TH,E PREStiRVATION OF YELLOW FEVER VIRUS* .BY W. A. SAWYER, M.D., W. D: M. LLOYD, M.D., AND S. F. KITCHEN, M.D. (From the International IIealtlz Division of the Rockefeller Foundation and The Rockefeller Institute for Medical Research, New York) (Received for publication, March"28, 1929) In preparation for a comparative study of the characteristics of strains of yellow fever virus from West Africa and Brazil, we found it necessary to seek a reliable method of preserving the virus over a considerable period of time. The ideal method for our purposes wouId be one which would permit the sending of active virus in small sealed containers on sea voyages lasting over a month, and would also allow storage in the laboratory for several months without serious loss of virulence. Such a method was found among several processes previously applied to the preservation of other viruses, and it has now been in use long enough in connection with our experiments to show that it can be depended upon. In the first work with cxpcrimcntal yellow fever in monkeys, Stokes, Bauer and Hudson (1) kept the virus alive for a time by passing it directly from animal to animal. They made inoculations by taking blood from the sick monkey, adding it to a citrate solution to prevent clotting, and injecting the mixture into a healthy animal. This method would have required an excessive number of monkeys to preserve even one strain of the virus during an extended study. The interval between successive inoculations had to be short for several reasons: the titrated blood could not be stored more than a few days without danger of Iosing the virus, the incubation period was usually brief, and the blood for trans- mission was drawn as a rule on the first day of illness. The difficulty was later diminished by these investigators by transferring the infection from monkey to monkey by means of mosquitoes, and thus lengthening to several weeks the time interval between successive monkeys in a series. This method is far from ideal, however, for the mosquitoes require much care and occasionally develop n high * The studies and observations on which this pnpcr is based were conducted with the support and under the auspices of the International Health Division of the Rockefeller Foundation. Through the courtesy of The Rockefeller Institute in New York laboratory facilities were made available. 1 2 PRESERVATION OF YELLOW FEVER VIRUS mortality which might result in the loss of the yellow fever strain. The USC of mosquitoes, moreover, necessitates special equipment and adds an extra hazard for the espcrimenters. Sellards (2) succeeded in preserving yellow fever virus in frozen monkey liver for twelve days while transporting it from West Africa to England. Keeping tissues in a frozen state rcquircs conditions difficult to arrange, cspecinlly on long voyngrs. The method, howver, was one which promised to be highly useful, if tests provctl it rclial)lc for cstcnsivc pcriotls, cspcci:~lly where tlic use of mos- quitoes is impracticable. TJlc Yellow Fcacr C!Y&S Used For the purpose of comparing the vducs of different methods of preserving yellow fever virus WC h~vc brought togcthcr in tables the results of our obscrvntions with two strains which proved to be prnctically identical in their charnclcristics and very sxtisfnctory for experimental work. Both were highly virulent and fairly constant in their cffccts. For one, the "French" strain ("I?" in the tables), we arc indebted to Dr. A. W. Sellnrds, who 1~x1 brougilt it to America -. from West Africa via ISnglnntl. It lid been trnnsmittccl to monkeys by Mnthis, Scllnrtls, and Lnigrct (3). `I'hc other is the "Asibi" strain (((L`\" in the tnblcs), with which stokes, IZauc`r, ,zn(l Hudson did most of their work. It was originally obtained 011 the Gold