs, is possibly a chemically non-separable element from Hydrogen... Of course not very probable, but still a very interesting suggestion; which should not be quickly dismissed”... 26 years later Tritium was discovered. Simultaneously with the isotope separation studies, I carried out among other things some tracer-work on the interchange between the atoms of lead compounds and lead, all in molten state. In 1921 Bohr’s institute was opened. Those working at the institute at its start were, besides its director, H. Kramers, H. M. Hansen, I. C. Jacobsen, James Franck, who was invited for a short visit, and myself. In my first study at the institute I measured the conductivity first of a single crystal of sodium nitrate and then after it was molten and resolidified into a polycrystalline mass. This crystalline conglomerate was found to have a specific conductivity, fifty times higher than the single crystal. From this result it was concluded that deviations from the ideal crystalline state promote electrolytic conductivity. While increase of temperature produces a reversible loosening of the lattice, we are A SCIENTIFIC CAREER aA here faced with an “irreversible loosening’ of the crystal structure. This was a most modest beginning in a field that later proved to be of great importance. Hafnium Bohr’s first fundamental papers, published in 1913, in which the quan- tum theory of the atomic structure was introduced, dealt only with the structure of the atoms of hydrogen, helium, and lithium. InJanuary, 1922, I learned during a walk together with him that he now had extended his theory to the entire periodic system, giving among other things an ex- planation of the appearance of the rare-earth elements in that system. Their number according to his theory was restricted to fourteen, from which it followed that the unknown element 72 cannot be a rare earth, it has to be a homologue of the titanium group. In the summer of that year I became interested in geochemical prob- lems. Returning to Denmark, I proposed to Coster, who previously had studied X-ray spectroscopy with Siegbahn in Lund, that he should teach me the technique and that we ought at the same time to have a look at zirconium minerals for the missing element 72. The first spectrum obtained by him demonstrated the presence of the element in zirconium minerals, and further studies revealed its presence in all commercial zirconium samples, which indicated a very close kinship between zirco- nium and the new element hafnium. By a very protracted fractional crystallization of ammonium zirconium hexafluorides, hafnium could be prepared in a pure state. The discovery of hafnium was not accepted without opposition. Urbain, in Paris, a few years earlier crystallizing crude ytterbium salts, observed twenty-six optical spectral lines not shown by the initial sample. He ascribed these lines to the presence of the previously un- known element 72 in his sample. After the discovery of hafnium, it was, however, demonstrated that none of these lines is to be found in the spectrum of hafnium. In spite of this fact, Urbain upheld his claim to have discovered element 72. Rutherford took great interest in our work —all our extensive correspondence with Nature passed through his hands—and suggested that I should send a paper on the chemistry of hafnium to the editor of Chemical News. He remarked in his letter that the editors of this periodical were strongly pro-French and I should not mind if they refused to publish my paper. In Sheffield, my friend the physicist Lawson (“interned’ as a prisoner of war in the Institute of Radium Research of Vienna) handed my contribution over to the editor of Chemical News, Professor Wynne. He remarked that he was pleased with the paper but they might have something to say about the name “hafnium,” adding: “We adhere to the original word Celtium given to bo bo ADVENTURES IN RADIOISOTOPE RESEARCH it by Urbain as a representative of the great French nation which was loyal to us throughout the war. We do not accept the name which was given it by the Danes who only pocketed the spoil after the war.’ The paper was, however, published by Chemical News without remark. Another opposition to the discovery of hafnium came from London. Alexander Scott, the chief chemist of the British Museum, could not iden- tify a fraction of a sample of an Australian titaniferous sand. After our discovery was announced, he thought this fraction to be hafnium. Scott’s paper induced the 7’%mes to publish in its February 2, 1923, issue an editorial under the title “Hafnium’’, stating: “Science is, and doubtless should be, international, but it is gratifying that this chemical achievement, the most important since the late Sir William Ramsay isolated helium in 1895, should have been the work of a British chemist in a London laboratory.” Scott’s sample, sent us for investigation, did not contain a trace of hafnium or zirconium. The determination of the hafnium content of a great number of zir- conium minerals and historical zirconium samples was a fascinating task. Berzelius determined the atomic weight of zirconium by analyzing its sulphate. This method supplies too low values for the atomic weight. This error was, however, compensated by the presence of hafnium, almost twice as heavy as zirconium, in hissample. Venable in South Carolina, who spent many years with the determination of the atomic weight of zir- conium, applied a modern method devised by Richards at Harvard. He could not find the reason why his determination led to a clearly too high value. After the discovery of hafnium, he sent us a sample of his zirco- nium, and, after taking into account its quite appreciable hafnium content—which we determined—he could correct the presence of hafnium in his sample and arrive at a precise value for the atomic weight of zirconium. Through my work with hafnium I came into contact with the great Austrian chemist Auer von Welsbach. He invested a part of his very substantial royalties obtained for his patent of cerium-iron alloys (applied in cigar-lighters among other things) in a beautiful estate in Carinthia on which he built a castle. The rough crystallization of rare earths was carried out in one of his nearby situated works, and the final crystalliz- ation was done by himself in his castle. He was at that date and for many years to come the only man who possessed highly purified samples of all elements of the rare-earth group. When staying with him, he expressed his astonishment that when separating hafnium from zirco- nium I had chosen to handle large amounts of fluorides, which are highly unpleasant compounds to work with. He achieved all his great success in the field of rare-earth chemistry by crystallizing double- sulfates. We found out later that there is no significant difference between the solu- bility of zirconium and hafnium double- sulfates, and if we had chosen A SCIENTIFIC CAREER I3 to crystallize these compounds, we would not have been able to separate hafnium from zirconium. All hafnium commercially available for the next twenty-five years was prepared by crystallizing the double- fluorides. V. M. Goldschmidt Auer von Welsbach presented me with small samples of octohydro- sulphates of all elements of the rare-earth group. This gift enabled me to measure the density of these compounds and to observe a systematic decrease of the size of the ions of rare-earth elements when proceeding from cerium to lutetium, a contraction which explained the extreme kinship of zirconium and hafnium, which are more closely related chemi- cally than any other elements of the periodic system. (When testing the rare earths for radioactivity, making use of Auer von Welsbach’s samples, we discovered that samarium emitted a-rays.) In Oslo, V. M. Goid- schmidt simultaneously observed the contraction of ionic size, proceed- ing from one rare earth to the next one and denoted this rare-earth con- traction as the “lanthanide contraction.’ Goldschmidt described his and my work in his posthumously published book Geochemistry, a most fascinating reading, like everything that he wrote. V. M. Goldschmidt was one of the most able men I ever met. Endowed with an immense knowledge and a fabulous memory, he was full of fertile ideas. A few weeks prior to the occupation of Norway, I spent a few days with him at his home on Holmenkollen near Oslo. He predicted the tragic happenings of the coming years, which very few foresaw. He mentioned that his pupil and former assistant Lunde soon would become a “Gauleiter’” of Norway. Lunde was later the Minister of propaganda in the Quisling government. Goldschmidt predicted that the Norwegian coast batteries would fail to fire at the invading enemy, which they in fact did with very few exceptions. He was also endowed with much humor. When Quisling came into power, Goldschmidt was imprisoned and all his property seized. Being short of phosphorus fertilizers, the government released him and instructed him to prepare phosphorus from Norwegian minerals. All his property, however, remained confiscated. When German colleagues passed en route to Rjukan, where they had to inspect the heavy-water works, they called on Goldschmidt. He invited them for dinner, encouraging them to eat with the remark : “Please go on eating, gentleman, all you consume is state property.” Tracers in Biology During the work with hafnium, I continued the tracer work and in 1923 applied radium D and thorium B as tracers in the study of the uptake of lead by bean seedlings and also in the removal of labelled lead by non- 24 ADVENTURES IN RADIOISOTOPE RESEARCH labelled lead from such seedlings. This was the first application of radio- active tracers in biological studies. The following year we extended these studies with my friends Christiansen and Lomholt to the distri- bution of lead and bismuth in the animal organism. Potassium is one of the few radioactive elements found in nature outside the members of the disintegration series. We were interested to find out which of the potassium isotopes is radioactive. For this purpose we carried out a partial separation of the potassium isotopes, applying the same method used when separating the isotopes of mercury. A few kilograms of metallic potassium were distilled and a heavy and a light potassium fraction obtained. From the difference in the activity of these samples and the difference in their atomic weight, the mass number of the active isotope could be calculated. Among other instru- ments that were used to measure the activities of our samples was the first counter built by Geiger in his institute at Kiel. The atomic weight of our sample was determined by Hénigschmid in Munich. From these data it was concluded that the mass number of radioactive potassium is 41. The first one to draw my attention to the fact that this result was probably wrong was Baxter, when I visited him at Harvard. He had found that, in constrast to all other atomic weight figures determined by Honigschmid, that of potassium was wrong. Baxter proposed to determine the atomic weight of our potassium samples. From his results it followed that the mass number of radioactive potassium is 40. The two greatest authorities in the field of atomic weight determination thus arrived at different results as to the atomic weight of our potassium samples. To reach a decision, we extracted the small calcium content of an old potassium-rich mica. If “K were the active isotope, then the calcium isolated should contain “Ca. Aston could not, however, find any “Ca in our sample. Thus “4K does not disintegrate and is not radio- active. Baxter was right. A few years later Fermi and collaborators observed the production of an artificial potassium isotope when bombarding potassium with neutrons. We obtained with Miss Hilde Levi Fermi’s product by bombarding scandium and also calcium with neutrons. As scandium has only one stable isotope, we could conclude from our investigations that Fermi’s radiopotassium has the mass number of 42. Activation Analysis Auer von Welsbach was very cautious in giving away his very valuable rare-earth samples, but one day when I was staying with him he was in a generous mood and told me to choose one of his samples, of which A SCIENTIFIC CAREER bo fw) | he said he was willing to give me a larger amount. I chose dysprosium without having any special reason to do so. Ten years later, after the discovery of artificial radioactivity, we exposed Auer’s dysprosium to slow neutrons and succeeded in producing an exceedingly strongly active radiodysprosium. No element is known that can be actived more inten- sively than dysprosium and europium. Exposure of Auer’s europium to a neutron beam also led to the formation of a very strongly active radio- europium, while no active gadolinium could be prepared by using radium- beryllium as the source of neutrons. At that time my friend Professor Rolla, of the University of Florence, who prepared a few kilograms of gadolinium oxide, sent me samples of this material which he wished us to analyze for europium by X-ray spectroscopy. We had earlier analyzed several of his samples quantitatively applying secondary X-rays, a method which was worked out in the Freiburg laboratory. Having no access to a Roentgen spectroscope in Copenhagen at this time, we tried together with Miss Hilde Levi to analyze the samples by exposing them to a flux of slow neutrons. All the samples contained some euro- pium. By preparing standards containing a known amount of pure gado- linium and pure europium, we could arrive at quantitative figures for the europium content of Rolla’s samples. This was the start of activa- tion analysis, which has since become an important tool in analytical chemistry. It was possible by this method to determine, for example, the minute amounts of sodium and potassium in a nerve fiber. Deuterium as a Tracer Urey’s epochal discovery of deuterium took place while I worked in Freiburg. Most kindly he promptly supplied us with some liters of waters containing 0.6 per cent of deuterium oxide. This low heavy- water concentration sufficed to study the interchange of the water molecules between goldfish and the surrounding water and also to determine the water content of the human body, making use of the principle of isotope dilution already introduced a few years earlier (1931) when we determined the lead content of rocks. The mean lifetime of the water molecules in the human body was determined as well. When I returned to Copenhagen in the fall of 1934, August Krogh called on me immediately upon my arrival. He wished to apply labelled water in his permeability studies. I initially intended, upon return to Copenhagen, to do work with deuterium on similar lines as later published by Schoenheimer and Rittenberg. The possibility of obtaining artificial radioactive isotopes, however, induced me to abandon this plan and to concentrate on the application of radiophosphorus in biological studies. 26 ADVENTURES IN RADIOISOTOPE RESEARCH Radioactive Phosporus As a neutron source we had only radon-beryllium, later radium-beryl- lium mixtures, at our disposal. When Niels Bohr celebrated his fiftieth birthday, his friends presented him with 600 milligrams of radium, which he most kindly put at our disposal. With such modest neutron sources, the only tracer of an element of physiological importance which could be produced having sufficient activity was radiophosphorus. We irradi- ated 10 liters of carbon disulphide from which carrier-free =P could be easily separated. All our preparations, however, had an activity below 1 ue. The first problem attacked was whether the mineral constituents of the skeleton are renewed or not during life. Labelled phosphate was administered to rats, and the specific activity of their plasma inorganic phosphorus and skeleton apatite phosphorus compared. The comparison indicated a 30 per cent renewal in the course of the first 24 hours. The amount of phosphate involved in this process exceeded twenty times the phosphorus content of the blood. Thus a large part of the phosphorus present in the soft tissues must have been released and applied in the replacement of skeleton phosphorus. This result demonstrated the dynamicity of phosphorus metabolism. These conclusions were pub- lished about the same time, in 1935, as the first paper by Schoen- heimer and Rittenberg appeared in which they demonstrated the dynamic nature of fat depots. It was followed by a great number of other most illuminating papers in which deuterium, and later heavy nitrogen, was applied as a tracer. Since 3P has a half-life of fourteen days only, happenings through life of a mouse cannot be followed using this tracer. However, applying Ca we succeded a few years ago in showing that only one-third of the calcium atoms of the skeleton of the mouse are replaced during life. The above-mentioned first application of an artificial radioactive isotope as a tracer was followed by our investigation of whether and to what extent the phosphatide molecules of the brain are renewed. These investigations were extended to other organs and to the formation of labelled phosphatides in the chick embryo following the injection of ®2P into the fertilized egg. We transfused labelled plasma of a rabbit to a sister rabbit and followed the rate of disappearance of the labelled phosphatide molecules from the circulation of the second rabbit and their accumulation in various organs. The next step was the study of the rate of renewal of the ATP, creatine, and similar molecules, partly in collaboration with Professor Parnas in Lwow, Poland. With Arm- strong and also with Krogh and Holst, we studied **P incorporation in dentine and enamel. In one of the early applications (1937), the penetra- tion of 22=P into yeast cells was traced and shown to be an almost one- way process. This investigation was made possible by co-operation with A SCIENTIFIC CAREER Za ‘ Linderstr6m-Lang and Olsen at the Carlsberg Laboratory. The first investigations of the uptake of 3*P by plants (1936—37) was also carried out in co-operation with them. In 1940 Professor Hasting, who had formerly visited Copenhagen, invited me to deliver the Dunham Lecture at Harvard University. Denmark was occupied, and messages to the United States could be sent only by the United States Legation in Copenhagen. When I called on the minister asking him to forward a cable to Professor Hasting stating, “I shall be in New York the 21st of June,” the Minister remarked, “Vou ’d better write ‘I intend to’’”’. It was a wise remark, as | did not succeed in getting to the United States and the Dunham Lecture was ultimately delivered by Schoenheimer. We observed, with Aten, that while phosphate penetrates comparat- ively slowly into erythrocytes, it is incorporated very rapidly into labile organic acid-soluble molecules. Thus the red corpuscles are a kind of trap, though imperfect, for ?2P, a fact which makes it possible to tag red corpuscles with =2P, re-inject these into the circulation, and from the dilution figure calculate the red corpuscle volume of the subject in the course of a day. This method of red corpuscle volume determina- tion found an extended application. The first clinical determinations could be carried out with the minute *2P activities prepared by us by irradiation of carbon disulfide with neutrons emitted by a radium- beryllium source. To investigate the formation of phosphatide or of casein in the milk of the goat, which was the subject of the dissertation of A. H. W. Aten, larger activities were needed. These were prepared by Martin Kamen, put at our disposal by the great kindness of Ernest Lawrence. He supplied us later also with **Na and #K. We used these isotopes, among other purposes, to study the rate of interchange of vascular with extravascular ions. We were much impressed by the observ- ation that within the first minute a very large fraction of the soglium ions of the circulation, for example, was replaced by extravascular sodium. Today we know that the exchange-rate values obtained by the tracer method supply minimum figures only. Maz von Laue’s and James Frank’s Nobel Medals My work was interrupted for only one day during the enemy occupa- tion of Denmark. When, on the morning of Denmark’s occupation, I arrived in the laboratory, | found Bohr worrying about Max von Laue’s Nobel medal, which Laue had sent to Copenhagen for safe-keep- ing. In Hitler’s empire it was almost a capital offence to send gold out of the country, and, Laue’s name being engraved into the medal, the discovery of this by the invading forces would have had very serious consequences for him. (Three years later the invading army occupied 28 ADVENTURES IN RADIOSIOTOPE RESEARCH Bohr’s institute.) I suggested that we should bury the medal, but Bohr did not like this idea as the medal might be unearthed. I decided to dissolve it. While the invading forces marched in the streets of Copen- hagen, I was busy dissolving Laue’s and also James Frank’s medals. After the war, the gold was recovered and the Nobel Foundation gene- rously presented Laue and Franck with new Nobel medals. The Nobel Prize In December, 1935, on their journey home from Stockholm, where they were presented by King Gustaf V with the Nobel prize, for their fundamental discovery of artificial radioactivity, Frédéric Joliot-Curie and his wife stayed for a while in Copenhagen. It was then that Joliot mentioned that he, his wife, and the third French Nobel laureate, Jean Perrin, proposed me for the Nobel prize and also that they failed to obtain the adherence of the Paris Academy to their proposal — the celtium-hafnium controversy was not yet forgotten. During the war Niels Bohr with his extreme kindness remarked to one of his friends that one of the numerous disturbances created by the war was that I could not receive the Nobel prize. The shocking refusal of the acceptance of the prize by Domagk, Butenandt, and Kuhn at the order of their ruler made the Swedish Academy of Sciences reluctant to distribute further prizes during the war. In 1944 the Academy decided, however, to award me the prize for 1943. With the war going on, no festivities were held, and the prize, contrary to the usual custom, was handed over to me in a meeting of the Academy of Sciences by the president. Radioactive Tracers in Radiobiology In. 1940 we got interested, with L. v. Hahn, in the formation rate of desoxyribonucleic acid, DNA. While the incorporation of *2P, for ex- ample, into adenosintriphosphate of the growing liver indicates mainly renewal of these molecules and not an additional formation, the incor- poration into desoxyribonucleic acid indicates the latter to at least a very large extent. By investigation of the effect of ionizing radiation on the incorporation of #P into DNA, it should thus be possible to find out if irradiation blocks DNA formation. Together with Professor Hans von Euler, we studied in Stockholm the incorporation of **P into the DNA of the Jensen sarcoma of rats and found in the investigated 100 rats exposed to Roentgen rays a marked depression of **P incorporation, and thus a marked depression in the rate of formation of DNA. Similar results were obtained when investigating *?P incorporation into the DNA in the various organs of growing rats. Indirect radiation effects ~were A SCIENTIFIC CAREER 29 observed by us as well. These were among the first application of radio- active tracers in radiobiological studies. Our joint investigations, among others, were extended to the determination of the number of fertilizing asp pollen, the atoms of which can be located in a seed. The incorporation of 2P into DNA of the nucleated erythrocytes of the hen was found, in collaboration with Ottesen about the same time, to be quantitatively conserved during the lifetime of the erythrocytes, which enabled us to measure the life-cycle of the red corpuscles of the hen. Prior to and during the war I saw a lot of August Krogh, famous physiologist and a man of great kindness, to whom I was much indebted. While staying in Stockholm, he wrote down a detailed program of further permeability studies in which radioactive tracers would have to be applied. It is much to be deplored that he could not witness the great further success of his eminent pupil Ussing in this field. Radioactive Carbon My chief activities since 1943 have been in Stockholm and, for some years after the war, in Copenhagen too. During the last years I have been attending solely to my laboratory in Stockholm. I extended the radiation studies to the measurement of C incorporation into DNA in the organs of growing mice, which was found to be depressed in contrast to incorporation into proteins. My colleague Forssberg and I studied the effect of irradiation on bicarbonate, glucose, and fatty acid meta- bolism and other problems, applying “C as a tracer. These studies, among others, led to the discovery of a fatty acid fraction of the liver having a very rapid turnover rate. For the last years we have been interested in physiological and clinical problems of iron metabolism. In 1953 I had the privilege to deliver the Aschoff Memorial Lecture, which is given each year in the University of Freiburg to commemorate the great pathologist. Aschoff was not only one of the great pathologists of this century but a man of great wisdom and vision. The British patho- logist Robert Muir wrote in his obituary- note on Aschoff, published during the war, “I think one may say that in the period since Virchow’s time, he has been the outstanding figure.’’ Aschoff showed some interest in our early work with lead and was quite enthusiastic about the determina- tion of the volume of the body water by applying heavy water as an indicator, which was the first clinical application of isotopic tracers. In my Lecture I mentioned that our investigations had led us to the conclusion, not unanimously accepted by the audience, that the forma- tion of haemoglobin is not radiosensitive, that so long as erythropoetic cells with an incomplete haemoglobin content are present in the bone marrow, even if the organism is exposed to Roentgen radiation, additional hemoglobin is laid down in these cells. Since then this conclusion has 30 ADVENTURES IN RADIOISOTOPE RESEARCH been fully corroborated by work carried out in our laboratory, and especially by the beautiful work carried out in Oxford by Lajtha and his associates. When we started with Paneth in the first days of 1913 to apply radium D asa tracer of lead, the word “isotope” was not yet coined. Groups of radioactive substances such as mesothorium and radium, or ionium and thorium, were denoted by Soddy as “‘chemically inseparable elements”. Much has happened since those days! Originally published in Z. anorg. Chem. 82, 322 (1913) 1. THE SOLUBILITY OF LEAD SULPHIDE AND LEAD CHROMATE GrorGE Hevesy and Fritz PAaANnrtTH From the Institute of Radium Research of the Vienna Academy of Sciences THe fourth decay product of radium emanation, RaD, is known to exhibit all the chemical reactions of lead; if RaD is mixed with lead or lead salts it cannot be separated from the lead by any chemical or physical method! and if complete mixing of the two substances has taken place then the same concentration ratio is maintained whatever amount of lead is withdrawn from the solution. Since RaD can be determined in much smaller amounts, owing to its radioactivity, than lead, it may be employed for the qualitative and quantitative estimation of lead to which it has been added; the RaD is an indicator of the lead. The lower limit for the qualitative detectability of lead in its most sensitive microchemical reaction? (precipitation of K,PbCu(NO,),) amounts to 3 x 10-9 gm; the limit for quantitative determination lies considerably higher and varies with the particular problem ; for ex- ample, the solubility of lead carbonate could be obtained from deter- minations of the conductance but KoHLRAuUSCH? was able only to make an approximate estimate for lead chromate in this way. With the aid of RaD as a tracer these solubilities can easily be determined directly ; an amount as small as 10-1 gm RaD may be measured, by means of an ordinary and not particularly sensitive electroscope, if one is content to measure the f-radiation of RaE which comes to equilibrium with the RaD after a few weeks. By awaiting the formation of a quantity of RaF sufficient for calculating the equilibrium amount, it is possible to determine quantitatively even 10-12 gm of RaD by means of the a- radiation. In radiolead from pitchblende there is about 10-7 gm of RaD per gm of lead and thus 1 mgm of radiolead can be detected with the aid of its f-radiation ; since much smaller orders of magnitude are involved 1A review of relevant experiments is contained in the paper by F. Paneru and G. Hevesy in Monatsh. Chem. 42, 1 (1913). 2J. Emicu, Lehrbuch d. Mikrochemie p. 80 (1911). °F. Koutrauscu, Z. phys. Chem. 64, 159 (1908). 34 ADVENTURES IN RADIOISOTOPE RESEARCH in the solubilities discussed above we must therefore prepare artificially radiolead by the addition of relatively large amounts of radium-D to lead nitrate. 1. DETERMINATION OF THE SOLUBILITY OF LEAD CHROMATE About 0.2 ¢ of emanation was allowed to decay in a closed flask over distilled water and the solution thus obtained, containing about 10~-* gm RaD in water, was added to a solution of approximately 10 mgm PbCl, in water. The lead was then quantitatively precipitated with potassium chromate, filtered, washed from the filter into a stoppered bottle and shaken with about 100 cm? of distilled water in a thermostat at 25° for a period of 24 hr. The mixture was immediately filtered, the first portion of the filtrate being rejected because of a possible change in its concentration as a result of adsorption on the filter, and 70 cm$ of the remaining filtrate was evaporated to dryness on a _ watch- glass-shaped nickel tray over the water bath. When equilibrium had been established between the RaD and Rak the activity on the tray was measured. The calculation was done as follows: 1 cm? of the RaD solution used for labelling the lead showed (also after establishment of equilibrium) a p-activity of 16.90 arbitrary units and, therefore, the whole solution, amounting to 120 cm’, contained 2030 units. This activity had been distributed on 9.69 mgm of lead chloride or 11.35 mgm of lead chromate and thus one arbitrary unit of RaD was associated with 11.35/20380 = = 0.00559 mgm lead chromate.The 70 cm of solution which had been evaporated had deposited an activity of 0.15 units on the tray and thus 0.15 x 0.00559 = 0.000839 mgm of lead chromate must be on the tray. Hence, the solubility of lead chromate at 25° C is calculated to be 1000 x 0.000839/70 = 0.012 mgm/I. A second experiment with the same solid phase also gave 1.2 x 10~ gm/I. The first experiments, carried out with smaller amounts of RaD and accordingly with a much lower accuracy, yielded values which varied between 3 x 10-5 and 6 x 10-5 gm. Lead chromate is therefore the most sparingly soluble lead salt; only the solubility of lead phosphate is of the same order of magnitude. Apart from a rough estimate by F. Kontravuscu! based on a measure- ment of conductance of the saturated lead chromate solution, there are no data available on the solubility of lead chromate ; IKoHLRAUSCH estimates the solubility as 10-4 gm/I. 1F, Kountrauscn, Z. phys. Chem. 64, 159 (1908). THE SOLUBILITY OF LEAD SULPHIDE AND LEAD CHROMATE 33 2. DETERMINATION OF THE SOLUBILITY OF LEAD SULPHIDE For these experiments 9.69 mgm of lead chloride (8.36 mgm in terms of sulphide) were labelled with 140 cm? of another solution of RaD which contained 66.2 arbitrary units per cm’. The lead was then quantit atively precipitated at the boil with a hot solution of Na,S, the PbS was filtered off, washed and shaken with distilled water as described in the case of lead chromate. The filtrate, the first part of which was again rejected, was completely clear and colourless; it contained 415 arbitrary units of RaD per litre. In this instance one arbitrary unit corresponds to 8.36/140 x 66.2 = 9.0 x 10-4 mgm of lead sulphide, and thus 1 1. of solution at 25° C contained 415 x 9.0 x 10-4 = 0.37 mgm or 3.7 X 10-4gm. The same value was obtained after filtering the solution once again; other experiments yielded the values of 300 and 320 ar- bitrary units per litre, i.e. 2.70 and 2.88 x 10-4 gm lead sulphide per litre. A part of the lead present in the solution probably occurs as hydroxide, owing to hydrolysis, instead of sulphide, as suggested by O. Waicer?. The very weak turbidity obtained when the completely clear saturated solution, prepared by shaking water with PbS, is treated with a stream of H,S supports this view. We have therefore determined the solubility of PbS in water saturated with H,S; the solution from which the PbS is precipitated cannot be used directly for determining the solubility since a portion of the PbS passes as a colloid through the filter; the once-filtered PbS, on the contrary, is already freed from the small particles passing through the filter and these do not recur when the solution is shaken with distilled or H,S-saturated water. In the solution which is saturated with H,S and PbS, the concentration of H,S is about one thousand times that of the PbS. The solubility of the latter is less than the value obtained in distilled water; calculated on the basis of 1 1., the arbitrary activity amounted to 148 and 173 and hence the amount dissolved was 1.33 and 1.56 x 10-4 gm, respectively. It is not possible to decide with certainty whether there is a decrease in solubility due to an increase of the S ion concentration or due to prevention of hydrolysis ; the first case, however, is improbable since the decrease in solubility is only very slight in proportion to the high concentration of H,S. In analytical practice, moreover, this problem need not be considered ; it is only of interest to know the amount of PbS which is present in solution in a clear filtrate ; our experiments give an average value for this of 3 x 10-4 gm in the absence of H,S and 1.5 x 1074 in a solution saturated with H,S. If the filtrate runs turbid through the filter, it is evident that the amount of PbS will be greater. In one instance we observed 1—2 mgm/l. 20. WEIGEL, Z. phys. Chem. 55, 293 (1907). 3° Hevesy 34 ADVENTURES IN RADIOISOTOPE RESEARCH W. Brirrz! determined the solubility of PbS by means of an ultra- microscopic method: When two equivalent solutions which produce a precipitate are mixed in a series of experiments at increasing dilution and the mixture obtained is observed with an ultramicroscope it is noted that, beyond the limit of macroscopic differences, the number of suspended particles of the precipitate becomes steadily less until, at a certain dilution, the mixture appears to be empty or no longer different from its components. This limiting value for the disappearance of the un- dissolved excess corresponds to the solubility of the substance produced, i.e. lead sulphide. Birrz finds a value of 1.3 mgm/I. for the solubility of lead sulphide at room temperature. He remarks that the determination is made more difficult, in the case of sulphides, because they form colloidal solutions which are almost optically transparent at a high dilution ; separate particles can of course be produced by adding salting- out electrolytes with dissimilar ions but at the same time this may cause an increase in solubility. The solubility determined by the ultra- microscopic method is therefore probably rather too large. Correspond- ingly, O. WetcEu! found that 0.86 mgm of freshly precipitated PbS dissol- ved in 11. by calculating the solubility of PbS from the conductance on the assumption that all the PbS going into solution is hydrolysed. Freshly precipitated PbS, however, undergoes a transformation and after about 20 hr have elapsed the solubility amounts only to about 0.43 mgm/I. The PbS used in our experiments was already transformed and the solubility of 3 x 10-4 gm in 1 1. which we found agrees very well with WEIGEL’s value?. RaD is not the only radioelement which can serve as an indicator for lead; careful studies by FrLEeck? demonstrate that thorium-B, radium-B and actinium-B also cannot be separated from lead. The last two cannot indeed be considered for our purposes but thorium-B, with its half-life of 10.6 hr, might well be applied with success as an indicator for lead. Besides lead, we know of two other elements with which a radio- element can be used in practice as an indicator, viz. bismuth and thorium. The former can be labelled with thorium-C or preferably with Rak,* while the latter may be labelled with uranium-X, radioactinium, radio- thorium or, best of all, ionium®. 1W. Bintz, Z. phys. Chem. 58, 288 (1907). 10. WEIGEL, Z. phys. Chem. 55, 293 (1907). 27. Bernrecp, Z. phys. Chem. 25 (1898) considers the PbS electrode to be a reversible electrode of the second kind and calculates the lead ion concentration to be 10~4 at the PbS electrode at one atmosphere pressure of hydrogen sulphide from the electromotive force of the cell Pb |1N Pb(NO,), | 1N NaHS | PbS. 3 A. Frecx, Proc. Chem. Soc. 29, 7 (1913). 4A. Frecx, Proc. Chem. Soc. 29, 7 (1913). 5 F,. Soppy, Radiochemistry. London (1911). THE SOLUBILITY OF LEAD SULPHIDE AND LEAD CHROMATE 35 An advantage of the indicator method is that, irrespective of impuri- ties, only the amount of the labelled element is measured, whereas in other very highly developed microanalytical methods of determination, e.g. by employing microweighing, there is always the danger of co-deter- mining invisible impurities. Apart from this, the sensitivity of the radioactive indicator methods is indeed significantly greater and, assu- ming the availability of adequate amounts of the radioactive substance, can be increased almost without limit. Summary The solubility of lead chromate at 25° in pure water has been determined as 1.2 x 10-5 gm/l.; for lead sulphide at 25° in pure water and in H,S-saturated water the values were 3 x 10-4 and 1.5 x 104 gm/1. respectively ; RaD was used as a tracer for lead. 3* Originally published in Z. phys. Chem. A 171, 41 (1934) 2. PLATINUM BLACK G. Hevesy and T. Somya From the Institute of Physical Chemistry, University of Freiburg For the preparation of platinum black, which is used in hydrogen electrodes and for other purposes, the electrolysis of hydrochloric acid solutions of platinum containing lead acetate is employed. The question then arose as to whether the presence of lead in the solution is essential to the preparation of good platinum black and, if so, as to the part played by the lead. To obtain an answer to the first question, we have electrolysed both hydrochloric acid solutions containing only platinum chloride and others containing also small quantities of lead. It was shown that platinum black cannot be obtained successfully by the electrolysis of a solution containing only platinum. On the contrary, a grey or light brown deposit is always obtained. On the other hand, the preparation of platinum black is accomplished from solutions which contain a corresponding amount of other heavy metals in place of lead. After this observation we proceeded to study whether lead is carried into the deposit when a solution containing lead is electrolysed and, if so, in what amount and form. DETERMINATION OF THE LEAD CONTENT OF PLATINUM BLACK Since the detection of small quantities of lead in platinum is very tedious we have made use of a radioactive tracer method. A known amount of lead acetate labelled with thorium-B was added to the platinum chloride solution and the lead content of the platinum black deposited on platinum electrodes having a surface area of 15.07 cm?, at a current density of 10 mA/cm?, was determined by measuring the intensity of the a-radiation emitted by the deposit. The amount of lead was calculated from this intensity measurement as follows: From the same radioactive lead acetate solution, of which a known volume was added to the platinum chloride solution, lead peroxide was precipitated (see below) after adding nitric acid and a further amount of inactive lead acetate and the a-radiation of this precipitate was compared with PLATINUM BLACK Sil that of the platinum black. Now if care is taken that the thickness of the deposits attains the range of the q-radiation in the material. which is 12.8 uw in lead and 30.6 w in lead peroxide, then the activity will provide a simple measure of the lead content. Denoting the activity of the platinum black electrode by S,, that of the lead peroxide electrode of the same size by S,, the density of platinum (21.3) by d,, the density of lead peroxide (8.9) by dy, the range of a-radiation in platinum (12.8 jv) by R,, the range of q-radiation in lead peroxide (30.6 “) by R,, the number of grammes of lead in 1 em? of the active lead acetate solution by p and the number of grammes of lead in the 2.5% lead acetate solution by P, then the required lead content of the platinum black, 2, expressed as a percentage, is given by a = S,pRd (at. wt. of Pb)100/S,pPR,d,(mol. wt. of PbO,) Therefore p was chosen to differ from P because it had been found preferable to produce the lead peroxide deposit from a solution with a lead content higher than that for the platinum black coating. The lead content of the platinum black, as determined, is found in the Figures in Table 1. It may be seen that the lead content of the platinum black rises sharply with increasing lead acetate content of the platinum chloride solution. TaBLE 1. — Leap ConrTEentT or Pratinum Buack as A FUNCTION oF THE Leap ACETATE CONTENT OF THE ELECTROLYSED 0.2 N HCl SoLtution Contarnine 3% PtCl, Lead Acetate Content Lead Content of the of the Solution Platinum Black (%) (%) 22 0.035 1.34 O.815 1.44 1.5 129 Tiel In order to decide whether the lead found in platinum black is present in solid solution or not we have compared the line distances obtained on Debye-Scherrer diagrams for different platinum black samples with those of pure platinum. The difference, as is evident from Table 2, was shown to be vanishingly small and therefore it must be assumed that the large majority of the lead occurring in platinum black is not present in the form of a solid solution. The measured line distances underwent a considerable increase when the sample was heated. Thus, the second sample recorded in Table 2 showed, after heating for 16 hr in a vacuum to 500°C, a line distance of 124.7 mm; after 44 hr heating at 625 the distance was 125.2 mm. 38 ADVENTURES IN RADIOISOTOPE RESEARCH TaBLE 2. — Link DIsTANCE OF THE (422) INTERFERENCE OF PLATINUM Buiack Lead content of the | Line distance before | Line distance after platinum heating | heating in vacuum (%) (mim) | (mm) 0 124.3 124.6 (500°) 1.5 124.4 125.2 (625°) TA 124.1 125.7 (625°) Heating to still higher temperatures resulted in a considerable evapor- ation of lead, as shown in Table 3. With regard to the values in Table 2, it should be mentioned that it was not possible to assess the line distance in the case of platinum grey (lead-free platinum deposit) with sufficient accuracy. The number 124.3 in the second column thus refers to platinum wire whereas the corresponding value in the third column was indeed obtained from platinum grey. After heating, the platinum grey did of course yield lines of adequate sharpness. The exposures for the Debye- Scherrer diagrams were obtained with the aid of the precision camera described by Sacus and Weemrts!; platinum wires 4 mm thick and coated with platinum black were used for the exposures. A Metalix tube with a copper anti-cathode was used as the source of radiation and was operated for as long as 13 hr at 45kV and 20 mA. The lead content present as a solid solution was calculated by means of Vegard’s law, according to which the lattice constant of the solid solution is a = (3.905c, + 4.93¢,)/100 where a is the length of the side of the unit cell of the alloy, c, is the number of atoms per cent of platinum and cy, for lead. The calculation showed that, of a total of 1.5 per cent lead, only 0.2 per cent was present in solid solution after heating (Table 2) and of the 7.1 per cent lead in another sample only 0.3 per cent was similarly in solid solution. It is hoped to study in more detail, by means of radiographs of platinum black containing thorium-B, the distribution of lead in platinum. THE EFFECT OF HEATING ON THE LEAD CONTENT OF PLATINUM BLACK It has already been mentioned that considerable amounts of lead evaporated when platinum black was heated at higher temperatures. Ina more detailed study of this point the alpha activity of platinum black, 1Sacus and Wererrs, Z. Phys. 60, 481 (19380). PLATINUM BLACK 39 obtained by the electrolysis of solutions containing lead acetate, labelled with thorium B, was determined before and after heating. The results of this experiment are seen in Table 3. TaBLE 3. — Errect or HEATING FoR 16 HR IN A VACUUM ON THE Lrap- ConTENT OF PLatTINUM BLACK ORIGINALLY ConTAINING 1.59 Leap Temperature (°C) 600—6 10 630—650 715—725 Loss of lead caleulated from the bo oh! 0) wo Zw or decrease in a-activity (%) The loss of lead is not entirely due to evaporation but partly also to diffusion of the lead contained in the platinum black into the platinum foil on which the coating was deposited. Differentiation between the loss by evaporation and diffusion is possible by making use of the y-radiation instead of the a-radiat:on for making the comparison; whereas the amount of lead removed by diffusion weakens the q-radiation to the same extent as does the lead disappearing through evaporation, this is is not the case when the y-activity is measured. For example, the decrease in y-radiation after 16 hr heating at 685 to 700°C amounted to only 42 per cent and thus considerably less than the decrease in a-radiation. THE QUALITY OF THE VARIOUS SAMPLES OF PLATINUM BLACK An attempt was next made to measure the easily traced adsorption of thorium B and thorium C from solutions of these radio-elements with a view to assessing the quality of the platinum black. Yet great difficulty was encountered in obtaining reproducible results. This method was therefore relinquished. It was then thought that a simple measure of the quality of the various platinum black coatings could be obtained by preparing hydrogen electrodes from the various platinum black samples and comparing their potentials. It was shown, however, that the potential of all the hydrogen electrodes prepared in this way was always the same within the experimental error of about 1 mV. We then changed over to determining how strongly the various samples of platinum black could be polarized with the same cathodic loading and to making use of the difference in polarizability as a measure of the quality of platinum black. The polarization was performed in N sulphuric acid solution with a current density of 20 mA/em? at room temperature for a period of 45 min; the area of one side of the electrode amounted to 1 em?. The polariz- 40 ADVENTURES IN RADIOISOTOPE RESEARCH ation potential was measured by using a normal hydrogen electrode. A constant value of the polarization potential was established after about 30 min. The result of the measurements is evident in Table 4. TaBLeE 4. — CatrHopic POLARIZABILITY OF PLATINUM DEPOSITS OBTAINED FROM SOLUTIONS WITH DIFFERENT LEAD CoNTENTS. POLARIZATION CURRENT Density 20 mA/cm? Lead content of the electrode (%) ......- | ‘fall 1.5 0.15 0.035 0 Polarization potential (MM Woooos ge desoaa05 84.4 77.5 $1.0 Clee) 103.7 The least polarizable and, therefore, the one of highest quality is platinum black with a lead content of 1.5 per cent, and it is interesting to notice that this sample of platinum black is identical with that obtained by electrolysing a solution containing 1 part of platinum chloride and 0.008 parts of lead acetate in 30 parts of water, and that LumMER and KuriBauM a long time ago used the electrolysis of a solution of this composition for preparing platinum black. This set of directions is also included in the Teatbook of Practical Physics by KonLRAUSCH and other similar works. Attempts were then made to heat the electrodes before they were polarized. In all instances the heating spoiled the quality of the platinum black. After heating the electrode containing 1.5 per cent lead for 16 hr at about 610°C the polarization potential rose from 77.5 to 88.3 mV, and after 16 hr heating at about 700° it became 183 mV. Changing the current density from 10 to 30mA/cm? when preparing the platinum black had no detectable effect on the quality. CONNEXION BETWEEN THE PARTICLE SIZE AND QUALITY OF PLATINUM BLACK The particle size of the platinum black was determined from the half breadth of the X-ray lines in accordance with Briu’s method?. The Debye-Scherrer camera used for this purpose had a diameter of 5.73 em. The diameter of the platinum wire covered with platinum black was 0.34 mm. Lines of the (220) and (311) faces were used for the investigation. The results of this study are shown by the data in Table 5. 1R. Brity, Kolloid Z. 55, 164 (1931); Z. Krist. 74, (147 (1930). PLATINUM BLACK 4] TaBLE 5. — DEPENDENCE OF THE PARTICLE S1ZE OF PratTinum Brack on irs Leap ContTENT Particle size ( A) Lead content of jae —- platinum (%) | Caleulated from oC te at ate a from | the (220) line ae the (311) line 7A | 62 | 64 1.5 81 | 75 0.15 68 68 0.035 58 0 61 57 It is evident from this table that the platinum black sample which has been found to have the best quality is distinguished by having the largest particle size. PREPARATION OF PLATINUM BLACK FROM SOLUTIONS CONTAINING GOLD It is evident from Table 6 that platinum black of good quality was prepared also from platinum chloride solutions which contained gold instead of lead and was deposited at a current of 30 mA/cm?. TABLE 6. — CatHopic POLARIZABILITY OF PLATINUM Deposits PREPARED FROM SoututTions Havine Various Gotp Contents. POLARIZATION C urRENT Density 20 mA/cm?. ELrectrotyTE 1 N H,SO, Platinum content of the | sole 5 me | Polarizati Sen OST Gold EOE ts) | olariz site ; the solution potential platinum black was ob- ry, y tained (%) Ce) | ae | 0.14 0.9 76.9 1.8 0.1 78.7 1.8 0.01 Wd 1.8 0.0001 86.4 We have also prepared platinum black from platinum chloride solutions which contained thallium, cadmium or zine in place of lead. Whereas thallium can substantially replace lead as far as the appearance of the deposit is concerned, the behaviour with cadmium is different inasmuch as a solution which contained about 0.02 per cent cadmium chloride yielded a fine black deposit while the electrolysis of solutions which contained only about 0.01 per cent cadmium chloride yielded a grey deposit instead of platinum black. Electrolysis of platinum chloride solutions containing zinc yielded a grey deposit in all cases. We then 42 ADVENTURES IN RADIOISOTOPE RESEARCH attempted to electrolyse hydrochloric acid solutions of pure platinum chloride at a high current density, i. e. at 100 mA/cm? and above. With this heavy loading it was no longer possible to obtain an adherent deposit. The deposited platinum powder fell into the solution, had a black-grey appearance and was extraordinarily fine grained. Summary The platinum black obtained from platinum chloride solutions containing lead in accordance with the directions of LUMMER and KuriBaumM, contains con- siderable amounts (1.5 per cent) of lead. Variation of the lead content of the platinum black with the lead content of the solution subjected to electrolysis was observed. The electrolysis of a solution containing 1.9 per cent of lead acetate yields a platinum black which contains 7 per cent of lead. By assuming the validity of VEGARD’s additive law for the lattice dimensions it is found that the greater part of the lead present in platinum black does not occur in solid solution. Of the various samples of platinum black the one prepared in accordance with the directions of LumMER and KuriBpaum was the least electrolytically polarizable and thus the most perfect. A determination of the particle size of the platinum black samples by means of the half-width of the X-ray interferences resulted in the fact that the best platinum black sample had the largest particle size. Platinum black was also prepared from platinum chloride solutions which contained other added metals instead of lead. Originally communicated in Nature, 128, 1038 (1931). 3. LEAD CONTENT OF ROCKS G. Hevesy and R. Hossre From the Institute of Physical Chemistry of the University of Freiburg In recent years various geochemical problems have arisen which make it important that our scanty knowledge of the lead content of rocks should be amplified and made more precise. To this end we have deter- mined the lead in a series of samples, representing in all about 220 rocks, some of which we owe to the kindness of Prof. ArtrHurR HotmEs of the University of Durham. The sample to be analysed was brought into solution ; silver sulphate was added and the silver and lead present in the solution were simultane- ously precipitated as sulphide. The precipitate was them brought into solution and the lead deposited electrolytically as peroxide. That the deposit was actually lead peroxide was confirmed by a colorimetric test, tetramethyl-diamino-diphenylmethane being added to the solution of the deposit. To ascertain that the total lead content was actually recovered, we used the method of radioactive indicators. We added to the rock sample a known amount of the lead isotope radium D, prepared from radium emanation, and checked the yield obtained by measuring the activity of the lead peroxide deposit. As the purest chemicals com- mercially obtainable were found to contain quantities of lead that would have influenced our results, all the chemicals used were first purified from lead. Moreover, every precaution was taken to avoid contamination of the samples by dust which might have contained traces of lead. The results obtained are listed in Table 1. The average value found is 16x 10-® gm lead per gm rock, a some- what larger value than that given by CLARKE and StrercErR! who found 7.5X10-§ gm per gm rock. As shown in the communication that follows, the amount of lead accumulated in the rocks since the solidifica- tion of the earth’s crust (as a result of the decay of uranium and thorium) is very much smaller. Thus, as between the atomic weights of rock- lead and ore-lead we have in most cases to expect differences only in the 1 CLARKE and STEIGER, J. Wash. Acad. Sci. 4, 58 (1914). 44 ADVENTURES IN RADIOISOTOPE RESEARCH Taste I. — Leap Content or Igneous Rocks a U { gm Lead per Rock Types - : gm Rock Basalt; GiantisuCauseway «cerca ssssoscse++s ses A SC lime Gabbros and related types (composite of 67 samples) oD Ome Essexites and related types (composite of 40 samples) 10 A0me Shonkinites (average of 2 samples) .............. 105 10m s Soda-granites and soda-syenites (composite of 26 | SAMPLES! cnevere) coy cre wer s euensieeyai cticusteneneleteae as toncue ores « SS Ome Potash-granites and potash-syenites (composite of 24 Cal IES) Gogacochocoosoouuooocbo6g dono oGacude 4" 1058 Amphibolite, inclusion in Kimberlite, Wesselton Manes Sct Atri Catt ey nteus eee, otra teu ioe tate ae merci eeoioie 150m Kimberlite (“ basaltic’ type). Dyke from 1350-foot level- Dutoitspanes Mine crm clic ce ieiecreieeeiae | a SC UO Lherzolite, Baltimore, Maryland ................ i LO Noppack, Die Naturwiss. 18, 761 (1930). LEAD CONTENT OF ROCKS 45 Taste IT. — Leap Content or Basic AND Utrrasasic Rocks Aanp or METEORITES gm per gm Rock Gabbrosm(aversize) merece = 0 f-rcrereienser 5 x 107-& Kamilberlitieperctterer us svarseys/erta oo cic ies fy s sy ee eherzo lute mrcey reer an sec eee cuneate 19 « 10-6 Stony meteorites (average) ......... 5 SANE Iron meteorites (average) .......... Ve Ob eal Oe share of the total lead available for partition, and that this uneven distribution has so far been compensated only slightly by the formation of lead from radioactive decay. 46 ADVENTURES IN RADIOISOTOPE RESEARCH COMMENT ON PAPERS I, 2, 3 RaDIoactTive tracers are very often applied in the solution of problems which can only be solved by making use of this device. Often this method is, however, applied not as a necessity but for the sake of convenience, for facilitating the solution of problems which can be solved by other methods as well, though more tediously. An example of the latter is the determination of the lead content of platinum black described in paper 2. By making use of labelled lead the analysis and the behaviour of platinum black under the effect of heat, of electrolytic polarization and so on could be carried out without even dissolving the sample The determination of the solubility of lead sulphide and chromate in water discussed in papers is a border case. The writer of the well-known text-book on physical measurements, Kohlrausch, succeeded in calculating the solubility of lead carbonate from the electrolytic conductivity data, but he arrived in the case of lead chromate at a rough estimate of its solubility only. In the determination of the lead content of rocks by making use of isotope dilution control an isotope of lead is necessarily to be used. This investigation was the first application of the isotope dilution method. Being interested in the abundance of the elements, we determined various constituents of an artificia air crust sample prepared by my colleague, the well-known mineralogist Schneider- héhn, who disposed over a very extended collection of rock and mineral samples. We usually applied the method of analysis of X-ray spectroscopy using a secondary X-radiation method worked out in the Freiburg laboratory in which these in- vestigations were carried out. Knowing the intensity ratio of two very closely situated X-ray lines, for example that of HfL, and Luf,, by adding to a pulverized sample to be analysed a known amount of LuO, we can, after taking an X-ray spectrum, calculate from the intensity ratio of the 2 above-mentioned lines the unknown hafnium content of the sample. When the X-ray spectrum is excited by cathode rays the sample gets hot and sputters easily ; for this reason a continuous X-ray spectrum is first produced on a metallic tungsten surface and the sample irradiated by this continuous réntgen radiation. Heating of the sample and consequent sputtering is now avoided. The application of the method in the above-mentioned case assumes the absence of significant amounts of the rare lutetium in the sample investigated. The sensitivity of this method (in the twenties of this century, when we applied it) did not suffice to determine the very small amounts of lead present in rock samples and correspondingly we embarked on a chemical determination of lead controlled by making use of isotope dilution. This device, discussed further on p- 96, proved to be a very useful one in chemical analysis. . Originally published in Kgl. Danske Videnskabernes Selskab. Mathematisk-fysiske Meddelelser. 14, 5 (1936) 4. THE ACTION OF NEUTRONS ON THE RARE EARTH ELEMENTS G. Hevesy and Hinpe Levi From the Institute of Theoretical Physics, University of Copenhagen THE action of neutrons on the rare earth elements can be followed up in two ways: by investigating the radioactivity induced in these elements under neutron bombardment, and by observing their absorbing power for a beam of slow neutrons. In this paper both these lines of attack will be discussed for the rare earth group and for yttrium and scandium. ARTIFICIAL RADIOACTIVITY OF THE RARE EARTH ELEMENTS The artificial radioactivity of some of the rare earth elements was investigated by AManpi, D’AGostino, Fermi, Pontrecorvo, Raserri and SEGRE (1), others were investigated by ourselves (2) by SuGDEN (3) by Marsu and Suepen (4) by McLennan and Rann (5) and by E. Rona (6). The neutrons used were produced by the action on beryllium of the a-rays from radium emanation and were in many cases slowed down by inserting layers of paraffin 10—20 em thick in the path of neutrons ; a GEIGER—M@ULLER counter was used to measure the activi- ties obtained. Scandium A sample of scandium oxide prepared by Prof. Srmrpa-B6HM and kindly presented to us by Prof. H6n1ascuMrp, who used the preparation in determining the atomic weight of scandium, was activated for a few days using an emanation-beryllium source of 200—300 MC. The oxide was then dissolved in dilute hydrochloric acid and 100—150 mgm sodium chloride as a carrier of 47K (cf. p. 48) and the same amount of calcium oxide were added. The filtrate obtained after precipitation with carbon- ate-free ammonia was treated with oxalic acid and the calcium oxalate- formed was removed. The sodium chloride which had been added was recovered, after the removal of the ammonium chloride content of the last filtrate, by evaporation and ignition. The activities of the three 48 ADVENTURES IN RADIOISOTOPE RESEARCH fractions, those of scandium oxide, sodium chloride, and calcium oxalate, were then determined. The two first mentioned preparations were found to be active, the activity of the scandium oxide decaying very slowly and that of the sodium chloride fraction having a half-life of 10 to 16 hours. The activities are due to the formation of 4§Sc and 221K respectively ; the reactions leading to these products are $3Sc + on = 21Se and 228C ote an = ‘fla + sa ; The mass numbers occuring in these equations follow from the fact that scandium has only one stable isotope, “Sc. The calcium oxalate investigated was inactive; we are thus unable to find any evidence for the reaction $3Se + 3n = 39Ca + tH which possibly takes place also. The activity which cannot be separated from scandium is presumably due to 48Sc; most of this activity decays with a period of about two months While “2K emits hard f-rays having a half value thickness of 0.19 gm/em? Al, 3{Se emits soft p-rays with a half value thickness of 0.01 gm/ em? Al. Yttrium We investigated (2) samples of yttrium oxide kindly given us by the late Baron AvER v. WELSBACH, by Prof. Pranpti, and by Prof. Roi. The two first named preparations were used some time ago by Honic- scumip to determine the atomic weight of yttrium and investigated by one of us on that occasion by X-ray spectroscopy. While the investi- gation of Baron AvsER’s preparation revealed the presence of some dysprosium, that of PranpTL was found to be of the highest purity. The great purity of this preparation and of that of Rota was also shown by their behaviour under neutron bombardment : No initial decay with the period of dysprosium (2.5 h.) could be observed, the sole period being one of 70 h., which we found to be the period of decay of yttrium. Aver’s preparation decayed initially with a half-life of 2.5 h., which was obviously that of dysprosium ; but afterwards it showed a 70 h. period like the other preparations. The molecular volumes of corresponding compounds of yttrium and dysprosium are only very slightly different*, so these elements are unusually closely related chemi- * The volumes of the octahydrosulfates differ by less than 0.8% (G. v. HEVEsy, Z. anorg. Ch. 147, 217 ; 150, 68 (1925) and the ionic radii by about the same amount (V. M. Gortpscrmipt, Uniricw and BartH, Oslo. Acad. Proc. Nr. 5 (1925). THE ACTION OF NEUTRONS ON THE RARE EARTH ELEMENTS 49 5 10 days 5 10 days Fig. 1 a) Decay Curve of a Pure and an Impure Yttrium Preparation. ’ b) Samarium Decay Curves (the two days’ period only; the weak period of 40 min is not visible). cally and their separation is attended with very great difficulties. Figure 1 shows the decay of a pure preparation and of one containing some dysprosium. Since yttrium has only one stable isotope, 8°Y, the artificial radio- activity obtained from it is presumably due to the formation {9Y. We find the intensity of the yttrium activity to be 0.005 timesas large as that of dysprosium, both preparations having been activated until saturation was obtained in a paraffin block of 30x 30x25 em edge ; the neutron source was placed on the top of the preparation which was covered by a thin shield of paper. The f-rays emitted by yttrium are absorbed to half of their initial value by 0.6 mm Al. Lanthanum Marsu and Suapen (4) find 1.9 days as the half-life of lanthanum and for the intensity of the f-rays emitted, a value amounting to 35% of that observed for the activity of praseodymium. As we find a value of 22 for the ratio of the radiation intensities of dysprosium and praseody- mium, the lanthanum activity works out at 2.0% of that of dysprosium. 4 Hevesy 50 ADVENTURES IN RADIOISOTOPE RESEARCH As lanthanum has but one stable isotope, La, the activity obtained is presumably due to the formation of 4%La. Fermi’s coefficient a, indicating the increase in activity when the bombarding neutrons are slowed down by a thick layer of paraffin or other hydrogen- containing substances, instead of being allowed to impinge directly from the beryllium source on to the substance to be activated, was found to be 12. Cerium No activity was observed after bombardment of cerium for several days with a neutron source of few hundred millicurie. Praseodymium AMALDI, Fermi, and others (1) found the artificial radioactivity of praseodymium to decay with a 19 h. period, the same value being found later by other experimenters (4), (5). Although only one stable isotope of praseodymium is known. “Pr, the above-mentioned investigators found a second period of decay (5 min) which in contrast to the first period is not hydrogen-sensitive. Neodymium Fermi and his collaborators (1) found that activated neodymium decays with a period of 1 h. ; we find (2) that this activity is 2500 times as small as that of dysprosium. MarsH and SuagpEn (4) found no activity, while according to McLENNAN and Rawnwn (5) the half-life is 35 min. Neodymium has the stable isotopes 142, 143, 144, 145, 146 and 148, and the activity observed is presumably due to the formation and decay of PONG Samarium The artificial radioactivity of samarium decays, as was found by Fermi (1), and later by us, with a period of 40 min. We find (2) the intensity of the activity to be 0.6% of that of dysprosium. Samarium has furthermore, as was first noticed by MarsH and Sua@pEn (4), a much longer period as well. We determined the period of this isotope to be 2 d., as can be seen from Fig. 1 and found its intensity to be <| of that of dysprosium, i.e. 2.0 on our relative scale. Samarium has the stable iso- topes 144, 147, 148, 149, 150, 152, and 154, and it is not possible to deter- mine the mass number of the active samarium isotopes with certainty. THE ACTION OF NEUTRONS ON THE RARE EARTH ELEMENTS 51 A very intense activity was obtained by SUGDEN (3) on bombarding europium with slow neutrons. It decayed with a period of 9.2 h. The intensity of the europium radiation was found by us (2) to be 80% of the dysprosium radiation emitted by the same amount of dysprosium, both preparations being activated until saturation was reached. Care was taken, too, that the neutron beam was weakened only to a small extent by the activation process, i.e. very thin layers were activated. Juropium has two stable isotopes 151 and 153 and the activity is possibly due to the formation of 2Eu. The europium: dysprosium activity ratio is found to be smaller for thick layers than for thin layers. The value 40 was found for the hydrogen-effect, a. The half-value thickness (2) of the B-rays emitted is 0.02 cm Al, and it was concluded from absorption measurements that energies up to 2.0-108 eV occur.* In addition, y-rays have been detected which are little absorbed by 4 mm. lead. Gadolinium Frermr and others (1) found gadolinium to decay with a period of 8 h. after neutron bombardment. McLENNAN and Rann (5) found a half-life of 6.4 h. and twice the intensity found for neodymium. The combination of the last mentioned figure with our intensity data leads to an intensity value which is 250 times as small as that observed for dysprosium. Marsn and SuGpEN (4) could not find any activity. Terbium The activity of terbium (3) decays with a period of 3.9 h. As this ele - ment has only one stable isotope, *g;I'b, the activity observed is presum - ably due to the formation and decay of 16071}. The intensity of the radiation (2) observed is 2.5 per cent of that of dysprosium. Dysprosium The activity of dysprosium (2), (4) decays with a period of 2.5 h. and is the strongest yet observed in the domain of artificial radioactivity. We have therefore chosen it (2) as a standard of comparison for the * R. Narpu and R. E. Srpuy (Proc. Roy. Soc. A 48, 332, 36) by using a cloud chamber determined recently the energies of the p-ray spectra and found that the maximum energy lies at 1.3 - 10® eV, while the upper limit of the spectrum is: 1265-108 eV. 4* iy ADVENTURES IN RADIOISOTOPE RESEARCH activities of the rare earth elements: we denote the intensity of dys- prosium arbitrarily by 100. It is of interest to remark that the 2.3 min activity of silver, which is considered a very strong activity, is 12 times as weak as the activity of an equal amount of dysprosium. The hydrogen effect (a) was found to be 100, the half-value thickness of the f-rays emitted was 0.025 em Al; and the upper limit of the continuous f- spectrum concluded from absorption measurements with aluminium has an energy of 1.4- 108 eV (2).* Dysprosium is one of the commoner rare earth elements of the yttria group and as it is very strongly active, activated samples of rare earth elements denoted as ‘‘erbia’’, ‘“‘holmia’’, ‘“‘yttria’”’, etc. often decay with the period of dysprosium. Holmium We found (2) the activity of holmium to decay with a period of 35 h., while E. Rona (6) recently found the value of 33 h. The half-life of 2.6 h. measured by Marsu and SugpEN (4) and later by McLENNAN and Rann (5) is presumably due to the presence of dysprosium in their preparations ; some of our impure preparations, too, showed an initial decay with the period of dysprosium. The samples of holmia investigated were given us by the late Baron Aur. Holmium has one stable isotope, 165 ; the activity observed is therefore presumably due to the decay of '$Ho, the intensity of the activity observed being 20 per cent of that of dysprosium. The hydrogen-effect (a) is much smaller (2) than that of dysprosium ; the half value thickness is 0.04 kgm/cm? Al ; and the upper limit of the f-ray spectrum has an energy of 1.6 - 106 eV. Erbium Erbium has a very weak activity of similar intensity to the 40 min samarium radiation, decaying with a 7 min period according to MAaRsH and SuGDEN (4), and with a 4.5 min according to McLENNAN and RANN (5). A second period (2) was found by us to be 12 h. ; the period of 2.5 h. ascertained by SUGDEN (3) using a commercial preparation is presum- ably due to the presence of dysprosium, and that found by Marsu and SuGDEN (4), 1.6 d., to the presence of holmium in the sample investi- gated. Recently Rona (6) has given the value of 13 h. for the longer period. The intensity (2) of the longer period of erbium is 0.35 per cent of that of dysprosium, and the half- value thickness of the f-rays emitted is 0.03 em Al. * R. Narpu and R. E. Sipay, loc. cit. found that the maximum energy lies at 0.75 - 108 eV, while 1.8 - 10® eV is the upper limit. THE ACTION OF NEUTRONS ON THE RARE EARTH ELEMENTS 53 This element shows an activity having a long- life as first stated by Rona (6) who finds that the f-rays emitted are half absorbed by 0.015 em Al. E. Neuntncer and E. Rona* found recently a period of 4 505%" ‘on Bei ie Lu 300 30 \ ° . ° . 20 20 \ , eS Oe ee 10 20 days 10 20 30 40 50 100 days Fig. 2. Lutecium Decay Curve. a) showing the short period and the beginning of the 7 days’ period. b) showing the measured points for the 7 days’ period and the curve obtained after subtracting the residual activity. (+ 1%) months. After bombarding 100 mgm TmO,, kindly lent us by Prof. JantscH, with about 100 me for 23 days we obtained 60 counts per minute, while 100 mgm of dysprosium activated to saturation with the same source gave about 4000 counts per minute. The activity of this preparation decayed with a period of about 3.5 months; a thulium preparation activated to saturation would therefore exhibit an activity about 1/,, of that of dysprosium. Ytterbium The activity of ytterbium (2), (3) decays with a period of 3.5 h. As ytterbium has the isotopes 171, 172, 173, 174, and 176, it cannot be decided whether the activity is due to the formation and decay of 33Yb * EK. NeEuNINGER and E. Rona, Wien. Anz. 73, 159 (1936). 54 ADVENTURES IN RADIOISOTOPE RESEARCH or of 73Yb. The ytterbium radiation (2) is somewhat weaker than that of erbium, and amounts to 0.3 per cent of that of dysprosium. The half-value thickness of the f-rays emitted is 0.04 cm Al. Lutecium Lutecium (cassiopeium) exhibits an activity of fairly long life (2), namely one decaying with a period of 6—7 d., and having an intensity of 1.4 per cent of that of dysprosium; there is a second activity of somewhat less intensity decaying with a period (2) of 4 h.; as lutecium (cassiopeium) and ytterbium are very closely related elements, and lutecium being usually contaminated with ytterbium, we considered it possible that the 4 h. period observed might be due to the presence of ytterbium in the sample investigated. A very pure lutecium (cassiopeium) preparation, however, prepared by AveER and kindly lent us by Prof. HOntescuMip, also showed the 4 h. period. Furthermore the intensity of this radiation was stronger than that emitted by a pure ytterbium preparation activated by a neutron source of the same strength. So we must conclude that both the periods observed are due to lutecium. The long period of decay has not been observed by any experimenter besides us, presumably because the times of exposure have been too short. For the shorter period McLenNAN and Rann (5) give a value of 3.6 h. and Rona (6) 4—5 h. The decay of the lutecium preparation lent us by Prof. H6niascumip is seen in Fig. 2, the time of exposure being 2.8 days. In comparing the intensities of the long and the short periods the former must be divided by 0.267 which value follows from a consideration of the relation J, =J, (l1—e~’**), where J. is the saturation value of the activity. J, the value obtained after t days, and 2 the decay constant (= half-life/In 2). As can be seen from the Fig. 2b a third long period is present in the activated lutecium which is possibly due to the presence of small amounts of thulium. ABSORPTION OF SLOW NEUTRONS BY RARE EARTH ELEMENTS Determination ef the period of decay from absorption data When faced with the problem of determining the period of very slowly decaying radioactive isotopes having half-lives of several months or years, decay measurements become very tedious. In such a case we can obtain information about the decay constant required by comparing the absorption of slow neutrons in the rare earth element in question with that in another rare earth element of known period. A knowledge of this ratio and of the activities obtained for both elements after a known l THE ACTION OF NEUTRONS ON THE RARE EARTH ELEMENTS time of exposure allows us to calculate the unknown period of decay provided we can assume that all the neutrons absorbed are captured by the nuclei of the absorbing element and that mainly thermal neutrons are involved in both cases. In the oxides investigated, only the nuclei of the rare earth element absorb, for oxygen nuclei capture only a small number of neutrons. Let us consider, for example, the case of scandium. Denote by R, the observed absorption ratio for equal numbers of scan- dium and dysprosium atoms, and by R, the ratio of the activities obtained renee _N-R,: 0.69 after an exposure of N days; then the half-life of scandium is : s D) Lis days. We compared the activity of 66 mgm of scandium and 100 mgm of dysprosium and found after an activation of 24 days an activity ratio of 0.9210. During this activation time, full saturation of the dysprosium activity was obtained, while the scandium was far from being saturated. For equal numbers of scandium and dysprosium atoms we found an intensity ratio of 0.40 x 107°. To compare the absorbing powers of scandium and dysprosium we inserted in the path of the neutron beam, which had been slowed down in the usual way by a block of paraffin, first, a layer of scandia (590 mgm/cm? Sc) and then a layer of dysprosia (840 mgm/cm? Dy) and measured the activation of a rhodium foil in the absence and then in the presence of the absorbing layer. The amounts of the absorbing material necessary to reduce the activity of rhodium in each case to 90% of its initial value were calculated to be 300 mgm/cm? Sc and 43 mgm/cm? Dy. A more satis- factory way to proceed in comparing the absorbing powers would have been to have used a dysprosium indicator to measure the absorption in dysprosium and a scandium indicator to measure the absorption in scandium, but the small activation of scandium after a few days’ expo- sure to neutrons rendered this infeasible. We have, however, applied the last mentioned method to compare the absorption of neutrons in dysprosium, europium, and holmium, as discussed in the next section. The comparison of the absorbing powers of equal numbers of atoms of dysprosium and scandium led to the result that the former absorbed 25 times as strongly as the latter. It follows from this result and from the comparison of the activities of the two elements, that the half-life of 3iSe is about two months. A similar value was obtained by decay measurements. Strongly absorbing rare earth isotopes forming stable products The unusually strong activities of some rare-earth nuclei are to be ascribed to the existence of strong nuclear resonance levels in the nuclei in question, these levels corresponding to energies of slow neutrons abundant in the neutron beam passing through them, and also to the fact 56 ADVENTURES IN RADIOISOTOPE RESEARCH TaBLE 1. — ABSORPTION OF SLOW NEUTRONS IN RARE EartH ELEMENTS (Amount necessary to reduce the activity of the indicator by ten per cent) Element Indicator mgm/cm? | = == : Europium | Europium 13 Dysprosium | Dysprosium | 40 Holmium | Holmium | 120 that the isotope formed by the capture process is not a stable one already known but an active one hitherto unknown. It is a matter of experience that a mass number cannot be occupied both by a stable and an active isotope of the same element, so that should the mass number 166 be occupied by a stable dysprosium isotope the high capturing power fcr slow neutrons shown by '32Dy would not lead to an active but to a stable dysprosium isotope. The appearance of a strong activity shows that at least one isotope of this element captures neutrons strongly, but high absorption does not necessarily imply strong activity because nuclei yielding stable isotopes can also be very strong absorbers of neu- trons. To obtain information about the existence of strongly capturing rare-earth nuclei not leading to the formation of radioactive products, we compared the activities of dysprosium, europium, and holmium with their absorbing powers for the same neutron beam as was used to activate them. The results of these measurements, in which the absorbing el- ement itself was used as indicator, are seen in Table 2. TaBLE 2. — ABSORPTION OF Stow NEvuTRONS IN RarRE EartH ELEMENTS (Amount necessary to reduce the activity of the indicator by ten per cent) Element | Indicator mgm/cm* ca —=— e Eee See Europium Rhodium | 16 Dysprosium Rhodium | 43 Holmium | Rhodium | 160 Gadolinium | Rhodium | 2 Samarium | Rhodium 12 Yttrium ~ Rhodium 500 Scandium Rhodium | 300 Cadmium | Rhodium | 18 While the activity of europium is slightly smaller than that of dyspro- sium its absorbing power is more than twice as big; europium thus absorbs slow neutrons to an appreciably larger extent than is to be expected from the activity of the radioactive europium isotope formed. To explain this discrepancy we have to assume that besides the THE ACTION OF NEUTRONS ON THE RARE EARTH ELEMENTS 57 9 h. period a second period, long and therefore not observed, is present. Samarium also shows an absorption stronger than is to be expec- ted from the activity of the known radiactive samarium isotope. Of the numerous isotopes of samarium not leading to the formation of active isotopes at least one must therefore have a strong resonance level for slow neutrons. In view of the fairly weak activity of samarium the absorption measurements could not be carried out by using a samarium indicator, so rhodium was used for that purpose. The results of these measurements and also of absorption measurements with other rare earths using rhodium as indicator are shown in table 3. TaspLtE 3. — PERCENTAGE OF InrvT1aAL INTENSITY OF THE NEUTRON BEAM PRESENT AFTER THE PASSAGE OF A ‘‘THICK’’? LAYER Element | mgm/em? Intensities 1 s id a Samarium | 580 28% Gadolinium 120 33% Dysprosium | 610 | 40% SS | Sisal gies Cadmium i 390 43% It is well-known that the activity obtained by the action of slow neutrons is not a trustworthy measure of the intensity of the neutron beam, because the neutron absorbing powers of different elements are very specific and depend very much on the neutron velocities. The ambi- guity arising from this fact can, however, be avoided by using the same element as indicator and absorber in absorption experiments. Should that not be feasible, as would happen if, for example, the absorbing substance did not show any or had only a very slight activity — this is the case with gadolinium — it is advisable to adopt the following pro- cedure. The maximum absorption obtained in a thick layer of gadolinium is measured using, say, rhodium as indicator; then the thick layer is replaced by a few milligrams of material and the absorbing power measured again. The first mentioned measurement gives the result that no more than 67% absorption can be obtained for the neutron beam in question through a thick layer of gadolinium, while the last mentioned measurement shows that 2mgm of gadolinium are necessary to reduce the intensity of the neutron beam by 10%. To arrive at a figure giving the amount of gadolinium necessary to reduce the intensity of neutrons of such velocities as are actually absorbed in gadolinium we must mul- tiply 2 mgm by 0.67 and thus obtain a value of 1.3 mgm. The correspond- ing figures for a few elements are given in Table 4 and 5. Of all the rare earth elements gadolinium —as can be seen from the table — has the 58 ADVENTURES IN RADIOISOTOPE RESEARCH TaBLeE 4. — ABSORPTION OF Stow NeEurrons In Rare Earru Elements (Amount necessary to reduce the activity of the indicator by ten per cent of that observed after passage of the neutrons through a ”’thick”’ layer) Element Indicator | mgm/cem? ae: oon eres ete! i Samarium Rhodium 10 Gadolinium Rhodium 1.6 Dysprosium Rhodium 30 Cadmium | Rhodium | 12 highest absorbing power; it is indeed, as has already been shown by DunninG, PreGRAM, Finx, and MircHety (8), the strongest known absorber of slow neutrons. In view of the very strong absorbing power of gadolinium great care must be taken in interpreting the results of absorption measurements on rare earth preparations which might con- tain traces of gadolinium. The presence of less than 1/, per cent of gadolinium in erbium, for example, would suffice to explain the whole absorption shown by erbium. As europium is often contaminated with gadolinium we used various preparations of europium to compare the absorption in europium and dysprosium. One of the preparations was kindly given us by Prof. Pranpri and was entirely free of gadolinium ; it gave a value only slightly lower than the other specimens investi- gated. The high values found by different observers for the absorbing power of yttrium are clearly due to the presence of impurities in the prepar- ations used. According to AMaupt1 and his collegues (1) the absorbing power of yttrium is 70 per cent of that of cadmium, and DuNnNtnG, PraraM, Fixx, and MrrcHE.u (8) give 39 per cent ; whereas using very pure preparations as described on page 48 we find that yttrium is a very Taste 5.— THe Revative AcTIVITIES OF THE RARE EartH ELEMENTS | Element Relative Element Relative Bonmbarded Intensity Bombarded Intensity Vaniaiboo co ogooos | 0.5 Merbvwm) <\..1<- - 2.5 Lanthanum .... 2 Dysprosium ... 100 Gwin “Soocscoc — olmium’ <>. 20 Praeseodymium . 4.5 IDigoyqbued, Saggan | 0.35 Neodymium .... 0.04 AMayvibysbay Sogoe | 12 Samarium ...... | 0.6 Ytterbium .... | 0.25 Huropium) =. -).... 80 Lutecium ..... 1.4.3 1 Gadolinium ..... | very low THE ACTION OF NEUTRONS ON THE RARE EARTH ELEMENTS 59 poor absorber, its absorbing power being only 4 per cent of that of cadmium and 0.3 per cent of that of gadolinium, if the absorption of equal numbers of atoms of the different elements are compared. In Table 6 are given the relative intensities of the activities produced in the rare earth elements by neutrons that have been slowed down by large amounts of paraffin wax. We are still investigating the intensities obtained under the action of fast and semi-fast neutrons and the possible existence of resonance levels. Comparison between the effect of neutrons on rare earth elements and other elements As is shown in this paper numerous radioactive isotopes of the elements of the rare earth group are formed under the action of neutrons, a result which was to be expected from the known existence of a large number of stable isotopes of these elements. Thus the reactions of neutrons with the rare earth elements show the same typical features as their reactions with elements of lower and higher atomic number. The most remarkable feature is perhaps the comparatively frequent occurrence of pronounced resonance phenomena, which phenomena are much commoner among the rare earth elements than in any other part of the periodic system. This fact may be considered as a simple consequence of Bonr’s theoretical considerations on neutron capture, since it would be expected that the distribution of resonance levels would be an especially close one in this region. In fact the product of the number of nuclear particles multiplied by the binding energy of a neutron in the nucleus reaches a maximum in the domain of the rare earth nuclei on account of the circumstance that the binding energy for higher particle numbers decreases considerably until — for the natural radioactive bodies — it has fallen to about half its maximum value.The more frequent occurrence of resonance capture in processes leading to the formation of stable isotopes, than in those giving radioactive isotopes is also in conformity with general experience and is easily explained by the theoretical con- siderations mentioned above since the distribution of levels will be much closer in the former case on account of the fact that the binding energy is considerably larger in processes of this kind than in those leading to the production of unstable isotopes. The use of neutrons in analytical chemistry The usual chemical methods of analysis fail, as is well-known, for most of the rare earth elements and have to be replaced by spectroscopic, X-ray, or magnetic methods. These methods can now be supplemen- 60 ADVENTURES IN RADIOISOTOPE RESEARCH ted by the application of neutrons to analytical problems by making use both of the artificial radioactivity and of the great absorbing power of some of the rare earth elements for slow neutrons. Qualitative analysis with the aid of artificial radioactivity is based on the determination of periods of decay. All rare earth elements have half-lives varying from a few minutes to a few month, so they can all be measured conveniently. The period of decay of 2.5 h., for example, is completely characteristic of dysprosium and is an unambiguous indication of its presence in the sample investigated ; as little as 0.1 mg can be determined without difficulty. We used the method of artificial radio- activity to determine the dysprosium content of yttrium preparations. The procedure was the following: we mixed 0.1%, 1% ete. of dyspro- sium with neodymium oxide, the latter being chosen because it is one of the cheapest rare earth elements, having a low neutron absorbing power as has yttrium, and determined the intensity obtained. The yttri- um sample to be investigated was then activated under exactly the same conditions, and a comparison of the dysprosium activities obtained gave 1% as the dysprosium content of the yttrium sample. Another very beautiful analytical method is based on the very different absorbing powers of the different rare earth elements. A sample, 5 mgm of which spread over 1 cm? absorbed a quarter of the slow neutrons falling on it, could be identified at once as gadolinium, no other element having so high an absorbing power. Unlike the method of artificial radioactivity, the absorption method is limited in its application by the fact that the absorption measure is the sum of the absorptions of the different elements present in the sample. This limitation is, however, largely due to the fact that our knowledge of the absorption of neutrons and still more our devices for producing neutrons of different energies are only inanembryonic state. The absorb- ing powers of different nuclei depend to a high degree on the energy of the neutrons in question and the future development of our knowledge of neutron absorption will presumably make it possible to apply absorption methods of neutron analysis of great simplicity and reliability. This method of analysis, as also that based on periods of decay, gives a direct means of identification of the nuclei involved ; this distinguishes them from all other analytical methods, chemical, spectroscopic, X-ray, and magnetic, which are based on the investigation of the electronic proper- ties of the atom in question. Effect of neutrons on minerals containing rare earth elements Many of the rare earth minerals, because they are products of residual magmatic crystallisation, contain rare earth elements, thorium, and uranium, along with beryllium and other light elements. The last THE ACTION OF NEUTRONS ON THE RARE EARTH ELEMENTS 61 mentioned element is far the most effective neutron source under bombardment with a-particles or with the y-rays emitted by uranium, thorium, and their disintegration products ; the nuclei of other elements, such as lithium, boron, magnesium, aluminium etc. are much less effective.* In minerals containing large amounts of strongly capturing rare earth elements, the neutrons produced in the mineral or in its surroundings are absorbed to a large extent in the element in question. The mineral gadolinite, for example, contains about 50% of rare earths, of which according to GoLpscHMipT and THOMASSEN** up to about 15% is gadolina ; this mineral often contains, too, other light elements including considerable amounts of beryllium, about 0.38% of thorium, and some uranium as well. 1 gm of thorium and its disintegration pro- ducts produces up to 108 neutrons per year or in all 101 neutrons since the formation of the minerals. If these neutrons are all absorbed in 1 kgm of the mineral in question and are absorbed primarily by the gado- linium content, 10!6 gadolinium atoms will be formed having an atomic weight one unit higher than before the absorption. As 1 kgm gadolinite contains about 10” of gadolinium atoms the equivalent weight of gadoli- nium will increase during that long span of time by but one unit in the fourth decimal place. While this result is only a very rough estimate it suffices to demonstrate that some of the rare earth elements which primarily form higher stable isotopes by capturing neutrons, increase in equivalent weight as time proceeds. Dysprosium on the other hand when decaying forms holmium, holmium forms erbium etc.; the process in such cases leads to an increase in the amounts of rare earths of higher atomic number and to a correspond- ing decrease in the amounts of those of lower atomic number. Such behaviour is not confined to the rare earth elements; during their presence in the earth’s crust many elements heavier than zinc will undergo increases, though small ones, of their equivalent weights o1 of their abundance relative to the lighter elements. The first named behaviour is shown primarily by even and the last named by odd ele- ments, because elements having an odd atomic number have always a few isotopes only so that the consecutive mass numbers are not filled by stable isotopes and the formation of radioactive isotopes through neutron addition is possible. In the case of several even elements like cadmium, tin, gadolinium, osmium, mercury, lead etc., a long series of consecutive mass numbers are filled by stable isotopes so that the capture of neutrons * We compared the activities obtained when dysprosia was bombarded with neutrons of a beryllium-radon and a magnesium-radon source in the presence of large amounts of paraffin wax and the figures obtained were as 100: 1. ** V7, M. Gotpscumrpt and L. Tuomassen, Oslo Vidskap Selskapets Skrifter I, Nr. 5, S. 44 (1924). 62 ADVENTURES IN RADIOISOTOPE RESEARCH leads chiefly to the formation of higher stable isotopes. It is therefore even elements that undergo an increase of their equivalent weights with time while the relative abundance of the elements of odd atomic number shifts towards the heavier elements. Below zine, conditions are very different : the result of neutron action in minerals leads here often to the formation of elements of lower atomic number and only to a smaller extent to the formation of heavier isotopes or heavier elements. For example, bombardment of aluminium leads to the formation of a magnesium isotope and to a sodium isotope; the branching ratio between these two processes depends greatly on the energy of the neutrons. Summary The artificial radioactivity of the rare earth elements including scandium and yttrium was investigated. The periods of decay of numerous radioactive isotopes produced lie between 5 min. and a few month. The biggest and smallest saturation intensities of the radiation emitted by these isotopes are in the ratio 10,000: 1. The half-value thickness in aluminium of the #-radiation emitted was measured in several cases, and, in some cases, the maximum energy of the continuous f-ray spectrum and FeErmi’s constant a as well. The absorption of slow neutrons in rare earth elements was measured with a view to discovering the presence of strongly absorbing nuclei not giving rise to active isotopes. The application of artificial radioactivity to analytical chemistry is discussed. It is shown that the combination weight of the rare earth elements occurring in minerals in which a continual production of neutrons takes place has undergone a slight change during geological time. References 1. E. Amaupr, E. Fermi et al., Proc. Roy. Soc. A. 149, 522 (1935). 