VACCINATION AGAINST YELLOW FEVER WITH IMMUNE SERUM AND VIRUS FIXED FOR MICE BY W. A. SAWYER, M.D., S. I?. KITCHEN, M.D., Ah'D WRAY LLOYD, M.D. (P'ront tlzc Yellow Fcvcr Lnborafovy of ~Jzc Inkrnczliounl IleaItJc Divisioll, Rockefelh Fouirdnlio~~, Ncu York> (Received for publication, &larch 22, 1932) The method here presented for vnccinntion against yellow fever was devised primarily to interrupt the long series of accidental infections of persons making Iaborntory investigations. In the 4; years which ha.ve clapscd since rlrcsrts m01~1~cj~ first came into use as cspcrimcntal animals in yellow fcvcr stuclics, Ihcrc h;l\,o l~cc~n rqx~rtod 32 infections with yellow fever in IaLor;LtOrics clcvotc(l to roswrch in this discasc, xcording to II)crry ant1 Kitchm (1 j, ant1 live scientists have lost their lives. Prclimt7rory Ex~t~ri7ur~ri.s :;liiJl O/h-r I'rllow I;`cwr Vnccincs The results of cspcrimcnts in this l:~l)or:~tor! anal the receipt of unfavorable reports from clsc\Vhc~rc ha\.c mxtlc us un\villing to recommend the immunization of persons against )~llo~ fc\ver \yith \xccincs prqxlrcd by the chemical treatment of the virus-containing tissues of infcctccl nnonke~x. \`accines of this general type have bcm devised by Ilindle (3). :\r:~gio (.;), I'ct tit and Stcfanopoulo (4)) and Monteiro (5). .Accorcling to Chag:ls (61. irregular rcsulls follo\ved the vaccination of 25,000 persons b!r the method of .Ar:lg,`co during the rcccnt cpidomic in Rio de Janeiro, and failures were also encountcrccl in the csi)crimcntaI vaccinnti6Ii `of monkeys. Okell (7) found that a vaccine of the sfme general tJ.pe lost its immunizing po\ver rapidly on storxgc, 2nd this would crl~lain the fnilures to immunize persons far from the ~)lace \rherc the \Ylccine ~~1s prepnrccl. Davis (S) mndc an experimental chloro- form-treated tissue vaccine \vhich WCS found, on laboratory test, to be sometimes infective and sometimes 1Tithout immunizing effect. Burke and Davis (9) recorded the case of a person in IIrazil \sho contracted a fatal infection with ycllo~ fever 5 months after \~nccin:ition 1\4tli rnonkcj~ tissues treated \Yith formnlin and ~~henol. 111 hlarch, 103 1, :1 commission of the French r\cadcni>~ of ASledicine (10) :lnnounced its conclusiou that 110 mcthocl of v:~cciri:ition against j~llorv fever had been sufiicicutly studietl for use in practice. Aflcuzrcztir~lr 0j lFi~,~~.s ilr ,Il0~k(,y .Tc,).llllz.---In this Inbor:~tory vaccination esperi- ments \Yith attenuated virus \vcre carried out in 1929 and 1030 with the cokbora- 945 946 YELLOW FEVER VACCINATION tion of N. &MI Hudson. The virus in highly infectious monkey serum was pro- gressively attenuated by exposure (a) to a temperature of 37OC., (b) to the effects of added formaldehyde (0.05, 0.1, or 0.2 per cent) or tricresol (0.3 per cent) at room temperature, or (c) to the effects of 0.3 per cent tricresol at a temperature of 37oC. or O.S"C. Portions of the serum under treatment were withdrawn at intervals for testing. Some of these portions were injected at once subcutaneously into mon- keys, Others were dried in the frozen state in test-tubes or ampoules by methods dcscribcd in a previous publication (1 l), and stored for a week or more before injec- tion into monkeys. Later the monkeys were given test inoculations of virulent yellow fever virus of the Asibi strain. The periods of exposure of the vaccine to the attenuating agents varied in the different experiments, but commonly ranged from an hour or less to several days. As a rule the vaccines exposed for the short- est periods to the attenuating agents produced y~.)Iow fever in the monkeys inocu- latcd and the specimens long csposcd f:tilcd to imr!