Coast, West Africa, and was sent to us from Lagos, Nigeria, by Dr. Henry Beeuwl~es, Director of the West iZfric,zn Y'ellow Fever Commission of the Rockcfder Foundation. In each instance the virus studicd was prescrvecl in the blood or the liver tissue of J1crctrc1r.s ~k.711.s monkq~s. After an animal \vns inoculated, the rectal tempcraturc \v:1s taken twice a cln>v, and when the temperature rose to 40oC. for the first time, n specimen of l~lootl \vns tnken. `l'hc nnimal was :mesthetizctl u4th ctlier ant1 the I)lood wns clrx\vn from the heart. The sick animals were allo~~tl to die or wcrc cliloroformcd when moribund. \\`hcn ncctlctl 2s n source ol lrirus, liver tissue 1~3s rcmovccl at nccrops~~ \yitli prc- cautions for stcrilit>y. In 211 crises \vllcre 311 :lnini;ll is rcportccl as h:~\~ing (lid of yCllO\v fwcr, the gross lesions of this cliscxx \vc`rc fountl on nccrops\., :lntl ch;lr- actcristic*chnngcs were seen in microscopic preparntions of tissues, cspcci;ill~ those of the liver. In n few instances the prcscnce of tulxrculosis or d:xntery com- IV. A. SAWYER, W. D. M. LLOYD, AND S. F. KITCHEN 3 plicated the picture and made it difficult to determine the onset and course of the yellow fever. Such cases were excluded from consideration in this study. Subcutaneous or intraperitoneal injections of fresh first-day blood were almost invariably followed by the development of experimental yellow fever, even when small doses were used. As results with material which had been kept less than nine days arc of little interest to this study, they have been omitted from the tables prescntctl, csccpt for the two partial titrations of fresh citratcd blood of Monkeys 9.5 and 11s included in Table I. All injections of virus into monkeys in this study were intrnperitoneal except in the case of the two titrations, in which they were subcutaneous. \\`e have noticed no difference between the results following these two methods of inoculation. The liver tissue proved less relinblc for inoculation than the first-day blood. This is as would be espcctcd when one considers that the virus diminishes rapidly in the blood after the first ~;LJ' or t\vo of fever, and probably decreases in the liver during the lalter cl:lys of tlic tlisc:lsc. `l'hc duration of the illness vnrics in the different cases ant1 the concentrntion of virus in the liver at death probably varies also. \\`l~:ttevcr t11c csplanntion, the liver tissue, though usually in- fectious, provctl to be less consistently so than the first-day blood specimens. Escept when specialI\. noted in the tnbIes, no specimens which failed to infect are included unless tfle original fresh materia1 was proven infectious by the inoculation of some other specimen derived from it. ..In a very few cases the inoculated monkq~ had prclViousl>r rcceivcd material related to ~rcllow fever, most often human blood two or three weeks old, and had shown no reaction. On account of the remote possibility- that the animals had been protected by this material the letter `(U," for "used," is placed after the number of each of these monkeys in the tables. The weights of monkeys are omitted from the tables, as moderate fluctuations seemed to play no noticeable part in determining the results of the tests. Specimens of monkey blood sent us from Lagos, Nigeria, are included in the tnbIes and identified by specimen numbers preceded by the letter "S" instead of the numbers of the monkeys from which the virus was obtained. The amounts of blood or liver are stated in the tables in terms of the fresh material. For example, 0.02 gm. of ciricd Mood would be recorded as 0.1 cc., because blood loses about S 1 per cent of its weight in drying. 