2. G. Hevesy and Hinpe Levi, Nature 136, 103 (1935) and Nature 137, 849 (1935). G. Hevesy, Nature 135, 96 (1935): Roy. Danish Acad. (Math.-fys. Medd.) XIII, 3, (1935). 3. S. SugpEN, Nature 135, 469 (1935). 4. J. K. Marsu and S. Suapen, Nature 136, 102 (1935). 5. I. C. McLENNAN and W. H. Rann, Nature 136, 831 (1935). 6. E. Rona, Wiener Akad. Anzeiger 27, (1935) and 73, 159, (1936). 7. J. CO. CuHapwick and M. GotpHaBER, Camb. Phil. Soc. 31, 612, (1935). 8. J. R. Dunnine, G. B. Pecram, G. A. Frnx and D. P. Mrrcneti, Phys. Rev. 48, 265, (1935). 9. N. Bour, Nature 137, 344 (1936). 10. H. A. Berne, Phys. Rev. 50, 332 (1936). Originally published in the Kgl. Danske Videnskabernes Selskab. Mathematisk fysiske Meddelelser. 15, 11 (1938) 5. ARTIFICIAL ACTIVITY OF HAFNIUM AND SOME OTHER ELEMENTS G. Hervesy and Hinpe Levi From The Institute of Theoretical Physics, University of Copenhagen ARTIFICIAL RADIOACTIVITY OF HAFNIUM SoME time ago we found that under the action of neutron bombardment a radioactive isotope of hafnium is produced, the activity decaying with a period of a few months (4). To determine the period of decay more exactly, we activated 280 mgm of hafnium oxide prepared by one of us (3) by placing it in a paraffin block together with radium- beryllium sources, containing 600 mgm of radium element as sulphate and twenty times as much metallic beryllium powder. After irradiation for three months the hafnium oxide was removed from the paraffin block and put into an aluminium dish having a surface of 1.2 em? and a height of 2 mm. The dish was placed directly below the aluminium window of our counter, the window having a thickness of about 20 wu. We followed the decay of the hafnium preparation for 200 days by comparing its activity with that of an uranium standard. The decay curve obtained is seen in Fig. 1 and Table 1. From the latter we can conclude that the half life of hafnium is 55 + 10 days (standard mean square deviation). Our initial activity was 20 counts per min, the natural effect being about 4 counts per min. We followed the decay curve until we had a net activity of 1 counts per min. From the fact that, in spite of the long activation, such a modest TaBLE 1. — DECAY-MEASUREMENT OF HAFNIUM | Date Nr. of Days | Counts/min — oa) — —-—— —— | fe NAGUE BYi6 4 boc 0 16.1 IHS, WADDTS 6Ib6 Gace ] 15.3 yy WAND B35 6556 7 oad I IEXe eo Oley erence 25 12.0 ; 73 7.8 64 ADVENTURES IN RADIOISOTOPE RESEARCH activity was obtained we can conclude that hafnium does not belong to the elements showing a strong artificial radioactivity. This is partly due to the fact that the capture of neutrons by most of the hafnium isotopes leads, as explained later, to the formation of a heavier stable ine) oO counts/min. 20 40 60 80 100 200 days Fig. 1. Decay Curve of Hafnium. isotope ; as stable isotopes 176, 177, 178, 179, and 180 are known and only the absorption of neutrons by the last mentioned isotope can lead to the formation of an active product. The relative abundance of the isotopes in the naturally occurring element hafnium, as determined by Aston, is seen from Table 2. TABLE 2. — RELATIVE ABUNDANCE OF THE Harnrum IsoToPEs | Mass number | Abundance 176 DG 177 19%, 178 | 28% 179 18% 180 30% ARTIFICIAL ACTIVITY OF HAFNIUM AND SOME OTHER ELEMENTS 65 We measured also the absorption in aluminium of the f-rays emitted by hafnium. The values obtained are seen from Table 3. TasBLE 3. — ABSORPTION IN ALUMINIUM OF THE f-RAYS Emirrep BY HarnrumM ——— ss Thickness of the Al-foil Counts/min. 0 15.8 11 mgm/cm2 10.2 16.5 mgm /em? 7.4 (half value thickness : 16 -- 1 mgm/cm?). From the figures in Table 3 follows that an aluminium layer of 16 mgm per cm? reduces the intensity of the p-rays emitted by a hafnium oxide layer of 230 mgm/cm? to one half of its initial value. The com- parison of the absorbing power of aluminium for the f-rays of hafnium and scandium, decaying with periods of 55 and 90 days respectively, shows no great difference ; the ratio of the two half-value thicknesses being 1.2. The softness of the hafnium radiation is partly responsible for the low activities obtained after long exposure of hafnium oxide with radium-beryllium sources of a few hundred millicurie, the £- radiation emitted being absorbed to an appreciable extent in the haf- nium oxide sample itself. In the case of hafnium, as already mentioned, every place between the mass numbers 176 and 180 is occupied by a known stable isotope; the formation of the active hafnium isotope is presumably due to the process “2 Hf + on = “Hf. On emitting P-rays according to the equation MAHI = Ta +f the active hafnium isotope becomes the only stable isotope of tantalum known. Hafnium is thus partly converted into tantalum under the action of neutron bombardment, while, as shown by us previously, hafnium is formed under the action of neutrons on lutecium. It is quite possible that, under bombardment with a powerful stream of deuterium or of neutrons, further decay periods of hafnium will be discovered. THE EFFECT OF NEUTRON BOMBARDMENT ON SCANDIUM A few years ago we embarked on the investigation of the effect of neut- ron bombardment on scandium, (4), (5), (6), chiefly in the hope of being able to prepare an artificial radioactive isotope of potassium and to ob- 5 Hevesy 66 ADVENTURES IN RADIOISOTOPE RESEARCH tain some information on the then not entirely elucidated nature of the natural radioactivity of potassium. We bombarded a few grams of very pure scandium oxide prepared by Prof. SrerBA-BOuM and used by Prof. H6n1iascuMipT in his work on the atomic weight of scandium. After neutron bombardment the scandium oxide was dissolved in dilute hydro- chloric acid and 100—150 mgm of sodium chloride as a carrier of #?K and the same amount of calcium oxide was added. The filtrate obtained after precipitation with carbonate-free ammonia was treated with oxalic counts/min wow an 4 @ Ww it 2 4 6 B10) 2 4s 1G: 318) 20) 22) 6245 26) 28 30) 32) 134i) 36) 88 40 42 days Fig. 2. Decay Curve of the Potassium Precipitate. acid and the calcium oxalate formed was removed. The sodium chloride which had been added to the solution of the scandium chloride compound was recovered after the removal of the ammonium chloride content of the last filtrate by evaporation. The activities of the three fractions, namely scandium oxide, sodium chloride, and calcium oxalate, were then determined. Only the two first preparations mentioned were found to be active. The activity of the scandium oxide decayed very slowly while the various sodium chloride fractions obtained in different ex- periments lost half of their slight activity within 10 and 18 hours. We had just finished the experiment mentioned when a note was published by Fermt and his collaborators (1) concerning the action of neutrons on potassium. They found that potassium captured neutrons by giving birth to a potassium isotope decaying with a half-life of 16 hours. The values found by us for the period of the slight activity of different potassium preparations obtained from irradiated scandium showed a half- life between 10 and 18 hours; we thought it justifiable, therefore, to identify the element found by us with that found by Ferri and his collaborators. The initial activities measured amounted usually to about 10 counts/min. In one case, through the courtesy of the Medical Radi- um Station and Dr. J. C. JACOBSEN, we obtained an unusually strong neutron source containing 600 millicuries radium-emanation. The decay curve obtained for potassium 42 in this experiment is seen from Fig. 2. ARTIFICIAL ACTIVITY OF HAFNIUM AND SOME OTHER ELEMENTS 67 The activity found by us in the filtrate of scandium precipitate could only be that of #K, as the presence of active impurities was excluded by the fact that the above mentioned very pure scandium sample was used. The possibility that we measured the half- life of radioactive sodium of 15 hours can be excluded with certainty not only for the reason mentioned above but also for the following reasons: Sodium, ?#Na, can be prepared either from 2%Na by simple neutron capture, or from magnesium if the capture is followed by emission of a proton, or from aluminium if the capture is followed by emission of ana-particle. From the first mentioned process with the neutron sources at our disposal only very weak activities can be obtained even when starting with pure sodium. To prepare measurable amounts of radio-sodium from a few grams of impure scandium oxide an appreciable amount of magnesium or aluminium would have had to be present in the preparation. 15 mgm of aluminium mixed with 150 mgm of ammonium nitrate, for example, gave after activation to saturation less than 0.5 counts/min. and the activity obtained by similar amounts of magnesium with the same sources as used when activating scandium was still smaller. The amount of radio-sodium obtained from magnesium is less, and that of sodium by neutron capture very appreciably less than that obtained from alu- minium. Scandium having just one stable isotope 3iSc, only the potassium isotope *#K can be produced under neutron bombardment according to the equation Sc + in — BK + He. In the case of neutron capture by potassium, on the other hand both reactions 9K +n and 44K + n can occur. While Fert and his collaborators left it open which of the two last named potassium iso- topes were produced, we could conclude from our experiments that the process witnessed by Frrmr was “K +n=#K, and also that the process 39K very probably leads to the formation of the potassium isotope #°K which is responsible for the natural radioactivity of potassium. Recently, WALKE (9), by making use of LAwRENCE’s powerful cyclotron, which supplies a many thousand times stronger neutron beam as ob- tained from our radium-beryllium sources, was able to follow the decay of #K through ten periods and determined its half life period to be 12.4 + 0.2 hours, i. e., a somewhat lower value than that following from the investigations of Fermi and from our Fig. 2. Besides preparing “IK according to the equation 5 1 421 aise + on = igK +a. we succeeded (4) also in preparing this isotope by the process goa + on = igK + 7H . 68 ADVENTURES IN RADIOISOTOPE RESEARCH Walke (10) while reproducing FrrMt1’s results and also ours as to the preparation of #K from scandium, was unable to reproduce our ex- periments in which “K was prepared from calcium. This negative result induced us to repeate our experiments, this time by bombarding with fast neutrons as muchas 1 kgm of calcium carbonate. These were dissolved in a minimum amount of HCl, precipitated by a minimum amount of ammonium oxalate, which sufficed to precipitate all calcium after dissolving 100 mgm of sodium chloride as carrier. Before we finished these experiments, a second paper of WaLKE (12), (13), was publis- hed in which he describes succesful experiments in producing “I< from ealeium, thus corroborating our statement. ACTIVITY OF SCANDIUM After the removal of the radio-potassium produced, the scandium was still showing a weak activity which could not be removed by chemical operations and which is possibly due to a radioactive isotope of scandium. The decay of the weak activity of scandium observed for 240 days is seen from Table 4, which shows the presence of a very weak activity decaying with a period longer than a year. We could not follow up this very weak activity further but concentrated our inte- rest on another period obtained after activating for 24 days in a paraffin block which contained emanation-beryllium sources of an average strength of 50 millicuries. The result obtained is seen from Fig. 8a; the half- life works out to be 90 + 5 days. TaBLE 4. — Activity oF A ScANDIUM SAMPLE AFTER REMOVAL OF PorasstuM Date Nr. of Days | Counts/min | 5 le mekGee sey | 0 23 E1086 Dis MEM Ie en Oy ees | 16 6.0 + 0.4 DAV A 302% 2 =. 50 7.2 0.4 ZOMVN NN aSO 22) 76 5.1 + 0.4 BamVsl aS Old oi.0 92 See oOed WAU, Gita oqac 117 4.9 + 0.3 NB} WAGDES Bio clos 161 5.3 + 0.4 24. X es Olena sta 233 4.4+ 0.3 In the next set of experiments we activated simultaneously three scandium preparations for 50 days with radium-beryllium sources of a strength of about 200 millicuries : one in the usual way inside the ARTIFICIAL ACTIVITY OF HAFNIUM AND SOME OTHER ELEMENTS 69 paraffin block, the second one in a paraffin block but with the preparation surrounded by a shield of cadmium, which absorbed nearly 100% of ) the C-neutrons, and the third one with fast neutrons. The result of the activation of the first named sample is seen from Fig. 3b. The investig- ation of the second sample led to the result that in the presence of 20 40 60 80 100 200 days Fig. 3. Decay Curye of Seandium Irradiated in a Paraffin Block for a) 24 Days b) 50 Days. . cadmium the artificial radioactivity of scandium is reduced to 2% of the value obtained in the absence of cadmium. From Fig. 3 there follows for the half-life period of scandium the value 90 + 5 days. Quite recently WALKE (11), by making use of LAWRENCE’s powerful cyclo- tron, bombarded scandium with deuterons and obtained a period of 85 + 2 days. We want furthermore to mention an early experiment in which we bombarded scandium with fast neutrons emitted by a mixture of 600 millicuries emanation and beryllium powder ; we observed a period of decay of about 50 hours. As shown by Poot, CorK, and THORNTON (8), 70 ADVENTURES IN RADIOISOTOPE RESEARCH and by WaALKE (12), under bombardment with fast neutrons the following two reactions occur as well : 2Sc-+ on = BSc + 3 On 318c + on = 31Sc + 2 on the scandium isotopes obtained emit positrons and have half-lives of 4 and 43 hours respectively ; it was presumably the last mentioned reaction which we observed. Scandium 44 was also produced (13) by the action of a-particles on potas- sium 41 and(11) by the action of deuterons on calcium 43, while scandium 43 was produced (11) by the action of a-particles on calcium 40, and (2) by the action of deuterons on calcium 42. WaLKE was furthermore successful in producing scandium 42 under the action of a-particles on potassium 39, and of scandium 41 under the action of deuterons on calcium. The list of the known radioactive iso- topes of scandium is seen from Table 5. Taste 5. — Active Isororrs or ScANDIUM (According to Walke [11]) Active Isotope | Particle Emitted | Half-life 418¢ | positron | 53 min. 2Sc | positron | 41 ch Sc positron 4.0 h 4Sc | positron 52 h 46Se electron 85 days and possibly also a period of about 1 year. We measured the reduction of counts when covering an active scandium preparation decaying with a period of 90 days with aluminium foils of varying thickness. The result obtained can be seen from Table 6. TaBLE 6. — ABSORPTION IN ALUMINIUM OF THE f-RAYS Emirrep By ScANDIUM Thickness of the Al-foil Counts/min | 0 PAS Ist experiment 5.5 mgm/cm? | 8.8 11.0 mgm/em? el | 0 61.8 2nd experiment 11.0 mgm/em? | 35.3 16.5 mgm/em? 25.8 . . F \ / 9 (half- value thickness: 13 = 1 mgm/cm*) ARTIFICIAL ACTIVITY OF HAFNIUM AND SOME OTHER ELEMENTS al In view of the softness of the f-rays emitted we used thin scan- dium oxide layers; about 50 mgm/cm?. In spite of the thin layers used the soft components were absorbed in the preparation to an appreci- ably greater extent than were the hard ones ; accordingly we have to reckon with the possibility that the radiation emitted by scandium is still softer than indicated by the figures of Table 6. THE RADIOACTIVITY OF EUROPIUM AND ITS ANALYTICAL APPLICATION In their fundamental research on the action of neutrons FrrMr and his collaborators (1) investigated also the activity of a gadolinium pre- paration bombarded by neutrons and found an activity decaying with a period of 8 hours. A few years later, SUGDEN (7), investigating the radio- activity of europium, discovered a very strong activity decaying with a period of 9.2 h. and, at that time, interpreted the above mentioned period of decay of gadolinium to be due to the presence of some europium in the sample investigated. Investigations carried out by us, in which we made use of different gadolinium samples prepared by Prof. Roiua and partly by Prof. PRanpT. and the late Baron AvER v. WELSBACH, confermed completely the conclusion arrived at by SuGpeEn, and this induced us to make use of the radioactivity of europium produced under the action of neutrons to determine the amount of europium present in gadolinium preparations. Prof. Rontia, being engaged in the prepara- tion of large amounts of pure gadolinium compounds, sent us several samples, the europium content of which he wished ascertained. We describe in the following the analytical procedure used by us. Thin layers of the gadolinium oxide samples to be investigated were fixed between two glass plates and placed within a paraffin block. Usually we investigated simultaneously the activation of 4 symmetrically placed preparations. It is of importance to bombard layers having the same thickness and to bombard them with neutrons in such a way that each preparation is hit by the same number of neutrons; the latter was achieved by arranging the sources in the block circularly. We used in these experiments radium-beryllium sources containing 600 mgm of radium, the neutron emission of which corresponds to that from about 400 millicuries of radium emanation ; in addition a beryllium-emanation mixture containing 300 millicuries emanation was also present. After irradiating the samples for 3 days they were homoginized and each sample placed in a small aluminium dish having a surface of 1.2 cm* and put below the window of a Geiger-counter. The intensity of the activity of the different gadolinium samples investigated is proportio- nal to their europium content. In order to arrive at a figure stating a2 ADVENTURES IN RADIOISOTOPE RESEARCH the europium concentration, we added 2° of europium oxide to a pure gadolinium preparation denoted as standard sample in Table 7 and compared the activity of the latter with that of the gadolinium prepara- tions of unknown europium content. The results are seen from Table 7. TaBLE 7. — Acriviry OF DIFFERENT GADOLINIUM PREPARATIONS (The sample labelled ‘‘Standard”’ is the Gd,O, to which 2° Eu,O, was added: samples 1—4 represent progressiv stages in the purification process carried out by Prof. Roza) Samples eounts/min “Standard” 124 1 60 2 60 3) 30 4 25 That, in spite of the large amount of radium and emanation used, the activities measured were not stronger is partly due to the high absorbing power of gadolinium, which reduces the density of thermal neutrons. This effect is especially marked on account of the fact that the thermal neutrons diffuse and are likely to pass through the preparation several times. The latter effect can be best estimated by comparing the activity of pure europium oxide with that obtained when this material is embedded in gadolinium oxide. We activated simultaneously 200 mgm of europium oxide and 200 mgm of gadolinium oxide containing 2%, of europium. If gadolinium absorbed to the same extent as europium, the first named preparation should be 50 times more active than the last mentioned one. Actually we find the ratio to be 200 from which it follows that the presence of gadolinium in our preparations reduced the activity of europium to 1, of the value which would have been obtained if the same amount of europium oxide had been subject- ed to irradiation. Summary The irradiation of hafnium with neutrons has been shown to produce a radio- activity with a half-life of 55 + 7 days which may be ascribed to 18)Hf. The intensity of the f-rays emitted is reduced to half of its initial value by an alumi- nium foil having a weight of 16 mgm/cm2. Scandium, $*Sc, was found to decay with a half-life of 90 + 5 days. The half value thickness for the absorption in aluminium of the f-rays from this element was found to be 13 mgm/cm?. The europium content of gadolinium oxide samples prepared by Professor Luret Rotia was determined by making use of the artificial radioactivity pro- duced under the action of neutrons on the europium present in his samples. E. . R. Friscu, Nature 136, 220 (1935). . Hevesy, Kgl. Danske Vid. Selsk. Math.-fys. Medd. 6, 7. S. 91 (1925). . Hevesy and Hinpe Levi, Nature 135, 580 (1935). . Hevesy, Kgl. Danske Vid. Selks. Math.-fys. Medd. 13, 3 (1935). . Hevesy and Hinpr Levi, Kgl. Danske Vid. Selsk. Math.-fys. Medd. 14. (1936). . SUGDEN, Nature 135, 469 (1935). . L. Poot, J. M. Corx and R. L. THornton, Phys. Rev. 52, 41 (1937). . WaLKE, Phys. Rev. 51, 439 (1937). . G. Hurst and H. Wakes, Phys. Rev. 51, 1033 (1937). . WaLKE, Phys. Rev. 52, 400 (1937). . WaLkE, Phys. Rev. 52, 663—669 (1937). . M. ARTIFICIAL ACTIVITY OF HAFNIUM AND SOME OTHER ELEMENTS 73 References Nw PIC AmaAutpi, E. Frrmr and others, Proc. Roy. Soc. A 149, 52: Zyw, Nature 134, 64 (1934). 74 ADVENTURES IN RADIOISOTOPE RESEARCH COMMENT ON PAPERS 4 AND 5 Usuatty the radioactive indicator must be added to the element the atoms of which are to be traced. It is, however, also possible to produce the radioactive tracer in situ by bombarding the sample with a neutron stream or other energy- rich radiation. In contrast to present days very few people disposed of rare-earth elements before World War II. Among those was my friend Professor Luigi Rolla, professor at the University of Florence. We used to analyse his samples by making use of the method of X-rays analysis described in the comment to papers 3 and 15. After preparing a few kilograms of gadolinium oxide he wished to find out whether or not his samples were free from Eu,O,, the most likely impurity present in Gd,O,. At that date we had no X-ray spectrograph at our disposal and in order not to disappoint Professor Rolla we tried to answer the above question by exposing 50 mgm of his sample in a paraffin block to the effect of neutrons emitted by a mixture of 600 mgm radium and_ beryllium and 300 Me of radon and beryllium. Neutron sources were placed in the paraffin block to obtain slow neutrons which are strongly captured by europium producing a radioactive europium isotope. They are strongly absorbed by gadoli- nium as well, their absorption leading, however, to the production of stable gadolinium isotopes and not to a radioactive isotope of this element ; the latter can only be produced by more intense neutron streams than applied in our experiment. The presence of an activity in Rolla’s exposed samples decaying with a half-time period of 9.2 hr indicated the presence of some europium in his preparations. To carry out a quantitative analysis we added to a known amount of pure Gd,O, (obtained from the great rare-earth chemist Auer von Welsbach) known amounts of pure Eu,O, (also obtained from him). The comparison of the activity of Rolla’s samples with those of these standard preparations lead to the result that Rolla’s purest gadolinium oxide sample contained 0.40, his least pure sample 0.96 per cent of europium oxide. We had already previously, as described in paper 4, applied this method of activation analysis in the determination of dysprosium present in yttrium samples. The determination of europium in gadolinium is unsurpassed in its simplicity and sensitivity. Europium being the element which can be determined with the greatest sensitivity by activation analysis. We were thus fortunate to be faced with the task of applying this newly introduced method in a case which proved later to be the most favourable one. The modest neutron flux emitted by our radium-beryllium sources allowed not less than 0.01 per cent of europium to be determined. By making use of the neutron flux of the cyclotron SEaBorG and Lrvinawoop could determine 6 p. p. m. of gadolinium in iron by activation analysis and after the availability of pile-emitted neutrons of great density such small amounts of europium could be determined as 10~™ mgm In the determination of gadolinium we availed ourselves of the very high absorb- ing power of this element for slow neutrons, thus of an absorbtion method. References Sraspora and Livinawoop (1938) J. Amer. Chem. Soc. 60, 1784. Originally published in Phys. Z. 15, 797 (1914) 6. THE PROBLEM OF THE ISOTOPIC ELEMENTS G. Hevesy and F. Panera From the Institute of Radium Research of the Vienna Academy of Science 1. THE [ISOTOPE CONCEPT It is well known that the separation methods of analytical chemistry have failed when dealing with some radioelements : Nobody has ever succeeded in separating radium-D from lead, mesothorium from radium, or ionium from thorium, nor has it once been possible in these and nu- merous other cases to achieve even a slight enrichment. As more un- successful experiments became known, the workers in this field adopted the view that they were concerned with an inseparability of quite a different kind from that operative with, for example, the rare earths. F. Soppy! was the first to give clear expression to this view by desig- nating such elements as “chemically and physically practically iden- tical” and also to search systematically for new examples of such in- separability among the radioelements?. Especially striking in connexion with the inseparable elements was the fact that they frequently have considerably different atomic weights which, since the a-particle was known to be identical with the helium atom, could be calculated in many instances with certainty ; for exam- ple, the end product of the uranium series, radium-G, which is generally regarded as lead, must have an atomic weight different from that of ordinary lead’. Confirmation of the correctness of this conclusion has been obtained from the recently performed determinations of atomic weights?, which demonstrated that the lead from pitchblende has in fact an appreciably lower atomic weight than ordinary lead and the Jead from thorium minerals. 1F, Soppy, J. Chem. Soc. 99, 72 (1911). 2A. Frecxk, J. Chem. Soc. 103, 381 (1913). 5 See, for example, G. Hrevesy, Phys. Z. 14, 61 (1913); F. Soppy, J. Chem. Soc. 105, 1402 (1914). 4M. Lempert, see K. Fasans, Z. Elektrochem., 1 June (1914) who suggested these experiments; O. HoOnrascumip, Ibid.; M. Curre, C. R. Acad. Sci., Paris June (1914). 76 ADVENTURES IN RADIOISOTOPE RESEARCH The question of the identity of different elements was given increased attention since this was the basis of arranging the radioelements in the periodic system!. K. Fagsans? has indicated that this idea can be carried throughout the periodic system and that the ordinary elements also are probably mixtures. Fasyans has given the tame ‘‘pleiade’’ to such a group of elements which occupy the same position in the periodic system; the separate members were called ‘isotopic’? elements by Soppy. The lack of an ionium spectrum in ionium-thorium samples* could scarcely be explained on any assumption other than that the isotopic elements show no differences in their spectra. The theory of isotopic elements was not readily acceptable to chemists and physicists; to the former, because ever since the formulation of the periodic system they had been accustomed to regard the atomic weight as a fundamental property of an element ; to the latter, because there was no known instance in which two different elements exhibited the same spectrum and such a hypothesis seemed difficult to unify with the prevailing ideas on the origin of spectrum lines*. These doubts were removed and the whole concept of the nature of isotopic elements was simultaneously given considerably more weight by the ideas, developed by E. RurHerrorp® and N. Bonr®, on the constitution of the atom, and by the experiments of MosELry’? on the X-ray spectra of the ele- ments. According to the Rutherford model of the atom, the mass of the atom is associated with an extremely small volume at the positively charged centre and the number of positive charges, and not the atomic weight, is primarily responsible for the properties of the corresponding element. Since the number of electrons which occupy the volume bet- ween the nucleus and surface of the atom is given by the size of charge on the positive nucleus and all chemical and physical properties of the element depend on the number and arrangement of these electrons ; gravitation and radioactivity are excepted. Instability of the nucleus results in radioactive phenomena and the fact that the nuclei of two atoms have the same charge and the same physical and chemical pro- perties but different mass and stability (e. g. radium-D and lead) agrees very well with the Rutherford-Bohr theory. 1A, 8. RussEtt, Chem. News 107, 49 (1913); K. Fasans, Phys. Z. 14, 136 (1913); F. Soppy, Chem. News 107, 97 (1913). 2K. Fagans, Chem. Ber. 46, 422 (1913). 3 F. Exner and E. Hascuexk, Sitz. Ber. Akad. Wiss. Wien 121, 175 (1912); A. S. RussELL and R. Rossi, Proc. Roy. Soc. 87, 478 (1912). 4A. Scuuster, Nature 91, 30 (1913). 5K. RurHerrorD, Phil. Mag. 21, 669 (1911). 6 N. Bour, Phil. Mag. 26, 1 (1913). 7H. MosEtey, Phil. Mag. 26, 1024 (1913). THE PROBLEM OF THE ISOTOPIC ELEMENTS TEA Determination of the charge on the nucleus can be made approxima- tely as a result of the experiments on scattering of a-particles by M. GEIGER and E. Marspen, and more accurately by the recently perfor- med study by H. Mosetry on X-ray spectra. A knowledge of the wave- length of the characteristic X-radiation of an element permits caleu lation of the nuclear charge when certain assumptions are made: it was thus found that this charge always increases by unity on moving from one position in the periodic system to the next higher!. Generally this means a climb to the element with the next higher atomic weight but, in a few exceptional cases, where the chemical properties force the element with the lower atomic weight to be arranged higher in the system (e. g. cobalt and nickel), the rule stated above still applies and thus demonstrates that the number of charges, and not the atomic weight, determines the position of an element in the periodic system. According- ly, the separate positions can be numbered by stating the nuclear charge ; aluminium, for example, thus acquires the atomic number 13, gold 79, ete., and between these all the available numbers except three are already representative of known elements. E. RurHERFORD and C. ANDRADE? have proved directly, by determining the X-ray spectrum of radium-B. which was found to be the same as that of lead, that there are ele- ments having different atomic weight but the same nuclear charge. Isotopic elements differ, according to this observation, only in the structure and mass of the nucleus. The structure does not enter into the ordinary physics and chemistry but is only of importance to the radioactivity. The radioactive properties, however, were the chief means of differentiating the isotopic element and, with a few exceptions (metaneon, the different kinds of lead), even now we are only aware of the existence of such isotopic elements in those examples in which at least one of them is radioactive. Separation by utilizing the radio- active differences does not seem to be conceivable ; it is otherwise with the second fundamental property of the nucleus, gravitation, which should permit both distinguishing and separating. It is useful in these discussions to distinguish between the gravita- tional and electronic properties ; in all applications of weighing (prima- rily determinations of atomic weight and of solubility, etc.) differences in weight of the atoms are directly of use, and diffusion in the vapour- state also depends noticeably on the mass and even permits separation ; Asron* has thus succeeded in fractionating metaneon and neon. Centri fuging also is a process in which mass plays a part and can be applied in 1H. Mosetry, Phil. Mag. 26, 1024 (1913); Ibid. 27, 705 (1914). See also A. VAN DEN Broek, Phys. Z. 14, 32 (1913). 2 EK. RurHerrorp and C. AnpRADE, Phil. Mag. May (1914). 3K. Fasans, Naturwissenschaften 2, 544 (1914). 4 Aston, British Association Report, Birmingham (1913). 78 ADVENTURES IN RADIOISOTOPE RESEARCH several cases for separation. On the other hand, the theory mentioned above considers the chemical properties as essentially independent of the mass, and this applies also to the spectrum and radius of the atom. Differentiation between gravitational and electronic properties is naturally only clear-cut in limiting cases; for example, the velocity of diffusion in liquids, which is primarily governed by the radius, is not independent of the mass! and, according to Bour, the same should also apply in respect of the Rydberg constant of the spectrum series? ; a difference in atomic weight of 1 per cent affects the latter quantity by about 0.05 per cent. The characteristic vibrations of the molecules in the space lattice, and consequently the specific heats, also are probably noticeably different in isotopes?. 2. CAN ISOTOPIC ELEMENTS REPLACE EACH OTHER CHEMICALLY ? From the above discussions it is evident that isotopic elements are certainly not truly identical ; the question now is whether they can be denoted as chemically identical, i. e. whether they can replace each other in their chemical mass action. It is well known that the concentration of substances taking part in all chemical reactions is important (law of mass action of GuLDBERG and WaaGe); if isotopes are chemically identical the concentration must be represented by the sum of the isotopic elements present. For example, the solubility product of ba- rium-free radium-mesothorium chloride would be written in the form [Ra** + mesothorium**] [Cl]? = K Now there is a particularly clear method of testing for replaceability. In electrochemical processes a jump in potential is determined by the concentration of ions of the metal involved ; now when two elements (A and B) are replaceable, the addition of ions of the element B to those of A should exercise the same effect on the potential jump as if the ele- ment A had been raised to the ionic concentration A + B. For example, the potential difference RaD metal/RaD nitrate solution should be changed to the same extent by the addition of lead nitrate solution, within the meaning of Nernst’s theory of the galvanic production of current, as if the ionic concentration of RaD had been increased, and vice versa. Instead of the electrode potential of a metal, the so-called decomposi- tion potential, which, according to Lz Branc, is of the same magnitude 1G. Hevesy, Phys. Z. 14, 1209 (1913). 2N. Bour, Phil. Mag. 27, 512 (1914). 3K. Fasans, Naturwissenschaften 2, 544 (1914). THE PROBLEM OF THE ISOTOPIC ELEMENTS TY and is the voltage at which the element can be deposited electrolyt ically, ‘an be considered. This was the first method which we adopted to solve the above problem, namely, to determine whether the decomposition potential of an element is displaced when an isotopic element is added to it. The sensitivity of radioactive methods permits the quantitative determination of even the unweighable amounts which always deposit below the decomposition potential, and this opened up a second method of testing the problem ; we studied the variation in these amounts on adding isotopic elements. The third method depended directly on mea- suring the potential difference shown by a RaD peroxide electrode. More details will be discussed below concerning the method of depositing RaD peroxide which we have succeeded in preparing from radium ema- nation in visible amounts. 3. STUDIES ON THE REPLACEABILITY OF ISOTOPES (a) The Decomposition Potential of Radioelements When determining the curve of the decomposition potential it is usual to measure the current passed by the cell as a function of the electrode potential. In plotting these curves it is always postulated that the current is carried essentially by the ion whose decomposition potential is to be determined and that current can pass continuously only when the potential difference attained at the cathode is equal to that which would be registered when the metal in question is immersed in the so- lution. This method of determining decomposition potentials is not applicable in radio-electrochemistry since the concentration of the radio- ions is not sufficient to carry the current exclusively. Therefore, we studied the amounts of the radioelements deposited during a time of 24 hr, under precisely the same conditions, as a function of the cathode potential. In the first method a sudden increase in the current strength occurred at the value of the decomposition potential ; in the second method there was a sudden increase in the amount deposited ; a further difference between the two types of decomposition-potential curves consists in that the deposition in the second type can be investigated at potentials even higher than the decomposition potential whereas in the the first type the cathode potential does not rise even when the current is increased. We have plotted in Fig. 1 a decomposition-potential curve of the second type for radium-E; the solution was about 10 9N in Rak (isotopic with bismuth). It is evident from the curve that some RaE is deposited at any potential! and that because of the sensitivity of the method this amount can be SO ADVENTURES IN RADIOISOTOPE RESEARCH determined quantitatively but that at —0.24 V (compared with the calomel electrode) there occurs a sudden increase in the amount de- posited. If bismuth nitrate is now added to the solution until the Bi + RaE normality of 10~4 is reached the characteristic increase takes place at —0.14 V i.e., about 100 mV lower (see Fig. 2). According to Nerwnstv’s theory it can be expected that a change in concentration of 0 106 +05 +0% +03 +02 +07 O Fig. 1. Cathodic deposition of radium-E. Concentration of the solution about 10~8N in Rak 40 90 80 0 106 405 40% 103 +02 +07 O -O1 -02 -O3 -04 -05 -06 Volt Fra. 2. Cathodic deposition of radium-E. Concentration of the solution 10-4N in bismuth isotopes, Bi + Rak. THE PROBLEM OF THE ISOTOPIC ELEMENTS 81 trivalent Bi by a power of ten will result in a lowering of the decom- position potential by about 18 mV; in the present example, therefore, 90 mV would be expected. The break in the curve for pure Rak is in- deed distinct but after all not so sharp as that for Rak + Bi; this is an effect which in the first case is connected with the fact that the elect- rode could not be covered with a layer of Rak even if all the Rak present were deposited. 1 90 sat “i i sob 40 30 20 10 | 206 07 108 +09 +70 «+71 «+12 +413 +14 415 +16 Volt Fra. 3. Anodic deposition of thorium-B peroxide. Concentration of the solution 10~12N in ThB. The lack of sharpness becomes still more pronounced with more dilute solutions, e. g. in the case of our experiments with ThB. The solution was about 10~!2N in ThB. The discontinuity for peroxide deposition, which can be traced more easily than that for metallic thorium-B, occurs at + 1.13 V (see Fig. 3). Since the decomposition potential in 0.001 N lead nitrate solution saturated with PbO, occurs at 0.87 V, the displace- ment amounts to 0.26 V. From the concentration difference of nine powers of ten a difference of 9 x 28 = 252 mV would be expected from theory, and thus the values agree very well!. Individual difficulties which have been encountered in these determinations will be examined in the discussion of the experimental details. 1 If the average value of 20 mV determined by CumMING and ApEaG (Z. Elektro- chem. 13, 19 [1907]) is assumed as the displacement per power of ten, then the agreement is less good, yet always passable in view of the large sources of error in these experiments ; a similar mean value is obtained from our measurements which are quoted later. 6 Tlevesy 82 ADVENTURES IN RADIOISOTOPE RESEARCH (b) Deposition below the Decomposition Potential As was shown some time ago!, a small quantity of any radioelement deposits even below the decomposition potential and can be measured with the aid of sensitive methods which are now available. Thus, for example, 4 parts per 1000 of Rak are deposited at about —0.17 V in 24 hr on an electrode 1 cm? in area when the stirring is thorough ; this deposition is not affected by the presence of foreign ions, apart from bismuth, in the solution. If the solution is made 0.01 N in Bi ions the deposition of RaE no longer takes place. At a higher concentration the percentage deposition should naturally be much smaller; 4 parts per 1000 of a 0.01 N bismuth nitrate solution would indeed amount to a few milligrams and thus would form a visible cover which cannot exist below the decomposition potential. This specific effect of bismuth ions on the deposition of RaE ions cannot be interpreted in any way other than by replaceability of these isotopes. We found similar results for ThB, irrespective of whether it was de- posited as metal or peroxide?. For example, at 1 V i.e. below the deposi- tion potential, 5 per cent deposited and the deposition was in no way affected by the presence of thallium or other ions near to lead. In 107° N lead nitrate solution, the deposit was already less than 14 part per 1000 and in 10-3 N the fraction deposited was no longer detectable. Clearly in this instance also, increased deposition occurs because of the high concentration, but the positions of most of the ThB atoms are taken by lead atoms, depending upon the concentration ratio of the two. (c) Measurement of a RaD Peroxide Cell Concerning the question of isotopy of the elements we are mostly limited to indirect methods like those described above, since no single instance is known in which both of two isotopes exist pure and in visible amounts. Visible amounts can be made available only from relatively long-lived elements ; when recovered from minerals they are always contaminated with isotopes, e.g. uranium-II with uranium-I, ionium with thorium, mesothorium with radium, and so on. Radium-D, which occupies a posi- tion midway between the long- and short-lived, is always mixed with about ten million times the amount of lead when obtained from pitch- blende ; the considerable quantity of radium emanation available to us, however, gave us the opportunity to obtain directly visible amounts of RaD, completely free from lead because of its formation from emanation allowed to disintegrate in carefully purified quartz vessels. 1G. Hevesy, Phil. Mag. 23, 628 (1912). 2The anodic deposition of ThB at strongly positive potentials was explained by the formation of (ThB)O, (F. Panrrx and G. HEvesy, Svtz. Ber. Akad. Wiss. Wien 122, 1027 [1913)). THE PROBLEM OF THE ISOTOPIC ELEMENTS 83 In the course of a few weeks the sealed flasks, which had meanwhile become coloured a deep brownish-violet, were opened, washed out with nitric acid which had been distilled through a quartz condenser, and the solution was evaporated. Until completion of the electrolysis care was taken to use only quartz and no glass vessels. According to the con ditions of electrolysis metallic RaD or RaD peroxide was obtained as a visible coating on small platinum wires; preliminary experiments allowed this result to be expected since we had convinced ourselves that amounts of lead smaller than 0.001 mgm, as peroxide, are still clearly visible and electromotively effective, i.e. they can be used for building a cell’. We have manipulated various quantities of emanation, 15 c on the avarage, but even 100—200 me are sufficient for carrying out an experiment. The activity of the wires, as checked by measuring the a- and f-radia- tion, was of the order of magnitude expected for pure lead-free RaD ; moreover, Our apparatus was free from lead to the extent that we were able to detect an artificial contamination of 10-9 gm Pb. We measured the electromotive force of the following cell : Pt/RaDO,/Ra(DNO,),, HNO,, RaDO,/KNO,/KCl, Hg,Cl,, Hg 10-N ) LOseN’ satd) LN IN sated: The potential of RaD O, was found to amount to —0.884 V. The PbO, potential measured in the same conditions was found to be —0.888 on the average?. In another series of experiments lead nitrate was added gradually and the following electromotive forces were found (at 20° C) (Table 1). TABLE lL PbO, (Raa), Total normality a = of the lead | Change in | 1 | Change in 3 eHg , | eHg isotopes | potencial “3 potential (Vv) Be (V) is | difference | | difference 10-5 0.906 0.906 1O=3 0.774 | 0.032 0.868 0.038 1071 | 0.837 | 0.037 0.839 0.030 Total change 0.069 Total change | 0.068 1 Refer to J. Koenrcspercer and W. J. Mier, Phys. Z. 6, 849 (1905) : Ibid. 12, 606 (1911). 2The RaD nitrate concentration could only be determined to the nearest order of magnitude and therefore importance should be attached only to the agreement of the two potentials and not to their absolute values. 6* 84 ADVENTURES IN RADIOISOTOPE RESEARCH It is evident that the cells are identical within the limits of experi- mental error. We attach less importance to this than to the fact that the addition of Pb ions to RaD nitrate solution exercises precisely the same effect on the potential difference of the RaD peroxide which, according to Nernst’s theory, the Ra D ions (and only they) should have. ; RT G This proves that ¢ in the Nernst formula E = a In Al is to be n understood as the sum of the concentrations of the isotopic ions present. A special peculiarity of this RaD peroxide electrode deserves to be mentioned. If it is allowed to remain in contact with air for some time it immediately shows, on immersion, a potential which may be one- or two-tenths of a volt higher than the constant electrode potential established after a certain time. This is probably connected with the strong ionization in the vicinity of the wire. (d) Experimental Details The curves described above for the decomposition potential of Rak were obtained as follows : Two gold electrodes, each 1 cm? in area, were immersed in 25 cm’ of 0.1N nitric acid solution and were polarized for a long time until the desired electrode potential had established a con- stant value. A steady motion of the solution was ensured by passing a current of nitrogen. After the attainment of constant potential a few- tenths of a cubic centimetre of a solution at the same nitric acid con- centration and containing radium-E or RaE and Bi was added! and the experiment was allowed to run for 24 hr. After this time the electrodes were withdrawn without interrupting the current, washed with distilled water, always in the same way, and measured in an electroscope ; 5 cm? of the solution were evaporated on a watch glass and likewise measured and hence the percentage deposition of Rak could be calculated. The experiments with ThB also were carried out similarly ; in this case the active solution was added to 100 cm? of 0.001 N nitric acid and the de- position was made on correspondingly pre-treated platinum electrodes with an area of 4 x 2 em?. The potential difference was measured by means of a Siemens compensating apparatus. The Rak solution was obtained directly from emanation, and the thorium-B by exposure of a platinum foil to radiothorium. Particular care was used in the latter case to exclude lead completely ; a part of the experiment was carried out in quartz vessels and with the use of water purified specifically for this purpose?. 1 The addition of small amounts of bismuth to make the solution about 1077 N often caused an initial change of the cathode potential by several millivolts. 2 The purification of the water was that usually employed for determirations of atomic weight (cf. O. Honiescumip, Mitt. d. Inst. Radiumforschung. 8, 8). THE PROBLEM OF THE ISOTOPIC ELEMENTS ae The effect of adding very small amounts of lead on the deposition of ThB below its decomposition potential was studied as well. Table 2 clearly shows the decrease in percentage deposition of ThB from a 0.001N nitric acid solution on platinum electrodes (-- 0.4 V, yz) With increasing concentration of lead ; in every experiment four electrode surfaces were measured and the mean value was taken, TABLE 2 | Amount of ThB y ; deposited, as a otal concentration of : ; percentage of that Pb isotopes Aye originally present (%) 2 On Ne 0.98 Oe | 0.75 NOs 0.86 Om> 0.105 LOms no longer detectable Thus up to a concentration of the solution of 10-7 N, the deposition is only slightly affected, at 10-5>N a marked fall is already noticeable, and at 10-3N the deposition is no longer measurable. This method, which can be still further refined by choosing smaller electrodes, still permits the detection of very small amounts of inactive lead, since the addition of another element, e.g. thallium, which is a neighbour of lead, has no noticeable effect on the deposition of ThB even at a con- centration of 10-°N TI. 100 90 60 70 60 50 40 30 20 $06 707, #08 «0090 «410 «+11 «©4720 99084 47S +76 VOlt Fic. 4. Anodic precipitation of ThBO,. The lead isotope concent- ration [Pb + ThB] of the solution 10~5N, 86 ADVENTURES IN RADIOISOTOPE RESEARCH The determination of the decomposition potential by means of the methods mentioned above is based on the assumption that the current strength is large enough to permit deposition of the whole quantity of the radioelement within the duration of the experiment. It is easily seen, e. g. in the electrolysis of a 0.001 N lead nitrate solution, that the above condition is far from being satisfied, since the electrode potential is attained in our apparatus at a current strength of about 3 x 10-6 A which, in the course of 24 hr, is capable of depositing only a very minute fraction of the lead ions present. This is particularly emphasized since, if this point is not taken into consideration, there will be found too high a value in determining the decomposition potential by the methods mentioned (sudden increase in the amount deposited). For example, Fig. 4 shows the apparent decomposition potential of ThB using 0.001 N solution ; it is considerably higher than the calculated value, and the explanation is probably to be found in the reason mentioned above. We hope to be able to revert to several of the points which have been discussed, particularly to the deposition below the decomposition potential. 4. DISCUSSION It has already been mentioned above that the difference in the atomic weights of individual isotopic elements exists without any doubt. Hence it follows that, in so far as gravitational properties are concerned, the isotopes are not identical and that by centrifuging, for example, meso- thorium should be easier to separate than its isotope radium from barium. On the other hand, a similar differential in the chemical pro- perties of isotopic elements is not observed, we have found replace- ability in the electrochemical behaviour. It is concluded that the electrode potential may be written in the form : IRAN LG == lin oe nF where Yc denotes the total concentration of all the isotopes present, and correspondingly the mass action law may be written in the form : _[Zisotope A]™ [Xisotope B]™ ... iE eee 1G [Xisotope A’]™ [Lisotope B’]":... The proposition that two atoms with different weights can replace each other in their mass action seems at first glance to contradict the second law of thermodynamics. The contradiction disappears, however, when the concept of chemical individuality, to which the mutual replace- ability is related, is considered more closely and is defined appropriately. THE PROBLEM OF THE ISOTOPIC ELEMENTS 87 We generally ascribe to each element a particular chemical character which varies discontinuously from one element to another. In their mass action silver atoms can replace only other silver atoms and not lead, thallium or other atoms. Accordingly, in the Nernst formula for the electrode potential of silver : e can be changed only by the addition of silver ions and not by others. If it is now found that atoms which, in spite of having different atomic weight, replace each other chemically and that ¢ must be understood as the total concentration of the isotopes, it then seems necessary to define the concept of chemical individuality such that this does not imply complete equality of the atoms involved but the mutual replace- ability of the two atoms. The correlative of replaceability seems to be equality of the nuclear charge numbers whose fundamental importance becomes more and more prominent. Summary Experiments have been made to discover whether isotopic elements can replace each other chemically ; the following electrochemical methods have been employed for this purpose. (1) The electrolytic deposition of radium-E with and without the addition of bismuth has been studied and it has been found that the decomposition potential is displaced by the addition of bismuth in the sense and by the amount which would be expected of the addition of the same (Rak) ions in accordance with Nernst’s theory; a study of the deposition of thorium-B with and without the addition of lead yielded the same result. (2) It has been shown that the deposition of the very small amounts of radio- elements which precipitate below the decomposition potential is hindered by the presence of isotopes (and only by these), and this likewise can be explained only by replaceability. (3) Radium emanation has been allowed to disintegrate in quartz and the radium-D formed has been deposited electrolytically as the peroxide on platinum wires; visible and at the same time electromotively active amounts (a few- thousandths of a milligram) have thus been prepared. The cell RaDO, | RaD(NO,), | | KNO, | KCl,Hg,Cl,,Hg showed the same electromotive force as a cell similarly made with lead peroxide, and furthermore the addition of lead ions to the RaD solution changed this e. m. f. in the same way a corresponding addition of RaD ions should change it according to NeERNst’s theory ; hence it is concluded that the ionic concentration ¢ in the Nernst formula must be understood as the sum of the isotopic ions. From our study, therefore, the conclusion must be drawn that isotopic elements are able to replace each other in their mass action. 88 ADVENTURES IN RADIOISOTOPE RESEARCH COMMENT ON PAPER 6 As shown by Nernst the electrode potential of a metal is proportional to the logarithm of the concentration of its ions present in the surrounding solution. The validity of this regularity was tested up to 0.0001 N ionic concentration. The application of labelled bismuth and labelled lead permitted us to demonstrate the validity of this regularity at much lower ionic concentrations than the above mentioned one. When this investigation was carried out in 1913 the notion of isotopes had just emerged, and it was thus of interest to demonstrate that the voltage at which RaE, for example, is precipitated on the cathode, is influenced by the addition of a bismuth salt to an extent to be expected on assuming the practical chemical identity of bismuth and RaE. It is not, however, influenced by adding salts of other metals. Below its decomposition voltage minute traces of RaE are deposited as well, this minute precipitation is also influenced by addition of bismuth salts, but not by addition of salts of other metals. The large amount of radon available at the Vienna Institute made it possible to obtain a visible RaD layer on a platinum wire. The electrode potential of a RaDO, electrode was found, measured against a calomel electrode undistinguishable from the potential of a lead peroxyde electrode. The experiments were carried out with peroxide of lead instead of metallic lead, as the electrode potential of metallic lead was found not to be sufficiently reproduceable. If it were possible to measure these electrode potentials to an accuracy of several decimals, we would presumably measure some difference between the electrode potential of lead peroxide and radium D peroxide, as isotopes are not strictly identical in their chemical properties. The very far-reaching practical chemical identity of isotopes of an element is, however, conspicuously demon- strated by the results of this paper. Originally published in Z. phys. Chem. 89, 294 (1914) 7. THE VELOCITY OF DISSOLUTION OF MOLECULAR LAYERS G. Hervesy and E. Rona From the Chemical Institute of the University of Budapest THE velocity of dissolution of finite layers can be followed quantitatively by considering the process of dissolution as being comprised of two partial processes ; one of these consists in the formation of a layer of saturated solution surrounding the solid surface and the other is a process of diffusion from this boundary layer into the liquid!. The velocity of dissolution is represented by the equation : dz/dt = DOF(c, — c)/6 where 6 denotes the thickness of the boundary layer, / a proportionality factor, D the diffusion coefficient, O the area, c, the saturation con- centration, and ¢ the concentration of the solution. The dissolution will therefore proceed more rapidly the smaller the thickness of the boundary layer, i.e. the greater the speed of stirring, the greater the diffusion velocity of the participating molecules and the further the solution is from the saturated state ; the formula also shows a parallelism between solubility and velocity of dissolution. Nernst and BrRuNNER? have shown that these ideas are quite generally applicable to heterogeneous reactions. The present communication discusses the course of the dissolution process of molecular layers, which could be also described as infinitely thin, and the extent to which the above simple equation is satisfied. THE PREPARATION OF INFINITELY THIN LAYERS It is well known that an infinitely thin layer of radioactive metal or its oxide, known as the so-called active deposit, can be obtained simply ; by the decay of the gaseous emanations metallic products are formed, which are isotopes of polonium, lead, bismuth and thallium and which 1 Noyes and Wuirney, Z. phys. Chem. 23, 689 (1897). 2 Nernst and Brunner, Z. phys. Chem. 47, 52, 55 (1904). 90 ADVENTURES IN RADIOISOTOPE RESEARCH gradually become deposited from the suspension in the air; this de- position process can be considerably accelerated by applying an electric field. In our experiments we made use of a radiothorium preparation, which provides a constant source, and the active deposit yielded by the emanation was collected on a quartz surface 1.6 cm in diameter. The quartz disk was covered with a mixture of ThB (lead isotope) and ThC (bismuth isotope) because the first decay product of emanation, ThA, decays very quickly with a half-life of 1/, sec. A simple calculation yields 5 x 10-4 gm as the total mass of the deposit, of which about 90 per cent consists of ThB and 10 per cent of ThC. In order to cover the surface completely with a molecular layer of lead 2 x 10~® gm would be necessary, i.e. 50,000 times the amount actually present; we can thus rightly consider the surface as having an infinitely thin covering of lead and bismuth. The velocity of dissolution was determined as follows: The quartz disk was allowed to stand for several hours after cessation of the activation until radioactive equilibrium had been established and the f-activity formed by the deposit, and showing the relative amounts of Pb and Bi present, was then determined; the disk was then placed in a bell- shaped vessel, provided with an outlet tube and cock, and after a certain time the 100 cm? of liquid in the vessel was drained out. Care was taken to attain a constant stirring speed, the disk being placed in the solution only when this speed had been established. The f-activity of the quartz disk treated in this way was then measured again 15 min after completion of the experiment and at various later time intervals, and it was thus possible to decide upon the amounts of Bi and Pb present, from the change of the activity with time, and to determine the percentage which had entered into solution. THE DETERMINATION OF THE VELOCITY OF DISSOLUTION The investigation was concerned with the determination of the effect (1) of the concentration of acid in the solution (2) the viscosity (3) of the speed of stirring (4) of the time on the velocity of dissolution of the molecular layer and finally on the effect due to isotopes of the corresponding elements in the solution. The percentages of Bi and Pb isotopes which dissolve in nitric acid in 60 sec, under the same experimental conditions, are summarized in Table 1. THE VELOCITY OF DISSOLUTION OF MOLECULAR LAYERS g} TABLE 1. — AMOUNTS DISSOLVED IN 60 sec ——_——— On In 10— N 1G Nes 10-2N. | 10-2 N 10-1 N N water | HNO, HNO, HNO, HNO, HNO, HNO. (%) | (%) (%) (%) (%) (%) (%) Bismuth | isotope 37 38 30 61 72 77 78 (ThC) Lead isotope 60 61 60 SO 81 83 84 (ThB) The velocity of dissolution is the same in 10~4 N acid as in conductivity water but increases with further increase of the acid concentration and in N acid amounts to about twice the above value. It is well known that the bismuth isotopes dissolve colloidally in water and from diffusion experiments the conclusion was drawn that this is no longer the case in 10-3 N acid?. It suggests itself to associate the sudden increase in the velocity of dissolution with this change. The velocity of dissolution of finite layers depends on, among other factors, the diffusion velocity of the products involved. In order to change this velocity glycerol was added to the nitric acid in order to cause a considerable increase of the viscosity and a corresponding lower- ing of the diffusibility of the hydrogen and other ions without otherwise changing the experimental conditions. The consequence of the glycerol addition was, as is clear from Table 2, a decrease in the velocity of dissolution. TABLE 2 Dissolved in 10-3 N Dissolved in - aa ah is 10—2 N HNO, ae 3 containing 25% glycerol Bismuth isotope (%) 61 52 Lead isotope (%) | 80 73 The viscosity of the glycerol mixture found by the ordinary outflow method was 1.650 relative to that of water as unity. Even after treatment for several hours with concentrated acids it is found that the quartz disk still retains about 20 per cent of its original ThB—ThC coating which, however, is not present on the surface but is situated inside the quartz where it has arrived through the so-called 1F. Panera, Kolloid-Z. 13, 1, 297 (1913); G. Hevesy, Phys. Z. 14, 1209 (1913). 92 ADVENTURES IN RADIOISOTOPE RESEARCH radioactive recoil process. A portion of the active deposit is laid down on the quartz disk in the form of ThA, the ThA then emits a-particles with the result, according to the principle of action and reaction, that a recoil of the atoms is required and thus some of the ThB atoms formed from ThA are deposited under the surface of the quartz. The range of such recoil atoms amounts to about 1/,) cm in air and, therefore, less than 10-4 em in quartz; this thin layer of quartz is quite sufficient to protect the fraction driven inward by the recoil effect from the reaction of the acid, although radiation, from which its presence can be inferred, still affects the electroscope through this layer. The portion of the active deposit found underneath the surface depends on the time and other conditions of exposure which have been chosen to be strictly the same in all these experiments. The portion of the active deposit occurring below the quartz surface, found experiment- ally to be 20 per cent, was not taken into account in compiling the tables, the values in which refer only to the soluble part of the active deposit. The determination of the effect of the stirring speed on the velocity of dissolution meets with difficulties. Because of the large velocity of dissolution of molecular layers the times of experiment must be limited to a few minutes and the unavoidable immersion and withdrawal of the quartz disk from the solution always acts as intense stirring. In our experience the effect of stirring velocity on the velocity of solubility of molecular layers was not considerable. The Relation between Velocity of Dissolution and Solubility As shown by the above formula, the velocity of solubility of finite layers increases with the solubility of the substance ; this is true also for infinitely thin layers, and thus the velocity of solubility of the lead isotopes is greater than that of the bismuth isotopes (Table 1), both in water and in nitric acid, corresponding to the greater solubility of the lead salts involved. ; Lead peroxide dissolves more slowly than metallic lead and lead monoxide, in agreement with its lower solubility in HNO,. Such peroxide layers! were produced by the anodic deposition of ThBO, on platinum. Only 20 per cent of ThBO, dissolved in the same conditions in which 80 per cent ThB entered the solution. Since dissolution from quartz is not strictly comparable with dissolution from platinum, a comparison was therefore made between the velocity of solubility values of ThBO, and ThB likewise deposited on platinum by a cathodic reaction ; in this case also more of the latter dissolved as is proved by the figures in Table 3. 1F. Panera and G. Hevesy, Wien. Ber. 122, 1038 (1913). 2F. Paneru and G. Hevesy, Wien. Ber. 123, 1050 (1913). THE VELOCITY OF DISSOLUTION OF MOLECULAR LAYERS 93 TABLE 3 Ratios of the amounts of ThB Ratios of the amounis of ThC and ThBO, dissolving in and ThCO, (?) dissolving in 1 min 1 mip 10—-2N 10-2N HNO, 10—2N 107AN HINO, HNO, + 10-1N | HNO, + 10—-1N oxalie acid oxalic acid 4.0 0.76 3.1 0.75 It is also evident that in the presence of a reducing agent such as oxalic acid the large difference between the dissolution velocities of the cathodically and anodically deposited lead isotopes disappears, in agreement with the ready solubility of the peroxide in oxalic acid solution. In the more recent development of electrochemistry particular atten- tion is devoted to the reactions which take place between the deposited products and the electrode material; the study of the velocity of solubility with the electrolytically deposited radioelements offers an easy method of approaching more closely to these problems ; thus, in the case of polonium it was proved that this element forms stable com- pounds more easily with Pt and Pd than with gold. Change in the Dissolution Velocity due to the Presence of Isotopic Ions of the Dissolving Metal in the Solution Isotopic atoms are interchangeable in their electrochemical reactions!. Now the dissolution of a metal is to some extent the reverse of its electro- lytic deposition and therefore it is to be expected that ThB will dissolve only to the same small extent as lead in nitric acid saturated with lead nitrate. Before discussing the behaviour of a molecular layer when dissolving in a solution which already contains some dissolved isotope, we would like first to explain in more detail the process which takes place between a solid phase and its saturated solution. Just as the equilibrium state between a liquid and its saturated vapour is regarded as dynamic, i.e. the assumption is made that the same number of molecules condense from the vapour and leave the liquid in unit time, the equilibrium state between a solid phase and its saturated solution is also regarded as dynamic, i.e. it is assumed that at the boundary of, for example, PbCl, solid water saturated with PbCl, a dynamic exchange of PbCl, molecules takes place between the two 1G. Hevesy and F. Panera, Phys. Z. 15, 797 (1914). G4 ADVENTURES IN RADIOISOTOPE RESEARCH phases. It is necessary to know the rate of this exchange because if it is very large the (ThB)Cl, molecules in the solid phase will exchange directly with the PbCl, molecules of the solution and thus simulate direct dissolution. An answer to this question would be possible if, for example, a saturated lead chloride solution could be shaken with solid PbCl,, the lead atoms of which were numbered or characterized in any other way without affecting their chemical properties, by finding in which phase the numbered atoms then existed. Such ideally labelled lead atoms are the radioactive isotopes (ThB, RaB, AcB, RaD, etc.). We need only add, for example, (ThB)(NO,), to the solution of lead nitrate, precipitate the Pb—ThB mixture as chloride and thus to obtain a ‘‘coloured”’ lead chloride. If 1 mgm of PbCl, were originally associated with one relative unit the detection of this relative unit in the saturated solution would allow the inference that 1 mgm of the lead atoms originally in the solid phase had then been transferred into the saturated solution, or vice versa. The application of radioactive indicators thus serves to permit tracing of the exchange of atoms of the same kind between two phases; the ‘colouring’ of the labelling atoms is a purely radioactive property and, although their mass is different from that of the labelled isotope (they have different atomic weights), their chemical reactions, with which we are concerned, are still the same. To determine the velocity of exchange between the solid lead chloride and its saturated solution, 250 mgm of ThB-labelled PbCl, was shaken with 25 cm? of a saturated solution of pure lead chloride in a thermostat at 20°C. After 24 hr the mean value of ten experiments showed that 1.2 mgm, or 4 per cent of the lead chloride originally in the solid phase had entered the saturated solution; after 48 hr the value was about 3/4, per cent. The lead chloride solution was prepared by cooling an originally slightly supersaturated solution in the thermostat and was therefore fully saturated ; the possible sources of error all tended to yield high values for exchange and the above value should therefore be regarded as an upper limit only. The velocity of exchange depends very markedly on the mechanical consistency of the solid phase, and this also applies to the velocity of solubility. It is thus imperative to compare these two quantities under the same experimental conditions. For example, on shaking 250 mgm 1These ideas do not apply strictly to diffusion processes or, therefore, to the exchange of atoms in the same phase, because the velocity of diffusion is dependent on the mass; this dependence is very slight, however. Considered from the stand- point of diffusion only those atoms which are both isotopic with and of the same mass as those under study are really ideal indicators. It appears, however, that UY may be such an ideal indicator for UX,. THE VELOCITY OF DISSOLUTION OF MOLECULAR LAYERS Qd5 of the same lead chloride 44 per cent had already dissolved in 1 hr, representing 44 per cent of the saturation concentration since the liquid volume amounted to 25 em? and 250 mgm are soluble in this volume. It is seen, therefore, that the velocity of exchange between the two phases is small compared with the velocity of solubility. It will be seen later that in the case of a molecular layer the exchange velocity becomes much larger when expressed on a percentage basis but is still smaller than the velocity of solubility, and thus a diminution of the velocity of dissolution of a substance can still be detected by the presence of its isotope in the solution. To this end we have compared the amounts of ThB collected on quartz which dissolved in water and in a saturated PbCl, solution, under the same conditions, in 1 hr. There was only a small difference since in the first case 79 per cent dissolved while in the second the value was 75 per cent. Shorter experimental times were then chosen and thus the presence of lead ions in the solution had a very considerable effect on the velocity of solubility of ThB. These experiments were performed with the same apparatus used for obtaining the values recorded in Table 1. a Taste 4. — EXPERIMENTAL TIME 60 sec rah . | f : | Amount dissolved in Amount dissolved - i E: i é = | 10-3 N HNO, saturated in 10 -?N HNO, | Gi a (o/) with Pb(NOs;), . /O (%) \ Bismuth isotope ThC .. 61 64 : ; | =; ta Lead isotope ThB .... | 80 67 The presence of lead ions in the solution diminishes the velocity of solubility of the lead isotope but not that of the bismuth isotope. It is a fortunate circumstance that ThB and ThC are simultaneously present in the same place on the quartz surface, and when, in spite of this, the velocity of solubility of only one is affected, this means that there is a specific effect due to the addition of the appropriate element ; for example, glycerol, which has no selective effect but which increases the viscosity of the solution, affects the velocity of solubility (Table 1) of ThB and ThC equally. In order to ascertain whether a small concentration of lead ions affects the amounts of ThB and ThC dissolved, we have performed experiments in 10~-3N HNO, solutions which were also 10°3N in lead. The time of experiment was 40 sec and the arrangement was different from that described above. The ratio of the amounts of Pb and Bi dissolved was found to be : 96 ADVENTURES IN RADIOISOTOPE RESEARCH in pure 10-3N HNO, 1799 in 10-8N Pb(NO,), +10-3N HNO, 1.49 which is still a significant difference. The conclusion to be drawn from these experiments is that the velocity of exchange between the solid phase and its saturated solution is already commensurate with the velocity of dissolution for a molecular layer, but, that when the experimental time is short an effect on the velocity of dissolution due to the presence of isotopic ions in the solution can be detected. The following experiment seemed to be of interest in connection with those described above: 200 mgm of Pb(NO,), labelled with ThB was added to a solution of PbCl, (200 mgm), a portion of the PbCl, was then allowed to crystallize out and the distribution of the different kinds of lead atoms between the chloride and nitrate was studied. After a few minutes required for performing the manipulations it was shown that there was a completely uniform distribution of all the lead atoms, within the limit of error amounting to 1 per cent. Z. IKLEMENSIEWICZ! has recently performed similar experiments. He studied the distribution of ThB and also RaB between a lead amal- gam and a mercuric nitrate solution and found a completely uniform distribution ; the accuracy of his experiments was greater, with an error of 1%, per cent. Summary The velocity of dissolution of molecular (infinitely thin) layers shows quali- tatively the same behaviour as that of finite layers. The velocity of dissolution of lead and bismuth isotopes in nitric acid increases with acid concentration, with lowering of the viscosity of the solution and with the solubility of the substance involved. The presence of lead ions in the solution lowers the velocity of dissolution of the lead isotope ThB without affecting that of the bismuth isotope ThC. The velocity of exchange between the molecules of solid lead chloride and a saturated lead chloride solution can be determined by labelling the lead chloride with ThB; in the case of a finite layer it is vanishingly small compared with the velocity of dissolution but in the case of a molecular layer the two properties ure commensurable. 1%. KLEMENSIEWICZ, C, R. Accd. Sci., Paris 158, 1889 (1914). Originally published in Phys. Z. 15, 797 (1915) 8. THE EXCHANGE OF ATOMS BETWEEN SOLID AND LIQUID PHASES G. HeEvesy From the Institute for Radium Research of the Academy of Sciences of Vienna When a liquid is in contact with its saturated vapour there will take place, in accordance with kinetic ideas, a constant exchange bet- ween the molecules in the two phases. Correspondingly, it is to be ex- pected that a kinetic exchange of the molecules will likewise occur when a solid phase is in contact with its saturated solution. Radiochemical methods permit an experimental study of this ex- change. For example, if the exchange between the molecules of a solid layer of lead chloride and a saturated solution of lead chloride is to be determined the following procedure is adopted: A known amount, in relative (electroscopic) units, of ThB is added to a solution of known Pb(NO,), content and the whole is precipitated with hydrochloric acid. In accordance with all previous experience the ThB can no longer be removed chemically from such a mixture of PbCl, and ThBCl,; if there is, for example, one atom of ThB mixed with 10! lead atoms on the average in this mixture, this ratio will remain the same after any che- mical operation and if a ThB atom can be detected electroscopically in the lead chloride phase which was previously free from ThB the con- clusion can be drawn that 101° of the lead atoms originally mixed with the ThB have also entered this phase. Thus the ThB or another isotope of lead serves as an ‘‘indicator’” for lead. If solid lead chloride labelled with ThB is shaken with a saturated (unlabelled) solution of lead chloride for 36 hr at 20°C it is found that less than 14 per cent has been transferred from one phase into the other. The determination of the velocity of dissolution of lead chloride having the same grain size showed that 44 per cent of the amount of PbCl, corresponding to saturation passed into solution within 1 hr; since the number of exchanged molecules (expressed as a percentage of all those present) is extremely small it appears that the velocity of exchange of lead chloride molecules between solid lead chloride and the saturated solution of PbCl, is vanishingly small compared with the velocity of dissolution of solid lead chloride. 7 Hevesy 98 ADVENTURES IN RADIOISOTOPE RESEARCH A different result is obtained, however, if a study is made of the ex- change, not between a finite layer of lead chloride and its saturated solution but between a molecular film of lead chloride, which can easily be prepared by a radiochemical method, and the saturated solution. It is then seen that the percentage of exchanged molecules is very con- siderable and the velocity of exchange becomes commensurable with the velocity of dissolution of the molecular layer. This result can be expected from kinetic considerations ; a rapid exchange can take place, in general, only in the superficial layers. A lead rod, 4 em long and 5 mm in diameter, was immersed for 1 min in 10 cm? of a lead nitrate solution labelled with a known amount of ThB, and then the quantity of ThB deposited on the lead surface was deter- mined. This gives the number of lead ions originally in the solution which have thus been transferred to the lead surface. The first column of Table 1 records the normality of the Pb( NO )., the second the number of lead ions per thousand originally in the solution and then occurring on the surface of the lead, the third the amounts of lead in grammes, the fourth the amount, expressed as a fraction, of that required, accord- ing to MULLER and KOENIGSBERGER!, to indicate the potential of lead peroxide. On the basis of the Loschmidt number the mass of a unimole- cular layer of PbO, is calculated? as 3.2 x 10-7 gm. According to the measurements of KOENIGSBERGER and MULLER? twice this mass is re- quired for optical detection and eight times to impart the PbO, poten- tial to an area 1 cm? in extent. TABLE 1 i x Promille of lead Number of molecular layers Normality of the | FE Amount of lead AE “ rot | of the solution | 1 em? in area which can be solution in | exchanged Pb(NO transferred to the (an covered by the amount of PN Os)s | solid phase | aay exchanged lead 10-5 4.3 | aA Om | 0.069 10-4 4.5 a6 xX LO 0.72 10-8 | 3.9 ASG l0me | 6.2 10-2 | 7 3.8 < 10-® | 59 107-1 ez Heth Se 266 1 0.4 | 0.4 x 10-4 | 625 According to our ideas on the process of the galvanic production of current, lead will either go into solution or will be deposited, when a lead rod is immersed in a solution of lead nitrate, according as the con- centration of lead nitrate is on one or the other side of the limit at which lef. W. J. Miuier and J. KornrGsBercerR, Phys. Z. 6, 849 (1905). 2 J. KoENIGSBERGER and W. J. Muxuer, Phys. Z. 12, 606 (1911). THE EXCHANGE OF ATOMS BETWEEN SOLID AND LIQUID PHASES O9 a potential difference of zero prevails between the lead nitrate solution and the metallic lead (absolute zero point of the electrolytic potential), Thus it might be thought that the amount of lead deposited in the above experiments does not represent the exchange between the two phases but is the result of such an unbalanced electrolytic process. The caleu- lation of the amount which can be expected to be deposited at an isolated electrode shows, however, that the value is much smaller than that actually observed. The capacity of the double layer at the metal-electrolyte boundary, from the measurements of KriGER and KrumMReEIcH!, amounts to 27 uF ; the potential difference of this condenser in the case of Pb/Pb(NO,), is always less than + 0.2 V ; thus the calculated charge of the condenser is 5.4 x 10-§C. This quantity of electricity corresponds to 5.6 x 10-9 gm Pb, which is considerably less than the deposit of lead found by experiment. In order also to confirm experimentally that an exchange of atoms between the two phases, and not a one-sided deposition, is involved, the following experiments were performed : A lead rod similar to those used in the experiments already mentioned was coated electrolytically with a layer of metallic lead labelled with ThB, immersed in 10 cm? of lead nitrate solution for 1 min and then, by determining the ThB content of the solution, the number of lead atoms transferred from the metallic phase into the lead nitrate solution was determined. Thus it was possible to establish that while 1.7 x 10-4 gm had depos- ited from 10 em? of a 0.1 M lead nitrate solution on an area of 2 cm2 in 1 min., 1.6 x 10-4 g Pb had correspondingly entered the solution in identical conditions in the experiment just mentioned. In our experiments, therefore, there is indeed an exchange of atoms between the metallic phase and the lead nitrate solution. Because of the magnitude of the amounts exchanged, which in certain conditions amount to one hundred times the unimolecular layer, the process cannot be a pure ‘‘kinetic’” exchange (exchange at complete thermodynamic equilibrium) but involves nonuniformity of the lead surface and pre- cipitation of the lead atoms at particular points from the solution. The observed exchange is essentially a result of ‘local currents’’. The velocity with which the exchange of the lead atoms takes place in the interior of the solid metallic phase can be computed approx- imately since it is equal to the velocity of diffusion of lead in solid lead. The diffusion velocity of lead in mercury according to M. v. WoGcav? amounts to 0.6 cm? hr7! at 18°C ; in solid lead the value is many times 1Krtcer and Krumreicu, Z. Elektrochem. 19, 620 (1913). 2M. v. Woacau, Ann. Phys. 23, 345 (1907). 100 ADVENTURES IN RADIOISOTOPE RESEARCH less since the viscosity of solid lead is very much greater than that of liquid mercury?. The viscosity of mercury is 0.016 at 18°C whereas the value for solid lead, according to KuRNAKOw and ZEMAczNyY?, is 3 x 1012; the velocity of diffusion of lead in solid lead calculated from these figures is 2 x 10-4 cm? hr-!. The minuteness of this diffusion velocity is best brought out by means of the following analysis : Consider a diffu- sion cylinder consisting of four equal parts, each part being surrounded by four molecular layers with an assumed thickness of 0.8 x 1077 em and with the lead atoms of the lowest four molecular layers labelled. After 1 hr there will be less than one per thousand of the labelled lead atoms in the uppermost part of the diffusion cylinder, in the third part only 1.6 per cent. The kinetic exchange during 1 min can extend only to the uppermost molecular layer and to a small extent to the second and third layers. Table 2 records the amounts of lead which have exchanged bet- ween a 0.001 N Pb(NO,), solution and a lead rod in various times. TABLE 2 { | = : | | Number of molecular |; Amount of lead | | | | | Time layers 1 cm? in area aN exchanged | : (sec) | which the exchanged (zm) | lead can cover 15 POSS MO Soll 15 2:3 x 1058 | | Seo 30 | 2.8 x 107—§ 4.4 30 | V285610=6" 4| 4.4 60 le 325 < Om 5.5 60 3.6 x 10—§ 5.6 Table 3 contains the results of such experiments in which a lead roc dipped into 10 em? of a 0.1 N solution of Pb(NOs),. TABLE 3 H | | i | Number of molecular Amount of lead | Time | | layers 1 em? in area which ee | exchanged : (min) : the exchanged lead can (gm) | : cover 1 WS NO 156 10 | alex Om 328 30 32561054 —| 564 ’ How far such an extrapolation is permissible for the solid state will shortly be discussed on the basis of experiments. 4 Kurnakow and Zemaczny, Jb. Radioakt. 11, 25 (1914). THE EXCHANGE OF ATOMS BETWEEN SOLID AND LIQUID PHASES 10} Different behaviour is found in studying lead peroxide surfaces immersed in a lead solution labelled with ThB. In this case the exchange is much less; between a PbO, surface 2 cm? in area and 10 cm3 of a 0.001 N Pb(NO,), solution, containing 0.001 N HNO, and saturated with similarly labelled PbO,, the following exchange takes place : TABLE 4 Expressed in fractions of | | Amount of lead | Expressed as | : ; | | the amount required to Time exchanged | molecular layers | . | ee | impart the PbO, poten- (gm) 1 cm? in area Z : P | tial to the area of 1 em? eve) + wr 1 1/ 10 see ey $2 ae in Is ; » WW = 1 1 1 min PE Se I Use G ie Wen 10 min eOSs< LOms 1.5 1}. 60 min 2.0 x 10-6 Sal he TaBLE 5, — THE TIME OF EXPERIMENT IN THIS CASE IS ALWAYS | min >; THE CONCENTRA- TION OF THE LABELLED LEAD NITRATE SOLUTION VARIES BETWEEN 1071! ann 107 6°N a a RN ESE SS ON SE SS a ar SE | | 5 Z j , Expressed in fractions of Normality of the Amount of lead | Expressed in | tl a Ee t # i % : ie amount required to solution of lead exchanged i molecular layers ara a ae note1 | ' npart the DU Ovel= nitrate (gm) | 1 em? in area | se B ie ee ; ; | | tial to the area of 1 cm 2 Se ——— » ah Te. ee = ES 10-6 e064 54 1058 M0 | tee te NS =s 1 10% Pre Ue) We /48 =) etc Vv eee 1 1/ 10 S00x 10m2 | ye hy 10-1 Daal Om 3.5 iy Here also an exchange rather than a unilateral dissolution is involved as is proved by the following experiments : This time a labelled PbO, surface 2 cm? in area is immersed in 10 cm? of a 0.001 N Pb(NO,), solu- tion, saturated with PbO, and containing 0.001 N HNOg, and it is found that the following amounts of lead (Table 6) have passed into solution from the solid phase : TABLE 6 | Expressed in fractions of Amount of lead | Expressed in . aa the amount required to Time exchanged | molecular layers : | Aaa impart the PbO, poten- (gm) 1 cm? in area tial to the area of 1 cm* 10 sec IES S< al Ome! tf, “lee 1 min 2S" Oe || 1/, SA 20 min OSs lo-°= | 1.2 1/, The experiments just described are made difficult owing to the break- ing off of invisible amounts of lead peroxide which fall into the solution 102 ADVENTURES IN RADIOISOTOPE RESEARCH and are co-determined when the solution is evaporated, thus producing an erroneously high exchange. In the experiments discussed here it was merely assumed that lead and ThB cannot be separated by chemical and electrochemical reactions, as was first proved by Friecx and later confirmed by many authors. If one phase contains on the average 10!° atoms per atom of ThB and if we ean detect a ThB atom in the other phase which was originally free from ThB then, as already mentioned in the introduction, the con- clusion can be drawn that 101° lead atoms also have been transferred from the first to the second phase. Our experiments do not indicate how many atoms have changed places more than once between the two phases. It should be mentioned that when diffusion processes are involved the presence of one ThB atom cannot strictly be taken to imply the accom- paniment by 10!° atoms of lead, since the diffusion velocities of ThB and Pb are not equal. As far as solid and liquid phases are concerned, however, in which the diffusion velocity is very little dependent on the mass, it is practicable to draw the above-mentioned conclusion and, for example, to equate the velocity of diffusion of lead isotopes in lead to that of lead in lead. Summary The exchange of atoms between two phases, for example, between metallic lead and a lead nitrate solution, can be followed by labelling the lead in one phase with one of its isotopes, for example, with ThB; the amount of labelled Jead transferred in a given time into the other phase can then be determined. In the case of Pb/Pb(NO,), the exchange is very rapid and depends mainly on the local currents. At particular points in the metal some lead goes into solution and at other places lead is deposited from the solution. The exchange between a surface of lead peroxide and a lead nitrate solution is much less; in the experimental conditions described in the paper it amounts to only one-third of a molecular layer of lead peroxide in a 0.001 N solution during the course of 1 min. The whole molecular surface layer is replaced only after 1 hr has passed. In using stable lead peroxide the ideal case of kinetic exchange is much more nearly approached — exchange with complete thermodynamic equilibrium between the two phases — than when metallic lead is used. Originally published in Ber. dtsch. chem. Ges. 53, 410 (1920) 9. THE INTERMOLECULAR EXCHANGE OF ATOMS OF THE SAME KIND GerorGE Hervesy and LAszLto ZECHMEISTER From the Chemical Institute of the School of Veterinary Medicine, Budapest THE present study is intended as a contribution to the answer of the question as to whether and when interchange of similar atoms takes place within a molecule and also between neighbouring molecules of like or unlike kinds. In considering a benzene molecule, for example, the question arises as to whether a carbon or hydrogen atom can move by exchanging places with another similar atom from one position to another in the benzene hexagon, or whether a certain hydrogen atom is al- ways bound to the same carbon atom. If two neighbouring benzene molecules are considered there is the further question as to whether carbon or hydrogen atoms which were originally present in the first molecule may or may not be found in the second molecule after a definite time. Such an exchange of positions could be produced either directly by the atoms vibrating within a molecule periodically entering into the sphere of attraction of another molecule, or indirectly in the following way: If there is dissociation such that a hydrogen atom splits off from each of two benzene molecules the dynamic nature of the dissociation process in which the atom is recaptured yields a 50 per cent chance that the hydrogen atom which originally was separated from the first molecule will enter the second molecule and thus be subjected to an exchange of position. Because dissociation and recombination processes take place very rapidly!, even the slightest dissociation in the liquid phase, where molecular collisions occur extremely often, will lead to such exchange in a short. time. Although this question cannot be decided by experiment with benzene, yet this can be done with lead compounds, for example, by means of radioactive methods. It is well known that there are different isotopes of lead which can be distinguished easily and with certainty through their radioactive properties, although they exhibit the same chemical behaviour. By preparing two different compounds of lead, the one from 1M. Le Branc and K. Scuick, Phys. Chem. 46, 213 (1903). 104 ADVENTURES IN RADIOISOTOPE RESEARCH ordinary and the other from radioactive lead, it is possible to distinguish any atom of lead in the one substance from any atom of lead in the other since they have distinct properties. By dissolving the two compounds and after a certain time separating them again a simple measurement of radioactivity will show whether each atom of lead is still in the same kind of molecule as before the experiment or whether an exchange of atoms has taken place. It has previously been demonstrated! that when equimolecular quantities of inactive lead chloride and active lead nitrate are dissolved and the latter subsequently recrystallized half of the active lead atoms originally present in the nitrate molecules transfer to the lead chloride. The originally inactive lead chloride was proved to be half as radioactive after the experiment as the lead nitrate was before. The same result was obtained with the following combinations : Lead nitrate (active) and lead chloride in pyridine ; lead formate (active) and lead acetate in water; lead acetate (active) and lead tetra-acetate in glacial acetic acid ; and lead tetra-acetate (active) and lead acetate in glacial acetic acid. In contrast it was found that there is no exchange of lead atoms when the lead is firmly bound to carbon. The behaviour of organically bound lead is illustrated by the following examples: Lead chloride (active) and tetrapheny! lead dissolved in pyridine ; lead acetate (active) and tetraphenyl lead in amyl alcohol ; and lead nitrate (active) with diphenyl lead nitrate in dilute ethyl alcohol. The original activity or inactivity of the dissolved substances in these instances was not modified by the experiment. The conclusion to be drawn from these findings is that exchange of atoms does not take place even if the lead is undissociated in only one of the two compounds. Exchange is even less likely when this type of binding of the lead exists in both components and particularly when these components are che- mically identical, in other words, when they are molecules of the same substance. There will be no exchange of lead atoms between two mole- cules of tetraphenyl lead. The results up-to-date suggest, therefore, that an intermolecular exchange of atoms (at least in the time required to perform chemical operations) is connected with the existence of an electrolytic dissociation. The existing experimental data are inadequate to prove whether two similar atoms of the same molecule are able to exchange in a measurable time, although it can be assumed probable that the opportunities for positional exchange within one molecule are similar to those arising between two neighbouring molecules. We intend to attack this problem more closely by introducing a radioactive and an inactive lead atom into 1G. Hevesy and E. Rona, Phys. Chem. 89, 303 (1915). THE INTERMOLECULAR EXCHANGE OF ATOMS OF THE SAME KIND 105 the same molecule but with different bonding. A further range of appli- ‘ation for our method might be opened up by splitting off one of the two lead atoms and making comparative measurements of radioactivity on the products. EXPERIMENTAL The radioactive lead was prepared in the following way: The active deposit of a strong radiothorium preparation, which had been recovered from a mesothorium, sample whose activity (gamma) corresponded to 5 mgm of radium, was collected on the surface of a negatively charged lead foil. The activated lead was then dissolved in nitric acid and the resulting nitrate was converted as required into compounds such as chloride, formate, acetate, etc. The salts obtained in this way were labelled radioactively with the lead isotope thorium-B. The activity was measured in the usual way with the aid of an a- electroscope. The substance to be measured was spread on a metallic surface and its activity compared with that of a control substance of the same weight and identical surface conditions. 