\~lnizc. Between these extremes, there wcrc founcl in sonic wscs csposures \vhich :rttcnuntccl the virus to such 311 cxtcnl th;tt the wccirl;ltc(l ;lnim;~ls sho\vctl no itsvcr but :quirctl an active im- munily. This zone in the scrics of csposurc tinn(aL` \vas inconstant nnd narrow. The irrcgul:~r rcsul t s con\~inccd us that mucll curtlicr rcsrnrch would be wct`s- snry il a safe and tIepc~~dnl~lc vaccine for human Iwings ~ws to be prepared by the chemical trentnicnt of virulent strains of ~.cllow fc:\w virus. Sir~rult~~~~c02~s Irlvjcifioll of T'irl(s urrd ImmIlil( Sc~zlw.-The practice of injecting virus and immune serum at the s:lmc time but in (IitTcrcnt plnces has been followed in \~accinating s\vinc ;Ig;linst Ilog cliolcr;L ant1 c-;lttlc agninst rindcrpest, and in immunizing :Ig:ninst other tIiswsc5 of :~llinl;LIs. 1 t is possible to immunize against at Icast one ol tllcsc cliscasc~, Ilog cholera, \vIlcn the irirus and serum have been misccl l)cforc injection, for !)u\ral (13) ll:ls sI~o\~n 1 hnt the dried virus and the dried immune sc'r11m nta~~ l)c misctl, s;torc(l, :Lnci usccl t:l,gctller \vitIlout the loss of anti- gcnic I)owtbr. h~l:~Il!~ \Yho, like oursclvcs, have h:\(I occnsion t 0 rn:\l;e protection tests of sc~ `91, using moul;c!5, h:l\ye obscrvcd the prcscncc of ;t ' : ;~vc immunitJr in those nnir,,.ds \\,liicIi have ljccn Ijrotcctcd ag:linst ftl injection of ~~ellow fever virus by a s; ill- tnncous or prc\.ious injection of immune serum. This observation (Theiler I ;Id Sell;1 rtls (1 .q , I< `incil;L!~ :lnd IIinclIc (l-1), .Arq$o (15)) l ifs suggested the possill;' :t.y of using the snmc general method in \~nccinating man. In one of our carI!. csperiments :I solid active immunity was produced by mix- ing ~~cllow felrcr virus of the potent Asibi strain in monkey serum, dried under vacuum in the frozen state and redissolived, with sirniIarl>F dried immune monkey scrlini, :Intl injecting the mixture sul~cutaneously into monkeys. The immunity Jv;ls ;Ic(luirctI in tlic ;~l~t;cricc of fc\.cr. z\lthougli the scruni and virus were in con- tact for 30 to 40 minutes at room temperature, thorough immunization was nccom~~Iishcd \Ylicn the amount of irttnlurle serum in the mixture wns 5 or 10 I ~IWY the minimum ncccssflr>r to protect against the virus. It seemed to us that it would be a ~~alunble safeguard if the living yellow fever virus sclcctcd for USC lvith immune serum in vaccinnting human beings were of W. A. SAWYER, S. I?. KITCHEN, AND W. LLOYD 947 lower viruIence for man than the strains now being carried in monkeys. In vac- cination against hog choIera, according to information given us by Dr. R. E. Shope, on rare occasions there occur unexplained failures of the immune serum to protect against the virus and as a result the vaccinated animals die ol the disease. A ttenuation of Virus in Mouse-Brain Tisszr, . 0 -When The&r (16) showed that yellow fever virus which had been adapted to mice had lost much of its virulence for monkeys, the hope was raised that the untreated virus in mouse-brain tissue, after enough passages in mice, would lose its virulence for man and would be safe to use as a vaccine. At the time of our experiments the French strain of yellow fever virus had been through over 100 successive passages in mice and was never- theless still able to produce fever in monkeys, although it had apparently long lost its power to kill them. Among 9 rJ~cszu monkrys which were inoculated with this strain, after 102 to 120 passages through mice, and were afterward given test inocu- Iations with virus from monkey source, 8 dev4oped fever and all 0 were subse- quently found to be immune. Attempts were