2 cc. of glyccrinated blood containing 1 cc. of actual blood would be recorded as 1.0 cc. Incubation periods are shown in the tables as th; interval between inoculation and the first day of fever; for esnmple, if an animal was inoculated in the after- noon and its first fever occurred in the afternoon of the second day after inocula- tion, the interval \voultl IX recorded as 2.0 days, but if the first fever was in the morning of the third (I;\\., the period would be 2.5 days. The intcrvnl between inoculation :1ntI the time of denth or recovery is simil:~rly recortlcd. In the case Of recovery the interval is rncasurecl to the lnst rccortltd tempernture of 40oC. or over, ant1 is f0lhwl by- the letter "IL" All deaths recorded were from yellow fever. 4 PRESERVATION OF YELLOW FEVER VIRUS In order to limit As far as possible the number of monkeys needed for our investigations, most of the tests of preserved virus for infectivity were made as required by other experiments. This will e?iplain the apparent lack of plan in some of the groups of observations presented. TABLE I Preservation of Yellow Fever Virus in Citratcd, Clotted, or Glycerinafed Blood rMethod of preservation source, monkey ani strain = I %:"o"d" t fresh' IZI > - Interval in days to Age of specimer Result of test Death inoculation Fever or recovery Obser-: vation period Citration 95 A 95 A 95 11 118 I: 118 I: 118 F 18 1: 32 1: 66 I s 103 A S 10412 s 10.5 A 66 F s 106 A GG I: 117 I: 6.5 A cc. 0.6 0.08 0 005 0.9 0.07 O.OOG 0.6 3.5 2.0 1.0 1.0 1.0 days 0 0 0 1 1 1 22 22 30 35 35 35 No. 99 100 to1 129 117 119 31 u 29 u 112 42 4.5 43 dnys 1.5 5.0 2.0 4.5 10.0 11.0 6.0 7.0 F Immune 2.0 4.0 -* 8.5 - - Not tested - - Not tested - - Succumbed - - Not tested 3.5 5.5 - - Not tested Clotting 2.0 1.0 2.0 1.0 1.0 30 35 111 46 24 24 30 28 28 30 - - Succumbed 21.0 24.5 Glycerinn- 30 105 u 6.5 10.5 tion 60 181 12.0 16.5 100 179 - - 29 Succumbed * No fever was observed before the temperature became subnormal, 6.5 days after inoculation. = - Citratcd aid Clotted Blood The most convenient method of making direct transfers of yellom fever virus from monkey to monkey was to prcvcnt clotting by the addition of sodium citrntc t.0 the freshly drawn blood and to inject the misture intr:Ir'critonc:11ly or subcutaneously. On the clay Of the bleeding, or even the nest day, the virus content of the citrnted l,lood seemed to remain high, as illustrated by the first six items in Table I. W. A. SAWYER, W. D. 3.5. LLOYD, AND S. F. KITCHEN 5 With longer storage the results became increasingly less dependable, and in the six tests of material from 22 to 35 days old there was only one successful inoculation (Table I). In preparing the citrnted specimens, from two to five parts of freshly drawn blood were mixed with one part of stcriic 2 per cent sodium citrate solution, usually in 0.9 per cent sodium chloride or Locke's solution. The specimens were then stored in the refrigerator at from 1" to 4oC. The three ci trated specimens received from Lagos (S 103, S 104, and S 105 in Table I) consisted of the blood of the same monkey plus nn equal quantity of 2 per cent citrate solution. The specimen which proved to be infectious after the long journey from Lagos to New York was one of the two (S 103 mti S 104) which had been kept in the ship's rcfriggator. The other specimen of the same blood (S 105) had been stored at a higher tcnipcrnture in the cool room. If infectious blood is drawn and aliowed to clot, it will retain its virulence for about the same length of time as if it had been citrnted, ~2s suggested in Table I by the one failure nftcr 30 days in storage and the one success after 35 days. The infectious specimen of clotted blood (S 106) hnd come from Lagos 2nd was from the same bleeding as the citrntcd specimens S 103, S 104, and S 105. It had been stored in the ship's cool room. It will be noticed that the incubation period in the case