1. Lead Nitrate (Active) and Lead Chloride in Pyridine An amount of chloride (0.76 gm) and 0.90 gm of nitrate were dissolved completely in 100 gm of boiling pyridine and the solution was kept hot for Y, hr. The lead chloride which crystallized on cooling was filtered at the pump, washed with a little cold pyridine and with ether to remove the solvent, and dried in a vacuum. The sample was quite free from nit- rate. Chloride prepared from the original nitrate was used as a stan- dard. Measurement: 0.268 gm of substance caused the following ionizations : Experimental sample, 2.76 scale divisions per min (calculated for com- plete exchange ¥ x 5.64 = 2.82 scale divisions per min). Standard sample, 5.64 scale divisions per min. 2. Lead Formate (Active) and Lead Acetate in Water A quantity of lead formate (3.00 gm) and 3.83 gm of sugar of lead (KKahl- baum) were dissolved in 25 ml hot water. After keeping warm for 1, hr the majority of the sparingly soluble lead formate was crystallized by cooling the solution; the crystals were filtered at the pump, freed from traces of adhering acetate by washing with alcohol and dried in avacuum. A sample of the active lead formate was used as a stan- dard. Measurement with 0.604 gm of each substance : Experimental 106 ADVENTURES IN RADIOISOTOPE RESEARCH sample, 32.26 scale divisions per min (calculated for complete ex- change, 32.43 scale divisions per min); standard sample, 64.86 scale divisions per min. In a further experiment the standard substance used was the residue obtained by evaporating the mother liquor from the precipitated for- mate. The accuracy of the experiment was sufficient to establish the uniform distribution of the activity between formate and acetate. In this experiment 1.00 gm of formate and 1.30 gm of acetate were dissolved in 20 ml of water. Measurement with 0.134 gm of each substance : Experimental sub- stance, 22.38 scale divisions per min; standard 20.61 scale divisions per min. 3. Plumbous Acetate (Active) and Plumbic Acetate in Glacial Acetic Acid Plumbous acetate (1.60 gm) was dissolved ina little glacial acetic acid and the whole was poured into 50 ml of hot glacial acetic acid containing 2.20 gm of dissolved crystalline plumbic acetate!. The solution, after clarifying by filtration, was kept for 10 min at 80°C, diluted with water to four times the volume and boiled. Measurements were made on the deposited lead peroxide after washing with dilute acetic acid and al- cohol and drying. The residue obtained by evaporating an aliquot part of the filtrate was used as standard. Measurements were made with 0.136 gm of each substance : Experi- mental sample, 20.00 scale divisions per min; standard, 21.95 scale divisions per min. 4. Plumbic Acetate (Active) and Plumbous Acetate in Glacial Acetic Acid The active plumbic acetate was prepared by the addition of an active sample of red lead, obtained by the oxidation of lead monoxide?, to glacial acetic acid. A portion of the beautifully crystalline acetate was used as a standard. Plumbic acetate (1.72 gm) and 1.48 gm of plumbous acetate (Kahlbaum) were dissolved in 15 gm of hot glacial acetic acid to an almost completely clear solution. The plumbic salt which crystallized on cooling the solution for some time was washed with cold glacial ace- tic acid, and dried by pressing and in a partial vacuum. Measurements on 0.400 gm of each substance : Experimental sample, 1.29 scale divisions per min (calculated for the case of complete exchange, 1.43 scale divisions per min); standard, 2.86 scale divisions per min. 1A. Hurcurnson and W. Potxarp, Soc. 63, 1136 (1893) ; Ibid. 69, 212 (1896). See also A. Cotson, C. R. Acad. Sci, Paris 136, 676, 891, 1666 (1903). 2L. Vanino, Prdparative Chemie 1, 488 (1913). THE INTERMOLECULAR EXCHANGE OF ATOMS OF THE SAME KIND 107 5. Lead Chloride (Active) and Tetraphenyl Lead in Pyridine Tetrapheny] lead, Pb(C,H;),, can be prepared in accordance with the description by P. Prerrrer and P. Truskrer!. The compound can be recrystallized from hot pyridine or amyl alcohol. Tetraphenyl lead (1.70 gm) and then 0.92 gm of lead chloride were dissol- ved in 95 ml of pyridine at the temperature of the boiling water bath and, after heating for 14 hr, the solution was cooled to about 35° ©. A mixture of the two substances crystallized from which the lead chlor- ide was extracted by boiling with water. The lead chloride showed the same activity as before the experiment. Measurement with 0.180 gm of each substance: Lead chloride before the experiment, 7.39 scale divisions per min; lead chloride after the experiment, 7.32 scale divisions per min. 6. Plumbous Acetate (Active) and Tetraphenyl Lead in Amyl Alcohol Some acetate (0.70 gm) and 1.00 gm of the tetraphenyl compound were dissolved in 70 ml of hot amyl alcohol and the solution was kept near its boiling point for 15 min. Crystallization of tetraphenyl lead took place gradually on cooling and became complete overnight. The filtered product was thoroughly washed with amyl alcohol, ethyl alcohol and hot water, in succession, and dried in a vacuum desiccator. This sample proved to be completely inactive whereas a sample of the lead acetate prepared as a standard was very radioactive. Measurement with 0.800 gm of each substance: Tetrapheny] lead after the experiment, less than 0.02 scale divisions per min; lead acetate (standard substance), 180.0 scale divisions per min. 7. Lead Nitrate (Active) and Dipheny] Lead Nitrate in Dilute Ethyl Alcohol Diphenyl lead nitrate crystallizing with two molecules of water, viz. (C,H;),Pb(NO,), + 2 H,O, was prepared by the method of A. Po.ts? by adding tetraphenyl lead to nitric acid. 1P, PrerrrerR and P. Truskier, Ber. dtsch. chem. Ges. 37, 1125 (1904); ct. K. A. Horrmann and V. Wourt, Ibid. 40, 2428 (1907). 2 A. Poxts, B. 20, 717 (1887). This information was confirmed by P. PFEIFFER and P. Truskier, Ber. dtsch. chem. Ges. 37, 1125 (1904). — The observation already made by Poms that the formation of diphenyl lead nitrate is disturbed by the appearance of dark coloured substances when the hot concentrated nitric acid is cooled a little below its boiling point is an interesting one. Only the boiling acid can therefore effect the smooth course of all intermediate steps since, other- wise, the reaction takes a different course. 108 ADVENTURES IN RADIOISOTOPE RESEARCH A weight of 1.00 gm of lead nitrate and 1.59 gm of the diphenyl compound were dissolved in 30 ml of 48% hot alcohol in the presence of two drops of dilute nitric acid. Since there was no deposit within 1% hr, the solution was evaporated almost to dryness on a water bath and the crystalline mixture taken up in 30 ml of 95% hot alcohol, lead nitrate then being deposited on controlled cooling. This was treated six times with boiling absolute alcohol and dried in a vacuum. A sample of the original lead nitrate served as the standard sample. Measurement with 0.355 gm of each substance : Lead nitrate before the experiment, 5.55 scale divisions per min; lead nitrate after the experiment, 5.80 scale divisions per min. Summary (1) It has been found that organically bound lead atoms do not undergo intermolecular place exchange in a homogeneous phase. (2) Such place exchange occurs to an extent corresponding to that calculated from probability when the lead atoms are dissociable. (3) Radioactive indicator methods were employed to obtain these results. LO9 COMMENT ON PAPERS 7—9 In paper 7 it was shown that a kinetic interchange takes place between the lead atoms of solid lead chloride and the lead ions of a surrounding saturated lead chloride solution. This problem was later studied in detail by Panreru. He found that the uppermost molecular layer of lead sulphate powder participates only in an interchange process. When investigating the behaviour of natural crystals of lead compounds in several cases he found just a fraction of the lead atoms of the uppermost molecular layer participated in a kinetic interchange, presumably positioned at the edges of the crystal. The other extreme case, an interchange of almost all atoms of a precipitate with those of the surrounding solution, was observed in the case of freshly prepared silver bromide by Langer and by Zimmer. In paper 7 the velocity of dissolution and that of interchange of massive and molecular layers was compared ; among other things, it was demonstrated that the velocity of dissolution of ThB is diminished in the presence of lead ions in the surrounding solution: however, the velocity of dissolution of the bismuth isotope ThC was not diminished. This investigation, carried out between 1913 and 1914, aimed at the demonstration of the identity of the behaviour of isotopes. When extending these studies to an interchange between the lead atoms of a lead foil and the surrounding lead ions, several hundred atomic layers were found to be involved in an interchange process presumably due to a dissolution of more clectropositive parts of the lead foil followed by a precipitation of lead atoms on more electronegative parts of the foil. An adsorption of lead ions on the metallic surface takes place as well, but its role is insignificant compared with the inter- change of lead atoms. An investigation of the behaviour of colloidal lead particles carried out by the author and M. Brurz in 1929 brought out a marked adsorption of lead ions by the colloidal lead particles and a slow interchange only between the lead atoms of the colloidal particles and the lead ions of the liquid phase. In a simultaneous investigation of a system composed of colloidal copper and silver ions (i.e. two metals showing a marked difference in their electrochemical potential) besides some adsorption of silver ions on copper colloids an intense replacement of copper atoms by silver atoms was observed. In papers 8 and 9 interchange of atoms between heterogenous phases was studied. Paper 9 contains a report on experiment aimed at the elucidation if and to what extent interchange of atoms takes place in a homogeneous phase. When dissolving in the same solute non-radioactive tetrapheny] lead and labelled lead chloride, or vice versa, no interchange of lead atoms was observed; this is in contrast to a solution containing 1 mole of non-radioactive lead nitrate and 1 mole of labelled lead chloride ; after subsequent separation by crystallization an equipartition of the radioactive lead atoms was found to take place between the chloride and nitrate of lead. The last mentioned result can be considered to be the most direct proof of the theory of electrolytic dissociation. Reference A. LANGER (1943) J. Chem. Phys. 11. 11 K. ZIMMER (1946) Arhiv f. Kemi, A 21, No 17. G. Hevesy and M. Bintz (1929) Z. Phys. Chem. B 3, 271. Originally published in Ann. Phys. 65, 216 (1921). 10. SELF-DIFFUSION IN SOLID LEAD J. Grow and G. HEevresy From the Chemical Institute of the School of Veterinary Medicine of Budapest We have recently shewn! that the velocity of self-diffusion, that is, the velocity with which the atoms (molecules) of molten lead change places, can be ascertained by determining the velocity with which a radioactive lead isotope spreads in molten lead. Experiments will now be discussed whose purpose is the determination of the self-diffusion velocity in solid lead. The extraordinarily high resistances which oppose place exchange in the solid state led from the outset to the expectation of very slow self- diffusion in solid lead ; we have, therefore, avoided setting up experiments at room temperature and have sought rather to determine the self- diffusion in lead heated and maintained about 40°C below its melting point. Several series of experiments lasting from 1 to 3 months showed that the self-diffusion velocity of lead at 280°C, that is, 46° below its melting point, is less than 0.001 cm?/day. A series of experiments was then performed in which lead filaments, about 2 cm long, were heated for more than 400 days; these filaments consisted, as will be described in detail below, of a 1.5 cm long inactive and a 0.5 cm long active portion of lead. No diffusion of the active lead isotope into the inactive lead could be detected even after this long period of experiment. The self- diffusion constant of the solid lead is accordingly still smaller than 0.0001 cm2/day, even at a temperature of 280°, since values of this order could still have been easily determined in the stated conditions. This result is not without interest, especially when it is compared with the well-known Roperts-AUsTEN experiments?. RoBERTS-AUSTEN allowed gold to diffuse into solid lead and found the diffusion constants recorded in the table below, which also includes our experimental result. Even at 251°C, therefore, the diffusion of gold into lead is at least three thousand times as fast as that of lead into lead at 280°C, the 1J. Grou and G. Hrvesy, Ann. Phys. 63, 85 (1920). 2W. ©. Roperts-Austen, Phil. Trans. 187, 404 (1896). SELF-DIFFUSION IN SOLID LEAD ee T Gold in lead D Lead in lead D CC) (cm?/day) | (em/*day) 100 0.00002 | -- 165 0.0045 -— 200 0.0075 _ 251 | 0.026 = 280 — =< 0.0001 latter temperature being somewhat higher and thus more favourable to diffusion. At this latter temperature, which is not very far removed from the melting point, the self-diffusion is still extraordinarily slow and should be incomparably slower, for example, at room temperature. When relating this result to the velocity of self-diffusion in other metals it must be borne in mind that lead is one of the softest metals and that self-diffusion should presumably prove to be much slower in the harder metals. Quantitative data on the diffusion in solid metals have been provided only by Roprerts-AvstTEN, but metallurgy is plentifully supplied with qualitative experiences which point out the relatively rapid diffusion of alloying solid materials, of which the rapid interpenetration of iron and carbon! supplies at 250°C the best-known example. Thus there exists a very considerable difference between the diffusion of two solid metals into each other and the self-diffusion in a solid metal, in complete contrast to the diffusion in the liquid media. Thus, we have obtained? a value for the self-diffusion constant of molten lead which is only slightly different from the constant for gold in lead. The main reason why self-diffusion in lead is so much slower than the diffusion of gold into lead appears to be that the gold diffusing into the lead loosens up the crystal structure and in this way facilitates its own transmission. It is found that the introduction of an impurity into the crystalline structure can have exactly the same effect as a rise of temperature in facilitating the place exchange of the ions (atoms, molecules). However, it is not stated absolutely that all other metals diffuse more easily into lead than do its own atoms ; we attempted to allow simultane- ous diffusion of the lead isotope radium-D and polonium (which is a homologue of tellurium) into lead, but no positive result was obtained. Diffusion experiments in solid bodies claim so much interest, for this and other reasons, because information can be obtained from the results concerning the magnitude of the resistance opposing the displacement of individual atoms in the crystal structure. Diffusion experiments of 1M. A. Corson, Ann. Chim. et Phys. 17, 221 (1846). 2J. Grow and G. Hevesy, Ann. Phys. 63, 85 (1920). 1 ADVENTURES IN RADIOISOTOPE RESEARCH the Roperts-AustEeN type, however, are not suitable for obtaining the desired information on this point. If it were required to decide upon the slowness of place exchange in solid lead from the Roperts-AUSTEN data a completely erroneous result would be given, whereas the app‘i- cation of radioactive indicators, i.e. the measurement of the velocity of diffusion of a lead isotope in lead yields the required information. EXPERIMENTAL METHOD The layers of active and inactive lead were joined together by the method already described!. The inactive lead was melted in a vacuum in one limb of the Y-shaped hard glass tube and, after it had solidified, the fused active lead contained in the other limb was poured on, thereby pro- ducing a cohesive metallic cylinder. While the active material used in the determination of the velocity of self-diffusion in molten lead was ordinary lead labelled with ThB, this procedure was no longer admissible in the present instance because the ThB decays with a half-life of 10.6 hr and the time of experiment amounted to more than 1 year. Joachimsthal lead, a mixture of ordinary lead, uranium-lead and RaD, has therefore been chosen as the active material. Of these three lead isotopes only the RaD is active and this only to such a small extent that its radiation is not suitable for determining the amount of RaD present; the a- rays of its daughter product, polonium, however, serve aS a convenient means for determining the RaD. Another point in which the experimental method followed here differed from that used for diffusion in the liquid was that, after joining together the layers of active and inactive lead, the boundary surfaces were fused together by means of a finely pointed flame in order to obtain complete contact between the two kinds of lead, this being clearly of great importance for uninterrupted diffusion. A completely cohesive column was thus obtained but of course mixing of the sharp boundaries of the active and inactive lead was unavoidable. In order to take account of this fact, we proceeded as follows : The column of lead, moulded in the manner described, was cut into two vertical sections with a toothed saw and one of these strips was sealed in an evacuated glass tube which was then placed into an electrical resistance furnace. After the experiment the strip was sectioned at three places marked with India ink and was thus separated into four equal parts. The second vertical strip had already been cut at the correspond- ing places before the experiment and was used as a control. If the active laver of lead is denoted by I, then layer II likewise showed some activity 1J. Grow and G. Hevesy, Ann. Phys. 63, 85 (1920). SELF-DIFFUSION IN SOLID LEAD 113 . on account of the mixing at the boundary ; this activity, found even before the diffusion, could, however, be subtracted from that found after diffusion and the mixing at the boundary could be taken into account in this way. Yet the correction described would only acquire importance in the case of an experiment giving a positive result ; since no activity noticeably in excess of the natural decay was found in layers IIT and LV, it was unimportant. In order to be able to measure accurately the a-activity of the in dividual sections we have also used here, as in the experiments which served for determining the self-diffusion in molten lead, rolled sheets of lead and measured the activity of the disk thus obtained in the a- electroscope. The total length of filament amounted to 16—20 mm. The a-activity of polonium indicated by the electroscope is a measure of the amount of the lead isotope (RaD) present only if the RaD and Po exist in radioactive equilibrium. The amount of RaD which diffused in the first four months had come to more than 80 per cent radioactive equilibrium when 14 months had elapsed ; that which diffused in the second 4 months had reached over 50 per cent after the same time, when the measurements were made. The absence of any appreciable activity in the layers III and IV made it possible to determine the diffusion constants of both lead in lead and polonium in lead at 280°C as being less than 0.0001 cm?/day, without awaiting exactly the establishment of the radioactive equilibrium between RaD and Po. It is intended, how- ever, to follow this process further during the next year and thus to be able to extend the observations beyond the determined limits of the diffusion constants mentioned above. We are also concerned in work- ing out other types of method which permit the determination of very much smaller diffusion constants than those mentioned. We may mention here that twelve lead filaments have been prepared as described above, and have been introduced separately into evacuated glass tubes and heated in an electric resistance furnace for 400 days. The furnace temperature, which varied between 270 and 290°C, was followed constantly with a quartz thermometer, since our experience with continuously heated glass thermometers in similar experiments has been unsatisfactory. Summary The self-diffusion velocity of solid lead has been determined by following the diffusion of the lead isotope radium-D in solid lead at 280° C for more than | year. The diffusion constant, even at this temperature which is only 46° below the melting point, is shown to be still less than 0.0001 em2/day. Self-diffusion in lead thus takes place at least two-hundred times more slowly than the diffusion of gold in solid iead at the same temperature. Ss Hevesy Originally published in Nature, 115, 674 (1925) li. SELF-DIFFUSION IN SOLID METALS G. Hevesy and A. OBrRUTSHEVA From the Institute of Theoretical Physics, University of Copenhagen THE “Sagacity” with which atoms, or groups of atoms, oscillating about fixed points in the crystal lattice, refuse to exchange position with neighbouring atoms, is often regarded as one of the chief characteristics of the crystalline state. On the other hand, numerous cases are recorded in which crystalline bodies, for example, solid metals, penetrate into each other, in which, therefore, a replacement of the atoms of one metal by those of the other takes place. The classical experiments of RoBERTS- AustEeN on the diffusion of gold in lead bars are widely known. At a temperature as low as 100° he found the diffusion coefficient of gold in lead to be 2 x 10-5 cm? day~!, being thus only about 100,000 times smaller than that of sodium chloride in water. Several cases of inter- penetration of solid metals have been recorded since, including the interesting case of the diffusion of thorium in heated tungsten wires, reported recently by Lanemure. But it must be noticed that from the rate at which one metal like gold diffuses in another like lead, no con- clusion can be drawn about the velocity with which the atoms change their position either in a bar of pure lead or of pure gold ; no conclusion can be drawn on the rate of self-diffusion in these elements. The idea of self-diffusion was introduced by MAxweE tt, when calculat- ing the rate of diffusion of gases. The calculation was very much simplified by considering the case in which the molecules of the two diffusing gases had the same properties, for example, the exchange of place of molecules in a column of nitrogen. The use of the radioactive isotopes of lead enabled one of the writers, in collaboration with J. Grow (Ann. d. Phys. 65, 216 [1921]) to realise a measurement of self-diffusion in the case of liquid and solid lead, the diffusion in liquids and solids being practically independent of the difference in the masses of the isotopes. For the rate of the self-diffusion in molten lead, namely, of thorium B in molten lead, close to the melting point, the value found was 2 cm? day~!. In the solid metal, however, after heating a bar, the upper part of which was composed of radio-lead, for about a year at 280°, and then analysing the lower part with the electroscope, no diffusion could SELEF-DIFFUSION IN SOLID METALS lld be found. It was, therefore, concluded that the self-diffusion in solid lead is, even at this high temperature, less than 10~4 cm? day~?, To increase the sensitiveness of the method, we prepared in the present work two thin foils, one of ordinary lead, the other with lead containing thorium B in homogeneous mixture, and pressed these together in vacuo. The thickness of the inactive foil was chosen slightly greater than the range of the a-particles to be measured ; therefore no scintillations originating from the radioactive lead could be observed when investigat- ing the inactive foil. But, on heating the aggregate of the foils, a diffusion of the active lead into the inactive one took place and the a-particles due to the diffused atoms or their successive products of disintegration produced scintillations on the observing screen. By comparing the number of these scintillations with the number of scintillations produced by the active foil at the beginning of the experiment, the rate of self- diffusion in lead was determined. The following values were found : 2 Din em? day—* t D in em? day— 260° OX 10s J Ope Oneal Oms © 280° Mey << MOR 320° 4:7 x 1055 S00 2b 10m > S240) V4 10-4 The diffusion rate 2° below the melting point is thus 10,000 times smal- ler than in molten lead. When investigating the diffusion of two very similar metals like silver and gold, or thallium and lead, into each other, we can expect to find conditions not very far removed from those encountered in the case of self-diffusion. By using a foil of thallium and one of active lead it was found that the coefficient of diffusion of lead in thallium amounts at 285°, ¢.e. 15° under the melting point of the latter, to 2 x 10-5 cm? day—!. On the other hand, when investigating the diffusion of two not simi- lar metals into each other, much more intricate conditions were to be expected. We determined the rate of diffusion of polonium, which is the highest homologue of sulphur, into both lead foils and single crystals. The coefficient was found about the same both in the foil and crystal (at 310° D=1.3 x 10-5 cm? day—!). In this connexion it may be men- tioned that, in discussing the discrepancy between the values of the period of decay of polonium found by different investigators, Mme. Cv- RIE has put forward the explanation, that during the long time of observation, the polonium in some cases diffused into the metal from the surface of which it was collected. Recently, Maracinganu (C. 2. 176. 1879, (1923)), working in Mme. Curtr’s laboratory, has obtained evidence that the apparent period of polonium is appreciably shorter if the lead on which it is collected is heated for a while. S* Originally published in Z. Phys. 79, 197 (1932) 12. THE HEAT OF RELAXATION OF THE LEAD LATTICE G. Hevesy, W. Serru and A. Karin From the Institute of Physical Chemistry, University of Freiburg THe heat of relaxation of the lead lattice (the work of release of the lead atoms) is determined from the temperature coefficient of the velocity of diffusion of a radioactive lead isotope in lead; a study is made of the sensitivity to structure of this quantity and the velocity of diffusion. Besides the heat of vaporization, heat of fusion and lattice energy there is another quantity of energy which is important to the crystalline state of aggregation. This is the heat of relaxation of the crystal lattice, or the energy of release of the lattice components. The latter is the energy of activation of self-reaction, which takes place between the atoms of a metal and results in place exchange of the lattice units. When the transport number is known the heat of relaxation of ions of an electro- lytic conductor can be calculated from the temperature coefficient of the conductivity. Direct measurement of the velocity of self-diffusion is the only course open in dealing with metals. The heat of relaxation, Q, is calculated from the diffusion constant, D, measured at various temperatures, by using the well-known formula b= Aen Wk? where A is a constant which is practically independent of the tempera- ture. In the measurement of self-diffusion it is usually necessary to follow the speed of mixing of two closely related metals, such as gold and silver or tungsten and molybdenum, which is then equated as a first rough approximation to the speed of self-mixing of one component. The velocity of self-diffusion in lead can be determined accurately, without such an uncertainty, by measuring the velocity of diffusion of a radioactive lead isotope in ordinary lead. Previous experiments of this kind have already been described!. This paper will deal with the result of an investigation carried out recently with the object of determining the heat of relaxation of the lead lattice and of ascertaining how far this quantity and the velocity 1G. Hevesy and A. OsprutscHewa, Nature 115, 674 (1925). THE HEAT OF RELAXATION OF THE LEAD LATTICE 77 of self-diffusion in lead are structure sensitive. The method used has been described in detail on a previous occasion!. The lead isotope ThB is condensed on a lead surface and the ionization caused by the a-radia- tion of the radioisotope (or its decay products) is measured before and after the course of diffusion. The deeper the thorium-B penetrates by diffusion into the lead, the greater is the absorption of the a-radiation and the smaller will be the ionization arising from it. The calculation of the velocity of diffusion of thorium-B in lead, which is the same as the velocity of self-diffusion of lead, is executed by means of a formula developed by R. Ftrru?, which correlates the ionization values before and after diffusion, the range of the a-radiation in lead and the time. In a more sensitive modification of the method which has also been described the recoil yield, instead of the ionization due to the q-radia- tion, is measured before and after diffusion. Both **Kahlbaum” lead and lead of very high purity, which was kindly made available by the Akkumulatoren-Fabrik A. G., Hagen, West- phalia, were used in the experiments. The lead was freed from its content of gases by prolonged fusion in a vacuum and was purified from oxide content by passing it through a capillary system. The lead single crystals were prepared by the method of Kyropoutos and were characterized, as compared with crystalline lead, by their remarkable resistance to oxidation by the air. All the experiments described below were carried out with complete exclusion of air either in an atmosphere of nitrogen or in a vacuum, and the small tubes containing the single crystals were broken in the evacuated apparatus. All the results recorded in Table 1] can be represented by the equation : Or == 516 << A10°e ie log D = 5.76 — 04343 (14025/7) or by a straight line (Fig. 1). Thus Q amounts to 14025 R = 27870 cal/mole and A = 5.76 x 105 and, within the limits of experimental error, there is no difference between the behaviour of the single crystals and the crystalline material. An investigation was then made as to whether destruction of the texture at the surface of the single crystal by a milling machine, with the specimen necessarily being exposed to the air for a short time, has a measurable effect on the self-diffusion constant. No marked effect on the velocity of diffusion due to this manipulation could be found. Table 2 shows the results of these measurements. The values thus obtained can also be represented by the above equation and by the straight line applying to the unworked material. 1G. Hevesy and W. Serru, Z. Phys. 56, 790 (1929); Ibid. 57, 869 (1929). > R. Firrx, Handb. d. phys. u. techn. Mechanik 7, 687 (1930). 118 ibe SreiLF-D TABLE ADVEN’' PURES IN RADIOISOTOPE RESEARCH IFFUSION IN Leap Sryente Crystats Dirruston constant (D) or ThB in Leap (Sevr-Dirruston Constant oF Lrap) No. | re | ees | (1/7')108 log D Remarks? 1 182 4.12 - 10-8 2197 7-39 2 196 Bee 108 2131 7004 3 207 Se SIMS 2083 —7.09 Crystalline lead 4 222 2.45'- 10-7 =) 72020 —6.6 Single crystal from 5 238 Tse Ome 1956 E61 themmnelt 6 245 O16, 10—7 1930 —6.0 7 245 6.57. -1057 1930 —6.18 Single crystal from the melt 8 258 L1--10-© | 1888") —5.96 9 | 259 3.4 -10-§ | 1879 By 10 | 268 | 23 -10-* | 1865 —5.64 11 275 6.0 -10-§ | 1824 5.22 12 290 TAOS Ae PTT 5.15 13 301 1.6--=10—5 1742 =rAn7 9 14 312 | 1.62-10-5 1709 4.19 15 317 2.82 - 10=5 1695 —4.55 — Single crystal from the melt 16 322 2.36 - 1075 1680 —4.63 | Crystalline lead 17 324 4.78 - 10-5 1674 AOD 14 1 [In all cases where there is no remark single crystals grown by the Kyropoulos method were used: in items 6, 8 and 13 these consisted of Kahlbaum lead and in all other items of lead from the Akkumulatorfabrik, Hagen, Westphalia. TaBLE 2. — Dirruston Constant oF ThB tn Cotp-WorKED LrEap ia ae eee Scam ace. (cm*/day) | oe aes eet = —— vs —— fc = = = => a 18 196 AAS 210-8) | 2130 —7.38 Milled single crystal 19 217 Ze tcl Ome! 2040 —6.54 Milled single 20 233 Aro Om! a 1976 —6.34 Milled crystallite 21 237 6375s- 10m | 1960 —6.17 _ Milled single crystal 22 254 2:99) 10-2 | 1897 == 564. Milled crystallite 23 270 Ania Ome® 1841 —5.34. Tempered crystallite The results discussed above relate to a temperature range which extends from the melting point of lead (327°) to 182°. Below the latter temperature the self-diffusion can be followed by making use of the recoil effect. As has al ready been mentioned, this method does not make use of the decrease in ionization after diffusion but the recoil yield is determined. Whi le the range of this q-radiation in lead is 3 x 10-3 cm the range of the recoil particles extends only to a thickness of about one hundred atom layers (4.7 * 1076 em). In corresponding degree to THE HEAT OF RELAXATION OF THE LEAD LATTICE 119 the shorter range of the recoil particles the latter method is indeed considerably more sensitive than the method first described. In the study of the self-diffusion of lead ions in lead iodide it has been found possible to determine that the measurement of a-radiation and recoil yield the same result ; in spite of the fact that the recoil measurements register processes in the top few hundred molecular layers they were 27 «6 OS a4 a de af 20 19 +16 #«17*~« Tr. 194 Fig. 1. Rate of Diffusion of labelled solid Lead. Schmelztemperatur— Melting point. as highly reproducible as the a-measurements. Recoil measurements at lead surfaces, on the contrary, indicated a high sensitivity of the metallic surface to external effects. For example, contact of the lead sampk with air for a short time was sufficient to lower the diffusion values and even the values obtained in a carefully purified nitrogen atmos- phere were rather lower than those determined when working in a vacuum. In spite of the uncertainty arising for these reasons in the recoil values, the experimental points obtained by this method also lie approximately on the straight lines obtained with the a-measurements as the basis (cf. Fig. 1); it must also be borne in mind that the recoil range in lead is not accurately known and that it must be calculated by extrapolation from the values measured in air. The temperature coefficient of the velocity of diffusion is in quite good agreement with the results of the already mentioned preliminary experiments, but the difference in behaviour of single crystals and crys- 120 ADVENTURES IN RADIOISOTOPE RESEARCH TasLe 3. — Dirrusion Constant (D) or THor1tum B IN Leap DETERMINED BY THE Recor, MrerHop ee ae Snes | Cra log:D ps ee eared 1 106 1.45 - 107-1 2640 10.84 Nitrogen 2 113 Nee 4 oo Oa 2590 9.85 Vacuum 3 114 2.60 - 10712 2584 —10.59 Nitrogen 4 20, Ooms t0 2544 = 9372) | Vacuum 5 28 = 235 elo 10 2494 — 9.63 Vacuum GF S29 Ora O28 2487 | — 9.71 Vacuum eA Son 3 Ale alO- 20 2444 — 9.47 Nitrogen BE |S 7 ale e459) 1020 2438 — 9.34 Vacuum 9 Weyl |p Bolo Tey 2438 — 9.47 Nitrogen LOS i) AT Ne AOS Oa, OAS — 9.97 Vacuum 11 1a a yee 2 tO 20cm O87 NT ie Or40 Vacuum 12 153 een PS <2 oll Ome LO | 9346 °°| 648 | Nitrogen tallites demonstrated in those experiments could not be reproduced, possibly because the single crystals in the preliminary experiments were unavoidably exposed for a long time to contact with the air (during counting of the scintillations). The present study shows much more pointedly that neither the heat of relaxation of the lead lattice nor the velocity of self-diffusion of lead atoms is structure-sensitive. This result is possibly connected with the ready recrystallizability of lead since in the molybdenum—tungsten system van LremMpT! was unable to find a structure dependence of Q, yet A and therefore the diffusion constant showed such dependence. He found A to be about eight times as large in polycrystalline material as in a single crystal, by measuring the veloc- ity of diffusion of molybdenum in tungsten, and even earlier a structure sensitivity of the electrolytic conductivity, which is closely related to the self-diffusion, had been demonstrated in salts?. VAN LigmMpT calculates the constant A from the equation A= 22°06 where x is the distance between lattice planes and y the vibrational frequency of the atom. In the case of molybdenum diffusing in tungsten single crystals he finds remarkably good agreement between the observed and calculated values of A. The value of A which we have measured in lead is, on the contrary, about one thousand times the value calculated 18. A. M. van Liempt, Z: anorg. Chem. 195, 366 (1931); Rec. Trav. Chim. 51, 114 (1932). 2G. Hevesy, Z. Phys. 10, 80 (1922); G. Tammanwn and G. VsEszi, Z. anorg. Chem. 150, 355 (1926); T. E. Purprs, W. D. Lansine and T. G. Cooks, J. Amer. Chem. Soc. 48, 112 (1926), etc. THE HEAT OF RELAXATION OF THE LEAD LATTICE [Ay by vAN Liempt’s method. The self-diffusion (self-reaction) in lead accord ingly represents, at least formally, an example of a chain reaction in which the chain length is independent of whether single crystals or polycrystalline materials are involved. The magnitude of the heat of relaxation is compared, in Table 4, with the energy content and the heats of fusion and vaporization. Taste 4.— Hear Properties or LEAD keal/g atom Heat of fusion Banish chou centor ean seer trcttey «rere ee ile Energy content at the melting point... 3.0 Flicatimoti relaxations ys steterem ceteris sesseksets) os « 27.9 Meath oh WwapOriZation! 92k yls ssa uk ake cla 36.2 Table 4 shows clearly that the heat of relaxation is not very mucl: less than the heat of vaporization but that it is very much greater than the energy content at the melting point and the heat of fusion. Summary The heat of relaxation of the lead lattice (heat of activation for the self-reaction of the lead atoms) amounts to 27,830 cal/mole. This quantity, like the constant A of the diffusion equation, is only slightly structure sensitive. Originally published in Z. f. EHlektrochem. 37, 528 (1931) 13. DIFFUSION IN METALS G. Hevesy and W. Sriru From the Institute of Physical Chemistry, University of Freiburg DIFFUSION in Salt-like compounds is facilitated by means of a relaxation process which occurs when the crystal is heated. This relaxation which exhibits some similarity to activation in the theory of reaction velocity depends chiefly on the size, valency, electron affinity and polarization properties of the lattice components. In silver iodide, for example, where the small univalent strongly-polarizing silver ion contrasts with the large iodide ion which has slight attraction for electrons and is easily polarizable, there is easy detachment of the silver ion and it is well- known that mobilities indeed exceed those occurring in aqueous solutions. In the pure metal the behaviour is altogether different. In such case there is only one kind of lattice component, a high co-ordination number and a high symmetry of charge distribution. Large diffusion velocities cannot therefore be expected in pure metals. Metal alloys are different. In the lead-gold system, for example, the small gold atom which has a high affinity for the valence electron contrasts with the larger lead atom which has a lower electron affinity, and hence there occurs a system which is readily subject to relaxation in which the gold atom easily vacates its position. Rosperts-AusTEN in his classical investigations has already been able to demonstrate that gold diffuses into lead even at moderate temperatures with a considerable velocity. The velocity of diffusion of gold in lead is attained through the speed of dissolution of gold atoms in the gold-lead phase and in its taking up a new position. The gold penetrates into lead but, on the contrary, lead is practically immobile in gold. The diffusion constant for gold in lead is 4x10°3 cm? day! at 150° whereas the diffusion of lead in gold-lead at a temper- ature of 141° amounts only to 3x10-! cm? day, i.e. it is smaller by seven orders of magnitude. Lead atoms accommodated near silver atoms are more easily dissolved than those considered above. It has been found that lead diffuses about twice as quickly in silver saturated with lead as in pure lead. If the gold in lead alloys is replaced by other elements whose properties become more and more similar to those of lead, then these elements show a DIFFUSION IN METALS 125 lr steadily decreasing diffusion velocity. The margin between the diffusion velocity of the added element and that of lead thus becomes steadily smaller and the unilateral diffusion becomes gradually less apparent, Considering now an alloy of ordinary and radioactive lead, both con- stituents will exhibit the same diffusion velocity. This is an example of self-diffusion and therefore a complete mutual replaceability in the 140 160 180 200 220 240 260 280 300 320 Diffusion in Bleilegierungen Schmelzpunkt des Bleis ; gS 10" "Tie ales fs Raa Ey > ee Os (a lf 17 Fie. 1. Diffusion in Bleilegierungen — Diffusion in lead alloys Schmelzpunkt des Bleis — Melting point of lead diffusion process. The gradual decrease of diffusion velocities of metals in lead in the sequence, Au— Pb, Pb—Pb, and also the step-wise increase of the heat of relaxation (heat of activation), can be seen in Fig. 1. The latter quantity is also recorded in Table 1. The marked preferential diffusion of one component is also encountered in salt-like compounds but there is no mutual replaceability of the components which, indeed, is a characteristic of metallic systems. The silver in silver chloride can only change places with silver; it is other- wise in metallic alloys. Considering the place-exchange processes in a saturated silver-lead phase, the readily detachable silver atom will leave its place with great frequency in unit time. The silver atoms intermit- tently seek out the lead atoms and occasionally also replace other silver 124 ADVENTURES IN RADIOISOTOPE RESEARCH TABLE 1 ———— ow ie eer an Au 30 Ome ~ 14,000 Ag 226 Om 15,000 Bi 3.2 - 1055 18,500 Au 1.9 - 1075 18,500 Sn |} 4.4-10-& | ¢. 28,000 Bb | Weegee | ¢. 30,000 atoms. The lead atoms on the contrary scarcely ever leave their places. At room temperature, a lead atom in pure lead changes its place once in 10 days while the silver atom in lead-silver alloy vacates its position 100 times per sec.! As the temperature is raised the difference diminishes. We have already been able to demonstrate, in an earlier paper?, in the case of silver alloys that the speed of place exchange of the atomsde- creases as the ideal metallic state is approached. The last step, however, i.e. the measurement of self-diffusion in silver, was not practicable in the systems mentioned. We have therefore applied our attention chiefly to lead alloys because they presented an opportunity for determining also the self-diffusion in lead. In determining the frequently very low velocity of diffusion, use was made of quantitative optical-spectroscopic analysis which has been found especially suitable for this purpose. Indeed this methods permits both the determination of very small amounts of metal and the use of very thin and strictly localized layers in the analysis. Samples of known composition are first prepared, e.g., lead-thallium alloys with the concentration varying between 3 and 0.01 per cent and the ratio of the intensities of suitable lines of the two elements is determined. Layers, each 0.05 mm thick, are cut after diffusion and these are analysed spectroscopically by comparing with the test samples in accordance with the method described by GERtacu?. The diffusion constant can be calculated since the concentration, as a function of the distance from the original boundary, is known. In the method described it is above all necessary to ensure that the metal whose diffusion constant is to be determined is distributed as atoms in the lead. If this does not apply then the analytically determined concentration is not identical with 1 Concerning the calculation of the frequency of place exchange of an individual atom, refer to H. Brauner, Z. phys. Chem. 110, 147 (1924): J. FRENKEL, Z. Phys. 35, 652 (1926): J. A. M. van Liempt, Z. anorg. Chem. 195, 366 (1931). 2G. Hevesy, Z. Elektrochem. 34. 463 (1928). 3.W. Geritacw and E. Scuwerrzer, Chemical Emission-Spectrum Analysis, Leipzig (1930). More details of the application of this method to measurements of diffusion will shortly be described by GuENTHER and Larrp. DIFFUSION IN METALS 125 the concentration considered in the diffusion process!. The silver-lead and gold-lead systems provide examples of the behaviour just referred to. Starting with a concentrated silver-lead alloy, in which the silver is as finely divided as possible, then the silver which has migrated by diffusion will constantly be replaced by dissolution from the grains of silver. The system therefore consists of a constantly saturated solution of silver in lead from which silver diffuses into pure lead. As will be seen later, the solubility of silver in lead can be determined by means of this behaviour. In only a few cases have we used an alternative to the spectroscopic method. The silver content of lead alloys has been determined, for example, by the usual method of volatilizing the lead and assaying the residual lead of silver. It was also necessary to employ radioactive methods in order to determine the self-diffusion rate. At higher temper- atures, where the selfdiffusion is already somewhat larger, the diffusion constant could be determined as follows: A radioactive lead isotope is condensed on the surface of inactive lead and the ionization due to the a-particles emitted by the radioactive lead is measured. The system is then heated to the experimental temperature and the ionization, which now has a lower value owing to the diffusion which has taken place, is measured again. The self-diffusion constant can be determined from this decrease due to diffusion. At a lower temperature this method, which indeed is very sensitive, proved not sensitive enough and had to be replaced by another. The recoil yield before and after diffusion was measured and the diffusion constant was calculated from the decrease by means of a formula developed by R. Firru?. While the first method enables the diffusion constant to be determined down to 1077 cm?/day, the use of the second method permits diffusion constants of 10-13 cm? day or less to be obtained. The solubility of a metal in a solid phase can be determined from the analysis of diffusion since only those particles of the metal which are distributed as atoms are involved in the diffusion process. The dis- cussion is best illustrated by an example. (a) Starting with a concentrated alloy the diffusion is allowed to proceed until all parts of the originally pure layer of lead are saturated with the diffusing metal. Since an increase in the diffusion time is no longer accompanied by an increase in concentration there now exists a saturated solution the analysis of which yields the solubility directly. At 288°C, for example, it is impossible to produce by diffusion a silver- lead alloy containing more than 0.13 atomic per cent of silver. 1 cf. G. GrusBE (Z. Metallk. 19, 438 [1927]), who has determined a series of diffusion velocities in high-melting metals. With regard to the problem of diffusion in alloys refer also to the many publications of G. TammMann and his school. 2 R. Firra, Handbuch der physikal. und techn. Mechanik 7, 687 (1930). 126 ADVENTURES IN RADIOISOTOPE RESEARCH This method, of course, requires long experimental times and a time of 3 months is needed for homogenizing a 5 mm layer even with silver which diffuses comparatively rapidly in lead. We have therefore often used the following consideration for determining the solubility : (b) Disregarding the initial layer (lead-silver alloy) at first, the silver concentration is determined at various positions after diffusion in the originally pure lead and the diffusion constant is calculated from these results. Now if the diffusion constant is known the corresponding con- centration of silver can be calculated and this in turn yields the solubility of silver in lead. Starting with a silver-lead alloy of 10 atomic per cent atomic per cent 0.5 atomic per cent we obtained the following values for the solubility of silver in lead 0.15 atomic per cent 0.12 atomic per cent 0.13 atomic per cent Hence, starting with silver-lead alloys of various concentrations, from which the same solubility values have been obtained, it may also be concluded that the silver from the macroscopic grains of silver was supp- lemented so rapidly that the original lead-silver alloy always remained saturated. If a diffusion experiment is performed, commencing with a 0.1 atomic per cent silver-lead alloy, the same diffusion coefficient found by the method described above is obtained both from the concentration of the silver in the initially pure lead layers and from the concentration decrease of the silver in the original silver-lead alloy. The method descri- bed for determining the solubility is important in so far as there are at present no other methods available for determining very low solubilities of one metal in another!. Summary The diffusion of one metal in another solid metal is in most cases an almost unilateral process. For example, the velocity of diffusion of gold in lead is very rapid but lead diffuses extremely slowly in gold. As the two alloying components become increasingly similar, for example, in passing from gold-lead, silver-lead, bismuth-lead, thallium-lead, tin-lead to lead-lead, the one-sided nature of the process gradually disappears. Diffusion measurements make it possible to determine very low solubilities of one metal in another. The solubility of silver in lead at 285°C was thus found to be 0.13 atomic per cent. 1 In the literature silver is stated to be insoluble in solid lead. Originally published in Phys. Z. 56, 790 (1929) 14. THE APPLICATION OF RADIOACTIVE RECOIL IN DIFFUSION MEASUREMENTS G. Hevesy and W. SErrH From the Institute of Physical Chemistry, University of Freiburg A bayer of thorium-B chloride placed on the surface of PbCl, shows a decrease in a-recoil yield after heating. The velocity of diffusion of the thorium-B ion in lead chloride and thus the velocity of self-diffusion of the lead ions can be determined from this effect. This extraordinarily sensitive method by means of which diffusion constants down to 10-8 em? day—! can be determined permits the measurement of the velocity of diffusion in PbCl, and PbI, in the vicinity of 100°C or at a higher temperature. Two different cases must be distinguished in diffusion in crystalline substances, heterogeneous diffusion and _ self-diffusion. The difference between these two cases exists also in other states of aggregation but is only slightly perceptible in the liquid and gaseous forms. During hetero- geneous diffusion in crystalline substances individual lattice components are replaced by foreign particles or the foreign ions (atoms) intrude into the interstices of the lattice. In self-diffusion the lattice components are replaced by identical particles. Considerable affinities between the diffusing and lattice-element substances often operate during heteroge- neous diffusion and we are confronted with a process which is a combi- nation of a chemical reaction, often proceeding with a significant decrease in entropy, and a true diffusion process. Self-diffusion produces merely a positional mixing of the lattice components without any practicable change of entropy. The phenomenon of self-diffusion is employed when informa- tion is required on the strength of binding of the individual ions (atoms) in the crystalline compound. In such measurement the method often used is to study the diffusion of an ion, e.g. a cation, in the crystalline compound whose cations are closely related to the diffusing one. For example, the diffusion rate of cuprous ions in silver salts may be measured or that of the cuprous ions in silver salts the cuprous and _ silver ions being considered to be nearly identical from the standpoint of diffu- sion. The binding strength of silver and cuprous ions in various compounds can be determined to a fair approximation from such measurements. On the other hand it is not possible to determine the 128 ADVENTURES IN RADIOISOTOPE RESEARCH binding strength of e.g. iodide in silver iodide, by similar measurements, and it must suffice to conclude from TuBANpb?’s! transport measurements and G. G. Scumrpt’s ionic emission experiments that the binding strength of the iodide ion in silver iodide is considerably greater than that of the silver ion. The indication of the last-mentioned result calls to mind another method for determining the velocity of self-diffusion, viz. by calculating from the electrolytic conductivity of the crystalline compound by using the theory propounded by Nernst for electrolytic solutions or by means of the Einstein diffusion equation. This method also, however, yields only the velocity of self-diffusion of the lightly bound ions, that is, silver and cuprous ions in silver and cuprous salts. The ideal of self-diffusion can be approached extraordinarily closely by allowing a radioactive ion to diffuse in the appropriate compound of an isotopic inactive ion, by applying the method of radioactive indicators, for example, by measuring the diffusion of ThB ions in lead chloride. Now the ions of the radioelements, except those of the thalium isotopes which are too short-lived (half-life always less than 5 min) to be considered, are multivalent. Multivalent ions, however, are always characterized by particularly strong binding?. From this it follows that the self- diffusion can be measured by the method sketched out above only with the aid of an extremely sensitive arrangement. The values thus obtained should indeed yield data on the binding strength of even this ion which has practically no share in the electrolytic conductivity and whose velocity of self-diffusion cannot therefore be calculated from conductivity data. With the usual measuring apparatus the procedure of STEFAN is followed by placing several, frequently three, equally thick layers of the diffusion medium on a layer of the diffusing substance. The diffusion constant is inversely proportional to the square of the layer thickness. The smaller the diffusion constant to be measured the less will be the chosen layer thickness. If a velocity of diffusion (D) of, for example, 10-8 cm? day~! is to be measured then, for an experimental time of 1 day, it is necessary to choose a layer thickness of about 0.01 mim. It is not practicable, however, to place three equally thick inactive layers on a 0.01 mm thick layer of radioactive lead chloride and to separate them again after diffusion for the purpose of radioactive anal- ysis. On the other hand the various radioactive methods yield opportuni- ties to attain such small layer thicknesses in another way. One of the authors with OprurscHEWA? has determined the velocity of self-diffu- sion of lead atoms, both in single crystals and in crystalline lead, by 1C. Tuspanpt, H. REINHOLD and W. Jost, Z. phys. Chem. 29, 69 (1927). 2 Compare the transport results of TuBANpDT, Z. phys. Chem. 29, 69 (1927) : see also E. Friepricu, Z. Elektrochem. 32, 576 (1926). 3G. Hevesy and A. OsrutscHewa, Nature 115, 674 (1925). THE APPLICATION OF RADIOACTIVE RECOIL IN DIFFUSION MEASUREMENTS 129 collecting atoms of the lead istotope on a lead surface and counting the number of scintillations shown by the infinitely thin radioactive coating before and after diffusion in the heated metal. The decrease in the num- ber of scintillations is a measure of the velocity with which the radioactive lead atoms have passed, because of diffusion, outside the range of the a-rays and into the interior of the metal. The layer thickness here required for the diffusion calculation is the range of the a-rays in lead, which amounts to about#/,, mm. Diffusion constants down to 10-8 cm? day! were measured with the help of this method. This sensitivity, however, was inadequate in the present study and we therefore used the radioactive recoil effect which can be expected to yield a considerable increase in sensitivity for detection of the diffusion. The range of a-recoil in lead amounts to only 3x10-> mm and thus the application of the recoil method permits measurement of the extraordinary small diffusion constant of 10-13 em? day~!. The velocity of diffusion of the ions of the lead isotope ThB (half-life 10.6 hr) was measured in different compounds lead. The recoil yield thus the activity of ThC, was determined before and after diffusion. The radioactivity measurements were made only after the establishment of the equilibrium between ThB and ThC, since the a-rays responsible for the recoil effect are derived not from thorium-B but from its daughter product thorium-C (half-life 1 hr). EXPERIMENTAL METHOD The radioactive substance was condensed from the vapour phase on the pellet to be used for measuring the velocity of diffusion. The pellets were prepared by pressing very carefully purified lead chloride or iodide. The pressure used was 1800 kgm/cm? and was applied for 1 min. The pellet was pressed on to the front of a 14 mm diameter brass cylinder and the two together were suspended in the apparatus (Fig. 1). This apparatus consists of two chambers A and B connected by means of a cock with a 2 em bore. Each chamber can be separately evacuated and filled with purified nitrogen. The brass cylinder with the pellet is fastened by means of a silver wire to a screw, C, situated above the chamber A. A phosphorus pentoxide tube is first attached to the standard joint. S, and the whole apparatus is filled with nitrogen. The ThB chloride is condensed on to the lead chloride surface in the following way: the active deposit from thorium is collected on a plati- num foil and the foil is then exposed to the action of chlorine gas. The foil is transferred into the vaporization apparatus, which is attached at S (see Fig. 1) while the pellet is situated in the chamber A. The vapor- ization chamber consists of a brass tube K, which can be cooled, into which the brass cylinder with the pellet just fits. The lower joint of 9 Hevesy 130 ADVENTURES IN RADIOISOTOPE RESEARCH this tube is connected with a matching glass joint, which contains two brass rods for the power supply and across the ends of which the activa- ted platinum foil is fastened horizontally. The chamber B and _ the vaporization chamber are evacuated and filled with nitrogen before the pellet is lowered from A until it is near the top of the platinum foil GC. At a nitrogen pressure of 1 mm the foil is brought to a white heat (about 900°) for 1 see and thus the vapour of ThB chloride is transferred on to the surface of the pellet. inceals The recoil product, ThO’’, emits f-rays whose strength constitutes an easily traceable measure of the recoil yield. The recoil product is collected as follows : After first waiting until radioactive equilibrium has been attained the pellet is placed in a tube attached at S above a copper foil charged to — 220 V, the pellet being earthed, and the recoil atoms are collected on the copper foil. A decrease in the pressure to 2 em aids the collection of the recoil product. Measurement of the f-activity of the copper foil gives the recoil yield before the diffusion. THE APPLICATION OF RADIOACTIVE RECOIL IN DIFFUSION MEASUREMENTS 13] The pellet can be brought up to the experimental temperature for a definite time by attaching at S a furnace containing a glass vessel with a standard joint. In the glass vessel is a hollow iron block (1.5 kgm) into which the pellet and the brass cylinder can be introduced such that the direct contact with the metal facilitates rapid equalization of the temper ature. The temperature is measured by Hoskins’ method by means of a thermoelement of high thermoelectric power fixed in a side hole in the iron block. This apparatus also is filled with nitrogen. The lead chloride was prepared from Kahlbaum purest lead chloride by repeated recrystallization from hydrochloric acid solution. It was dried by heating nearly to the melting point in a current of HCl. The lead iodide was obtained from an active acid solution of HI and Pb(NO,),, purified by decantation and dried over P,O,;. Sublimation was avoided since PbI, prepared thus always contains traces of iodine. CALCULATION We are indebted to Professor R. Firru of Prague for the formulae used below. The calculating procedure is as follows: If the recoil activity before the experiment is equal to 1 and after the experiment to A, then 1—A atoms of lead have diffused so far in to the pellet that their recoil products are no longer able to leave it. If all the recoil particles moved perpendi- cular to the surface, then all those issuing from lead atoms which had not diffused deeper than the range, a, of the recoil particles would be able to escape from the surface. The relationship between the number A and the diffusion constant D is therefore aemror @ (1) where Z is the time and z is the distance of the particle from the surface. By using the Gaussian error function 9 2) = =| Cn a7 (2) ! 0 the equation A Sane | (3) "\2/pz is obtained, whence D can be calculated. g* 132 ADVENTURES IN RADIOISOTOPE RESEARCH It must be borne in mind that in the present instance the recoil par- ticles are expelled in all directions, such that the particle can reach the surface only when the distance x of its starting point from the surface satisfies the condition wat < cos a (4) where a is the angle between the normal to the surface and the ray. The ratio of the number of particles reaching the surface from a point C to the total number of particles issuing from that point in all directions is equal to the ratio of the surface of the spherical cap of height a—a to the surface of a sphere of radius a, whence it follows that 27a(a — 2X) a =a = 1 (5) ate (IE a To take account of this concept the integrand in equation (1) must be multiplied by (1—a/a), and thus A 7 Dee Cae ana: (6) ) \(aDZ) a or, by substituting & —a/2\(DZ) (7) the result is 1 A = y(é) — —— (l—e-*) 8 p(s en ( (8) This equation is evaluated graphically and from the value of € thus obtained D is calculated by means of equation (7). Since the diffusion constant has been calculated in some cases from the decrease in the a-ionization it may be appropriate to discuss the calcula- tion for that method. The a-activity was measured by selecting a parallel beam normal to the pellet surface such that equation (1) could be applied. The shutter had an air gap of 5.38 em and thus only the a-rays of ThC”’, with a range of 8.4 em, were able to penetrate. The conditions for the calculation were thus simplified. In measuring D by means of qa-radiation it must be noted that the particles entering the electroscope do not all have the same ionizing effect since this is dependent on the path already travelled by a particle. A particle which has come from the interior of the pellet has less effect than one which has started from the surface. If the decrease in ionization due to a particle which has travelled a distance x in the Pbl, pellet in relation to the effect of one which has started from the surface is represented by J = q(x) (9) THE APPLICATION OF RADIOACTIVE RECOILIN DIFFUSION MEASUREMENTS 133 and if also the retarding effect due to the air column of the shutter is equal to that of a PbI, layer with a thickness b, then a—b : ] A = y(é) — — w(x) e774 22 . dx (10) : } V@DZ) 0 In the above equation — = (a—b)/2\/(DZ) (11) and the expression is evaluated graphically. THE DIFFUSION OF LEAD IONS IN LEAD CHLORIDE If an attempt is made to determine the diffusions of lead ions in lead chloride with the ordinary apparatus of STEFAN, by pressing together a 3 mm deep inactive lead chloride pellet and a 1 mm deep ThB-labelled lead chloride layer, no diffusion of the radiactive lead ions can be detected after 4 days at 480°C. It is not feasible to raise the experimental temperature since the high vapour pressure of lead chloride at the above-mentioned temperature (10-1 mm) already causes disturbance. The interference can indeed be partly restrained by carrying out the experiment in a pressure bomb under a nitrogen pressure df 200 atm, but cannot be wholly eliminated. For the reasons mentioned it also seemed hopeless to prolong the duration of the experiment, possibly by replacing ThB by the long-lived RaD. For the same reason, the determination of the diffusion constant of lead chloride, in the vicinity of the melting point, hy means of the decrease in a-radiation after heating ThB chloride col- lected on the surface of a PbCl, pellet was also a failure. Even at 370C° the amount of ThB which evaporates can be detected by radioactive me- thods. The diffusion constant of lead ions in lead chloride must therefore be measured at lower temperatures at which only the very sensitive recoil method can be considered. The results of the measurements obtained by this method are shown in Fig. 2 and Table 1. The circles relate to a series of experiments in which the active deposit was treated with chlorine. The circles combined with strokes relate to experiments where ThB oxide or sulphide, instead of the chloride, was condensed on the PbCl, pellet. Another series of results not quoted here yield the same graphical pattern. The time of experiment was so chosen that the decrease in the recoil yield after diffusion amounted to about 50 per cent. This circumstance is most favourable both for carrying out the experiment and for calcula- tion. It is necessary to know the range of the recoil rays in lead chloride in order to be able to calculate the diffusion constant from equation 8 on 134 ADVENTURES IN RADIOISOTOPE RESEARCH page 130. This quantity is determined as follows: The retarding power of lead and of chlorine for a-radiation is known, whence the range of a- radiation in PbCl, can be obtained. The ranges of a-radiation and recoil particles in air are known. On the assumption that the ratio of the ranges in air is equal to the ratio in lead chloride, the range of recoil particles in lead chloride calculated from the above data is 7.5 X 10-8 em. Sh) 125° MG? 5° 200 B25? 250°" 275° FOO Fra. 2. Self-diffusion of Pb in PbCl, and PbI,. a Strahlen = a-1adiation Rucks toBstrahlen = recoil radiation In the discussion of the experiments with PbI, a method will be described which permits experimental testing of the correctness of the above value. In order to make sure that vaporization effects have not influ- enced the results experiments have also been performed at reduced pressure and these have yielded the same results as those at the ordinary pressure. Attempts to condense ThB oxide or sulphide instead of the chloride were made, in order to obtain information on the effect of a possible incomplete formation of ThB chloride on the experimental results. It is evident from Fig. 1 that the results were not essentially different, owing to the fact that the lead ion surrounded by many chlorine ions soon loses its oxygen partner. The treatment of the PbCl, pellet with chlorine or HCl, after condensing the active material, is also without effect on the result. THE APPLICATION OF RADIOACTIVE RECOILIN DIFFUSION MEASUREMENTS 135 Taste 1. — Setr-Dirrusion or Pb [ons ty PbCl, — rr | t | 1/7 D log D | Remarks 166 0.002277 | 1.47- 10-12] —11.83 180 2207 4.20 11,38 180 2207 | 4.44 S35 183 2193 6.60 == H1418 198 2123. 2.38 - 107-11 10.62 201 2108 | 2.72 Wes ity 210 2070 | 5.79 10.24 211 2068 | 4.69 —10.33 | Sulphide distilled 216 2045 | 9.68 10.01 | Oxide distilled 217 2040 9.93 —10.00 220 2028 | 1.28-10-10) — 9.89 223 2018 | 1,22 9,91 225 2008 1.63 — 9.79 | Oxide distilled 225 2008 | 1.85 Pees O88 235 1969 | 3.44 — 9.46 249 1916 | 9.00 — 9.05 268 1847 | 2.60- 10-9 — 8.58 | 270 8.50 | 1841 | 3.16 — The curve of Fig. 1 can be represented by the equation D = 1.060 X 107 = @~38-120/RT This yields 38,120 cal for the molecular heat of relaxation of lead ions and the value 1.06 x 10~7 for the constant A.The heat of relaxation of the chloride ions in lead chloride is found to be 10,960 cal, from the conduct- ance of PbCl,. The large difference between the heats of relaxation immediately renders intelligible the result of TUBANDT, according to which practically all the mobility in lead chloride is due to the chloride ions. The Jead ions require a much greater energy content than the chloride ions to enable them to take part in place exchange processes. The transport number of lead ions in lead chloride is found to be 107% at 270°C. The investigation of self-diffusion in single crystalls of lead chloride, in which the present experience suggests a smaller diffusion, will be discussed later. 136 ADVENTURES IN RADIOISOTOPE RESEARCH THE DIFFUSION OF LEAD IONS IN LEAD IODIDE Lead iodidé has a smaller electrolytic conductance than lead chloride ; yet a relatively high diffusion velocity of lead ions in lead iodide would be expected, in spite of the low conductance, since TuBANDT found a high value (0.67) for the transport number of the lead ion in PbI,. In agreement with this expectation, it is evident from Table 2 that the diffusion constant can be measured even a few degrees above 100°C. It was always necessary to use annealed pellets in order to obtain repro- ducible values. We find the molecular heat of relaxation to have a value of 30,000 cal and the constant A to amount to 3.43 10°. TasLte 2. — Sevtr-Dirrusion or LEAp Ions ry PbI (Recomm MetrHop) t 1/7 D loz D 114 0.002585 6.31 - 10712 —— le) 122 2532 9.59 —11.02 124 2518 1.47 - 1070 —=—1OLS3 137 2440 4.23 LOLS 147 2382 NEN S NO mee —= O98} 165 2283 6.35 — 9.20 The diffusion constant can be represented by the formula D = 33 5/4b33 x 10° e@30,000/RT Since there is an appreciable mobility of the lead ion in PbI, even at temperatures which are very far removed from the melting point it was also possible to apply the decrease of ionization after diffusion, due to the a-particles, for measuring the diffusion constant. The results are shown in Table 3 and Fig. 2. The diffusion constant can be represented by table following equation : D = 9.11 « 105 e—30-140/R7 The heat of relaxation, which amounts to 30,140 cal/mole, does not differ appreciably from the value yielded by the recoil experiments (30,000 cal). This agreement is also expressed in the parallel courses of the curvesin Fig. 2. The fact that they do not quite coincide is probably due to some uncertainty attaching to the value of the recoil range, as already mentioned above. The two straight lines could be superposed by assuming the range of the recoil particles in lead iodide to be 0.11 4 instead of 0.075 wu. THE APPLICATION OF RADIOACTIVE RECOILIN DIFFUSION MEASUREMENTS 137 Table 3. — Srtr-Dirrusion or Leap Ions in PbI, (a-ParticLe Merxop) t 1/7 | D | log D } 955. | ‘0.001895 | 3:63:- 10? —6.44 260 W876 | 5.30 6.28 301 1749 | 3.42- 10-8 ae yy, 302 1739 | 4.26 255537 315 1701 | 6.70 57 The diffusion constant of lead ions in PbI, can also be calculated from the electrolytic conductance of this compound and the transport number, At 390°C, for example, the calculated diffusion constant is 0.9 10~¢ em? day~! while the recoil measurement and a-ray measurement yield 0.9 x 10-6 and 2.2 x 10-8, respectively. The behaviour at low tempera- tures, where the mobility of the iodide ions controls the conductance, is discussed in the subsequent paper. Summary The velocity of diffusion of lead ions in lead chloride and iodide has been measured by making use of radioactive recoil. The values obtained are Dppci, = = 1.06 x 107 e~ 3812/27 and Dppr, = 3.43 X 104 e— 3900N/RT | The diffusion velocity values for the lead ion in lead iodide were confirmed by other methods. The high value for the heat of relaxation of lead ions in lead chloride (38,120 cal/mole) explains TuBANpT’s result, viz. that in lead chloride the chloride ions whose heat of relaxation amounts only to 11,180 cal are practically the only mobile ions. The transport number of the lead ions in lead chloride at 270° is cal- culated to be 107°. The velocity of diffusion of lead ions in lead iodide calculated from the electro- lytic conductance and from the transport numbers at 290°C as determined by TuBANp?, is in good agreement with our experimental value. 138 ADVENTURES IN RADIOISOTOPE RESEARCH COMMENT ON PAPERS 10—14 Wuen faced with the task of calculating the diffusion rate of gaseous oxygen in gaseous nitrogen, to facilitate the calculation MaxweELu made the assumption that the molecules of oxygen and nitrogen have the same radius and same mass ; he thus arrived at the notion of self-diffusion. The introduction of the labelling principle made it possible to measure a diffusion rate close to self-diffusion. In connexion with the discussion of the interchange between lead atoms of a lead foil and the surrounding lead ions the problem of the diffusion rate of solid lead in lead was first raised in 1915 (paper 8). The first experiments in this field described in paper 10 were, however, carried out a three years later only. Prior to this investigation Grou and the writer measured the rate of diffusion of labelled molten lead in non-labelled lead. Self-diffusion in liquids cannot be expected to lead to results which cannot more or less be foreseen. The rate of diffusion of molten lead in molten lead does not differ much from the diffusion rate of cadmium or thallium in molten lead. In contrast, the rate of self-diffusion in solid metals cannot be foreseen. The diffusion of a solid metal, even a closely related one, in another metal produces changes in the crystalline state which may strongly facilitate penetration. We found that the atoms of solid labelled lead diffused into solid lead about 200 times slower than thallium atoms and about 10,000 times slower than gold atoms diffuse into solid lead. In the first investigation on the diffusion in solids described in paper 9 labelled lead was soldered on a non-radioactive lead rod. After keeping this system at 280°C for up to 400 days slices from the rod close to the place of soldering were prepared and their radio- activity compared. The figures obtained permitted the calculation of the upper limit of the diffusion rate of lead in lead. In a later investigation (paper 11) carried out with Mrs. OpruTsHEVA, wife of the well-known Russian physical chemist FrRuMKIN, we increased the sensitivity of the method by pressing in vacuo a non-radioactive lead foil on one labelled with thorium B. The thickness of the inactive foil was chosen slightly greater than the range of a-particles to be measured ; therefore no scintillations originating from the radioactive lead could be observed when investigating the inactive foil. But, on heating the aggre- gate of the foils, a diffusion of the active lead into the inactive one took place and the a-particles emitted by the succession products of ThB (ThB emits no a-rays, but comes rapidly into exchange equilibrium with a-particles emitting desintegration products) produce scintillations on the observing screen. In further investigations with Srrru we replaced the counting of scintillations by ionization measurements. The range of a-particles emitted by the disintegration product of ThB in lead amounts to 3 x 107% em. A replacement within this thickness of some of the ThB atoms by non-radioactive lead atoms due to an interchange process leads to a decrease in the ionization measured. The range of the recoil particles emitted by the ThC, the disintegration product of ThB, is still appreci- ably (almost 1000 times) shorter than that of the a-particles. The measuring of the decrease of the recoil yield with time of a with ThB covered surface permits to determine as low a diffusion rate as 10~13 em?day~!. By making use of this me- thod self-diffusion in solid lead taking place at 106°C or at a higher temperature was measured (paper 12). From the change of the rate of self-diffusion with temperature the value prevailing closely to the melting point was calculated and compared with the rate observed after melting took place. The ratio wor- ked out to be 10,000. The measurement of the rate of self-diffusion of lead ions in a solid lead salt permitted the determination of the transference number of the ions of solid lead halogenides (paper 14). In his very beautiful investigation TuBANpr found that while both ions have a large part in the conduction of electricity through the solid lead iodide, alone the movement of chloride ions is responsible for the passage of an electric current through solid lead chloride. The ionic mobility of Pb?” in PbCl, calculated from its self-diffusion rate determined by making use ‘ao part of the electrical current passing through solid lead chloride the movement of lead ions of the very sensitive recoil method indicates that for about is responsible. References {. Grou and G. Hevesy (1920) Ann. Phys. 63, 85. G. Hrevesy and W. Serru (1929) Z. Phys. 57, 869. G. Tusanpt, H. Remnnonp and W. Jost (1928) Z. Anorg. Chem. 177, 254. Originally published in Nature 125, 744 (1930) 15. SEARCH FOR AN INACTIVE ISOTOPE OF THE ELEMENT 84 (POLONIUM) G. Hevesy and A. GuENTHER From the Institute of Physical Chemistry, University of Freiburg THE elements 81 (thallium), 82 (lead), and 83 (bismuth) have both radioactive and inactive isotopes, whereas the elements 84—92 are only known in an active form. Several attempts have been made to find inactive isotopes of the latter elements. AsTon, using his mass spectro- graph, tried to discover a stable isotope of radon in the atmosphere, and Hawn made extensive researches to find an inactive isotope of radium. All these attempts failed. We have recently tried to extend the series of inactive elements by searching for an inactive isotope of the element 84 (polonium), which follows bismuth. Through the work of the discoverer of this element, Mme. Curie, and her co-workers, as well as of Marckwaup and of many others, the chemical properties of polonium were found to be intermediate between those of bismuth and tellurium. Hence it is obvious that if a stable isotope exists, it must be associated in nature with tellurium or bismuth. We looked for the elements 84, therefore, in the following tellurium and bismuth minerals: Hessite, calaverite, nagyagite, tetradymite, and bismuth glance as well as native bismuth. The minerals were dissolved, and a known amount of polonium added as radioactive indicator. On removal of the polonium from the solution, it was to be assumed that any isotope present in the solution would accompany the active polo- nium. By special methods devised for the purpose, it was possible to regain the added polonium electrolytically on molybdenum electrodes, the deposit weighing only about 1/10 mgm. X-ray investigations, carried out by the secondary ray method to avoid the possible volatilisation of the substance under the action of the cathode rays, have shown that the deposit cannot contain more than 1/2 per mille of the element looked for. The X-ray line searched for was polonium L,,, the wave-length of which was calculated from Moseley’s law to be 1111 X. U. All the lines on the plate could be identified as belonging to lead, bismuth, silver, mercury, or tungsten. As we started with about 400 grams of each of the minerals mentioned, 1 gm of each mineral cannot contain more than 107 SEARCH FOR AN INACTIVE ISOTOPE OF THE ELEMENT 84 (POLONIUM) 141 om. of the element in question. This negative result is in agreement with generalisations arrived at by Dr. A. S. Russeu. There is very little hope of finding an inactive polonium isotope, or in general of extending beyond bismuth (83) the series of stable elements. 142 ADVENTURES IN RADIOISOTOPE RESEARCH COMMENT ON PAPER 15 Tue search for unknown elements is much facilitated by adding a radioactive isotope of the desired element to the solution of minerals in which the element is most likely to be present. In paper 15 the conclusion was reached that a heavier stable element than bismuth is unlikely to be found. This conclusion is supported by later experience. In an earlier investigation carried out in 1926 in Copenhagen an unsuccessful search was made for a stable isotope of the element 87 by trying to detect a-rays emitted by MsTh, or f-rays by radon. Both processes should lead to the formation of the element 87. A radioactive isotope of this element was later discovered by Perey and Lrcorn. A more detailed presentation of the results stated in paper 15 is to be found in Z. f. anorg. Chem. vol. 194. References G. Hrvesy (1926) Kgl. Danske Vid. Selsk. Mathem.-fysiske Medd. 7, 11. M. Perey and I. Lecorn (1939) Nature 144, 326. G. Hevesy and A. GuENTHER Z. f. anorg. Chem. 194. 162 (1930). Originally published in C. R. Acad. Sci., Paris 178, 1324 (1924) 16. RADIOCHEMICAL METHOD OF STUDYING THE CIRCULATION OF BISMUTH IN THE BODY I. A. CHRISTIANSEN, G. HEvEsy and S. LomMHOLT From the Institute of Theoretical Physics, University of Copenhagen DwRin@ recent years bismuth has acquired increasing importance in the treatment of syphilis”. In order to study the conditions of absorption, distribution in the body and elimination, we have used a radiochemical method which was first proposed by Hrvesy and by Panern®, The principle of the method is as follows: The medicament is prepared from a mixture of a bismuth salt solution and a solution containing a radioisotope, radium-E, of this element. It is well known that a final measurement of the quantity of radioisotope present in the sample suffices for calculating the quantity of inactive bismuth. Radium-E was extracted with a hydrochloric acid solution from radium-D, which had been produced by disintegration of the emanation from a quantity of radium corresponding to 2—4 c. The experiments were performed on rabbits. From time to time small quantities of an oil suspension of the medicament were injected intramuscularly. The rabbits were killed after about 15 days. The following samples were examined: (1) the places of injection; (2) the most important organs; (3) the daily amount of urine; (4) the daily amount of faeces; (5) small known amounts of the suspension used in the experiment. All the organic tissues were prepared for analysis by charring with small quantities of fuming nitric acid; the ash was dissolved in dilute nitric acid and the acid was evaporated of in a petri dish; the radio- activity of the small quantity of salts remaining at the bottom of the capsule was finally determined electroscopically by measuring the f-rays of the radium-E. The a-rays from the polonium present in the residue were absorbed by means of an aluminium foil about 0.05 mm thick. Control experiments have shown that the maximum error of the various experiments performed by this method is about 10 per cent. 1 Sazerac and Levapiti, C. R. Acad. Sci., Paris 172, 1391 (1921); Ibid. 173, 338 (1921). 2 See Aston, The Isotopes, London (1923); F. Paneru, Z. angew. Chem. 35, 549 (1922); G. Hevesy, Biochem. J. 17, 441 (1923). 144 ADVENTURES IN RADIOSIOTOPE RESEARCH Nine rabbits were used in these experiments. Quinine bismuth iodide’ was used in five cases and bismuth hydroxide in the remaining four. The results obtained were in reasonable agreement. We shall limit the results presented here to those from one of the experiments with quinine bismuth iodide. These results are summarized in Fig.1. The heights of the vertical columns represent the quantities of bismuth injected daily; the shaded parts of the columns represent the quantities found 7p) = a> 5 8 5 g oe mg oO 5 2 no) O22 05 a6 ay oa) a ivowe) 014 o ne 0,2 0,t Holl A= eres. rit.) — mg 2,65 2.50) 2,00 SO 1,00 + Foeces 0,50 0,O1 Urine 8/2 %2 2 "Wf Di-tribution of bismuth in the rabbit at the corresponding points of injection. The upper black columns represent the quantities found in the daily faeces and the lower black columns the quantities found in the urine. The black rectangles placed in the upper part of the diagram give the contents of the various organs. The general results of all the experiments can be summarized as fol- lows: © bismuth is eliminated chiefly in the urine; the quantity of bis- muth found in it is double the amount passed in the faecal matter; it increases during the period of treatment in the urine but this is not so clearly demonstrated in the faecal matter; (2) the heart and lungs contain only a small amount of bismuth; the liver contains quite a small quantity and the kidneys a fair amount, generally more than double the amount in the liver; only a very small quantity of bismuth has been found in 50 em? of blood. METHOD OF STUDYING THE CIRCULATION OF BISMUTH IN THLE BODY 145 The results obtained show that bismuth should only be used with great care, because of its quite slow and irregular distribution leading possibly to a danger of poisoning. Its resemblance with mercury from the aspect of circulation. Precautions are required in the simultaneous use of the two metals, which cause accumulation of the toxie effect, See Svenp Lomuott, Brit. J. Dermatol. (1921); Arch. Dermatolog. u. Syphilis 126, 154 (1918). 10 Hevesy Originally published in C. R. Acad. Sci., Paris 179, 291 (1924). 17. RADIOCHEMICAL METHOD OF STUDYING THE CIRCULATION OF LEAD IN THE BODY I. A. CHRISTIANSEN, G. HEVESY and Sv. LOMHOLT From the Institute of Theoretical Physics, University of Copenhagen Ar the meeting of the Academy on 7 April 1924 we presented some work performed by a radiochemical method on the circulation of bismuth in the body. Since then we have used the same method for some experi- ments on the circulation of lead, the first results of which are presented here. In using bismuth in medicine it is of very great importance to know the rate at which the bismuth injected into the body is distributed and eliminated. This question can be resolved by using a substance with a quite short half-life (5 days), such as radium-E. It is quite different in the case of lead, since all the medical interest in this element centres around the chronic poisoning which is caused gradually by the absorp- tion of small quantities of lead during a long period of time. This is the reason why we have used radium-D (half-life 20 years) in our experi- ments, whereas Hrvesy, in his experiments on the distribution of lead in plants, obtained satisfactory results with another isotope of lead, thorium-B, having a half-life of only 11 hr. Since radium-D emits only soft B-rays, which are difficult to measure with the electroscope, we have counted the f-radiation of radium-E in equilibrium with the radium D and, consequently, have measured the various products of analysis only at the end of several weeks. The experiments were performed on rabbits and guinea pigs. The method described previously has been modified slightly: Instead of evaporating the solution of the organic matter, which is decomposed by means of nitric acid (or potassium permanganate in acid solution), it is diluted, treated with 100 mgm inactive lead nitrate, and lead sulphide is precipitated. After filtering at the pump on a plane filter, the dried filter is placed in a petri dish and the activity of the deposit is finally determined. One example only is given here from our experiments. The lead hydrox- ide, mixed with olive oil and a little carbon black, contained a quantity 1. G. Hevesy, Biochem. J. 17, 441 (1923). METHOD OF STUDYING THE CIRCULATION OF LEAD IN THE BODY 147 of radium-D derived from the disintegration of an amount of emanation = : 1/ 1 = SE er | eee ae see corresponding to 4%—1 c. The results are shown in Fig. 1. The heights of the vertical columns represent the quantities of lead injected daily; the shaded parts of these columns indicate the deposits of corresponding injections. The upper black columns represent the w @ ce) © oO — w oO o a o > iS ln Ye = D ms ® == v=o Eo fo) c o c On DAG SO Co) 5 = ne EL 5 2o oo = =] oy < Oe Jeo YoRre} ports — © is = oO 1,00 OLUrin Hate IM 4 I9 20 2! 22 23 24 25 26 27 28 Fig. J Distribution of lead in the rabbit quantities found in the daily faeces and the lower black columns show the quantities present in the urine. The black rectangles in the upper part of the diagram give the contents of the various organs. It will be evident that there is quite a substantial difference between the results for bismuth and for lead; the amounts of lead stored in the liver and eliminated in the faeces are larger, at the expense of the amounts found in the kidneys and urine which are the major participants in the case of bismuth. We have recovered 90 per cent of the amount of lead injected. 10* 148 ADVENTURES IN RADIOISOTOPE RESEARCH COMMENT on PAPERS 16, 17 In the early twenties dermatologists became much interested in the therapeutic application of bismuth compounds. This induced us, together with CHRISTIANSEN and Lomuo tr (the latter being a dermatologist) to investigate the distribution of administered bismuth in the rabbit. We applied in our study RaK-labelled bismuth. This work was then extended to the study of the distribution of labelled lead. The first-mentioned investigation was the first application (1924) of radio- active tracers in animal physiology, shortly following their first application in plant physiology. While participating at the Liverpool meeting of the British Association for Advancement of Science, the writer learned that the gynaecologist Blair Brui obtained good results by applying lead salts in cancer therapy. This induced us (Hrvesy and WaGner, 1930) to compare the distribution of lead between normal and cancerous tissue, applying labelled lead. The great Freiburg patholo- gist ASCHOFF on my request delegated a Japanese collaborator of his to help us in this work, and later ScHOENHEIMER to assist the latter. This was the first experience of SCHOENHEIMER in the field to which he later made, jointly with his eminent colleague RrrrenBerG, a very great number of unsurpassed classical contributions. References G. Hrvesy and O. H. Wacner (1930) Arch. Exp. Pathol. and Pharmacol. 149, 336. R. SCHOENHEIMER (1942) The Dynamic State Of Body Constituents, Cambridge, Mass. Originally published in Nature 13, 754 (1935). 18. RADIOACTIVE INDICATORS IN THE STUDY OF PHOSPHORUS METABOLISM IN RATS O. Curevitz and G. Hrevesy From the Institute of Theoretical Physics, University of Copenhagen RECENT progress in the production of radioactive isotopes by neutron bombardment makes the radioactive isotope of phosphorus {3P easily accessible. This isotope, which has a half-life value of 15 days, can he utilised as an indicator of inactive phosphorus in the same way that the radioactive isotopes of lead, bismuth and so on were formerly used Percentage of active phospnorus tn urine (0) and faeces (+ 3 it I 15 19 23 27 3I Number of Days Fra. 1. On the first day 7.4 per cent of phosphorus was found in the faeces and 5 per cent in the urine as indicators of these elements. If, for example, we add _ active rl og to 1 mgm of inactive phosphorus in such quantity that the Geiger counter registers 1000 impulses per minute, carry out with the phos- phorus activated in this way any sort of chemical or biological reaction and then find that the product obtained gives 1 impulse per minute, we may conclude that 1/1000 mgm of the phosphorus originally intro - duced is present in the product investigated. 150 ADVENTURES IN RADIOISOTOPE RESEARCH Rats were fed with a few milligrams of sodium phosphate containing #sP as indicator. The radioactive phosphorus present in the urine and faeces was then investigated for a period of a month. The result is shown in Fig. 1, which shows the percentage of the 2 mgm of phosphorus taken, found daily in the excrements. The rat was killed, and, after ignition, the phosphorus content of the different organs was investiga- ted. The result of an experiment in which the rat was killed 22 days after being fed on active phosphorus is seen in the first column in Table 1. The largest part of the phosphorus taken is present in the bones, and the smallest in the kidneys. When, however, we take into account the very different weights of the different organs and calculate the phos- phorus content of the latter per gram after drying, we obtain a very different picture, as seen from the second column in Table 1. The spleen, kidneys, and the brain are found to contain per gram most of the active phosphorus. During one of the experiments, the rat produced six off- spring on the seventh day, of which five were eaten by the mother; this caused a large increase in the active phosphorus content of the excreta in the following three days. The presence of 2 per cent of the 2 mgm active phosphorus taken by the mother was revealed by the analysis of the remaining offspring. TABLE |, — DISTRIBUTION OF THE ACTIVE PHOSPHO- RUS IN THE RAT Per cent | Per cent EMU) sgaqonscagaouDGde | 26.3 per gm PASCOS ats esol cisccieiessieye'ss- | 31.8 Brain and Medulla .... | 0.5 | 14.7 Spleen and Kidneys ... | 0.2 | 18.2 Davel fee we ee ee | 17 | 13.9 Bloodless ratte ae | 0.4 | 1.8 Skeletonwrssacecicsen men | 24.8 | 2.8 ! 