made to transfer the infection from 4 of thcsc animals to normal monkeys by inocul:ltion with blood taken at the bcgin- ning of fever, but none of the animals rccciving the blood developed fever; half of them wcrc afterward shown to be susceptible, and the others immune, to yellow fever virus from monkey source (;\sibi strain). Two ctintro1 monkeys inoculated with normal mouse-brain tissue rernaincd free from fe\.er. The gradual diminution of virulence for monkq~ as yellow fever virus is passed through mice is illustrated I,>, our espcricncc with the Asibi strain. A monkey iuoculatcd \vith virus of the Sth passage in mice died of callow fever. Of 2 mon- kq,s inoculated \frith \.irus of the 10th IUSSI~`, 1 died of callow fever and the other had feircr followccl by* immunit~~. One \yhich received the 11 th-passage virus clicd of ~~110~ fe\~r. hlankq~s inoculntetl l\.ith \Grus of the lSth, 20th, and 25th passages dcvclopctl fc\.cr ant1 \Ycre immunized. The French strain in the 176th passage in mice still produced delinitc attacks of fever in monkeys. ?`herc: is other ciridcncc of the persistcncc of :I dcgrcc of virulence for monkeys and man in ~.ello\v fever \Grus after man!. passages through mice. Sellards (17) has shown that such \,irus \vill cause cnccphalitis in monkeys if injected intra- cercbrnll>~. hlorcover, there are on record 3 mild cases of yellow fever contracted b\r persons through contact lrith espcrimentall~~ infected mice or their tissues (1). Yellow fever \G-us after 100 or more passngcs in mice is probably of low virulence for man, but still capable of causing mild attacks of callow fever. If used in vaccinating persons, it seems advisable, therefore, that the virus be attenuated or given with a potent immune serum. Two methods of attenuating the i7iru.s in mouse brain were tried, those of Semple (18) and r1livisntos (19) for preparing nntirabic vaccine. Modifications were necessary, and in some instances the vaccine was dried in the frozen state and stored before use. The vnccine of the Scmplc tJ,pc, as prepared and adminis- tered to monkq5, was apparcntl>~ devoid of immunizing power. That of the t\li\risatos tJ.pe produced fc\.cr and immunit>~ in monkeys l~hen the brain tissue had been esposcd to ether at 0.5"C. for 6 or 24 hours, but was inert when the exposure had been for 48 or 72 hours. 948 YELLOW FEVER VACCINATION Vaccinatiosz with Living Yellow Fever Virm Fixed for Mice and .Tmmwze Human Serum The preIiminary experiments with various ways of vaccinating against yellow fever seemed to show that the most dependable and cffectivc of the methods tried was the injection of living yellow fever virus with a simultaneous or preceding injection of potent immune serum. Moreover, it appeared that yellow fever virus, after many passages through mice, had lost most, but not all, of its virulence for monkeys, and probably for man, although retaining its power to immunize. It was proposed, thcreforc, to test in monkeys the safety and the immunizing power of a vaccine composed of living virus fixed for mice and human immune serum, anti, if thcsc tests gave satisfactory rcsuIts, to commence immunizing persons exposed to yellow fcvcr. The vaccine used throughout the csperimcnts nnd in vaccinating people con- sisted of two parts, (a) the virus in mouse-brnin tissue suspended in n portion of the inlmune serum or in normal serum, and (21) immune human serum for separate injection. In manor csperimcnts, hon*e\~er, both parts were combined. The description which follo\vs gives the technic used in making most of the lots of the vaccine en~plo~wl in immunizing persons, and at the cntl there is a statement of ways in which the I>reparntion of the first vaccines, i'nccines A, l3, 1, and 2, and the last, \`accine 6, differed from the description. The essentials of the method of prcpnr:ltion of cach \.accinc used ;\re given in Tnble I. 7'11~ Co7~/wrrclll Calrluilriug TVirzrs.