of the successful inoculation with clotted blood was 21 days, the longest in our experience. Although the virus cnn be kept in cIottcci blood about as Iong as in titrated blood, titration is the more convenient method, for it is necessary to grind the Clot with sand and physiologicni salt solution in ;L mortar before inoculation. If there is a sufficient amount of serum, this can be used without the clot. In our esperience it :rppcars to be the esception for stored titrated or clotted yelIow fever blood to retain its virulence as long ns a month, and the method is obviously unsuited for prolonged storage. \Yc made no experiments to determine whether the virus would persist longer in titrated blood under vaseline or oil, or in clotted blood kept continunliy frozen. Glycerirtaled Blood-or Liver As glycerine had proved useful in the preservation of a number of viruses, we matIe preparations of blood and Ever, containing 50 per cent of glq'cerine and tested them for the presence of active virus after storage in a refrigerator at 2' to 6oC. The results are shown in Tables I and II. The blood specimens were prcpnrcd by adding freshly drawn blood to ati equal volume of glvcerine in n test tube and shaking vigorously. The mixture w,zs a thick fluid which usually clotted gradunlly in storage until it wns of a firm, jelly- 6 PRESERVATION OF YELLOW FEVER VIRUS like consistency and required grinding with sand to bring it into solution in physiological sa1t solution. The liver preparations were of two kinds. The second specimen of glycerinntcd liver listed in Table II consisted of pieces of liver tissue &Ix! of ;pccimer days 12 1.5 1s 20 21 30 30 100 g:ln. 11- 1 t- 1 -I- 3 + I+ 1 .o 1.0 1.0 1 .O 1 .o 1 0 If 1.0 0.1 0 0 1 0.1 0.1 1.0 1.0 1 0 1.0 Olmrvn~ tion pcriotl days 48 Result of test Ik!atl1 or inoculation I.`c \`er RXOVCry -- - - Immune 6. 5 9.0 s.5 0.5 1: Not tcstcd 5.5 s.5 4 5 7.0 5 0 7.5 - - Not tested - - iVot tested ;:o .:o (Jo 1y 6. 0 11.0 12.5 15.0 I< Immune 9. 0 12.0 1: Immune 1.5 5.0 1: Not tested 30 30 30 30 23 9 19.0 22.5 9 5.5 s.0 9 - - Not tested 30 10.0 14.5 GO - - Not tested 1 63 - - Succumbed 100 7.0 S.Oi 100 4. 0 5.0 150 6. 5 s.0 - - - " So control inoculation \a5 mntle \Vitli the original liver tissue of this monkey to prove it infectious. i Chloroformed earl>? in the tliscnsc to secure material for pathologic31 study. $ Tubed \\ith dcium chloride. - droppet into SO per ccn t gl>.cerinc in 1,0&e's solution. The other specimens wew prepxed by grinding liver in an cqu:ll vo~umc of L~ckc~~ solution, using sand, and then adding to the resulting suspension an equal volume of glycerine. The results with gl>*ccrin:ltcd blood lvcre distinctly better than with citrnted or clotted blood, as \Vill be wcn in the tnblcs. The samples of glycerinated blood TV. A. SAWYER, W. D. M. LLOYD,' AND S. F. KITCHEN 7 which had been kept 30 and 60 days were shown to contain active virus, but the incubation period for the 60-day specimen was prolonged. The loo-day speci- men failed to infect and produced no immunity. With the liver specimens we obtained infection in every instance, but reCovery .followed infection from one of the two JO-dny specimens and from the 60-&y and loo-day specimens. The virus of yellow fcvcr nppnrcntly can bc preserved in 50 per cent gIycerinc for a longer time than in titrated or clotted blood. Infec- tion has even been obtained with glyccrinated material after storage for 100 days, but there is a distinct falling off in virulence with stor- age after 30 days. Stronger conccntrntions of glycerine have not been tried. I~rrm,r IAfY The method of prescrvirlg tlw virus in frozen liver proved to be reliable for periods of at Icxst one month (Table II); but the single specimen tested after 100 days failed to infect. Pieces of liver in test tubes were plnccd in the freezing compartment of an ordinary household electric refrigerator. 