7.4 Muscles and fat ...... | 17.4 The active phosphorus content of the urine and faeces shows great fluctuations during the first few days after the intake of the preparation. Later, it becomes fairly constant; and we have obviously to deal with the excretion of phosphorus which has already been deposited for a while in the skeleton, the muscles, or other organs, and which has been displaced again. From our experiments, it follows that the average time which a phosphorus atom thus spends in the organism of a nor- mally fed rat is about two months. This is also supported by the fact that rats killed about a month after the intake of phosphorus contain only about half the active phosphorus found in those killed after a weeks time. This result strongly supports the view that the formation of the PHOSPHORUS METABOLISM IN RATS 151 bones is a dynamic process, the bone continuously taking up phosphorus atoms which are partly or wholly lost again, and are replaced by other phosphorus atoms. In the case of an adult rat, about 30 per cent of the phosphorus atoms deposited in the skeleton were removed in the course of twenty days. In another set of experiments we investigated the different parts of the skeleton. No conspicuous differences in the active phosphorus content could be found, with the exception of the teeth. The front teeth, which grow rapidly in rats, contained a larger part of the 2 mgm _ phos- phorus taken than the average of the whole skeleton, the ratio being about 10:1 in the case of adult and 6:1 in that of half-adult rats, whereas the molar teeth took up less than the average per gram of the skeleton, the ratio being 1:2 in the most extreme case. A detailed account of these and further results will be published elsewhere. Originally published in Kgl. Danske Videnskabernes Selskab. Biologiske Meddelelser 13, 9 (1937) 19. STUDIES ON THE METABOLISM OF PHOSPHORUS IN ANIMALS O. CHrevitz and G. HEVESY From the Institute of Theoretical Physics, University of Copenhagen In a recent letter to Nature! we communicated the results of some experiments on the metabolism of phosphorus using a radioactive phosphorus isotope as indicator. What follows is a more detailed descrip- tion of some of our experiments, carried out chiefly on rats but partly also on human subjects. PRINCIPLE OF THE METHOD USED Disregarding hydrogen, the only element which is ever met with in a nuclear state (as a proton) in chemical reactions, isotopes do not separate to a measurable extent during chemical or biochemical processes. It follows from this inseparability that when a known amount of radio- active phosphorus is added to, for example, 1 mgm of phosphorus the presence of the former will always indicate the presence of the latter, we can thus distinguish for example between the phosphorus atoms taken in with the food (to which we add some radioactive phosphorus) and those already present in the system. The use of isotopic indicators is not dependent on an absolute inseparability of isotopes by chemical methods. We know indeed that minute separations almost always occur. It is sufficient that, within the analytical accuracy claimed, no separation takes place. Phosphorus has only one stable isotope 3!P but we can prepare un- stable radioactive isotopes of phosphorus having atomic weight of 30 and 32; the latter has a half-life of about a fortnight and is very suitable for use as an indicator. It was used by us in many experiments of different kinds. 1Q. Curevitz and G. Hevesy, Nature 136, 754 (1935). STUDIES ON THE METABOLISM OF PHOSPHORUS IN ANIMALS ] Vt w PREPARATION OF RADIOACTIVE PHOSPHORUS . . 9 . A . Radioactive phosphorus #3P can be prepared from chlorine or from sulphur under the action of fast neutrons, or from ordinary phosphorus under the action of slow neutrons; the nuclear reactions are: Clin Pi! Sele 32¢ 1 32p 1 ie on = 73P +1H 31p 1 32p i5t + gn = 451 Using neutrons liberated from mixtures of radium and _ beryllium, P can be prepared most conveniently from sulphur. We found it advis- able to use carbon disulphide instead of the elementary sulphur used by Fert and his colleagues in their original experiments. About 10 litres of carbon disulphide were exposed to neutrons from radium- beryllium mixtures and a fortnight later the carbon disulphide was distilled off. The residue contained the radioactive phosphorus formed, along with some of the decomposition products of carbon disulphide. The residue was oxidized and the phosphoric acid obtained converted into the phosphate compound wanted. We used chiefly sodium radio- phosphate in our experiments. The weight of the radiophosphorus produced is extremely minute; using a source containing 100 mgm of radium, less than 10°18 gm of radiophosphorus is obtained. By adding a Suitable quantity of sodium phosphate to the sodium radiophosphate solution we obtain the “radioactive” (“labelled’’) sodium phosphate desired. To concentrate the radiophosphorus obtained by neutron bombard- ment of carbon disulphide other methods besides that outlined above were used. A very convenient way to prepare nearly pure radiophos- phorus is the following. Under the action of the radiation some decompo- sition of the carbon disulphide takes place anda partly orange-coloured precipitate is formed which settles on the glass walls. This slight preci- pitate contains a large part of the radioactive phosphorus formed. The precipitate is possibly identical with the red sulphur described by Maenus as far back as 1954, which was found to consist of a mixture of sulphur and organic sulphur compounds. We are engaged on the investigation of this precipitate. In a third method of preparation the phosphorus formed was removed from the carbon disulphide solution by shaking the latter with diluted (20: 1) nitric: acid. 154 ADVENTURES IN RADIOISOTOPE RESEARCH DETERMINATION OF THE RADIOACTIVE SODIUM PHOSPHATE The radioactivity of the samples of blood, bones, ete. to be analysed is in most cases too feeble to be measured even by means of a very sensitive electroscope. GEIGER-MULLER counters, much more sensitive instruments, are therefore utilised for measuring purposes. We use for the most part tubes having an available surface of about 1.5 em?. The sample to be measured must accordingly be spread over about the same area. The P-rays emitted by the radio phosphorus are fairly penetrating and are not much weakened when an aluminium dish of 1.5 em2 surface is filled to a depth of a few millimeters with a bone sample weighing 100 mgm. We want to know what percentage of the radio- active phosphorus taken is to be found after a certain time in, for exam- ple, the bones. The procedure is as follows. We take a solution of active sodium phosphate, use 99 per cent of it for feeding the animal and keep 1 per cent as a “‘standard’’. We kill the animal, separate a bone sample, ignite it, and measure its activity. Should the latter be, for instance, half as large as that of the standard which is measured simultaneously, then we can conclude that 0.99 0.5 per cent of the active phosphorus atoms eaten are actually present in the bone sample investigated. Although the /-radiation from radioactive phosphorus atoms is not much weakened in penetrating through 100 mgm of bone ash, we can entirely eliminate the possible error due to this absorption by adding 100 mgm of calcium phosphate to the standard solution; this has the same absorbing power as the bone sample. It is advisable to make the standard as similar to the sample to be measured as possible. In dealing with urine, faeces, muscles, liver etc. we first destroyed the organic matter by one of the usual methods; in several cases, however, these were replaced by treatment with fuming nitric acid. Then calcium phosphate and calcium oxide were added if necessary to make the sample more Similar in its composition to our standard preparation and finally the sample was ignited. To demonstrate the utility of the isotopic indicator method we will first consider the problem of the origin of the phosphorus in the faeces. ORIGIN OF THE PHOSPHORUS IN THE FAECES Chemical analysis enables us to determine the phosphorus content of the excreta but not to decide to what extent the phosphorus found in the faeces is undigested material and what fraction of it is phosphorus having its origin in the organism. The investigations described in this paper have revealed that a fairly rapid interchange takes place between STUDIES ON THE METABOLISM OF PHOSPHORUS IN ANIMALS 15: ol the phosphorus present in the different bodily organs and that present in the blood. A part of the latter finds its way, when the digestive fluids are formed, into the intestinal tract and is thus added to the faeces. The following experiment permits us to distinguish between food phos- phorus and that originating from the blood. We add a known amount of radioactive phosphorus to the diet and determine what percentage of the latter is to be found in the faeces. In a separate experiment we inject a known amount of radioactive phosphorus (sodium phosphate) into the blood and determine what part of this phosphorus appears in the faeces. The combination of the two results enables us to deter- mine what part of the phosphorus found in the faeces is due to incomplete digestion of the food eaten. In Table 1 the amount of radioactive phosphorus eliminated through the kidneys and the gut is given for the case of a patient fed on a normal hospital diet to which 0.5 mgm of labelled sodium phosphate was added. Within 5 days 21.7 per cent of the phosphorus was eliminated TaBLeE 1. — RADIOACTIVE PHOSPHORUS GIVEN TO Human Sussect Per Os | Percentage of original Number of days after Diuresis eaionee ee taking P |; in gm | | in 1 gm of in total | the urine ash urine ee eee | 1880 | 1.23 11 [Ore Rai / 1800 | 0.31 2.8 W==A GEOCdOCHSD CONS | 1620 | 0.31 2.8 SS Ay Be Set eee ene h 16700 4° «6.26 2.4 iy SooutopooUUe0o | 1540 | 0.29 2.7 Beet see eeeeayisy | 1860 | 0.25 1.8 | in total | faeces OMS ca | = = 0 SE omen OOO OC — | ~~ 7.0 PE AGt ieee BRE ear oe — | _ 5.6 Bob eo OCOD ODD —- — 1.8 Ar Nakevekevolotairovershel sy aces -— ~- out = ONR See shoe) sheieyen sisi es — — 0 in the urine and 15.5 per cent in the faeces. Similar results were obtained in other cases. Table 2 shows the results obtained when the radioactive phosphorus was injected into the blood of the same patient. Within days 20.5 per cent was lost through the kidneys and 2.5 per cent through the gut. Thus about 1/8 of the phosphorus atoms eliminated from the blood pass through the gut. By combining the above results it follows 156 ADVENTURES IN RADIOISOTOPE RESEARCH that of the phosphorus found in the faeces about 20 per cent was not undigested material but was phosphorus which had already had a share in building up the organism and had left it by entering the digestive liquids and thus getting into the faeces. In the case above, 22.3 per cent of the radioactive P left through the kidneys within 6 days and in other cases values varying between 20 and 25 per cent were obtained. TABLE 2. — RADIOACTIVE PHOSPHORUS INJECTED INTO THE BLOOD oF A PATIENT Percentage of original radioactive P Number of days after Diuresis | injection in gm in 1 gm of in total the ash urine See rrererote taretetsroras 1650 0.78 12.5 DD poise Gide wee e 1510 0.20 3.1 ea eee a a 1850 O1.5 2.9 3—4 J be) eC MCR Ee 850 0.16 1.6 BO) oe aie sue svae oe) oka | 1450 | 0.13 2.2 U3 socaceaacoccds S00 0.09 0.6 SB =O)“ sirsy suet ovat sbeusesers ete 2000 0.10 1.8 inlgmof in tota! the ash faeces Od. 6 era eeSas Sere -- 0.085 0.24 MD cera han shal serererouore -— Ome 13 OD See es oc a 0.072 0.37 a Ae aval aycreneteatents eas -- 0.072 0.56 AOE nega elarekar ev onene ete -- 0 0 In carrying out experiments like those described above, the most satisfactory procedure would be to replace by radioactive labelled phosphorus atoms the normal phosphorus present in all the foodstuffs administered. By bombarding the material in question with a strong source of slow neutrons we could turn some of the phosphorus atoms into radioactive phosphorus; but such a process always leads to a dis- ruption of the molecular bonds of the phosphorus atoms which become activated and so to a destruction of the chemical compound. We must therefore content ourselves with adding inorganic radioactive phosphate to the food consumed and try to obtain a mixture of radioactive inor- ganic phosphate and food as uniform as possible. In our experiments earried out with human subjects the sodium radiophosphate was admi- nistered in a large volume of milk. Milk contains 0.0795 per cent of inorganic phosphorus and about half that amount (0.036 per cent) STUDIES ON THE METABOLISM OF PHOSPHORUS IN ANIMALS —_ or ot | of phosphorus in organic form. Although the latter does not exchange with the atoms of the inorganic radioactive phosphate, the bulk of the phosphorus (0.0795 per cent) reaches a state of kinetic equilibrium with the radioactive phosphate added and becomes radioactively indi- cated. During the digestion process the 0.036 per cent will be set free from its molecular binding and only at this stage will it have an oppor- tunity to become thoroughly mixed (in an atomic sense) with the radio- active phosphate atoms. While, as has already been mentioned, it would be preferable in investigating phosphorus metabolism to utilize food in which all the phosphorus atoms are labelled, it is not probable that the information obtained with such material would be appreciably different from that obtained in the experiments described in this paper. Experience shows that the retention of phosphorus does not depend on the form in which the phosphorus is present?! in the food, on whether it is present as inorganic and thus exchangeable phosphate or as non- exchangeable. Ducks reared on diets containing phosphate only in inor- ganic form matured normally and laid 85 to 795 eggs during the first summer?. About 15 per cent of the phosphorus present in meat, more than half that present in milk, and the greater part of that present in vegetables, i.e. the bulk of the phosphorus eaten, is present in inorganic and thus exchangeable form. Rats are inclined to eat their offspring and they could easily be fed on young rats born by a mother fed on radioactive phosphorus, but the chief source of phosphorus would in this case, too, be inorganic phosphorus, namely that present in the skeleton. ELIMINATION OF PHOSPHORUS BY RATS We carried out numerous experiments with rats which were fed on a normal diet to which radioactive phosphorus was added. In some cases we added 0.1 mgm or less in the form of sodium phosphate dissolved in a few drops of water which was then soaked up by a small piece of bread given to the animal. The average of several experiments gave a total excretion of 26 per cent through the kidneys and of 32 per cent through the gut. In some other experiments calcium phosphate was administered, mixed with butter, which was given to the rat on a small piece of white bread. The result of such an experiment is seen in Table 3, which contains the results of the analysis of the urine and the faeces collected during 19 days. The urine was concentrated by evaporation, treated with fuming nitric acid, and ignited ; a known fraction of the ash 1M.,.Sperrs and H. C. SHerman, J. Nutrit. 11, 216 (1936). 2G. FINGERLING, Biochem. Z. 38, 448 (1911). 158 ADVENTURES IN RADIOISOTOPE RESEARCH obtained was then introduced under the Geiger counter. 19 days later the rat, which weighed 256 gm, was killed, the corpse was treated with fuming nitric acid to destroy organic compounds, the fatty residue was treated with conc. sulphuric acid, and then ignited in an electric TaBLe 3. — 1.5mcGm RADIOACTIVE Catcium PHOSP- HATE ADDED TO NorMAL Diet or ADULT Rat : | Percentage of original rad. P Number of days after taking rad. P | in the urine in the faeces | O27 Binshe ate ata ke he 13.1 Sos elle ware tein Sue elas 3.9 4.7 ALO) isis wrererererevereieteres Del 2.4 TSANG tm tee ase ene ort | 18 0.93 NSPE Gr eacre cere ert | ge) 121 MGSO Fae o-stietchs orexeeue oe | 1:2 1.88 Rotallieyere. Does 24.0 1 Faeces contaminated by urine. oven. 50.2 per cent of the phosphorus given was found in the ashes, which were to a large extent composed of calcium phosphate, and had a total weight of 5.84 gm. In some cases we added large amounts of calcium phosphate contain- ing active phosphorus to the diet. When for example 18 mgm of phos- phorus as calcium phosphate were given — this corresponds to about four times the phosphorus present in the normal diet — 41 per cent of the active phosphorus was eliminated through the gut in the course of 19 days and only about 10 per cent through the kidneys. Furthermore an analysis of the active phosphorus content of the corpse and the excreta revealed that when large amounts of phosphorus were added to the diet the animals would eat only part of it, however, carefully it was administered. We decided therefore to study the effect of the intake of large amounts of phosphorus on dogs. The phosphorus atoms absorbed have ample opportunity to enter into kinetic exchange with the phosphate ions present in muscles, bones, and other organs and also to a certain extent to enter organic molecules and replace the phosphorus atoms present there. Many of the last mentioned processes are dependent on enzymatic action. The rate at which the active phosphorus enters the blood corpuscles, the parti- culars of this process, and the distribution of the radioactive phosphorus between the blood and the different organs were investigated by Pro- fessor LUNDSGAARD and one of us and the results will be published shortly. STUDIES ON THE METABOLISM OF PHOSPHORUS IN ANIMALS 159 PHOSPHORUS EXCHANGE IN ADULT RATS A preliminary investigation revealed the following distribution in adult rats killed three weeks after eating the radioactive phosphate administered in the form of 0.5 mgm sodium phosphate added to the normal diet. TABLE 4. — DisTRIBUTION OF Rap. P In Aputt Rats Kintep 3 WEEKS AFTER EATING IT \Whsins: 25 bee Go oon adda s oe Zoro INRYOES Sooo obdneuonalD ous | 31.8 Skeletomiererercrercicl ier bereneren 24.8 Musclestand" fat <<... < > = 17.4 Wiverwanrcci cece ose tae S [viens Brain and Medulla ....... 0.1 Kidneys and Pancreas .... 0.1 In interpreting the results obtained it is convenient to compare the radioactivity of equal weights (say 100 mgm) of the ashes, of the bones, the teeth, the liver, and so on. These all contain about the same percent- age of phosphorus (17 per cent, 17 per cent, 16 per cent); the phos- phorus content of the ash of the blood is rather different, but as was stated above the behaviour of the active phosphorus in the blood was not investigated to any great extent in the course of this work. In a series of experiments we gave the same amount of radioactive phosphorus to 6 rats. One pair of rats was killed after one week, a second pair after two weeks, and a third pair after three weeks. The results are Seen in the following table. The weights of the different skeletons vary to an appreciable extent; the weights of the animals were 225, 210, 200, 215, 235 and 220 gm before, and 220, 205, 200, 205, 235 and 220 gm resp. after the experi- f rad. P f 1 Animal killed weeks after Dee ee guns eating rad. P | in the skeleton in the incisors | UA BE a Ore ee REO 34.2 2.1 Le ter sce ctatchere seca eiar a yeiies ene | SDL Zeal DOEe RUE AE Na srt, eet eo | 32:2 2.8 PATS iter Fh nee HEREC PC RCRCNC | 27.2 Del Deel hen ee meestithod evens Ae cliche vevans 24.6 2.8 ob) enact acer ened tavenske atistons 25.4 2.7 160 ADVENTURES IN RADIOISOTOPE RESEARCH ment. In comparing the rad. P content of different organs of the same rat we are independent of the assumption that all the rad. P given was actually eaten by the animal, though we are not, when we compare the rad. P content of organs from different rats. The greater rad. P content of the bones of the animals killed after the lapse of only a week cannot, however, be due chiefly to such a reason as this, because in that case the rad. P content of the incisors would also be appreciably higher in the case of rats killed after the lapse of one week. This is not the case, as can be seen from the figures in Table 5. We must therefore conclude that the rad. P taken up by the bones, and in exactly the same way all the phosphorus taken up by the bones, has a certain chance of being lost again. Indeed an uptake of phosphorus atoms by the bones of an TABLE 6 p. c. of rad. P weight of p. c. of rad. P | taken, present ashes of the taken, present in 100 mgm organ in in the total of ashes mgm ashes a) rat killed after 1 Week IB OMNES (dic ielere acne sheus/soa8l 0.8 4300 34.3 Molars” oki sieteistsreis susa0a 0.2 100 0.2 UMCISOTS ici eke elite sits 1:3 253 3.3 ABAVIETE Vereey awit ois etess erent | ee 1031 —— bb) wat killed after 2 Weeks BONES: sa seares etierd spate ere 0.7 4200 29.5 IMOlarsy << tciccuevetware se O72 100 0 MMCISOTS 4) pehic ete wate ihetene 19 215 4.1 WIV CT evens cel ario ie etaieesra 2.0 210 4.2 adult rat can only be explained by a corresponding process in the opposite direction. Another example of the decrease in the active phosphorus content of the bones with time is seen in Table 6. While the bones show a decrease in their rad. P content with time and the molars no change to within the accuracy of experiments, the incisors show a marked increase. The incisors of adult rats show a very pronounced growth. The discussion of their behaviour is therefore better postponed and will be dealt with in the next chapter, where experiments on young rats are described. The results of an experiment carried out with two rats both killed after 5 days time are seen in Table 7. 1 The weight of the ashes of the liver was found to be very variable. STUDIES ON THE METABOLISM OF PHOSPHORUS IN ANIMALS 161 TABLE 7 p. c. rad. P taken found in 100 mgm of ashes Boas! Gaaoonoboadcn dot 1.3 1.4 Molarse a 2.4) sen inno 0.24 0.34 IMCiISOLs ae ieee 2.4 a} NGL VO Tee eiatere cishoren teenie oreo 2.7 lez) INDEOCS Goopososecacae | ed 1.8 [siatoh Gm oiiom oOo OOD 0.46 0.58 As is seen from the above figures the muscles show a somewhat larger content of rad. P than an equal weight of the bones. The active P content of the brain ash is decidedly lower. To ascertain if the phos phorus atoms present are not acid soluble, phophorus compounds are also replaced by active P atoms, and the brain treated with 6 per cent trichloracetic acid solution. By this means all the acid soluble phosphorus was removed. The operation was carried out with great care. After igniting the filtrate and residue, the activity of both fractions was measured. We found both fractions to be active, the activity of the phosphatide fraction being about 1¥/; of that of the trichloracetic acid extract. We are engaged in following up this point in greater detail, using more trustworthy methods of separation. EXCHANGE OF PHOSPHORUS BY GROWING RATS The uptake of phosphorus shown by different organs of rats about 2 weeks old is seen in Table 8. The rats were killed three days after being fed with radioactive phosphorus added to their normal diet. Focusing our attention first on the bones we notice that 100 mgm of ash contain more than ten times as much radioactive phosphorus as TABLE 8 | c _ nae | Rat T (weight 27 gm) Rat Ll (weight 24 gm) —— p. c. of rad. P p. c. of rad. P taken present weight of ashes weight of ashes taken present | | | in mgm | in 100 mgm in mgm in 100 mgm of ashes of ashes Bones" (Leg)! stcae. eotine ae: 65.4 10.5 BS 10.9 TMCISONS! bs. syeis: steve syorevetstoegois esse | — 5.8 — 5.8 Molars.j.iu tas -su tee eee ols | 39.4 2.9 33.8 2.8 IN GUIS) aiareoore bo ob o5-0-0 ace | _- 11.0 - _ BOOP iesees «rete «ts eaters rete eee | _ 2.8 — 2.6 1] Hevesy 162 ADVENTURES IN RADIOISOTOPE RESEARCH was found in the ease of adult rats. The high radioactivities of the bones are due to the fact that in this case an appreciable part of the bones are actually grown from blood of high radioactive phosphorus content; a rapid formation of new cells takes place, in whose building up radio- active phosphorus participates. A very conspicuous difference is found between the active phosphorus content of the molars of rapidly growing and of adult rats, the great difference being due primarily to the low exchange values in the latter. The brain as a whole was found to contain 0.5 per cent of the active phosphorus taken by the animal. The ratio between the rad. P content of the muscles and the bones is nearly unity in the case of the young rats, while in adult rats the muscles show a higher rad. P content. When we compare the radioactive phosphorus content of the bones of growing rats, we find for example more activity in 100 mgm of the ashes of the bones of animals killed after one week than in those killed after two weeks. This is due chiefly to the fact that the phosphorus atoms present in the bone at a certain time will soon be found in an entirely new part of the growing skeleton, and will also have a certain chance of leaving the skeleton entirely. If we want to obtain information on the latter point we must compare the “radioactive” phosphorus contents of whole skeletons. We carried out such experiments, comparing the whole of the leg material. Five very young rats having a total weight of only 25 gm were fed on their normal diet plus some radioactive phosphorus (0.50 mgm each). Two were killed 2 days later and three 65 days later. 10 mgm of the ashes of the leg bones of animals killed after 2 days contained 8.4 times as much radioactive phosphorus as that of rats killed after 65 days. The active phosphorus atoms were in fact distributed all through the greatly increased amount of bone tissue ; the leg bones increased in the course of 63 days to about ten times their original weight, as can be seen from Table 9. When we compare the radioactive phosphorus content of the total bone material of the legs, the difference between the rats killed after 2 days and after 65 days is much less; the difference still present is due to the loss of phos- phorus atoms by the bone material. The phosphorus atoms which were present in the bone for a while and left it again will be found partly in the excrements but to some extent also in some of the organic compounds building up the organism. In the course of two months about one third of the phosphorus atoms originally present left the skeleton. A comparison of the behaviour of the active phosphorus present in the incisors with that in the bones is difficult in view of the rapid using up and replacement of the incisors. Prof. Horst, Prof. Krocu STUDIES ON THE METABOLISM OF PHOSPHORUS IN ANIMALS 163 and one of the writers of this paper are at present engaged on an investigation of the exchange of phosphorus in the incisors on differ- ent lines. TABLE 9 , : : Weight | D.C. Of Period between taking of 2 | ; : of bone ash (legs) | radioactive P radioactive P and killing in mgm H present DP OANS csie tess os lous weieie 65.4 7.4 PAS Be MS Bre lois oS. 16 GcomoN 59.0 1.5 65 4 Bowdoooan btoooc 440 4.1 65 soocoraootee ac 514 D1 OD bsss Miacediics cake kee: 613 DD, UPTAKE OF PHOSPHORUS IN PREGNANT RATS AND IN HUMAN PLACENTA In Table 10 the result of the investigation of adult normal and pregnant rats is seen. Those designated I were killed after a lapse of one week, those marked II after two weeks. As can be seen from the above figures the different organs of the preg- nant rats took up less rad. P than normal rats, the difference being found at least partly in the foetus and placenta. In the first rat, which was in an advanced stage of pregnancy, the foetus and still more the placenta had a high content of rad. P, higher than any organ of the mother, We find here again a very conspicuous illustration of the differ- ence between the taking up of P through an exchange process and through TABLE 10 ] { Normal rat p. c. | Pregnant rat p. c. of rad. P taken | of rad. P taken | . | j present in | present In | 100 mgm ashes | 100 mgm ashes MBoneswer re cece ene 0.78 0. 49 SBoneser ents tis ae. eate 0.74 | 0.52 I IGN. S55 ocoaddsee 1.3 | 2 JOU JUNKO Sconocodccoc 1.9 | 1.7 EEMolamsystnc 4 do tak O21 ha i) 0.12 IT A (ol hehe outs Bp bec oere 0.23 0.16 IIe 1B aie oe ue eee 2.0 | 1.6 NOG eriera sie five tees 1.94 1.0 MSH OStaIe cits wwicie eee —- Dall Moet ap seacscc ci adler. — 0.54 Placenta seine sa cictes -- | 4.0 Le Plagemtakerss crac ons _ | Dy ee 164 ADVENTURES IN RADIOISOTOPE RESEARCH actual growth, the latter being much more effective in introducing rad. P into the tissue. An appreciable part of the foetus has actually been built up by utilising the circulating rad. P and has correspond- ingly a high #P content. This is still more the case for the rapidly growing placenta. In the case of the second animal, pregnancy occurred at a much later date than the intake of rad. P. The foetus was nourished by blood poor in rad. P, and correspondingly the rad. P content of the ash of the foeta was much less. Whereas in the first case the weight of all foeta was 345 mgm, in the second case it was only 52 mgm, the weight of the placenta ash being 43 and 12 mgm res- pectively. We also had an opportunity to find what was a comparatively very high rad. P content for the placenta of a human subject; as much as 0.095 per cent was found in the ash of the placenta, which weighed 133.8 mgm. We can estimate the total ash which the patient in question should give on ignition as 2800 gm. The weight of the placenta ash thus amounted to less than 4¥/s 999) of the total ash, while the rad. P con- tent was as much as 1/,,,, of the total amount of rad. P given, showing a concentration of rad. P in the placenta ash more than twenty times as great as that in the average ash of the body. One might try to explain the high rad. P content of the placenta by its high blood content. That this explanation fails is seen, however, from the following. The ash of the placenta was found to weigh 133.8 mgm and the ash of about 5 ec. of blood would weigh the same. But as early as 8 hours after the injection of rad. P such a volume of blood was found to contain less than 1/,9999 of the latter’, and after the lapse of a few days — when the placenta were removed — still less. The high rad. P content of the pla- centa cannot therefore be due to their blood content. No activity could be detected in the ash of the few weeks’ old foetus removed in the course of an operation, but the weight of this sample amounted to only a few mgm. UPTAKE OF PHOSPHORUS BY RACHITIC RATS We carried out a set of experiments on two months’ old rachitic rats, which had been used by FrepErIcA and GuDJONSON in their experi- ments on the effect of vitamin A and D deficiency on rickets. The rats were fed before and during the experiments on a diet free from or poor in vitamins A and D. The weights of the animals before the experiment were 89, 83, 85, 93, 90, 95 and 103 gm. The results are seen in Table 11 1 In the case of another subject we found | ce. of blood to contain 0.0027 per cent of the phosphorus injected after the lapse of 12 hours, the blood corpuscles containing 11 times as much active phosphorus as the plasma. STUDIES ON THE METABOLISM OF PHOSPHORUS IN ANIMALS 165 TABLE 11 ——— te p. c. from the rad, P taken Killed found in 100 mgm ashes Weight in mgm Bones fanteors | Melars | Liver | ees Incisors | Molars Live. BAWieel ease silars sci ise oie LDN CBEST COM a SO 1 CYST SAS 8 0 am 135 dl esl EC tt en ep een Aes 358 0 1d 5.9 | 329| 105 | 76 103 DeWieesks testes vaee se 3.0| 41] 0.9] 5.0 | 403 11355 86 OAT Bett Bae San ais MEME eet 3.5} 3.7] 0.9 | 5.0] 361 96 | 64 84 OT een age te ee ee DT nGeOu|) yuledal wed Su) eSdsel a WS: | c6O9 |) 168 MT ey aati 22| 3.6| 11| 1.2; 419| 109 | 57 | 145 ee 29) 4a il O19.) 1-8) 422) 5} Si. | 205 The above bone figures show a marked difference as compared with normal rats of the same age (cf. Table 6). We are engaged in carrying out further experiments on rats with rickets. GENERAL CONSIDERATIONS The rapid entrance of the labelled phosphorus into the bone is in no way puzzling. If solid calcium phosphate, one of the chief constituents of the bone, is in contact with the solution containing labelled phosphate ions a rapid distribution of the latter takes place between the surface of the solid phase and the liquid phase, as was seen from the following experiment. 3950 gm _ freshly precipitated Ca,(PO,). were shaken with 5 ce. of water saturated with Ca,(PO,). at room temperature and containing an infinitely small amount of labelled sodium phosphate. After lapse of four hours 84.1 per cent of the labelled phosphate ions were found in the solid phase and only 15.9 per cent in the solution. The calcium phosphate of the bone tissue being in a very intimate con- tact with the blood stream, i.e. with cells containing labelled phosphate, a similar exchange to that described above will take place between the unlabelled phosphate of the bone and the labelled phosphate present in the liquid phase. Beside the above mechanism we have to consider two others just as important. During growth, the bone tissue formed will be built up from labelled phosphorus as long as the blood stream contains the latter. Finally we have to envisage a third possibility, namely the entrance of labelled phosphorus into the bone through a constant break-down 1The weight of the total skeleton is obtained by dividing the figures obtained for the legs by 0.26. 166 ADVENTURES IN RADIOISOTOPE RESEARCH of the bone tissue already formed and the formation of new tissue in the case of adult animals as well. The following examples may help to make the three ways of entrance of the labelled atoms into the bone easier to understand. 1) When solid salts are in contact with labelled ions of the solution within a short time a distribution equilibrium of the labelled ions between the surface layer of the solid and the solution will take place, as is seen for example in the experiment described above. This phenomenon was studied extensively by PAnerH and his collaborators! in the case of lead salt which were shaken with solutions containing labelled (radio- active) lead ions. 2) If we deposit for example lead electrolytically from a solution containing labelled lead ions, the metallic deposit will be a labelled one, just as the bone grown from blood containing labelled phosphorus will contain labelled phosphorus. 3) In investigating the exchange between metallic lead and a solution of labelled lead ions, or vice versa, we find? a different behaviour to that described above in the case of lead salts. The exchange in the case of metal is not restricted to the uppermost atomic layer of the lead surface; many atomic layers are involved in the exchange process. This is due to the fact that the lead actually goes into solution from certain parts of the surface, while lead ions are discharged, at other parts. This is a much more effective process in bringing about an exchange between the lead atoms in the solid and in the liquid phase than that observed in the case of solid salts where only the uppermost atomic layer is involved (within any resonable time) in the exchange process. The entrance of labelled phosphorus into the bone will also be much facilitated if it is not only the uppermost phosphate layer that is invol- ved in the exchange process; if in fact the bone is destroyed at certain places and rebuilt at others. In view of the important enzymatic actions? going on in the bone tissue such a reversible breakdown process will easily occur. Summary By adding radioactive phosphorus (phosphate) to the diet of rats, the metabolism of the phosphorus atoms taken in with the diet can be followed up in the animal body. An appreciable part of the phosphorus taken finds its way not only in growing but also in adult animals into the bones, teeth, muscles, and different bodily organs. 1K. Panera and W. Vorwerk, Z. phys. Chem. 101, 445, 480 (1922). 2G. Hevesy, Phys. Z. 16, 52 (1915) 3.R. Rosison, The Significance of Phosphoric Esters in Metabolism. New York (1932). STUDIES ON THE METABOLISM OF PHOSPHORUS IN ANIMALS 167 In growing animals it was found that the atoms already present at an early stage of the formation of the skeleton become distributed in the course of time over the different parts of the skeleton and other organs demonstrating thus the dynamical nature of the building up of bone tissue. Some of the phosphorus atoms present in the bones leave the skeleton for good, being eliminated through the kidneys or the bowls or becoming located in other organs of the body. The replacement of individual phosphorus ators by other phosphorus atoms also takes place in the bone tissue of adult animals including that of the teeth. It was ascertained that about one-seventh of the phosphorus found in the faeces of a human subject is due to material which has entered the intestines through the digestive juices after being located in the blood stream or in the organs of the body for a shorter or longer time. Originilly published in Kgl. Danske Videnskabernes Selskab. Biologiske Meddelelser 13, 13 (1937) 20. INVESTIGATIONS ON THE EXCHANGE OF PHOSPHORUS IN TEETH USING RADIOACTIVE PHOSPHORUS AS INDICATOR G. Hevesy, J.J. Horst and A. Krocu From the Institute of Theoretical Physics, Dentistry School and Zoophysiological Laboratory, Copenhagen ANATOMICAL INTRODUCTION THe hard part of a tooth is composed of three distinct substances viz. the dental substance proper, dentine, the enamel, and the cement. The dentine constitutes by far the largest portion; the enamel is found in a comparatively thin layer partly covering the dentine; and the ce- ment covers the surface of the root in a thin layer. In the case of the canines of cats we found the weight of the enamel ash to be 11.2% of that of the dentine ash, the weight of the enamel before ashing being equi- valent to about 9.7% of that of the dentine. The dentine is penetrated throughout by fine tubes (dentinal tubes) starting from that side of the dentine which faces the pulpa cavity; they have an initial diameter of 2 to 8 and do not much diminish in Size at first as they approach the surface; the distance between adja- cent tubules is about two or three times their width. From the tubules numerous immeasurably fine branches are given off and penetrate the hard intertubular substance. Near the periphery of the dentine, the tubules, which by division and subdivision have become very fine, terminate imperceptibly in free ends. It is reported that tubules have been observed passing into the enamel in the teeth of marsupial animals, and to a less marked degree in human teeth. In this case they pass, not into the enamel prisms, but into the inter-prismatic substance. The ena- mel is made up of microscopic columns, very hard and dense, arranged close side by side, and fixed at one extremity on to the subjacent sur- face of the dentine. The enamel columns have the form of six-sided prisms. Their diameter is about 0.005 mm. They are united by a small amount of substance which appears to be similar to the intercellular substance of an epithelium. The small amount (about 1%)! of 1 J. H. Bowes and M. M. Muray, (1935) Biochem. J. 29, 12, 2721. EXCHANGE OF PHOSPHORUS IN TEETH 169 organic matter in the enamel is probably found to a large extent in the above mentioned connective substance. In marsupials and some rodents there are regular canaliculi in the interprismatic sub stance. The central cavity of a tooth is occupied by a soft and very vascular dental pulp, containing cells, blood-vessels, nerves, and fine connective- tissue fibres. The cells are partly disseminated in the matrix and partly form a stratum at the surface of the pulp. These superficial cells, the odontoblasts, send out elongations into the tubules in the dentine. It is through the intermediary of the pulp that constituents of the blood get into the hard tissues of the teeth. Chemical composition of the teeth a) Dentine. On analysing a great number of dry human dentine samples BowEs and Murray! found a loss in weight of the fresh tissue on ignition amounting to 29—29.79%%. The losses on ignition found in some of our experiments can be seen in Table 1, in which we have also included for the sake of comparison the values found for the tibia and jaw. TaBLE 1. AtBINO Rat 200 g ’ | Loss on ignition Organ in % of fresh weight jeyxopsuoatall rel Gageaccage 33.6 Incisors Distalbendirrercccecus ccc ei 25.0 | IANVCTAC Cares vi vegethe euch ono, ks | 26.4 MOL GTS) +. 4c1steh siasevauh: oxoeh se ee cretel elaine eis ES 27.0 er FV eadla lovers orc caret ee ester | 79.1 Hae \| Average: onic. Wapnites es os | 63.2 Cat 4 kgm MN CISOIS sre aercpenstevs ssa et edeKe isa hot ataiorev exe 32.0 Wamime we iecsieus even sia olay shove eraheaeliste eibin louse 35.0 MOL aS Rice cicierotoysie sieieverete sie ota te oad sien 38.0 SVAN groko gots scenes st GLa icls fas tales Risse eeke Sake | 50.4 IMIDE) oN oho oes ema OOOO OAc 66.8 PAD IACIAID YSIS )..¢ oxsieterepaieets «| ais shee dss | 36.7 The average values found for the chief constituents of the dentine by Bowes and Murray! are seen in Table 2. 1T. H. Bowes and M. M. Muray, (1936) Biochem, J. 30, 1977. 170 ADVENTURES IN RADIOISOTOPE RESEARCH TABLE 2. ANALYSES OF DENTINE oF HUMAN TEETH (% in Dry DENTINE) Slight hypoplasia Severe hypoplasia NCI sR tae eA = 1209 70.28 Of ee Ge LOR eae oct iia rs,o 27.79 26.96 Eee a Go ae nse ace oka aso mreke 13.81 133355 COS inakersnsccvoneneneesioeares aoe 3.18 3.10 Mio ss cocked alciereeahsrtenetos 0.835 0.728 ©) GRA Certs eens rman ~ 0.023 Bowes and Murray give the following average figures for the compo- sition of the enamel: TaBLE 3. ANALYSES OF ENAMEL OF HUMAN TEETH Slight hypoplasia Severe hypoplasia INS Dagens cpeiers eens tacicee custo: 95.38 94.67 Oe Omnic Seen on omer 37.07 35.81 Pe creavassL tae eretetars Oe eee 17.22 17.72 | Slight hypoplasia Severe hypoplasia COM ar as eet tee | 1.952 2.434 Miettinen ee eee | 0.464 0.477 (OUT Sera aoa (aya 0:8 0.19 As is seen from the above figures phosphorus is the second most abundant mineral constituent of the teeth, its share in the dentine amounting to 13.5—13.8% and in the enamel to 17.2—17.7% while in the dentine ash 18.2—18.4% in that of the enamel 18.4—19.4% were found. In the ash of the incisors of rats an even higher phosphorus content of 20°4 was found. Bone ash contains an only slightly lower amount of phosphorus than tooth ash, the values found varying between I 9eand) 18.5%: In distinction to the chief constituents of the teeth the minor consti- tuents vary within wide limits. The composition of the mineral consti- tuents of the teeth corresponds approximately to a mixed crystal of the minerals hydroxide-apatite and carbonate apatite, the former predomi- nating strongly. As in apatite minerals the OH ions of the tooth apatite can be replaced to a certain extent by F ions for example. The degree of replacement of OH by F , will depend primarily on the fluorine content of the blood during the development period of the tooth and also on that which circulates in the fully calcified tooth. The fluorine EXCHANGE OF PHOSPHORUS IN TEETH ilya | content of the blood will depend on the fluorine content of the food and water taken up. It is thus easy to explain why the fluorine content of the teeth varies within wide limits (see Table 4). The high fluorine content of the teeth of human beings living at Colorado Springs is due to the high fluorine content of the water which amounts to up to 2 mgm per liter. The high fluorine content of the teeth of some North African sheep is to be explained by the high fluorine content, above 0.02%, of the soil on which they graze. On such soil plants of high fluorine content grow, are eaten by the sheep, and lead to an abnormally high fluorine of the blood plasma, which in turn leads to an abnormally high replacement of OH by F in the teeth’? > 4. TaBLeE 4. FLUORINE CONTENT oF TEETH ASH oO Oo Wikia ore cis Sc ce. cars cd bere Cro Dic OIdiote.s Guimond or teeth 0.03 Wirmine iin Sooosgcoapeoncendeaoboou0G | teeth 0.69—0.74 1 SAC Ur ote Reset nha Ba iran PR eae - teeth 0.006—0.03 Wan Mer Modis Gacécodaosnousndsddodo0c | dentine , 0.065 Win ING? MOE coscanooscudcoodcGddagcac _ enamel 0 Man? (ColoradomSpringsis a. «cece chee «1+ «1-1 dentine 0.112 Mane Colorado Sprimgsi 2). cere elsiel leis «61 hei enamel 0.065 INTE ele Rae a gh Biase ea RA | dentine 0.030 IVE pra Seeaereratecns caret ccetsteve ates a apes bavetelist aheivs..er' st olin | enamel 0.005 Wall estanpyer tine. er aterstorensictenal we oretal er enedeyavls sien ike | dentine 0.022 CRISP Gs coo s.Som obs clo m Go Gmina c'cld.o Ooi Ooic | enamel 0.0057 SAMS GG op oo dneded6 oe GOS Ob a dic Ocoee) ob | teeth 0.89 Sheep, young from neighbourhood of Norwe- | gian aluminium factory where fluorides are UHMINGECE ooaoogoccuns on dan dD DO b OO UO MDS incisors 0.45—0.49 Sheep?) North) Africa ys. -1. te 2-< oss 2 > teeth 0.04 Sheep® North Africa, attacked by fluorine BLISS ey OOOO aD HOTS DIS DISLGNDO Oo NO NON teeth 0.32—0.45 The much higher fluorine content of animals living in sea water is also due to the comparatively high fluorine content of the latter. In the same way that F replaces OH in the apatite lattice, magnesium for example replaces calcium. Human dentine ash has a magnesium content® of 1.18—1.39% whereas human enamel ash has only 0.42%. While the calcification of the tooth tissue is presumably the result of specific cell activity, it is quite possible that later on a replacement of GQ) R. KiemMent, Naturwiss. 21, 662 (1933). @2)R. R. SHarpiess and E. V. McCouiuvum, J. Nutrit. 6, 163 (1933). (3) H. BorsSEVAIN and W. F. Drea, J. Dent. Res. 13, 495 (1933). () K. Ronotm, Fluorine Intoxication, p. 260. Copenhagen (1937). G) M. Gann, A. CHavnot and M. Laneuars, (1934.) Bull. Inst. Hyg. Maroc. Nos. I—II. (6)M. M. Murray, Biochem. J. 30, 1568. (1936.) ye ADVENTURES IN RADIOISOTOPE RESEARCH calcium by, for example, magnesium takes place, governed chiefly by solubility (chemical affinity) conditions, and it is quite conceivable that in the course of time more and more calcium is replaced by magne- sium if the magnesium: calcium ratio is in favour of such an exchange. The enamel being characterised by a decidedly poorer lymph circulation than the dentine, the much lower magnesium content of the former can easily be accounted for by a reference to the above considerations. The 0.42% magnesium found in the human enamel possibly got into the latter wholly or to a large extent during the formation of the enamel tissue. The presence of as much as 4°/, of magnesium in elephant den- tine is possibly due to a high magnesium content of its food or to a high magnesium retention in its blood. It is also of interest to remark that the magnesium content of the teeth found in prehistoric skelet- ons is only one third of that found in teeth of recent generations, fur- thermore that carious teeth! show a greatly increased magnesium con- tent. Besides the elements discussed above, spectroscopic investiga- tion? revealed the presence of traces of Na, Ag, Sr, Ba, Cr, Sn, Zn, Mn, Ti, Ni, V, Al, Si, B and Cu in dental] tissue. That the concentration of the minor constituents of the teeth does not fluctuate between still wider limits is due to the narrow limits within which the concentration of most e!ements in the blood plasma is restricted. This is caused partly by a prevention of the resorption of excessive quantities of the elements, conspicuously shown in the case of calcium, and partly by prompt removal chiefly through the kid- neys of excessive amounts of the mineral constituents present in the plasma. But even in spite of this levelling mechanism of the blood plasma some of the mineral constituents are deposited to a remark- able extent in the tooth tissue, as is seen above in the case of flu- orine, and it is quite possible that even an excessive replacement of, for example, the calcium by magnesium, sodium, or potassium might lower the resistance of teeth to disease. While the conclusions given above are based partly on hypothetical assumptions, in the case of lead, which also replaces calcium in the crystal lattice, the accumulation in the teeth with time can clearly be shown. While small children have only negligible amounts of lead in their teeth, the lead content increases with age®, the increase being markedly greater in the case of carnivorous than herbivorous animals, presumably on account of a greater lead intake in their normal nourish- ment. In the case of lead poisoning the lead content of teeth is greatly OT Francia, Ann. Clin. Odoniat. 8 695, (1931); M. M. Murray and J. H. Bowers, Brit. Dent. J. 61, 473, (1936). () &. Lowatrer and M. M. Murray, Biochem. J. 31, 837, (1937). (3) F. Prriemr, (1934) Arch. f. Hyg. 111, 232. EXCHANGE OF PHOSPHORUS IN TEETH es increased. All these observations support the hypothesis, that even in fully formed teeth an exchange of mineral constituents is regularly taking place. To test this hypothesis we have studied the exchange of phosphorus by means of labelled phosphorus atoms. PHOSPHORUS EXCHANGE IN TEETH We investigated the movement of the phosphorus atoms both in the teeth of fully grown and growing animals by using labelled phosphorus atoms as an indicator. By adding radioactive phosphorus, prepared from sulphur by the action of neutrons, to food administered to animals at a known date, it is possible to distinguish the phosphorus atoms which were present in the food sample and which have been retained and deposited in the organism, from those already present in the body and the teeth at the start of the experiment. We can thus follow the movement of the phosphorus atoms taken in for example a glass of milk and investigate if and to what extent these particular atoms get into the teeth and how they are distributed there. The dentine contains 14% and the enamel 17.59, of phosphorus in the form of phosphate (PO,). It is the movement of these phosphate radicles which we actually investigate. For the sake of brevity we shall often use the word phosphorus in discussing the behaviour of the phos- phate radicle. We may recall that the phosphorus taken with food, amounting in the case of an adult to somewhat more than 1 gm. per day, is to a large extent (in most cases up to about 80°) absorbed from the gut and gets into the blood stream. Adult human blood contains 44--50 mgm%®% of phosphorus of which only 2—5 mgm% are present as inorganic P. Very different views have been put forward on the for- mation of the bone and tooth tissue, but they all consider the blood plasma as saturated or nearly saturated with calcium phosphate and the precipitation of the latter from the plasma as being of paramount impor- tance for the ossification process. The solubility of calcium phosphate in the plasma is very strongly affected by the presence of proteins, carbo- nate and bicarbonate ions, and possibly also other constituents. It is also dependent on the acidity of the blood, slight changes in which may be sufficient to produce precipitation. It seems very probable that it is not simple calcium phosphate but a complex salt of the apatite type, a solid solution of hydroxide apatite and carbonate apatite, that precipitates. In addition to the inorganic phosphate, blood contains a phosphoric ester at a comparatively high concentration which is mainly found in the corpuscles; as it cannot yield phosphate ions by dissociation, this ester does not affect the saturation of the blood with respect to calcium 174 ADVENTURES IN RADIOISOTOPE RESEARCH phosphate. However, as Ropison! discovered, the cartilage and osteid contain an enzyme, phosphatase, which hydrolyses this ester, thus setting free inorganic phosphate, whereby the concentration of the phosphate ions increases and a supersaturation occurs, followed by a precipitation of the calcium phosphate in the matric of the tissue. With the discovery of the bone phosphatase a second agency (in addition to the acidity change) of great importance was found, regulating the cal- cium phosphate precipitation leading to ossification. Ropison found that the enzyme had the greatest activity in ossifying cartilage, bones, and teeth of very young animals, the activity per unit weight of tissue decreasing with age. Although the plasma contains on-an average only 0.5 mgm of phosphorus present as phosphoric ester per 100 cc. this is completely hydrolysable by the bone phosphatase and thus supplies phosphate ion amounting to about 1/, of the inorganic phosphorus present in the plasma, an amount amply sufficient to bring about a supersaturation and a subsequent precipitation of calcium phosphate, or more correctly of the apatite-like bone substance, from the already nearly or fully saturated plasma. The conclusions arrived at in this paper are independent of the special mechanism assumed for the ossifi- cation process. DISTRIBUTION OF LABELLED PHOSPHORUS IN THE INCISORS OF RATS The rapidly growing incisors of rats are very suitable for studying the distribution of phosphorus. According to FrrpeRiIca and GuDJON- sons? the average extrusive incisor growth per week is 2.7 mm. in the case of adults and 3.4 mm. for young rats. As seen in Fig. 1 A the cross sec- tion of the pulpa is very large at the lia proximal end and gets narrower toward the distal end, the last millimetres of the teeth being free of pulpa. The a problem we have to investigate is how A the distribution of newly formed cal- cium phosphate in the incisor takes place. = ‘ Two extreme cases must be envisaged: a) the labelled phosphate is deposited erase: 1R. Rostson (1912) The Significance of Phosphorus Hsters in Metabolism, New York. 2L. S. Frrperica and S. V. Gupsonsons, Kgl. Danske Vid. Selsk. Briol. Med. 28, 813 (1931). EXCHANGE OF PHOSPHORUS IN TEETH Wis. in close proximity to the pulp from which it is derived, while the tissue formed at an earlier date is pushed along in the direction of growth; b) the labelled phosphate is equally distributed throughout the incisor. Cutting incisors tranversally into pieces and analysing these separa- tely revealed the fact that the largest part of the labelled phosphate is found in those regions of the incisor where the pulpa is strongly deve- loped, but that some of the labelled phosphate is found all through the incisoral tissue (Tables 5 and 6). TABLE 5. DISTRIBUTION OF LABELLED PHOSPHORUS, CONTAINED tN THE NorMAL Diet, FoUND IN THE INCISOR AFTER 2 DAYS. WEIGHT OF THE Rar 210 gm + DeENorEes Upper — Lower TEETH Prostata uincieor Weight of ash | % of ence P | % of the labelled in mgm | taken found P per mgm ash Proximal Po caccccewase 38.2 | 0.42 oon 1Aropcianeyl MMS cesooadgasnee 40.8 0.47 0.012 1Prropenatl Moe ZS oeceoosaae 29.2 0.37 0.013 lPRorenmeyl I” sossbeonocse hee, 0.38 0.014 Which tpente: a Pee ane 92.6 0.125 | 0.00135 is teal Ms ale i aia ctsteve as alee ster=e WV5-2 | 0.072 0.00063 TiS peal SMe ee occchic, snare lee cue cece 36.4 0.008 0.00022 Percentage of labelled P found in the total incisors = 1.85. Average per 1 mgm ash= 0.005. Biggest ratio between proximal and distal end = 60. TABLE 6. DistTRIBUTION OF LABELLED PHOSPHORUS, ADMINISTERED IN THE NorMAL DIET, FOUND IN THE INCISOR AFTER 7 Days. WEIGHT OF THE Rat 240 em a | Weight of ash | % of labelled P | % of the labelled | Part of the incisor i taken found -P per mgm ash | in mgm | Promimalele tes ot vast eke a: | 950 | 0.28 | 0.011 Proximal fai eects es 23.6 | Oi) ee) 0.013 Broximalol (22.6. 220i gece ee 21.8 | 0.29 0.013 roxanne SO ae erotics 31.6 | 0.32 0.010 Middlee ers sisson do oc 81.6 | 0.204 0.0032 Micclioe sadam eI Li ta cok GSO etn 206 0.0031 Distal amd fis se eo ete 29.3 0.020 0.00074 Percentage of labelled P found in the total incisors = 1.69. Average percentage per 1 mgm ash = 0.006. Biggest ratio between proximal and distal part = 18: 176 ADVENTURES IN RADIOISOTOPE RESEARCH In the experiments now to be described the distal part of the incisor was removed by operation one day before labelling the phosphorus present in the blood. In these experiments the radioactive P was not added to the food but given in the form of subcutaneous injections. 2 days, 5 days and 8 days after the administration of the labelled phos- phorus the end part of the freshly grown incisor was again removed by operation and its radioactivity ascertained. The distal parts removed were all outside the range of the pulp. The figures obtained are seen in Table 7 and those from a similar experiment in Table 8. TABLE 7. 2 — { ee °% of the labelled ‘ p ; : | Weight of the | ‘° : Davs after intake of labelled P ye i ' P found in 1 : | tissue in mgm | : | mgm fresh tissue 2 42.8 | 0.00089 Fete Rete Sega ete ane 16.4 - 0.00030 eee Cee eis er! 2 ae 20.4 0.00066 NeiGeh po eUUKe Yeh} Giisnieoo a tie come 26.6 0.00090 Percentage of the labelled P found in mgm of average incisor tissue — 0.0076. The removed distal ends contained 8 to 25 times less labelled P than the average tissue. PABLE 8): | | . °% of the labelled , . : | Weight of the i i Days after intake of labelled P eae; ; P found in 1 : | tissue in mgm | ? ‘ | mgm fresh tissue ie ede irk Sera ey eee 14.4 9.00040 OU Sed Byard UN de WR Se eee 11.0 | 0.00044 13 tratekilled): “c- 64.02. seas 21.3 0.00062 Percentage of the labelled P found in 1 mgm of average incisor tissue = 0.0062. The removed distal ends contained 10 to 16 times less labelled P than the average tissue. Though the figures in the tables above clearly show that the deposition of labelled phosphorus is not restricted to the regions in the vicinity of the pulp, but that the labelled phosphorus is to be found even in the most remote part of the incisors, we attempted to obtain incisors with an appreciably larger pulp-free part. As is well known, rats, being rodents, grind their teeth and thus continually remove parts of the pulp-free end of the growing incisors. By eliminating the upper incisors the animal was prevented from gnawing and incisors were thus obtained in which the distal pulp-free end had a length of 10.5 mm. as shown EXCHANGE OF PHOSPHORUS IN TEETH ] ~] ~l ) TasBLE 9. DISTRIBUTION IN THE INCISOR AFTER 3 DAys. LABELLED PHosPHoRUS INJECTED SUBCUTANEOUSLY. WEIGHT OF THE Rat aBoutT 200 gm tibia, Loss Bact ve. pie | fresh tissue | P found in | P found in EL 3 see Fig. 4 : | | ignition ; in mgm 1 mgm tissue 1 mgm ash By see eeeeee 215.4 0.0028 0.0134 79.1 joy cuoapoeoc 64.4 0.0033 0.0064 48.3 ’ & poodoouoac 78.8 0.0019 0.0033 42.0 Is. Gordaoder 33.1 0.0015 0.0028 47.5 a avaviels $1.5 0.0014 0.0034 58.7 182 ADVENTURES IN RADIOISOTOPE RESEARCH WABI 1D. ne eee eS eee eee : | Labelled P found | in 1 mgm ash | Labelled P found | in 1 mgm fresh Loss in Organ | tissue as a per- | as a percentage | Weight on centage of the | of the amount | ignition | amount given given | MT CISOM eee eee ere eee a tes oss" 0.00046 0.00062 26.0 WIG ENE: fais olga ee arora Cro IOROnG 0.00025 0.00034 27.4 ANWie ee tere eens see a ae wie aia | 0.00125 0.0020 36.3 Miloiamineddl: a':o2.ns< «+ acer | 0.0024 | 0.0077 | 68.7 4iilloy sh asic lbXe a poo oa eae aS | 0.00068 0.0014 | Dell cut in 5 pieces the labelled P content of which is seen below, I denoting the proximal end. TABLE 16. Weight of Seana P found eight | ee | in 1 mgm fresh “ie _ | tissue as a per- ssue in a a | centage of the - | amount given ] Lees Poe Ae 6.1 0.0062 10 ee aie arrays | 14.4 0.0027 MTT eo. Be sta 236.2 0.00040 DVenteaneestan 33.0 | 0 Wag fates pe 19.6 | 0 One cc. of blood contained 0.5% of the activity injected; assuming a blood content of 10 ce., only 5% of the labelled phosphorus injected was present in the circulation after the lapse of 1 hour. Only 1 hour after administering the labelled phosphorus the tibia phosphorus was found to be 1000 times less active than the blood phosphorus, while for the molars the corresponding ratio was found to be 5000 and for the incisors (inclusive of growth) 2700. We also determined the activity of the acid soluble phosphorus extracted from the muscles of the rat and found 1 mgm to contain 0.042% of the labelled P given. From this figure and those found for the activity of the tooth and tibia phosphorus to be seen in Table 16 it follows that a comparatively fast phosphorus exchange is taking place in the muscle compared with that ascertained in the bones and the teeth. EXCHANGE OF PHOSPHORUS IN TEETH 183 EXCHANGE OF PHOSPHORUS IN THE TEETH OF CATS For the teeth of young cats! killed a few hours after the subcutaneous injection of the labelled phosphorus the results seen in Table 17 and 18 were obtained. Tasie 17. Car WetcHInG 2 KGM KILLED AFTER 3!/, Hours 6 : °4 of injected | % of the labelled Weight of ash Tooth : labelled P present) P per mgm ash | in mgm y ; in the tooth | of the tooth (Wp perpmolaryr).)t-)- | 123 0.016 0.00013 Lower molar ......... | 110 0.014 0.00013 Upper canine ........ | 108 0.044 | 0.00041 Lower canine ........ 91 | 0.040 | 0.00044 TaBLeE 18. Cart WeIGHING 2.5 Kem., KILLED Arter 1!/, Hours eae % of injected | % of the labelled | Weight of ash Tooth | labelled P present) P per mgm ash mgm i | in the tooth of the tooth ph a = — a a aes OMincrsorsieseeereio er 94.8 | 0.0032 | 0.000034 Caminemesienicicnince sce | 148 0.019 | 0.00013 Wie eee eee | 126.3 | == | 0.00037 We also investigated fully grown cats. A cat weighing about 4.5 kgm and killed three days after administration of the labelled P gave the figures seen in Table 19. In this experiment the labelled P injected was not of negligible weight but amounted to 15 mgm (corresponding to about 75 mgm sodium phosphate). The labelled phosphorus used in this experiment was kindly presented to us by Prof. LAWRENCE and was prepared by the action of high speed deuterium ions on phosphorus and accordingly contained a comparatively large amount of normal phosphorus. The injection of 15 mgm P into a cat leads to an accelerated excretion and the figures are thus not entirely comparable with those of the last described experiment, which was furthermore carried out on a growing cat. 1The heads of the cats were kindly given to us by Professor LUNDSGAARD; they were obtained in the course of an investigation on the distribution of labelled phosphorus carried out by him and one of the present writers. In the first men- tioned case 1 mgm plasma P was found to contain after 314 hours 1.6% of the activity injected, i. e. about 2300 times as much as that present in 1 mgm of the upper molar P. 184 ADVENTURES IN RADIOISOTOPE RESEARCH Tasie 19. Car WeIcHING 4.5 KGM., KILLED AFTER 3 Days. SSS SSS | ree | | ° EF, 0, i Fi Mewarene of ash |” of injected | % of the labelled | labeled P present! P per 100 mgm Bag in the teeth | ash of the tooth — Es pee a area = aes Molars aire eeeeern 690.5 | 0.0080 | 0.0012 Wpper canines’ 725... - | 768.4 0.0076 | 0.0010 Lower canines ....... | 635-2 0.0068 0.0013 The corresponding enamels weighed 29.3, 34.3 and 55 mgm. The canine enamel was found to contain less than 1/4) of the labelled P con- tent of the corresponding dentine. In another experiment a strong preparation was administered in three portions, 5 days, 2 days and 1 day before killing the animal, each portion containing 40 mgm P. The results are seen in Table 20. TaBLE 20. Cat WEIGHING 4 KeM., KILLED AFTER 5 Days. | Weight of Weight of | % of injected | % of the labelled | | teeth | ash | labelled P present| P per 100 mgm | in mgm in mgm | in the teeth | ash of the tooth Molar 5 site eustoete eters sR | 323.3 | 186.0 | 0.0027 | 0.0015 Gaming, Sates tas: | 422.2 | 274.3 | 0.0038 | 0.0014 Gh lbaensanes sogoocoueue | 172.5 Ie | 0.0021 0.0018 | The enamel obtained is discussed on page 183. In investigating the incisors of rats we found the activity to be due almost exclusively to the phosphate of the mineral constituents, the pulp being only slightly active. In the earlier experiments conditions were however very diffe- rent from those obtaining in the above mentioned case. The uptake of labelled P in the teeth of a cat is much smaller than in the incisors of a rat and correspondingly the ratio of labelled P in the plasma to labelled P in the teeth is much larger in the case of the cat. Now a high blood activity will lead to a comparatively high pulp activity and we must expect a greater share of the pulp! in the total activity of the tooth in the case of cat teeth. To test this point we removed the pulp of some of the canine teeth and compared the activity of the dissected and the total canine. We found an activity ratio of 3 : 4, showing that a quarter of the activity of the canines of a fully grown cat is due to the pulp. A comparison of the figures of Tables 17 and 18 with those of 19 and 20 shows that the uptake of labelled P in young animals is greater than 1 Human tooth pulp was found by H. C. Hoper, Proc. Soc. Exp. Biol. Med. 35, 53 (1936) to contain 0.70% phospolipins besides other phosphorus compounds. EXCHANGE OF PHOSPHORUS IN TEETH P85 in fully grown ones and also that while in the former case the canines take up 3 to 4 times as much labelled P (per mgm ash) as the molars. in the latter case no such difference is found. As has already been men- tioned above the figures for the two sets of experiments are not entirely comparable, but no objection can be raised against a comparison of ihe ratio of the canine and molar uptake, which differs very markedly in the case of growing rats from the ratio for fully grown animals. The following is a possible explanation of this difference: the labelled P uptake in the teeth of young rats is due partly to a growth of the teeth and not to an exchange process; since in the cat the canines grow faster than the molars the uptake is greater in the former case. One would be inclined to object:to this explanation in view of the short dura- tion of the experiment, as the growth in the course of few hours may be entirely negligible. This objection is however unwarranted. The molars of the growing cat weighed 116 mgm and those of the fully grown animal 691 mgm. It does not take longer than a few years for the growing cat to become fully grown so the yearly growth of a molar will be above 100 mgm. Let us now calculate the amount of tooth ash formed on _ the assumption that the labelled phosphorus found in the tooth is due to growth. A molar of the growing cat took up 0.016% labelled P during 3.5 hours. The labelled P which we injected into growing cats had in most cases a negligible weight originally, but very soon after the injection it mixed with the inorganic phosphate of the plasma (corresponding to about 5 mgm P) and from that moment we must consider the labelled P as having a weight of about 5 mgm 0.016% of the labelled P will therefore correspond to 0.0008 mgm P. The next step is that a large part of the labelled phosphorus leaves the plasma and is replaced by other phosphorus atoms coming from different bodily organs and also from the blood corpuscles. The result is that 0.016% of the activity given no longer represents 0.0008 mgm P but a greater weight, our scale of indication becoming less and less sensitive. From the experiences of Prof. LUNpSGAARD and one of us onthe exchange of phosphorus present in the plasma we canestimate roughly that the amount of P which corresponds after the lapse of 3.5 hours to 0.016% of activity is about 0.008 mgm in the case discussed. To transform from phosphorus weight to ash weight we have to multiply by six. The weight of the tooth thus increases by 0.04 mgm in 3.5 hours and about 100 mgm in a year. The order of magnitude of the growth observed and that calculated on the assumption that the uptake of labelled P is due to growth is thus the same. A very simple but instructive calculation can be carried out in the case of a fully grown cat into which as much as 120 mgm labelled P was injected. We can calculate how many milligrams of these 120 mgm 186 ADVENTURES IN RADIOISOTOPE RESEARCH are to be found after the lapse of 5 days in a single tooth. Making use of the figures quoted in Table 20 we find that a canine takes up 0.005 mgm and a molar 0.003 mgm. THE BEHAVIOUR OF THE ENAMEL The difference in the mechanical properties of dentine and enamel is very pronounced. The hardness of anterior enamel is nearly half as great as that of hardened toolsteel, while dentine compares closely with brass!. The hardness is taken as the pressure in kilograms necessary to push a steel ball into the test piece. The above mentioned difference is not due to a pronounced difference in the relative abundance of the mineral constituents of dentine and enamel, as discussed on p. 5, but to the following conditions. The amount of organic constituents +water found in dentine is about six times as large as the amount present in enamel, the calcification of the enamel tissue being thus carried through much more effectively than that of the dentine tissue. Bowes and Murray? found organic matter in human enamel to an extent of only 1%. As there is more organic matter? in enamel near the junction with the underlying tissue, the dentine, than in the part equidistant from the dentine and the surface of the teeth, the outer part of enamel must contain even less than 1% organic matter. The latter appears to bet a protein containing tyrosin and resembling reticulin. Another outstanding difference between dentine and enamel seems to be the size and degree of orientation of the crystallites present in these. As to the orientation it has been stated’ that enamel of high quality gives X-ray diagrams of a high degree of orientation, while enamel of poor quality does not. On igniting dentine an X-ray diagram characteristic of B-Ca,(PO,). is often but not always observed; this is never shown by ignited enamel. As it was found? that -Ca,(PO,), is formed when an excess of PO,-ion is present, it was concluded that the dentine apatite often adsorbs an excess of phosphate ion which promotes the formation of B-Ca,(PO,), on ignition. In the case of enamel forming larger crystallites, no excess of PO,-ions being present, no p-( az(PO,); 1H. C. Hopes, J. Dent. Res. 15, 251 (1936). 2 J. H. Bowrs and M. M. Murray, Biochem, J. 29, 721 (1935). 30. F. BopeckErR, J. Dent. Res. 6, 2, 117 (1923). 4P. Pincus, Nature 138, 970 (1936). 5J. Tuewuis, Naturw. 25, 42 (1937). 6 W. F. Bats, M. L. Lerevre and H. C. Hones, Naturw. 24, 976 (1936). 7G. Trémuet and H. Moétier, Z. anorg. Chem. 206, 227 (1932). EXCHANGE OF PHOSPHORUS IN TEETH 1&7 formation was observed on ignition. While important information may be obtained by the study of X-ray diagrams the interpretation of the latter must be made with care. PHOSPHORUS EXCHANGE IN THE ENAMEL In view of the connection found between the content of organic matter and phosphorus exchange in the teeth it did not appear very promising to look for a pronounced exchange in the enamel. The enamel investig- ated by us was in some cases removed mechanically while in others we succeeded in separating the enamel of cat teeth after igniting the tooth very carefully. The enamel, having a different expansion coeffi- cient from the dentine, splits off during the ignition process and can thus be removed. The method of separation used recently by various workers!, in which the tooth is pulverized and placed in an organic liquid of suitable density when the heavier enamel settles to the bottom of the tube, is not suitable for our purpose. The reason is that some dentine often sticks on the pulverized enamel; assuming that the dentine is strongly active and the enamel not, we see that the presence of traces of dentine in the enamel might falsify the analysis. We made several experiments with the enamel of cat teeth but in most cases with negative results, the exchange in equal weights of enamel being at least 20 times as small as that found in the molars of cats. In one case we got a positive effect, the canine of a fully grown cat five days after injecting the labelled phosphorus showing a radioactivity of 26 relative units (counts per minute), one enamel sample showing 0.6, and another 0.7 counts. The first mentioned enamel was separated by grinding it off from the dentine, while the second one was obtained by the same method from the uppermost enamel layer. The ash weight of the canine was 277.3 and that of the enamel samples 33.1 and 19.1 mgm. We are however reluctant to accept this positive result. On account of its smaller weight and greater distance from the underlying dentine, the outermost layer should be less active than the second enamel layer, unless the labelled phosphorus present in the saliva (which 13.4 mgm. % P 100 ec.) can interact with the outer layer of the enamel. We intend to follow up the problem of the phospho- rus exchange in enamel using phosphorus preparations of greater activity. 1Comp. P. J. BrRekHus and W. O. ArmstronG, J. Dent. Res. 15, 23 (1935). 1M. KarsHan, J. Dent. Res. 15, 388 (1936). 188 ADVENTURES IN RADIOISOTOPE RESEARCH EXCHANGE OF PHOSPHORUS IN HUMAN TEETH Other things being equal the exchange of phosphorus in teeth will be determined by the efficiency of lymph circulation in the tooth. Exchange experiments can thus be carried out to obtain information on the latter point. It does not look improbable that the growth of caries will be facilitated by a poor circulation; to decide this point we compared the phosphorus exchange in two teeth of the same individual (16 years old) removed simultaneously, one on account of caries, the other, a healthy one, to space the patients teeth better; about a two hundred thousandth part of the labelled phosphorus was found in each of the teeth investigated, a quantity sufficient to be measured but not large enough to permit the exact comparison necessary to decide the point discussed above. The weights of the whole fresh teeth were 800 and 540 mgm and of the ash obtained on ignition 465 and 330 mgm this corresponds to a loss on ignition of 58 and 61%. The time which elapsed between the injection of the radioactive phosphorus and the extraction of the teeth was 7 days. Through the very great kindness of Professor LAWRENCE we were able to continue these experiments using a much stronger radioactive phosphorus sample Aen Zale LABELLED PHOSPHORUS IN THE TEETH OF A 25 YEAR OLD PaTIENT a) Necrotic Roots. asl ieee | | Relative labelled P content | Fresh weight | Ash weight ke | | | | In total | In 100 mgm | root | root ash = Paes tae - b as a MRS otitis asco AO ae 2287 eleS.los | ameed 1.96 De slate CMe SRT AGT 284.1 190.1 | 4.9 272 SN a RA ev a eerie 199.4 L271 2.5 1.97 Bh EG Soin on de wssteey shasta 230.5 143 .0).°) © wake! 0.98 Bers wc kts torts eee fatioe 124.8 16.5 | 3.9 5:13 Gi aSe's cp ch cieioe ake 435.2 268.5 | 5.9 2.19 [er AOR OE Ree 205.7 eae 3.86 CL | See i Sy tense 169.5 LOSmauae ake ee Ty, OS oh ae A 172-55 | 0 10G{6nn) ae 3-8 Sa TOWMP itoery eee sires ote poeee 183.5 115.0 | Zl | 1.83 prepared by him with the aid of his powerful cyclotron. 900 mgm labelled sodium phosphate per os were administered to a patient 25 years old. 4 days later 10 necrotic teeth and 5 days later still, three more, fairly well preserved, living teeth were extracted. Of the 2.5 - 108 EXCHANGE OF PHOSPHORUS IN TEETH relative radioactive units we can estimate that about 1.8- 106 absorbed. As is seen in Table 21/6 relative units were found in a fairly well preserved tooth on an average, showing that about 1 : 300.000 part of the labelled phosphorus atoms enter a single tooth; in the case b) Necrotic Crowns a ae | | Relative labelled P content Fresh weight | Ash weight Alisa Tih. | ; In total | In100mgm crown | crown ash One single crown........... | 65.8 | 39.1 Sill 9.4 | aN : Fragments of several crowns 241.7 149.8 We 8.2 c) Almost normal roots —— aia EEE Relative labelled P content Nr. Fresh weight, Ash weight | — | | In total |In100mgm root | root ash | | | Deb revevecay cases tener irc avel orolarciatensvancbers onde4 | 240-0) 4 2.8 | 1.16 De ee a asco Ws: Foes 6511, F436. | 2 say 7) 40:9 Ce eee cL cee 685.9 | 430.2 | 6.7 | 1.56 d) Almost normal crowns rc re — oe on eS ES See rs i eS | Relative labelled P content Nr. |Fresh weight | Ash weight mi | | ; In total |In10 mgm crown crown ash be 11 8. OPPCRORSE OID ED EOE 533.5 338.7 1.9 0.56 DW a aleirete Tore tales sayentvetaheieke teehee sri 862.9 670.9 1.8 0.27 Spee aiate ota ey ere Seek hie cle ie arahs 464.1 429.9 1.8 0.42 of a 16 years old boy about 1: 200 000 was found. In the latter an activity of only 0.5 units (counts per minute) was shown single tooth, and the estimate was accordingly only a very rough From the above result it follows that about 1 : 300,000 189 were case by a part of the phosphorus taken up with the food finds its way into each tooth of an adult. 190 ADVENTURES IN RADIOISOTOPE RESEARCH Summary It has been shown that an exchange of phosphorus atoms present in the teeth with those present in the blood plasma takes place. During the growth of the incisors of rats the newly deposited phosphorus atoms are to a large extent found in close vicinity of the dental pulp, but even in the most remote part of the incisor presence of newly substituted phosphorus atoms can be established. An exchange of phosphorus atoms thus takes place even in those parts of the incisors which are entirely outside the range of the pulp. The exchange in the molars was found to be less pronunced than that in the incisors, this being presumably due to the fact that these do not grow. In the teeth of young cats within fews hours., besides an exchange of phos- phorus atoms, an increase in the labelled phosphorus content due to the growth of the teeth could already be ascertained. An exchange of phosphorus has also been proved for human teeth, 1: 300,000 of the phosphorus administered being found in each tooth. The replacement of 1%, of the phosphorus content of a human tooth by phosphorus atoms taken up with the food takes about 250 days. Originally published in Biochem. J. 34, 532 (1940) 21. RATE OF REJUVENATION OF THE SKELETON G. CH. Hevrsy, H. B. LEvi anp O. H. REBBE From the Institute of Theoretical Physics, University of Copenhagen Tue first experiments in which labelled (radioactive) phosphorus *P was applied as an indicator [Curevirz and Hevesy, 1935], showed that some of the phosphorus atoms of the mineral constituents of the bone exchange rapidly with those present in the plasma. This result was corroborated and extended by later work on this subject [HEvesy ef al:. 1937: Cook eé al., 1937; Dows et al., 1937; 1939; Arrom e al., 1938; CoHN and GREENBERG, 1939; LEFEVRE and Bate, 1939]. The question as to what extent the P contents of the mineral constituents of the bone are replaced in a given time by plasma P remained unan- swered however. This is an important question, as the rate of this repla- cement is a measure of the rate of rejuvenation of bone tissue. The extent to which bone P is replaced by plasma P in the course of a given time can be determined by comparing the activity of 1 mgm of bone P with that of 1 mgm of inorganic plasma P. The activity of the plasma P, i.e. its =P content, changes appreciably, however, with time. The *P atoms, like all P atoms present at any moment in the plasma, exchange with the P atoms present in the various organs and in doing so are removed from the plasma and replaced by tissue P. The application of the above- mentioned consideration implies a constancy of the activity of the plasma inorganic P. To secure such a constancy, we administered labelled phos- phate all through the experiment, instead of doing so at the start of the experiment, as in all investigations mentioned above. By taking blood samples at intervals, we ascertained that the activity of the inorganic P of the plasma remained constant. At the end of the experiment, the bone sample was purified from all non-mineral constituents and the radioactivity of 1 mgm of bone P compared with that of 1 mgm of plasma inorganic P. EXPERIMENTS WITH FROGS Experiments of 5 to 240 min duration were carried out with frogs. As early as 5 min after injecting 0.3 ml. physiological NaCl solution containing a negli- gible proportion of labelled sodium phosphate into the lymph sac, the mineral 192 ADVENTURES IN RADIOISOTOPE RESEARCH TABLE 1. LABELLED P ContTENTS OF PLASMA AND TIBIA OF A FROG Wt. 45 gm: temperature 22°; Determinations 5 min after the start of the experiment | | #P content per Fraction |mgm FP (specific | activity) IDEN), SoonooGoacaacac 100 ID DMS coasccaocg40c 0.026 IDTV OME Soougaconda ue 0.013 Yo 0-30 Epiphysis O- iaphysi ’ 20 Diaphysis Epiphysis Diaphysis 0-10 fe) I SD [i Hr. Fic. 1. Extent of replacement of frog’s bone P by labelled P constituents of the tibia contained some labelled phosphate, as shown in Table 1. With increasing time, the ®P content of the mineral constituents of the tibia increases (Fig. 1). After the lapse of 1 hr., the specific activity of the epiphysis P amounts only to 1/600 of the corresponding magnitude of the inorganic P of the plasma. Thus only 1/600 or less of the epiphysial P was replaced by plasma P within 1 hr. In the next 3 hr. a further 1/900 part of the epiphysis P exchanged. The first point in the curve was obtained by analysing the right, the second point by analysing the left tibia. For the diaphysis, the corresponding figures were found to be 1/900 and 1/1200, respectively. In the course of 4 hr., therefore, only a minute part of the tibia P is replaced by plasma P. A still smaller replacement is found when the frog is kept at 0°. RATE OF REJUVENATION OF THE SKELETON 193 EXPERIMENTS WITH RABBITS After the lapse of 2 hr., 1/530 and 1/1800, respectively, of the tibia epiphysis P and diaphysis P were found to be replaced by plasma P. With increasing time, «un increase of the replacement of the bone P takes place, as shown in Fig. 2 and Table 2; this increase diminishes, however, with increasing time, as would be excepted. The bone tissue contains numerous small crystals formed by mineraliza- tion of the matrix. The crystals are built up on similar lines to the mineral apa- tite.t While the atoms situated on the uppermost layer of the crystals [ Pa- neth, 1922] exchange easily with those present in the surrounding liquid, those situated inside the crystal are prevented from doing so, except at very high tem- peratures. The exchange between bone phosphate and plasma phosphate which we observe in experiments of short duration is due to a replacement process bet- ween the phosphate ions situated on the surface of the apatite crystals and those of the plasma or lymph. Should the surface exchange be exhausted, an increase of the time of the experiment may at first have no effect at all on the extent of replacement. If, however, in the course of time a dissolution and rep- recipitation of the apatite crystals takes place, new and far-reaching possibilities TABLE 2. Extent oF REPLACEMENT OF THE BONE P OF THE RaBBiIt BY LABELLED P The labelled phosphate was injected intravenously during the experiment Fraction PA Vey YS | 4 hr. % Epiphysis ............. | 0.180 | 0.200 WDiephysisa act sh tio | 0.056 | 0.106 % 30 Epiphysis Diaphysis O 5 10 20 30 40 50 days Fig. 2. Extent of replacement of rabbit’s bone P by labelled P ‘From X-ray measurement, it was concluded [Caglioti, 1936] that the inorganic part of the bone has the composition of about 3Ca.(PO,),-CaCO,-xH,O with the hexagonal structure of hydroxyapatite, the length of the axes being a= 9°2 x 10~*cm. and ¢ = 6'9 x 10~—° em, the axis being oriented parallel to the length of the bone. The organic parts consist of polypeptide chains, supported and stretched. 13 Hevesy 194 ADVENTURES IN RADIOISOTOPE RESEARCH of an exchange between plasma P and bone P will arise. From these considera- tions it follows that an exchange taking place within a long interval cannot be extrapolated from results obtained in experiments of short duration. We have, therefore, carried out experiments in which we kept the activity of the plasma inorganic P of rabbits at a constant level for several days or weeks. EXPERIMENT OF LONG DURATION To obtain a constant level of the plasma inorganic P, the first day every 30 min and later twice every day, labelled sodium phosphate of negligible weight was administered by subcutaneous injection to rabbits. After removal of the marrow, the bone was first extracted for 12 hr. with hot ether-aleoho]. The bone was then treated with hot alkaline glycerol solution for further 6 hr. The fractions obtained were dissolved in HNO, and their P contents precipitated as molybdate. The molybdate was dissolved in dilute NH, and precipitated as ammonium magnesium salt. An aliquot of the sample obtained was used in the colorimetric P determina- tion, while another aliquot was reprecipitated as ammonium magnesium phosphate and its radioactivity measured with a Geiger counter. In prolonged experiments the analysis of the plasma inorganic P is conveniently replaced by that of the urine P. In Tables 3 and 4 the specific activities of the different bone P fractions are recorded. TaBLeE 3. ExTENT oF REJUVENATION OF THE TIBIA IN THE CoURSE oF 9 Days Wt. of rabbit: 2 kgm % P rejuvenated Fraction wes ae | (Specific activity) PIP iy Sisal ssvserilere is 11.2 Dia phiysise eects seers 3.2 Tibia phosphatide P... | 74.8 Marrow phosphatide P. | 80.1 In the course of 9 days therefore, only 11% of the epiphysis and 3% of the diaphysis are rejuvenated, while most of the phosphatide molecules present in the marrow and in the bone are newly formed. In the course of 50 days, only 29% of the epiphysial and 7% of the diaphysial mineral constituents were replaced (Table 4). The tibia and femur show about the same behaviour. About half of the scapula remained unchanged. The almost complete replacement of the apical and medial parts of the incisor dentine P can hardly be interpreted as due to an exchange between dentine P and plasma P, since the replacement rate of the dentine P was found to be lower than that of the tibia P [Hrvesy et al., 1937; LerevrE and Bate, 1939] and since the tibia P, as seen above, was replaced only to a restricted extent. The high *?P content of the incisor dentine must be due to an actual growth, to a formation by a calcifi- cation process. As a plasma containing *8=P was instrumental in calcifying the newly grown parts of the incisor, the P of the latter was bound to have the same specific activity as shown by the plasma P. As seen in Table 4, the P of the apical part of the incisor dentine investigated has, within the errors of experiment (-+ 5%). RATE OF REJUVENATION OF THE SKELETON 195 TaBLe 4. Extrent oF REJUVENATION OF THE SKELETON OF A RABBIT IN THE CoURSE OF 50 Days eee Fraction | % Prejuvenated Femur epiphysis inorganic P .........-..-- 29.7 Femur epiphysis phosphatide P .......... 100 Femur epiphysis glycerol extract* P ........ 51.0 Femur diaphysis inorganic P ............. | 6.7 Femur diaphysis phosphatide P .......... ca. 100 Femur diaphysis glycerol extract* P ...... | $4.5 Tibia epiphysis inorganic P .............. | 28.6 Tibia diaphysis inorganic P .............. | 7.6 (CORI socucdéconcege coooubaMuurcD oC UDG OOS 27.5 Nicholl: GopgcnonucdoSoMUGEnsoo” O10 Oo Mor | 43.8 lbaetsare lero, hake a5 SeoocmouDD bos booT | 103 Imeisor dentine, amediailiyr. <1. 1s cite; -letel 98.5 Ibavertsjore Gleiahnuates saree Sob sanuocc00vuGaGKuS 41.2 Incisor enamel, apical -- medial .......... 82.0 lbnxomor emer, maensal Godin an ouocoppo Goon | 6.6 * This fraction presumably contains some mineral P. he same specific activity as that of the plasma P. This part, having a length of about 0.9 mm., is entirely newly formed during the experiment. The bulk of the medial part of the dentine was freshly grown as well, while the incisal part, having a length of 1.2 mm., is only partially newly formed with participation of the labelled plasma. About half of the P atoms present in the incisal part of the dentine were not labelled; they must thus be those which were located in the apical or medial region of the incisor before the start of the experiment. The tissue containing these atoms was pushed forward in toto. Partly before this “slipping” process and partly during it, some of the P atoms of the dentine have the oppor- tunity to exchange with labelled P atoms and, therefore, the P of the incisal part of the dentine shows an activity which amounts to about 1/3 of the specific activity of the plasma P. We see here an interesting case of tissue formation in which macroscopic aggregates are “‘slipped’’ from one place to another in toto, experiencing only a restricted atomic or molecular replacement. This effect is much more clearly shown in the growth of the enamel. The apical and medial parts of the enamel are formed by a calcification process from labelled plasma and, therefore, these parts of the enamel became strongly active. From the fact that the incisal part of the enamel is only slightly active, we have to conclude that this fraction is not formed in the course of the experi- ment through a calcification process. Its crystals were formed at an earlier date from non-labelled plasma and the whole fraction “slipped”? in toto during the course of the experiment from the position in which it was calcified into the place it took up at the end of the experiment. The incisal end of the dentine is probably to a large extent also formed by “‘slip”’ in toto of the medial parts, though this conclusion is not supported as clearly by the activity figures arrived at in the case of the enamel. A part of the incisal dentine P had an opportunity to exchange to an appreciable extent before the “slip’’? took place and also during that process. Enamel P exchanges only to a minute extent [ARMsTRONG, 1940}. 9 It is also of interest to note that the activity figures exclude the possibility that 13* 196 ADVENTURES IN RADIOISOTOPE RESEARCH the incisal part ot the enamel is formed by extensive calcification of the outer region of the dentine. In that case, the enamel could not be much less active than the corresponding dentine part. The foundation of the incisal enamel is laid in the apical end and reaches its final position without interchange with the dentine. As both the P and Ca of the teeth have their origin in the plasma, the application of labelled Ca as an indicator can be expected to lead to similar results to those found above. EPIPHYSIAL AND DIAPHYSIAL BONE TISSUES The epiphysis was found to exchange the P content of its mineral constituents at a higher rate than the diaphysis, as seen in Table 5. TasBLE 5. Ratio OF THE SPECIFIC ACTIVITIES OF THE EPpIPHYSIAL AND DIAPHYSIAL P OF THE TIBIA OF RABBITS The level of the activity of the plasma inorganic P was kept constant all through the experiment Time | Ratio | Diver ed ccatte forty Pevat ep aten seas steer ols 34 Be SOS. Fess Ayal edetlas'st ofa Ste aueienoheteehe 1.9 NOSdayStreis tooseeasaeioc te 3.5 BO! At, a aerate catesepsey: Bice: be 138 In experiments with frogs, in which the labelled phosphate was injected into the lymph sac at the start of experiments, taking 1— 22 days, the ratio 1.3—1.6 was found. In these experiments, the tibia and femur were both investigated and the average was taken. In an investigation of the tibia P of rats, to which the labelled phosphate was administered at the start of the experiment, the ratios 3.1, 2.9, 2.5, 1.7 and 1.8 were found after ; hr., 4, 10, 50 and 110 days, respectively. It is of interest to remark that from the finding that the diaphysial P is only about half as active as the epiphysial P 110 days after the start of the experiment, we can conclude that an appreciable part of the skeleton has not been renewed within 110 days. In experiments on chickens, Dots et al. [1939] found the above ratio to be 3, 22 hr. after administration of labelled phosphate; rachitic chickens gave the ratio 2.5. The more rapid exchange of the epiphysial P is just what would be expected. The epiphysis is characterized by a poorer mineralization of the matrix than is the diaphysis and contains more organic matter and water than the diaphysis. The circulating lymph, containing the labelled P, will therefore reach the apatite surface more easily in the first-mentioned case. Should the size of the apatite crystals be smaller in the epiphysis than in the diaphysis and therefore the ratio surface: volume be larger in this type of bone tissue, one would also expect a more rapid exchange of the mineral constituents of the epiphysis. Whether such a difference in the size of the apatite crystals actually occurs is not known. X-ray investigations [BALE et al. 1934] lead to the result that the size of the ultimate RATE OF REJUVENATION OF THE SKELETON 197 crystals of dentine and bone is about 10 ~® cm.; while the very effectively mi- neralized enamel contains larger crystals (10> cm.). It is, therefore, quite pro- bable that the difference in the extents of mineralization of the epiphysis and diaphysis manifests itself in a moderate difference in the sizes of the ultimate crystals in the two types of bone tissue. It is of interest to note that the difference in hardness of the different types of bones is, to a large extent, the result of a different degree of orientation of the crystallites of the bone. X-ray patterns in- dicate that orientation occurs during growth and first in those bones where the need for solidity (e.g. leg bones) is greatest [CacLirotr and GiGanrE, 1936). Summary Labelled (radioactive) phosphate was administered to rabbits and frogs repeat- edly during the experiment in order to keep the radioactive plasma inorganic phosphorus at a constant activity level. The comparison of the activity of 1 mgm bone inorganic P with that of the plasma inorganic P permits us to conclude to what extent the mineral constituents of the bone were renewed during the expe- riment. Within 50 days, 30% of the femur and tibia epiphysis were found to be renewed, while the corresponding figure for the diaphysis amounted to 7%. Half of the mineral constituents of the scapula were found to be unchanged. The phos- phatides of the bones and the marrow were entirely renewed. The phosphorus of the apical part of the dentine of the rabbit’s incisor was found to have the same activity as the plasma phosphorus. From this result it follows that this part of the dentine was grown with the participation of labelled plasma phosphorus during the experiment. The greater part of the incisal end of the dentine was not formed by a calcification process in situ but by a “slip”? of the apical part of the dentine. The same behaviour is shown even more pronouncedly by the enamel. No interaction of any significance takes place between the incisal part of the enamel] and the dentine either during their formation or at a later date. The atoms pre- sent in the former are to a vely large extent those which were previously located in the apical (medial) part of the enamel. References ARMSTRONG (1940) J. biol. Chem. (in the Press). Artrom, SaArzAna and SEGRE (1938) Arch. int. Physiol. 47, 245. Bate, HopGe and WarRREN (1934) Amer. J. Roentgenol. 32, 369. Cageuriotr (1936). Atti Congr. naz. Chim. pura appl. 1, 320. CaGLiotri and GiGanteE (1936) R. C. Accad. Lincei, Classe sci. fis. 23, 878. Curevitz and Hevesy (1935) Nature, Lond., 136, 754. Cutevitz and Hevesy (1937) Kgl. Danske Vidensk. Selsk. Biol. Medd. 13, 9. CoHN and GREENBERG (1939) J. biol. Chem. 128, 116 and 130, 625. Coox, Scorr and ABELSON (1937) Proc. nat. Acad. Sci. 23, 528. Dots and JANSEN (1937) Proc. Acad. Sci. Amst. 40, 3. Dots, JANSEN, S1z00 and VAN DER Maas (1939) Proc. Acad. Sci. Amst. 42, 2. Hevesy, Horst and Kroeu (1937) Kgl. Danske Vidensk. Selsk. Biol. Medd. 1 AeA LEFEVRE and Bate (1939) J. biol. Chem. 129, 125. PanetH (1922) Z. Elektrochem. 