-The \rirus used in the preparation of the vaccines \vns of the French strain, 105th to 176th passages in mice. Only persons immunized against ~dlo~~ fcvcr by illness or vaccination wx allowed to work Jvith the virus. ;1 10 per cent suspension of the infectious mouse-brain tissue in im- munc serum was centrifuged for i hour at about 3,000 1C.P.N. The supernatant fluid ~3s then passed through n Scitz filter. 11s tests of sterility sets of aerobic and anaerobic cultures in slightly nlknline meat infusion peptow`. broth were made beforc and after filtration. The filtrate was plxed in amounts of 1 cc, in sterile, plugged test-tubes 2nd uxs dried in the frozen state binder vacuum. The tubes were then senled in the blast lamp ant1 stored in the refrigerator. Some of the \r;Iccinc, r~~lissolvccl in clistillcd water, was injcctctl intrncercbrnlly inlo 0 susceptible ivhitc mice in :lmounts of 0.03 cc. and intrnperitonenlly into 4 mice in amounts of 0.2 cc. In alI cases some' of the mice given intracercbral injec- tions died of J-ellow fever encephnlitis, showing that there was living virus in the prepnmtion. `The mice rccci\dng intrapcritonenl injections remnined IYell. At least 1 monkey was vaccinated with each lot of the material together with the required amount of immune serum and was tested later for immunity. W. A. SAWYER, S. F. KITCHEN, AND W. LLOYD 949 The Immune Serzcnt.-The immune sera used for the sep`&Zte injections at the time of vaccination, and also in suspending the virus in the preparation of the vaccine, were obtained from 6 members of the laboratory staff who had had yellow fever as the result of accidental laboratory infections (l), and from 3 other members who had been immunized by vaccination during these studies. Blood was with- drawn, defibrinated, and cIeared of cells by centrifuging. Aerobic and anaerobic Mclhod of Prcparatioir oj /he Vucciircs Used Designa. tion of vaccine A 11 1 2 3 4 5 6 Strain of virus used and No. of passages in mice French-105 Prcnch---106 French-107, 116 Frcncll-117 Frcncb-146 French-1 68 ITrcncli--1 i-t FrcnclI- 176 - Human serum in which virus was suspended No. 3, immune 10 No. 3, immune 10 No. 1, immune 1, 2.5, 5 No. 2, immune 10 I'001 ~1, immune 10 No. 4. immune 10 1x0. -5, inlmllJle 10 No. 6, norrnnl 10 -_~ Percentage of virus* in suspension - - I _ - Filter used Not fil- terccl Not fil- tered Not fil- tcrcd Ikrkcfcld N Seitzjl Seitz Seitz Sci tz . * Fresh nlousc-brain tissue containing \*cllow fever virus. R&Sllcpr Fresh$ Dried Dried Dried Dried D ricd Dried Dried Dried - - 1 - - \iortality ratio in micet o/12 6111 7/18_ A,/12 n./11q :: /`l: I"/12 St; !,' :(9 _.,. .- t hIice \j*ere inocul:ttctl intr:~ccrebralIy. The ratio sl~ows the proportion of mice d>ving from 4 to 10 ti:i~5 xftcr inoculation. 1 \Vhcn 3 weekl\~ injections of this fluid iwzcine were giLTen, the 2nd anri 3rd doses were freshly prepared by the same method but with new lots of virus. This vaccine ~~1s fresh in Esperimcnt I and dried in Experiment IV (Table II). $ hIortnlity ratio for 1 per cent virus, 2/6; for 2.5 per cent, l/6; for 5 per cent, 4/G. 11 Part through 13erkcfeId S, part through Seitz. li The brain of 1 ~OLISC was examined microscopically, and lesions of yellow fever encephalitis were found. - cultures in broth ivere made as tests of sterility. To the serum for USC in the sep- rate injections was added 0.2 per cent of tricresol in ether, by