11s the temperature in the compart- ment was - 12oC. the liver tissue was frozen hard. On one occasion the refrigerating rn~~chineq~ did not function and the tempcra- ture in the comp~~rtment rose suflicicntI>V to permit the liver tissue to thaw. As `a result the spccitncns then in storage, the first four in Table II and also the seventh, were under rcfrigerntion for t\vo dnys without being frozen. The freezing mcthotl is simple and valual,le when mxchinery is at hand for main- taining a constant tempcrnture low enough to keep the tissues frozen hard. It was not until U-C triecl drying blood or liver under vacuum in the frozen state and storing it in scnlctl lulx2s thnt ~vc wcrc able to preserve the virus of yellow fever for long periods. The method lens suggested to us by. Dr. `I`. Sf. Rivers, who had found it sntis- fnCt0q in the prescn3tion of sc\rcrnl Iyiruscs including vaccine virus ant1 \rirw3 III. \Yhnt was csscntinlly the same mcthotl had been used by ITarris anal Sh~cl~cll (4) in 1911 in preserving ral)ics virus; and. Rous (5) rcportecl tlint in tlic snmc year the nctive virus of :L transmissible snrconln of folvls was stored successfully in drictl and powdered tissue by ?rfurphy. Xlorc recently the method of drying \vhilc in the frozen state KL~S npp~ietl to the prcser\.ation of bacterial cultures by S\\rift (G). 8 PRESERVATION OF YELLOW FEVER VIRUS Our usual procedure in preserving the yellow fever virus in dried blood com- mences with the drawing of 1.5 to 20 cc. of blood from the heart of a monkey under ether ancsthcsia on the first day of fever. The needle of the syringe is then thrust through several layers of gnuzc which cover R sterile, cylindrical glass ev:+ornting dish 12 cm. in dinmctcr and 6.5 cm. deep, and the blood is ejected. the thin layer of fluid blood on the bottom of the dish is then quickly frozen by setting the dish in n sh:illo~ pan containing nlcohol and pieces of solid carbon dioxide. At least an hour bcforc the IAcetiing, a desiccator of the IIcmpel improved type with an internal diameter of about 15 cm. is packed in an ice-salt mixture in n rcct:mgulnr metal p:In. In the ~roo\~c in the upper pnrt of the tlcsiccntor is about 130 cc. of fresh conccntrate(l sulphliric acicl to absorb the moisture given off by the l~lood. In the lx~ttoni is cnougll gl>~ccrinc to cover a porcclnin platform ;uxl make a goocl contact \Vitli tlic c\.aporating clish. When the I~lootl in t11c v\-:1por:\t inK tiish is thoroughly frozen, the dish is IiftctI from the nlcol~ol, t.lic liquitl is c]uicl;ly \\.ipcd from the bottom, nntl the dish is plncetl in the tlcsiccn tar. `I'hc co\`cr of the desiccator is then given nt least one coniyjlctc turn to mal~c sure that n gootl contnct is made. hi estrcmcly small quantit\- of ;I suit:il~lc thick lril,ric.;lllt II:15 been applied to the contact surfaces in X~\TlIlCC?. `I`he clesicc:ltor is tlicn c\.;icu;Ite(l \vith 311 electric air pump until the prcssIlrc ll;ts f;lllcn to 1 or 2 mm. 0i rnCrc~Ilr?-. `I'lic clcsiccntor in its pan of ice nntl salt is Ijut into 3 rcfrig:cr:Lt~~r to J>rct\.cbn1 raI)itl melting of the ice. I'rom 16 to 20 hours later the \`:lcullm is rc~lcxcvl ant1 tlic clcsiccator is opcnccl. The evaporating dish is t:iI;cn oul, \vipctl free of gl!.ccr-inc ( anti pIaced on ;L sheet of paper under a sm:\ll glass-tol)pctl fr:lrncl. `I'hc oljject of the frame is to prcvcnt the distribution, 1)~. :lir currents, of infcitious (lust \vhich mi~:ht 1~ inl~dd hy the operator. Rul)lw gl0vc'S arc \Yorn \\.llilc the clriccl mntcri:d is being tubed. :\hut ten 1~111g:ficcl sterile test tubes mc:lsuring 12 1)~ 200 mm. arc numbercrl nnci \\~eiglied in advance. `I'hc~- nrc taken one 1,~ one, warmed slightly in