28, 113. Originally published in the Svedberg p. 456. Uppsala (1944) 22. RETENTION OF ATOMS OF MATERNAL ORIGIN IN THE ADULT WHITE MOUSE By G. Hrvesy From the Institute of Theoretical Physics, University of Copenhagen and the Radium Station in Copenhagen Wuar percentage of the atoms present in the new-born organism is retained during the later phases of life and what percentage is inherited by the subsequent generations? An attempt was made to answer these questions by following the fate of the phosphorus atoms in the white mouse, radio-phosphorus being used as an indicator. The phosphorus atoms present as constituents of different compounds in the body of the new-born animal are released successively from the compounds in which they are present. The phosphorus atoms thus released are either excreted or re-incorporated into various compounds present in the body, the latter process being much more frequent. Although, for the sake of simplicity, we speak of phosphorus atoms, practically no phosphorus atoms but only phosphate radicals are released from and built into such phosphorus compounds. The organism is sup- plied with phosphorus in the form of phosphate radicals and the phos- phorus atoms adhere, as far as is known, to their partners throughout the numerous metabolic processes in which they participate. The white mice used in the experiment were kept on the following diet. Wheat flour, oatmeal and a small amount of milk were administered, while on alternate days only Cootry’s standard food was provided. Once a week, cabbage or lettuce was administered as well. About 0.1 mgm of labelled sodium phosphate with an activity ofa few microcuries was administered by subcutaneous injection to a pregnant mouse. As a result of introducing radiophosphorus into the organism of the pregnant mouse, we obtain offspring of which the phosphorus contents were labelled. A litter consisted of about 8 offspring having almost the same weight, as shown in Table 1. The radio-phosphorus contents of the offspring may therefore also be expected to be almost equal. This fact makes it possible to determine the total *=P content of the offspring at any date by measuring the total activity of any member of the litter. One offspring was killed shortly after birth and dissolved in concentra- ted nitric acid. The phosphorus content of a known aliquot of the solution RETENTION OF ATOMS OF MATERNAL ORIGIN IN THE ADULT WHITE MOUSE 199 was precipitated as magnesium ammonium phosphate. The precipitate was filtered through an aluminium dish of 1.1 em diameter having a perforated bottom covered with filter paper. The dish containing the precipitate was then placed unter the Geiger counter. By comparing TaBLE 1. — WEIGHT oF 6 NEW-BORN OFFSPRING OF A MOUSE | Relative No. Weight in gm ee | activity = 51 i = 1 33 | 100 2 | Se ae e98t5 3 val 99.0 4 le?) 93.5 5 133 99.5 6 V3} the activity of an offspring killed at a given date with the activity of another killed at a later date, it was possible to calculate what percentage of the phosphorus atoms of maternal origin was lost in the interval between the two dates. All offspring were killed successively, dissolved, and treated in the way described above. The writer is much indebted to Mr. K. ZeRAHN for dissolving the mice and precipitating their phosphorus content. All offspring were killed and investigated within about three months. After the lapse of this time, the activity of the phosphate precipitates had decreased so greatly that it could no longer be measured with suffi- cient accuracy. The activity of the first offspring was compared with that of the second, the activity of the second with that of the third, and so on. The mouse obtains its =P content not only by birth but also by lactation. In order to simplify the problem, to reduce the *2P content of the offspring mainly to such **P as was obtained by birth, the active mother was replaced by an inactive mouse soon after gestation. As the replacement of the mother was not made immediately after the birth of the offspring, we actually measured the loss of ®2P acquired by birth plus the #?P acquired by lactation in the interval between birth and replacement of the active mother by an inactive one, i.e. within a few days. 200 ADVENTURES IN RADIOISOTOPE RESEARCH RESULTS The result obtained are shown in the following tables. TABLE 2. — MOTHER INJECTED FEBRUARY 9, DATE OF GESTATION: FEBRUARY 18. REPLACEMENT OF THE ACTIVE BY AN INACTIVE MOTHER: FEBRUARY 22. (EXPERIMENT 1.) | | : No. of | i Relative | Weight ae | Killed His |: olfspring | | activity , in gm 1 22/2 100 3 2 3/3 82 a 3 16/3 73 15 4 30/é 48 18 5 13/4 A 225 23 6 13/5 40 Loss of *P in the course of 81 days: 60 per cent. In about 3 months,! a time sufficient for mice to reach adulthood, TABLE 3. — MOTHER INJECTED FEBRUARY 9. DaTE OF GESTATION: FEBRUARY 16. REPLACEMENT OF THE ACTIVE BY AN INACTIVE MOTHER: FEBRUARY 23. “soa Ae (EXPERIMENT 3.) v. < > gt < \ ‘ iy Au of | Killed | od a ate | WIE \ . . i, offspring | | activity in gm 1} ; Y | E ; 7s i =a ff? ae a | 24 /6 j | 92 Ween” } =] ] 24 /2 100 | 2.8 \ ass. /.O/ 2 5/3 90 6.9 “ ay * 4 YR of 3 19/3 | 70 13 oa J ; ws ie L @ i 2: 2/4 4 15 — 5 16/4 53 20 6 18/5 | 51 16 7 | 7/6 | 39° |" 29 Loss of *P in the course of 103 days: 61 per cent. only about 60 per cent of the 32P content of the mouse acquired by birth is thus eliminated. The phosphorus content of the new-born mouse is found to be about 4mgm the amount of P excreted by the mouse in the 1 Tf only the last two values in Tables 2 and 3 are considered the average loss of ®P in 80 days works out at 57 per cent. If all mice killed in April, May and June are considered, the average loss is 56 per cent in 72 days. RETENTION OF ATOMS OF MATERNAL ORIGIN IN THE ADULT WHITE MOUSE PO] course of three month about 1 gm The fact that nearly one half of the maternal phosphorus is retained in the body, in spite of the comparatively large amounts of phosphorus excreted by the mouse in the course of 3 months, is due mainly to the protection of a large proportion of the phosphorus of the bones against interchange with the phosphorus atoms in circulation. The uppermost atomic layers of the bone apatite crystals interchange easily with the phosphorus atoms of the plasma or the lymph; furthermore, a kind of biological ‘‘recrystalli- zation’ takes place, i.e. dissolution of some molecular apatite layers followed by new formation of such layers through crystallization. .\1] these processes, however, do not affect, or affect only at a very slow rate, large parts of the bone apatite which thus retain their P atoms. During the formation of the skeleton, a large proportion of the **P atoms present in the organism will find their way into the bone apatite and be fixed there to a very appreciable extent during the time of the experiment (3 months) or even for the lifetime of the mouse. When comparing the *P content of several rats injected simultaneously with labelled phosphate at different dates, it was found (HEvEsy 1939) that the =2P, in so far as it was not excreted, accumulated to a very large extent in the skeleton. This fact is illustrated by the following table. ; TABLE 4. — PERCENTAGE *P PRESENT IN THE Bopy FounpD IN SOME ORGANS OF THE Rat | Time after which the rat was killed Organ SSS ST —— = SSS % hour , 4 hours | 10 days | 20 days | 30 days | 50 days | 98 days | | | | IN RUK hs. Ose Sie wee | 67.6 OIE onc & O:Grd Daa ic eS Bol. A AROS IS SIO It SESE eee | 188 mgm TIT, Bee 8 i as cate May Oey Ad ae ra en Wate Aas ors | EY sO Oiic. OSB. loa cent CeO TA cle cate eRe ee See Re 24 Combustion value (calculated according to Rubner) .... | 400 cal. * In Sweden, mice and rats are fed almost exclusively on these cakes, the exact composition of which was hitherto unknown. The author is much indebted to Professor E. BRUNIUS and Mrs. ESTHER SIHLBOM who most kindly made the analysis of these cakes at Statens Institut for Folkhiilsan. 224 ADVENTURES IN RADIOSIOTOPE RESEARCH moderate changes in heterophylous values could be detected. In our experiments, no effect on growth or fertility due to the presence of #Ca could be observed. Our main litter size was 5.7. As shown by Russexi@*), the litter size of mice at term is reduced as a result of irradiation during preimplantation stages with 100 r or more, and when exposed shortly after implantation, by a minimum dose of 200 r. The composition of standard cakes fed to our mice is seen from Table 2. RESULTS The results of experiments in which mice born from mothers kept on a ®Ca diet for weeks prior to and after parturition, and continuously kept on a Ca diet till they reached an age of about 100 days, thus were fullygrown, are shown in Table 3. The mean conservation of Ca by the uniformly labelled skeleton of the mice in the course of 390 days, representing a mean value of the duration of the experiments, works out to be 64.7 +. 7.34 per cent, the standard error of the mean being 2.78. If we disregard the last experi- ment in which the mice were kept on a high calcium diet, the mean value is 67.2 + 7.86 per cent, the standard error of the mean being 3.23. Thus, 2/, of the calcium atoms present in the skeleton of the out- 45Ca content {O00 200 300 400 500 G00 days Fie. 1. Loss of Ca, obtained from labelled mother at birth, during lifetime of the mice. Each point indicates the “Ca content of another member of the litter killed at the stated time CONSERVATION OF SKELETAL CALCIUM ATOMS THROUGH LIFE 995 TABLE 3. — Loss or ®CaAa BY THE UNIFORMLY LABELLED SKELETON OF Mice witH TIME INDICATED BY MEASUREMENTS O01 THE RADIOACTIVITY OF THE SKELETON OF DIFFERENT MEMBERS or A Lirrer Kittep at Various Times. THE MICE WERE Born FROM ActTivE MoTHERS AND WERE ADMINISTERED ®CA UNTIL THE First MEMBER OF THE LITTER WAS KILLED a | No. of litter | Age in days | *5Ca content | | 111 100 14 Si 2 a AEA re ERO oA neg ee 329 66.7 | 519 | 57.0 | / 108 | 100 | (ae aOR eee Bae, tee oe eo rte e320 s00e7 | 517 78.8 | ey LOS: Si S100 Mee eryeetee tel etc nee ok oe eee Se e326) eee ass | | 501 | 69.4 | | | as 100 1 fi eae Centar Oe eee a Sea | 220 | 79.9 | | 393 (64.4 | 706: || V / 106 | 100 Se Re eS, cor ee ae Cer ee Bee ge | 325 63.8 | | 56 100 Ilia ape yaa en et Bei ic Braja (Sian bec | 129 81.4 | b 266 | 69:7 99 | 100 eee ee araan VETS sisa's5 iareyeiavecais sates Sooke Gceasen rat emints ake cree | 308 | 55.5 WeeS 92/0 ens. ee 503.- 9) pos * Cheese and egg shells were added ad lib. to the standard ‘‘gard-bred” diet. grown mice are present after the lapse of more than a year and can thus be considered to be unreplaceable during life. Figs. 1 and 2 and Table 4 show the results of some of our experiments in which the litter, born from active mothers, was kept from birth on a %(Ca-free diet. These experiments include the results obtained between the third and the 560th day after birth, thus almost the lifetime of the mouse. The percentage of “Ca lost between day 3 and day 560 works out to be 53 per cent and 44 per cent, respectively. The mean loss of 15 Hevesy 226 ADVENTURES IN RADIOISOTOPE RESEARCH 45a observed in experiments lasting 100 to 180 days amounts to 43 per cent (Table 4). The loss of “Ca during the first three days of life is less than 10 per cent; thus half of the maternal calcium atoms are preserved during life. 100 80 60 45Caq content 40 20 100 200 300 400 500 600 days Fic. 2. Loss of Ca, obtained from labelled mother at birth, during lifetime of the mice. Each point indicates the Ca content of another member of the litter killed at the stated time The calcium content of our newly born mice, weighing 1.23—1.37 em. varied between 0.28 and 0.85 per cent of the body weight, not much differing from the calcium content of the new-born rat (4.7 gm) for which data varying between 0.27 and 0.35 per cent are reported”. The calcium content of 1 gm fresh weight of newly born mouse amounts to 0.3 times that of 1 gm of the adult animal, which is 1.05 per cent. If all the maternal calcium atoms had the same chances to supply cal- cium to the offspring, and all were labelled, we would find 1 gm of newly born mouse to be 0.3 times as active as 1 gm of the mother. We find the “Ca content of 1 gm of new-born mouse to amount to 1.7 times that of 1 gm of the adult mouse. The #Ca taken in by the mother has thus only an opportunity of interchanging in the average with about 4/, to 1/, of the body calcium before being utilized in the building up of the embryo. bo bo co | CONSERVATION OF SKELETAL CALCIUM ATOMS THROUGH LIFE TABLE 4. — RETENTION OF MATERNAL CALcIUM ATOMS BY THE OFFSPRINGS | | Per cent of | as on. Number of 4 Rug | , Weight mothers’ activity No. of litter | Ageindays| . ; ~ | maternal atoms | in gm | present in the | | | freOring present | | offspring ay 4 2.95 8.25 100 | 31 13.9 6.59 80 1 eS chats es oromicneroic 39 16.8 5.00 67 43 18.6 6.0 UP 103 24.3 4.25 52 | | 1 2.20 6.95 100 ye ors ese 129 41.9 | 4.95 71 | 181 3] 3.96* 57 ne f | Wie Si lSOk aly 1023 100 Guia eltelielel.eve eee) e1e, ere l | 131 38.1 | 7 4 72 1 | 1.45 9.91 | 100 VN gate Veet 128 39.6 7.25 73 | 181 40.0 5.69 57 | 3 Su ae 8.55 100 ; | pueee 2 es | onell-5 | 6.98. | 82 CEN he Os Aan SO te cote 95 | 921.0 6.67 78 | 42 24.2 6.27 73 DISCUSSION a) Conservation of the calcium atoms of the outgrown skeleton through life The fact that a very appreciable part of the skeletal calcium is preser- ved in the outgrown animal throughout its lifetime results from experi- ments carried out by SrncER, ARMSTRONG and PremER™, by CaRLson®**” and by Baver“), Similar results were obtained in investigations on the renewal of the mineral constituents of the skeleton, performed by the present author and his assoc. who used ®2P as an indicator®”. From specific activity data of the plasma and the skeleton of the outgrown rat, the percentage of renewable skeletal sodium was calcu- lated by Bauer” to amount to 30—40 per cent of the sodium present (disregarding the extracellular sodium); a similar figure — 45 per cent — is reported by Everman“ and by BapEN and Moore”. Since sodium is mainly an extracellular element, the specific activity of plasma sodium decreases only slowly with time, not so the specific activity of calcium. The calculation of the percentage of renewable skeletal calcium from bo bo GO ADVENTURES IN RADIOISOTOPE RESEARCH specific activity data is therefore encumbered with difficulties (cf. p. 221). From data collected during five days, Bauer") estimates, however, that less skeletal calcium than skeletal excess sodium is exchange- able in the rat, thus less than 30—40 per cent. As it was shown above (p. 217), the mobilization of some further skeleton calcium is still going on in the mouse after the lapse of more than 100 days and the non- exchangeable part of the skeleton amounts to 67 per cent. It is interesting to note that, when injecting ®Ca at the start of the experiment interperitoneally to outgrown rats whose skeletal calcium content was increased appreciably during the experiment, SINGER and Armstrone™) found a Ca retention of 42—45 per cent in the skeleton after the lapse of 52 days and the release of only small amounts of radiocaleium after that date. BucHANAN@”?, who exposed mice to air containing “#CO,, found that 30 per cent of the bone carbonate are replaced within 12 days, while 45 per cent only are renewed in the course of three months. b) Conservation of maternal calcium atoms by the offspring through life Our results demonstrate the very pronounced ability of the skeleton {o conserve maternal atoms. In the first mentioned experiments, one third of the #Ca content of the outgrown mouse was found to be replaceable by inactive food calcium. In the latter experiments with growing mice, released #Ca had a further outlet, viz. utilization in the formation of additional skeleton, which takes place in the growing organism. Investigations were carried out earlier on the loss of **P through the lifetime of mice born from active mothers’?. Some results of these investigations are shown in Table 5. TaBLE 5. — Loss or =P THROUGH THE LIFETIME or Micr Born From ActivE MotruHers. MoTrHer Insectep Wirth #P on Fresruary 9. GESTATION: FEBRUARY 18. REPLACEMENT OF THE ACTIVE BY AN InactivE MorHer: FEBRUARY 22. f ‘ : | Killed : | Relative Weight No. of offspring | Abe ive | date | activity {| in gm 8 enh OR eee ere 22/2 | 100 3 DUS ONS a eee 3/3 82 7 Beye As. sieht. 5 36 16/3 | 73 15 1 SENS Tea ae een 30/3 | 48 Is Stns wana See yen Anan 41 25 Gaeta. ek 13/5 | 40 35 * Incl. C*? of 3 offspring. CONSERVATION OF SKELETAL CALCIUM ATOMS THROUGH LIFE 299 The fact that a very appreciable percentage of the maternal phosphate is preserved —though less than of the maternal calcium —is presumably due to the lower share of the bone phosphorus in the total body phos phorus than the part of bone calcium in body calcium. 17 per cent of the phosphorus content of the mouse are present in the soft tissues, but only 1 per cent of its calcium content is located there. The phosphorus and calcium atoms present in various components of the soft tissues with the exception of desoxyribo nucleic acid phosphorus of some tis- sues—are poorly conserved and, consequently, maternal calcium may be expected to be better conserved than maternal phosphorus. From the fact that during the first 40 days of life—thus during a phase of intense skeleton formation—only less than a third of the mater- nal calcium atoms of the mouse is lost, we can conclude that the largest part of the calcium atoms leaving the circulation is utilized to skeleton formation and remains largely conserved in the skeleton. LesBitonp and assoc.” injected labelled phosphate into newborn rats and followed the *P uptake by the humerus and the lower jaw. Denoting the total **P taken up by the humerus in the course of the first hour by 100, the uptake after eight hours was found to be 150, after one day 117, and after three days 116. In spite of the rapid growth of the humerus, the *?P present after the lapse of a day is thus conserved through the following days; similar results were obtained in investiga- tions on the **P uptake by the lower jaw. The incorporation of calcium atoms in the rapidly growing bone tissue can also be studied by following its uptake into the incisor of outgrown animals. CarLson®”** °°” performed extensive and highly instructive studies on the calcium metabolism of outgrown rats, among others with the result that the calcium atoms incorporated with the rapidly growing incisors are conserved to a very large extent in contrast to those incorporated with the outgrown skeleton. It is rather difficult to determine the calcium intake and excretion by the suckling mouse. Our adult mice (36—37 gm), however, were found daily to consume 4+ 0.6 gm of standard bread containing 8.3 + 1.2 mgm. calcium; further 0.2 mgm calcium was contained in the 4 ml. of daily consumed water. The calcium recovered daily in the faeces amounted to 8 mgm. A very appreciable part of the faeces calcium may be assumed to be of endogenous origin, thus having passed the circulation before excretion. The share of endogenous phosphorus in the faeces phosphorus was calculated from the specific activity of faces P and urine (plasma) P“**» these calculations lead to the result that 74 per cent of the phosphorus of the human food and 72 per cent of the rat food are absorbed into the circulation. About the same percentage of the food P can be expected to be taken up by the mouse. As to the utili- (43) zation of calcium, data are available only for the uptake by humans”, 230 ADVENTURES IN RADIOISOTOPE RESEARCH Here, the mean percentage uptake was found to be 56. From the above data it follows that, out of the daily uptake of 8 mgm calcium by our mice, at least 4 mgm. have passed the circulation, representing a mini- mum amount of 2 mgm in the course of 500 days. From our results it thus follows that these 2 mgm were prevented from interchanging with 2/, of the 370 mgm calcium present in the skeleton of a mouse weighing 36 gm. The protected part of the skeleton calcium did not come into contact with the plasma or lymph and, correspondingly, an exchange between the unlabelled food calcium and labelled skeleton calcium could not take place; the same is true for the new-formation of the protected apatite crystals of this part of the skeleton under participation of food ‘alcium. A possible rearrangement within the protected area would not manifest itself in our experiment. The inaccessibility of parts of the skeleton minerals manifests itself also by the observation that radium, which like calcium is a strongly bone-seeking element, can find a life-long abode in the skeleton. The fact that a large fraction of radium administered to human subjects remains for decades in the skeleton is due presumably to the incorpora- tion of the radium into parts of the skeleton which are covered by apatite layers and thus become inaccessible and, even if released, are incorpo- rated again with the apatite structure. AuB and associates“® report a case in which no decrease in the radium content of a woman was found to take place between 1934 and 1945. This woman had been administered radium in 1924. CONSERVATION OF ANCESTORAL ATOMS The radiocalcium atoms going over from the first generation of mice into the second (cf. p. 226) do not indicate the total amount of maternal calcium atoms passing from the mother to the offspring, since the mother is not uniformly labelled. Chemical data indicate a passage of about 1.3 per cent. Since the calcium of the second generation is uniformly labelled, the passage of the ancestoral calcium atoms from the second into the third generation is properly indicated by the radioactive tracer. About one thira of the “Ca content of the second generation is lost prior to gestation, while about 0.5 per cent or less of the remainder passes into the third generation. From the calcium atoms present at birth in each generation, thus 1/35, part or less goes over to the following generation. As our mice contained 6 . 10?! calcium atoms, the eleventh generation did no longer contain a single ancestoral calcium atom. It is of interest to compare the life cycle of the ancestoral calcium atoms of the mouse with that of easily accessible water molecules. Applying deutoriated or tritiated water as an indicator first the half CONSERVATION OF SKELETAL CALCIUM ATOMS THROUGH LIFE as life of water molecules present in the rat was found to vary between 3.6 and 2.5 days®*®, that of the mouse is expected to be somewhat shorter. Thus, in the course of 165 days, all 1024 water molecules present at the start of the experiment in the mouse are replaced. About 4 per cent of the maternal water molecules go over to the offsprings and, from these, the second and third generations of offsprings will take up a share which depends on the age of gestation; the fourth generation, however, will hardly contain any more ancestoral water molecules. When the rate of disappearance of labelled water was followed in the rat during a long period, which was made possible by using tritiated water as an indicator, it was observed” that, after the lapse of 30 days, the labelled water disappeared at an appreciably slower rate than with a half-life of 2.5 days. The controlling factor of the dissappearance of labelled water from the organism is now the release of firmly bound tissue tritium which again becomes a constituent of the water molecules. Due to this fact, it lasts 60 days until the number of labelled water molecules of this type, present in the mouse, decreases to a 107*th of its initial value. If we disregard those water molecules whose hydrogen atoms were temporarily incorporated in tissue constituents and released appre- ciably later to become constituents of water molecules again, then all ancestoral water molecules are lost by the mouse during two generations. While the loss of ancestoral calcium is determined mainly by the loss at birth, many ancestoral water molecules are lost during the lifetime of a generation, none of them reaching the third generation of offsprings. Summary Since it was desirable to obtain uniform labelling of all calcium present in the skeleton of the mouse, CaCl, was added to all water administered to mice for weeks before and after gestation. Such water was also given to new-born mice after weaning until adult age was reached. The members of the litter, having almost the same radiocalcium content, were then sacrificed at different dates within 560 days. From the labelled calcium atoms present in the skeleton of the outgrown mice, 67.2 +. 7.9 per cent were found still to be present in the skeleton of sister mice sacrificed after the lapse of 390 days. When administration of Ca was interrupted after the birth of the litter, and its members reared by inactive mothers were sacrificed at different dates within 560 days, a mouse killed shortly after birth contained 8 per cent of the maternal 45Ca atoms, another mouse killed after 510 days contained 4 per cent. Half of the calcium atoms present at birth is thus conserved during the lifetime of the mouse. From the figures obtained of the passage of labelled calcium from one genera- tion to the next, it follows that the eleventh generation does not contain a single calcium atom present in the first generation of its ancestors. }. 2, Bie ele 4. 5. ADVENTURES IN RADIOISOTOPE RESEARCH References O. Cuievitz and G. Hevesy, Nature 136, 754 (1935). . O. CutEvitz and G. Hevesy Dan. Biol. Medd. 13, No. 9 (1937). M. Manty and W. F. Bate, J. Biol. Chem. 129, 125 (1939). R. 8S. Manty, H. C. Hopes, end M. Manty, J. Biol. Chem. 134, 293 (1940). F. Paneru, Z. Elektrochem. 28, 113 (1922). G. Hevesy, Phys. Z. 16, 59 (1915). 3. G. Hevesy and M. Birtz, Z. phys. Chem. B 3, 270 (1929). . O. ERBACHER, Z. phys. Chem. A 166, 23 (1933). G. Hevesy, Les Prix Nobel, Stockholm 19401944, p. 95. C. P. Lesitonp, G. W. W. Wiuxinson, L. F. BELANGER and J. Rosison, Amer. J. Anat. 86, 289 (1950). C. L. Comar, W. E. Lotz and G. A. Boyp, Amer. J.Anat. 90, 113 (1951). . H. E. Sxreper, C. Noxtan and L. Smpson, J. Biol. Cem. 189, 159 (1951). . B. Kipman, B. Rayner, M. L. Turr and J. M. Vaucuam, J. Pathol. and Bact. 64, 453 (1952). 3. B. Encretpt, A. Enestrom and R. Zerrerstr6mM, Biochim. et Biophys. Acta 8, 375 (1952). . R. Amprino and A. Enestr6m, Acta Anat. 15, 1 (1952). . G. C. H. Bausr, Acta Orthop. Scand. 23, 169 (1954). F. Panetu, Radio-Elements as Indicators p. 55. McGraw-Hill, New York (1928) . W. D. Armstrone, J. A. JoHNsoN, L. Singer, R. I. Lienxe and M. L. PREMER, Amer. J. Physiol. 171, 641 (1952). . M. FaLtKEenHEIM, W. F. Neuman and H. H. Carpenter, J. Biol. Chem. 169, 713 (1947). M. FaLKENHEIM, F. E. UNDERWOOD and H. C. Honae, J. Biol. Chem. 188, 805 (1951). . S. B. Henpricxs and W. D. Hir1, Metabolic Interrelations p. 141. Josuah Macy Jr. Foundation (1951). . M. Turt, B. Kipman, Brayner and J. Vaucuan, Brit. J. Exp. Pathol. 33, 207 (1952). . H. G. Hones, E. Gavetrt and J. THomas, J. Biol. Chem. 163, 1 (1946). 2. B. HAstines, Metabolic Interrelations p. 40. Josuah Macy Jr. Foundation (1951) a. M. A. Logan and P. O’Connor, J. Biol. Chem. 127, 711 (1939). 3. A. Cartson, Acta Physiol. Scand. 31, 30 (1954). . G. Hevesy, [Isotopic Indicators p. 421. Interscience Publishers, New York (1948). 5. L. A. Hann, G. Cu. Hevesy and E. C. Lunpscaarp, Biochem. J. 31, 1705 (1937). ». M.J.L. Dots, B.C. P. Jansen, G. J.S1zoo and G. J. vaAN DER Maas, Koninkl. Nederl. Akad. Vetenschap. Proc. 42, No. 6, 1 (1939). . H. E. Harrison and H. O. Harrison, J. Biol. Chem. 185, 857 (1950). . A. Carutson, Acta Pharmacol. 7, Suppl. 1 (1951). . W. Minper and T. Gorponorr, EHaperientia 8, 71 (1952). . O. ZETTERSTROM and M. LuuNGGREN, Biochim. et Biophys. Acta 8, 283 (1952). . G. Hevesy, H. Levi and O. H. Resse, Biochem. J. 34, 532 (1940). . E. L. Srvmons, L. O. Jacosson, E. K. Marxs and E. Lorenz, Radiology 52, 371 (1949). 3. L. B. Russett and W. L. Russet, J. Cel?. Comp. Physiol. 43, Suppl. 1, 103 (1954). CONSERVATION OF SKELETAL CALCIUM ATOMS THROUGH LIFE 233 34. L. Sincer, W. D. ArmstronG and M. L. Premer, Proc. Soc. Exp. Biol. Med. 80, 643 (1952). 35. G. E. Boxer and D. Sterter, Jr., J. Biol. Chem. 155, 237 (1944). 35a. A. Caruson, Acta Physiol. Scand. 26, 200 (1952). 36. R. C. THompson, J. Biol. Chem. 200, 731 (1953). 36a. G. C. H. Bauer, Acta Physiol. Scand. 31, 334 (1954). 37. G. Hevesy, The Svedberg Volume p. 456. Stockholm (1944). 38. I. S. Eprtman, A. H. James, H. BapEn and F. D. Moors, J. Clin. Inv. 23. 122 (1954). 39. H. BapEn and F: D. Moors, J. Clin. Inv. 23, 122 (1954). 40. L. Srncer and W. L. Armstronec, Proc. Soc. Exp. Biol. Med. 76, 229 (1951). 41. D. L. Bucuanan and A. Naxao, J. Biol. Chem. 198, 245 (1952). 42. H. H. Donatpson, The Rat p. 314. Philadelphia (1915). 43. M. Burau, H. Spencer, J. Swernov and D. LAszu6, Science 120, 1029 (1954). 44. G. Hevesy, L. Haun and O. Resse, Dan. Biol. Medd. 14, No. 3 (1939). 45. K. KserutF-JENSEN, Acta Physiol. Scand. 3, No. 1 (1941). 46. J.C. Aus, R. O. Evans, L. H. Hempetman and H. S. Martianp, Medicina Se 22 (1952). Originally published in the Scientific Monthly 83, No. 5 (1956) 25. PATH OF ATOMS THROUGH GENERATIONS G. HEVESY From the Institut for Research in Organic Chemistry, Stockholm THE number of atoms inherited from the mother whichis passed on to the next generation depends on three factors: (i) the fraction of the total number of atoms of the mother which were passed on to the newly born; (ii) the number of inherited maternal atoms that are replaced before birth of the grandchild by atoms from nutrients; and (iii) the number of maternal atoms still present at pregnancy which can be used for development of the embryo of the next generation. The quantity mentioned first depends essentially on the weight ratio between the mother and the newly born. Sodium, chlorine, and sulfur are preferred by the body of the newly born, whereas potassium, calcium, magnesium, and phosphorus are preferred by the body of the mother. The greatest discrepancy from an equilibrium distribution is seen in calcium. The concentration of calcium is about 3 times larger in the body of the mother than in the body of the newly born. The reason for this is the calcium deficiency of the developing skeleton of the newly born. The loss of inherited atoms before reproduction of the animal and the stability of the still-present, inherited atoms for the development of the embryo of the next generation vary decidedly from element to element. It has long been recognized that food supplied to the body is not only used to supply energy but is also of great importance in the replace- ment of used parts of the body. However, only in the last decades has interest begun to develop in the quantitative side of the second function of the nutrients and in determining the lifetime of molecules and atoms in the bodies of animals and plants. REPLACEMENT OF SODIUM ATOMS OF THE BODY As the first example for the replacement of atoms present in the body by atoms absorbed from food, we shall consider the fate of the sodium atoms in the organism. If radioactively labelled sodium—for example, in sodium chloride—is supplied to the body, the supplied sodium ions PATH OF ATOMS THROUGH GENERATIONS I25 mix rapidly with the sodium ions circulating in the body fluid. There- fore, the percentage of elimination of radioactive sodium is equal to the percentage of elimination of the total amount of circulating and easily exchangeable bound sodium in the body. Each day about 4 percent of the radioactive sodium present in the human body is eliminated (1—3). Therefore, the half-life of the sodium atoms circulating in the body is about 2 weeks. About 54 grams of sodium are present in the extracellular fluid of an 80-kilogram person. To this must be added an intracellular amount of about 33 grams. Of the latter, about four-fifths is contained in the mineral constituents of the skele- ton (4). The intracellular sodium and part of the mineral skeleton sodium enters comparatively fast into exchange with the circulating extracellu- lar sodium. Two-thirds of the mineral skeleton sodium, about 18 grams, is anchored so strongly that it is retained to a large extent throughout life (4—7). Of the sodium that is not strongly anchored—a total of 70 grams (2x 1074 sodium atoms)—not a single atom is present in the body after 162 weeks. These atoms are all replaced by sodium atoms from the food. All the chief ingredients of the body, with the exception of hydrogen, nitrogen, sulfur and iron, take part in the build-up of the apatite in the bone. The amount of conservation of a part of the individual atoms of an element in the body for a long or very long time depends foremost on the factor of how much of this elements enters into the bone apatite and is retained there temporarily or finaly. The evaluation of the degree and the speed with which atoms of the mineral bone skeleton are replaced by atoms of the blood fluid or the lymph is therefore of the greatest importance for our problem. EXCHANGE OF PHOSPHORUS ATOMS OF THE SKELETON STRUCTURE The question of whether and to what degree the atoms of the skeleton structure are exchanged with those of the blood and the lymph was raised only after radioactive phosphorus became available as an indi- cator for phosphorus atoms. Immediately after the discovery of artificial radioactivity by Frédéric and Irene Jontot-Curig£, radioactive phos- phate was produced. This was first used to answer the question of whether an exchange takes place between the phosphate ions of the bone and those of the blood (8). A few minutes after radioactive phosphate had been given to an adult rat it could be traced in the bone structure. The content of phosphorus-32 in the bone skeleton increased very rapidly at first; after 1 hour, however, the increase was much slower. The obvious interpretation of these observations was that a rapid exchange takes place between the marked phosphate ions of the fluid 236 ADVENTURES IN RADIOISOTOPE RESEARCH blood and the phosphate ions lying on the surface of the apatite crystals of which the bone structure is built. Hand in hand with this exchange goes a biological recrystallization of the skeleton. Crystals of the bone structure go into solution, and the new crystals, which have been formed from the labelled fluid blood, have to be radioactive. The biological recrystallization—that is, the renewal of the bone skeleton—governs ‘ Fic. 1. Phosphorus-32 absorption by the tibia of an adult rat 5 minutes after intravenous injection (left) and 120 minutes after intravenous injection (right). (3) [Photos courtesy of C. P. LeEBLonp this exchange process nearly completely after a short time. Autoradio- graphs taken by LeBLonp and _ co-workers (9) illustrate clearly that radioactive phosphate is absorbed by the epiphysial plate of the adult rat 5 minutes after injection (compare Fig. 1). After 2 hours, the absorp- tion is very distinct. It follows from the classical investigations of PANETH (10) that the ions in the topmost molecular layer fraction of a crystal powder enter into an exchange equilibrium with the ions of a surrounding solution. This statement holds only for a part of the ions which are located in the topmost molecular layer of a well-developed crystal surface—of a mine- ral, for example. Even if only a small of the topmost molecular layer of the bone apatite took part in the exchange proceedings, it would be sufficient for the removal of an important part of the phosphorus-32 from the plasma which was added to the blood fluid. Three percent of the bone phosphate, or maybe even more, settles in the topmost molecular layer of the bone apatite, and the phosphorus content of the latter is about 700 times greater than that of the blood plasma. If after a while PATH OF ATOMS THROUGH GENERATIONS 937 1 milligram of the phosphate of the bone skeleton shows the same radio- activity as 1 milligram of phosphate of the blood fluid, it could be concluded that the total phosphate content of the bone skeleton was renewed during the experiment. This, however, is not all the case. The experimental conditions for determining the part of the bone skeleton which is renewed are very unfavourable. The specific activity of the plasma phosphate falls at first very rapidly, later on very slowly, and shows fluctuations during the necessarily long duration of the experiment. In order to simplify the conditions for the experiment, the specific activity (activity per milligram) of the bound phosphate of a rabbit was held constant for 50 days by repeated daily injections. It followed that about one-third of the soft epiphysial bone skeleton was renewed during the duration of the experiment (11). From the hard diaphysial bone structure, a considerably smaller fraction is re- newed. The exchangeable part of the bone structure of the rabbit has therefore to be limited to about one-third. FATE OF CALCIUM ATOMS In order to be able to pursue the fate of calcium through several generations of mice, we had to know what fraction of the calcium ion in the bone skeleton of mice is renewed during a lifetime. Numerous experiments concerning the calcium metabolism of the skeleton have been described (12—16). In determining the fraction of the exchange- able skeleton calcium, we continued the afore-mentioned experiments, and we replaced the very tedious method by another. We did not investi- gate the replacement of the inactive bone calcium by labelled calcium in the fluid blood. We bred animals that were evenly marked with radioactive calcium, calcium-45. We determined then the fraction of radioactive calcium ions that left the animal during its lifetime and that were replaced by inactive calcium ions from the food. We supplied the mother with radioactive calcium of limited activity (less than 0.5 microcurie) so that no radiation damage to the animal had to be feared. In this way we obtained an evenly activated generation of mice, which cannot be obtained by any other method. Each member of a single generation had the same calcium-45 content within a few percent. The mice received radioactive calcium with their food till they were grown up. One of the siblings was killed at birth, and the radioactivity of its skeleton was determined. The remaining animals were killed at other times and examined. It became evident that after more than 1 year, which represents a considerable part of the life span of a mouse, 6.7 + 7.9 percent of the original calcium atoms of the skeleton could still be found. Therefore only one-third of the bone skeleton is renewed. 238 ADVENTURES IN RADIOISOTOPE RESEARCH In another series of experiments we examined newly born mice right after birth. The other mice were transferred to an inactive foster-mother and analyzed at different times in the following 1.5 years. Figure 2 shows the results of some of these experiments. One should expect that during the intense growth of the first week of life a large part of the atoms taken over from the mother are disposed of and replaced by those from 100 80 fo) Oo fs. (e) %Ca Content 20 100 200 300 400 500 600 doys Fic. 2. Loss of inherited calcium atoms during the life of a mouse. We determined the total activity of siblings whose mother was fed calcium-45. The offspring were killed at different times food. In the first 50 days, the loss of atoms from the mother is consi- derable. However, in the whole life of a mouse it does not amount to more than 50 percent. We found about 1/300 of the labelled calcium atoms that the mother of the second generation received at its own birth, in the newly born of the third generation. The loss in calcium atoms of the ances- tors before propagation and through limited transmission to the off- spring is repeated from generation to generation. It can therefore be found by extrapolation that, of the 610! calcium atoms to be found in a mouse weighing 30 grams, not one is present in the llth generation of its descendants. The loss of labelled calcium atoms in the transition from the second to the third generation of one strain of mice amounted to 1/200 only. Calcium atoms from the mother should no longer be traceable in the 12th generation of these animals. That the calcium atoms of the ancestors can be traced in such a long series of descendants can be ascribed to the fact that about 99 percent PATH OF ATOMS THROUGH GENERATIONS 239 of the body calcium is located in the skeleton. The skeleton is therefore able to preserve an important part of its building materials. However. only a fraction of this calcium is available for the structure of the body of the descendants. FATE OF WATER MOLECULES The calcium and phosphorus atoms of the ancestors of the mouse (similar conditions should be valid for human beings) are traceable in a long series of generations in the descendants. However, the water molecules that the mother transmits to the descendants disappear during the life of the first generation. Shortly after the discovery of heavy water, we were able to determine, thanks to the support of H. C. Urry, who discovered this isotope, the half-life of water molecules in the human body. We found that this amounts to about 10 days in the body (17) for normal water intake. The half-life depends on the amount of water consumed. SCHLOERB and his co-workers (18) also found recently a half-life of 10 days with a normal daily intake of 2.7 liters. In the case of a rather large daily water intake of 12.8 liters, the lifetime fell, however, to 2.5 days. In the case of a normal water intake, after 810 days not a single molecule of the 210°? water molecules originally present remains in the human body. The half-life of water molecules in rats was determined to be 2.5 to 3.5 days. In the mouse it should amount to about 2.5 days. After 340 days, therefore, not a single maternal water molecule is present in the mouse. Owing to the large part that the element calcium plays in the devel- opment of the skeleton, it is preserved best in the descendants. However, even of this element, not a single atom of the ancestors is present in the 11th and 12th generations. This evidence illustrates the independence of the hereditary pattern from an atomic share of the forefathers. It is well-known that the hereditary pattern depends on the ability of the organism to group in atoms, molecules, and higher cellular units in a certain manner. A protein composed of 20 different amino acids and having a molecular weight of 100,000 is able to appear in more than 10° isomers, as calculated by STAUDINGER (19). Therefore, incompar- ably more types of proteins can exist than the number of water molecules (10%) which are present in the oceans of the world. Since hereditary patterns are tied to the reproducibility of individual proteins or nucleo- proteins, there is room for new individual hereditary patterns as long as the number of human beings has not reached 10!?7° or even a larger figure. 240 WwW — MYQQPyPrHoOWOomn ADVENTURES IN RADIOISOTOPE RESEARCH References .G. Hevesy, Acta Pysiol. Scand. 3, 123 (1942). .G. Burcu, P. REAsSER and J. GRONWNICH, J. Lab. Clin. Med. 32, 1169 (1947). . W. Stri, Isotopic Tracers and Nuclear Radiations. McGraw-Hill, New York (1949). . H. Minter et ai., in J. E. Jounsron, Editor Medical and Physiological Appli- cations, Radioisotope Conference, 2nd, Oxford, 1954, Vol. 1. Academic Press. New York (1954). .G. C. H. Bauer, Acta Physiol. Scand. 31, 334 (1954). S. EpEtman et al., J. Clin. Invest. 23, 122 (1954). 5 ) . Bapen and F. D. Moors, Jbid. 23, 122 (1954). . Cutevirz and G. Hevesy, Nature 136, 754 (1935). ). P. LEBLonp et al., Amer. J. Anat. 86, 289 (1950 . Panetu, Z. Hlektrochem. 28, 113 (1922). . Hevesy, H. Levi, O. H. ReBBrE, Biochem. J. 34, 532 (1940). . SINGER and W. L. Armstrone, Proc. Soc. Exp. Biol. Med. 76, 229 (1951). . CarRLson, Acta Pharmacol. Toxicol. 7, Suppl. A (1951). . AMpRINO and A. Enestr6Om, Acta Anat. 15, 1 (1952). . CarRLtson, Acta Physiol. Scand. 31, 308 (1954). C. H. Bauer and A. Cartson, Ibid. 35, 67 (1955). . Hevesy and E. Horerr, Klin. Woch. Schr. 13, 1524 (1934). . R. ScHOERB et al., J. Chem. Inv. 29, 1926 (1950). . StauDINGER, Les Prix Nobel (1953), p. 115: Naturw. Rundschau 9, 8 (1956). Originally communicated in Acta Chemiqa Scand. 11, 261 (1957) 26. NOTE ON THE CHLORIDE CONTENT GF THE MINERAL CONSTITUENTS OF THE SKELETON G. HEVESY From the Institut for Research in Organic Chemistry, Stockholm IN view of the only slightly differing size of the hydroxyl and fluoride ion, the hydroxyl ions of the bone apatite can be replaced by fluoride ions. Furthermore some of the fluoride may be present in the bone mineral as calcium fluoride. The fluoride content of the bone apatite is determined by that of the plasma, which in turn depends on the fluoride content of the food. The fluoride content of the earth’s crust is much lower than that of the seawater, the mineral constituents of the skeleton of mammals living in the sea is, correspondingly, very much, about eleven times, larger than that of mammals living on land, which contain about 0.05% fluoride only. The skeleton of fish living in the Baltic, which has a low fluoride content, have a much lower fluoride percentage (0.06 %) than the ske- leton of fish living in the Atlantic (0.43%). Incorporation of fluoride into the mineral constituents of the bone was in recent years much investigated, mainly in connection with the observation that the presence of fluoride in the mineral constituents of teeth increases their resistance to caries. The radius of the chloride ion is much larger (1.81 A) than that of the fluoride ion (1.33 A) and calcium chloride being very soluble we can expect to find slight amounts of chloride in the mineral constituents of the bone. While the X-ray diagram of fluoroapatite is almost identical with that of hydroxy-apatite that is far from being the case for chloro- apatite”. The fluids circulating in the bone tissue having a high chloride content (about 300 mgm/100 ml); this has to be quantitatively removed prior to the determination of the amount of chloride incorporated into the mineral constituents. Such removal by chemical treatment of the bone encounters great difficulties. However, when labelling the skeleton all through with radiochloride and placing the animal, e.g. the mouse, on diet containing non-radioactive chloride for several months, all exchangeable radiochloride will be removed and excreted, the radio- chloride fixed in the mineral constituents alone remaining in the ske- leton. 16 Hevesy 242 ADVENTURES IN RADIOISOTOPE RESEARCH If the human half of the exchangeable chloride is removed in the course of 14 days® and replaced by chloride of the food and after 6 months, i.e. after 12 periods, all exchangeable chloride initially present will be practically absent. In the mouse, with its high metabolic rate, the removal rate of chloride can be expected to be still higher than in man. The removal can be accelerated by increased chloride feeding. Besides the chloride present in the standard biscuits fed we added 0.2% NaCl to the water the mice were drinking after they were put on non-radioactive diet. To the water administered to the mice in the first phase of the experiment °°C] of 0.67 « C activity per liter was added as sodium chloride weighing 9.3 mgm. The labelled chloride was adminis- tered to pregnant mice about 2 weeks prior to gestation. After gestation the administration of labelled chloride was continued for 4 months when the animals were fully grown. One member of each litter was then killed and its total 36Cl content and that of its skeleton determined. The remaining members of the litters were investigated 6 months later. The bone was ashed in the presence of sodium carbonate. The activity of 100 mgm, thus of an infinite thick layer of the samples obtained was determined, the counts registered being multiplied by the total weight of the ash-sodium carbonate mixture. The total ash of the first member of the litter of the mice investigated had a total activity of 2780 counts per min, mean value 2910 + 452 counts, that of the the mineral constituens of the skeleton of the first investigated offspring 56.6 counts. This was prior to biological removal of all exchangeable chloride from the skeleton but after the removal of some of the latter in the course of the isolation of the mineral con- stituents. By keeping the mice on an activity-free diet for 6 months the activity of the mineral constituents declined in the average to 21.2 counts (ef. Table 1). The total chloride content of a 35 gm mouse taken to be 48 mgm, the acti- vity of 1 mgm of the body chloride prior to removal of the active food was 60.6 counts per min. As the total skeleton after biological removal of all exchangeable chloride had an activity of 21.2 counts per min. the sequestered chloride content of the bone mineral amounted to 0.35 mgm or 0.73% of the total body chloride. The sequestered fraction of the bone calcium of the mouse, that non- replaceable by circulating calcium, was found®) to be 67.2 + 7.9%, The corresponding figure for bone sodium is stated to be 60—70®), 654, 60© and 69 by different authors. Thus about a similar percentage of excess sodium and of excess calcium is prevented from interchanging with their circulating atoms. The sequestration of bone constituents is }resumably due to the fact that a contact between these constituents and the circulating body fluids is obstructed. To arrive at the total CHLORIDE CONTENT OF THE MINERAL CONSTITUENTS OF THE SKELETON 243 excess chloride content of the bone of the mouse we correspondingly have to multiply the figure of 0.35 mgm found for the non-exchangeable bone chloride by about 1.5, thus arriving at the result that the total ex- cess bone chloride content of our 35 gm mice amounts to about 0.53 mgm. TaBLE 1. — Counts PER MIn. *6Cl ACTIVITY OF THE SKELETON OF 33—36 G MiIcE AFTER BEING KEPT ON NON-ACTIVE Drier, THUS AFTER BIOLOGICAL REMOVAL oF EXCHANGEABLE RADIOCHLORIDE, FOR 6 MONTHS 211.6 + 10: 21-2 — 2.208 m.eV.: = -—_ 0.208 In contrast to the bone sodium which is to a large extent present as excess sodium in the skeleton it can be shown that the chloride present as excess chloride makes out 1/10 only of the total bone chloride. From 234 m-equiv. sodium present in 1 kgm of dry human bone 84.9%, thus 46 gm, was found by Edelman et al. to be excess sodium in a 70 kgm man. For the dog Edelman and associates found 89.5% of the bone sodium to be excess sodium and a similar figure is stated by Miller and associates”. 1 kgm of rat bone was found to contain 125 m-equiv excess sodium®). As to the total chloride content of 1 kgm fat free bone this was found to amount to 19 m-equiv. only®, thus to less than the extracellular bone sodium which makes out 25 m-equiv. While excess bone sodium is to a marked extent responsible for the difference between total and extra- cellular body sodium, for chloride this difference is almost entirely due to the presence of intracellular chloride in the soft tissues. References G. Hevesy, Kgl. Danske Videnskab. Selskab Biol. Medd. 22, No. 9. (1955). 3.G. C. H. Bauer, Acta Physiol. Scand. 31, 334 (1954). 244 ADVENTURES IN RADIOISOTOPE RESEARCH 4. R. E. Davies, H. L. Kornpere and G. M. Witson, Biochim. et Biophys. Acta 9, 403 (1952). .J. S. Eprertman, A. H. James, H. BapEn and F. D. Moore, J. Clin. Invest. 33, 122 1(1954); 6. W. H. Berestrom, J. Clin. Invest. 34, 997 (1955). 7. H. Minuer, D. S. Munro, E. ReENscHLER and G. M. Witson, Radioisotope Conference, Oxford, Vol. I, p. 138. 8. W. H. Berestrom and W. M. Watuace, J. Clin. Invest. 33, 867 (1954). 9. R. Waxttacys and G. CHANDRON, Comt. rend. 230, 1867 (1950). Cu COMMENT ON PAPERS 24, 25, 26 THe availability of Ca at a later date much facilitated the investigations of processes taking place in the bone apatite. Experiments taking years could be carried out, which had not been possible before. Furthermore, in contrast to the body phosphorus, which is only partly found in the skeleton, 99 per cent of the body calcium of the mouse is concentrated in the latter. In paper 24 experiments are described in which Ca was administered to pregnant mice and also to the offsprings until they were fully grown. After that date, they were kept on a non-ra- dioactive diet. Mice were then killed at intervals, and the 4°Ca content of their skeleton determined. In the course of 2 years, which covers the largest part of the life span of the mouse, 33 per cent of the skeleton 4°Ca, and thus of the skeleton calcium, was found to be replaced by calcium atoms of the food or present formerly in the soft tissue. When #5Ca was administered to the pregnant mouse alone, the offspring were found to conserve 50 per cent of the maternal calcium atoms through life. When similar experiments were carried out with #P (paper 20) 60 per cent of the *P acquired by the mouse at birth was found to be lost when reaching maturity (paper 23). The calcium atoms have a fairly stable position in the skeleton so that it takes a great number of generations until the last calcium atom originating from the first ancestor is lost. In contrast with calcium atoms, the last ancestoral water molecule is lost already after a few generations, as described in paper 25. The above-mentioned renewal figures indicate replacement of skeleton calcium atoms by such taken up with the food or formerly located in soft tissues. Changes in the apatite crystals without participation of food calcium or soft tissue calcium would not be indicated by the methods described. In the course of a symposium which took place in the Ciba Foundation in 1951, the question was raised of how much chloride is to be found in the mineral consti - tuents of the skeleton. This question, which could not then be answered, induced an investigation the results of which are found stated in paper 26. The non-exchange- able chloride fraction was found to remain very much behind the non-exchange- able sodium fraction of the mineral constituents of the bone. Originally published in Scand. Arch. Physiol. 77, 148 (1937) 27. THE FORMATION OF PHOSPHATIDES IN THE BRAIN TISSUE OF ADULT ANIMALS L. Hauww and G. HEVESY From the Institute of Theoretical Physics, University of Copenhagen Ir 1s generally assumed that no regeneration of the brain tissue of adult animals takes place. To test the validity of this assumption we investi- gated whether any formation of phosphatides takes place in the brain tissue of adult animals. This problem cannot be attacked by ordinary chemical methods because these do not permit the making of a distinction between phosphatide molecules formed at different dates; this is, how- ever, possible if we introduce a labelled phosphate into the animal body (CHtEviTz and Hevesy, 1935, 1937), labelled sodium phosphate for example, and investigate whether the formation of labelled phosphatides can be established in the brain of the animal. We carried out experi- ments on rats, mice, and rabbits. Labelled phosphorus .can. be obtained by adding radioactive phos- phorus to ‘‘normal” phosphorus. If we dissolve for example 1000 rela- tive (radioactive) units of the radioactive phosphorus isotope #?P in a solution containing 1 mgm of phosphorus, say as sodium phosphate, and administer this solution. to an animal, then the presence of 1 relative radioactive unit in a part of the animal tissue will prove the presence of Yigg Of the total number of phosphorus atoms administered. The radioactive #2P used in our experiments was obtained by bombarding ‘arbon disulphide with fast neutrons from a mixture of radium sulphate and beryllium. Phosphatides are the second most abundant constituents of the brain tissue. The composition of the latter is shown in Table 1. TABLE 1. — COMPOSITION OF THE DRY BRAIN TIssSuE oF ADULT ALBINO Rats Prot CU ee ashe ee octane ieee stn) Gael eroeveucleice 48.5 Pinos pirat Ges ievererererertererersiey stele teiatieta) oer 26.5 Lipides not containing P .......... oe Esters and inorganic constituents ... 9.8 FORMATION OF PHOSPHATIDES IN THE BRAIN TISSUE OF ADULT ANIMALS 247 The total phosphorus present, amounting to 1.39%, is distributed between protein, phosphatides and acid soluble compounds in the follow- ing manner. Protein P 6.8% Phosphatide P 67.6% Acid soluble P 25.6% the largest amount of phosphorus being thus present as phosphatide. Let us assume that labelled sodium phosphate is introduced per os or by injection into the animal body and that after the lapse of some time some of the inorganic phosphorus present in the brain tissue is found to be labelled. Such an observation is not to be interpreted as a formation of new brain tissue because the inactive phosphate ions ean be replaced by labelled ones through a simple exchange process. The phosphate present in the lecithin molecule cannot, however, be replaced through a simple exchange process i.e. the labelled phos- phate can only enter the lecithin molecule during a synthesis of the latter. The presence of labelled lecithin molecules in the brain is there- fore a proof that a synthesis of lecithin has taken place after the intro- duction of the labelled sodium phosphate into the animal body. Though it was highly improbable that the phosphate present in the lecithin- molecule could be replaced by labelled phosphate through a physical exchange process, we tested this point by carrying out the following experiment. We shook 10 cc. of cats blood at 37° for 4 hours with 2.5 ee. of isotonic sodium chloride solution containing labelled phos- phorus. 10 ce. of blood contain about 1.2 mgm of phosphatide and 0.4 mgm of inorganic phosphorus, the latter being labelled by the activity added. The lecithin was then extracted. The extraction was made with ether + alcohol and the extract was shaken for several hours with calcium phosphate to remove any inorganic labelled phosphate which might be present in the extract. Of the 10,000 radioactive units only 3 units were found in the blood phosphatide extracted. Even this very slight exchange is probably due to en- zymatic actions occurring in the blood. From an experimental point of view blood seemed to be a very suitable liquid to carry out an exchange experiment and as the exchange observed was only a very slight one it did not seem of interest to pursue the subject further and carry out exchange experiments in liquids from which enzymes had been removed. In this connection we may, however, mention an experiment in which blood containing labelled phosphate was allowed to circulate through an isolated liver. Professor LuNpsGAARD being engaged on liver perfusion experi- ments kindly added a solution (3 cc.) containing labelled phosphorus of negligible weight to the blood used in his experiments. In 10 ce. 248 ADVENTURES IN RADIOISOTOPE RESEARCH of blood used for 4 hours, 7 of the 10,000 radioactive units added were found in the isolated lecithin phosphorus; the liver perfusion has thus a positive effect on the formation of labelled lecithin. We may also mention that in 1 cc. of the blood of a cat killed 11, hours after injecting a negligible weight of labelled phosphorus, we found a lecithin phosphorus activity amounting to 2% of the activity found in 1 ce. of plasma, while the acid soluble phosphorus present in 1 ce. of blood corpuscles contained, as Professor LunpsGAarRD found, an activity of 12.5% of that of the plasma, about 2/; of this being present in the phosphorus esters. A detailed study of the distribution of the labelled phosphorus between plasma and blood corpuscles is being carried out by Professor LunpsGaaRpD and one of the present writers. EXTRACTION OF PHOSPHATIDES The usual method of extracting phosphatides is by means of alcohol- ether mixtures. In this method of extraction a small part of the inorganic phosphorus present is dissolved as well; but in view of the preponderance of phosphatide phosphorus in the brain the error thus introduced can generally be disregarded. Under the peculiar condi- tions which prevail in the investigation of the formation of labelled phosphatides the error mentioned above can however become very embarassing. We introduce labelled inorganic phosphorus into the animal body per os or by subcutaneous injection. Now it is possible that only a very small amount of this is converted into labelled phos- phatide phosphorus so that even of a trace of the labelled inorganic phosphorus is extracted by the alcohol-ether mixture, our results can be seriously falsified. This is best seen from the following example: brain tissue is shaken in vitro with a solution of labelled inorganic phosphorus containing 10,000 relative radioactive units, and 0.1 mgm phosphorus; the solution is then removed and the dried tissue extracted with alcohol + ether. The presence of as little as 1074 mgm of in- organic phosphorus in the extract corresponds to 10 relative radioactive units and may be partly or wholly responsible for the activity of the extract. For this reason, although conditions in experiments in vivo are much more favourable than those in the example above, we chose a method of extraction more suitable for our special case than the alcohol- ether extraction. The procedure adopted by us was as follows. We dried the brain tissue with acetone and extracted the phosphatides by prolonged shaking with carefully dried ether; the extract obtained was evaporated to dryness and dissolved a second time in ether in the presence of a large excess of finely powdered dry sodium phosphate. We had found that FORMATION OF PHOSPHATIDES IN THE BRAIN TISSUE OF ADULT ANIMALS 249 by shaking a solution of labelled sodium phosphate with a large excess of finely powdered unlabelled sodium phosphate the former can be removed from the solution. A distribution of the phosphate ions between the liquid and solid phase takes place and the chance of a labelled phos- phate ion being in solution is entirely negligible on account of the over- whelming excess of the solid phase; this is especially so when the whole procedure is repeated. Another method of purification of the ether extract from the labelled inorganic phosphorus was as follows. Unlabelled sodium phosphate was dissolved in the extract and precipitated as ammonium magnesium phosphate. By repeating this procedure it was possible to get rid of the slightest trace of labelled inorganic phosphorus present in the ether extract. The activity of the ether extract had then to be measured. The amount of material being small (2.3 mgm) it was advisable to add a carrier. An inactive commercial lecithin preparation was used for this purpose, some of it being dissolved in the ethereal solution before evaporation. Before destroying the lecithin, calcium oxide was added to bring about the formation of calcium phosphate. The activity of the latter was measured by means of a Geiger—Miiller Counter. INVESTIGATION OF THE BRAIN OF RATS All our investigations were carried out on fully-grown adult rats. One animal was killed after a lapse of 5 days, a second one after a lapse of 3 days, and the third one after one hour. The results obtained are seen in Table 2. TABLE 2 | ; j ay f labelled P found Fresh weight} Dry weight Borer iiabe :00 labelles ca Rat killed after of brain | of brain | in mgm | inmgm | in the brain | im brain phosphatide = = We | es BO days: co.o8 1800 | 380 710-2 | 3.7-1072 3 ay Goo oo | 1440 300 ilOm= 2.41052 1/4 y osco0S 1430 290 6.8°107 2 0.42°10° 2 The above figures show clearly the formation of labelled phosphatide in the brain of adult animals. Though it could hardly be doubted that what we extracted and tested was actually phosphatide we obtained further evidence of this preparing from the brain tissue of rats to which labelled phosphorus had been administered, the highly characteristic chlorocadmium compound of lecithin and tested its activity. ° 250 ADVENTURES IN RADIOISOTOPE RESEARCH PREPARATION OF CHLOROCADMIUM LECITHIN The ether extract of the brain was evaporated to dryness, the residue was dissolved in alcohol, and a saturated solution of cadmium chloride in methyl aleohol added. The chlorocadmium lecithin which precipitates from the solution was further purified in the following manner: the precipitate was suspended in chloroform and a solution of ammonia in methyl alcohol added, whereupon the lecithin remained in solution while the cadmium is precipitated. The next step was to evaporate down the solution containing the lecithin, and to dissolve the latter in alcohol; the whole purification process was then repeated. We ascer- tained in our preliminary experiments that the above lenghty process can be carried out with a yield of about 50%. In an experiment with a rat’s brain extract which showed an activity of 2.4-1072 °, of the total amount given to the animal, we recovered, after the preparation of the chlorocadmium compound, lecithin containing 1.2-107? %, of the total activity administered. INVESTIGATION OF THE BRAIN OF MICE As has been mentioned already it is of great importance to investigate the brains of animals no longer growing. Through the kindness of Pro- fessor KrocuH and Dr. HagEpORN we obtained mice of unusually high age and investigated their brain. The animal was killed 21 days after injecting the labelled phosphorus. A result obtained was the following: TABLE 3 Percentage of labelled phos- | Weight of fresh | Weight of | phorus administered found brain dry brain | in brain lecithin | | in the brain 814 mgm | 86.2 mgm | 1.35°107 1 | Deda Oma INVESTIGATION OF THE BRAIN OF RABBITS The investigation of a brain of a rabbit killed 27 days (I.), resp. 4.5 hours (II.) after injection with radioactive phosphorus gave the follow- ing result. TABLE 4 | Percentage of labelled phos- Weight of Weight of | phorus administered found fresh brain dry brain }—__— a - -—— in the brain |} in brain lecithin I 8940 mgm | 1400 mgm | 6.9:°10°2 | 2.3:107? II | | 1.81073 6.3°1073 FORMATION OF PHOSPHATIDES IN THE BRAIN TISSUE OF ADULT ANIMALS 251 RELATIVE ABUNDANCE OF LABELLED PHOSPHORUS IN THE BRAIN COMPARED WITH THE AMOUNT PRESENT IN OTHER ORGANS The relative abundance of labelled phosphorus in different organs (activity per mgm phosphorus) is seen from Table 5. TABLE 5 | ; | Brain F 7 “ Total Brain | pre A Animal Killed after | Lecithin Bone (A) B (B) | REN Helio RCA ONERO NC ROR 5 days 2.12 | 2.34 0.9 1 (tibia) Ratner nny aicitia eee oats | Sundays: ay 8212S ee O70) ALO I a Piaiticketetel oo sce etstowee acters 1 hour Pale 0.40 5983 i MOUS retacks wis sisters cle etaee:s 21 days 0.94 | 1.08 0.9 1 | | } 1—3.5* FRAUD DIG lees oa loreisy cis aususes 27 days 6.10 | 6.10 1.0 * 1 = Tibia diaphysis, 2.6 = Jaw, 3.5 = Tibia epiphysis. As is seen from the above figures the ratio between labelled phosphorus in the whole brain and labelled phosphorus in brain lecithin shifts in favour of the latter with increasing time; furthermore that the percent- age replacement of phosphorus atoms in the brain by labelled phosphorus atoms is not very different from that found in the case of the phosphorus atoms of the bone. NoTE ADDED IN PROOF: Professors ArroM, PERRIER, SARSANA, SANTAGELLO and SEGRE most kindly sent us a manuscript of a paper in which the formation of lipidic phosphorus in different organs has been demonstrated by using radioactive phosphorus as indicator. They found the most marked metabolism in the liver, the intestinal mucosa and the kidney, the least in the brain and the muscles. WEIGHT OF THE NEWLY FORMED LECITHIN So far we have only calculated lecithin phosphorus formed as a per- centage fraction of the total activity given to the animal. In what follows, we will try to calculate the amount of newly formed labelled lecithin in grams. This is a more difficult problem. If we inject 1 mgm of phosphorus showing an activity of 1000 units we know that the presence of | unit of activity indicates 1/,,,, mgm of the phosphorus atoms injected. Now in the blood plasma of a rat we have for 17 ce. bo OU bo ADVENTURES IN RADIOISOTOPE RESEARCH of blood with an inorganic phosphorus content of 4.5 mgm per 100 ec., 0.75 mgm of phosphorus; if we inject radioactive phosphorus (1000 units) of negligible weight, 1 activity unit will indicate 0.75/1000 mgm of the phosphorus originally present in the plasma. In the blood plasma, however, numerous fast processes take place in which the phosphorus atoms are involved; this is shown by investigations carried out with labelled P. We get beside other reactions, a partition of the phosphate ions between the plasma and the phosphate in the skeleton, and in view of the very large preponderance of the latter a large part of the labelled phosphate ions will soon be present in the skeleton. Similar considerations apply to the muscles and other organs. The com- paratively large amount of phosphorus ester present in the blood cor- puscles will also take part in a dynamic interchange with the labelled phosphorus atoms present in the plasma. In the course of the last two years Prof. E. LuNpsGaarp and one of us have carried out and extended investigation of this point, to be published shortly. As a result of the processes just mentioned, only a few, let us say 10, units remain behind of the 1000 activity units introduced into the plasma, while the total inorganic phosphorus content of the plasma is unchanged and amounts as before to 0.75 mgm. The effect will be that 1 activity unit will now indicate as much as 0.75/10 mgm (i.e. much more) phosphorus. On account of the rapid dynamic happenings in the animal body our scale of indication will rapidly change and, the function representing this change being a very intricate one. One way to obtain some infor- mation is the following. We make an experiment of very short du- ration in which 1 activity unit does actually indicate 0.75/1000 mgm of labelled phosphorus; then we make a longer experiment and determine experimentally the activity of the inorganic phosphorus present in the plasma. If only 10 activity units are now present then 1 activity unit indicates 0.75/10 mgm P at the latter stage. The activity accumulated at the early stage will be correctly interpreted by making use of the first mentioned scale, that accumulated at the last stage by making use of the last mentioned scale. In some cases we have followed the labelled P content of the blood continuously. Let us now consider the rat killed after 1 hour, the blood plasma of the animal containing 0.75 mgm inorganic phosphorus, a 4.2-10~8th part i.e. 3.2-10°% mgm of this was converted into brain lecithin phosphorus. This being a very small part of the brain lecithin phosphorus, the reverse reaction can be disregarded and the amount of labelled brain lecithin phosphorus formed in 100 hours can be taken as a hundred times as much, i.e. 3.2-10~1+ mgm i.e. 8 mgm of the brain lecithin is newly formed in the course of about 4 days. The calculation of the weight of labelled phosphorus was carried out on the basis of an analysis of the activity of the blood and of the brain lecithin after the lapse of one hour. Now one part of FORMATION OF PHOSPHATIDES IN THE BRAIN TISSUE OF ADULT ANIMALS 253 the labelled lecithin was already formed after the lapse of a few minutes, mass of inorganic P at which time the ratio —— in the blood was smaller activity and 1 activity unit therefore indicated a smaller mass of phosphorus than after the lapse of one hour when the measurements were actually made. Hence the actual amount of labelled lecithin phosphorus will be less than that calculated above but is definitely higher than 0.08 mgm. PHOSPHORUS EXCHANGE IN BRAIN LECITHIN IN VITRO We shook a freshly removed rat brain for 5 hours at 37° with 3 ce. of an isotonic sodium chloride solution containing 0.09 mgm of phos- phorus labelled by the addition of radioactive phosphorus. The brain was carefully washed with cold acetone and dried at room tempera- ture and the lecithin extracted as described above. The ethereal solu- tion was shaken for several hours with sodium phosphate to remove inorganic labelled phosphorus and after this operation had been repeat- ed four times, the sodium phosphate was found to be entirely inac- tive. The solution showed an activity of 2200 units and we found 18 units in the lecithin extract; this corresponds to 0.00076 mgm of label- led phosphorus or about 4/599) part of the total lecithin phosphorus content of the brain. Thus the formation of a very small amount of labelled phosphorus also takes place in the freshly removed brain tissue in vitro, presumably under enzymatic action. We intend to follow up the exchange problem in vitro in greater detail. DISCUSSION OF THE RESULTS From the results above it clearly follows that brain lecithin is con- stantly being synthesized in the brain tissue of adult animals. Pre- sumably a part of the lecithin is constantly broken down under enzy- matic action and rebuilt again, thus making it possible for the labelled phosphorus atoms present in the blood to enter the lecithin molecule. BELFANTI, Conrarpr and Ercotrtr (1936) give the following scheme according to which lecithin is supposed to decompose under the action of lecithases. It is possible that a reaction takes place in both directions according to this or a similar scheme, so that the phosphorus atoms present in lecithin, are rendered exchangeable when enzymes are present although they undergo no exchange in the absence of these. The study of the mode and rate of action of the different lecithases may be much facili- tated by following up exchange process in lecithin and its decomposition products in the presence and absence of the different enzymes. 254 ADVENTURES IN RADIOISOTOPE RESEARCH Lecithin ~Q,,. ae 2 : "0, ar ae = le ie ‘: ape ZE _ bag, A > Ss Fatty acid ——— —— ——(Fatty acids Te Fee aa ester Lysochitin of Cholin | Cholinphosphatase Cholin -+ Glycerophos- phoric acid | Glycerophosphatase Glycerin -|} Phosphoric acid The figures in our experiments show clearly that the ratio of labelled lecithin phosphorus to labelled phosphorus other than that from lecithin in the brain increases with time. The amount of labelled lecithin produced within 1 hour in the brain of a rat can be estimated as lying between 0.08 and 0.0008 mgm. In starting this research we contemplated the possibility of the for- mation of labelled brain lecithin being influenced by nervous action. No effect of nervous actions on chemical processes in the brain has yet been ascertained; the dependence of the latter on the former must be described by a curve where a very large increase in the abscissa (nervous action) corresponds to a minimal change in the ordinate (chemical effect). To test this possible effect of nervous action it would be necessary to carry out a very large number of experiments. Summary By using labelled (radioactive) phosphorus as indicator, it was found that one hour after the subcutaneous injection of labelled sodium phosphate, labelled lecithin was already formed in the brain tissue of fully grown rats. Similar experi- ments were also carried out with fully grown mice and rabbits. The result proves that a constant breakdown and building up of lecithin takes place in the brain tissue presumably under enzymatic action. References O. CmiEvirz and G. Hevesy (1935) Nature 136, 754. O. Cutevirz and G. Hrvesy (1934) Kgl. Danske Vidensk. Selsk. Biol. Medd. XGHE 9): S. Betranti, A. Conrarpr and A. Ercori (1936) Ergebn. Enzymforsch. 5, 213. C. Artom, L. Perrier, M. SAanraGEeLLo, G.Sarzanaand E. Srert (1934) Nature 139, 836. Originally published in Nature, 140, 275 (1937) 28. LECITHINAEMIA FOLLOWING THE ADMINISTRATION OF FAT G. Hevesy and EK. LunDscaarp From the Institute of Theoretical Physics and the Physiological Institute, University of Copenhagen Axsourt two hours after the administration of a meal containing fat, the fat content of the blood begins to rise. Bloor! found that when olive oil is administered to a dog, besides an increase in the neutral fat content of the blood an increase in its lecithin content also takes place. The average increase was found to be about 20 per cent. A maximum is reached after four hours. Bloor was inclined to ascribe the lecithin formed after the administration and resorption of the neutral fat to a synthesis occurring inside the red blood corpuscles. Other explanations might, however, be suggested as well, namely: (1) The lecithin is synthesized in the intestinal mucose and resorbed into the blood. (2) The synthesis takes place, after the resorption of neutral fat, in the liver, or some- where else outside the intestinal tract. (3) The increase in the lecithin content of the blood is due to mobilization of preformed lecithin afte1 the resorption of the neutral fat. To decide which of these suggestions is to be accepted we repeated Bloor’s experiment, but administered simultaneously with the oil- labelled (radioactive) phosphorus in the form of sodium phosphate. In the case denoted by (1) the additional blood lecithin should contain chiefly labelled phosphorus; in case (2) the additional lecithin should contain only small amounts of labelled phosphorus; in case (3) the ad- ditional lecithin should contain ordinary phosphorus only. We determined the normal P present in the blood lecithin, which was extracted by the usual procedure, by the method of Fiske and Subbarow, and the labelled P by means of a Geiger counter. While, as seen in the table, the lecithin phosphorus content of 100 cc. of blood increased by 2 mgm four hours after administering the oil, that of labelled P only increased by 0.096 mgm. We must, furthermore, take into account the fact that half the labelled phosphorus administered two hours before the oil produced 0.028 mgm labelled lecithin P during that time. We must therefore deduct 2 x 0.028 mgm from the ’oil effect’ of 0.096 mgm, obtaining 0.04 mgm per 100 ce. of blood for the maximum value of the “oil effect’. 256 ADVENTURES IN RADIOISOTOPE RESEARCH An important objection can, however, be raised to our conclusion; it may be argued that the intestinal tract might contain large amounts of phosphorus other than the labelled phosphate administered by us, the presence of which must be accounted for when carrying out the above calculation. To investigate this point and to ascertain to what extent the labelled phosphorus was resorbed, we killed the dog after the last experiment, the results of which are seen in the table. We washed the intestinal tract with water and determined both its total P content and its labelled P content. We found by activity measurements 39.6 mgm labelled P and by chemical determination 175 mgm normal P. Within six hours as much as 259.4 mgm of the 300 mgm administered to the dog was thus resorbed. The 135 mgm. non-labelled phosphorus reached the intestine, presumably along with the digestive fluids, so that the 40 mgm labelled P were ‘diluted’ to 175 mgm. We determined also the total acid-soluble phosphorus content of the intestinal mucose; it was found to amount to about 40 mgm, bringing the above figures up to 215 mgm. But even if we make the assumption that this dilution was present during the whole of the resorption process we should get the result 5.2 x 0.064 = 0.21 mgm per cent lecithin P, while an in- crease of 2 mgm per cent was found in the blood lecithin P. Lecithin phosphorus found | Labelled total P found Time : P P | in 100 ec. blood |———— = a in Labelled P given in mgm. | ia ee aa oa as in the total blood hours in 100 ce. blood hee Total Labelled el garnets of the dog 0) 150 — — a = 2 150 (++ 50 gm oil) 16.0mgm | 0.028 mgm 1.03 mgm 6.18 mgm 4 oad 15s on) OLO48n ters ma 12.18 6 9 2=="(259.4.mem. resorbed) | 18.0 Be 0.096 a POO) is | 12.00 o It is of interest to compare the labelled P content resorbed with that actually found in the blood stream of the dog. Six hours after the beginn- ing of the experiment, as is seen in the table, only 4.6 per cent of the amount resorbed was found. This result illustrates beautifully the great rapidity of the phosphorus exchange in the body. As observed by us in numerous cases, the individual phosphorus atoms present in the blood stream exchange their places rapidly with others present in the different organs. For this reason we can conclude with certainty that during our experiments the ratio labelled phosphorus to ordinary phosphorus must have been appreciably higher in the intestinal mucose than in the blood. The only moderate increase in labelled phosphorus in the blood leci- thin after administration of oil, an increase which nevertheless in all our experiments exceeds the increase observed after the radioactive LECITHINAEMIA FOLLOWING THE ADMINISTRATION OF FA‘ 257 phosphorus was administered alone, leads to the conclusion that during the absorption of neutral fat, lecithin is formed outside the intestinal tract. A comparatively rapid formation of labelled lecithin in several organs in the course of normal metabolism has in fact recently been observed?. References 1. W. R. Buoor, J. Biol. Chem. 23, 314 (1915), 24, 448 (1916). 2.C. Arrom, G. Sarzana, M. Sanracetro and E. Secret, Nature 139, 836 (1934). Jomp. also: L. Haun and G. Hevesy, Scand. Archiv. f. Phys. (Aug. 1937). 17 Hevesy Originally published in Biochem. J. 32, (1938) 29. FORMATION OF PHOSPHATIDES IN LIVER PERFUSION EXPERIMENTS L. A. Haun and G. CH. HEvessy From the Institute of Theoretical Physics, University of Copenhagen Asour 2 hr after a meal containing fat, the fat content of the blood begins to rise. This alimentary lipaemia is followed by lecithinaemia | RercHER, 1911; Biroor, 1915], the phosphatide content of the blood increasing more or less parallel with the fat content. A maximum is reached after about 4 hr and after 8 hr the fat and phosphatide contents of the blood are almost at the initial level. As to the origin of the phos- phatides responsible for alimentary lipaemia the following possibilities exist: (1) the phosphatides are synthesized in the intestinal mucosa and resorbed into the blood; (2) they are synthesized in the blood; (3) they are mobilized under the influx of lipaemic blood from the liver or other organs and possibly wholly or partly formed in the former during the influx. To obtain further information on the above problem, oil together with labelled (radioactive) sodium phosphate, were administered to a dog [Hevesy and Lunpsca arp, 1937]. If the increase in the phosphatide con- tent of the blood which amounted to 15 % after 4 hr. was due to phos- phatides taken up from the intestine, the phosphatides extracted from blood should have shown a marked radioactivity. The latter was, how- ever, much smaller than to be expected on this assumption. It follows that while phosphatides are synthesized in the intestinal mucosa [ARTOM et al., 1937; SINCLATR and SmiruH, 1937] and some do enter the circulation from the bowels [HimmMeEricu, 1934; StULLMANN and WILBRANDT, 1934; Freeman and Joy, 1935] the bulk of the phosphatides which are responsible for the alimentary lipaemia must originate from outside the intestinal tract. We next tested the possibility that the additional phosphatides are formed in the lipaemic blood [HAHN and Hevssy, 1938]. A few ml. of dog blood were shaken with labelled sodium phos- phate under the usual precautions for 4.5 hr. The phosphatides ex- tracted after the experiment were only slightly radioactive, the labelled FORMATION OF PHOSPHATIDES IN LIVER PERFUSION EXPERIMENTS 259 phosphatides formed amounting to only about 0.1% of the total amount present. No difference was found in the behaviour of normal and lipaemic bloods. FORMATION OF LABELLED PHOSPHATIDES IN PERFUSION EXPERIMENTS Through the great kindness of Prof. LuNpsGaarp and Dr Buiixen- CRONE, who carried out perfusion experiments on isolated livers, we were enabled to test the formation of labelled phosphatides in the blood of cats circulating through an isolated liver and also in the liver tissue. To 120—160 ml. of cat blood diluted to about twice its volume with physiological NaCl solution a minute amount of active sodium phos- phate was added. The blood was then defibrinated and allowed to circu- late through an isolated liver for 2.5 hr. The labelled inorganic P present in the blood was determined at the start and at the end of the experiment and also the labelled phosphatide P of blood and liver at the end of the experiment. The ratio of the specific activity (activity per mgm P) of the blood phosphatide P to that of the blood inorganic P is seen from Table 1. In interpreting the figures of the table we should recall that if all phosphatides molecules present are newly formed the specific activities of the inorganic P and phosphatide P should be equal. The TABLE 1 Specific activity of blood phosphatide P Specific activity of blood inorganic P y a (average value) | | | | 12x 10-8 | | | Normal blood (av. of 3 exps.) | OS xel0m: Oy << Ome | 16.6 10m? S452 10m) 22D Lipaemic blood (av. of 2 exps.) el Ome figure of 0.851073 for the ratio quoted, for example, shows that the new formation of phosphatide molecules within the experiment amounts to only 0.085%. In the course of the experiment inactive inorganic P of the liver and also a part of the P present in the organic P compounds of the liver exchange with the active plasma inorganic P and lower the specific activity of the latter. The specific activities of the plasma inor- ganic P being thus different at the start and at the end of the experiment we have calculated the ratio (seen in Table 1) for the beginning of the experiment (col. 1), for the end (col. 2), and also an average value (col. 3). The figures of Table 1 for normal blood hardly differ from the figures obtained in the experiments in vitro (average value 0.8 107%). While, ‘ 260 ADVENTURES IN RADIOISOTOPE RESEARCH however, in the experiments in vitro no definite difference was found in the formation of phosphatides in normal and lipaemic bloods, in the perfusion experiment about three times as many newly formed phospha- tide molecules were found to be present in the lipaemic blood as in the normal. This result is supported by figures obtained when investigating the labelled phosphatides extracted from the livers used in the perfusion experiments. Here also (see Table 2) a greater part of the phosphatide present became labelled when lipaemic blood was used. TABLE 2 Pe ) _ | Spec. activity of Phosphatide P | , | phosphatide P . | Total P per gm , Liver per gm fresh tissue) ri E suk x 100 fresh tissue mgm | aa mgm | | Spec. activity of | | inorganic P Perfusion with normal blood | a | 0.96. | 2.6 1.25 Pe) Rs ke} 2.9 1.53 3 0.88 | 2.0 1.78 Perfusion with lipaemic | blood Al 0.75 | 2.4 2.75 5 | UE a 2.2 | 2.59 6 | | 335.0 2019 1.32 The livers used were taken from fasting cats, except in Exp. 3. In Oxp. 6 the specific activities of the ester P of the liver and the protein P (remaining P after extraction with ether-alcohol and trichloroacetic acid) were determined as well. The relative figures obtained were; speci- fic activity of the inorganic P, 1; of the ester P, 0.218; of the protein P, 0.068. The radioactive P atoms can only enter into the phosphatide molecules by a synthetic process. If the radioactivity of 1 mgm organic P of the liver were equal to that of 1 mgm inorganic P, all organic P atoms would have been replaced. If the organic P were not radioactive at all, none of the organic molecules could be newly formed. If all the phospha- tide molecules present in the liver after the perfusion experiment had been newly formed the value of the ratio in the last column would be 1. From the figures it follows that 1.5% of all phosphatide molecules pre- sent in the experiment with normal blood and 2.7% in that with lipaemic blood are formed in the course of the experiment. During the perfusion experiment some molecules may have decom- posed thus introducing an uncertainty into all conclusions based on the determination of the amount of phosphatides present. Our conclusions FORMATION OF PHOSPHATIDES IN LIVER PERFUSION EXPERIMENTS 261 are not influenced, however, by this source of error, since they are not based on determinations of the phosphatide content before and after perfusion but on the ratio of labelled and non-labelled phosphatide molecules present in the liver. We may, therefore, conclude from the results described above that lipaemic blood is more effective in the formation of phosphatides in the liver than is normal blood. This result suggests that one of the main reasons for alimentary lecithinaemia is that during the influx of lipaemic blood phosphatide formation in the liver is increased and phosphatides are discharged into the circulation |cf. AYLWARD et al., 1935]. As the lipaemic blood is changed into normal blood the excess phosphatides are taken up by the liver and other organs, until the ‘‘normal” phosphatide content of the blood is reached. Summary In experiments in vitro the amount of labelled phosphatide obtained in shak- ing blood with radioactive sodium phosphate is the same whether normal or lipaemic blood is used. In perfusion experiments the lipaemic blood is found to contain more labelled and thus newly formed phosphatide than the normal blood. The same result apples also to the phosphatides extracted from the liver in the perfusion experi- ment. References ArTOM, SARZANA, SANTAGELLO and SEGRE (1937) Nature 139, 836. Ayt Warp, CHANNON and WILKINSON (1935) Biochem. J. 29, 172. Broor (1915) J. biol. Chem. 23, 317. FREEMAN and Joy (1935) J. biol. Chem. 114, 132. Hanw and Hevesy (1938) Mem. Carlsberg Lab. 22, 188. Hevesy and Lunpscaarp (1937) Nature 140, 275. HimmericH (1934) Amer. J. Physiol. 116, 342. ReEIcHER (1911) Verh. Kongr. inn. Med. 28, 327. Srncuatr and SmirH (1937) J. biol. Chem. 121, 361. SttrMann and WiLBRanptT (1934) Biochem. Z. 270, 52. Originally communicated in Kgl. Danske Videnskabernes Selskab. Biologiske Meddelelser, 15, (1940) 30. RATE OF PENETRATION OF PHOSPHATIDES THROUGH THE CAPILLARY WALL G. Hevesy and L. Haun From the Institute of Theoretical Physics, University of Copenhagen Tons or molecules of crystalline substances present in the plasma can easily penetrate through the capillary wall. As soon as a few minutes after injecting labelled sodium ions (4Na*) into the jugularis, we find these ions proportionally distributed between the sodium (3Na*) ions of the plasma and those of the interspaces. On the other hand, colloidal particles like those formed by the proteins of the plasma under physiolo- gical conditions pass through the walls of the capillaries at very slow rate only. The phosphatides present in the plasma can be expected to have an intermediary position as to their penetrability through the capillary wall between the crystalline constituents and the proteins present in the plasma. To determine the rate of penetration of the plasma phosphatides through the capillary wall, we introduced labelled phos- phatides (phosphatides containing radioactive P) into the plasma and measured the rate of their disappearance from the circulation. The labelled phosphatides were obtained in the following way. Label- jed sodium phosphate was administered to a rabbit (A). The phosphati- des formed, after the start of the experiment, in the liver and other argans of this rabbit become labelled; a part of these labelled phospha- tides is liberated into the plasma. By injecting plasma of this rabbit (A) into the circulation of another rabbit (B), we introduced labelled plasma phosphatides under strictly physiological conditions into the circulation. To avoid the increase of the plasma volume of rabbit B, we removed, previous to the injection of the labelled plasma, for example, 20 ec. blood of rabbit B. This blood was, after addition of heparin, gently centrifuged to separate the bulk of its plasma content which was then replaced by the labelled plasma of rabbit A. The blood thus obtained was injected into the jugularis of rabbit B. This rabbit, thus, gets its own corpuscles reincorporated, combined with the corresponding amount of labelled plasma of the other rabbit. An aliquot part of the plasma of rabbit A is kept to be analysed. The labelled phosphatide molecules introduced into the circulation of rabbit B become distributed in the total plasma of the rabbit almost RATE OF PENETRATION OF PHOSPHATIDES THROUGH THE CAPILLARY WALL 263 at once, the next step being the continuous escape of the labelled phos- phatide molecules through the capillary wall and their replacement by other phosphatide molecules, originally located in the organs, which diffuse in the opposite direction, namely through the capillary wall, into the plasma. Since the phosphatide content of the plasma remains practically constant during the experiment, the exodus of a certain quantity of phosphatides must be followed by the influx of about the same amount. In view of the very minute turnover of phosphatides O = first experiment %#= second experiment sma re) re) Percent of labelled phosphatides present in total pla 20 60 100 140 180 220 250 min. Fic. 1. Rate of disappearance of labelled phosphatide molecules from the plasma. in the blood, the number of labelled phosphatide molecules which are decomposed in the plasma during the experiment can be neglected. The processes described above are going on under strictly physiological conditions. The replacement of ordinary phosphorus (3!P) by radioactive phosphorus (#2P) in some of the phosphatide molecules can certainly not be considered to entail the introduction of a non-physiological com- ponent into the circulation, as such a replacement cannot influence the chemical behaviour of the phosphatide molecules to any significant extent. The rate at which the labelled phosphatides escape from the plasma of rabbits is seen in Tables 1 and 2, and also in Fig. 1. The figures of the tables were obtained by comparing the radioactivity of the phospha- tides present in 1 cc. plasma samples of rabbit B, taken at different intervals, with that of the phosphatides of an equal plasma volume 264 ADVENTURES IN RADIOISOTOPE RESEARCH of rabbit A. The phosphatides were extracted by making use of BLooR’s method. After being converted into phosphate by wet ashing, an aliquot part of the solution obtained was used in the colorimetric measurement of the P content, another to secure an ammonium magnesium phosphate precipitate, the activity of which was determined by a Geiger counter. The calculation of the amount of labelled phosphatides present in the total plasma of the rabbit from that found in 1 ce. necessitates the knowledge of the total plasma volume. This was calculated from the blood volume and the known haematocrit value. The blood volume was determined by making use of a method recently described (HAHN and Hevesy, 1940). This method is based on the measurement of the dilution of a known volume of corpuscles containing labelled organic P compounds in the circulation of the animal, the blood volume of which is to be determined. In experiments on rabbits, the injection of foreign plasma was preceded by the removal of a corresponding volume of blood, as described above. In experiments on chicks, however, no blood was removed beforehand. EXPERIMENTS ON RABBITS (@) Some of the results obtained were previously published by us in a note to Nature 144, 204 (1939).