the method of Krum~~ictlc and Rnnzhnf (20). Each strum was tested for potency by injecting it s~~lx~~tnneously into ~I~CSMS monke~~s in nmounts of 0.3 cc. or more per kilo of body weight and thereafter injecting 0.4 cc. of monkq~ blood containing yellow fever virus of the Asibi strain! virulent for monkeys, under the skin in another place. i\mounts of serum ~vhich protected the nnirnals against J-cllow fever wercconsidercd suit:&lc for USC in the s:lmc rfos:~ge 11cp kilo in vnccin%ting m8n, 950 YELLOW FEVER VACCINATION Serum 1 consisted of several lots from W. A. S., who had been infected from monkey source. This serum had a titre of l/128 when tested in mice by the intra- peritoneal protection test (21). Serum 2 (titre, l/256) from M. T. and Serum 3 (titre, l/128) from S. I?. K. came from persons who had been infected with virus from mice. The donor (K. S.) of Serum 4 (titre, l/256) had been infected with virus from monkeys. L. M. M., the source of Serum 5 (titre, l/128), had been immunized by vaccination. Serum 6 was normal serum without protective power. The blood of cnch donor had been found to give a negative Wassermann reaction. Serum Pool A (from W. L., K. S., and `I'. N.), Pools 1 and 2 (W. A. S., S. I?. K., M. T., T. N., and K. S.), and Pool 3 (W. A. S., S. I?. K., M. T., and T. N.) were made up of sera from persons who hnd experienced attacks of yellow fever. Pool 2, in the amount of 0.3 cc. per kilo of body weight, protected a monkey against the A'sibi strain. Pool 3 did not protect either of 2 monkeys when injected in the amount of 0.4 cc. per kilo of body rycight. It protected when used in the amount of 0.5 cc. Pool 4 (titre, l/2.50) JY:IS :1 mixture of the sera of 2 vaccinated persons, M. Il. and I,. 13. `1.`. 1Vhcn injcctcd 0 hours before the virus, this serum protected onl-y 1 of 2 monkqrs in a dosage of 0.3 or 0.4 cc. per kilo, but it gave protection in the dosage of 0.6 or 0.S cc. per kilo. Dcviafions from /AC Tcchic I)c,sc-rib&-In preparing l'accines A, 13, and 1, the suspension of brain substance in immune serum ivas neither centrifuged nor filtered. Vaccine 1 was prcpnrcd in 3 strengths containing 1, 2.5, and 5 per cent of brain tissue, respectively. \:nccine 2 1~s filtcrccl successively through Berkefeld cylinders V and N, instead of through the Seitz filter. In preparing Vaccine 6, the ~~irus-containing mouse-brain tissue 1~2s suspenticd in normal human serum instead of in immune serum, and the \~:lccine \vns dried in ampoules in 0.5 cc. amounts instead of in test-tubes in nniounts of 1.0 cc. A tluli~lislrt~fin~~ of tJw l~vt2i-tei)it*.p-m In the vaccinxtion experiments in which mon- ke1.s \vcre USC~, all the serum zlntf the \virus \vcrc given as a rnixtllre except in ;L few instances, in I. unusual sensitivity to mouse-l)rnin tissue. 1 per cc11 t suspcIisions, csIxxinl1~~ if lilterecl, produced only slight transient reactions. 111 the light of this experience and the observation that severe immediate reactions are seldom if ever encountered after t`he'first injections W. A. SAWYER, S. F. KITCHEN, AND W. LLOYD 951 of the spinal cord or brain tissue of rabbits in antirabic treatment it was decided that the preliminary intradermal test was unnecessary. Vaccination of Monkeys Nine experiments in the vaccination of ~JZCSZLS monkeys with mixtures of virus fixed for mice ,znd immune human serum were performed with the vaccine prep;lr~tions nnd scra nlrcady described. The results ,zre brought together in Tables II and III. . Most of the animals showed no reaction to the vaccinati0n during an observa- tion period of at least 30 dnys. Some had rises of temperature to 40oC. or above, which are recorded in Tables II and III as fever. The temperatures of the mon- keys were taken twice each clay, and the tissues of animals that died were examined histologically. Ijlood tlrnn9 long enough after vaccination to rule out passive immunity ivas tcstctl in monkcJ,s or mice, or in both, for its protective power against yellow fever virus. I~ \vi thin 11 d:l~x. l':issi\re immunit!