the flnmc to prwcnt. conticnsntion of moisture, partially filled with the fl&es of dric(l l~loo~l. fl:lmctl nt the mouth, rcpluggecl, and ngxin \r~cigh&l. The gain in IYciglit in grams rcpresen,ts tlic \\cight of the dried blood, Lvhicli is approximately one-fifth of the u,ciglit of the fresh blood or of its volume in cubic centimeters. In the cvnporating dish the dried blood is in the form of n thin, red, brittle wafer, with fissurcs running through it. It lies 100s~ from the glass and can easily bc brc)ken into bits and transferrctl to the test tulx; \vith forceps and narrow sp;ltul;L. It is light 2nd porous 2nd can be rcxlily crushed in n mortar to n smooth powder before it is tubed, but this prohbl!~ increases the risk of nccidentnl infection nntl otxlinarily is of no ndvnntngc. ;\ftcr the weighing, the tops of the plugs are cut off \vith scissors nntl the rcni:iindcr pushcd do\\n to the surfncc of the dried blood with a rod. To m:lkc certain of complete dryness, granular calcium chloride is poured in above the plugs lwfore the tulxs nre sdeti in the ffnme of n Mast lamp. The tubes arc stored in the rcfrigcrntor. The LISC of calcium chloride, although W. A. SAWYER, IV. D. M. LLOYD,, AND s. F. KITcREN 9 perhaps advisable, is not necessary, since the oldest specimens tested were put up without it. The method of putting in the calcium chloride has been described by J, H. Brown (7). When the dried blood is required for inoculation, the tube is opened by scrntch- ins with a file at the middle of the cotton plug nnd touching with a hot piece of glass or metal. The plug is cnrcfully rcmovcd and n smnll amount of 2 per cent sodium citrate in physiological salt solution is put in to prevent scattering of dust. The contents nrc then cmpticd into n mortar, the adhcrcnt particles are scraped out of the tub; with n platinum wire, and the material is ground up with addi- tional fluid. Without the sodium citrntc in the physiologica salt solution the dissolved blood will clot in the mortar and make injection diflicult. Small quantities of blood,, 1.0 cc. or 0.5 cc., are easily dried in n. test tube or ampoule. The dry blood forms n smnll htton which lies loose in the bottom of the container. `l`hc blood nl:L~r I)c put up itI nmpoules \vith snfcty ant1 with great simplicity of tcchniqnc. 1,111 this nd~~xrit;~g:e 01w the use of the cvnpornting dish nn(l test tul)c is more th:ln on'sct 1)). the grcxtcr (lifficulty of getting the material out. \Vherc solitl carbon tIiosi(tc ("(lq. iw") is not avnilnt~le the blood may bc frozen in small amounts (1 .O or 0.5 cc.) in ;L numlxr ol tat tubes in an ice-salt misturc before desiccntion. To prescrvc the virus in liver tissue, from 10 to 20 gm. of the fresh tissue is ground \\ith sterile s;intl in :l ni~~rt:lr \~ithout nclclcti lluid. `I'hc resulting past42 is scraped in to tllc% glass cl-npor:l tin :: tlish :lnti tlic stcrilc gnuzc cover is rcplnccd nnd tied with string. Gcntlc tapping: of the tiish on n table top will complctc the even distribution of the mxtcrial over the bottom of the dish. The material is then frozen, tlriccl, and sc:~lctl in tr~l~cs l,>, the s;tmc technique as the one used for blood. f\S tllC tOt31 \vci,qht Of the original matcri:ll is known, the amount of liver tissue in each tulK, in terms of fwsh liver, is (leterrnined by divitling the original \YCig:ht of the liver in proportion to the finnl \ycights of liver and sand in the severnl tubes. Inocul:~tions arc mnclc ri,itfi the material 3s in the cnse of blood, but physiological sxl t solution is usccl instead of citrate solution. The dried liver is soft, porous, nnd friable, and has 3 light greyish yellow color due to the presence of much fat. " Freezing `is not nlxolutcl~ ncccssq for the successful preservation of blood. Specimen 157 r\ (`l'ablc