— F. E. Haven and W.F. Bate [J. Biol. Chem. 129, 23 (1939)] injected emulsions containing labelled phosphatides prepared from the liver of the rat into the circulation of another rat and found the labelled phospha- tides to accumulate mainly in the liver and the spleen. As seen in Tables 1 and 2 and also in Fig. 1, half of the labelled phosphatides introduced into the plasma leave the circulation by penetrating through the capil- lary wall in the course of about an hour. As the non-labelled phosphatides can be expected to show the same behaviour as the labelled ones, we can conclude that, from all phosphatide molecules present at the start of the experiment in the plasma, half will no longer be present after the lapse of about an hour, and will be replaced by others which were previously located in the organs. Taste 1. — Rare or Escape oF LABELLED PHOSPHATIDES THROUGH THE CAPILLARY Watt or A Rassit WEIGHING 2.4 kgm First experiment | | Per cent of labelled phos- : phatides injected into the Time ie ; ; jugular vein, present in the total plasma (Oo ohbaleaeoigs Sone orci 100 SOGM serrun ds Mofe-a 10 : e = ie — liens Phosphorus in Plasma 'e () O) O 10 O 10 O 60 120 150 180 210 Age of the chicken in days Fig. 1. Phosphorus distribution in chicken blood. were obtained by analysing the blood of a large number of white Leghorn chickens. The analyses were repeated once a month or oftener beginning at the time when the chickens were 1 month old and continuing through the periods of growth, egg production, and subsequent molting. The results present very instructing data, they show that the phosphatide phosphorus alone, especially that of the plasma, changes very markedly with the age of the chicken, a rapid rise in the latter taking place after the lapse of 5 months at the time of production, this high level being held under the entire production period with some fluctuations and dropping quickly as production ceases and molting season approaches. 18* 276 ADVENTURES IN RADIOISOTOPE RESEARCH We determined the blood phosphorus of the laying hen denoted as I, the result being seen from Table 3. The blood phosphorus of another hen is discussed on p. 274. TABLE 3. — P-coNTENT oF HENS BLOOD mgm % mgm % mgm % in plasma [ieorpuecles| in blood iphosphatide IP... 24 .). ..-) 20.0 22.6 | 20.7 limonene IP Saas pacces ac 5.4 | -- | — Total acid soluble P..... | -- | Seale 21.3 Rest (Protein) P ........ 94) 93108.) 168 In the blood of non-laying hens® after 24 hours fasting an average phosphatide P content of 16.8 mgm% was found the total plasma P amounting to 13 mgm%, the plasma inorganic P to 4.6 mgm%. EXPERIMENTAL METHODS The yolk was dried by adding ice cold aceton, the dry yolk was carefully pulverised and the powder obtained shaken for 15 min with 150 cc. ether, the last mentioned procedure being repeated four times using fresh ether. The ether was than carefully evaporated, the residue taken up with dry ether, the latter removed by evaporation, this time in a Kjeldahl flask, and the residue ashed. The phosphatides of plasma, corpuscles and total blood were extracted by an ether-alcohol mixture after Bloor. The extract was several times carefully evapo- rated to dryness and taken up with ether or petrol ether. The residue of the first extraction was treated with trichloracetic acid (10 cc. of 10% solution for each cc. of blood) and from the filtrate obtained the inorganic P precipitated as ammo- nium magnesium phosphate; the esters present in the filtrate were hydrolised and the phosphate produced by the hydrolysis of the esters precipitated as ammo- nium magnesium phosphate. Though the extraction and the neutralisation of the acid solution were both carried out at —9°, some of the inorganic phosphorus present may be due to decomposition of the esters and we therefore gave in the table only the total acid soluble phosphorus present in the corpuscles which includes both the inorganic and the ester phosphorus. The liver was minced, dried in vacuo at room temperature, pulverised, dried again in vacuo and extracted with ether-alecohol (1:3), the latter being left to boil for 15 sec. In one case we extracted with ether alone to com- pare the active P content of the ether soluble phosphatides such as lecithin with that of the total phosphatides. The acid soluble P was extracted from the dried liver powder by treatment with cold (—10° to —15°) solution of trichloracetic acid, first with a 10/9 solution for 10 min and than twice with a 5°/, solution each for 5 min. The inorganic and organic constituents of the acid soluble phosphorus were separated as stated above. The P con- tent and activity of the residue obtained after extraction of the phosphatides and the acid soluble P was also investigated. 1H. M. Dyer and [. H. Ros, J. Nutrit. 7, 623 (1934). ORIGIN OF PHOSPHORUS COMPOUNDS IN HENS’ EGGS Pe =I ~I We determined the phosphorus content of a known fraction of the inorganic phosphate solution obtained in the above described procedures by the colorimetric method of Fiske and Suparow. The phosphate content of another fraction of the phosphate solution was precipitated in the form of ammonium magnesium phos- phate and its activity determined by making use of a Geiger tube counter. Let us say we have administered a hen a labelled phosphate solution containing | mgm P and showing an activity of 10° counts per minute. We want to know what per- centage of this labelled phosphorus will be found in the yolk lecithin. To arrive at this figure we take from our solution containing the labelled phosphorus as much as corresponds to !/; 999 Of the amount administered to the hen and _preci- pitate the phosphate, denoting the precipitate obtained as our standard preparation, while we will call the precipitate obtained from the yolk lecithin as lecithin prepa- ration. Before precipitating both the standard and the lecithin preparation we add to the solution a known amount, usually about 80 mgm, of inactive sodium phosphate, by so doing we diminish the amount of labelled phosphate possibly remaining in solution after precipitating with the magnesium citrate reagents and furthermore we obtain a standard and a lecithin preparation of equal weight. The /-rays emitted by the active phosphorus being to an equal extent absorbed in the two preparations the activity of which is to be compared, their weight and thus the thickness of the layers investigated being the same, there is no need to pay attention to the absorbtion of the /-rays in the samples investigated. Nor need the decay of the radioactive P be considered, as both the preparations to be compared, the lecithin and the standard preparation, decay at the same rate. The yolk residue obtained after removal of the lecithin was treated in similar way and also the white of the egg, while the shell was ignited and dissolved after ignition in hydrochloric acid, the solution being treated in the way described above. The samples were placed in small aluminium dishes having a surface area of 1.1 em? and were placed immediately below the aluminium window of the Geiger-counter used. Before discussing the results obtained we recall some facts about the circulation of labelled phosphorus in the blood. Sensitivity of labelling Let us start from labelled sodium phosphate preparation of such activity that when the later was first put into the blood, 1 mgm P will show 10000 activity units. As a result of a rapid exchange going on chiefly between bone phosphate and the inorganic phosphate of the blood 1 mgm will soon correspond to less than 10000 activity units. The total inorganic phosphate content of the blood remains constant, except in the case which we will not consider at present where a comparatively large amount is injected, while the individual phosphate ions will very soon be replaced to a large extent by other phosphate ions which were hitherto located in the skeleton or in other organs. After some time we shall find a large part of the labelled phosphate in the organs and the probability that the labelled phosphate leaves the organs and gets back again into the blood will increase, the effect of this re-entrance into the blood will be that with increasing time the net rate of decrease of the inorganic labelled phosphate content of the blood will be less and less. Loss of phosphate by excretion and by the formation of organic phosphorus compounds in the blood and in the organs will further complicate the curve representing the labelled P content of the blood as a function of time. We determined the latter experimentally for the blood of different ani- mals and also of human subjects, but not for the hen, (Compare, however, the results given on page 281). The conclusions drawn in this paper do not necessitate ADVENTURES IN RADIOISOTOPE RESEARCH he ~I 02) the knowledge of the change of the labelled phosphate content of the hens blood with time, it is for our present purpose sufficient to bear in mind that an initial rapid decrease of the labelled inorganic P content of the plasma occur and_be- comes slower. In the first experiments described in this paper, in contrast to most of our experiments, we administered large amounts of P, of the order of magnitude of 100 mgm. The very strongly active phosphorus preparation (of a strength of about 10® counts) used in these experiments was a generous gift of Professor LAWRENCE and was prepared by the bombardment of few grams of red phosphorus by high speed deuterium ions generated in Professor LAWRENCE’s powerful cyclo- tron. The active P was thus mixed with a comparatively large amount of inactive phosphorus. In the experiments to be described, in contrast to some other experi- ments, the comparatively large amounts of phosphorus did not interfere, their presence in the active preparation has even the advantage that we can fix exactly the limit within which the sensitivity of our indicator, the number of mgm of total inorganic P indicated by | count activity, varied throughout the experiment. The 100 mgm P administered had an activity of 10® counts. If the labelled P had not been diluted by non-labelled P of the organs we should have found after the lapse of 28 hours, the time of the experiment discussed on page 276, a specific acti- vity of the plasma blood inorg. P—activity per mgm P—amounting to about 1% of the total activity administered. (The amount of inorg. P present in the total plasma is only about 5 mgm and thus much smaller than the 100 mgm P administered.) As seen from Table 9, however, only 0.01% was found, showing that from the inorganic P atoms present in the blood of the hen after the lapse of 28 hours only 1% were those actually administered, the rest being ones originat - ing from different organs and partly also from the food taken within that time. We carried out three types of experiments: a) Administration of labelled sodium phosphate to a hen and investigation of the eggs layed at different dates. b) Administration of labelled sodium phosphate, killing the animal, removal of the yolks and investigation of these yolks, the blood, the liver and other organs. c) Experiments in vitro in which eggs were placed in labelled sodium phosphate solutions for few days and investigated as to what extent the labelled P penetra- ted inte the egg. We will first discuss experiments of the type a). a) Investigation of the labelled phosphorus content of eggs laid at different dates We injected radioactive phosphorus as sodium phosphate subcutaneously to hens and investigated the radioactive phosphorus content of the different parts of the eggs laid at different times. The first egg was layed 414 hours after admi- nistering the radicactive (labelled) phosphorus. We found the albumin to contain 0.0015% of the 40 mgm of phosphorus injected, a similar amount 0.0014°%, being present in the yolk. As the total phosphorus content of the yolk was found to be 100 mgm and that of the albumin only 4 mgm, the specific activity (active phos- phorus per mgm normal phosphorus) was twenty-five times larger on the albumin than in the yolk. We found the lecithin phosphorus to be 53% of that of the total phosphorus of the yolk and to be entirely inactive. No synthesis of lecithin mole- cules took place in the yolk therefore within the 44% hours preceding the laying of the egg, as in that case some active lecithin molecules should have been formed: taking this fact into account the specific activity of the other than lecithin phos- phorus present in the yolk works out to be thirteen times smaller than that of ORIGIN OF PHOSPHORUS COMPOUNDS IN HENS’ EGGS ITY the albumin P. As 40 mgm active phosphorus were injected and only 0.0006 mgm are found in the albumen we can conclude that the formation of albumen from inorganic blood phosphorus in the course of the last 41% hours which the egg spent in the oviduct is a very moderate one, even when we take into account that the 0.0006 mgm active phosphorus found in the yolk passed through the albumen into the yolk bringing the amount of labelled phosphorus present at least temporarily in the albumen to 0.0012 mgm and that a large part of the active phosphorus injected gets rapidly replaced in the blood by non-active phosphorus present in the skeleton and other organs. In the shell of the egg we find 10 mgm phosphorus by chemical analysis (colori- metric method of FiskE and SuBBAROW) and 0.1 mgm of the labelled phosphorus administered by radioactive determination (measurements with a Geiger-counter). 1% of the shell phosphorus originates thus from the labelled phosphorus adminis- tered, which got into the shell in the course of the last 414 hours before laying the egg. The labelled phosphorus content of eggs layed at different time is shown by the figures of the tables 4 to 6. In what follows we discuss the significance of these figures. That the specific activity of the shell is very much higher after 0.17 days than at a latter date is due to the fact that shortly after the administration of the labelled P the acti- vity of the inorganic P of the plasma is very high and it is the latter which is incorporated into the shell. As found by us in numerous cases the active P content of the plasma decreases first rapidly and later at a decreasing rate the difference between the specific activity of the plasma and that of the tissues becoming less and less. The specific activity of the shell phosphorus is a measure of that of the inorganic plasma P at the time of formation of the shell and vice versa. The low specific activity of the albumin P in the egg layed after 0.17 days comes possibly for the following reasons. The white of the egg was already to an appreciable extent formed before the administration of the labelled P. The phosphorus compound of the plasma, presumably the plasma protein which mainly enters into the white was at such an early date after the administration of the labelled P active only to a small extent. The synthesis of labelled organic compounds takes some time and therefore shortly after the administration of labelled P the specific activity of the inorganic plasma P is much higher than that of the organic P. On the other hand the labelled organic P disappears at a slower, usually even much slower, rate from the plasma than the labelled inorganic P, the latter having a unique opportunity to exchange with the inactive tissue, especially bone tissue P. When comparing the yolk figures with those of the albumin we have to bear in mind that contrary to the albumin which is formed within the day preceding the laying of the egg the greater part of the yolk was already present when the active phosphorus was injected and therefore the labelled phosphorus of the yolk was diluted by the unlabelled phosphorus already present in the yolk. With increasing time we should expect the amount of active phosphorus in the yolk to increase. Labelled P administrated at different dates In another set of experiments we were interested in producing strongly active egg-lecithin to find out whether after feeding the latter as dry yolk to rats, the presence of active lecithin in the blood of the rats can be ascertained. This was found not to be the case. In these experiments we administered to the hen on several days active phosphorus which made the interpretation of the activity- measurement of the yolk removed from the ovary rather difficult. A comparison of the activity of the shell of the yolk with its fluid interior revealed large diffe- 280 ADVENTURES IN RADIOISOTOPE RESEARCH rences. The semi-solid yolk shell formed from very active blood was found in one case to be seven times more active than the fluid interior of the yolk, and five times in another case. With decreasing size of the yolk the difference between the specific activity of the yolk phosphorus originating from the inner and the outer part of the yolk diminished and finally vanished. TABLE 4. — ActivE PHospHORUS CONTENT oF Eces. HEN I. Percentage of active phosphorus administered found in: — Egg laid after ad- | | | | | | ministration of active Shell | Albumin Total yolk | Yolk lecithin phosphorus | | ! = aA ie Seat oe. =a | oa Oslidanys: ve issc.e 42 0.24 |. «0.0015 ~)' 020014 | --.0.000 IO RM csy srene atevsre | 0.052 0.032 | 0.109 0.014 | | | | BU Aa Miers 0.036 0.030 | 0.42 Oe 4.5 A Som eiainc 0.026 | 0.027 0.95 0.34 O25g) Tosael\ beneieroe 0.022 | 0.020 | 0.85 0.35 TaBLeE 5. — Hen I. Percentage active phosphorus administered found in 1 mgm egg phosphorus x 10%. (Specifie activity x 108). Yolk after | Egg laid after administra- a ; ee ter | Yolk ; 2 i Shell Albumin removal of 7 tion of active phosphorus : phosphatides | phosphatides | aa pan Ness nia _ ORE Kast Sore byersverecere 24.0 OS 8icg| 0.03 | 0 | | KOK esteem etek! Ss hi ed Whe t8 OL is} 2.0 0.26 RD ee okra eee ae he S268 ea Bes) 5.1 3.3 nb ae et areas Ne de ae2kG | 12.6 | 6.4 eS | m | : _ 6.5 sot aeeparcts sharrestens | 2-2 | 5.0 10.4 7.0 | i] TABLE 6. — DISTRIBUTION OF THE ACTIVE PHOSPHORUS ADMINISTERED BETWEEN DIFFERENT PARTS OF THE EGG Yolk after Aphis Egg laid after administra- | Shell | Albumin | : é soy Yolk lecithin 2 Tee: - * | . extraction of Tat late tion of active phosphorus 6 | h phosphatides ke? phosphatides % em el USE GENES moaooaaue Ge 98:9" | O26 0.5 | 0.0 Osama eae Beet 270° | tO 9a) e4ea 7.2 SO en ela secs eae lin 720 | 6.0 | 50.9 | 36.1 LA ls ae ac ae |, e265 oe | 60.5 34.2 6.5 oF. uae SDIO CRO OIOIOe | 7445) | Ded | 56.0 | 39.3 vem rials Ose CNS od6eoosgucer | 46.1 35.0 18.9 | —_—_—_—_—_—_—_—_—_—_—__=—_—X—«X—«X«K¥—«Kx—=n”n'F5SsS5vV“""_— c 4 oooccusognoad | 30.5 6.8 38.2 24.5 | Le ir) eaten. 15.0 G2amele) 656-4 32.4 6 Noe ee ee 2.8 | ORIGIN OF PHOSPHORUS COMPOUNDS IN HENS’ EGGS PR] b) Specific activity of yolk phosphorus We administered to a hen 100 mgm. P as sodium-phosphate showing an act ivity of 106 counts and killed the hen after the lapse of 28 hours. From the ovary 34 yolks were removed and from the oviduct one egg. The weights of these are recor- ded in Table 7. TABLE 7 Weight of yolk Specimens present About 3 OpsrIa PATA Rss) o ekoueuenereqeneraiey cere | 20 100 boodoucdoDooo obo 9 HOO Aree ae es coke oes | 1 ZO OR iil. eveeunsi aims oO edeteva olane 1 TIM SOUOR Meme ase. oats Gey tite. l C0 7500 AG BGs. gibi O85 KOEI ] Se LS OOON e551 Mvens tere len shevere stateress isle 1 CE) 1S 0.0 Oley enadecsc) ens suerotessyeyskere 1 The specimens of 700 mgm and more were treated separately while averages of the 30 mgm and the 100 mgm. yolks were taken. The lecithin was extracted by ether and the residue brought into solution as described above. The results obtained are seen in Table 8. The specific activity of the total P shows a maximum in the case of the 2500 mgm yolk. This result, puzzling at first sight, can be easily understood after consi- TaBLeE 9. — Speciric AcTiviry oF YOLK PHOSPHORUS (Percentage of the activity administered present in 1 mgm P) = Phosphatide P | Non Sela Weight of yolk * es | Total P “A % | 30==LOO mpm 4.24726. -0ee 0.00055 | 0.018 0.0073 700 en seer Sede 0.00814 | 0.0173 | 0.0129 Sa) ey Ae mm | 0.0147 | 0.0186 | 0.0164 4600 een ru eet See | tod | i 0.0090 7700 nee er ee | = | = | 0.0055 [EGU je eee eee | Bk | = | 0.0044 dering Fig. 2 taken from a paper of H. Geruartz!, in which the daily increase in weight of the yolks of a hen is recorded. The yolk grown from active blood and thus active will be diluted by the non-active yolk already present and this dilution will lead to a decrease of the specific activity of its P content. The dilution being least in the case of the 2500 mgm yolk, (comp. Fig. 2) its specific activity is bound to be highest. It takes some time after administration of the labelled sodium phosphate until labelled lecithin is transported into the plasma whereas 'H. Geruartz, Arch. dtsch. Ges. Physiol. 156, 215 (1914). 282 ADVENTURES IN RADIOISOTOPE RESEARCH inorganic P of very high activity is present almost at once after injecting the active sodium phosphate. The non phosphatide P of the yolk is partly inor- ganic P which gets into the yolk in the early stage of the experiment when its specific activity is very high; we must therefore consider the lecithin P and not the non- lecithin or total P content as a proper measure of the growth of the yolk. From the fact that the lecithin P of the 30—100 mgm yolks became active only to a very slight degree we 15 must conclude for example that they had hardly grown within the last 28 hours. When discussing the labelled lecithin P (phospha- 12 tide P) present in blood and in i some of the organs we shall find definite evidence that the yolk eo lecithin is drawn from the plasma 9 lecithin. 8 It is of interest to remark that a6 the ratio of the total active leci- 5 thin content of small yolks, such 6 as would require 10 days or more 5 to attain completion, is a quanti- J tative measure of their relative growth since the administration 3 of the labelled P. When, however, 2 comparing the lecithin P activity of a small yolk which increases its weight in the course of a day only to a slight extent with that Fo (ith (Mey 3) 2 | ; : days of a large yolk growing at a rate of few gm per day the ratio of the Fie. 2. Increment of the weight total activities will not always be of yolks in the course of 12 days a correct measure of the growth before completion of the yolk since the administration of the la- according to Gerhartz. belled P. It may happen (comp. Fig. 2) that the growth of yolk per hour is larger at the end than in the beginning of the experiment the latter process determining thus to a larger extent the total growth within the time of experiment. From which it follows, that if at the beginning of the experiment the specific activity of blood lecithin happens to be greater than at the end, we underestimate the growth of the large yolk. It is, however, the determination of the slow rate of growth of the small and tiny yolks, often present in a very large number in the ovary, which can be of special interest and which can hardly be determined by any method other than that outlined above. Investigation of blood phosphorus Plasma and corpuscles of the hens blood were separately investigated using the experimental method described on page 276. The results obtained are seen from Table 9 which contains data on the specific activity (activity per mgm P in percent of that injected) and also the total phosphorus present in the hens blood under ORIGIN OF PHOSPHORUS COMPOUNDS IN HENS’ EGGS 983 the assumption that the volume of blood of the hen amounted to 150 cc. and the volume of the blood plasma to 100 cc. TapBLe 9. — Speciric Activiry AND ToTAL PHOSPHORUS CONTENT oF THE HEN’s BLoop pe Specific | Total phosphorus Fraction | activity content Plasma inorganic .......2........-- 0.0104 5.4 Plasma phosphatide .............-. 0.0125 20.0 Corpuscles phosphatide ........-.-. | 0.0046 11.3 Corpuscles acid soluble ............ | 0.00386 | 26.5 Corpuscles proteim ................. | 0.0031 | 15.9 That the specific activity of the plasma phosphatide P is greater after 28 hours than that of the inorganic P is due, as discussed on page 279, to the rapid disappea- rance of the individual inorganic P atoms from the plasma. In the experiment discussed on page 286, in which the hen was killed only 5 hours after the admi- nistration of the labelled P, the specific activity of the phosphatide P was found to be only 42% of that of the inorganic P. It is of great interest that the specific activity of the plasma phosphatide P is several times larger than that of the corpuscle phosphatide which shows that a much smaller percentage of the corpuscle phosphatide than of the plasma phos- phatide is renewed in the course of the experiment. This is an interesting result as it definitely disposes of the often discussed possibility that the blood phospha- tide is synthetised in the corpuscles. Some of the corpuscles being formed during the experiment from labelled plasma are bound to contain labelled phosphatides; labelled phosphatides can furthermore easily get into the stroma of the corpuscles which are partly composed of phosphatides. A very suggesting change in the phosphatide content of hens blood at the time of production was ascertained by HELLER, Pau, and THompson (comp. Fig. 1). The most interesting feature of the curves recorded by them is a gradual increase in the total P of the blood at the time of production, this high level being held during the entire production period with some fluctuations and dropping quickly as production ceases and molting season approaches. The increase is due to that of the lipoid P and is much more conspicious in the case of the plasma than in that of the corpuscles; the lipoid P content of the plasma is higher all through than that of the corpuscles, at the peak of production the former value being nearly three times higher than the last mentioned one. As about 2/, of the blood volume is composed of plasma it follows, that from the total lipoid P present in the blood */, are to be found in the plasma. The predominance of phosphatide P ip the plasma found for laying hens is entirely unique as seen from the figures of Table 10, but understandable if we envisage the great strain put on the organism of a hen as to lecithin supply. A hen laying daily has to produce about 60 mgm lecithin! P a day: taking a total plasma volume amounting to 100. cc the total lecithin P of the plasma works out to be 20 mgm. If the lecithin found in the yolk is, as suggested from our results, taken from the plasma lecithin then the plasma has to give off three times its total lecithin content in the course of a day thus 1 Lecithin plus other phosphatides. 284 ADVENTURES IN RADIOISOTOPE RESEARCH putting an appreciable strain on the lecithin circulation. A strain which would be still more pronounced in the case of a lower plasma lecithin content. TaBLE 10. — PHOSPHATIDE P IN PLASMA AND CELLS OF DIFFERENT ANIMALS mgm % P in Plasma | Cell heey eee | | plasma Ua Ditokershets rekace (oe (os catersy s 26 10 3.8 URMalls lo icbienrerstss epee one “oes eters 3° 12 3.6 WiaMe aie teiete reso ae S508 Gunre 9 19 ail TD OO ereyeusyers, 0 Gisy eee, 0s) oe 14 14 1 Wanyamne: shenl sr <<) 1-1 14—20 8—23 0.87 Protein phosphorus in the hen’s blood After removal of the phosphatides and the acid soluble phosphorus, the remain- ing P is generally assumed to be present as protein P. The protein P content 31.8 mgm%, found in the corpuscles of the hen in the 28 hours experiment is much higher than in the corpuscles of the blood of other animals, the corpuscles of the rabbit containing for example, as found by Mr. Arren, 4.4 mgm %. The same considerations apply to the protein P content of the plasma, which was found to amount to 9.4 mgm% for the blood of the hen in question and of 7 mgm% in the case of the hen discussed on page 276 while the plasma of a rabbit, for example, was found to contain only 0.03 mgm % piotein P. From the high value of the specific activity of the protein P in the 5 hours experiment it follows that the protein phosphorus compounds present in the plasma were renewed even at a higher rate than those of the phosphatides. This result suggests a great participa- tion of the plasma phosphorus protein in the formation of the egg. To arrive at a final conclusion as to the relation between the phosphorus protein compounds of the plasma and those of the yolk and white is difficult because of the fact that we lack simple methods of separation of the protein compounds. Vitellin, for example, can only be isolated by a very tedious and lengthy process and the isolation and separation of the blood protein phosphorus compounds are still more difficult, partly because only small quantities of these substances can be secured in the experiment. The fact that we have to base our conclusions on the amount of phosphorus present in the residue, remaining after extraction of the phosphatides and the acid soluble phosphorus compounds makes the result obtained less trustworthy than those arrived at when investigating the phospha- tides, for example. The high value for the protein phosphorus of the corpuscles found by us, which may to some extent be due to an incomplete separation ot the phosphatides and acid soluble P, is supported by the data obtained by HELLER, Paut and THompson. They find for the total P present in the cells of laying hens about 100 mgm %, but only about 40 mgm % for the sum of inorganic acid soluble and lecithin P. The discrepancy suggests the presence of a further not investigated P fraction, which might be protein P. In the case of the plasma phosphorus the curves of HELLER and his colleagues show the anomaly mentioned only to a smaller degree; the total phosphorus found by them is not very much larger than the sum of the acid soluble and phosphatide P. ORIGIN OF PHOSPHORUS COMPOUNDS IN HENS’ EGGS IRD The high protein P content of the blood plasma of a laying hen has presumably the same biological significance as the high phosphatide P content, namely a reduction of the strain put on the protein resp. phosphatide producing and carrying system in the organism of laying hens. Phosphatide content of the liver We extracted the total phosphatide content of the liver of a hen 28 hours after the administration of the labelled P, using the method described on page 277. Since we were interested in seeing whether lecithin soluble in ether shows the same specific activity as the total phosphatides we extracted another part of the liver tissue with ether alone. We found no marked difference, as seen from Tables 11 and lla, which also ccntain data on the specific activity of inorganic and acid soluble (other than inorganic) P of the liver. As seen from Table 11 the specific ativity of the liver phosphatide P is 56% of that of the inorganic P, from which it follows that about one half of the phos- phatide molecules are labelled and thus formed after administration of the labelled TaBLE 11. — Speciric ACTIVITY OF THE LIvER P (Activity per mgm. P) Fraction Specitie activity Total phosphatides (ether-alcohol extract) . | 0.0152 Mecithine (ether vex tract) acct lle ele i-ierere = 0.0158 Jao ReAING oo SoS sabe bic ob Soca DO OOOO obo OC 0.0272 Acid soluble, other than inorganic ........ 0.0224 sodium phosphate. This result must, however, be interpreted with great caution. As already mentioned on page 277 in the early stage of the experiment the specific activity of P of the plasma is much higher than in the latter stages and the inorga- nic P of the liver was also more active at the early stage. This change of the speci- fic activity with time would not affect our results if the specific activity of the phosphatide P should decrease with time at the same rate as does that of the inorganic P. That is, however, not the case. The phosphatide molecules can mainly escape from the circulation at an appreciable rate into the yolk, while the indivi- dual inorganic P atoms present in the plasma can rapidly exchange with such present, for example, in the skeleton, the latter being a much faster process in view of the huge extent of the skeleton. Therefore, when drawing conclusion TasBLE Ila — PERCENTAGE OF LABELLED P ADMINISTERED FOUND IN PLAsMA, CORPUSCLES AND LIVER ; P Total Plasma Total Corpus- Total Liver Fraction (100 gm) % cles (50 gm) % (44 gm) % Phosphatide B=... --,. 0.25 0.052 0.608 Mbaxorsernance JB So oe56hoc 0.056 — | Total acid soluble P — 0.100 1.643 Jeadoteival IP econ scosoe 0.176 0.050 | Total EP voy rere eee 0).482 0.202 2.251 DR6 ADVENTURES IN RADIOISOTOPE RESEARCH from the comparison of the specific activities of the phosphatide P and the inorganic P as to what extent the phosphatide molecules got labelled we are apt to get values which are possibly too high. A trustworthy value could be obtained by keeping the specific activity of the inorganic P of the plasma constant by continuous injection of labelled phosphate of varying concentration and by thus avoiding a decrease in the specific activity of the inorganic P of the plasma, which is used for the synthesis of the phosphatide molecules in the liver and elsewhere. In the above case we can, however, conclude that a very appreciable part of the liver phosphatide molecules most have been re- newed within the 28 hours of the experiment. In experiments on isolated livers in which the skeleton and other organs are not present it is easy to calculate from the ratio of the specific activities of inorganic P and phosphatide P the amount of newly formed phosphatides. In an isolated liver of a cat in the course of 2.5 hours about 1° of the phosphatide molecules present are newly formed. If in the course of 2.5 hours in an isolated liver of a cat about 1% of the phosphatide content is renewed there can be hardly any doubt that in the liver of a living hen in the course of 28 hours a large part of the phos- phatide found is synthetised during that interval; in the liver of a living animal the enzymatic and other actions necessary for the synthesis of phosphatides are certainly as abundant as in an isolated liver and the phosphatide formation in the liver of a laying hen could hardly be less than in that of a cat. We are led to the same conclusion by the following consideration. The daily amount of phos- phatide P tranfered from the plasma into the ovary is, in the case of the hen in question, which was laying one egg every other day, about 50 mgm. The main source of phosphatide production is, as we will see, the liver, and an amount not very far from 60 mgm must therefore have been produced daily in the liver of the hen. Since the latter containing 38 mgm of phosphatide P, a Jarge amount of the liver phosphatide must have been renewed during the experiment. A similar conclusion applies to the plasma phosphatides, the 50 mgm phosphatide P being carried by the plasma, the total content of which is 20 mgm, the plasma phospha- tide molecules must have been renewed repeatedly. We are thus led to the result that the main source of phosphatides in a laying hen is the liver and that more than one half of the phosphatide molecules present in the hens liver were newly formed during the 28 hours preceding the admini- stration of labelled phosphate. That the greatest part of the phosphatides is formed in the liver of a laying hen and reach the ovary through the plasma is very clearly shown in an experiment in which the hen was killed only 5 hours after admi- nistration of the labelled phosphate. The acid soluble P of the liver, other than inorganic, mainly derived from P ester, shows, as seen in Table 11, a higher specific activity than the phosphatide P present in the liver. Experiment on a hen killed after five hours 3.8 ec. of physiological sodium chloride solution containing 10 mgm labelled sodium phosphate were injected subcutaneously into a hen which weighed 1800 gms. The hen, which layed previously one egg daily weighing about 45 gm was killed after the lapse of 5 hours. The results obtained are seen in Table 12. Two separate determinations were carried out, the values found and also their average are given. As seen from Table 12 the specific activity of the phosphatide P, which is a measure of newly formed phosphatides, is by far the greatest in the liver and markedly higher than that of the plasma phosphatide P. Contrary to the 28 hour ORIGIN OF PHOSPHORUS COMPOUNDS IN HENS’ EGGS 287 experiment, where the percentage of newly formed phosphatide molecules in the plasma nearly reached that found in the liver, in the 5 hours experiment the concentration gradient in the flow of labelled phosphatides directed from the liver into the plasma is very clearly shown (comp. Fig. 3). The percentage of label- led molecules in the ovary phos- phatide the other hand, much smaller than in the plasma phosphatides. From this it follows that the labelled phosphatide mo- lecules present in the ovary were within 5 hours only partly replaced by ones present in the plasma. We investigated a yolk weighing 1.0 gm. The figures obtained are isseon given in Table 12. A second yolk Pies investigated weighed 2.7 gm. and its lecithin P had a specific acti- — spieen — — =a Pe ag vity of 0.0050. The specific act- fe et esc GS = AS vity of the yolk lecithin of the first mentioned yolk was found to be of that of the plasma lecithin. From these figures it follows that about 1/,, of the 1.0 gm i. e., 0.09 gm of yolk were grown within 5 hours. The actual growth was, however, presumably greater than 0.09 since in P 1/ about 14/,, Fie. 3. The heaviness of the sha- gm, ding indicates the specific activity the early stages of the experiment the plasma phosphatide was only very slightly active and so was the yolk tissue formed in this phase of its development. The fact that TaBLe 12. — Spectric ACTIVITY OF of the lecithin P and thus the per- centage of the phosphatide mole- cules formed within the last five hours in the total phosphatides of the organ in question. PHOSPHATIDE P Specific activity (% of activity given, found in 1 mgm P) | Relative specific activity ; that of the Organ | Se —} inorganic plasma P Single values Average taken i (| 0.09 WIVErs sib4sc ne ste eee ) : ( 0.088 0.54 ) 0.082 \ 0.06! j Blasmaeessrerccscrescae oY ( 0.069 0.43 0.069 \ nae 0.006 Ovary \ 064 | 0.0064 0.039 ) 0.0064 — § y : 0.005: = Yolk o jevelene| Siclahs lel eie, © } ¢ 3 0.0064 0.025 t} 0.0075 § ¢ 0.018 Impestine srrenaeernere 01 ( 0.018 0.11 } 0.018 § SY OMG Wipe alain dic < 0.02 =< ().02 0.1 288 ADVENTURES IN RADIOISOTOPE RESEARCH the specific activity of the ovary phosphatide was found to be low, as low as that of the yolk, proves definitely that the role of the ovary is not produc- tion of phosphatides but their extraction from the blood plasma together with other suitable constituents. The combination of phosphatides with proteins giving the characteristic composition and consistency of the yolk, is evidently one of its principal functions. In the experiment described above the specific activity of the P of the yolk soluble in trichloracetic acid was found to be 0.035, thus !/,, part of that of the inorganic P of the plasma the latter being 0.16. Making the assumption that most of the acid soluble P originates from the inorganic P of the plasma we find a growth of the yolk amount- ing to 1/, , part of its weight of 1.0 gm during the experiment. While the above mentioned figure of 1/,, was, as already mentioned, a lower limit of the part of the yolk newly formed within 5 hours, the figure of 1/4.; is a higher limit. A part of the acid soluble yolk phosphorus was formed at an earlier stage when the speci: fic activity of the plasma inorganic P was appreciably higher than at the end of the experiment, and as our calculation is based on the specific activity of the plasma inorganic P at the end of the experiment it gives too high a value for the amount of yolk formed during the experiment. The phosphorus of the white of the egg removed from the oviduct had a low specific activity, namely 0.0013. This is an interesting result in view of the strong activity shown by the phosphorus compounds present in the plasma (comp. Table 13). A possible explanation of this result is that some of the phosphorus present in the protein or other compounds of the oviduct tissue is utilised to produce the phosphorus compounds present in the white of the egg. In the course of five hours perhaps the compounds present in the tissue of the oviduct get labelled only to a slight extent. An other explanation is that while the average plasma protein P has a high specific activity 0.15 after the lapse of five hours, the specific activity of the phosphorus of one of the components of the protein mixture might be low. We are now engaged in the investigation of the origin of the phosphorus present in the white of the egg. Taste 13. — Speciric ACTIVITY OF PrasMA PHOSPHORUS Fraction | Specific activity Ibravergeryovker 12 Seogcoaccac 0.16 Wecithinwk eerie cies 0.069 Protein #2 cams eecrac O.14 c) Experiments ‘n vitro We placed eggs in a neutral physiological sodiumphosphate solution containing 30 mgm P for 24 hours and investigated the activity of the different parts of the eggs, the results being seen in Tables 14 and 15. The comparatively high labelied P content of the shell is due to phosphate exchange processes between the large shell surface and the solution and possible also to the formation slight amounts of calcium phosphate trom the carbonate of the shell. An investigation of the activity of the lecithin extracted from the yolk gave an entirely negative result, this in agreement with the observation recorded on p. 279 that after the egg left the ovary po more lecithin formation takes place. ORIGIN OF PHOSPHORUS COMPOUNDS IN HENS’ EGGS 289 TABLE 14. — Ratio oF THE Specific AcTIvVITY OF Eaa P anp SoLturion P | Shell | Albumin Yolk | a — | | Kee I (Total P)..... lS elO met eas SolOme2 9 Sx Lom Egg II (Total P)..... PAO SAN ae |i) ess Se Oat | 4.0 x 1075 TABLE 15. — DISTRIBUTION OF THE ACTIVE PHOSPHORUS TAKEN UP BETWEEN THE DIFFERENT Parts OF THE Eao | Shell Albumin Yolk MB proms Wayeystern ad Seveh dic tehisate 99.46 | 0.44 0.10 1 Det 0 Ra ees a nen 99.40 0.41 0.19 DISCUSSION We saw that by investigating the labelled phosphorus content of eggs or yolks we could draw conclusions as to the growth of the egg or yolk since the date of administration of the labelled P. It is, for example, possible to show that while the egg is in the oviduct not only is no more yolk formed but also no new lecithin molecules are synthetised. Should suitable enzymes be present, new and thus labelled lecithin molecules could be formed without any growth of the yolk. The ovary of a laying hen contains numerous tiny yolks growing at a slow rate; by comparing the incorporation of labelled P by such yolks, we get a quantitative measure of their relative growth since the administration of the labelled P. When comparing the growth of small yolks with large ones we can usually not obtain strictly quantitative results as to the relative growth because of the much more rapid relative growth of large yolks compared with that of small ones. Placing eggs in a solution containing labelled P for some days we find the shell to contain an appreciable part of labelled P, while the amount shown by the white and especially by the yolk is very small, though easily measureable, even in the case of the yolk. No formation of labelled lecithin is, however, found in the yolk. As to the formation of lecithin in the growing yolk, we arrive at the following result: The phosphatides found in the yolk are synthetised at least to a large extent in the liver and are transported through the plasma to the ovary which extracts the phosphatides. This is most clearly seen in the experiment in which the hen was killed only 5 hours after the administration of the labelled sodium phosphate. In this experi- ment the specific activity of the liver phosphatide P reached 54% of 19 Hevesy 290 ADVENTURES IN RADIOISOTOPE RESEARCH that of the plasma inorganic P, while the specific activity of the plasma phosphatide P was appreciably smaller, amounting to only 43%; that of the ovary was very much smaller, namely 3.9%, and about as large as that of the strongest yoik phosphatide P. In the 28 hours experiment as to be expected, the difference in the specific activities was much smaller, the specific activity of the liver phosphatides being only some- what higher than that of the plasma phosphatides. In the 28 hours experiment on the hen which used to lay one egg every other day the amount of phosphatides passing through the plasma on the way into the ovary was, in the course of the experiment, about twice the amount of phosphatides present in the plasma. In the 5 hours experiment, in which the hen experimented on was laying one egg daily, the amount, of phosphatide passing the plasma on the way into the ovary was about half the amount present in the plasma. From the low specific activity of the phosphatide P, that is from the low percentage of newly formed phosphatide in the ovary, it follows that in this organ only an insigni- ficant amount of phosphatide can be formed. We have also to consider that a part of the labelled phosphatides found in the ovary is due to the presence of blood containing the latter. The specific activity of the plasma phosphatide P being appreciably smaller than that of the liver the labelled phosphatides must have come from the liver into the blood and not vice versa. By carrying out experiments in vitro with blood containing labelled sodium phosphate we found only a slight formation of labelled phosphatides, which is in accordance with the above con- clusion. ; The formation of phosphatides in the intestinal mucose by using radioactive phosphorus as indicator was first shown by Arto, Perrier, SANTAGELLO, SARZANA and Srcrb®. They found in an experiment, carried out on a rat, that after injecting labelled sodium phosphate the phosphatides extracted from the gut after a few days showed a specific activity only exceeded by that of the liver phosphatide P, the ratio of the specific activities being 1.2. The phosphatide production in a laying hen is larger than in any other animal of similar size, as the amount produced daily to be incorporated in the yolk is as much as about 2 times that present in the liver which contains more phosphatide than any other organ. The laying hen is, therefore, a very suitable animal for studying phosphatide formation. In our 5 hours experiment the specific activity of the intestinal phosphatide P is much smaller than that of the liver phosphatide P and also than the plasma phosphatide P. The bulk of the labelled P present in the plasma can, therefore, not originate from the intestinal phosphatide and the latter can not be the chief source of the yolk phosphatide. The phosphatides formed in the () Nature 139, $36 (1937). ORIGIN OF PHOSPHORUS COMPOUNDS IN HENS’ EGGS 29] intestine can, however, have and presumably actually do have a role in the supply of the plasma phosphatides. The presence of phosphatides in the intestinal lymph was repeatedly shown™ in experiments on dogs. The amount of phosphatides reaching the hens circulation by the influx of intestinal lymph could be ascertained by measuring the amount of intestinal lymph produced and also its phosphatide content. In Fig. 3 we show the specific activities of the phosphatide P in the organs of the hen killed 5 hours after the administration of the labelled sodium phosphate. The heaviness of the shading indicates the specific activity, A hen laying daily deposits about 60 mgm phosphatide P in the yolk or about 3 times as much phosphatide as present in the plasma. In the course of a day the phosphatide content of the plasma of a laying hen must therefore be replenished three times. In view of this great strain on the phosphatide circulation in the plasma it is very significant that the plasma phosphatide content of a laying hen is higher than in most other animals. If the laying hens plasma should show such a low phos- phatide content as does a rabbit or a rat (per cc.) the plasma lecithin would have to be replenished as much as 17—22 times a day. It is significant that the high phosphatide content is maintained only during the laying period and that the red cells contain less phosphatide than the plasma, a behaviour not shown by the blood of any other animal investigated. We find furthermore that in the course of 28 hours taken by the experi- ment a much greater part of the phosphatides found in the plasma is labelled than of that contained in the corpuscles. This is a significant result as it demonstrates clearly that lecithin is carried to the ovary by the blood plasma and not the blood cells which obtain their lecithin in various ways. Labelled phosphatide could be taken up by the cell membrane, possibly diffuse through the cell membrane; labelled inorga- nic phosphorus which was found by us to diffuse at a moderate speed into the corpuscle could lead to the formation of labelled phosphatide phosphorus inside the latter, finally the lecithin could get into the cor- puscles at their birth. If they are formed from labelled plasma the newly formed corpuscles should become labelled as well. As to the formation of labelled phosphatide from labelled inorganic P in blood, we found, in experiments in vitro that such a formation actually takes place, though only on very minute scale. As to the rate of formation of blood corpuscles, some information on this point could be obtained by inject- ing labelled plasma and investigating the radioactivity of the phosphorus compounds isolated from the corpuscles after the lapse of some time. If after the lapse of a day, for example, only 1% of the corpuscle phos- phatides were found to be labelled we could conclude that the rate of () H. E. Hamericu, Amer. J. Physiol. 114, 342 (1934); S. Freeman and A. C. Joy, loc. cit 110, 132 (1935). 19* 292 ADVENTURES IN RADIOISOTOPE RESEARCH formation of the corpuscles per day is less than 1% of the total corpuscles present. As to the white of the egg, we find that at least a large part of its phosphorus content is drawn from organic phosphorus compounds, possibly from protein phosphorus. We arrived at this result by compar- ing the specific activity of the phosphorus of the white of the egg with that extracted from the shell. The latter derives its phosphate content from the inorganic P of the blood plasma and is accordingly a convenient measure of the activity of the latter. The shell is formed at about the same time as the white of the egg, the great discrepancy between the specific activity of the shell P and albumin P exclude the possibility that they are of common origin. Summary By administering labelled sodium phosphate to laying hens the share of the labelled phosphorus administered in the formation of the volk, albumin and shell of the egg can be followed by aid of radioactive measurements. The comparison of the specific activity (activity per mgm P) of the phosphorus extracted from blood plasma phosphatides with that extracted from the liver, the ovary, and the yolk phosphatides leads to the result that the bulk of the phosphatides of the yolk originate in the liver. It gets from the liver into the plasma and is then taken by the latter to the ovary. No formation of phosphatides takes place in the oviduct. After the egg leaves the ovary no more active phosphatide is formed. No formation of labelled phos- phatide in the yolk can be ascertained in experiments in which an egg is placed for a day in a labelled sodium phosphate solution. In the last mentioned experi- ment in vitro slight amounts of labelled phosphorus are found in the yolk, appre: ciable quantities in the white, and large amounts in the shell. The specific activity of the phosphatides extracted from the blood corpuscles was found to be only 1/, of that extracted from the plasma. Therefore, we conclude that the phosphatides formed in the liver and other organs are carried to the ovary by the plasma rather than by the corpuscles. The latter apparently play no important role in this process. Originally published in Biochem. J. 32, 2147 (1938) 32. THE ORIGIN OF THE PHOSPHORUS COMPOUNDS IN THE EMBRYO OF THE CHICKEN G. C. Hevesy, H. B. Levi and O. H. Resse From the Institute of Theoretical Physics, University of Copenhagen SEVERAL of the numerous compounds containing phosphorus present in the embryo of the chicken! occur in the yolk and the white of the egg. Those which do are chiefly phosphatides and nucleoproteins but, as Table 1 shows, other phosphorus compounds also occur in those parts of the egg. TABLE 1. — [Purwmer + Scort, 1909]. PERCENTAGE oF THE Toran P (94 mgm) AT THE BEGINNING AND THE Enp or INcUBATION OF A HeEN’s Eae | | | Beginning End 7 | ame i. e Itevongeraauke; 12) Sd oogonouceanoec Trace | 60 Water-soluble PB oo. 5......24. leteGeaie Pale 2 S86 Bther-soluiblepearaeererseeicies aoe | 64.8 "| 19:3 Watte [iris Bases te tcnee a. css as iat "20 KUGLER [1936] has lately found that, on the twentieth day of incu- bation, i. e. the last day but one, only 25 mgm of the 65 mgm of lipoid P originally present in the yolk remained there; 8 mgm were found in the embryo, and the remainder had been hydrolysed yielding inorganic P. About two-thirds of the phosphatides present were found to be lecithin and one-third kephalin. In view of the large store of phosphatides present in the yolk even shortly before the egg is hatched, we should expect the embryo to avail itself of this store when it needs phosphatides to build up its nervous system and other organs containing these substan- ces. We can test this point by introducing labelled (radioactive) sodium phosphate into the egg before incubation and investigating if and to what extent the phosphatide of the yolk and of the embryo become 1A detailed investigation of the soluble phosphorus compounds present in the embryo of the chicken was recently published by NeepuHam et al. (1937). 294 ADVENTURES IN RADIOISOTOPE RESEARCH labelled. If the yolk phosphatide remains unlabelled while that of embryo becomes radioactive, we can conclude that the phosphatide molecules present in the embryo have not come from the yolk but have been built up in the embryo with the participation of labelled inorganic P. Similar considerations apply to certain other compounds occurring in the embryo. METHODS The phosphorus content of a series of solutions is usually determined colorimetrically. For example, the inorganic P present in one sample of an acid-soluble fraction can be determined in this way, and then in another sample the phosphagen-P present can be converted into inor- ganic P, so that colorimetric determination now supplies the value for the inorganic P + phosphagen-P. In our experiments this was inade- quate. We had to measure not only the P content but also the activity of the various fractions, so we had to obtain precipitates in each case. To obtain sufficient precipitate when dealing with eggs only incubated for a few days, it was necessary to work with several eggs simultaneously. We precipitated the phosphorus, after bringing it into the inorganic state, aS ammonium magnesium phosphate. The precipitate was then dissolved in 0.1 MN HCl and an aliquot part was sucked into a glass cuvette. This was placed below the Geiger counter used to determine the activity of the preparations, while another aliquot part was utilized for the colorimetric determination of the phosphorus content. The glass cuvettes were covered with a thin mica window (5—6 mgm _ per cm?) which only absorbed to a negligible extent the B-rays emitted by the radioactive phosphorus; the area of the mica window was 1.1 cm? and the liquid content of the cuvette amounted to about 0.5 ml. We were interested in the determination of the activity of 1 mgm P prepared from different phosphorus compounds present in the embryo or in the remains. Accordingly we were not concerned with quantitative determination of the P compounds present but concentrated our efforts on obtaining the various fractions in a pure state—to avoid, for example, traces of inorganic phosphate remaining in the phospha- tides extracted from the yolk. As the phosphatides of the yolk were found to be but slightly active, while the inorganic P was strongly active, even a small contamination of the former by the latter was to be avoided. The white, the yolk, the embryo and, in some cases, the amniotic and allantoic liquids were worked up simultaneously. As regards the white we were only interested in the total activity present after incubation. The white was ignited (reduced to ash) and its phosphorus precipitated as ammonium magnesium phosphate. ORIGIN OF THE PHOSPHORUS COMPOUNDS IN EMBRYO OF CHICKEN 29: Or The yolk was dried with acetone and the phosphatides extracted three times from the dry product with a 3:1 alcohol-ether mixture. The alcohol and ether were then evaporated off at about 50° in vacuo and the residue was taken up with light petroleum and filtered. The filtrate was evaporated in vacuo, the residue ignited, and the phosphorus precipitated as ammonium magnesium phosphate. Another part of the yolk was treated as follows. The acid-soluble compounds were extracted, then the phosphatides were removed as described above, and the residual part containing mainly vitellin-P and nucleoprotein-P was ignited; the P content of this last part was determined as ammonium magnesium phosphate. The embryos were dropped, immediately after being removed from their eggs, into liquid air and were subsequently pulverized. The embryo powder was then extracted several times with cold trichloracetic acid— in the first two extractions a 10% solution was used, and later one of 5%. The extract was filtered into cold concentrated NaOH solution and divided into three parts, (a), (b) and (c). From (a) a sample of the average acid-soluble P of the embryo was secured, (b) was precipitated with 25% barium acetate solution at pH 6.5. The cold precipitate was washed with a dilute barium acetate solution, centrifuged and dissolved in a few drops of cold HNOs. The inorganic P present was then precipi- tated by adding Fiske’s reagent. The remaining filtrate was hydrolysed with N HCl at 100° for 7 min. to split the two labile phosphate radicals of adenosinetriphosphoric acid. The phosphorus set free was finally precipitated as ammonium magnesium phosphate, Barium hydroxide was added to the filtrate from the barium precipitation to remove any inorganic P, the precipitate was separated by centrifuging and ethyl alcohol was added to the remaining liquid until an alcohol concentration of nearly 60% was reached. The precipitate obtained after addition of alcohol [OstERN et al., 1936] contained the hexosemonophosphate. Its P content was determined in the usual way. The third part, (c), was hydrolysed with NM HCl in the presence of 0.1 J/ ammonium molybdate for 30 min. at 40°. In the course of 30 min. most of the phosphagen present decomposed, so that the inorganic P originally pres- ent as such, and that obtained by the decomposition of the phosphagen', were secured together in this fraction. After removal of the acid-soluble P the embryo was thoroughly treated with an alcohol-ether mixture, as described above, to remove the phos- phatides. The residue, containing mainly nucleoprotein-P, was ignited with concentrated sulphuric and nitric acids and the P precipitated in the usual way. (YOn the phosphagen content of the embryo of the chicken, cf. Lehmann and Needham [1937]. ADVENTURES IN RADIOISOTOPE RESEARCH RESULTS Eggs incubated for 6—18 days. The results of the determination of the specific activities (activities per mgm P) of the different fractions extracted from seven embryos and from the remaining parts of eggs incubated for 11 days are shown in Table 2, while Tables 3—5 give the results obtained with eggs incubated for 18, 16 and 6 days. In addition to the specific activity (activity per mgm P, with that of the P extracted from the white of the egg taken as 100), we have also recorded in Tables 2 and 3 the activity (in counts per minute or in % of amount injected) and the P content of the fraction—this last quantity being determined, in all cases, by the method of Fiskr and SuUBBARROW. Ta BLE 2. — Speciric Activiry oF P EXTRACTED FROM DIFFERENT FRACTIONS OF AN Eaa IncusatTeD FoR 11 Days. (Spreciric Activiry or P ExtTRACTED FROM THE Waite Taken as 100) Teton vain, iB Sore Sase te per min activity Imbryo: Average acid-soluble P ...............5- 0.074 3.5 59 IraveryspeW JE S66 bcGu5ougbGGuOOC0DnDeSObOE 0.077 3.1 51 Adenosine-P +- inorganic P .............. 0.121 6.0 | 63 Oree tine= Pie icy oils sade ne apie see eames 0.171 S17 | 60 Ine solNyCIE” Gouoadcuonseods UOoueDU AOS 0.561 29.6 67 Residuals (?*nucleoprotei”?) Py...j.0.. 2... 1.49 85.6 72 Yolk: Phos hatide— lee sreroleeiels elon -'= elalaliste ele) cue) sels 10.4 0.55 0.067 TasBLe 3. — Speciric Activiry oF P ExTRAcTED FROM DIFFERENT FRACTIONS OF AN Eac IncuBATED FoR 18 Days. (Speciric Activiry or P ExTRACTED FROM THE WHITE TAKEN as 100) Fraction mgm P eer ie Embryo: Average acid-soluble P ................- 19.7 53.5 | 19 Inorganie (without skeleton) P .......... 10.91 DTe2M ai | Wh AM ovis) Eas aKsod\bbele) SeonocoecdoaqddaccddaC 4.50 7621) ANoleVebaie” Gaqgnoo doce ddoo oo oU np Oop Cou 0.048 0.14.4, 20 Rhospha tid ee earyetae et nic eie cetencnele erororelene 1.08 | 1.7 ll Residual (’’nucleoprotein’’) P ...........- 0.204 | 0.3 10 Mollow pee cid=soluble (Po cinete seiisieiacetee ais ate eicieus 0.828 | 1.3 11 pos pina tide-P\wacl a. le cave eee ae eee 17:50- 41 170228 0.11 esidualePs 02): <.¢.5h\ eee eee cee ete 2.16 i. ) 0124) 75040 The figures for the specific activities (activities per mgm P) of different fractions extracted from an embryo and from the remaining parts of an egg incubated for 18 days are shownin Table 3. The P content in mgm, ORIGIN OF THE PHOSPHORUS COMPOUNDS IN EMBRYO OF CHICKEN 297 the percentage of the injected activity present in the fraction and the relative specific activity are recorded; the specific activity of the P extracted from the white of the egg is taken as 100. The specific activities obtained when the eggs were incubated for 16 and 6 days respectively are seen in Tables 4 and 5. TasBLe 4. — Speciric Activity oF P ExTrRAcTED FRACTIONS or AN Eac IncUBATED FoR 16 Days. (THE SPEctIFic Activity oF P ExTrRACTED FROM THE WHITE TAKEN As 100) Fraction Specific activity Embryo: Average acid-soluble P ......... 14 Inorganic (without skeleton) P .. | 14 Abilevey ChaKel uveyrayboee” Ge dogoouusco0n 15 Creatine Patera tackler sricrioncre ors orsiols 14 Hexosemonophosphate-P ........ | 19 Phosphatide=Pyy errr cere «=! eleteler «lore « | 12 Residual (,,nucleoprotein”’) P .... | 16 Yolk: INGIclxo ONE IP GobscdacucoDs ocDd | 12 IPhosphatid 6aPae ever sierererelelelelelsele | 0.14 IN@sChe! IP Lecacauascanoscca00C 1.22 TaspLe 5. — Speciric ACTIVITY oF P ExrRAcTED FROM DIFFERENT Fractions oF 10 Eees IncUBATED FoR 6 Days. (Spectric Actrviry oF EmBpryo PHOSPHATIDE P TaKEN As 100) Fraction | Specific activity Embryo: sehosphatide Pieri. jos sic erie = | 100 Average (phosphatide) P........ 113 Yolk: Ibakoreghae 12 SoooobcacaposomooKD | 60 Acid-soluble minus inorganic P .. | 34 Rhosphatidem bya or. ar telels corer =< 0.032 FES Gel J eewars arse eietedccctens cheversl as lap: As the figures show, the phosphatides extracted from the yolk are only slightly active, while those extracted from the embryo show strong activity 1 mgm of embryo phosphatide-P is at least 100 times as active as 1 mgm yolk phosphatide P. Furthermore, the specific activity of the embryo phosphatide-P is about as high as that of the embryo inorganic P, showing that an inorganic P atom reaching the embryo has about the same chance of entering the skeleton as of being incorporated in a phosphatide molecule by an enzymic process—which of the two systems it enters is governed solely by probability considerations. From this it follows that the phosphatide molecules in the embryo are 298 ADVENTURES IN RADIOISOTOPE RESEARCH not identical with those derived from the yolk, but are synthesized in the embryo. The formation of labelled phosphatides in growing eggs was investigated by Hevesy and Haun [1938]. It was found that the phosphatides present in the yolk are taken up from the plasma by the ovary and incorporated into the latter; as soon as the yolk leaves the ovary no more change occurs in the content or composition of its phosphatides. When labelled phosphate is administered to a hen after the yolk has left the ovary and is located in the oviduct, the egg takes up active phosphate but no active phosphatide is formed. In experiments in vitro as well, eggs placed in radioactive sodium phosphate solution take up active phosphate but no active phosphatides are formed. The slight activity of the phosphatides present in the yolk of incubated eggs is presumably due to the influx into the yolk of small amounts of active phosphatides synthesized in the embryo. This view is supported by the fact that the ratio of the specific activities of the embryo phosphatide P and yolk phosphatide-P was much larger (3000) in the 6 days experi- ment than in the 16 days experiment (100). The activity of the residual P of the yolk, which is mainly composed of vitellin and nucleoprotein, was larger than that of the phosphatides; this can be understood if we admit the possibility that the extraction of the strongly active, non- protein constituent of the yolk is not quantitative, for in this case the specific activity of the residual P would be increased. The embryonic residue obtained after extraction of the acid-soluble and ether-soluble constituents is composed chiefly of nucleoproteins. That the specific activity of the nucleoprotein-P is the same as that of the inorganic P extracted from the embryo is not surprising, because much less nucleoprotein is present in the yolk than in the embryo (Table 1). The greater part of the nucleoproteins present in the embryo must therefore have been built up in the course of incubation; during this process labelled phosphate has an opportunity of entering the nucleo- protein molecules. Distribution of radioactive phosphate in the egg The greater part of the sodium phosphate injected into the white is still found at the end of the 6 days experiment in that part of the egg. The distribution of the activity between white, yolk, connecting fluids (which were not, however, free from white and yolk) and embryo is seen in Table 6. The low activity of the yolk might possibly be due to a slow rate of penetration of the vitellin membrane by the phosphate ions; this point is under investigation. Another possible explanation is that the inorganic P content of the yolk is lower than that of the white. If a distribution ORIGIN OF THE PHOSPHORUS COMPOUNDS IN EMBRYO OF CHICKEN 299 equilibrium is reached, the activity should be proportional to the amount of inorganic phosphate present in the phase in question, since the inor- ganic P, among all the P compounds present in the yolk and white, is practically the only source of activity; in the 6 days experiment, for TABLE 6. — DISTRIBUTION OF INJECTED AcTIVE PHOSPHATE BETWEEN DIFFERENT PARTS OF THE EGG | pine oe Fraction % activity incubation | 5 6 days | White | 61.6 Yolk | 10.3 Liquids l 26.00 Embryo | eA 16 days White | 14.9 Yolk | 17 Liquids | 19.8 Embryo 63.0 example, 10% of the 10.3% activity found in the yolk was presentas inorganic P. Finally we have to envisage the possibility that a part of the inorganic phosphate injected is not freely movable in the white—it might be precipitated as calcium phosphate or attached to proteins, its mobility being lowered thereby. We have also carried out experiments in which 0.1 ml. physiological NaCl solution containing a negligible amount of labelled sodium phosphate was injected into eggs which were not incubated. After the lapse of 5 days the distribution of the activity in different parts of the egg was determined; 97% was found in the white and 3% in the yolk. As was of course to be expected, a still greater preference for the white was shown by the active phosphorus in this experiment; the duration of the experiment was shorter than that of those discussed above, and transport of phosphorus from the white to the embryo was absent. To test whether the water injected encountered any hindrance in its propagation through the egg, we injected 0.2 ml. heavy water into the white of the egg; after the lapse of 5 days water was distilled separa- tely from the white and from the yolk and the densities determined. We are much indebted to Mr O. Jacopson for carrying out the density determinations using Linderstrom—Lang’s float method. He found that the water prepared from the white had a density exceeding that of normal water by 484 parts per million, while the corresponding figure for the water obtained from the yolk was 437. The deuterium content of the water distilled off from yolk was found to be only about 10% lower 300 ADVENTURES IN RADIOISOTOPE RESEARCH than that of the water from the white, showing that in the course of 5 days the water injected was very nearly evenly distributed throughout the egg, in contrast to the injected active phosphate. The anomalous behaviour of the latter, while of interest in the study of the circulation of phosphate ions in white and yolk, in no way influences the investigat- ion of the main problem discussed in this paper—namely, if and to what extent the molecules of the different phosphorus compounds present in the embryo are built up there or are drawn, from the yolk. Introduction of labelled hexosemonophosphate into the egg to be incubated In one set of experiments, instead of following up the fate of labelled inorganic P in incubated eggs, we introduced radioactive hexosemono- phosphate. Prof. Parnas very kindly presented us with this compound (prepared by Dr Ostern) in the form of barium hexosemonophosphate, from which, by treatment with sodium sulphate in the cold, the sodium compoundof the ester was obtained. 0.2 ml., containing about 0.2 mgm P as hexosemonophosphate salt and about 3 mgm. sodium sulphate, was injected into the white of each of the eggs to be incubated; to avoid decomposition of the ester, the solution was kept ice-cooled until it was injected into the egg. Of the 10 eggs receiving this treatment, only two supplied living embryos. After a lapse of 14 days, 7.7% of the activity injected was found to have been incorporated in the embryo (5.8% in the yolk) and a large fraction was also to be found in the white and in the connecting liquids. If, of the various fractions extracted from the embryo, we had only found activity in the fraction containing hexose- monophosphate, we should have had to conclude that the hexosemono- phosphate does not decompose in the egg but enters the embryo as such. In view of the results obtained in the experiments carried out with labelled inorganic phosphate, however, such behaviour was hardly to be expected. Furthermore, Kay [1926] found that in the embryo the phosphatase activity of the developing bone was extremely high, the phosphatase decomposing the hexosemonophosphate. We isolated the hexosemonophosphate from the embryo, as described on p. 295, and compared the specific activity of this fraction with that of the inorganic phosphate (-+creatine-P). We also isolated the phosphatide fraction and the residual phosphorus fraction containing mainly nucleo- protein-P. As Table 7 shows, no conspicuous difference can be seen between the specific activities of the different fractions of the embryo, with the possible exception of the residual P. In these experiments small activities had to be measured and the differences found between the first three fractions lie within the errors of the experiment. The results obtained suggest the explanation that active inorganic P splits off ORIGIN OF THE PHOSPHORUS COMPOUNDS IN EMBRYO OF CHICKEN 301 from the labelled hexomonophosphate injected and is incorporated in the different phosphorus compounds of the embryo, while the hexomono- phosphate molecules extracted from the embryo are not those syn- thesized by Dr OstrrNn but are molecules built up by the chicken’s embryo. TasBLeE 7. — Speciric Actriviry oF P FROM DIFFERENT FRACTIONS FROM Two Eaas INCUBATED FOR 14 Days AFTER THE INJECTION OF RaAapIoacTIVE HrExXoSEMONO PHOSPHATE: (Speciric Acriviry oF P ExtrrRAcTED FROM THE WHITE TakEN As 100) Fraction | Specifie activity Idrenlomyog Ibarordernme IB soogacoosaccoogGc0C | 24 Hexosemonophosphate-P ........ | 26 IPhosph atid CPi vary salelerereialore evolves | 20 Residual (‘‘nucleoprotein”’) P .... | 11 Yolk: IfaxonyegMane IP Goo angecopousodgco0K 36 Hexosemonophosphate-P ........ | 18 Phosphatide + residual P ....... | 0 The low value found for the residual P of the embryo may possibly be due to the building up of a part of the nucleoprotein fraction at an early date before much of the active hexosemonophosphate introduced has decomposed. The phosphatide-P and residual P extracted from the yolk were found to be inactive. These fractions were found to be only slightly active even after the injection of strongly active inorganic P, and the absence of activity after the injection into the egg of the much weaker hexosemonophosphate was only to be expected. The hexose- monophosphate fraction isolated from the yolk had a specific activity of 18; the inorganic P, 36. The larger value found for the specific activity of the inorganic P is possibly to be explained in the following way. Some active hexosemonophosphate diffuses into the yolk and partly decomposes into active inorganic P: this is the source of most of the active inorganic P which we isolated from the yolk. The hexosemono- phosphate, isolated by the method outlined on p. 296, contains, besides the active hexosemonophosphate, some non-active hexosemonophos- phate and possibly also some other acid-soluble P compound separated simultaneously, which diminished the specific activity of the ‘hexose monophosphate”’ fraction isolated from the yolk. In the embryo, on account of the strong enzymic action prevailing there, all phosphorus- compounds become labelled; on the other hand, in the yolk, as we have just mentioned, no such labelling takes place. 302 ADVENTURES IN RADIOISOTOPE RESEARCH On the phosphatide synthesis in the embryo of the chicken We saw that the phosphatide molecules present in the chicken’s embryo are not identical with those formerly located in the yolk, but that they were synthesized in the embryo. The work of ScH6NHEDTER and RiTTENBERG [1936] gives us important information about the units which are utilized in the synthesis. They found, by making use of deute- rium as an indicator, that the developing hen’s egg forms no new fatty acids and their result excluded also the possibility that unsaturated fatty acids present in the egg had been hydrogenated during develop- ment. NEEDHAM [1931], on the other hand, found that a marked desatu- ration occurs in an aqueous emulsion of embryonic tissues mixed with the corresponding yolk and vigorously shaken. The embryo must thus make use of the fatty acids present in the yolk to build up its phosphati- des: in doing this it possibly gives some preference to the less saturated fatty acids. The fatty acid components of the phosphatides extracted from the embryo are found to be less saturated than those extracted from the yolk residue. This, at first sight puzzling fact that the embryo, instead of using the phosphatide molecules found in great abundance in the yolk synthesizes its own phosphatide molecules, becomes less puzzling when we envisage the likely possibility that the synthesis of phosphatide molecules is a step in other chemical processes which occur simultaneously in the growing embryo. Summary Radioactive sodium phosphate was injected into hen’s eggs which were then incubated in some experiments for 6, and in others for 11, 16 and 18 days. While the phosphatide-phosphorus extracted from the embryo always showed a high specific activity (activity per mgm P), that extracted from the yolk was hardly active at all. The phosphatide molecules present in the embryo could not there- fore have been taken from the yolk only, but must have been synthesized in the embryo. The investigation of the ’’acid-soluble”’ and residual (mainly nucleoprotein) phosphorus extracted from the embryo led to a similar result—namely, that the ratio in which the labelled inorganic phosphorus atoms are incorporated into the different phosphorus compounds present in the embryo is governed solely by probability considerations. Practically all the phosphorus atoms present in the various compounds of the embryo must pass through the stage of inorganic P; only the inorganic phosphorus present in the embryo is taken as such from the yolk or the white. In some experiments, instead of radioactive sodium phosphate, labelled hexo- semonphosphate was injected into the egg before incubation. The hexosemono- phosphate-phosphorus extracted from the embryo had about the same specific activity as the inorganic and the phosphatide phosporus extracted. This result suggests that inorganic phosphate radicals which were split off from the hexo- semonphosphate and from other compounds present in the yolk and the white, rather than the hexosemonophosphate molecules introduced into the latter, are utilized to build up the phosphorus compounds of the chicken’s embryo. ORIGIN OF THE PHOSPHORUS COMPOUNDS IN EMBRYO OF CHICKEN 303 References Hevesy and Haun (1938) Kgl. danske vidensk. Selskab. Biol. Medd. 14, 1. Kay (1926) Brit. J. exp. Path. 7, 177. KuGier (1936) Amer. J. Physiol. 115, 287. LEHMANN and NegepuHaAm (1937) J. exp. Biol. 14, 483. NrEDHAM (1931) Chemical Embryology, Vol. Il, p. 1171. Univerity Press, Cam- bridge. Neepuam, Nowrnski, Drxon and Coox (1937) Biochem. J. 31, 11. OstERN, GuTHKE and TERSZzAKOWEC (1936) Hoppe-Seyl. Z. 243, 9. PruwmMer and Scort (1909) J. Physiol. 38, 247. ScHONHEIMER and RirreENBERG (1936) J. Biol. Chem. 114, 381. Originally published in Nature 142, 111 (1938) 33. FORMATION OF MILK A. H. W. ATEN and G. Hrevisy From the Institute of Theoretical Physics, University of Copenhagen We have administered labelled (radioactive) sodium phosphate to goats and investigated to what extent phosphorus present in different com- pounds extracted from the blood and the milk became labelled. In two cases the goat was killed after the experiment and the phosphorus compounds present in the organs investigated as well. Some of the results obtained are seen in the accompanying table. Activity per mgm P in milk. (Activity of | Activity per mgm phosphatide P extracted plasma inorg. P after 4% hours taken as 1) | from milk and organs, after 44% hours Interval after | | aan | sae the start of the | Fraction a cureliy Dek Fraction oulby sae | | mgm P | per mgm P ale ee | ee a 4 = =a ae | inorg. iP 0.68 Mal 2 ipia ceitas oueress | 0.09 0—2 hr. | Casein P | 0.54 | Plasma......... 0.02 | Hster) P | 0.32 Corpuscles <2 =. 0.01 | | Milk gland ..... | 0.13 WtnorgssE | 179° | Liver 2005.55 2- | 0.09 2—4% hr. | Casein P =| 1.71 | Kidney ......... 0.11 ster (116) ee | ‘ec ae Activity per mgm P of mail TREE NA cero. |) Coatsin 12 | = eee ester P accumulated in 0—3 Witeeree |. .082. uN eu | Imores 2 | 0.49 lebyebxoll, Y iar S6occac 0.76 2326 hr. | Casein P | 0.55 | Hydrol. 60 min ...... 0.68 | Ester P | 049 Remaining fraction .... 0.34 | | : _ mad: - While, mayer, IE Gaoogpac 1.48 Inorganic phosphorus The inorganic phosphorus extracted from the milk produced in the first two hours after the subcutaneous injection of the labelled phos- phorus, shows considerable radioactivity. Sbould the milk contain FORMATION OF MILK 305 only those inorganic phosphorus atoms which were located in the plasma at some time after the start of the experiment, the Specific activity of the milk inorganic phosphorus should be as high as that of the plasma inorganic phosphorus. In making such a comparison, it must be borne in mind that the specific activity of the plasma inorganic phosphorus rapidly decreases with increasing time through interaction of plasma phosphate phosphorus with that of bone and other tissue. No definite conclusion can therefore be drawn from comparing a single value of the specific activity of plasma and milk phosphorus. By following up, however, the change of the specific activity of the plasma inorganic phosphorus and milk inorganic phosphorus with time, we find that it takes 3—4 hours for the milk inorganic phosphorus to be almost entirely composed of individual atoms which had been present in the plasma after the start of the experiment. In milk produced shortly after the start of the experiment, a large part of the phosphorus atoms present were those which were located in the milk gland when the labelled phosphorus was administered. The replacement of the gland inorganic phosphorus by plasma inorganic phosphorus is thus comparatively slow because of a slow rate of penetra- tion of the phosphate ions through the cell walls. Heavy water, on the other hand, injected simultaneously with the labelled phosphate was already, after a short time, equally distributed between plasma and milk, because of the low resistance water molecules encounter when penetrating through cell walls. Casein phosphorus The comparatively high specific activity of the casein phosphorus is only compatible with the assumption that the phosphorus atoms utili- zed in the synthesis of the casein in the milk gland are drawn from the inorganic phosphorus of the plasma. From the difference in the rates at which the active casein phosphorus and the active inorganic phos- phorus present in the milk are formed, the time of formation of the casein in the gland cells can be estimated to be about 1 hour. Ester phosphorus The rate of formation in the milk gland of the average labelled phos- phorus ester molecule is lower than that of the average casein molecule (cf. table). 1144 hours after the administration of radioactive hexose- monophosphate (kindly presented to us by Prof. Parnas) injected into the veins of the goat, an appreciable amount of labelled ester was found in the milk, while another larger part of the activity was found in the inorganic milk phosphate. This result shows that a rapid enzymatic 20 Hevesy 306 ADVENTURES IN RADIOISOTOPE RESEARCH breakdown of the hexosemonophosphate and rebuilding of ester mole- cules takes place in the gland. The milk gland contains thus enzymes having the same action on hexosemonophosphate as Ropison and Kray’s? bone extracts; however, the bulk of the esters present in the milk are acted on by enzymes present in the gland at a much slower rate. Similar behaviour is shown by the mixture of phosphorus esters present in the blood?. Phosphatide phosphorus The formation of active phosphatide molecules is, as seen from the table, a slow process. The individual phosphatide molecules present in the milk were mainly built up in the milk gland and not taken up as such from the plasma (as is the case with the yolk phosphatide). This follows from the fact that the specific activity of the phosphatide phos- phorus extracted from the milk gland and also from the milk itself is higher than that secured from the phosphatide of the plasma. The view is often encountered that the milk fat originates from the plasma phosphatides which decompose in the milk gland, supplying fat and inorganic phosphorus. This view is entirely incompatible with the results obtained by us. To mention only one argument, we find the phosphatide phosphorus of the milk to be slightly, the inorganic phosphorus present to be strongly, active. The latter can therefore only originate from the highly active inorganic phosphorus of the plasma. It is well known that different milk fractions, secured consecutively within a short time, have a markedly different fat content. As we find? that the inorganic phosphorus extracted from these fractions has a different specific activity, we have to conclude that these fractions cannot originate from an initially homogeneous liquid. So we arrive at the result that some of the milk gland cells give off milk much more readily than others, but that some even of the first-mentioned cells retain a large part of their solid milk constituents, particularly the phosphatides (and fats). Not only are phosphorus compounds present in the milk not formed during the act of milking, as often assumed, but such compounds contained in the last fraction secured during the act of milking are partly of earlier date than those present in the im- mediately preceding milk samples. References L. Haun and G. Hevesy Nature i40, 1059 (1937). R. Rosrson The Significance of Phosphoric Esters in Metabolism (New York, 1932). 1A detailed account of the experimental results obtained will be found in the dis- sertation of A. W. ArEen, jun., to be presented to the University of Utrecht. 307 COMMENT ON PAPERS 27—33 In 1935, after mailing paper 16 to Nature we embarked on the study whether and to what extent the constituents of the brain are renewed during adult life. After the administration of 32P to rats an appreciable incorporation of the tracer into brain phosphatides was observed after the lapse of 1 hr or more (paper 27). These results were published simultaneously with those of ArTom et al. (1937) and of CHaAIKOFF et al. (1937) who demonstrated the incorporation of *P into the phosphatides present in a great number of organs. We subsequently concen- trated our interest on the origin of phosphatides. Among the most fascinating applications of isotopic tracers ranges the study of the origin of body constituents. With LunpsGaarp we fed dogs with oil containing labelled sodium phos- phate in order to find out whether an appreciable part of the increased lecithin content in the blood is built up in the intestine was labelled (paper 28). These experiments showed that intestinal mucosa is not the chief place of synthesis of plasma phosphatides. It was the results of perfusion experiments (paper 29) which first indicated that the liver releases labelled phosphatides to the circula- tion. In other experiments (paper 30) not the removal of the labelled phospha- tides-from the liver but the uptake of these from the circulation by the liver was followed. These were the first experiments in which blood containing in vivo syn- thetized labelled compounds was transfused. They led to the result that not only is the rate of turnover of phosphatides in the liver very high, but the exchange of phosphatide molecules between the liver cells and the plasma takes place at a much higher rate than the corresponding process between other organs and the circulation. The liver was found to be the main source of formation of plasma phosphatides. This was most spectacularly demonstrated by CHarkorr and his group (1946) who in the course of their very extensive and important studies on phosphatide metabolism have shown that ina hepatectomized dog, after administ- ration of labelled phosphate, the formation of labelled plasma phosphatides prac- tically ceases. In animals almost devoid of the higher unsaturated acids there is no diminution in the phospkatide turnover in the liver. An enhanced turnover rate is observed in the muscles of fat-starved rats (Hevesy and SmepLEy-MAcLEAN, 1940). That the chick builds up its own phosphatides and does not avail itself of the phosphatides in the yolk could be concluded from the following observation. After injecting labelled phosphate into the fertilized egg, the phosphatides extrac- ted from various tissues of the chick were strongly radioactive while the yolk phosphatides remained inactive (paper 32). That the phosphatide molecules of the milk in contrast to those of the yolk do not originate in the blood plasma but the former are built up in the milk gland could easily be proved (paper 33). The specific activity of the milk phosphatide phosphorus was found to amount to almost four times the specific activity of the plasma phosphatide phosphorus, but was lower than the corresponding value of the milk gland phosphatide phosphorus. The experiments were carried out at a time following administration of labelled sodium phosphate to the hen in which the activity of the plasma phosphatides was still increasing. In such experiments a milk phosphatide phosphorus specific activity which is higher than the corres- ponding value of the plasma phosphatides, excludes the plasma phosphatides 20* 308 ADVENTURES IN RADIOISOTOPE RESEARCH as a main source of milk phosphatides. A detailed description of these and a great number of additional experiments are described in the D. Sc. thesis of A. H. W. Aten Jr. (1939). References C. Arrom, C. Perrrer, M. SAanTANGELLO and E. Srcre (1934) Nature 139, 836. I. PertMAN, S. RuBEN and I. L. CHarkorr (1934) J. Biol. Chem. 122, 169. A. H. W. ATEN Jr. (1939) Isotopes and Formation of Milk and Egg. Dissertation, Utrecht. C. Enrenman, I. L. Coarkorr and D. B. Zinversmrr (1946) J. Biol. Chem. 160, 5. G. Hevesy and I. Smeptey-Macriean (1940) Biochem J. 34, 903 . Originally published in Kgl. Danske Videnskabernes Selskab. Biologiske Meddelelser. 15, 5 (1940) 34. TURNOVER OF LECITHIN, CEPHALIN, AND SPHINGOMYELIN G. Hevesy anp L. HAHN From the Institute of Theoretical Physics and Institute of Physical Chemistry, University of Copenhagen PHOSPHATIDE molecules present in the body have been taken up with the food or have been built up in the organism. A spectacular proof of the synthesis of phosphatides in the body is given by the fact that ducks raised in diets containing phosphorus only in inorganic form laid 85—195 eges during the summer™. These eggs contained 200—400 gm phosphatides (corresponding to 8—16 gm phosphatide P), and this very appreciable amount was synthesised by the organs of the ducks. On the other hand, phosphatides can enter the circulation from the intestine. The amount of phosphatide which is daily led by the intestinal lymph into the circulation of the rabbit®) on normal diet was calculated to be about 50 mgm. This is only about 1/5 of the amount daily synthesised in the liver (comp. p. 323); one must further consider that an appreciable part of the above mentioned 50 mgm was synthesised in the mucosa of the small intestine. Thus, the phosphatide molecules of the organs will be only to a small extent obtained directly from the food, the overwhelming majority being built up in the body. CONCEPT OF TURNOVER The ultimate aim of the investigation of the origin of the phosphatide molecules present in the body is to be able to state in which form the hydrogen, carbon, nitrogen, oxygen, and phosphorus atoms present in the phosphatide molecules were taken up by the body and in what steps they were involved until ultimately incorporated into phosphatide molecules. This exacting task can hardly be solved at present, and we must content ourselves with the determination of the place and rate of formation of the phosphatide molecules in the body from glycerol, Q) G@. FINGERLING, Biochem. Z. 38, 448 (1911). (?) H. StirumMann and W. WitBrRANpDT, Biochem. Z. 270, 52 (1934). 310 ADVENTURES IN RADIOISOTOPE RESEARCH fatty acid, choline (or another organic base), and phosphate. We will denote, in what follows, as turnover rate the rate of synthesis of phos- phatide molecules from inorganic phosphate and other components independent of the actual mechanisms involved, and we shall measure this rate by determining the extent to which labelled phosphate present in the cells of an organ is incorporated into these newly formed phos- phatide molecules. As the phosphatide content of an organ is usually constant, we can follow that with the formation of new phosphatide molecules the decomposition of an equal or similar number of old mole- cules goes hand in hand. The possibility must also be envisaged that new formation and decomposition of phosphatides do not take place in the same organ, but that the newly formed molecules are synthesised in one organ and carried into the other by the circulation. The turnover rate can also be measured by following the rate of incor- poration of fatty acids or of choline, for example, into the phosphatide molecule. The turnover rates measured by using different indicators need not necessarily be identical. It would be conceivable, for example, that the incorporation of the phosphate radical into the phosphatide molecules would be preceded by the formation of glycerophosphate and that this process would be a comparatively slow one in contrast to all other steps involved in the synthesis of the phosphatide molecule. In this case, the turnover rate measured, using labelled P as an indi- cator, would be slower than that found when using labelled fatty acids or labelled choline. The opposite would be the case if the reorganisation of the phosphate bond were to take place at a faster rate than the corre- sponding release and incorporation of fatty acids or choline into the phosphatide molecules. The question if and to what extent the rate of phosphate incorporation into the phosphatide molecule differs, for example, from that of the fatty acid incorporation into the latter cannot be answered at the time being. Feeding cats with mixed glyceride, the acids of which were composed to 85 per cent of elaidic acid, Stncuark™ found 12 hours later the plasma phosphatide fatty acids to contain 19 per cent of elaidic acid. In our experiments we found (comp. p. 326) that, after the lapse of 16 hours, about 4 per cent of the phosphatides extracted from the plasma of rabbits contained labelled phosphate. Q) R. G. Srncuair, J. Biol. Chem. 115, 215 (1937). TURNOVER OF LECITHIN, CEPHALIN AND SPHINGOMYELIN 311 INDICATORS APPLIED IN TURNOVER MEASUREMENTS a) Change of the degree of unsaturation of fatty acids Since the phosphatides contain both saturated and unsaturated fatty acids, the change of the composition of the fatty acids of the organ phosphatides after ingestion of cod liver oil, for example, can be utilised to get information on the rate of the phosphatide turnover in the organ in question. A change in the iodine number of the phospholipids extracted from the liver of dogs” and cats® after the ingestion of cod liver oil and the disappearance of the changes within 24 hours and 2 to 3 days, re- spectively, was observed at an early date. b) Incorporation of iodized fatty acids into the phosphatide molecule Iodized fatty acids, whether injected intravenously or given by mouth. enter the phosphatides of the liver, the blood®’, and the milk for example, c) Incorporation of elaidic acid into the phosphatide molecule This method was repeatedly used in the investigation of the turnover of phosphatides. The rate of entrance of elaidic acid into and disappearance from the phosphatides was found to be rapid in the liver and the intesti- nal mucosa and comparatively slow in the muscle. The process was found to be essentially complete in the liver within a day, but in the muscle only after the period of many days®. d) Incorporation of fatty acids, labelled by introduction of heavy hydrogen, into the phosphatide molecule Linseed oil was deuterated and the heavy” fat obtained fed to rats. The investigation of the deuterium content of the phosphatides extracted from different organs gives information on the phosphatide turnover in the organ in question‘®. (2) @. Ioannowics and E. P. Prick, Wien. Klin. Wochenschr. 23, 573 (1910). @ R. G. Srycuarr, J. Biol. Chem. 82, 117 (1929). Comp. also R. G. Srxcrarre, Phys. Rev. 14, 351 (1934). (3) ¢. A. Arrom, Arch. int. Physiol. 