~ could not have been present even in the first S da~rs of Ir in the amount of 1.5 cc. per kilo in Experiments I and II, and .3 cc. per kilo in I5 in I':sperimcnt I', 7,; dnys in Ksperimcnt \,`I, 47 days (Monke~~s 29 :tnti 31) :ttld 92 dryers (T\lonkc!x 30 and 32) in I 14 7 C-1 14 12 (k) 21 6 `l&J 9 -3 C-1 7 7 (*> 10 0 c-1 7 7 l-1 11 7 (*) 11 10 t-1 14 17 (+I 21 - : I 1st -aosi ti vc Last legativ result1 days 0 0 7 0 14 -3 -6 -45 05 st pasi- tivet days 19 23 23 15 23 14 14 80 X0 84 34 36 * IVhen inconclusive results are shown, there was an earlier negative result. t Two monkc~~ were used in Case 5, in other cases only 1. Numbers preceded b~v minus signs show the number of days before vaccination. Jo Two test monkeys were used in each case. Control monkeys receiving virus nnd normnl serum died of yellow fever. _ In :ln ndtfitional test 21 d3ys after vaccinntion 1 monkey survived and 1 died of yellow fever. /J Serum given G hours before the virus component. power. This evidence 3ntI the observation (`I`sble IlI) that even one one-hun- dredth of the usunl amount of virus would produce strong immunity in monkeys, led us to fix the quantity of virus to be used in wccinating human beings, irrespec- 956 YELLOW FEVER VACCINATION tive of weight, beginning with Case 16, as that which is present in 0.5 cc. of a 10 per cent suspension of infective mouse brain in normal human serum after filtration through a Seitz disc and storage in the dried state in the refrigerator for not more than 6 months. A man weighing 70 kilos, ior example, would now receive only 0.0007 gm. of mouse brain per kilo, minus the loss by filtration. While body weight would no longer determine the amollnt of virus, it would still decide the amount of immune yerum. The first pcrsp" to be vaccinated was admitted to the Hospital of The Rockefeller -f rlstitute for Medical Research through the courtesy of Dr. Rufus Cal:- and Dr. T. ht. Rivers, and was closely observed for us by Dr. G. P. IIcl-ry of the hospital stnff. Six hours after vaccination there was deciclctl : l:ndcrness in the nbd!~minnl wall where the unfiltered virus component of the vaccine had been injected, but none where serum alone hat1 i)cen given. The tcnclcrness disappeared in 2 days. There wcrc no sllj,jcctivc symptoms R~I!I no ahnormalities of tempera- ture, pulse, blood pressure, heart action zs shown by eIectrocnrdiogram, or urine, during 10 tf:t~~ in the hospitd, 2 before- ,vaccinntion and S after. The other persons vnccinntccl rcrnninccl on duty in the I&oratory. Outside the hospital the vnccinntcd pcx:-sons took their own tempera- tures at intcrv& of 4 hours during the (!:~y for LZ I>criod of 2 weeks and recorded nrly subjective syn~pton~s. `I`he symptoms following vac- cination were few and may bc bricflj, summarized in 4 groups as Collows : I. Ikrrly Rc~zctious to tlzc I)ljcCttd ll10zlsc-13rl.i )z Tiswc.---These symptoms were most in evidence during the evening following \i;\ccination, and they disappeared in from 1 to 3 days. They were (a) tenderness :{t the site of injection of the virus suspension, sometimes ~3th slight redness and s:*:clling, and (2)) rises of tempernture, with hcndnchc or dull pin in the b;lcli or legs it, 3 instances. Tenderness was pre- sent in all cases except Nos. 8, 12, and 16, nnc'! was severe in Cases 4 and 6. In 0 of the 16 cases the temperature rose above nl):.rnal, but in only 1 did it go above 37.4"C. Except in Case 6, these cnrly sympt( ~!lls xxrc slight or absent when the injected virus suspension hncl lxcn liltcrcrl. I.