III) wx preparcti in T,agos 1)~. the IVest African \`cllow Fcvcr Commission of the Rockefcllcr l~ountl:~tion anal I\`;IS brought to us through thC courtesy Of l)r. Oskar l.s (S). In preparing Specimen 1.57 :I the mixed blood of two monkeys \vns dried in a ~wxuni o\`cr calcium chloride nn(I se&d 10 PRESERVATION OF YELL& FEVER VIRUS TABLE III Pveservalion of Yellow Fcvcr k'irus irt Dried Blood ZZZ - f Age 0 slwci- me 11 Interval in days 10 hIethot1 of prcser vation IX= - ZZZ ., 7 hlonkr) inocu- lllld = , I- Death or recover OhSCK vatior perioc - - Drying -__ cc. 2.5 0.5 0.1 1.0 1.0 0.1 0. 1 6+ 0. 3 0. 3 0.3 0.5 0. 5 0. 1 0.1 1 0 1.0 0.1 0.1 0.1 0.01 0.1 1.0 1.0 1.0 1 0 - lid JfS ivn. 12 202 1s 205 13 95 2s 2s 28 20 30 11X 3 1 1oi 32 103 42 144 u 42 143 U 42 1.15 u 60 14i u GO 157 U 61 150 7s 152 X6 209 91 12% u 100 1SO 100 15% 101 176 101 177 102 173 109 151 U 127 210 151 235 155 231 - , Fcvcl - - 3.5 7.0 - 4.0 6.0 3.5 -t 2.5 4.0 3.5 1.5 3.0 3.0 2.0 1.5 3.5 - 9.0 - 3.5 3.0 - 4.0 16.5 - dtJYS 32 32 - - - - - 7.5 9.5 - 6.0 10.0 7.5 4.0 4.5 7.5 7.5 3.0 4.5 5.0 2.011 3. 5 10.0 - 11.0 - 5.0 5.5 - 6.5 19.0 = Not tested Not tested 30 Not tested 30 37 30 Succumbed GIccumbed vat tested Result of test inoculation + This specimen, v~hcn G days 01~1, producccl fever in a monkey in 2 dnys and death from yellow fever in 3.5 days. t No fever olxxrved. $ Tubed with calcium chloride. The other tubes wcrc without it. $ Rc-tulxtf \rith cnlcium chloricle Jvhen 4.1 clays 01~1. jj l~illcd to oi)t:lin mxtcri:11 for stud\, of early lesions. 11 Dried \vithout freezing. W. A, SAWYER, VW'. D. M. LLOYD, AND S. F. KITCHEN 11 TABLE IV Groups of Specimem of the Saw Material Preserved in DiJerent Ways and for Varyiq Lengths of Time Source, monkey and strain 32 I: 55 A 65 A 66 1; 117 I? Method of preservation Drying (;lyccrinc Drying Citration Clotting Glycerine Drying Glycerine Drying 4mount, fresh cc. 0. 1 0.1 0.1 0.01 1.0 1.0 1 0 0.1 0.1 0.1 1.0 2.0 30 2.0 30 2.0 SO 0.1 30 0.5 60 0.1 100 1.0 0. 3 0.5 1.0 1 0 GO `12 60 S6 O-1 days I - - - p;;;;t ICVM? + + -I- - + + - + + + - - - + + + - + -I- + + + fifethod of Amount Age of Ixcscwation fresh ipecimer Freezing Drying m. days 3f 20 1.0 9 0.1 9 0.01 9 0.1 60 1.0 100 1.0 150 Drying Glycerine I )rying Liver Result. Yellow fever? + + + - + + 1.0 30 + 1.0 100 - 1.0 30 + 1.0 30 +R 1+ 100 3-R 0.1 30 + 1.0 100 + 1.0 1.0 60 -63 +R 12 PRESERVATION OF YELLOW FEVER VIRUS in a glass tube. It was transported from Africa to America at-room temperature. A part was used in a successful inoculation when the material was 32 days old. The remainder was retubed with calcium chloride and was found to be highly virulent when 102 days old (Table III). It infected also after 155 days, but the incubation period was prolonged. The dried blood of this specimen consisted of small black glassy scales in striking contrast to the soft porous red pieces of the blood which was frozen before being dried. Specimens 86 and 87, from Lagos, were also dried without prcliminnry freezing. One of the two was found to be infectious when 28 days old (Table III), but the other was inert. The specimens of blood and liver which had been frozen and dried seldom failed to produce infection. Two of the three specimens of blood which were tested when 100 days old were found to be virulent when injected in amounts as small as the equivalent of 0.1 cc. of fresh bloocl. One of these two failed to infect at 127 days. l'hc other was tested again when 153 days old, in a dosage equivalent to 1 .O cc. ~(1 produced ~cllow fever after an incubation period of only four days. Tests after longer pcriocls hxve not yet been made. The dried liver gave similar results, but the 4xcrvntions were fewer (Table II). Our oldest spccimcn (from 3Ionkq~ 32) failed to infect when 60 days old when a dose