36,191 (1933); C. A. Arrom and G. Pererrt, Arch. int. Physiol. 36, 351 (1933). (@) F, X. Aytwarp, J. H. Buackwoop and J. A. B. Smirx, Biochem. J. 31, 130 (1937). ()) R. Srnciarr, J. Biol. Chem. 111,270 (1935); 121, 161 (1937), M. F. Koun, J. Biol. Chem. 126, 709 (1938). (6) B. Cavanacu and H. S. Rarer, Biochem. J. 33, 17 (1939). 312 ADVENTURES IN RADIOISOTOPE RESEARCH e) Incorporation of analogues of choline, in which arsenic replaces nitrogen, into the phosphatide molecule Arsenic can be detected in the lecithin fraction isolated from the liver and the brain of rats kept for 21 days on a diet containing arseno- choline chloride®, f) Incorporation of labelled phosphate into the phosphatide molecule This method will be discussed in detail. Most of the methods outlined above were successfully applied to show that a marked turnover takes place in some of the organs, and the appli- cation of the methods a), c), and f) lead to the result that the rate of the phosphatide turnover is much faster in the intestinal mucosa and in the liver than in the other organs. None but the ’’phosphate method”’ was applied, however, to arrive at quantitative data as to the rate of rejuvenation of the phosphatide molecules present in the different organs. QUANTITATIVE DETERMINATION OF THE TURNOVER RATE BY USING LABELLED PHOSPHATE The formation of phosphatide molecules containing *2P inside the tissue cell can only take place when the process of phosphatide formation was preceded by a penetration of ®?P into the cell, and the same applies to all indicators used in turnover experiments. This point was hitherto not considered. Its great importance is best seen by the following. To arrive at a proper figure for the turnover rate we have to compare the percentage of *2P in the total inorganic P of the cells with the per- centage of 32P in the total phosphatide P extracted from them. If these ratios, which correspond to those of the specific activities of the inorganic P and the phosphatide P, are found to be equal, we can conclude that all phosphatide molecules were renewed during the experiment. In this case, a proportional partition of 32P between the inorganic P and the phosphatide P present in the cells took place. This is only possible if the phosphate radical of all the phosphatide molecules was split off in the course of the experiment, a process which was then followed by a synthesis of phosphatide molecules with incorporation of other phosphate radicals in which #2PO, was represented proportionally to its total number present. If the specific activity of the phosphatide P is found, @) A. Wetcu, Proc. Soc. Exp. Biol. Med. 35, 107 (1937). TURNOVER OF LECITHIN, CEPHALIN AND SPHINGOMYELIN ols to be, for example, 10 per cent of that of the inorganic P, we can conclude that 10 per cent of the phosphatides were renewed during the experi- ment. Due regard must, however, be given to the change of the specific activity of the cellular inorganic P in the course of the experiment. By administering the labelled phosphate in several portions of suitably varying quantities in the course of the experiment, we can maintain a constant specific activity of plasma and interspace phosphate. As to the cellular concentration of °2P, which is nought at the start of the experiment and then gradually increases, we determine the change of concentration with time experimentally and compare the specific acti- vity of the phosphatide P extracted at the end of the experiment with the average value of the specific activity of the inorganic P which pre- vailed during the experiment. When determining the specific activity of the cellular inorganic P, due regard must be taken to the fact that a part of the tissue inorganic 32P is of extracellular origin. As the extracellular volume of the tissue is known and the specific activity of the extracellular P does not differ much from that of the plasma P, we can easily correct for the presence of the extracellular P in the tissue inorganic P. Since the extracellular phosphate in the case of the muscle tissue, for example, amounts to only about 1/90 of the cellular inorganic P, the correction mentioned above becomes only significant in experiments of short duration. If the rate of penetration of the inorganic phosphate differs greatly in the cells of different tissues, as it actually does, for example, in the case of the liver and the muscle, we do not get proper information on the relative rate of turnover of the phosphatides in these organs by comparing the specific activity of the liver phosphatide P with that of the muscle phosphatide P. Conclusions based on such a comparison will greatly underestimate the relative rate of phosphatide turnover going on in the muscle cells into which the inorganic P diffuses at a slow rate, in contrast to its penetration into the liver cells. We will arrive, however, at correct figures by comparing the ratio. specific activity muscle phosphatide P specific activity muscle inorganic P with the corresponding ratio of liver products. If we wish to draw quantitative conclusions from experiments carried out with elaidic acid as an indicator, we have to compare the elaidic acid content of the organ phosphatides with that of the elaidie acid content of the fatty acid mixture present in the corresponding cells in freely dispersed state. The latter magnitude is not known and the same consideration applies to the work with deuterated fat as an indicator. 314 ADVENTURES IN RADIOISOTOPE RESEARCH We may get some, though very restricted, information by comparing the heavy hydrogen (D) content of the organ phosphatides with that of the organ glycerides. After the lapse of 10 hours, the ratio liver phosphatide D __ kidney phosphatide D liver glyceride D ~ kidney glyceride D where D denotes the relative heavy content of the total “‘non-exchange- able” hydrogen, was found to be 1: 2. EXPERIMENTAL PROCEDURE The labelled phosphate of negligible weight, dissolved in physiological sodium chloride solution, was injected into the vena jugularis of the rabbit drop by drop during the experiment. Per hour 2.5 cc. were injected; the experiment took usually 06 ©) he) Specific activity 20 50 100) 150 200 250 min. Fic. 1. Change of the specific activity of the plasma inorganic P during continuous intravenous injection of labelled phos- phate to a rabbit. (Specific activity = per cent of the labelled P injected, found in 1 mgm. P). 4 hours. By taking small samples from the ear vein at different intervals, the change in the activity of the plasma was followed. In several cases, we extracted the inorganic P of the plasma and measured its specific activity (activity per mgm P), in others we contented ourselves with the measurement of the total activity of the plasma which, in experiments of short duration, is solely due to the inorga- nic phosphate present. The labelled P was injected drop by drop into the vena jugularis in order to obtain a comparatively small and easily accountable change in the activity level of the plasma (see Fig. 1). If all the labelled P is injected at the start of the experi- ment, as in our early experiments and in all experiments carried out by other workers with labelled P, the activity level of the plasma is very high at the begin- ning, and it is slow at the end of the experiment (see Fig. 2). If the labelled P is given by subcutaneous injection or by mouth, the activity of the plasma first increases with time and later decreases (see Fig. 3). The sensitiveness of the radio- Per cent of labelled P injected +2 +] Log of labelied P injected TURNOVER OF LECITHIN, CEPHALIN AND SPHINGOMYELIN Fig. 2. Change of the specific activity of the plasma inorganic P after sub- cutaneous injection of labelled phos- phate to a rabbit. 20 60 100 200 300 Fig. 3. Change of the logarithm of the labelled P content of the plas- ma with time after intravenous in- jection of labelled phosphate. 50 100 150 200 Hours 315 316 ADVENTURES IN RADIOISOTOPE RESEARCH active indicator, thus, changes very appreciably in the course of the experiment. If we are successful in keeping the activity levei of the plasma constant during the experiment, we can eliminate great difficulties otherwise encountered when calculating the turnover rate of organic phosphorus compounds. The changes in the activity of the plasma, shown in Fig. 1, can be further reduced by injecting amounts decreasing with time. In our later experiments we have chosen this procedure and varying amounts of labelled P were adminis- 0,15 Oo (oe) Specific activity 0,05 50 100 150 200 250 min. Fig. 4. Change of the specific activity of the plasma inorganic P during continuous subcutaneous. injection of labelled phosphate to a rabbit. (Specific activity = per cent of the labelled P injected, found in 1 mgmP). tered by subcutaneous injection. In an experiment taking 12 hours, for example, labelled P was injected every 20 min. In experiments taking several weeks, in the later phases of the experiment injections were made twice a day. The change in the plasma activity in such an experiment taking 4 hours is seen in Fig. 4. In experiments taking several hours or days a constant activity level could be easily obtained. The determination of the turnover rate of the phosphatides present in the different organs necessitates the determination of the specific activity of the inorganic P and phosphatide P extracted from the organ. This determination was carried out in the following way. At the end of the experiment the animal was killed by bleeding. The organs were at once placed in liquid air, minced, and extracted with cold 10 per cent trichloracetic acid. The inorganic phosphate present was precipitated as ammonium magnesium phosphate at 0°. Muscle samples were taken before death. To secure the phosphatide present in the organs, these were first dried with cold acetone and then treated with ether, later with boiling alcohol. The ether-alcohol extracts were evaporated in vacuo and taken up seve- ral times with petrol-ether; the phosphatides were then converted into phosphate by wet ashing. The procedure applied when isolating lecithin, cephalin, and sphin- gomyelin will be discussed on page 330. TURNOVER OF LECITHIN, CEPHALIN AND SPHINGOMYELIN 317 CALCULATION OF THE TURNOVER RATE In most of our experiments only a minor part of the phosphatide molecules present in the organ became labelled; we can, therefore, con- sider the reaction leading to the formation of labelled phosphatides to be a one-sided one and disregard the decomposition of labelled phospha- tides during the experiment. As already mentioned on page 312, to arrive at the value of the rate of the phosphatide turnover, we have to compare the specific activity of the phosphatide P extracted from the organ at the end of the experiment with the average specific activity of the cellu- lar inorganic P found in the course of the experiment. The value of the activity of the cellular inorganic P is obtained from that of the tissue inorganic P after subtraction of the share due to the extracellular fluid. The correction to be applied for the presence of extracellular P in the tissue inorganic P is, in most cases, a small one. In the liver of the rabbit, for example, out of 30 mgm. inorganic P only about 0.6 mgm is located in the interspaces. We arrive at this figure by assuming that the inter- spaces make out® 22 per cent of the weight of the liver and the inorganic P content of the interspaces is 3 mgm per cent. The specific activity of the liver extracellular P is, after 4 hours, 2.5 times higher than the specific activity of the tissue inorganic P; correspondingly, 5 per cent of the total inorganic P activity of theliver is due to extracellular P. In the case of the muscle, we arrive by an analogous consideration at the result that 25 per cent of the activity of the tissue inorganic P is of extracellular origin. The extent of the correction to be applied increases with decreasing length of the experiment, since in experiments of short duration only a small amount of labelled P penetrates into the cells. With regard to the considerations stated above, one must recognise the possibility that some of the phosphorus which one identifies, even after the most careful experimental procedure, as inorganic P, was in fact present in the tissue in the form of very labile, not yet known, organic phosphorus compounds. Very labile P compounds of that kind, if pre- sent, would probably be in fast exchange equilibrium with the inorganic P present, and their presence would therefore not influence much the calculation given above. The labile P of adenyltriphosphoric acid comes, for example, very quickly into exchange equilibrium with the inorganic P of the tissues or the corpuscles; it is often permissible to replace the specific activity of the inorganic P by that of the above mentioned labile P. The behaviour of creatinephosphoric acid is discussed on page 325. When calculating the turnover rate of phosphatides, we must consider the average specific activity of the cellular inorganic P prevailing during the experiment. This value is obtained by determining the specific Q) J. F. Manery and B. Hastinas, J. Biol. Chem. 127, 657 (1939). 318 ADVENTURES IN RADIOISOTOPE RESEARCH activity of the tissue inorganic P and the plasma inorganic P at diffe- rent intervals. The change of the specific activity of the tissue inorganic P is seen in Table 1, that of the plasma inorganic P is discussed on page 315. TaBLeE 1. — Speciric ACTIVITY OF THE ORGAN Inorcanic P as PERCENTAGE OF THAT OF THE Prasma InorGaAnic P Organ 100 min 240 min Liver :ffataeeeeeeeee [i erat Fh) ae ae oa Mruscles: “cjapeiveneeieveters «aie 0.8(Q) 4.6 Intestinal mucosa ..... 14.7 42.8 Brain()ix teed tak | 0.32 | 1.4 NGChiVeS GaccooKgoousGe 85 90 © In spite of all precautions taken, some creatine P may have been split off before the extraction of the inorganic P. The creatine P being, in experiments of short duration, less active than the inorganic P, such a ge composition may partly be responsible for the low value obtained in the experiment taking 100 min only. 2) Comp. p. 336 It is of interest to remark that, in the case of the kidneys, after the lapse of 100 min an almost proportional partition of 32P between plasma and cellular Pis reached. When investigating, after 4 hours, the inorganic P of the marrow of the kidney, which makes out only a minor part of the total inorganic P of the kidney, the specific activity was found to be only 48 per cent of that of the plasma. CELLULAR AND EXTRA-CELLULAR FORMATION OF PHOSPHATIDES The turnover rates recorded in the fourth column of Tables 3 to 9 are calculated on the assumption that the formation of phosphatide molecu- les takes place inside the cells with participation of cellular inorganic P. Let us assume for a moment that the formation of phosphatide molecules takes place on the cell wall facing the interspaces. Then, not the cellular but the extracellular phosphate radicals® would enter the newly formed phosphatide molecules. As the specific activity of the extracellular inorganic P is often much higher than that of the cellular inorganic P, in the last mentioned case more active P atoms would take part in the synthetic process than in the first mentioned one. A high activity of the newly formed phosphatide would then not indicate such a high @) From this view-point, it is without any significance whether the phosphate radical is directly incorporated into the phosphatide molecule or through inter- mediary stages. TURNOVER OF LECITHIN, CEPHALIN AND SPHINGOMYELIN 319 * turnover as it would if the formation of the phosphatide molecules took place with participation of the less active cellular P. It is obvious that the sensitivity of our radioactive indicator will be very different in the two cases mentioned above. Though it is much more probable that the turnover of the phosphatide molecules takes place inside the cells we have also recorded, in the fifth column of the above mentioned tables, the turnover rates calculated on the assumption of an extracellular for- mation of the phosphatide molecules. The values thus obtained give the lower limit of the turnover rate, while those obtained in column 4 give TABLE 2. — Spectric AcTIVITY OF THE INORGANIC P AND PHOSPHATIDE P ExtrraAcrep FROM THE ORGANS Rabbit I. — Weight: 2.4 kgm Intravenous injection during 4 hours SR SS A A Se | Specific activity Fraction in relative units Dncinn, aimemeiane IP ssodgcdocneqsoceaas do ctbouuoGuoUEDoSUOONT 100 Liver tissue inorganic P at the end of the experiment .......... 36.2 Liver tissue inorganic P corrected for the change in plasma activity Ghorauaver. {aXe} ~b-q eluUMSAh moo goUooO ODO DUO DONONOOOOCDOCbOUOOO UOC 44 Liver cellular inorganic P at the end of the experiment corrected as DOOWE oadwarbossdoun tobes Cor Oas Obi db domo hm Somo oA epoUcen os | 40.8 Liver cellular inorganic P average value during the experiment .. 20.4 IDINFEW FOLNOR AMM IP oSanogboous odd udcnnDOUD CODD UDC OO ODD OUOUS | 3.0 Kidney tissue inorganic P at the end of the experiment ......... 67.7 Kidney tissue inorganic P corrected for the change in plasma activity Chunar alate) Cxgqosminxyinn cognogaddagouUHUDUU Ubon oD ooo ouUnoGGodE 82.3 Kidney cellular inorganic P at the end of the experiment corrected as MOV oooocooogsc odo osodnooDS HOODOO COCO UOdaU SOND OUUDNUDOOD | 82.0 Kidney cellular inorganic P average value during the experiment | 73.5 Kadneys pliosphatices ene nee <1 rer, eh aetianls ce tae aso sinks. vio 6s | 5.5 the upper limit. It is conceivable that some of the phosphatide molecules are renewed inside the cell wall. In that case the inorganic P entering the newly formed phosphatide molecules will have a specific activity being the intermediary between that of the extracellular and the cellular P. A conti- nuous drop of the specific activity of the inorganic P in the cell wall may therefore take place while the phosphate penetrates from the inter- spaces into the cells. In the corpuscles the phosphatides are known to be practically con- centrated in the stroma, and the thickness® ofthe latter to correspond @) B. N. Erickson, H. H. Witurams, S.S. Bernstein, J. Arvin, R. L. JONES and J. G. Macy, J. Biol. Chem. 122, 515 (1938). ) DaNnteLul, J. Cell. Comp. Physiol. 7, 393 (1936). 320 ADVENTURES IN RADIOISOTOPE RESEARCH to that of a few molecular layers. It is, therefore, quite conceivable that in the outer layer of the stroma a slow rejuvenation of the phos- phatide molecules takes place with incorporation of plasma P. In the case of finding an organic P fraction extracted from the cells or the corpuscles to show a higher specific activity than the cellular, respecti- vely corpus cular inorganic P, we would be justified to conclude that the synthesis of the organic compound in question did not take place inside the cells, respectively the corpuscles. Investigations in the above men- tioned direction may bring forward results of histochemical interest. RESULTS OF EXPERIMENTS Investigation of the total petrol-ether soluble phosphatide mixture Experiments with rabbits TABLE 3. — Speciric ACTIVITY OF THE CELLULAR INORGANIC P AND PHOSPHATIDE P EXTRACTED FROM THE ORGANS Rabbit II. — Weight: 2.6 kgm Intravenous injection during 215 min | Specific activity Percentage of phosphatides | = = = - = 4 | renewed during the experi- Organ Inorganic P | Phosphatide P ment average during | at the end of the a l = the experiment experiment AQ) | Be) aes SE - Srl eee ute ee Es | . —— Liver atues amcor oe ete 100 | 19.0 19.0 3.86 Kidneys .aoseceenetescts oe | 382 | 18.3 | 4.8 20 Smallompbestimerncesiecsicmetac my 7.9 Tl 1.61 LOMA CIs erevctehede cleveds eierelsneiei eres 58 | 4.46 rier 0.91 Heart oven nat astseaee ee | Se all 153 2 0.31 PEC tesoip.om ena on doamLuaGrcot | 66.3 | 4.04 6.1 0.82 SS PLSOMA Gens erste ave shoneueicr seeneke siete | 70.2 | 3.65 | 5.2 | 0.74 Marrow UR aa tee hee | 40.8 1.63 | 4.0 | 0.33 MS VOUT Cs Fatiniid coe suchatersl ete tsispsistets | - 0.06 | — = “) Calculated on the assumption that the formation of phosphatides took place with incorporation of cellular inorganic P. : : : ) Calculated on the assumption that the formation of phosphatides took place with incorporation of extracellular inorganic P. soe 8) In several experiments the specific activity of the marrow inorganic P was found to be surprisingly low, even lower than that of the ester P. These low values were presumably due to the presence of traces of only slightly active bone P in the marrow sample. Critical remarks In Tables 2—9, data were given for the turnover rate of phosphatides in different organs of the rabbit. When calculating those values we assumed that the labelled phosphatides present in the organs were synthesised én situ. In the following, we will discuss bow far this assump- tion is justified. TURNOVER OF LECITHIN, CEPHALIN AND SPHINGOMYELIN 321 Tape 4. — Spreciric AcTIVITY OF THE CELLULAR INORGANIC P AND PHOSPHATIDE P ExtrrRAcrTED FROM THE ORGANS Rabbit ITI. — Weight: 2.3 kgm, Intravenous injection during 234 min Specific activity | Percentage of phosphatides renewed | = eas | during the experiment Organ Inorganic P Phosphatide P | average during | at the end of |~ aa = the experiment | the experiment AQ) Be) | i LIKI 4 So oso odcooobeouo oO oIDS 100 16.3 16.3 3.2 NIKO ootcndsccbcooegausKd 7.8 0.56 7.2 0.11 @) Turnover rate calculated on the assumption that the formation of phosphatides took place with incorporation of cellular inorganic P. “) Turnover rate calculated on the assumption that the formation of phosphatides took place with incorporation of extracellular inorganic P. TaBLe 5. — Sprecriric AcTIVITY OF THE CELLULAR INORGANIC P AND PHOSPHATIDE P ExtrRaActTED FROM THE ORGANS Rabbit IV. — Weight: 2.5 kgm, Intravenous injection during 215 min Specific activity | Percentage of phosphatides renewed during the experiment Organ Inorganic P Phosphatide P average during | at the end of | ~ the experiment | the experiment AQ) Be) eI yastaterenets esis less s 6:sie sche retons 100 14.8 14.8 | 2.9 ISICINEN? Gaooggcnb He mGU OO OOo | 374 23.2 6.2 | 4.6 Small intestine (mucosa)..... 107 20.0 18.7 | 3.9 | | i | a lEGtini hig ba cose gongacnooes ens 64.6 3.47 | 5.31 0.68 ruins etree sae rae oes = 763k = Tete =| 10.1 | 1e5t | | | ES TVET tosh yey oher Ore fare Sake ret Dols hee Cie — | 0.175 — ] — () Turnover rate calculated on the assumption that the formation of phosphatides took place with incorporation of cellular inorganic P. “) Turnover rate calculated on the assumption that the formation of. phosphatides took place with incorporation of extracellular inorganic P. Liver phosphatides Let us first consider the liver phosphatides. Apart from the liver, an intense turnover is going onin the intestinal mucosa, and the possibi- lity must be envisaged that the labelled phosphatides were carried into the liver from the intestine by the plasma. The plasma was found to contain only small amounts of labelled phosphatides, the specific activity of the plasma phosphatide P being, after the lapse of 4 hours, only 1/7 of that of the liver phosphatide P. This fact excludes the possi- bility that a substantial part of the labelled liver phosphatides was led from the intestine or any other organ into the liver. Large amounts 21 Hevesy o22 ADVENTURES IN RADIOISOTOPE RESEARCH TaBLe 6. — Sprciric ACTIVITY OF THE CELLULAR INORGANIC P anpd PHOSPHATIDE P ExrrRaAcTED FROM THE ORGANS Rabbit V. — Weight: 2.1 kgm Intravenous injection during 250 min | : | Specific activity Percentage of phosphatides renewed ae during the experiment | Phosphatide P Organ Inorganic P | | average during | at the end of | | the experiment | the experiment AQ) | Be) | | TANGO sls bee oe | 100 18.6 18.6 | 2.76 | | » | x KGChAGN/ ca goedUAaoouoodooeD6 | 364 22.8 | 6.3 3.58 Small intestine (mucosa) .... | 115 | 23.6 | 20.5 | 3.54 | I as | = = IMiTSCle trey. arc bot ereletaiee creer. | 12.0 | 0.87 | leo O.11 @) Turnover rate calculated on the assumption that the formation of phosphatides took place with incorporation of cellular inorganic P. (2) Turnover rate calculated on the assumption that the formation of phosphatides took place with incorporation of extracellular inorganic P. TABLE 7. — SprecrFic ACTIVITY OF THE CELLULAR INORGANIC P AND PHOSPHATIDE P ExrrRaAcTED FROM THE ORGANS Rabbit VI. — Weight: 2.6 kgm Subcutaneous injection during 255 min a a a a SS SE | | Specific activity Percentage of phosphatides renewed | > | during the experiment Organ | Inorganic P Phosphatide P | average during | at the end of =. x | the experiment | the experiment | AQ) | Be) ' os = - z Waivers Weaceuctcrereiey oleseieve siaes | 100 14.8 14.8 3.2 Corpusclesirtapesrercto circle => 29.0 1.51 5.2 0.33 ©) Turnover rate calculated on the assumption that the formation of phosphatides took place with incorporation of cellular inorganic P. ) Turnover rate calculated on the assumption that the formation of phosphatides took place with incorporation of extracellular inorganic P. of water can be led from one pond into the other by a narrow channel; salt water, however, (salt corresponding to labelled phosphatides in our case) cannot pass the channel without the water of the channel becoming salt as well. The concept of ‘‘specific activity” proves, thus, to be of great use when putting forward considerations such as those discussed above. One may say, in respect of these considerations, that, while the specific activity of the average plasma phosphatides is low, one of the phos- phatide fractions (phosphatides represent a mixture of numerous com- pounds) might be synthesised at a very fast rate in the intestinal mucosa, and the labelled molecules formed in this process might have rushed through the plasma at a fast rate into the liver without raising much TURNOVER OF LECITHIN, CEPHALIN AND SPHINGOMYELIN 323 Tasie 8. — Specirric ACTIVITY OF THE CELLULAR INORGANIC P AND PHOSPHATIDE P ExrrRacrED FROM THE ORGANS Rabbit VII. — Weight: 2.4 kgm. Subcutaneous injection during 11.5 hours ———— errr | | Specific activity Percentage of phosphatides renewed —— a a ae - | during the experiment Organ | Inorganic P Phosphatide P | | average during at) the end of |=...) | the experiment | the experiment AQ) | Be) | | | | NGI OT ps os alta) tayate ete’ aj shsen ates) sleeps | 100 25.2 25.2 14.9 CinywEOES cosocsuvcsogudner | 25d | 4.03 15.8 2.39 WIMSOES” Sig ope coeeeeeconodac | 14.7 | 1.31 | 8.9 0.78 PEST terion er cv arene ics ahohg aVieres 0? 32 | — 0.55 | — = | © ‘ | = | IMPATTOW er ciclo wis Sars exw scdas sess 36.53 | 31.8 | 87.0 | 18.8 @) Turnover rate calculated on the assumption that the formation of phosphatides took place with incorporation of cellular inorganic P. (@) Turnover rate calculated on the assumption that the formation of phosphatides took place with incorporation of extracellular inorganic P. (3) As the presence of traces only of bone P in the marrow sample investigated lowers the specific activity of the marrow inorganic P, the recorded figure for the inorganic P of the marrow may be too low and that recorded for the rate of renewal of the phosphatide P of the marrow, correspondingly, too high. TABLE 9. — Extent oF RENEWAL OF PHOSPHATIDES Rabbit IX. — Weight: 2.5 kgm. Subcutaneous injection during 50 days | Percentage of phosphatides not renewed Organ | AQ) Be) NGA OTe tr sher Totsi yo oser a tele shaves - | 0 0 IMTS Clee shah eres arta niche) staiaiciers 73 64 | IMATPOWikee sims eesverierscerer ot reais 0 0 WorpuSClesiescnt tetsicreiciciere craven 3 3 @) Rate of renewal calculated on the assumption that the formation of phosphatides took place with incorporation of cellular inorganic P. @) Rate of renewal calculated on the assumption that the formation of phosphatides took place with incorporation of extracellular inorganic P. the specific activity of the average plasma phosphatide P. As shown on page 340, the specific activity of the phosphorus present in different phosphatide fractions can differ substantially, but, in spite of exhaustive fractionation processes no fraction of extremely high or extremely low specific activity was found. Furthermore, the total amount of labelled phosphatides formed in the intestinal mucosa in the course of 4 hours amounts to only 1/5 of that formed in the liver during the same time. In this connection it is of interest to remark that, according to the results obtained by StrumMann and WinsBraNnptT which are discussed on page 309, the intestinal lymph carries up to 0.1 mgm _ phosphatide 21* o24 ADVENTURES IN RADIOISOTOPE RESEARCH P® per hour; but, even if this amount of newly formed phosphatides is quantitatively led from the intestine into the liver, it would not suffice to account for the presence of the amount of newly formed phospha- tides found in the latter which corresponds to more than 0.5 mgm phosphatide P per hour. An entirely different argument against the intestinal origin of the labelled phosphatides found in the liver is the following. The labelled phosphatides present in the plasma were not found to leave the blood stream at a very fast rate, half of the labelled phosphatides present leaving the plasma in the courseofan hour, 30 per cent of these phospha- tides being found in the liver® thus, a rapid rush of labelled phosphatides through the plasma does not take place. That the labelled phosphatides found in the liver are, at least to a large extent, formed in situ, was also shown in experiments on isolated perfused liver Such investigations were formerly®) carried out by us on isolated cat livers in which, after the lapse of 2.5 hours, the specific activity of the liver phosphatide P was found to be about 1.5 per cent of that of the liver inorganic P. A further proof that the phosphatides present in the liver were formed there was brought about by CHarKorr and his colleagues who found that, in experiments on rats, the removal of tissues very active in phospholipid turnover, namely the gastrointestinal tract and the kidneys, does not markedly influence the phospholipid turnover in the liver. Muscle phosphatides After discussing the origin of the labelled liver phosphatides we shall put forward similar arguments as to the origin of the labelled muscle phosphatides. The specific activity of the plasma phosphatides is found to be about 3 times higher after the lapse of four hours than that of the muscle phosphatides. Considerations based on the comparison of the specific activity of the plasma phosphatides and the muscle phospha- ™ When oilis fed to the rabbit, twice that amount was found to be carried by the intestinal lymph. The feeding of oil raises the rate of turnover in the intestinal mucosa and the liver as well, as shown in experiments on rats (C. Artom, G. SarzaNna and E. SeGre, Arch. Int. Physiol. 47, 245 (1938); B. A. Fries, S. RuBEN, J. PerR~tMAN and J. L. CuatKxorr, J. Biol. Chem. 123, 587 (1938) and also on isola- ted perfused cat liver, where the turnover rate was found to be about twice as high as in experiments in which non-lipemic (normal) blood was used (L. HAHN and G. Hrvrsy, Biochem. J. 32, 342 (1938). @) L. Haun and G. Hevesy, Nature 164, 72 (1939). (3) L. Haun and G. Hervesy, Biochem. J. 32, 342 (1938). (@) B. A. Fries, S. Rupen, J. PertrmMan and J.L. Cnatxorr, J. Biol. Chem. 123, 567 (1938). TURNOVER OF LECITHIN, CEPHALIN AND SPHINGOMYELIN oY) blo Or tides do not, therefore, exclude the possibility that the labelled phospha- tides present in the muscles were carried into them from other organs. This possibility is, however, excluded by the result of experiments based on the rate of entrance of labelled phosphatides into the muscles, While, in the course of 4 hours, phosphatides showing a relative activity of 0.54 units pass from the plasma into the muscles, phosphatides having an activity of 160 units were found to be present in the muscles after the lapse of the same time. In experiments of short duration the creatine P of the muscles gets only partly labelled and, therefore, a decomposition of creatinephos- phoric acid prior to the extraction of the inorganic P will lead to a “dilution”’ of the activity of the inorganic P present as such in the muscle tissue. The possibility that in our experiments, taking only a_ few hours, too low values are obtained for the specific activity of the muscle inorganic P cannot, therefore, be entirely discarded. As the extent of the new formation of the muscle phosphatides is calculated by comparing the specific activity of the phosphatide P with that of the inorganic P,a too low value of the specific activity of the inorganic P will manifestly lead to a too high value of the rate of new formation of the phosphatides. Kidney phosphatides Kidney phosphatide P is found in experiments of short duration to be more active than the phosphatide P extracted from all other organs. From this fact we may, however, not follow that the kidney phospha- tides are renewed at a faster rate than the phosphatides in the liver or the intestinal mucosa. The labelled inorganic P of the plasma diffuses with a remarkable speed into the kidney cells (see Table 1). This is in no way surprising in view of the role of the kidney cells as to excretion and re-absorption of phosphate. A result of this fast penetration of active phosphate into the kidney cells will be a formation of active phosphatide molecules already in the earliest stages of the experiment. This is not the case in the cells of such organs into which the labelled phosphate diffuses at a slower rate. Labelled phosphatides of the plasma The renewal of phosphatides in the plasma can only be determined in experiments in vitro ; in such experiments, taking 4.5 hours, the specific (0 L. Haunand G. Hevesy, Nature 144, 204 (1939); Kgl. Danske Vidensk. Selskab, Biol. Medd. 15, 6 (1940). @) L. Haun and G. Hevesy, Mem. Carlsberg 22, 190 (1937). 326 ADVENTURES IN RADIOSIOTOPE RESEARCH activity of the plasma phosphatide P was found to be smaller than 1/1000 of that of the inorganic P. In experiments in vivo, an exchange between plasma phosphatides and organ phosphatides takes place and, as in some of the organs labelled phosphatides are formed at a fast rate, we will soon after the administra- tion of labelled phosphate find labelled phosphatide molecules in the plasma, which were released from the organs. In fact, almost all phospha- tide molecules found in the plasma were synthesized in the organs. The labelled phosphatide content of the plasma, at different times, is seen in Table 10. In this experiment, the labelled inorganic P content of the plasma was kept constant during 9 days. TasBLeE 10. — Sperciric AcTIVITY OF PHOSPHATIDE P AND InorGANIC P oF THE PLASMA eee ee | Relative specific activity Time | | | Inorganic P | Phosphatide P | | ADIN OUTS it ors ofa os sss es) sna) 5 5 fo, wens 6 | 100 0.53 GW OUTSH se crac ciete stocteroreters s0. ey | 100 3.8 WF WOWGES. GooanoomsodooooudGS 100 Sal STP OUTS etn sxexegotn oe eee tae ayetences 100 | 15.0 A MOUS enero evo lenetalersie tokens owls 100 | 22.0 in) MVOTUES Ge odococcK¢o00KddDC 100 21.5 Or daysics dra frie ecuie ak oso 100s 81.6 Three consecutive processes have to precede the appearance of label- led phosphatides in the plasma. Labelled inorganic P has to diffuse into the cells of the liver and other organs in which the plasma phosphatides are formed. The building up of the labelled phosphatide molecules represents the second process, their release into the plasma the third. In view of the time taken by these processes, it is easy to understand that in the early stages of the experiment the change of the labelled phosphatide content of the plasma has a more rapid than linear depend- ence with time. Since a large part of the phosphatide molecules found in the plasma originated from the liver, it is of interest to compare the amount of the active phosphatides found in the plasma with that present in the liver at the end of the experiment. As seen in column 3 of Table 11, after the lapse of 12 hours, the acti- vity of the plasma phosphatides reached 3/4 of that of the liver phospha- tides. A large part of the liver phosphatides is, however, not yet renewed and a further substantial increase of the activity of the plasma phospha- tides can only be expected by a corresponding increase in the active phosphatide content of the liver and other organs. TURNOVER OF LECITHIN, CEPHALIN AND SPHINGOMYELIN SPAT TaBLE 11. — ActivE PHOSPHATIDE CONTENT OF THE LIVER AND THE PLASMA OF RABBITS a | Extent of partition of Duration of the Ratio of active phosphatide| labelled phosphatides experiment content of liver and plasma} between liver phosphatides | and plasma phosphatides AL InYoWbRs) boo o6 GC 94 0.16 IPA iaVoybbeS “5 oocobe 18 | 0.76 OR Gastar. eiooy 14 | 1.0 Phosphatide turnover in the corpuscles Compared with the phosphatide turnover going on in the organs, the phosphatide turnover taking place in the corpuscles is but little. This is also shown by results obtained when investigating the origin Taste 12. — Extent or Partition oF LABELLED PHOSPHATIDES, ORIGINALLY PRESENT IN THE PLASMA, BETWEEN THE PHOSPHATIDES OF THE CORPUSCLES AND OF THE PLASMA IN EXPERIMENTS 7n vitro (Plasma of a rabbit containing labelled phosphatides shaken with corpuscles of another rabbit) | Extent of par- Animal | Time in hours tition (Percen- | tage) ‘ = a a ae se | 0.5 | 1.8 ie 1.5 3.6 Rab bitmermaces secre | 3.0 | 4.0 | 4.5 5.0 | | 1.5 5 IBIS Goocdcradoacds dor | 2.0 20) | 3.0 1.5 of the yolk phosphatides®. In these experiments, 28 hours after admi- nistration of the labelled phosphate, the specific activity of the corpuscle phosphatides was found to be only 1/3 of that of the plasma phosphati- des; showing the corpuscles to be, so-to-say, a by-path of the liver and other organ phosphatides on the way through the plasma into the yolk. The labelled phosphatide molecules of the corpuscles have various origins. Some of them were incorporated in the course of the red cell () G. Hevesy and L. Haun, Kgl. Danske Vidensk. Selskab. Biol. Medd. 14, 2 (1938). 328 ADVENTURES IN RADIOISOTOPE RESEARCH formation into a tissue containing labelled phosphatides. Some of the labelled phosphatide molecules came into the corpuscles after they reached the circulation. As seen in Table 12, in which the results of some experiments iw vitro are recorded, a part of the phosphatide molecules of the corpuscles exchanges easily with those of the plasma. Presumably those situated in the outermost layer of the stroma take part in this exchange process. It is, however, rather difficult to interpret the comparatively high specific activity of the phosphatide P extracted from the corpuscles in experiments in vivo without assuming that a phosphatide turnover takes place in the corpuscles, though the rate of this turnover is small compared with that of most of the acid-soluble P compounds present in the corpuscles (see Table 14). TasBLe 13— Extent or Partition oF LABELLED PHOSPHATIDES, ORIGINALLY PRESENT IN THE PLASMA, BETWEEN THE PHOSPHATIDES OF THE CORPUSCLES AND OF THE PLASMA IN EXPERIMENTS in viv0 Extent of par- Animal Time in hours | tition (Percen- tage) 24 16 24 18 Rabbit (2—2.5 kgm) 24 17 | 25 16 | 42 34 ——_—— — ae Semel =e ———— a ore | 18 6.0 Chicken (100—150 gm) _)| 29.5 81 In experiments in vivo with rabbits (see Table 13), in the course of a day, the activity of the corpuscle phosphatide P was found to be only about 1/6 of that of the plasma phosphatide P. A still greater difference was found when investigating chickens blood. Using elaidic acid as an indicator, SrncuatR® found, 8 hours after ingestion of the elaidic acid, 15 per cent of the fatty acids extracted from the plasma phosphatides to be composed of this distinctive fatty acid, while the corpuscles contained no more than traces of the indicator. When iodised fatty acid was used as an indicator, it was found® not only in the phosphatides of the plasma but also in those of the cor- puscles. In the latter, the concentration of iodised fat was even higher (3.3 per cent of the total fatty acids) than in the former (2.0 per cent). The application of iodised fatty acids leads, thus, to a result which is in () R. G. Srncuarr, J. Biol. Chem. 115, 211 (1936). @) CQ. Arrom, Arch. Int. Physiol. 36, 101 (1933). TURNOVER OF LECITHIN, CEPHALIN AND SPHINGOMYELIN 329 contradiction to that obtained by using labelled phosphate or elaidic acid as indicators. Phosphatides containing iodised fatty acids are possibly selectively taken up by the corpuscles, another explanation being that the molecules of these compounds present in the plasma were decomposed at a faster rate than those incorporated into the stroma. Phosphatides containing iodized fatty acids represent non-physiolo- gical compounds and, as shown by the above example, the results obtai- ned by using such indicators must be interpreted very cautiously. TABLE 14. — Spreciric AcTrvITY OF PHOSPHATIDE P AnD AcIp SoLuBLE P or THE CORPUSCLES A I A Relative specific activity Fractien after 4 Ronee i 12 hours a iBhospharide: PAs cctete od cates ene eclesea= | 2.6 9.6 limos IP epqandoooboooDe Ondo ODO OOOGNS 100 100 evar} avolsy olathe) IP oo Gaccgooundo0co0KoGGODaT | 99.5 | 100 Hydrolyzed by 1 n H,SO, in 7 to 100 min. 100 | Hydrolyzed in 100 min to 12 hours ...... 100 100 INon=hyadrolyzedi: tes: s «act oo 6)s oe cies eines | 87 | | ©) The active phosphatide molecules are partly such ones which were taken up from the plasma by a exchange process. In this connection, the observation® should be also mentioned that in lactating cows during fasting a marked decrease in the concentration of plasma P lipids takes place which persists for several weeks after reali- mentation, but there is no significant change in the amount of red cell phos- phatides. This result also shows the absence of an intense interaction bet- ween plasma phosphatides and phosphatides present in the corpuscles. Agnes sie Investigation of lecithin, cephalin, and sphingomyelin We discussed above the rate of renewal of the average petrol-ether soluble phosphatide molecules; in the following, we wish to describe some experiments in which lecithin, cephalin, and sphingomyelin were separately investigated and their turnover rate determined. Chemically, cephalin differs from lecithin by containing aminoethanol instead of choline. The biological consequence of this replacement is very signifi- cant®), Cephalin is highly active in accelerating blood clotting, whereas @ J. A. Smrru, Biochem. J. 32, 1856 (1938). ©) Comp. E. Cuarcarr, M. Zire and B. M. Hoae, J. Biol. Chem. 131, 35 (1939). 330 ADVENTURES IN RADIOISOTOPE RESEARCH lecithin is not. It was evenreported™ that cephalin prepared from cattle- blood or cattle-brain enhances, while lecithin inhibits the clotting of rabbits blood. The role of the phospholipids as transport agents of fats was much discussed, this role being often ascribed to lecithin alone. In our first experiments, we determined the turnover rate of lecithin and cephalin in the organs of rabbits 4 hours after intravenous injection of labelled phosphate. We found the turnover rate of cephalin extracted from the liver, the intestinal mucosa and other organs to be pronoun- cedly faster than that of lecithin. Simultaneously, CHarcarr® found the rate of rejuvenation of cephalin extracted from the liver and the intestinal tract of rats to be slower than that of lecithin. We were first inclined to explain this difference in the findings of CHARGAFF and ourselves by the fact that the former investigated the turnover process, in contradistinction to us, in carnivorous animals. We soon found, however, that it is the duration of the experiment which is decisive for the higher or lower rate found for the cephalin turnover. We will, in what follows, first describe the experimental procedure used. Experimental procedure The tissue is dried with cold acetone and extracted first with ether and then with boiling aleohol. The second process is repeated several times. The solutions obtained were evaporated to dryness and taken up by petrol-ether in the presence of pulverised dry sodium phosphate. The latter was added in order to remove traces of active phosphate possibly present. The process was then repeated in the absence of phos- phate and the dry residue taken up in ether. The next step was to pre- cipitate the cephalin from the solution by adding 96 per cent alcohol. The filtrate obtained was evaporated and the residue containing lecithin extracted with ice-cold alcohol. This procedure was repeated and the purified lecithin obtained precipitated as chloro-cadmium-lecithin. The compound obtained was thoroughly washed with ether in order to remove traces of chloro-cadmium-cephalin possibly present. The cephalin was prepared from the alcoholic precipitate obtained in the early treatment of the phosphatide mixture. To obtain pure cephalin the precipitate was repeatedly dissolved in ether and precipi- tated with alcohol. To secure sphingomyelin the fraction insoluble in petrol-ether was collected and treated alternatively with ether and ice-cold alcohol. The last residue thus obtained was dissolved in a mixture of methyl alcohol and chloroform. By adding ether to this solution purified sphingo- WY, Oxarmura, Mitt. med. Ges. Okoyama 48, 1585 (1936). @) B. Cuarcarr, J. Biol. Chem. 128, 592 (1939). TURNOVER OF LECITHIN, CEPHALIN AND SPHINGOMYELIN 331 myelin was precipitated. A further purification of this product was obtained by repeating the procedure described above. When sufficient amounts were available, the sphingomyelfm was recrystallised from pyridin, Experiments with rabbits In the experiments, the results of which are given in Tables 15 and 16, all the labelled phosphate was administered at the start of the expe- riment. In all later experiments, labelled phosphate was administered TaBLE 15. — Specrric Activity or INorGANIC P AND or DIFFERENT PHOSPHATIDE FRACTIONS Rabbit X. — Weight: 2.9 kgm All labelled phosphate was administered at the start of the experi- ment by stomach tube. The animal was killed after 19 hours Specific activity relative to the | Plasma | Inorg. P of Fraction inorg. P | the organ found at the end of the experiment IPlasniag) Ce1uinimeesretetcretehchete sl ieters eiele 39.1 -= IbtkyGie tuavoreeFONe 12) 55> 5ooonoaGKedoUr | 89.7 100 averelecitiimesly terre siete eet ier t-vel | 46.3 | 51.6 Mivere cep malin tees. cfets: | 4.93 | 4.39 | 4.931 ©) Tf, instead of considering the known mean phosphatide specific activity during the experiment, we consider 1/, of the end value, 4.83 is obtained. of the phosphatides present in the cell nuclei. BoLuMan et al. (1948) arrived at a figure of 5.3 for the percentage renewal of the liver phosphatides in the course of 2 hours. These authors raise doubts as to the applicability to their case of the method of calculating the turn- over rate as outlined above (HEvEsy and Haun, 1940 a) and perform their calculations by applying the method worked out by ZmVERSMIT et al. (1943). That the discrepancy between the percentage turnover of the liver phosphatides of the rat as found by BoLuMAN ef al. and by the present authors is not due to a difference in the method of calculating the experimental figures obtained, is demonstrated in Table 1, in which the results of the evaluation of the experimental data TURNOVER OF PHOSPHATIDES 353 of Bor~tman et al. are calculated both according to their and our method, almost identical figures being obtained in both cases. The method of calculating the experimental results can thus not be responsible for the difference in the renewal percentage obtained by different workers and we have thus to consider other explanations. As shown in the present communication, the ratio of the specific activities of the orthophosphate P and phosphatide P of the liver varies with the age of the rat, and these variations can be assumed to be at least partly responsible for the discrepancies mentioned above), EXPERIMENTAL To each of more than 100 rats of known age, kept on normal diet, 0.1 ml physiol. NaCl solution containing *P of 2 ucurie activity and a negligible *1P content was administered by subcutaneous injection. Two hours later, batches of 4 or more rats were pooled, the animals were killed by decapitation, bled, and the isolated organs frozen with solid CO,. An aliquot of the liver, spleen and kidney samples was used to determine the specific activity of the inorganic P, another for the total P, while a third was used for the determination of the specific activity of phosphatide P. In the case of the liver, the labile P of ATP was investigated as well. All organs were cut into small pieces and average fractions were obtained. To measure the inorganic P values, fractions (0.2—0.3 gm) were extracted with cold CC],COOH. The total P was obtained by wet ashing of about 0.2 gm_ fresh tissue, while a few gm were used for the extraction of phosphatides. Before extracting phosphatides the tissue was treated with 200 ml acetone for 15 minutes. The filtrate was dried in a CO, atmosphere and the residue extracted with ether. The acetone treated tissue was extracted by grinding it in a mortar twice with 150 ml ether and once with 1:3 ether-alcohol mixture for 15 minutes. The residue was then extracted for 8 hours in a Soxhlet flask with 150 ml boiling ether-alcohol mixture (1:3). All ether and alcohol fractions were united and dried in a CO, atmosphere. To free the phosphatides from traces of inorganic P and other P compounds the residue was dissolved in 300 ml ether and shaken with 450 ml 0.1 n HCl +. 0.01 n NaCl solution in a separating funnel. This procedure was repeated four times, as suggested by Haun and Tyren (1945). The ethereal solution was evaporated in a Kjeldahl flask and then ashed by a mixture of H,SO, and HNO,. An aliquot of this solution was used for colorimet- ric determination and another precipitated as magnesium ammonium phosphate and its radioactivity determined. To determine the specific activity of the labile P of ATP about 8 gm of liver tissue were extracted with 3 volumes cold CCl,COOH solution. The cooled filtrate was neutralized to phenolphtalein with cooling by adding solid Ba(OH),. The precipitate containing adenosine triphosphate, adenosine diphosphate, orthophosphate, and some other minor fractions of organic P compounds was washed with a little Ba(OH), and neutralized with CCl,COOH. Subsequently it was dissolved in 15 ml n HNOg. To the solution, as oie by Sacks and ALTSCHULER (1942), NH,NO, was Haddad until a concentration of 5 per cent was () That the spec. activity of the liver P of mice declines from 3.7 to 1.6 when the age increases from 6 to 24 weeks was observed by FALKENHEIM (1943). 23 Hevesy 354 ADVENTURES IN RADIOISOTOPE RESEARCH obtained, followed by 2 ml 10 per cent ammonium molybdate solution. The inorga- nic P was precipitated overnight; the filtrate was then hydrolyzed for 20 minutes at 100°C and cooled. The precipitate contained the labile P of ATP. This was dissolved in 15 ml 5 per cent NH, and its P precipitated as magnesium ammonium salt. The precipitate was dissolved in 0.1 n HCl, an aliquot being used in the colorimetric essay, while another was precipitated as magnesium ammonium salt and reserved for the radioactive measurements. Percentage Turnover of Liver Phosphatides The lower limit of the percentage turnover of liver phosphatides per hour, which is calculated from the percentage ratio of the speci- fic activities of the liver phosphatide P at the end of the 2 hour-experi- ment and the mean value of the orthophosphate P during the experi- ment (which was by 10 per cent less than the end value) and divided by 2, is given in column 2 of Table 2, while column 3 contains values TaBLE 2. — Lower Limit oF THE PERCENTAGE TURNOVER PER Hour or THE PHOSPHATIDES OF THE Rat Liver. ATP P,, Scate | Percentage turnover per hour Age of rats | Without considering Considering | repeated renewal | repeated renewal Ay lictatn.c date eNO | 12a ee | gy JES) NOs clas es aaeny etek: OOO | LOA te Aig, Slee eee ect | 1002. oO | 10-52 Tel SORlnct.t Se ee eee aes | 9.6 + 1.0 101-32 1.0 MOU peceacienak eee Syoucece | Fat ae ORS | S21 0x8 Lbevviear(}) errieeg csicrelac | 5.7 + 0.6 | 6.0 + 0.6 ©) Calculated from inorganic P value. corrected for the repeated renewal of phosphatide molecules during the experiment. The correction is obtained by calculating the percentage ratio of the average specific activities of the phosphatide P and the orthophosphate P during the experiment. The average ratio of the spe- cific activity of the phosphatide P was taken to be half of the end value (cf. footnote, p. 353). The turnover rate is seen to decrease with the age of the rats, the percentage turnover rate of the phosphatides of 1.5 year-old rats being only about half of that obeserved in 4 day-old rats. In Table 3 the percentage turnover calculated by comparing the speci- fic activity of the phosphatide P with that of the inorganic P is shown. BoLuMAN et al. (1948) found the percentage turnover per hour to be about 5. From the fact that their rats weighed 200 gm we have to conclude that the animals investigated were fully grown rats for which we arrive at a corresponding figure of 8 to 6. Furthermore, we have to TURNOVER OF PHOSPHATIDES 355 consider that in contrast to the present authors, they investigated fasting rats. Pharr and Porter (1947), when comparing the turnover of phos- phatides in the liver of fed and fasting rats, found the former value to be about 1/; larger than the latter. That the diet influences the percentage renewal of liver phosphatides of the rat was also shown recently by CAMPBELL and KOSTERLITZ (1948). While we cannot calculate turnover rates from the ratio of the speci- fic activities of the phosphatide P and total liver P, this ratio indicates to what extent phosphatide P atoms are renewed compared with the average total P atoms. We listed these ratios in Table 4 along with the total P content of the livers investigated. We have not listed the inor- ganic P nor the phosphatide P contents as we were interested primarily in phosphatide P and inorganic P fractions of high purity and the ex- tended purification processes entailed an appreciable loss. The absolute amount of phosphatides renewed during a given span of time increases, in contrast to the percentage renewal, with the age of the rat, as both the weight of the liver and its phosphatide content are increasing with age. The phosphatide content of the liver of the TasBLE 3. — Lower Limit oF THE PERCENTAGE TURNOVER PER Hovur oF THE PHOSPHATIDES OF THE Rat Liver. INORGANIC P ScaLe Percentage turnover per hour Age of rats | ; — Without considering | Considering repeated renewal repeated renewal eo lege ae Re as | 102+1.0 | LOSPe Lal NOB yeryrcks. sectel =ctecs erence 9.8 + 1.0 | 10.3 + 1.0 17 Lay ear oki esa 10.4 + 1.0 10S 11 DO AG sports oAeaseieoionsiaetya Si | 7.9 + 0.8 | 8.3 + 0.8 iti een ten | 7TA006 * | 6.0 + 0.6 Taste 4. — Speciric Activiry or LiveR PHOSPHATIDES AS PERCENTAGE OF THE SPECIFIC ACTIVITY OF THE ToTAL P OF THE LIVER Inge otieate per eaneS specific Total P of liver in mgm activity p.c. 7G) ae OO 67 | 324 WIN Ch ee cossonocssnedap 64 318 leh Xolivo OG eiae cron on eCe Oo 74 363 BWiClabaers oboe SuOOr 63 348 UEC WS trot olinecco no oer 43 374 POW A cot eterekatoeeretece rave oats 58 370 356 ADVENTURES IN RADIOISOTOPE RESEARCH 90-day-old rat (83200 mgm%) is 4/, times that of the 4-day-old rat (2800 mgm%) and the weight of the Jiver increases simultaneously from 0.24 gm to 4.0 gm (LANG 1937). The phosphatide content of the 90-day-old rat is thus 20 times that of the 4-day-old rat. The percentage turnover of the 4-day-old rats being 1.5 times that of the 90-day-old animals, the amount turned over in the course of 1 hour in the 90-day-old animals is Percentage phosphatide turnover per hour + InorganicPscale O ATP, P23 scale 0 10 20 30 40 50 60 70 80 90 540 Age in days" Fra. 1. Percentage incorporation of orthophosphate P resp. ATP P, , per hour into liver phosphatides of rats of different age 15 times that renewed in the 4-day-old rats, the figures being 10.25 mgm. and 0.69 mgm, respectively. If between the phosphatide and the ortho- phosphate (or ATP) molecule a phosphorus compound were interposed which was formed at a comparatively slow rate, it would be the ultimate precursor of the phosphatide molecule ; then the above mentioned figure would represent the lower limit of the amount of phosphatides turned over only. Spleen Phosphatides Since we do not know the mean value of the orthophosphate P of the spleen during the experiment we cannot calculate the lower limit of turnover rate. Assuming the mean value to be 2/; of the end value TURNOVER OF PHOSPHATIDES 357 for the 90-day-old rat, we arrive at a lower limit of percentage turnover of 3 per hour by making use of the data given in Table 5. This is a rough estimate which indicates that the turnover rate is about 1/, of that found for the liver phosphatides. Table 5 contains data of the relative specific activities of the spleen total P and phosphatide P and the plasma inorganic P, the spleen inorganic P value being taken to be = 100. TABLE 5. — SpEeciFIc ACTIVITY OF SPLEEN PHOSPHATIDES EXPERIMENT TAKING 2 Hours | Percentage ratio of the specific activity of phosphatide P to that of Age of rats in days | Spleen inorganic Spleen total | Plasma inorganic | P P P AOE AL os kane sans 4.8 e347 4.9 1 igen cf Creme ore 5.2 ye Lf) 4.8 1 OS ae Pe CRC ae 5.0 | 18.3 | — | SOre st aate eee | 3.9 18.5 | 3.4 YO a i ee ee 3.8 ae | 2.3 HAO ALs costars oi. arehe, erers 3.4 16.0 1.6 TABLE 6. — Sprecrric ActTIviry oF KIDNEY PHOSPHATIDES EXPERIMENT Lastinc 2 Hovurs Percentage ratio of the spec. activity of phosphatide P to that of Age of rats in days - a Kidney inorganic Kidney Plasma inorganic P total P iz | OMS: Syren bake ois 14.0 | 44.0 | 10.3 Tae ee PEt y ae Re | 14.4 (Bs 10.6 3 On eae oe ee | 15.2 41.3 13.5 SOm eee fae 121 EO ae? AOD try eee tee ees | 12.1 37.8 11.6 | The percentage ratio of the specific activities of the phosphatide P and spleen inorganic P measured at the end of the experiment decreases with the age of the rat, the decrease being more pronounced if we con- sider the ratio of the percentage activities of the phosphatide P and plasma inorganic P, as seen in Table 5. This decrease indicated a decreas- ing permeability with age of the spleen cells to inorganic P, a fact previously observed by AnHtsTROM ef al. (1944) and by ANDREASEN and OTTESEN (1945). 358 ADVENTURES IN RADIOISOTOPE RESEARCH Kidney Phosphatides The ratio of the specific activities of kidney phosphatide P and inor- ganic P is appreciably higher than the corresponding spleen values, though it falls below the liver values. We found the mean value of the specific activity of the kidney inorganic P to be 1.35 times the end value and, consequently, we have to divide the figures of column 3 of Table 6, which denote data of experiments, taking 2 hours, by 21.35 in order to arrive at an estimate of the lower limit of the percentage turnover per hour. The pronounced decrease in the phosphatide turnover with age shown by the liver phosphatides is not exhibited by the kidney phosphatides. Turnover of Lecithin and Cephalin In the investigations described above the rate of incorporation of 82P into the total liver phosphatides was determined. Lecithin and cephalin are renewed at a not very different rate. PLarr and Porrrer (1947) found, for example, 6 hours after administration of labelled phosphate to full-grown rats kept in usual diet the ratio of specific activities of lecithin P and cephalin P to be 1.5, while in the fasting rat 1.3 was found. These authors state, furthermore, that administration of choline increases the turnover rate of lecithin while ethanolamine promotes the turnover of cephalin formation. In the first mentioned case a maximum increase of 33 per cent, in the latter case a maximum increase of 87 per cent of the turnover rate was observed. They inter- pret the increased formation of labelled phosphatides following admi- nistration of choline and ethanolamine, respectively, as a mass action, the assembly of the phosphatide molecule being promoted by an increase in the choline and ethanolamine concentration, respectively. In experiments on dogs, ZILVERSMIT and assoc. (1948) recently found the mean specific activity ratio of lecithin P and cephalin P to be 1.2. TABLE 7. — Sprecrric ACTIVITIES | Fraction | Specific activity? | Plasma orthophosphate Wee cic cs cle) ier=isiete) syne 0.546 ILabyxeiP> vel aNo} ol aotsolaRHKeY IE GS GuGanbboube0DooR DOC | 0.550 Giivicerophospha embassies che etter teeters | 0.376 ARG Cena pv ep crete: Monoisia)s ciccnsrstacickehens. sta heeeieeserca ts | 0.180 Motalwtissucsphosphatidewh yee. 1) seelaeie t-1-t 0.0705 Mitochondria ‘:phosphatide P ................. 0.0473 CellSnucleiphosphatide Ee V9. 2 y-)s)s)-seieteeratoyens | 0.0296 1 Activity of 1 mgm P in percentage of the activity administered to the rat. TURNOVER OF PHOSPHATIDES 359 Phosphatide Turnover in Cell Nuclei and Mitochondria The figures obtained by different investigators for the turnover rate of phosphatides (or for the turnover rate of lecithin, cephalin) are average figures, as the turnover rate in different types of cells and even in different parts of the same cell (cf. Hrvesy, 1947) differ. In Tables 7 and 8 the specific activities of the various isolated fractions resp. the percentage replacement of phosphatide P of the total liver tissue of the cell nuclei and of the mitochondria of the liver by tissues inorganic P, glycerophosphate P and total liver, resp., P is stated. Isolating glycerophosphate we made use of the method applied by ENTENMAN et al. (1948). 6 wcurie of #2P were administered by subcuta- neous injection to each of 6 rats weighing about 150 gm. The animals were killed after the lapse of 2 hours. The phosphatide P turnover of mitochondria makes out 67% only of the corresponding figure of the total phosphatide P while the corre- sponding figure for the cell nuclei is 42 only. The last mentioned figure was in a former investigation found to be 65. The discrepancy is due to the fact that in the method used formerly of separation of cell nuclei (Dounce, 1945) the nuclei containing fraction contained also mito- chondria, in which as seen above, phosphatides are turned over at a more rapid rate than in the cell nuclei. The activity of 1 mgm phosphatide P in percentage of the activity of 1 mgm liver orthophosphate P listed in Table 8 indicates the lower limit of the percentage replacement of phosphatide P in the course of the experiment taking 2 hours. By a fortuitous coincidence, the end value and the average value, of the specific activity of liver orthophos- TABLE 8. — PERCENTAGE REPLACEMENT OF THE PHOSPHATIDE P OF THE LIVER Activity of 1 mgm phosphatide P in percentage of the activity of 1 mgm P of Phosphatide fraction ——— ae = Plasma liver | | ; . -ArTO- | | eek eer | orthophosphate | ortophosphate | phosphate | | Mota tissues wee a. 13.4 12.8 18.75 | 39.2 Mitochondria ...... 9.37 8.95 1k33ql | Zed Celli nuclei... i. | 5.64 5.38 7.89 16.5 phate are almost identical in an experiment taking 2 hours. This is not the case for the specific activity of gyleerophosphate P, as the labelled glycerophosphate accumulates gradually in the course of the experiment by incorporation of labelled orthophosphate P. We can assume the average value of the specific activity of glycerophosphate P during the 360 ADVENTURES IN RADIOISOTOPE RESEARCH experiment to be about half its end value. Correspondingly, we have to multiply the figure 18.75 listed in Table 8 by 2 to arrive at the per- centage renewal of the liver phosphatides, assuming glycerophosphate to be the last phosphatide precursor; when, on the other hand, assuming orthophosphate P to be a relevant precursor, a value of 12.8% is obtained. The turnover rate calculated on glycerophosphate “basis” is thus about 3 times the value obtained, supposing that orthophosphate P is the relevant precursor. In experiments on dogs, ZitveRsMIT et al. (1948) found a similar ratio (3.5) for the turnover rate of liver lecithin calculated on the assump- tion that glycerophosphate P resp. orthophosphate P is the ultimate phosphatide P precursor. Assuming glycerophosphate to be the pre- cursor of phosphatides, the turnover time of the liver phosphatides of the 150 gm rat works out to be 5 hours. As stated on p. 359, several arguments were put forward in support of the view that glycerophosphate is the ultimate precursor of phospha- tides and that the comparison of the specific activity of the liver phos- phatide P with the corresponding value of the liver glycerophosphate ~ P supplies a 3 times as high, and correct, turnover rate, as does compari- son with the liver orthophosphate P. This would indicate that it is the formation of labelled glycerophosphate which takes comparatively tong time, while the incorporation of labelled glycerophosphate into TaBLE 9. — Errect or FEEDING oF LABELLED GLYCEROPHOS- PHATE RESP. LABELLED ORTHOPHOSPHATE ON THE FORMATION OF LABELLED PHOSPHATIDES IN THE LIVER OF THE Rat (Artrom and Swanson 1948) | | Labelled glycero- Labelled ortho- Specific activity (arbitrary units) | phosphate fed phosphate fed see en = | eS 2 Liver glycerophosphate P .... | 45.3 | 18.0 Liver orthophosphate P...... | 21.6 | 23.4 Liver phosphatide P......... | 8.1 6.9 Plasma phosphatide P ....... | 0.5 6.8 the phosphatide molecules is performed at a comparatively rapid rate. This conclusion is not easy to reconcile with a recent finding by ARTOM and Swanson (1948). These authors state that 6 hours following feeding of labelled glycerophosphate, the specific activity of liver glycerophos- phate P was much greater than the corresponding value of orthophos- phate P of the liver and also much greater than the orthophosphate P value of the liver following feeding of labelled sodium phosphate. The specific activity of phosphatide P of the liver was, however, not signi- ficantly greater in these experiments in which a high liver glycero- TURNOVER OF PHOSPHATIDES 361 phosphate specific activity was observed. The presence of a highly active glycerophosphate in the liver did thus not markedly accelerate the formation of labelled phosphatides. The above stated specific acti- vity figures were found by Artom and Swanson 6 hours after feeding of the labelled compounds. In view of the fact that the results of only one experiment is stated, the results of this investigation, which aimed at the elucidation of a very different problem, can however not be considered to invalidate ZitversmMir et al. conclusions. A closer investigation of the phos- phatide formation in ‘the liver following feeding of labelled glycero- phosphate would not be without interest. Search for the Existence of a Small Phosphatide Fraction of Rapid Turnover Rate It is conceivable that a small fraction of phosphatides is present in the liver, which is renewed at a more rapid rate than the average phosphatide molecule present. The existence of such a fraction could influence much the conclusions drawn in the last paragraph which are based on the number of phosphatide mole- cules turned over during a time unit. TaBLE 10. — Sprciric ACTIVITIES OF P FRaAcTIONS oF MICE 10 Minutes FoLtLoWINnNG INTRAVENOUS INJECTION oF 32P Specific activity? Fraction aS _——- 1 2 3 | | Plasma orthophosphate P ...... | 20.4 32.6 31.8 Liver orthophosphate P........ no — 6.56 Liver ATP Pog .....-----+-e- 4.30 4.26 _ Liver phosphatide P........... | 0.0983 | 0.0898 0.0813 1 Activity of 1 mgm P in percentage of the activity administered. Let us assume that of 100 phosphatide molecules 99 are renewed to an extent of 18% in 1 hour, while 1 phosphatide molecule is renewed 100 times during the experiment. Our experiments would hardly reveal the presence of such a fraction responsible for 118 renewal processes, while our turnover experiments would reveal a 19% renewal of the average phosphatide molecules, only. One hundred renewals per hour is a large figure in view of the fact that the very rapidly rejuv- enated ATP P,, in experiments in vitro is found to be renewed 72 times per hour cnly (Meyveruor et al., 1938). But even a 1% phosphatide fraction renewed 10 times in the course of 1 hour would be responsible for 10 renewal processes beside the 18 performed by the rest of 99% phosphatide molecules, increasing the number of molecules turned over to 28. The presence of a small rapidly renewed fraction should become noticeable, in experiments of very short duration. We carried out experiments in which about 14 microcurie *P was injected into the tail vein of each group of 10 mice; the (22—26 gm) animals were killed by decapitation after the lapse of 10 minutes, the organs pooled, and the specific 362 ADVENTURES IN RADIOISOTOPE RESEARCH activity of the plasma orthophosphate, liver inorganic P, liver ATP P.., and phosphatide P determined. Some of the results obtained are listed in Table 10. The phosphatides were purified according to Levin’s method as modified by Haun and Tyrtn (1946). The above results do not indicate the presence of a rapidly renewed phosphatide fraction. Experiments of Long Duration As observed at an early date (Hevesy and ATEN, 1938; ZinveRsMiT et al., 1943, 1948) in experiments of long duration the specific activity of phosphatide P exceeds the corresponding value of orthophosphate P, the last mentioned magni- tude decreasing at a slower rate than the orthophosphate P. As seen in Table 11 the specific activity of orthophosphate P resp. labile ATP phosphorus declines between !/,. day and 84 days (the maximum value is observed about 1/,. day after subcutaneous injection) from 0.546 to 0.000603. Thus, out of 900 22P atoms in the TABLE 11. — Spectric Activiry oF LivER FRACTIONS AT DIFFERENT Dates FOLLOWING ADMINISTRATION OF *2P a Specific activity’ after | | | | | 1/12 day | 1 day 12 days 21 days 44 days | 84 days Fraction ee - — | pat ys a | | | | | a | | Orthophosphate P ... | 0.546 | 0.108 0.0153 | 0.00849 | 0.00134 | — ATP Pog oe eee eeeee | = || SO ENOS 0.0185 | 0.00481 | at 0.000603 Phosphatide PP ..«...'. | 0.00703 | 0.121 0.0252 | 0.00518 | 0.00169 | 0.00066 | | | 1 Activity of 1 mg P in percentage of the dose administered, obtained by comparing the radioactivity of a known aliquot of the solution injected with that of a known amount of phosphorus of the fraction investigated. nitial maximum state, only 1 is present after the lapse of 84 days (beside the loss due to radioactive decay). During the same interval, the beginning of which does not correspond, however, to the maximum value of **P content of phosphati- des, the P content of phosphatides decreases to 1/1) of their initial value. The decline of the 22P of liver orthophosphate, which in view of the high perme- ability of liver cells corresponds closely to the fall of #P content of plasma orthophosphate, is due partly to incorporation of *°P into the tissues and partly to its excretion. With increasing time excretion becomes more and more the sole way of escape of #P from the circulation. While on the first day, the =P content of orthophosphate decreases to 1/; oi its 2-hour-value, a decline to 1/, of the first day’s value takes 12 days, a decline to 1/, of the 12 day-value 9 days, the decrease of the 21st day value to almost 1/, takes not less than 28 days, while a decrease of half the 44 day-value requires 35 days. The renewal of the phosphatides takes place at a slower rate than the escape of 2P from the plasma, however, after a long sequence of days this difference diminishes due to the reduced loss of #2P by the circulation. The values were obtained from pooled organs of 4—6 rats weighing about 160 gm. In the 84-day experiment 0.1 millicurie was administered to each rat, in the experiments of shorter duration correspondingly less. The administration of 0.1 millicurie or more involves risks in experiments of long duration as to bio- logical action of the radiation emitted by the administered P, as the radiation dose to which the rats are exposed during the experiment may amount to a few hundred rep. TURNOVER OF PHOSPHATIDES 363 Fat Metabolism and Phosphatide Turnover A possible connection between fat metabolism and phosphatide turn- over was repeatedly discussed. BoLLMAM and FLock (1946) compared the amount of phosphatides turned over in the liver and plasma with the amount of fat metabolized by the rat and arrived at the following conclusion. Assuming, as found by these authors, 0.175 mgm of phosphatide P to be renewed per hour in a rat liver weighing 5gm, an equal amount of phosphatide P must have been metabolized in the liver or have left this organ during that time, the latter amounting to 0.048 mgm. These figures account for a sufficient phosphatide turnover in the liver and in the plasma to metabolize or transfer fat equivalent to only 3 per cent of the caloric needs of the rat and indicate that phosphatide forma- tion apparently is not an obligatory step in fat oxidation or transfer. For the turnover rate of the 150 gm rat we found about twice the value given by the above authors. Thus, according to our results, the amount of phosphatides turned over should be twice the above figure and, if glycerophosphate is assumed to be the ultimate phosphatide pre- cursor, even three times the last mentioned value, but still only 18 per cent of the caloric needs of the rat. Consequently, our findings do not contradict BortMan and FrLock’s conclusion. In the above consideration no account was taken of the possibility of the existence of a minor phosphatide fraction which may be renewed very rapidly, as discussed on p. 361, nor was the phosphatide turnover in other organs than the liver accounted for. Summary (1) The extent of incorporation of intracellular orthophosphate P of the liver phosphatides of the rat decreases with increasing age. While it amounts to 12 per cent per hour for the liver phosphatides of a 4-day-old rat, the corresponding figure for a 1.5-year-old rat is 6, intermediate figures being obtained for rats of intermediate age. (2) The calculation of the ‘“‘rate of turnover’’ from the ratio of the end value of the specific activities of phosphatide P and the mean value of the specific activities of orthophosphate P during the experiment supplies, in experiments lasting 4 hours or less, almost identical figures with those obtained when, accord- ing to ZILVERSMiT et al., the calculation is based on the change of the specific activity of phosphatide P with time. (3) Replacement of the specific activities of liver inorganic P by corresponding values of ATP P,., of the liver leads to very similar turnover rate values. (4) The ratio of the turnover rate of the phosphatides of the total liver tissue, the mitochondria, and the cell nuclei of the liver is found to be 1 : 0.67 : 0.42. (5) The percentage of labelled P administered present in | mgm liver phospha- tide P of the rat, which is as high as 0.121 after the lapse of 1 day, declines to 0.00066 after the lapse of 84 days. 364 ADVENTURES IN RADIOISOTOPE RESEARCH (6) The lower limit of the turnover of kidney phosphatides is not markedly dependent on the age of the rat and amounts to about 5 per cent in the course of 1 hour. Lower and with the age of the rat decreasing values are obtained for the spleen phosphatides. The percentage rate of incorporation of orthophosphate P per hour for a three-month-old rat can be estimated to be about 4. References . Anuustrom, H. Euter and G. Hrevesy (1944) Ark. Kema, A 2 _ ANDREASEN and J. Orrrsen (1945) Acta Physiol. Scand. 10, 2 . Artom Arch. int. Physiol. 36, 101. . Artom, G. Sarzana,C. Perrter, M. Sanranceto and E. Srere (1937) Nature 139, 836. CG. Arrom, G. Sarzana and E. Srere (1938) Arch. int. Physiol. 47, 245. Cc. Artom and N. A. Swanson (1948) J. Biol. Chem. 175, 871. J. L. Bottman and E. V. Frock (1946) J. Lab. Clin. Med. 31, 478. J. L. Bottman, E. V. Frock and J. Brerxson (1948) Proc. Soc. Exp. Biol. N.Y. 67, 308. R. M. CampesBecy and H. W. Kosreruitz (1948) J. Biol. Chem. 175, 989. I. L. Cuarxorr and D. B. Zitversmit (1948) Adv. Biol. and Med. Physics 2, 322. M. FatKENHEM™ (1943) Amer. J. Physiol. 138, 175. L. Haun and G. Hevesy (1937) Skand. Arch. Physiol. 77, 148. L. Haun and R. Tyr@n (1946) Sv. Vet. Akad. Ark. Kemi, A 21, No. 11. G. Hevesy and L. Haun (1940) Kgl. Danske Videnskab. Selsk. Bviol. Medd. 57 Na: be G. Hevesy and L. Haun (1946 b) Kgl. Danske Videnskab. Selsk. Biol. Medd. 15, Nr 6: G. Hevesy (1947) Sv. Vet. Akad. Ark. Kemi, A. 24, No. 26 G. Hevesy (1948) Radioactive Indicators, Interscience Publ., New York. G. Ivanovics and E. P. Prex (1910) Wien. Klin. Wschr. 23, 573. A. Lane (1937) Z. Physiol. Chem. 246, 219. I. Pertman, S. Ruspen and J. L. CHarxorr (1937) J. Biol. Chem. 122, 169. A. P. Puarr and R. R. Porter (1947) Nature 160, 905. I. Sacks (1949) Cold Spring Harbour Symphosia 13. I. Sacks and E. H. AurscHuLer (1942) Amer. J. Physiol. 137, 1750. R. G. Srncuare (1936) J. Biol. Chem. 114, 94. R. G. Srncuarr (1941) Biolog. Symp. 5, 82. D. B. Zitversmit, C. ENreNMAN and M. ©. FisHier (1943) J. Gen. Physiol. 26, 325; Ibid. 26, 333. D. B. Zitversmir, C. Enrenman and J. L. CHAIKOFF (1948) J. Biol. Chem. 176, 193. I Noy me 5. QQBn 365 COMMENT ON PAPERS 34—35 We calculated in paper 29 and the following papers the rate of renewal of phospha- tides from the specific activity of the phosphatide phosphorus at the end of the experiment, and of the mean specific activity of the cellular inorganic phosphorus during the experiment. The latter is obtained from the specific activities of the total inorganic P after correcting for the share of the extracellular phosphorus in the total inorganic P activity. The possibility was considered in these studies that it is not the cellular but the extracellular **P which participates in the synthe- sis of the organic phosphorus compounds present in the tissues and that ‘It is conceivable that some of the phosphatide molecules are renewed inside the cell wall.”? Correspondingly, all turnover data were calculated, assuming once parti- cipation of cellular #2P and than of extracellular #*P (paper 34). That phosphate enters the cells, at least partly, by the formation of ATP and other intermediates of the cycle on the cell membrane and that inorganic phosphate within the cell arose from the dephosphorylating reactions of the cycle was shown later by Sacks (1951). In paper 36 in which the calculation of the turnover rate is discussed we find the following remark: ‘‘... If the incorporation of the phosphate radical into the phosphatide molecules would be preceded by the formation of glycerophos- phate and this process would be a comparatively slow one, in contrast to all other steps involved in the synthesis of the phosphatide molecule, in this case the turnover rate measured, using labelled phosphorus as an indicator, would be slower than found when using labelled fatty acids or labelled choline...’ and that “the question if and to what extent the rate of labelled phosphate incorporation into the phosphatide differs, for example from that of the fatty acid incorporation into the latter cannot be answered at the time being’’. The question thus raised was answered by CHAIKOFF, ZILVERSMIT, and their associates (1941) who demon- strated that glycerophosphate is the pertinent precursor of phosphatide synthesis and that the calculation of the turnover rate of phosphatides from the specific activity of inorganic phosphorus leads to too low a value for the turnover rate. In experiments with rats, taking 2 hr, the mean specific activity of the ortho- phosphate phosphorus of the liver was found to be three times that of the corre- ponding value of glycerophosphate phosphorus (paper 35); thus, the turnover rate calculated assuming the liver orthophosphate P to be the phosphatide pre- cursor has to be multiplied by three to arrive at a correct turnover rate value. The turnover rate of phosphatides present in the different sub-units of the liver cell was found to differ markedly (paper 35 and Hrvesy, 1947). Applying palmitic acid — 1 — 4C as a precursor Chaikoff et al. could also demonstrate that the liver is the principal site for the formation of fatty acid ester bonds of plasma phos- phatide molecules. References G. Hevesy (1947) Ark. Kemi, A 24, No. 26. J. Sacks (1951) Arch. Biochem. 30, 423. D. B. Zitversmit, C. ENTENMAN, M. C. Fisuurr and I. L. Cuarxorr (1941) J. Gen. Physiol. 26, 325. D. S. Gotpman, I. L. Cuarxorr, W. O. Rernnarpt, C. ENTeEnNMAN and W.G. DauvBEN, (1950) J. Biol. Chem. 184, 727. Originally communicated in Nature 141, 1097 (1938) 36. MOLECULAR REJUVENATION OF MUSCLE TISSUE G. Hevesy and O. REBBE From the Institute of Theoretical Physics and the Zoophysiological Laboratory University of Copenhagen Tue decomposition of creatine phosphoric acid during muscular action and its rebuilding during rest, has been the subject of numerous detailed investigations. We were interested in the problem, if, and to what extent, creatine phosphoric acid molecules are decomposed and after- wards rebuilt, or ‘rejuvenated’, in the resting muscle. This problem can be easily solved by injecting labelled sodium phosphate, for example, into frogs, and determining if, and to what extent, creatine phosphoric acid extracted from the muscle of the frog becomes labelled (radioactive). Phosphorus atoms present in creatine phosphoric acid and other organic compounds do not exchange spontaneously with other phosphorus atoms present, and thus the fact that labelled creatine phosphoric acid can be isolated from the muscle is a proof that this was synthesized after the administration of labelled sodium phosphate. The muscle was placed at once after removal in liquid air, the acid soluble components extracted with trichloracetic acid kept at —9°, and the inorganic phosphate present in the solution precipitated as ammonium magnesium salt. The next step was the decomposition of creatine phosphoric acid remaining in the filtrate from the last- mentioned precipitate. The decomposition was carried out by adding sulphuric acid (17) and ammonium molybdate (1 per cent) to the solution. The phosphate ions were then precipitated as ammonium magnesium salt. The phosphorus content of the latter was determined by the colori- metric method of Fisk and SuBBARow, and its radioactivity by making use of a Geiger counter. The results obtained for this and some other fractions are seen in the accompanying table. A specific activity of the creatine phosphoric acid phosphorus amount- ing to 49 per cent of that of the inorganic phosphorus indicates that 49 per cent of the creatine phosphoric acid molecules present in the resting muscle were split and newly synthesized through enzymatic action in the course of the last 3 hours before the frog was killed. As the total number of creatine phosphoric acid molecules present in the muscle can be assumed not to have changed during that time, in the resting MOLECULAR REJUVENATION OF MUSCLE TISSUE 367 muscle we are faced with a molecular rejuvenation of the creatine phos- phoric acid to the above extent. The adenosin and the hexosephosphate molecules are rejuvenated to about the same extent as those of the creatine phosphoric acid. With increasing temperature, as is to be expected, the rate of molecular rejuvenation increases, and in the course of less than a day practically all creatine phosphoric acid molecules are renewed. PuospHorus IsoLaATED FROM FROG KILLED 3 HOURS AFTER SUBCUTANEOUS InsEcTION OF LABELLED Sopium PHOSPHATE Relative specific activity (activity per mgm P) Frog kept at 2° | Frog kept at 21° IGnormeaione 12 Soocccvasouadudd aoe Doc 100 | 100 Cream IP Gogoccgecago0duoucouduone 49 | 78 Adenosin P (7 min hydrolysed at 100°) | 50 | -— «‘Hexose”’ P (30 min hydrolysed at 100°) 52 | 78 Product of 100 min hydrolysis at 100° 45 | —- Non-acid soluble residual fraction .. 9 38 The new formation of some of the ‘acid soluble’ phosphorus compounds present in the blood also takes place to a very appreciable extent. In human blood 2 hours after intravenous injection of labelled sodium phosphate, the specific activity of the total acid soluble organic phos- phorus, kindly extracted by Mr. A. H. W. Aten from the blood cor- puscles, amounted to 20 per cent of that of the plasma inorganic phos- phorus. In experiments in vitro? in which dog’s blood was shaken for 2.5 hours with labelled sodium phosphate, 1/25 of the total acid soluble molecules was found to be labelled and thus split and resynthesized under the action of enzymes. In the same én vitro experiments the for- mation of only very minute amounts of labelled phosphatides (less than 0.1 per cent) could be ascertained. Also in experiments in vivo labelled phosphatides were found to be present to an appreciable extent in the blood only after much longer time. The specific activity of phospha- tide P extracted from human blood corpuscles 24 hours after administra- tion of labelled sodium phosphate was found to be 40 times less than that of plasma inorganic P, showing the very low rate of rejuvenation of the phosphatide molecules present in the blood. There is thus a conspicuous difference in the rate of rejuvenation of some low molecular water soluble compounds, as for example, creatine phosphoric acid, adenosin phosphoric acid, hexosephosphate, and non-water soluble products like phosphatides, nucleoproteins and similar 368 ADVENTURES IN RADIOISOTOPE RESEARCH compounds, present in the blood. This difference is closely connected with the fact that the first-mentioned compounds are at least partly rejuvenated through enzymatic action in the blood itself, while the latter are principally rejuvenated in the organs and carried from these in the blood stream. The measurement of the rate at which, for example, adenosintriphosphoric acid molecules are renewed can be conveniently used to determine the amount of enzymes the presence of which enables the exchange reaction to take place. References Q) K.Loumann, Biochem. Z. 194, 306 (1928). (2) L. Haun and G. Hevesy, Bull. Lab. Carlsberg 22, 188 (1938). Orginally communicated in Kgl. Danske Videnskabernes Selskab. Biologiske Meddelelser, 15, 7 (1940) 37. RATE OF RENEWAL OF THE ACID SOLUBLE ORGANIC PHOSPHORUS COMPOUNDS IN THE ORGANS AND THE BLOOD OF THE RABBIT 1 G. Hrevesy anp L. Haun From the Institute of Theoretical Physics, University of Copenhagen W the rate of renewal In a paper published recently in these Proceedings of the phosphatide molecules present in various organs of the rabbit and other animals was discussed. In the present publication, data on the rate of new formation of acid soluble phosphorus compounds are commu- nicated. The acid soluble organic P compounds represent a great variety of chemically very different bodies: esters as, for example, hexosephos- phate, nucleotide compounds as adenosintriphosphate, phosphagen, and other compounds. These compounds” are renewed at a comparatively fast rate in the organs in contradistinction to the phosphatides and desoxyribo nucleoproteins®., Furthermore, while the rate of new forma- tion of the phosphatides in the circulation is almost negligible, the acid soluble phosphorus compounds are renewed at a remarkable rate in the corpuscles. These facts justify the consideration of the acid soluble phosphorus compounds from our view-point as a definite group of the phosphorus compounds present in the body. EXPERIMENTAL METHOD Labelled P as sodium phosphate was administered by intravenous or subcutaneous injection to rabbits all through the experiment in order to keep the activity of the plasma inorganic P at a constant level. After the. lapse of some hours or days, the animal was killed by bleeding. The fresh organs were placed in liquid air and were extracted immedia- tely with cold 10 per cent trichloroacetic acid. The inorganic phosphate () G. Hevesy and L. Haun, Kgl. Danske Vidensk. Selskab. Biol. Medd. 15+ 5 (1940). ‘) With the exception of adenylic acid [T. Korzyssxr and J. K. Parnas, Z. physiol. Chem. 255, 195 (1938)] and, possibly, of other not yet known minor components of the acid soluble P mixture. (9). Haun and G. Hevesy, Nature, April 6, 1940. 24 Hevesy 370 ADVENTURES IN RADIOISOTOPE RESEARCH present in the extract was precipitated as ammonium magnesium phos- phate. The filtrate obtained was then hydrolysed with 1 n H,SO, for 7 min at 100° to split off the labile P which was then precipitated as ammo- nium magnesium salt. The filtrate obtained after the last mentioned operation was hydrolysed 100 min to split off the phosphate radical of the hexosephosphate present. In order to avoid several consecutive precipitations of ammonium magnesium phosphate which lead to an accumulation of very appreciable amounts of ammonium salt in the soluble fraction, we usually divided the filtrate obtained after precipi- tation of the inorganic phosphate present as such in the tissue into aliquot parts. One aliquot part was hydrolysed for 7 min, the phosphate split off was precipitated, and the filtrate obtained was hydrolysed for 100 min. Another aliquot part was hydrolysed for 7 min, the filtrate obtained was hydrolysed for 12 hours, and so on. The phosphate of the creatinephosphoric acid was split off by heating the solution for 30 min to 40°. In some cases, the total acid soluble organic P was converted into phosphate and was investigated in toto. The ammonium magnesium phosphate precipitates obtained were dissolved in diluted hydrochloric acid and an aliquot part was used for a colorimetric P determination. To another aliquot part about 80 mgm non-active sodium phosphate was added; the total P present in the solution was then precipitated as ammonium magnesium salt. The radioactivity of these precipitates was determined by the aid of a GricER counter. Though the separation of the different acid soluble P compounds described above is far from being quantitative, it sufficed in most cases for our purpose. In the experiments with blood, as anti-coagulent, ammonium oxalate was used. The corpuscles were centrifuged off and washed twice with a physiological sodium chloride solution. In experiments in vitro, the blood was kept in a CO,—O, atmosphere and was shaken, after addition of labelled sodium phosphate of negligible weight, for 30—190 min in a thermostat at 37°. RATE OF NEW-FORMATION As labelled phosphorus atoms can only be incorporated into organic molecules in the course of a synthetic process, the radioactivity of the organic phosphorus compounds isolated from an organ is a measure of the rate of its total or partial resynthesis. It is, however, not permitted to compare the specific activity (activity per mgm P) of the hexose- monophosphate extracted from the kidney and the muscle, for example, and to conclude from the fact that the hexosemonophosphate extracted from the kidney is much more active than that secured from the muscle, RENEWAL OF ACID SOLUBLE PHOSPHORUS COMPOUNDS IN ORGANS OF RABBIT 371 that the rate of new formation® of hexosemonophosphate is correspond- ingly larger in the kidney. The incorporation of labelled P atoms into the hexosemonophosphate molecules must be preceded by a penetra- tion of the labelled inorganic P into the cells of the organ. If this process is slow, the rate of formation of labelled hexosemonophosphate molecules is bound to be comparatively slow, in spite of a possibly very fast rate of new formation of the hexosemonophosphate molecules inside the cells of the organ in question. In fact, the labelled inorganic phosphate molecules penetrate very much faster into the kidney cells than into the muscle cells. To get proper information on the rate of renewal of an organic compound in an organ, we have to compare the specific activity of the P isolated from the organic compound in question at the end of the experiment with the average value of the specific activ- ity of the cellular inorganic P prevailing during the experiment. The TABLE 1. — Extent oF RENEWAL OF THE ToTAt ORGANIC ActpD SOLUBLE P IN THE ORGANS OF THE RABBIT Rabbit II. Weight 2.6 kgm. Intravenous injection during 215 min od | | | | | A B C | D Specific | Specific Average | Specific D sx. 100 aD) x 100 | activity of | activity of | specific ac- | activity of | } | B O | the tissue | the cellular | tivity of the | the organic | PPE | | . See ey Siete | inorganic P | inorganic P | cellular P | 12) Upper mae Lower limis | | of the per- | of the per- | at the end at the end Garnceine | zt the end | centage centage | of ee ex- | of ane Ges experiment | of wae ex- | renewed renewed periment periment, | periment | IDG 56 jo peo Comoe 100 -- — _ — | — CORDUSCLES top-ten l= 12 ey l2ete | yb Med 199 100 GGChitw? Gohooacosouoc | 87.4 Ip eexedic! 17.8 33.6 43.2 38.6 Small intestine, | | 7 | = 99 & . 7 mS ANUC OSAleieta}sretete stash 47.4 | 45.2 | 22.6 | 24.0 106 Dork MWAVCLE ceter sonore dic les sks sei s 44.0 | 40.6 i 20.4 14.3 70.2 | 35.2 Dina tea pna Ae es | BOLO me | ee26Oe) i ASH Ob 1 ULOF =| 85.3 | | | | Saleen sa sbococdnooGe Ors | 28:5 | 14.3 | — | = | -—- | | Stomaehintsi. alc 0s 5 ss |. 25.9 | 23.6 TS i) 16:9 71. 258-5 5 | 201 | | ‘ | IB) cae see ce oeue | 25.50) | 21.4 10.8 | 8.6 79.6 40.2 Leiiiee Ae AY ee | 1.32 | woe =e | OPS 608 — | = ©) The inorganic P extracted from the heart contains partly such inorganic P atoms which were formed through decomposition of creatinephosphoric acid prior to the extraction. As the specific activity of the creatine P is, after the lapse of 4 hours, lower than that of the inorganic P (comp. the muscle values in Table 3!), the specific activity of the cellular inorganic P of the heart is in fact higher than that stated above and, correspondingly, the values of the rate of renewal of the organic acid soluble P compounds in the heart are smaller than those stated in the last and the last but one column of Table 1. { The significance of the notion of the rate of new formation is discussed in the paper by G. Hrevesy and L. Haun, Kgl. Danske Vidensk. Selskab, Biol. Medd. 5 (1940). 24* = «6 3S ADVENTURES IN RADIOISOTOPE RESEARCH considerations mentioned above are discussed in detail in the publication cited above on the turnover rate of phosphatides. In this paper is de- scribed the method which permits us to calculate from the specific activity of the tissue inorganic P. the specific activity of the plasma inorganic P, the extracellular space of the organ, and the inorganic P content of the TasBLE 2. — Extent oF RENEWAL OF DIFFERENT FRACTIONS OF THE ORGANIC ACID SoLUBLE P Rabbits IT, III, and IV (average) Intravenous injection during 4 hours | Spec. activity of the or- | Spec. activity of the or- | | ganic P at the end of the | ganic P at the end of the experiment: Average experiment: Specific |pysees Ys _| Specific activity of the activity of the cellular Time of hydrolysis in 1 N 3 2 P i bie Organ MeO eran | cellular inorganic P during | inorganic P at the end ea i | the experiment. of the experiment. | | (Upper limit of the per- | (Lower limit of the per- | centage renewed) centage renewed) | \ : 5 : | me ILIMGR SeacogonoGnOG ot | O0—7 min | 152 76 An EWS SSoaoua So.0 odes | non-hydrolysed _ 66 | 33 Kidney, cortex. -. <7... | O0—100 min 64 57 Kidney, cortex 2): 1. | 100 min—12 hr. | 47 42 we ! | . Kidney, cortex ......: non-hydrolysed | 29 26 | | | | : | hydrolysed in 1 N | ; Kidney, cortex <..:... if s y | 48 » 43 \! NaOH at 80° If i] | TasBLe 3. — Specrric Activiry or Acip SoLuBLE P FRAcTIONS EXTRACTED FROM THE ORGANS OF THE RABBIT Rabbit VII. Weight 2.4 kgm Subcutaneous injection during 11.5 hours | Specific activity Fraction | at the end of the experiment Plasma inorganic oP saeyhe oe eens aes kel esenehe ee tokoroate eaten keraier= | 100 Corpuscle amorganie Ge cytere ore1e ie ol ofellol et sein ol aot! ol ails) ol 25 Corpuscle P hydrolysed 15 hours in 1 N H,SO, at 100° | 25 Corpuscle P hydrolysed 15—120 hours in 1 N H,SO, at 10° 25 Corpuscle non-hydrolysed residue ..........+.-++-+++---- 13.0 WhISCley miatorteth avy 12 So aoaaascoc ono aopdoccoddsdasacodc 15.5 Mnseclesicneatime We 25 <2 scsi ciis cierene cicie cactepeeneds eres yes au siis= 8.5 Marrowannoroanic JP) 5. ssc care toe ieheteeeieiete ions «mapas | yal Wisnaroniy OHeshaiKe) IEA Goo oo ob db ooo o0D ooo Do odOOa MODOC oC 36.8 [sieioa ino 12 Sotoogadaacsocogdes coco boddoodonoo G6 3.0 loyal Oe VaNS) 1 Go ooooUUeUORUooOONbGoUawOagmaooNsOC oC 2.3 @) The low value is presumably due to the presence of traces of slightly active bone P in the marrow sample- RENEWAL OF ACID SOLUBLE PHOSPHORUS COMPOUNDS IN ORGANS OF RABBIT Sis organ and plasma the average specific activity of the cellular inorganic P during the experiment. From the latter magnitude and the specific activity of the P of the organic phosphorus fraction at the end of the experiment, we can calculate what percentage of the organic compound in question is newly formed during the experiment, if only the extent of new formation is restricted. If a large fraction of the hexosemonophosphate molecules, for example, is newly formed during the experiment, we can no longer disregard the number of hexosemonophosphate molecules which were decomposed and resynthesized more than once during the experiment. If such a repeat - ed new-formation takes place, it will have the effect that the active hexosemonophosphate molecules present at the end of the experiment cannot be longer considered as having been formed with participation of inorganic P which had an activity corresponding to the average acti- vity prevailing during the experiment. The inorganic P atoms, which had an activity corresponding to a late stage of the experiment, will TaBLE 4. — Speciric Activiry oF AcripD SoLuBLE P FrRactTIoNns EXTRACTED FROM THE ORGANS OF THE RABBIT Rabbit VIII. Weight 2.0 kgm Subcutaneous injection during 9 days Specific activity Fraction at the end of the experiment aes eee 2 = ne ae ane = lAleison) rhayogegManKo 12. oo dooangecnonononapcdds Colac GO cn | 100 Corpusclestotaleacids soluble rr. ce sact-nielsiioie + lela oles | 94 Musclemimorcanic:—eneatime Beiter oterers) ponceeae o's) erode oie oe | 40 IMiaSCIEWES Letes bararcy cate econ e ous okcis chavs the sy chausheon eveterenessceve e eua tree | 18.7 Brain inorganic’ -|- creatine PB .......-..-2...<, (oadanod¢ | 18.8 BS TAI CS TOT PEE 29) nee the wre iste Seve ep orotots (hee apadeeeiadeds a auspouaie oharste | 17.3 TaBLE 5. — Speciric Activiry oF AcipD SoLtuBLE P FRAcTIONS EXTRACTED FROM THE ORGANS OF THE RABBIT Rabbit IX. Weight 2.5 kgm Subcutaneous injection during 50 days Specific activity Fraction at the end of the experiment Plasmaein orp Anica isis cla ate re ake crete boey> co orahous cf aVeareeetee | 100 Corpusclestotaleacidesolulblen I ii. cero i= a cvarssiaoi elicitin ici 100 Musclesinorganicsrereatimey by leiei-- ic aici oe) siela eeierelsuci elle | 88 IMniscle: ester Meme mer acter ster sterei ols stekeke ts, ojala cies Wis .0 Sieteucce at Braininorganicy— = eneatune) HPs v.teeicds +/a)e ie G 4 2 5 4 Time in hours Percent of administered 32 P present in the total plasma ne Fic. 2. =P content of the frog’s plasma after injecting **P into the lymph sack. T = 21°. In another set of experiments, we injected large amounts of phosphate (corresponding to 6.1 mgm P) into the lymph sack of the frog. The inorganic P content of the plasma at different times is seen in Table 2. TABLE 2. — PContTENT OF THE PLASMA AFTER INJECTING 6.1 MGM. Pinro THE LyMPH SACK OF THE FROG | | | | ee : - Excess P present in | Weight of the frog | P content of the | _ ah Time in hours A j | the plasma mgm in gm | plasma mgm per cent | : | | per cent | | 0.5 | 76 7A AT | 18.1 1.5 | 78 23.9 | 20.3 19.0 | 50 | 10.3 6.7 RELATIVE RATE OF UPTAKE OF LABELLED PHOSPHORUS BY THE BONE AND THE MUSCLE In our preliminary experiments, the results of which are reported in Tables 3—10, we compared the uptake of labelled phosphorus by fresh bone and muscle samples of equal weight at 2° and 22°, respect- ively. 1 gm muscle tissue was found to take up less =P than 1 gm bone tissue, though the difference is much less than would be expected if 2be 388 ADVENTURES IN RADIOISOTOPE RESEARCH TasLE 3. — Froe I, Kepr at 2°, Kinrtep arrer 12 Days Ash weight: 3.235 gm | Per cent labelled P | Weight of fresh tissue Weight of ash | administered found : in 1 gm fresh tissue | | \ | Bones mgm mgm Right femur:...... 427.8 126.3 4.33 heftriemaumy oe... <. 442.8 123.6 4.74 Right tibia ....... 469.5 149.2 4.76 IGGIAy qlee) Sagogoge 432.6 | 141.1 5.23 Museles gm | Right thigh ....... | 3.989 — 0.284 Wefitetbie hy. .-...-1- 6 3.467 | = 0.262 Right calf ........ 1.401 = 0.237 IGEN Chl Sac genooe | 1.336 = 0.199 1 mgm bone P had the same chance to be replaced by labelled P as 1 mgm muscle P. 1 gm bone tissue contains about 30 times as much Pas 1 gm muscle tissue. The comparatively low activity of the bone P is due to the fact that those phosphorus ions which are located on the TasiLe 4. — Froe II, Kerr at 2°, Kimtep artrer 12 Days Ash weight: 3.014 gm | | | Per cent of labelled | Weight of fresh tissue Weight of ash P administered found in 1 gm fresh tissue Bones mgm mem Right efemur <..-..: 417.0 123.9 4.74 Keftekenvure cee cit 425.4 116.1 4.38 Richt, tibiawe see 450.6 150.8 5.01 Ibi nly Goooaoes 450.7 152.2 4.74 | Muscles gm. Right thigh’ 22.2... 5.461 -- 0.24 Meh cthiohmeee ce see 4.635 — 0.23 Rightucalf sec. ss 1.670 = 0.28 Wetiecaliges eric 1.679 = 0.31 surface of the apatite crystallites containing the mineral constituents of the bone can be replaced by a physical exchange process with the labelled phosphate in the plasma or the lymph, while the phosphate ions present inside these crystals cannot be replaced. Labelled phosphate ions can only be incorporated into the inside of the apatite crystallites CIRCULATION OF PHOSPHORUS IN THE FROG 389 TABLE 5. — FroG ILI, Kerr at 2°, KizneED AFTER 2] Days Ash weight: 3.104 gm en | Per cent of Inbelled Weight of fresh tissue Weight of ash P administered found in 1 gm fresh tissue Bones mgm. mgm Right tibia epiphysis .......... 155.1 29.2 6.03 Left tibia epiphysis ........... 142.9 29.1 5.90 Right tibia diaphysis .......... 205.7 77.0 6.34 Left tibia diaphysis: 5....-...... 200.4 76.1 6.50 Museles em [Ruvedous, Tle SG aaa coo Gobmo UDO 1.269 = 0.51 Hee Ki C Aiken mae nctenctic ests hers sre etelsnehere e252 _ 0.42 during the formation of such crystallites from a plasma containing labelled phosphate. The dissolution of ‘‘old’” apatite crystals and the formation of ‘‘new” ones is, however, a comparatively slow process and is restricted to a fraction of the bone apatite. In contrast to the TasBLEe 6. — Froc IV, Kept at 2°, Kintep arrerR 22 Days Ash weight: 3.347 gm | } \ | | | Per cent of labelled P |Weigt of fresh tissue Weight of ash | administered found | in 1 gm fresh tissue Bones mgm mgm Right tibia epiphysis .......... 144.2 33.1 3.68 Left tibia epiphysis ........... 158.7 39.6 4.50 Right tibia diaphysis :......... 205.8 86.6 5.07 Left tibia diaphysis ........... 215.2 80.8 4.52 Muscles om. laueditir, WawvedM oo oo BO eb Ob ao OOOO 4.116 — 0.26 refit pila go atnetoyescpoicte Vis ro:s =e 10 or 4.109 — 0.29 Righitg callittgerenyertsttsc siete. 22-04 +: 1.477 -- 0.34 Left calf AdnoatosépnepoowudGec 1.548 — 0.36 bones, the greatest part of the phosphorus in the muscles is present as a constituent of organic compounds, as seen in Table 1 b. The bulk (about 80 per cent) of the organic P is present in the muscle of the frog in the form of acid soluble phosphorus compounds and, in a corresponding manner, the rate of activation of the muscle P depends 390 ADVENTURES IN RADIOISOTOPE RESEARCH mainly upon the rate of formation of active acid soluble P compounds. This process is much faster than the formation of the apatite crystals of the bone tissue and this fact explains why the replacement of phosphorus in the muscle tissue takes place at a much more rapid rate TABLE 7. — FrRoc V, Kept at 20—24°, KinteED arreR 8 Days Ash weight: 3.65 gm | | Per cent of labelled P | Weight of fresh tissue | Weight of ash administered found | | in 1 gm fresh tissue | | Bones mgm mgm Right femur epiphysis ......... | 172.9 33.5 3.30 Left femur epiphysis .......... 183.9 | 36.9 Be22 Right femur diaphysis ......... | 412.7 HRY 2.99 Left femur diaphysis .......... | 436.4 123.9 2.84 Right tibia epiphysis .......... | 265.9 | 59.6 2.92 Left tibia epiphysis ........... 237.1 | 62.4 3.16 Right tibia diaphysis .......... | 374.4 | 141.9 3.02 etki atibiadiaphysist 11). ole | 358.7 130.2 2.91 Muscles gm | SukeddMeqndKeAN oseG og agen oonaaD Ae | 5.372 —- 1.04 Left thigh: .fc.ss0 sees | 5.632 | = | 0.92 Right, Cal€ oh Sess. dnc ehs ensvslelalotets Goss 1.925 1 Deehbecale ac) ate eae ee | L980 a | Ll TaBLe 8. — Frog VI, Kept at 22°, KitLtED AFTER 8 Days Ash weight: 3.46 gm | | Per cent of labelled P Weight of fresh tissue Weight of ash administered found | in 1 gm fresh tissue | Bones | mgm ) mgm Right femur epiphysis ......... | 150.6 | 37.1 | 4.44 Left femur’ epiphysis 3.-......- | 169.8 | 37.6 2.66 Right femur diaphysis ......... | 398.9 130.4 | 3.26 Left femur diaphysis .......... | 382.0 122.3 3.20 Right tibia epiphysis .......... 263.9 60.2 2.97 Left, tibia sepiphysis! <4........-% 252.4 60.7 3.12 Right tibia diaphysis .......... 304.3 | 122.6 | 3.66 Left tibia diaphysis ........... 282.5 | 120.9 3.42 om | Richter call opm re tae oe eee 1.520 == eh Mert Galtier. Aoctameicen trocar | 1.584 — 1.2 CIRCULATION OF PHOSPHORUS IN THE FROG 391 TABLE 9. — FroG VII, Kerr at 20—24°, KintepD arrer 12 Days Weight: at the start 56.5 gm; at the end 49.5 gm Ash weight: 2.359 gm ee Per cent of | sbelled Ig Weight of fresh tissue Weight of ash administered found | . . | | in 1 gm. fresh tissue Bones m2m mem Right femur epiphysis ......... 79.8 Wiel 4.27 Left femur epiphysis .......... 77.6 16.7 4.35 Right femur diaphysis ......... Sieg 49.1 8.02 Left femur diaphysis .......... 103.1 485 4.70 Muscles em. Ela, Wali > oo ocoU DUE OOOO OC 2.608 = al ID Tae, CoecoeobcedsouceouuT 2.139 = ites Redes Call ce oop tiaeeoebe cog 5b oF 1.067 --- 0.9 ILE. GEN ooeoondoaomdbd duouUDt 1.630 = 1.1 than the replacement of phosphorus in the bone tissue. From the fact that the ratio of the uptake of 2P by 1 gm muscle tissue and 1 gm bone tissue decreases much with increasing temperature (see Table 10), we can conclude that the temperature coefficient of the penetration TasLe 10. — Comparison OF THE LABELLED P CONTENT oF BonES AND MUSCLES ; Duration of the expe riment Ratio of the labelled P contert of EOS | days | bone and muscle of equal weight Temperature: 2° Shoo dsc6.cnIo9 12 19.4 IE Goose conse 12 ari 1D OL Ge peor 21 12.8 (epiphysis) NIE Segoe ot hiexchoy ie 21 13.9 (diaphysis) IW eoeaococbn as 22 13.2 (epiphysis) IY: Goce aeons 22 15.5 (diaphysis) Average value... 15.4 Temperature: 22° Wf socogancodde 3.0 (epiphysis) Wf oS60dcaq0n0¢ 2.9 (diaphysis) WAlicacctoaedend 2.9 (epiphysis) AA Re eis omiaromoc 8 2.9 (diaphysis) WANE gadkedocooor 12 4.1 (epiphysis) WELD, Ss Se sraetetee 12 4.6 (diaphysis) Average value... 3.4 392 ADVENTURES IN RADIOISOTOPE RESEARCH and subsequent incorporation of labelled phosphate into the organic compounds of the muscle of the frog is much greater than the tempera- ture coefficient of the formation of apatite crystals. EXCRETION OF LABELLED P In a few cases, we determined the percentage #?P which was excreted by the kidneys of the frog. In one experiment, 1.5 cc. 0.6 per cent sodium chloride solution containing 0.008 mgm P as phosphate was injected into the lymph sack of the frog weighing 55 gm and kept at 18°. Urine was collected during 14 hours and the ##P content of the urine was determined. It was found to make out 10.6 per cent of the *#?P injected while, in other experiments, 7.1 and 5.8 per cent, respectively, was found. UPTAKE BY THE FROG OF *P FROM A SOLUTION CONTAINING LABELLED SODIUM PHOSPHATE A frog weighing 88 gm was kept at 18° in 100 cc. 0.6 per cent sodium chloride solution containing 4 mgm labelled P as sodium phosphate. The solution was renewed every day. After the lapse of 24% days, the frog was washed, killed and its P content extracted. It was found to be 695 mgm or 7.9 mgm per gram of fresh weight of the frog. The specific activity of this P was found to constitute 1/450 of the specific activity of the P of the solution in which the frog was kept. Thus, in the course of 214 days, 1/450 of the total P of the frog was replaced by solution P. We investigated, furthermore, the activity of the inorganic P extracted from the liver of the frog which was found to show a specific activity amounting to 0.99 per cent of the specific activity of the solution P. It was, thus, 4.5 times more active than the average P of the frog. RATE OF RENEWAL OF THE PHOSPHORUS COMPOUNDS IN THE MUSCLE In the preceding sections, experiments were described in which the percentage of the administered 32P present in the skeleton and the muscles was determined. In the following, we wish to discuss the rate at which the organic phosphorus compounds present in the muscles : 7 . 1) of the frog are renewed. We shall consider those cases of renewal‘ ( It is conceivable that molecules are renewed whithout the splitting off and reincorporation of phosphate group. CIRCULATION OF PHOSPHORUS IN THE FROG 3923 incorporated into the organic molecules. For example, creatinephosphoric acid is degraded under splitting off of phosphate and resynthesized under uptake of phosphate radicals. If labelled phosphate ions are pre- sent, they will have the same chance to be incorporated as have non- labelled ones. Let us assume 10% free phosphate ions present in the muscle cells to contain 10 **PO, ions while from 10° P atoms isolated from hexosemonophosphate of the muscle tissue only 1 is **P, then we have to conclude that 10 per cent of the hexosemonophosphate mole- cules were renewed during the experiment under incorporation of free phosphate. The ratios of the specific activities of the inorganic P and the organic P are, thus, a measure of the extent of renewal of the organic P compound which took place during the experiment. When trying to arrive at quantitative data we encounter the following difficulties: (a) The free phosphate extracted from the muscle tissue is partly cellular and partly extracellular phosphate; it is, however, the specific activity of the cellular phosphate only which is to be considered when calculating the rate of renewal. (b)The specific activity of cellular phosphate changes during the experiment, the change being due, for example, to an increas- ing influx of labelled phosphate into the muscle cells. In this connection it should also be mentioned that the method permits to distinguish between renewed and non-renewed, between ‘‘old”’ and ‘‘new’’ molecules; but no information is supplied on the point whether the molecules are repeatedly renewed in the course of the experiment or not. As to point (a), to account for the share of the extracellular phosphate in the total phosphate of the muscle tissue, we must know the specific activity of the plasma phosphate which we assume to be identical with the specific activity of the extracellular phosphate. We must also know the phosphate content of the plasma and that of the muscle tissue and, finally, the size of the extracellular space. The last mentioned magni- tude can be determined in each case by administering simultaneously with the labelled phosphate labelled sodium, or it can be assumed that the extracellular space makes out 14 per cent of the weight of the muscles. Another procedure which we used repeatedly is the following. We remove one leg of the frog 1 hour after the start of the experiment and determine the specific activity of the free muscle phosphate P. After further 3 hours, we extract the phosphate of the other gastrocnemius and determine the specific activity of the free phosphate P. If, within 1 hour, a proportional partition of 82P between plasma phosphate and the extracellular phos- phate took place, then the increment of the specific activity of the tissue phosphate between 1 hour and 4 hours is solely due to an increase in the specific activity of the cellular inorganic P. By this method, we can determine the percentage of cellular P which was replaced in the muscle © Comp. G. Hrvesy and O. Resse, Acta Physiol. Scand. 2, 171 (1940). 394 ADVENTURES IN RADIOISOTOPE RESEARCH by plasma P between 1 hour and 4 hours after the start of the experi- ment. The fact that the inorganic P of the tissue is partly of extracellular origin will lead to an overestimation of the activity of the cellular inor- ganic P and, thus, to an underestimation of the renewal figures of the organic P compounds. This source of error is mainly to be considered in experiments of short duration carried out at low temperature. On the other hand, even if the greatest precautions are observed, we risk a decomposition of some of the creatine phosphate present in the tissue prior to the separation of the inorganic P. Such a decomposition will lead to a decrease in the specific activity of the inorganic P, the inor- ganic P originating from creatine P being on the whole less active than the ‘free’? phosphate P. We shall, thus, underestimate the specific activity of the inorganic P and, correspondingly, overestimate the rate of renewal of the organic P compounds. This error will also be larger in experiments of short duration carried out at low temperature. We wish to mention a further possible experimental error. If the free phos- phate is not precipitated quantitatively, we risk to find some strongly active phosphate in the creatine phosphate fraction. A non-negligible amount of phosphate may remain in solution in cases in which the amount of P to be precipitated is very small. The following objection can be put forward regarding the calculation of renewal rates from the ratio of the specific activities of the inorganic P and the P split off from organic compounds. The P secured as inorganic phosphate might even after the most careful handling of the tissue have been largely present not as free phosphate in the tissue cells but incorporated in very labile compounds which were decomposed in the course of the extraction process. It is possible that this is the case, it is even quite possible that a large part of the inorganic P extracted as such from the muscle cells was originally present incorporated in very labile compounds and was decomposed during the extraction process. General experience indicates, however, that very labile phosphorus compounds are renewed at a fast rate and we can, therefore, expect the P of such labile phosphorus compounds to obtain within a short time a similar specific activity as shown by the inorganic P present in the cells. Should that not be the case, then the comparison of the specific activity of the ‘inorganic’ P with that of the P split off from the organic compound in question, would obviously lead to an overesti- mation of the rate of renewal. The specific activity of different P fractions is seen in Tables 11—18. Though all precautions were taken to prevent decomposition of creatinephosphorie acid it is difficult to state whether the variations in the values obtained for the rate of renewal of creatinephosphoric acid molecules in some of the experiments are genuine or are due to a more CIRCULATION OF PHOSPHORUS IN THE FROG 395 or less successful prevention of the decomposition of the creatinephos- phoric acid prior to the removal of the inorganic phosphate of the muscle tissue. The results of an experiment, in which the frog was kept at 20° for TABLE 11. — Sprctric Activity oF PHOSPHORUS FRACTIONS EXTRACTED FROM THE GASTROCNEMIUS OF A FrRoGc, 4 Hours Aarrer INJECTING LABELLED Soptum PxHos- PHATE INTO THE LyMPH SACK. TEMP.: 2° | | Activity in per | Per cent of total Relative Fraction PI content | cent of the stan- | activity adminis- specific ' in mgm | - | f | dard preparation | tered permgm P activity I, Jbaxorgefannxe IE) Gouagoccccuemeece 0.313 69.5 0.888 100 II. Inorganic -++ creatine P ....... 0.935 100 | 0.428 48.2 III. Creatine P calculated as II—I 0.622 30.5 0.20 22:1 IV. Creatine P (isolated) ......... 0.362 iD efel 0.19 21.4 V. Pyrophosphate +- hexose P ... 0.215 8.3 0.16 17.4 VI. Acid soluble residual P....... 0.113 Be 0.11 eT WillieeNonsacidesoluble (2. ce... - 0.500 1.6 0.013 1.4 We denote as pyrophosphate + hexose P the inorganic phosphorus obtained after the hydrolysis of a fraction for one hour at 100° in 1 N. H,SO, after the removal of the inorganic and creatine P. 4 hours and then for 1 hour at 0°, is seen in Table 18. The muscles were immersed in liquid air and treated with cold 5 per cent trichloracetic acid. The extract was sucked through a glass filter into cooled Fiskn’s reagent. These operations took 2 minutes. In this experiment, we tested TaBLE 12. — Speciric ActTiviry oF PHOSPHORUS FRAC- TIONS EXTRACTED FROM THE GASTROCNEMIUS OF FROGS. TEmMpP.: 2° piace ihe | Fraction Specific activity experiment hours | : | ” : t Ibayoyyed nano, JE Spo aaa oo oer 100 4 Creatine nes + secusis © eperesl vets. 2 7 14.5 4 Iboterpeeaie IP) Ge cacoocogngs 100 4 CreatinetP we. seme Soda eres 21.4 3 Ihave! cosoooomnconadc 100 3 | Creatine P................. | 8.8 to what extent the inorganic P is precipitated by Frskr’s reagent. After precipitation of the ‘‘free’’ phosphate, 60 mgm sodium phosphate were dissolved in the filtrate, the phosphate was then precipitated and its activity tested. If the first precipitation was strictly quantitative. this second precipitate should be inactive. The counter registered 228 counts while, in the case of the first precipitate. 2500 counts were regis- 396 ADVENTURES IN RADIOISOTOPE RESEARCH tered. When the 228 non-precipitated counts are considered, the specific activity of the creatine P fraction works out to be 14.1 instead of 15.6. The same technique was used in the following experiments. The lowest value found for the percentage renewal of creatinephos- phoric acid molecules in the course of 4 hours at 0° is 9 and in the course of 17.5 hours 10 while, in most experiments, appreciably larger figures were found. The rate of renewal of the creatinephosphorie acid molecules Taste 13. — Speciric AcTIviry oF PHOSPHORUS FRAC- TIONS EXTRACTED FROM THE GASTROCNEMIUS OF A FROG 40 Hours AFreR INJECTING LABELLED SopDIum PHOSPHATE. TEMP.: 2° Fraction Brae aes NeWbevorteimane 1) scgaagnobooddobcUneuDUHC | 100 II. Inorganic -- creatine P............... | 62.7 III. Creatine P (calculated as II—I)...... | 33.7 iV. Creatine (determined!) =... 7... | 34.4 V. Pyrophosphate +- hexose P .......... 22.8 WilerA cid@solulble resi dialer peer sensielels 11.3 WADE, Whorey exonrcl collinleles 12 cooagscucuccouddc 2.1 Taste 14. — Sprciric Acriviry or PHospHorus FRAc- TIONS EXTRACTED FROM THE GASTROCNEMIUS OF A FROG 24 Hours Arrer INJECTING LABELLED SODIUM PHOSPHATE. TEMP.: 20° Sees —— ew Relative specific | Fraction | Bs activity | Ibnxoseaon Je GaeonocceboodocdacaGomoouoGd | 100 @reaitine ye ei werccasiews osretenetore obttncshals occvenerelsnens | 95 P hydrolysed in the course of 1 hour..... | 91 Acide solublemresvaluall Pe iene erence crete eve hen | 36 TaBLe 15. — Spectric Acriviry oF PHosPpHORUS FRAC- pions ExTRACTED FROM THE GASTROCNEMIUS OF A FROG 400 Hours arrer InsEctTING LABELLED SODIUM PHOSPHATE. TEMP.: 20° : | Relative specific Fraction aa activity Iboyergeenoe JP GooscpococbocogDode dc dad000 000 | 100 (inch) IOS A owadopuaTonaecolo Coe OaOUC ODD 99 P hydrolysed in the course of 1 hour..... | $7 INGOIPEOK! Howls San oondocuaccdqdao0udoe 16 CIRCULATION OF PHOSPHORUS IN THE FROG 397 TaBLeE 16. — Sprciric ACTIVITY OF THE PHOSPHORUS FRACTIONS OF THE MUSCLES or A Froae Kept at 20° ror 4 Hours anpD SUBSEQENTLY AT 0° ror 1 Hour —eeeeeeeeeeeeeeEeEeEeEeEeEeEeEeEeEeEeaeaeaeaeaeaeaeeEeEeEeEeEeaeaaeEe—Ee—e—E—=—=—EeEeEeEeEeEeEeEe— ‘ ‘ 7. Relative | P content | Counts Specific ; Fraction Ng | : : specific In mgm per min. | activity | | | : | activity | Left gastrocnemius ihoverRegnove: 12" soo golaclagnouponueedde doooud 0.055 | 427 7770 100 Oreahine wm beet eee te cretscchelocch oie. wa fersuens evsteseye ite 0.240 290 1210 15.6 Jakes doe JE) G oS 6 Soi CRE OOD CRUCICID IC rc Race 0.143 | 98 | 686 | 37 Right gastrocnemius -+- sartorius Trayeyeeg sy ated bs ante .¢ cial Cle Ckoao Caco HERO RCN ROIORTTE | 0.350 2500 7150 | 100 (Oise Mine, 12-6 o.pols Cociolec oOo UDOere ob DOOR | 0.908 | 1010 1110 15.6 Residue after 17 hours hydrolysis ....... 0.372 | 45 120 | 1.68 of the resting frog is, thus, quite appreciable even at 0° though not as high as stated in a preliminary note. At 20° the lowest figure found after 4 hours is 16 per cent. Rate of interaction of the plasma phosphate and the cellular phosphate of the muscle tissue In the preceding section, we calculated the rate of renewal. of the organic P compounds present in the muscle tissue by comparing the P content of the tissue inorganic P with the *P content of the phos- TABLE 17. — Speciric ACTIVITY OF THE PHOSPHORUS FRACTIONS OF THE MUSCLES oF 2 Froas Kept at 0° For 17.5 Hours Specific Frog Fraction | P content activity inorganic

siete laieic rose Sees ore 0.694 10.1 G) G. Hevesy and O. Resse, Nature 141, 1097 (1938). 398 ADVENTURES IN RADIOISOTOPE RESEARCH TaBLeE 18. — Speciric Activiry oF PHOSPHORUS FRAC- TIONS ISOLATED FROM DIFFERENT ORGANS OF A FROG. AFTER ADMINISTRATION OF LABELLED PHOSPHATE Durine 45 Hours at 20° Fraction | Specific activity PS asrin ey mew vayela «Siar tys eas aseotet et erste ehetolasetsteeeyenrte | 100 Conawisol: 1 Sonn cedcocupoccuoccnoopbuSoC 3.6 Castrocnemimus Morgamicw! Vrs stets cle velotelee = 4.9 Gastrocnemius creatine +- pyrophosphate P | 5.3 DGAVOr BP! ee ac % Ses tone teonRei ones olor ah/si/ol's) one! ops eleysTareusns | 10.1 1D) oh Vanya et in OG o Bin'o'd 08 OE o.oo OS OGG on 0.35 Dia ph ysis: (Bi ierga roan dete wettetekitc cre keieets 0.20 TasBLe 19. — Speciric Activiry oF PHosPHORUS FRAc- TIONS IsoLATED FROM DIFFERENT ORGANS OF A FROG AFTER ADMINISTRATION OF LABELLED PHOSPHATE During 4 Days at 22° Fraction Specific activity Pl ASTMaM Pe amicyac ke ate ane See Stas | 100 Gastrocnemius inorganic IP 4 .......55+.-:- | 8.3 Gastrocnemius creatine (2 Meise eee cae oe | 7.4 Sartorius total acid soluble 2 2a.ceceses + < | 8.5 Gastrocnemius phosphatide P ............ | 15 phorus extracted from the compound in question. In the following, we shall discuss the interaction of the plasma phosphate with the cellu- lar phosphate. This is clearly a very different problem, the rate of inter- action between the plasma phosphate and the cellular phosphate being determined by the permeability of the cell membrane. The low rate at which phosphate ions migrate through the membrane of the cells of the gastrocnemius is seen in Tables 18 and 19. In the course of 4 days at 22° only somewhat less than 1/10 of the P atoms present in the labile P compounds got replaced by plasma P. The molecules of the labile P compounds were repeatedly renewed during this interval and many P atoms present in the muscle cells interchanged lively; however, the interchange between cellular and extracellular P took only place on a restricted scale. The results of further experiments in which the activity of the plasma was compared with the activity of the muscle is seen in Table 20. To keep the plasma activity at an approximately constant level throughout the experiment, 0.4 ec. solution was injected at the start of the experiment and further 0.08 cc. every hour. As seen in Table 20, within 1 hour and 4 hours the activity of the plasma changes only CIRCULATION OF PHOSPHORUS IN THE FROG 399 slightly, the average being 108, taking the end value to be 100. The values obtained for the specific activity of the tissue P are seen in Table 20 and Fig. 3. In some cases, very low values were obtained for the distribution ratio of labelled phosphate between plasma and muscle tissue. The fact 200 Gastrocnemius 150 100 100xspecific activity 50 Tibia epiphysis Tibia diaphysis | 2: 3 4 Time tn hours Fic. 3. Specific activity of tissue P. that in these experiments frogs kept through the winter were used, the experiment being carried out in the spring, suggests the explanation that poor circulation may be responsible for the low values obtained. TasLe 20. — Activity OF DIFFERENT FRACTIONS OF THE FrRoG 1 Hour anp 4 Hours, RESPECTIVELY, AFTER THE START OF THE EXPERIMENT Temp.: 22° ’ ‘ | Fresh weight | P content | sp cUIyIeY See Time Fraction ie y | ’ | per mem activity in mgm in mgm : : - | 9 | > | fresh weight of P ee = — S31 | eee SP: WOLDS TC Pe ae a | 98.7 | 0.00344 116@) 116 ey eae GASELOCTICMIUUS 15). slai ajo @ elas = 861.8 | 1.3855 32.7 0.72 | BI DIPWYSISi eis Uitsn vos Nec oee a aate 94.1 | 6.100 224 0.12 Deu Siseuaraiers ici ocle save orale | 67.6 | 6.830 235 0.08 | | IE VeLSUNT EU eter tel tetetotet fetcketerstolsycie <1 1267.5 | 0.044 100 100 AL Tarbes: || EESindorenteranvls soo pbooonodGOr aeslosde 1300 | 93r5 2.03 | MEIN SIS horaravaeqevetactes sisters sche 38.8 3.61 755 0.28 hDiaphysisicc: sie eae ere | 60.1 | 6.410 578 | «(0.19 ©) Taking the 4 hours value to be 100. 400 ADVENTURES IN RADIOISOTOPE RESEARCH INCREMENT IN THE SPECIFIC ACTIVITIES BETWEEN 1 Hour anv 4 Howvrs | | Value corrected | for the change Fraction | Experimental ; 2 | | in the specific value | we : activity of the | | plasma : oe = ; ae <: Gastrocnemius ........ | 1.34 | 1.21 | | EAP Lny;SIS teres ele eis leneiseaNe 0.16 | 0.148 IDKeV Msg conosco o Doc | 0.11 0.102 32P content of the liver fractions As in the case of mammalia, in the frog the liver phosphate interacts at a much faster rate with the plasma phosphate than does the muscle phosphate. The fast rate of renewal of the acid soluble P compounds TaBLeE 21. — Speciric ActTiIviry OF THE PHOSPHORUS FRACTIONS EXTRACTED FROM THE ORGANS OF THE FROG Kepr at 15° AND THE ORGANS OF THE RasBsir 10 Hours AFTER ADMINISTRATION OF LABELLED PHOSPHATE Fraction | Frog | Rabbit PV ASI ae. age ercieuoye cuceuerencuener eo ovenene Ke1Gl's | 100 | 100 Gastrocnemius inorganic P ......... | 2.11 | 15.5 Gastrocnemius total acid soluble P . | 1.49 | 11.0 Ibi hse thavoyegshovte, IBS oinop onan oaoaOR dG | 12.9 85 iver pyrophosphate By n c Fs Skin oO Total tissue 50 40 °) 30 20 10 Se 7 15 30 45 60 min. c= = Heart ne = Total tissue fo} 6) on oO b oO 30 20 7 1S 30 45 60 min. Fie. 6. Effect of dinitro-cyclo-pentylphenol on the incorporation of 44C into total tissue of skin (a) and heart (b) of the mouse, 14C injected as CH,4COONa. was found by Goutp ef al. (2) to decline during the same time- interval to 1/3 of its initial value as well. We can expect a very rapid metabolic rate in the brain in view of the high oxygen consumption of that organ. The striking decline in the activity of the fat free brain with time is seen in Fig. 5. After 30 min only, the activity of the acid, soluble and protein fractions is less than half of the value observed after 10 min. The slight increase in the “C content demonstrated by the last 416 ADVENTURES IN RADIOISOTOPE RESEARCH part of the upper curve is presumably due to an increase in the activity of the protein fraction, which as shown by the lower curve is increasing with time. d) Skin and Heart Muscle For the skin and heart muscle we determined the effect of D.P-P. injection on MC incorporation into the total tissue only. The curves plotted in Fig. 6 (a) representing the results obtained for the skin, show a similar trend to the corresponding curves obtained for the skeletal muscle. The first part of the curves which show MC incorporation into the heart tissue 6(b), indicates an accelerated rate of 4C incorporation into the total organ under the effect of D.P.P. administration. The further trend of the figures is quite complicated, which may be due at least partly to the fact that we deal with the total tissue which contains fat and acid soluble fractions of appreciable “C content and less active protein and glycogen fractions. The trend of the curve denoting the change in the specific activity of the heart muscle tissue under the effect of D.P.P. markedly differs from that of the curves obtained for the skeletal muscle. This difference suggests a specific effect of D.P.P. on the heart. Summary The rate of incorporation of “C after injection of carboxyl labelled acetate in the liver, brain, skin, skeletal and heart muscle of control mice is compared with the values obtained with organs of mice to which dinitro-cyclo-pentylphenol (D.P.P.) was administered. Administration of D.P.P. increases the rate of incorporation of “C into liver fat and total liver tissue in the first 10 minutes of the experiment and decreases in the later part. Such a behaviour is expected if D.P.P. increases the rate of ace- tate metabolism in the liver. Administration of D.P.P. has a slight effect only on “C incorporation into brain fat or total brain tissue. The D.P.P. injected mice take up less ¥C in the skeletal muscle tissue and skin tissue than do the controls. References 1K. Biocu and D. RirrenserG, J. Biol. Chem. 159, 45 (1945). 2R. G.Govutp, F. M. Srnex, I. N. Rosenserc, A. K. SoLomon and A. B. Hastines, J. Biol. Chem. 177, 197 (1949). 3A, Prot, K. Brocu and H. S. Anxsr, J. Biol. Chem. 183, 441 (1950). 4G. Hevesy, Nature 164, 1007 (1949). 5@. Hevesy, R. Ruyssen and M. L. Beecxmans, Experientia in print. 6(. Heymans and H. Caster, Arch. int. Pharmacodyn. 50, 20 (1935). 7B. Terapa and M. L. Tarnter, J. Pharmacol. and exp. Therap. 54, 454 (1935). 8 W. F. Loomis and F. Lipmann, J. Biol. Chem. 173, 807 (1948): 179, 503 (1949). ° J. Forcu and D. D. Van SuisKeE, Proc. Soc. Exp. Biol. Med. 41, 514 (1939). Originally communicated in Hap. Cell Res. 3, 191 (1952) 41. DETERMINATION OF THE RATE OF RENEWAL FROM THE RATE OF DISAPPEARANCE OF LABELLED MOLECULES GEORGE HEVESY From the Institute for Research in Organic Chemistry, University of Stockholm THe rate of renewal of a type of molecules is usually calculated from the rate of incorporation of the labelled atoms of the pertinent precursor in the molecules considered. If we wish, for example, to known the percentage of desoxyribonucleic acid molecules of the rat spleen, which are formed in the course of two hours, we administer labelled sodium phosphate to a rat and two hours later we compare the specific activities of the desoxyribonucleic acid P, and inorganic P extracted from the spleen. If the ratio of these specific activities is found to be 0.02, the specific activity of the inorganic P remained constant during the experi- ment, and this P can be considered to be the pertinent precursor of desoxyribonucleic acid P, we can conclude that in the tumour 2 per cent of the desoxyribonucleic acid molecules are present which were formed during the experiment or, more correctly, that at least 2 per cent of these molecules were formed in the course of two hours. If it takes some- time that the labelled precursor reaches the site of desoxyribonucleic acid syntesis the first phase of the synthesis of this compound will not be indicated by the tracer and we shall correspondingly underestimate its rate of synthesis formation. If we protract the experiment, the specific activity of the inorganic P of the spleen decreases more and more and this decrease is followed by a decrease in the specific activity of desoxyribonucleic acid P. We can also calculate the rate of renewal by comparing the specific activities of desoxyribonucleic acid P and inorganic P in this declining activity phase of the experiment. The rate of loss of #2P by desoxyribonucleic acid molecules in the late phase of the experiment is independent of the precursor problem, but parallel with a loss *P by strongly active ‘‘old’”’ desoxyribonucleic acid molecules the formation of less active ‘‘new’’ molecules takes place for example in the spleen and the rate of incorporation of **P in these molecules is partly determined by the specific activity of the pertinent precursor. Thus by replacing the calculation of the renewal rate from 27 Hevesy 418 ADVENTURES IN RADIOISOTOPE RESEARCH increasing values of the specific activities of desoxyribonucleic acid with time by a calculation based on decreasing specific activity values with time, we cannot fully eliminate the difficulty arising from the lack of knowledge of the pertinent precursor of desoxyribonucleic acid P. We meet, however, very different conditions when faced with the task to calculate the rate of renewal of a carbon labelled compound from specific activity or similar data. 14) of rapidly metabolized carbon compounds such as acetate, glucose and so on, is within a comparatively short time exhaled to a very large extent as carbon dioxide. Correspondingly the rate of loss of “C by fatty acid molecules in a later phase of the experiment in which, for example, labelled acetate was administered to the rat, is no longer a resultant of the disappearance of “old” strongly labelled molecules and the formation of “new” less strongly labelled ones but, at least in the first approxima- tion, the result of the decay of labelled molecules only. By following the rate of decrease of the “C content of fatty acids extracted from the organs of the rat we can thus calculate the renewal rate of fatty acid molecules in an analogous way to that, in which we calculate the period of decay of a radioactive body from its “decay curve.” In the present note the calculation of the renewal rate of the fatty acids of the liver from the rate of decrease of the “C content of the fatty acids in the rat injected with acetate labelled in the carboxyl group is described. EXPERIMENTAL 44 rats weighing 190 to 244 gm were injected intraperitoneally each with 0.2 ml of phys. sodium chloride solution containing acetate labelled in the carboxyl group of 10—26 / C activity. The animals were killed by decapitation and the total fat, fatty acids, neutral and phosphatide fatty acids, and also cholesterol of the liver, secured as described previously. (2, 8) The activity of the samples was compared without conversion into barium carbonate. RESULTS In an early investigation, using deuterium as an indicator, STETTEN and Boxer (12) found the half life-time of fatty acids of the rat liver to amount to 1.9 days. Rrrrenpere and Brioc3g, (11) feeding 130) label- led acetate, observed a more rapid incorporation of C0 into saturated than into unsaturated fatty acids of the mouse liver. In a more recent work, Prnt et al. (8, 9) compared the “C content of the fatty acid carbon (at the end of the experiment taking many days) with the average value THE RATE OF RENEWAL FROM RATE OF DISAPPEARANCE OF LAB. MOLECULES 4]9 of the 4C content of the precursor carbon prevailing during the experi- ment. The saturated fatty acids of the liver were found to reach half of their maximum isotope concentration in less than one day, the unsatu- rated acids in about two days. When investigating the effect of muscular exercise on the incorporation of 4C into liver fats, the present writer (5) found after the lapse of 41% hr the 4C content to be only one third of that detected 20 min after injecting labelled acetate into the mouse; however, according to the above men- tioned data, in the course of two hr. the loss of “C by liver fatty acids should be less than 10 per cent. Obviously, a fatty fraction is present in the liver which has a much shorter half-life time than one day. Ruyssen, BrECKMANS and the author (6) investigated the renewal rate of fatty acids in the liver of the mouse shortly after administration of labelled acetate. They obtained the result that the “C content of the liver fatty acids increase rapidly during about the first 30 minutes, this increase being followed by a rapid decrease. Furthermore, BEECKMANS and Exxiort (2) could demonstrate that in both the saturated and unsatura- ted fatty acid fractions this early increase is followed by a rapid decline in the “C content. The liver must thus contain one or more rapidly renewed fatty acid fractions. To discard the effect of this rapidly meta- bolizing fatty acid fraction we studied the decrease in the “C content of liver fatty acids one or 14 days after administration of labelled acetate only and followed it till the lapse of 4 or 4; days. As we in our determination of the renewal rate disregard those labelled fatty acid molecules which are formed during the experiment we have obviously quite apart from the above considerations to wait for about 1 day after injecting labelled acetate before securing the samples. GouLp et al. (4) found that after the lapse of three hr. only 85 per cent of the acetate 4C injected into the rat is already exhaled as “CO. In the later phase of the experiment the rapid loss of the exogenous acetate 144) may to some extent be compensated by formation of labelled endo- genous acetate. Pru et al. (9) kept the activity of body acetate of the rat at a constant level by feeding labelled acetate; in these experiments an increase in the body acetate activity could be observed after the lapse of ten days which was due to the formation of endogenous acetate of appreciable C™ content. Under our experimental conditions (injection of the labelled acetate at the start of the experiment) the activity of endogenous acetate formed during the experiment was, however, negli- gible. In Fig. 1 the decrease in the specific activity of fatty acid carbon with time 1} to 4% days after injection of labelled acetate is plotted; also data obtained by Prut ef al. (9) for the increase in the percentage of labelled fatty acids with time are seen. The figure contains thus both ’’rise curves” and ’’decay curves’. From the data of Pram et al. 27* 420 ADVENTURES IN RADIOISOTOPE RESEARCH follow that the half life-time of saturated fatty acids is less than one day and that of unsaturated fatty acids about two days; our data indicate practically the same result, 0.8 and 2.2 days, respectively. A closer coincidence is hardly to be expected in view of the fact that even when comparing fatty acid turnover in the liver of rats of the same race, age and weight, very appreciable fluctuations appear. Ame- Incorporation of acetate “C into saturated X polyacies Fic. 1. Rate of incorpo- Sx ration of acetate C4 50 me (Pru et al.) and rate of ~ loss of 4C by saturated fatty acids extracted from 25 ne ae rat liver. (‘Rise curve”’ and ‘“‘Decay curve’’.) 0 \ 2 3 4 days 1005 X . \ incorporation of acetate *C into 7 unsaturated fatty acids ~ Troasi«*S 1 > as sO ~~ Loss of 94C by unsaturated fatty acids Fig. 2. Rate of incorpo Rik ration of acetate “C aS ms (Prat et al.) and rate of loss of 4C by unsaturated O fatty acids extracted from \ 2 3 4 days rat liver. lioration of purification or measuring methods would hardly lead to more accurate mean renewal times, such could be obtained only by investi- gating an appreciably larger number of animals. The significance of the data obtained is restricted, as both the satu- rated and unsaturated fatty acids represent a mixture of components having different turnover rates, Some components of the unsaturated fatty acid mixture as linoleic or linoleic acid are not synthesized in the animal organism, and thus are not labelled. While PrHL and Bioc# [8 | state that the linoleic acid content of the rat liver is almost negligable, these authors find neutral fatty acids to contain 16 per cent, phospha- tide fatty acids and 8 per cent of linoleic acid. From the fatty acids ex- tracted from the liver of rabbits Porpsak and Brrckmans [10] found that when the acetate was given to the animals for 20 hr the specific THE RATE OF RENEWAL FROM RATE OF DISAPPEARANCE OF LAB. MOLECULES 42] activity of phosphatide fatty acids was larger than that of the glyceride fatty acids, in experiments of longer duration, however, no significant difference in the @C content of the two types of fatty acids could be observed. In all our experiments 1 to 4 days after injecting the labelled acetate the specific activity of the total fatty acids of neutral fat of the rat liver was found to be a few per cent higher than the specific activity of phos- phatide fatty acids, while according to Prot and BuiocH (8) 3 days after feeding labelled acetate to rats kept on lipid free diet, the phos- phatide fatty acids are 6 to 13 per cent more active than the total fatty acids of neutral fat. As mentioned above, investigation of the incorporation of “C into fatty acids of the liver shortly after the administration of labelled acetate to mice revealed the existence of a rapidly metabolising fatty acid fraction. A rapid decrease in the 4C content of total fat, neutral and phosphatide fatty acids of the liver of the rat is also observed shortly after the ad- ministration of labelled acetate but is less pronounced than in the case of the mouse liver. Some results obtained when injecting several rats each of which being killed at a different time after intraperitoneal injection of the labelled acetate are shown in Tables Ia and b. These and similar results are to be interpreted cautiously, among others because — in contrast to the experiments on mice where pooled livers of a large number of mice where extracted — in the present inves- tigation each point indicates the result obtained when extracting a single liver only. The fatty acid metabolism may strongly vary from animal to TABLE 1. — PERCENTAGE INJECTED ACETATE MC PRESENT IN 1 mgm OF NEUTRAL TOTAL FATTY ACIDS OF NEUTRAL FAT EXTRACTED FROM THE LIVER OF 182—190 gm Rats |Percentage injected ~ 103 in Time after injection lity mgm total fatty acids of neutral fat a) Aa oatt ote) BARE pry can cootess Seo SO GRR sng cB renbecsiner ke 51.4 LET Og es ytd oie wt: 63.0 EE OR Poy ce Ph re Yall | IZA ae Bea ees oe Danes DA ee srtaines tt. Lk ee, | 28.3 | _b) ZO GO ee ae Sey ceeccesen cls 56.2 504 Se een 66.0 DENLE i eee | 79.2 MW 5 . volume expressed in Time blood water in parts | of 1 cc. heavy water ee igo ‘ . percentage of body per million in the circulation : | weight Diy WSEGa aie 1 034 506 34 S80) sece ae 857 | 612 40.8 J Seemann ere 794 661 44.1 8) anmiidlg goo || 719 | 7127 | 47.8 - a | : | ; eZ) ANU es 570 921 | 61.4 7, Oemaine 7 495 | 1 056 | 70.4 133) senha, 56 490 1 070 71.3 2222, TTT = tall eat Ree a ~ L Pr OL = EOP a | | | | | | Z a. aye le oievra Ons ¢ rT Stimulated ....... | 415.7 135.4 32.6 2040 4.92 6.237 , . | 72 pe . Non-stimulated ...) 476.2 | 154.1 0.216 32.4 2110 | 4.04 US Lao TABLE 2. — Wercut or Cat 2.95 keM. Lerr Nerve STIMULATED FOR 5 Min. AT THE Enp or THE SEconD MiInuTtvE 0.3 MinuicuRIgE 2?2P InsEctTED INTRAVENOUSLY. INoRGANIC PLrasmA P = 4.42 MGM P. Cc. gm inorganic plasma P gm ugm P extracted from non-stimulated nerve stimulated and non- stimulated nerves ugmnerve P to 1 stance Ratio of activities of 1 gm Fresh weight of sciatic nerve in mgm Dry weight in mgm dry stimulated and 1 Ratio of activities of 1 P content in “gm mgm P per gm fresh nerve Percentage activity of 1 Percentage dry sub- | ape | ; | | Stimulated! seo. | Satpal NOE 2 33245) 1320) 4218 2.46 1.98 | 0.080 Non-stimulated ... | 394.6 WE BAN) |) IB 3.36 0.040 | the activites of 1 w4gm phosphorus extracted from the stimulated and non- stimulated nerve and finally the percentage activity of 1 “gm total nerve phosphorus relative to that of 1 wgm inorganic phosphorus. It will be seen from figures in Tables 1 to 4 that the irritated nerve in each case took up more #P than the non-irritated nerve, the ratio 456 ADVENTURES IN RADIOISOTOPE RESEARCH 32p varying between. 1.18 and 2.46. For the ratio of the quotiens —— for y q sip the irritated and non-irritated nerves figures varying between 1.10 and 1.98 were obtained. TABLE 3. — WeicutT oF Cat 2.4 kKGM Lerr NERVE STIMULATED FOR 5 MIN. AT THE Enp or THE SEcoND Minute 0.4 Minzicurte 2?2P Ingectep INTRAVENOUSLY ] 2 ] a mene | 8a. w ¢ | = | | > } © %O& mo] Was ° bo I S | ap || tel tel le Seah = Ay 5 a ee aa fae 1) ee eee ela | 80 = a a Pi eae |) eoreiol es A Sey a) 9s =] = gq hh o — wm i + eS 2s = = a Sb 2 | 298/298 | Ba 2 | Bs SI ba = So eB eS ey eas = ae s A | g | fey teh Ide Se pipes oo P=} © bo «1 He lean ea) oe & i Qp Se =| w | SHalasasagd| wa a | a i = = D } a os ane an S | 5 Es = 2 = Ce el Staa | oo P | ra ® | =I rat ora Caoo} O9NR 6 g b = 8 rm | Zo} 12) Sy |) Gel Seer 5 2 8 2 = S 2 a BUAl|BAas aot oo ee A = oy Ses ea} | a an Fé ee aan | t Pe | fod | - | ~ Stimulated ....... — 117 — | 1474 | 1.50 | 0.109 | } | Non-stimulated ... | = — WSS 1350 | | 0.075 | | | Taste 4. — WeiIcuHT or Cat 1.8 kam Lerr NERVE STIMULATED FOR 5 Min. AT THE END OF THE SECOND MINUTE 0.4 MiniicuRi&£ 22P INJECTED INTRAVENOUSLY. DURING THE Last MINUTE INTERMITTENT STIMULATION (2.5 Sec. IRRITATION AND 2.5 SEc. Rest) ' | | Lo} 4 4 | Sell I | = | ee} | Bhs 5 be | iS} | q 7p 1Gey SS | 2S Ee ey o | oo | =| a | Loh) ~feu 20> St | ow ie =| | ete a |} nos no™~ o res ~ |= | =z = | ops Og = Fos =, fs| | sa 8 et pede Cale yoy js! =) Ge | Ss ar) c | g s=ylaey ys | Pom) way ae UF ee | om ouSis || oP Tile ek Set ie tee || St a se re) to A : | se eet $s OS ame OS tay | ea eat One ae s S| = Pals Sys snes i sy 34 oo & ae = as Se Fe. Sy Seats, So) eee so Eo E 5s = 2% Wises Ka el |) 8 Se Zo ee, oe 3 + oe g | Sees | aE £3 2 2) = | © 7 1S sel pes i? Sele an eeoeeey 3 Bqh |2<87/ 385 2 | A | | ce Soe Stigniaiisy | ee) So de : | Fas me AS = | a A = L = —— = = = = | | | ee | Stimulated ....... po — | LIS 7 — | 1048); — | 1355) O57 3 | E | 2 | = Non-stimulated ... | — 101 | — | 1074 ==) 0.089 | | If we wish to know the percentage of P present in the nerve tissue (cf. above) that was taken up in the course of the experiment, we must compare the specific activity of the nerve P with the average value for the inorganic P of the plasma. This magnitude does not necessarily correspond to the amount that has penetrated from the plasma into the tissue during the experiment, as it is conceivable that a part of the P migrating from the plasma into the tissue cells has found its way back again. Considering that the amount of plasma P located in the nerve tissue constitutes only a small percentage of the total P content of the tissue and in view of the rapid participation of intruded phosphate in phosphorylation processes, we may suppose the amount of labelled P (plasma P) located in the nerve tissue to be practically identical with that penetrating from the plasma into the nerve tissue during the ex- periment. Assuming that the specific activity of the plasma inorganic P of the cat declines after intravenous injection at the same rate as in THE EFFECT OF EXCITATION ON NERVE PERMEABILITY 457 a rabbit of the same weight, the average specific activity of the inorganic P in the plasma works out in a 3-minute experiment to be about 3 times that of the experimentally determined amount. For the amount of P penetrating during 3 minutes into the resting nerve tissue (cf. Tables 0.08 0.09 i 3 and 4) we hence obtain the values —— and 5 as percentage activity e e of nerve P to plasma P. In the experiment recorded in Table 1 the 0.22 approximate value —_— is obtained. As discussed above the average 5 specific activity of the plasma P is higher than its end value which was determined experimentally. To account for this decrease, we have in the experiment taken 5 minutes to divide the percentage activity of nerve P to plasma P with the figure 6 which should give an approx- imately correct value. The amount of plasma P passing into 1 gm fresh resting nerve tissue in the course of 1 minute proves on the average to be 0.01 per cent of the total P of the nerve tissue, thus about 0.4 wem. The amount of phosphorus penetrating in 1 minute into the nerve tissue is smaller than the corresponding amount penetrating into the muscle cells. KaucKarR ef al. (1944) estimate that, in the course of 1 minute, 1 wgm phosphorus penetrates into the cells of 1 gm fresh muscle tissue of the rat. A simular figure is reported by Hrvesy and H. v. EULER (1942). Sodium Permeability In these experiments labelled sodium with an activity of about 0.5 millicurie was injected into the circulation. The cat was killed 5 minutes after the injection. In each experiment, as seen in Table 5, the irritated nerve was found to take up more 24Na than the resting nerve. TaBLE 5. — UPrake oF “74Na By Scratic NERVE Activity of 1 mgm dry nerve as per- | Ratio of uptakes | Centage of activity of 1 mgm dry Weight of cat in kgm | by irradiated and | plasma | resting nerves : ae = irritated resting i} | rc nctelerete tyerara tasers 3345 1.40 0.77 0.55 i fd (RAIS react STE 1.85 1.56 1.38 | 0.81 fl Bt evict Ate Hf) 1.06 2.67 2.48 EV ries one aes eke 222, 1.98 4.80 2.42 AV ispees hots aartene Gorekone 1.85 1.32 1.79 1.36 Assuming 1 mgm. dry plasma to contain 40 “gm. sodium and the mean activity of sodium during the 5-minute experiment to amount to3 458 ADVENTURES IN RADIOISOTOPE RESEARCH times that of the activity measured at the end of the experiment, 1 mgm dry resting nerve takes up on an average 0.2 “gm sodium per minute. In view of the very rapid change in the *4Na content of plasma which follows intravenous injection, this figure represents only a rough estimate of the amount of sodium taken up by the nerve. Potassium Permeability In these experiments labelled potassium chloride having a specific activity of 0.3 millicurie was injected. The injection took 1 minute, the ‘at was killed 2 minutes later. The results obtained are seen in Table 6. TaspLe 6. — Urpraker or ?2K sy Scratric NERVE Activity of 1 mgm dry nerve | Ratio of uptakes as percentage of activity of 1 | ’ Weight of cat in kgm by irritated and mgm dry plasma resting nerves - irritated | resting | IB nedancocadoc 3.0 2.29 6.62 | 2.88 Blip te ake aes 1.9 1.51 8.60 5.83 By a similar consideration as put forward in the case of sodium we arrive to the conclusion that in the course of 1 min. about 0.02 “gm pot- assium penetrates in 1 mgm dry nerve tissue or 6 “gm in 1 gm fresh tissue. It is of interest to compare this figure with the data recently obtained by Hopexin and Huxtery (1946) when determining the number of moles of potassium which leak through 1 cm? of membrane of the axons from Carcinus maenas in one impulse (1.7 + 10-1") and the amount of potassium re-absorbed during the period of recovery. When the ex- ternal potassium concentration is increased threefold, t3.10~1° Y mol em? sec-1 or 0.7 wgm potassium per minute were found to be re-absorbed. Bromine Permeability In these experiments 30 mgm bromine as sodium bromide, with an activity of 0.2 millicurie was injected. TABLE 7. — Uptake oF ®2Br sy Scratic NERVE | ae : Activity of 1 mgm dry nerve Ratio of uptakes| “S percentage of activity of 1 mgm dry plasma Weight of cat in kgm by irritated and resting nerves Ti ara | irritated resting | fy patie | 1 ae bianion tdi oc 1.98 1.34 3.4 2.9 THE EFFECT OF EXCITATION ON NERVE PERMEABILITY 459 Assuming, as in the case of 24Na, that the average plasma activity during the experiment amounts to about 1/, of the determined end acti- vity, 1] mgm. dry resting nerve took up 0.01 “gm. bromine. Similar amounts of radiobromine were found to be taken up by the basal ganglia, cere- breal cortex and the medulla oblongata of the cat. In no single case did the stimulated nerve fail to show an enhanced uptake of the ion investigated. When we compute the mean of all deter- minations (13 cases) of the ratio of uptake of various ions by the stimu- lated and the resting nerves, the value 1.55 + 0.10 is obtained. The uptake by the stimulated nerve was also found to be enhanced when the cat was curarized before stimulation in order to avoid muscular movement. Summary The effect of stimulation on the amounts of phosphate, sodium, potassium and bromide taken up by the sciatic nerve of the cat was investigated with the aid of radioactive isotopes as indicators. Stimulation was effected by condensor shocks at a rate of 50 per sec. giving maximal motor reactions. In each case investigated, including the curarized animal, the stimulated nerve was found to take up more labelled ions than the resting nerve, the mean ratio of the uptakes being 1.55 + 0.10. References J. L. Botiman and E. V. Frock (1943) J. Biol. Chem. 147, 155 L. Haww and G. Hevesy (1941) Acta Physiol. Scand. 2, 154. G. Hevesy and H. v. Ever (1942) Svensk Vet. Akad. Ark. Kemi 15, A No. 15. G. Hevesy and O. Resse (1946) cf. A. Kroau (1946) Proc. Roy. Soc. B 133, 195. A. L. Hope@x1n and A. F. Huxtey (1946) Nature 158, 376. H. Katcoxar, J. DEHLINGER and A. MruteEr (1944) J. Biol. Chem. 154, 275. T. R. Noonan, W. O. Fenn and L. Harce (1941) Amer. J. Physiol. 132, 612. Originally Communicated in Acta Physiol. Scand, 16, 20 (1948) 46. NOTE ON THE INORGANIC PHOSPHATE OF BLOOD PLASMA GRACE DE C. Exttiot, L. HAHN anp G. HEVESY From the Institute for Research in Organic Chemistry, Stockholm. In tracer work with radiophosphorus, we often add labelled sodium phosphate to blood plasma or inject it into the circulation and follow by means of radioactive measurements the path of the tagged phosphate into the red corpuscles, their passage through the capillary wall or other phase boundaries. For this type of work it is of great importance to know whether the inorganic phosphate, added or injected, shows the same behaviour as the endogenous inorganic phosphate present in the plasma. Should that not be the case, the calculation of the amount of phosphate penetrating from the plasma into the red corpuscles, for example, from the distribution date of the ®2P added and the inorganic phosphate of the plasma would clearly lead to wrong results. To inves- tigate if and to what extent added and endogenous inorganic phosphate show a different behaviour, we added to heparinized human or cat plasma a tracer dose of sodium phosphate of negligible weight and, after rotating the plasma at 37° for 1 hour in a thermostat, we electrodialysed the plasma. At intervals varying between 1 and 18 hours, we took samples and determined by radioactive measurements and colorimetric determi- nations, respectively, at what rate the added labelled phosphate and the endogenous inorganic P, respectively, is removed from the plasma. EXPERIMENTAL The dialysator applied was of the Hahn—Tiselius (1943) type. As membrane a thin sheet of cellophane was used. To 20 ml heparinized plasma an equal volume of physiological sodium chloride solution containing about 10 y labelled phosphate of 44 microcurie activity was added. To keep the salt content of the plasma at constant level a diluted sodium chloride solution was added at intervals. Through the electrode cells a borate buffer solution of pH 7.6 circulated. The potential applied to the ice-cooled dialysator was 20 Volts; the current intensity amounted to 30 milliamp. The inorganic phosphate was extracted from the plasma with 10% trichloroacetic acid. We carried out extractions both at 0° and at 20°. Imperfect extraction would not influence our results, as aliquots of the same extract were applied to the radioactive measurements and to the colorimetric determinations. NOTE ON THE INORGANIC PHOSPHATE OF BLOOD PLASMA 461 In some cases, the amount of the total acid-soluble P present and its activity were determined as well. In these experiments, the acid-soluble extract was ashed and an aliquot used to colorimetric determination of the total (inorganic ~- organic) acid-soluble P present, another aliquot to radioactive measurements. RESULTS Numerous experiments were carried out leading to the result that a difference is present between the behaviour of the inorganic phosphate added and the endogenous inorganic phosphate, but the difference is 100 60 w - Percentage activity O = Percentage inorganic P 40 Percentage inorganic P resp. activity 20 O 5 10 15 20 Time in hours Fic. 1. Rate of disappearance of the endogenous inorganic P content of human plasma and of the radiophosphorus added. not pronounced. As seen in Fig. 1 and other similar figures obtained by us, the radioactivity of the plasma sample declines at a somewhat more rapid rate than its inorganic P content. After the lapse of 18 hours 10% of the inorganic P is still present, but only 4% of the activity added. Thus, about 6% of the endogenous plasma inorganic P do not dialyse and, consequently, do not get into exchange equilibrium with the labelled phosphate of the plasma. In some experiments, up to 10% of the endo- 462 ADVENTURES IN RADIOISOTOPE RESEARCH genous inorganic P were found not to dialyse. The non-dialysing in- organic phosphate is presumably combined with proteins. That a minor part of the inorganic P of the plasma is bound to proteins is also made very probable by previous work in which other methods were used and which are discussed below. As seen in Table 1, the total acid-soluble P content was found to be appreciably higher than inorganic P content, about ¥/; of the acid- soluble P being present in organic binding. The specific activity of the total acid-soluble P was found to be lower than the corresponding value of the inorganic P, indicating that the labelled phosphate does not interchange or interchange to a minor extent only with the organic P present in the plasma. PREVIOUS WORK ON THE PRESENCE OF PHOSPHATE-PROTEIN COMPOUNDS IN THE PLASMA MACHEBOEUF and S@RENSEN (1925/27) in their studies on the com- position of egg-albumin were led to the result that between the large complexes which are constituting egg-albumin molecules few phosphorus containing complexes are present; these are bound so firmly to the other groups of the molecule that they have to be considered a con- stituent of the latter. The amount of protein-bound phosphorus was found in albumin, resp. globulin to amount to 7.5 resp. 2 mgm per gm total nitrogen. This phosphorus could not be separated from albumin, resp. globulin, by electrodialysis. S#RENSEN (1925/27) investigated on similar lines the phosphorus content of serum-albumin and serum- euglobulin. Serum albumin and euglobulin were found to contain 2—40, resp.0.15—0.3 mgm P pergm total nitrogen. In contrast to the phosphorus of egg proteins only a small percentage of the phosphorus of the serum proteins could be precipitated by alcohol. This and other observations induced SgrENSEN to regard the serum protein P as an accessory con- stituent of the serum proteins only. No conclusions can be drawn from these investigations if this accessory phosphorus is getting into exchange equilibrium with the comparatively large amounts of inorganic phos- phorus simultaneously present in the plasma or not. Masker et al. (1942) analysed fractioned centrifuged horse serum and found that the inorganic phosphorus concentration increases progressively with the protein concentration. Differences of 0.14 and 0.15 mM of P per kilo water were obtained between the top and bottom fraction, the top fraction containing 11% less, the bottom fraction 12% more inorganic P than the unfractionated plasma which contained 1.3 mM per kilo of H,O. These authors conclude from their results that phosphate-protein compounds norma'ly occur in horse serum. NOTE ON THE INORGANIC PHOSPHATE OF BLOOD PLASMA 463 That protein bound inorganic phosphate is present in small amount only in the serum of the dog is also borne out by the work of Smrru e¢ al. (1943) who determined the percentage of ultrafilterable inorganic serum phosphate in 13 cases. The average percentage of ultrafilterable in- organic P found in their experiments works out to be 96%. TABLE 1. — InNoRGANIC P CONTENT AND ACTIVITY OF PLASMA SAMPLES SECURED FROM THE DIALYSATOR AT VARIOUS TIMES Mime GacHours ORaUG P conrent in geiatiye Ey a relative units eo [Rete ACC relative waite a mgm%%)| relative units = 200 counts per min Olaeriasc ants | 100 100 Cy ee | 69.3 63.2 3) onugcGonscGac 58.5 56.7 ib Amoeoono mood or | 47.1 45.6 Cae ae | 42.6 39.2 (Gao dwooodadogs | 32.7 29.3 a Ee ce a a EEN | gO Pe LSU aea ceysyeieters sei « 9.8 3.2 SL eatateWel ol uelcrer svete | ell | 3.9 LOSS Pe mem eta: | lest | 4.4 | | Specific activity of | aides | Total acid-soluble P Wout Bonclssollelollsy IE Specific activity of in- organic P Wereteeicdcporatevats (exe | ei | 0.80 1 Extracted at 20°. 2 Extracted at 0°. Not only a combination with proteins, but also the formation of colloidal phosphate will prevent ultrafiltration of inorganic P or its removal into the electrodialysate. GROLLMAN (1927) found in an early work that, while the inorganic P of normal pig serum is entirely ultra- filterable, a successive increase in the calcium content of the plasma from 9.4 to 32.2 mgm % makes the inorganic phosphate less and less ultra- filterable and, finally, only 5% are found in the ultrafiltrate. Ample evidence was brought by different authors that only excessive quan- tities of calcium phosphate salts lead to the formation of detectable amounts of a colloid complex (cf. ScHmrpT and GREENBERG, 1935 ; McLEan and HiInprRicus, 1938). GOVAERTS (1943, 1947) compared the specific activities of the in- organic P of plasma and urine shortly after intravenous injection of labelled phosphate into the dog. In the first 11%, hours, the specific activity of the urine P was found to be greater than the corresponding value of the plasma P; after the lapse of that time no difference was found. GovAERTs interprets these results as indicating that the greater 4164 ADVENTURES IN RADIOISOTOPE RESEARCH part of the acid-soluble P of the plasma does not get into exchange equilibrium with the injected inorganic phosphorus. Our results do not contradict those of GovaEnrts. Phosphate identi- fied after treatment of the plasma with trichloracetic acid as inorganic phosphate but actually present in the plasma in a labile low molecular organic binding may show a similar electrodialytic behaviour as does inorganic phosphate. Proofs of the presence of a labile phosphorus com- pound in the plasma is yet outstanding. Summary To determine whether inorganic labelled phosphate added to plasma gets into exchange equilibrium with inorganic phosphate present previously, labelled phosphate of negligible weight was added to human plasma. The plasma was electrodialysed, samples were taken from the dialysator at intervals, their inorga- nic phosphorus content and its radioactivity determined. No pronounced difference was found in the rate of disappearance of the inorga- nic P content and of the radioactivity of the sample. As, however, after the lapse of 18 hours, only 4% of the original activity, but 10% of the original inorganic P content were present, we have to conclude that a small percentage of the plasma inorganic phosphate, possibly combined with proteins, does not interchange with the labelled phosphate added. References J. Govarrts (1943) Bull. Acad. Med. Belg. (6), IX. J. Govaerts (1948) Arch. Int. Pharmacodyn. 75, 201. A. GROLLMAN (1927) J. Biol. Chem. 72, 565. L. Haun and A. Tisexius (1943) Biochem. Z. 314, 336. M. Macuesoeur, M. Sorensen and S. P. L. Sorensen (1925—1927) Medd. Carlsberg Lab. 16, No. 12. F. C. McLean and M. A. Hinpricus (1938) Amer. J. Physiol. 121, 580. A. V. Masxket, A. Coanutin and S. Luprvie (1942). J. Biol. Chem. 143, 763. C. L. A. Scumipt and D. M. GreenBere (1935) Physiol. Rev. 15, 297. P. K. Surru, R. W. Ottayos and A. W. WINKLER (1943) J. Clin. Inv. 22, 1431. S. P. L. SoreNsEN (1925—1927) Medd. Carlsberg Lab. 16, No. 8. Originally published in Nature 132, 952 (1938) 47. FATE OF THE SULPHATE RADICAL IN THE ANIMAL BODY A. H.M. Atens jun. and G. HEvEsy From the Institute of Theoretical Physics University of Copenhagen PHOSPHORUS enters as phosphate in the numerous compounds in which it is to be found in the animal body; in connexion with the investi- gations carried out in recent years concerning the fate of ingested phos- phorus atoms in the organism, it seemed to be of interest to determine whether or not, in the course of the numerous metabolic processes in which phosphorus is involved, the phosphate radical exchanges its oxygen content with other oxygen atoms present in the body. This question could be answered by injecting into an animal sodium phosphate which contained heavy oxygen (!8O) as an indicator and then deter- mining if the phosphate recovered in the urine, for example, contained more than the normal amount of 180. As, however, it was recently found! that “heavy-oxygen phosphate” can be obtained by dissolving sodium phosphate in “heavy-oxygen water’ and vice versa, it is apparent that the oxygen atoms present in phosphate radicals exchange their places freely in water and there can be scarcely any doubt that the probability is extremely small of a phosphate radical leaving the body coupled to the same oxygen atoms with which it entered. Sulphate ions, on the other hand, have been found? to exchange oxygen atoms either not at all or at a very slow rate with neutral water, even at 100° C., and it seemed of interest, therefore, to investigate whether sulphate ions during their circulation in the body participate in chemical reactions which loosen the oxygen bonds sufficiently to make an oxygen exchange possible. In the experiments we wish to report here, sodium sulphate containing heavy oxygen was prepared from heavy-oxygen water, kindly presented to us by Prof. Urry? having a density 740 parts in a million greater than that of normal water. The reaction used for the preparation of the “heavy sulphate” was that which takes place between SO,Cl, and heavy- oxygen water in the presence of traces of iodine as a catalyst. 1 gm. of the dry material, converted into 50 cc. of solution, was injected into a rabbit. The urine of the rabbit was then collected for 24 hours, its sulphate content recovered as barium sulphate, the oxygen content 30 Hevesy 466 ADVENTURES IN RADIOISOTOPE RESEARCH of the latter converted into water, and the density of this deter- mined. The preparation of water from the oxygen of the sulphate was carried out in the following way. The barium sulphate preci- pitate was dried at 400° C. in a stream of nitrogen and then reduced with purified carbon at 900° C.; the gases evolved were mixed with a great excess of hydrogen and stored over oil in a gasometer; and, finally, the gas mixture was led over a nickel catalyst at 310° C. and the water formed collected. The density determination was kindly carried out by Mr. O. JAcoBsEn, using Linderstrom-Lang’s floating-drop method. Should the sulphate oxygen, during its stay in the animal, enter into exchange reactions with other oxygen atoms present in very great excess in the body, the oxygen of the heavy radicals would be replaced by normal oxygen atoms and the water prepared from the sulphate re- covered from the urine would show the density of normal water. If, on the other hand, the sulphate ions injected retain the oxygen atom with which they start, the water prepared from the urine sulphate should show an excess density of 370 parts per million if no secretion of normal sulphate took place. The water prepared from the sulphate isolated from the urine after injecting heavy-oxygen sulphate has shown a very appreciable density excess — 240 parts per million. When com- paring this value with the one calculated on the assumption that no exchange of sulphate oxygen took place, we must consider the following fact. Besides the heavy-oxygen sulphate — 0.84 gm of sodium sulphate being secreted in all during the day following injection — the urine contains also sulphate, even when no injection is given, the amount of which we found to correspond to 0.23 gm per day. The latter is normal sulphate and its presence reduces the density excess of the water prepared from the urine sulphate. From the high density found for the water prepared from urine sul- phate, one must conclude that most of the individual sulphate ions injected into the rabbit are recovered in their original form, and from this it follows that at least the greatest part of the sulphate administered leaves the body unchanged. References 1. BLUMENTHAL and HERBERT, Trans. Faraday Soc. 38, 849 (1937) 2.8. C. Darra, J. N. E. Day and C. K. InGo3p, J. Chem. Soc. 1968 (1937). 3. HuFFMANN and Urey, Ind. Eng. Chem. 29, 531 (1937) 4. Mantan, Urry and BLEAKNEY, J. Amer. Chem. Soc. 56, 2601 (1934) 467 COMMENT ON PAPERS 42—47 WHEN first investigating the rate of exodus of 24Na from the plasma of the rabbit, we were struck by the rapidity with which this takes place, by the swiftness of the interchange between vascular and extravascular sodium. After the lapse of 1, min. about half of the former was replaced by the latter. We know today that the figures obtained in such investigations indicate only the lower limit of the speed with which such an interchange takes place. The injected labelled sodium has first to diffuse into the capillaries before leaving the circulation, and the interchange between vascular and extravascular sodium which takes place during this early interval is not indicated by the tracer. When carrying out our first experiments we had at our disposal potassium of very low specific activity only. To carry out an experiment on the rate of exodus of 47K from the circulation, we had to inject very rapidly 10 ml of physiological potassium chloride solution. Though the blocking effect of potassium on heart beat is well known, we were much impressed by the momentary fatal effect of such a rapid injection of a potassium chloride solution. When in possession of potassium of higher specific activity, we compared the rates at which sodium and potassium ions leave the circulation. That the latter was found to take place at a more rapid rate is presumably mainly due to the following fact: The labelled ions leave and re-enter the circulation. The re-entry is facilitated in the case of sodium by the restricted extravascular pool, sodium being mainly an extra- cellular element in contrast with potassium. Potassium ions which left the plasma and entered tissue cells have less chance of re-entering the plasma than have sodium ions. As shown in paper 43 in 44 min half of the labelled water leaves the plasma of the rabbit. This is, for the reason mentioned above, an upper limit of the half- time that the water molecules remain in the plasma. After the lapse of 1 hr an almost complete equipartition of the water molecules between the circulation and the extravascular water was observed. Since radio-sulphur was not available at that date for the study of the fate of the sulphate group in the organism, we applied with 18O labelled sulphate {paper 47). 30* Originally published in Nature 133, 495 (1934) 48. DIPLOGEN AND FISH G. Hevesy and E. Horer From the Institute of Physical-Chemistry, University of Freiburg In recent months we have been carrying out experiments on the be- haviour of fish in heavy water. We find that goldfish (Carassius auratus) behaved quite normally in the heavy water in which they were kept. As heavy water was to be used as indicator of normal water, we had to carry out our experiments in water containing only 0.5 mol. per cent of diplogen, and it is therefore still possible that a higher concentration of this isotope in water exerts effects upon fish. The aim of our experiments was to follow the exchange of water between the fish and their surroundings, using heavy water as an indi- cator of the movement of the total water. The use of radioactive iso- topes for such purposes is well known. While the latter are practically chemically identical, and as such are entirely trustworthy indicators, that is not the case with the isotopes of hydrogen. Heavy water is, therefore, only to be used with great caution as an indicator of ordinary water. However, when using very dilute solutions of heavy water, we may expect that the rate of exchange of heavy water molecules between the fish and its surroundings will not be very different from that of the normal water molecules. By measuring the speed at which the heavy water enters the body of the fish we can therefore conclude at what rate approximately the exchange of water between the fish and its surroun- dings takes place. Some twenty fish having a total volume of about 10 cc. were kept in about 60 cc. of water containing 0.5 mol. per cent diplogen water. After TasBLE 1. — Rate or ENTRANCE OF HEAvy WATER INTO FISH | | Decrease expected in the case of | | 33 p.c. | 30 p.c. Decrease of the heavy ner | a ; ; equal distribution of the heavy Time in hours water content of the A ; | ; | water between fish and surrounding surrounding water water I | 1 | SZC: 30 p.c. If | 4 | 32 p.c. 290 p-c- III | 15 DIPLOGEN AND FISH 469 a certain time the fish were removed and the decrease of the density of the surrounding water was determined. The fish were then placed in normal water, and the rise in the density of the latter due to the entrance of heavy water molecules leaving the body of the fisch was determined. The results are shown in the accompanying tables. TaBLE 2. — Raves or Loss or HEAvy WATER BY THE FISH | | | | | | Decrease of the 4 ; pig be | Decrease expected in the case of ‘ / Initial heavy | heavy water con- | Ae Ak : Time in | s . | equal distribution of heavy water content | tent of the fish c : hours | , : | water between fish and surroun- of the fish after the . | f | ding normal water experiment | | | liye mal 0:20 p-c.8' |) 68 pics |) 51 pie. , | an II 4 0.27 p.c. 68 p.c. | O7-p:c: INGE), 1 0.26 p.c. 86 p.c. | 86 p.c. It follows from the above that, at least in a small fish, within a few hours all the water molecules leave the body of the fish, making way for water molecules derived from the surrounding water. It should be borne in mind that most fish contain about 80 per cent water. 470 ADVENTURES IN RADIOISOTOPE RESEARCH COMMENT TO PAPER 48 Urery’s discovery of heavy water was bound to impress the tracer-minded scien- tists, although their number was very restricted in those days. The present writer at once approached Professor Urry who most generously mailed a few litres of water containing 0.5 mol. per cent heavy water. In view of the great sensi- tivity with which the density of water can be determined, this strongly diluted heavy water sufficed to study the interchange between the water molecules of the goldfish and the surrounding water, and also to carry out studies described in paper 48, and presented more in detail by Hevesy and Horgr (1934). Within 1 hr an almost quantitative interchange between the water molecules of the fish and those of the surrounding water was found to take place. The amount of deuterium incorporated into the organic components of the fish, compared with the amount of deuterium entering as deuteriated water into the fish, was found to be small. The present writer intended, if successful in obtaining more concentrated heavy water, to study this type of incorporation. Shortly after- wards, SCHOENHEIMER and RirtTENBERG embarked on the study of this problem and solved it in a masterly way (cf. p. 403). The discovery of artificial radio- activity induced the present writer to abandon his original plans and to find out if and to what extent the mineral constituents of the skeleton are in a dyna- mic state ef. Radioactive Indicators in the Study of Phosphorus Melambolism in Rats. In paper 48 it is stated that the goldfish behaves in the same way in the heavy water employed in the experiments described as in tap water, though it may behave differently in more concentrated heavy water. In experiments with HAGexKvist carried out in recent years (1958), we found that the life-span of the fish investigated was reduced from years to 10 days when kept in 40 per cent heavy water. When the fish were placed in 50 per cent heavy water, they tried to escape by jumping out from the vessel in which they were kept. When our paper on the interchange of the water molecules of the goldfish and those of the surrounding water was published, Urry had not yet proposed a name for heavy hydrogen, while RurTHERFORD discussed the possibility of calling it diplogen. This explains why we chose ‘‘Diplogen and fish”’ as the title of our paper. References G. Hevesy and E. Horer (1934) Z. Physiol. Chem. 225, 28. Received by the Editor on 20 March 1933. G. HAcexvist and G. Hrevesy (1958) Acta Radiol. 49, 321. Originally published in Kgl. Danske Videnskabernes Selskab. Biologiske Meddelelser. 14, 5 (1939) 49. INTERACTION OF PLASMA PHOSPHATE WITH THE PHOSPHORUS COMPOUNDS PRESENT IN THE CORPUSCLES G. Hevesy and A. H. W. ATEN gr. From the Institute for Theoretical Physics, University of Copenhagen LIST OF SYMBOLS p —total amount of plasma-phosphate: ce —total amount of acid-soluble phosphorus in corpuscles Sp — Specific activity of plasma phosphate (activity per mgm phosphorus); S; —— specific activity of inorganic phosphorus in corpuscles: Se — average specific activity of total acid-soluble phosphorus in corpuscles; a, — total activity in plasma phosphate; ae —totalactivity of acid-soluble phosphorus in corpuscles; Sp — value of s, at the end of an experiment; S; — value of s; at the end of an experiment; S. — specific activity of acid-soluble organic phosphorus in corpuscles at the end of the experiment: S. — value of s, at the end of an experiment, etc.; A, — total activity of plasma ester at the time; A, — total activity of plasma ester at the start of the experiment; a —coefficient of penetration; k -—rate-constant of the monomolecular decomposition of hexosephosphate in blood. Tue distribution of inorganic phosphate and of inorganic acid-soluble phosphorus compounds between plasma and corpuscles deviates from equipartition. This difference in the distribution of the phosphate ion could be due to the fact that during the life time of the corpuscles equi- librium between its contents and those of the surrounding plasma is not yet reached, but it is more probable that we are faced with a case in which the partition coefficient of the ion in question between corpuscles and plasma actually differs from unity. The distribution coefficient of inorganic phosphate and also that of the acid-soluble organic phosphorus compounds between plasma and 472 ADVENTURES IN RADIOISOTOPE RESEARCH corpuscles fluctuates within wide limits. HALPERN® investigated the in- organic P content of the plasma and corpuscles of rabbit blood in 33 cases. In 29 cases the inorganic P content of the corpuscles was found to be less than that of the plasma of equal volume, the ratio between the inorganic P content of the corpuscles and that of the same volume of plasma varying between 0.86 and 0.38. In our experiments we find an average content of the plasma inorganic P amounting to 7 mgm % and of the corpuscles inorganic P to 4.5 mgm % making the above mentioned ratio equal to 0.64. When taking into account that the water content of the corpuscles amount to only 70% of that of the plasma, we obtain for the distribution coefficient of the inorganic P between the corpuscle water and plasma water a value differing not much from unity.©) While the determination of the inorganic P of the plasma is not difficult, as the acid-soluble plasma P is mostly composed of phosphate ions, the analysis of the corpuscles often gives less reliable results. Some of the organic phosphorus compounds present in the corpuscles may decompose“ in the course of the formation of additional inorganic phosphate. On the other hand, when the corpuscles have to be obtained quite free of plasma as in our experiments, it is necessary to wash them with a suitable solution free of phosphate, for example with an isotonic sodium chloride solution in the course of this operation some phosphate can be lost from the corpuscles by a diffusion process. In view of the great importance the plasma phosphoric esters play in Rostison’s theory of bone calcification, he and his collaborators” made a careful study of the amount of phosphoric ester present in the plasma from human and rabbit bloods and ascertained an average value of about 0.5 mgm%. The problem in which we were interested was the determination of the rate at which phosphate ions and also the phosphoric ester molecules of the plasma penetrate into the corpuscles and vice versa. The usual procedure employed to obtain information on the permeability of the (JL. Hatpern, J. Biol. Chem. 114, 747 (1936). In this paper earlier literature on this subject is to be found. R. T. Brain, H. O. Kay and P. G. MarsHayi | Biochem. J. 22, 628 (1927)] found the inorganic P content of the human blood plasma to be 4.1 mgm. %, that of the corpuscles 2.4 mgm. %. () R. T. Brain, H. O. Kay and P.G. MarscHatu [Biochem. J. 22, 629 (1928) ] find for human blood the same distribution coefficient as found by us for canine blood, namely 0.91. (3) MarrLAND, HansMAN and Rosison [ Biochem. J.18, 1152 (1924)] have shown that, of the blood is made acid to pH — 7.3, there is hydrolysis, if made alkaline to pH — 7.35 there is for a short time synthesis of the esters; this, however, soon gives place to hydrolysis and to a corresponding increase of the inorganic phosphate; comp. also H. Lawaczeck, Biochem. Zeitsch. 145, 351, 1924. (4) R. Rosrson, The Significance of Phosphoric Esters in Metabolism, p. 68., New York (1932) Comp. also R. T. Brarn, H. O. Kay and P. G. MarsHA Lt, loc. cit. INTERACTION OF PLASMA PHOSPHATE WITH THE PHOSPHORUS COMPOUNDS 473 corpuscle membrane to phosphate ions is to introduce sodium phosphate into the plasma and to investigate if and to what extent the phosphate and phosphoric ester content of the corpuscles is increased. By using this line of attack Hatpern found that at 3° inorganic P does not enter or leave the blood cell to any appreciable extent in the course of 9 hours. Above 23° a very slow, at 37° an appreciable penetration of the additional phosphate into the corpuscles was observed. A very convenient way of the study of exchange between phosphorus components present in the plasma and the corpuscles is opened by the application of labeled (radioactive) phosphate. By introducing active sodium phosphate of negligible weight into the plasma, all the phosphate ions present in the latter get labelled and, if after the lapse of some time radioactive phosphorus compounds are found to be present in the cor- puscles, we can conclude that these penetrated during the time in question from the plasma into the corpuscles. We carried out experiments both in vivo and in vitro, introducing active sodium phosphate into the plasma and investigating, after the lapse of a few hours, the activity and the concentration of the inorganic phosphate and also of the phosphoric esters present in plasma and corpuscles. In other cases active hexo- semonophosphate was introduced into the plasma and the activity and concentration of the above mentioned P compounds were measured. In view of the very slow rate of the formation of labelled ’’non-acid soluble’? phosphorus-compounds present in the blood (phosphatides and phosphorus containing proteins) ascertained in our former work” we left those substances out of consideration in this investigation. Some of the acid-soluble phosphorus compounds, of which a great variety occurs in the corpuscles, were found to be labelled to a large extent after the lapse of a short time. The problem which first occurs is whether the labelled organic phosphorus compounds (phosphorus esters and others which we will in what follows denote as ‘‘phosphorus ester’) are formed within the corpuscles from active inorganic phosphate or whether active esters diffuse from the plasma into the corpuscles. As we will see later, the labelled esters found in the corpuscles are, at least to a large extent synthesized within the corpuscles. As to the nature of phosphoric esters present in the corpuscles, the presence of various compounds has been recorded, such as adenosintri- phosphate, hexosephosphate, triosephosphate, mono- and diphospho- glycerate, glycerophosphate, and phosphopyruvate. The composition of the corpuscles of different animals was found to be markedly different ; while the corpuscles the blood of sheep contain,®) forexample, 80% of esters which are hydrolysed by boiling 1 n. HCl within 3 hours, the @ L. Hann and G. Hevesy, Mem. Carlsberg Lab. 22, 188 (1938). @) H. v. Euter and K. M. Branpt, Z. physiol. Chem. 240, 215 (1936). A474 ADVENTURES IN RADIOISOTOPE RESEARCH corresponding figure amounts, in the case of rats’ blood, to only 30. The average ester P content of human blood corpuscles is stated to be 20 mgm per 100 ce. blood, of which about 68% is present as phospho- glycerate, 21% as hexosephosphate, and 11% as adenyltriphosphate.™ Few data are available as to the phosphorus ester content of the plasma and its composition, the presence of small amounts of hexosemono- phosphate being recorded®). The phosphoric ester content of the plasma varies within wide limits, the average value being about 0.5 mgm% In the normal human plasma values varying between 0.0 and0.9 mgm% and an average valve of 0.33 mgm% were recorded®). DIFFUSION OF PHOSPHATE IONS INTO THE CORPUSCLES Radioactive sodium phosphate containing a negligible amount of phosphorus is added to 10 cc. of heparinised rabbit blood. The sample is shaken in a thermostat at 37° under a mixture of oxygen and carbon dioxide, after the lapse of few hours plasma and corpuscles are sepa- rated by centrifuging, the corpuscles washed 2—3 times with a physio- logical sodium chloride solution. The acid soluble components of the plasma and also those of the corpuscles are isolated in the usual way (extraction with ice-cold trichloro-acetic acid). While the acid-soluble fraction of the plasma is practically (90% or more) composed of in- organic P, the corpuscles contain mostly organic phosphorus compounds and, in addition, some inorganic P. The latter can be isolated by pre- cipitation as ammonium magnesium phosphate. The organic phosphorus compounds present in the filtrate are then converted into inorganic salts and precipitated as such. When carrying out such experiments, we find that in the course of few hours an appreciable part of the labelled plasma inorganic phosphate penetrates into the corpuscles. At the same time we find a formation of labelled organic phosphorus compounds in the corpuscles. What actually happens is that the individual inorganic phosphate ions of the plasma diffuse into the corpuscles and are converted in the latter into phosphorus esters. The question, which now occurs, is which is the faster process, the diffusion of HPO; into the corpuscles or the ester formation. This can be decided by comparing the specific activities of the different phosphorus fractions isolated from the corpuscles. Such a comparison is seen in Table 1. After the lapse of only half an hour about half the @ BE. Warwec and G. Stearns, J. Biol. Chem. 115, 567 (1936). S. E. Kerr and A. Antakxt, J. Biol. Chem. 121, 531 (1927). ) Comp. R. Rosrson, The Significance of Phosphoric Esters in Metabolism p- 69. New York (1932). () R. T. Brain, H. D. Kay and P. G. MarsHatu, Biochem J. 22, 635 (1928). INTERACTION OF PLASMA PHOSPHATE WITH THE PHOSPHORUS COMPOUNDS 475 ester was in exchange equilibrium with the inorganic P present in the corpuscles. The re-synthesis of the acid soluble organic P compounds must thus be a fast process and, from the fact that with increasing time the ratio between the specific activity of the ester P and inorganic P of the corpuscles only slightly increases and yet strongly differs from 1, TaBiLe 1. — AcTIVATION IN ViTRO OF AcCID-SOLUBLE PHOSPHO- RUS PRESENT IN THE CORPUSCLES Relative specific activities of corpuscles’ phosphate and of organic acid-soluble phosphorus present in the corpuscles, taking the spe- cific activity of the plasma-phosphate at the end of the experi- ment to be = l. SSS | : Rel. spec. activity Rel. spec. activity | eee ae ee | of corp. - acid- | of corp. phosphate | Pr nie ee age Time | pa soluble P Ss = & eo : 0) SeewhNs Sooanoe 0.25 O14 Rabbit G a ens SX0) teat. coo me A 0.38 0.21 | aX) Tai Gogo éo5 0.27 0.11 Rabbit H DOsmine. oh5.022 0.46 0.23 IU Getbht, “Gog aoud 0.69 0.36 we must conclude that only a part of the diverse organic phosphorus compounds present in the corpuscles is renewed and thus activated in the course of the experiment while the other part, composed of diphosphoglycerate, hexosephosphate and other compounds remains, at least practically, inactive. This conclusion is supported by the results of the following experiments. Instead of destroying the total esters we hydrolysed them with In. HCl or H,SO, for 100 min at 100° and deter- mined the specific activity of the hydrolysed P. While the average ester P secured from the corpuscles of a rabbit (G, comp Fig. 1) was found to have a specific activity amounting to 55% of the corpuscle inorganic P the corresponding figure for the hydrolysable ester was 80%. In the case of another rabbit (H), the figures were 53° and 100% respecti- vely®. From the facts mentioned above, it follows that the exchange () Under these conditions, diphospho-1-glycerate and also hexosediphosphate are only hydrolysed to a negligible resp. small extent (G. Warwec and E. Srrarns, J. Biol. Chem. 115, 567 (1936). ©) The difference between the rate of activation of the ‘hydrolysable’ and ‘“nonhydrolysable” fractions is still better brought out when comparing the specific activity of the pyrophosphate, obtained from adenosintriphosphate after 7 min hydrolysis, to that the residual P, as found in a recent investigation the result of which will be published shortly. 476 ADVENTURES IN RADIOISOTOPE RESEARCH reaction between inorganic phosphate ions and the hydrolysable esters is a very fast process. That the inorganic P present in the corpuscles does not reach exchange equilibrium with the plasma phosphate in the course of a few hours is due to the fact that a large part of the active phosphate ions are incorporated into the organic compounds of the corpuscles, while simultaneously non-active phosphate ions are freed to take place of the active ones and “dilute” the active inorganic phosphate 0,5 Fic. 1. Ratio of the specific activities of the total acid- soluble phosphate and the inorganic phosphate present in the corpuscles.