equivalent to 0.1 gm. of fresh liver was injected, but infected after 100 days and after 150 dn:;s when the equivalent of 1.0 gm. was injected. Another speci- men 100 days old was tested and found to be infectious. The process of freezing and drying \vill preserve hnctcrin 8s well as the virus of yellow fever, :uicl one \~oultl cspcct to get bnctcrinl infections occasionally from the use of dried liver, since pr~thogcnic orgnnisms frequently invade the liver lnte in the tliscasc. Such infections would occur very seldom when dried blood is used. As the result of inoculation xrith a specimen of fresh yellow fever liver we lost two monkeys from n streptococcus peritonitis, and iqoculntion with another specimen, which hnd been sent us under refrigeration, caused three animals to develop tubcrculous peritonitis. A close comparison of the vnluc of the methods can only be rnxde by submitting the snme original m,ztcri,zl to the several processes. One specimen of fresh blood or liver m,zy contain many times as much virus ns another. We have, therefore, brought together in groups in Table IV the specimens derived from each of sevel;al'lots of blood or liver. Ench of thcsc specimens ,zppe,zrs nlso in one of the previous tables, whcrc r~tlditional dntn :lt-c given. CONCLUSIONS 1. The virus of yellow fever mazy be preserved for ,zt Ic,zst 1.54 days in the blood or liver tissue of infected monkeys if the material is dried W. A. SAWYER, W. D. M. LLOYD, AND S. F. KITCHEN 13 in a vacuum while in the frozen state and kept in the refrigator in seaIed glass containers. A gradual diminution of virulence is notice- able in the oIder specimens. 2. If infectious blood is dried in a vacuum at rcxxn temperature, instead of in the frozen state, and is stored in sealed containers in the refrigerator, the virus may survive as long as 155 days. 3. The virus may bc preserved for at least 30 days in liver kept continuously frozen. 4. Storage of blood or liver in 50 per cent glycerine in the refrigerator will usually keep the virus alive for 60 cla;~s and may do so for 100 days, but with the injection of the older material there is a marked tendency toward lcnglhcning of the incubation period and increase in the num- bcr of rccovcrics. 5. Uellow fever virus in citrntcci or clotted bIood, when kept in the refrigerator, c-lies out rnpidly. 6. In our espcricncc the most satisfactory method of prcscrving strains of yellow fever virus in the laboratory consists of freezing and drying blood taken from a rnonk~~- OJI the lirst day of an attack of esperimcntnl ycllo\v fC\rcr ;111(1 storin, (7 the dry material in scnlcd glass - tubes in a colt1 l)lace. 1. Stokes, A., Tklucr, J. TI., xn(l TIudson, N. I'., T3sperimentnl trnnsmission of yellow fever to lnhnrntor~~ animals , :1 ??I. J. l'rop. illed., 1928, 8, 103. 2. Sellards, .A. I\`., nncl IIindle, E., The prcscr;-ation of yellow fever virus, III&. ,lfcrf. J., 10'S, 1, 713. 3. hInthis, C., Scllarcls, A. \\`., and Laigret, J., Sensil~ilit~ du M~ICUCIU ~I~CSIIS au virus dc In Gvre jaune. Cotrcpf. rc)rtl. shrccs :lcad. Sci., 1928, 186, 604. 4. Hnrris, D. L., nnd Shxkell, I,. F., `I'he effect of vacuum desiccation upon the virus of rabies with remarks upon a new method, J. .~l?rz. Pub. IIcalfh i4ssrz., 1911, 1, 52. 5. Rous, I'., A w-coma of the fowl transmissible by an agent separable from the tumor cells, 1. E.rl,. McYI., 1911, 13, 397 (footnote, page 399). 6. Swift, Homer F., T'reservntion of stock cultures of bxterin by freezing and drying, J. Erp. Med., 192 1, 33, 69. 7. Brown, J. II., Vacuum tulxs for the storage and shipment of bacteria, Scicru, 1926, 64, 429. 8. Hudson, N. 1'. , and Klotz, O., The preservation of the yell&~ fever virus; an unpublished report to the Intern;ttionnl IIenlth Division, I~OCkcfellCr Foundation, Nox~embcr, 1928.