~ucotr~~t osis wx commonly present. 7 S~llzf~f0l~I.s Su:;,qc'.sf ix (Jj Ltrfc Rr,clc-/ioll.s !iJ !;Ic I!!-;(,: 1. ful ilfousc-IJrniu TI'ssu~.-- (a) 7 d:l~5 after vnccin:Ltion, in Case 2, n I~~nlI)li nocle near one of the sites of injec- tion of unfiltcrcci virus suspension bccamc slig:iltIy enlnrged and tender. On the W. A, SAWYER, S. F. KITCHEN, AND W. LLOYD 957 following day the r;l;in at the site of the intradermal injection of the same suspen- sion became reddened and elevated in an area 1.5 by 2 cm. and was surrounded at a distance of a fetv rentimeters by 6 small papules (2 by 3 cm.) topped by vesicles 1 mm. in diameter. 1 day later similar papules appeared here and there on the arms and legs. `I`::,\ temperature was 37.2OC. on the 8th and 9th days. (b) 7 days after \.nccination, in Case 12, one of the sites of injection of lT!lfiltered virus suspension lwcame red and slightly swollen, and the temperature rose to 37.2"C. The swe!liylg mcnsured 4 by .5 cm. It was not tender, but there was itching. The swell~~lg disappeared in 3 days. (c) A mild attack of arthritis began 14 days after vaccination in Case 10. The fever lasted 3 days There was tenderness for 4 days in several joints, including the knees, the hiI:-.: an elbow, and the temporomaxillary articulation. Blood taken on the 2nd da>, of fcvcr and injected into a rhcszts monkey did not immunize the animal against a later injection of callow fever virus. The vaccinated person had had a similar nf tack of arthritis 11 months earlier, and he has had another such attack along \vith a "cold" 53 months after vaccination. (d) 13 da>Ts after Ivaccination, in Case 1.5, there was slight soreness of an ankle on motion. (A simil:tr condition in the right elbow was reported as being present in the case of II. W. 1;. from 6 to 13 da~rs after vaccination in Nigeria with i'accine 3 and Serum Pool 2.) In the 2 instances in which there were late reactions at sites of injection of the virus suspension, ttllb injcctcd mntcrial had not been centrifuged or filter-cd and therefore contained tissue particles. It seems probable that these inflammations _ of skin and lymph gl:~nd and also the arthritis were late reactions, of thta serum disease type, to the injected foreign tissue. lZccortling to Boots and Swiit (23), serum sickness is In:lnifcst ccl during the 2nd anal 3rd ~vccks following serum trent- ment b!r fever, l~ml~h notle cnlnrgcmcnt, urticnria, transitory leucopcnia followed by Icucoqrtosis, nncl less frcqucntl~~ 1)~. joint pain and tenderness. 3. S~`~rplotns I'rt~luhi~l JIcl:li)rg :Yo Rt~ltrfior: !n I/R: r'ilc-c-iilnliotr.-In this grr *rip are included (a) rises of temperature follo\ving the estrnction of infected teeth in Cases 1 and 8, and (h) symptoms of "colds" in Casts 6, 11, 14, and 16. 4. Sy~?~p~onzs Possil)i_v J&In/cd fo /hc l'irus Iujcctcd.--In this group are in!-luded symptoms which ma!' have been due to an immunizing yellow fever infection held in check by immune serum, although some of them may have been late reactions to foreign tissue, like the symptoms mentioned under "2." The leucopenia disrwsed in the following section should be classed nith this group. In Cast 1, on tlw 11 th da>, after vaccination there was chilliness, and on the 12th, a rise of temperature to 37.4"C. with hendnche. In Case 6 there were rises of temperature to between 37.2" and 37.9"C. almost daily from the 8th to the 13th days. In Case 7 the tempcra- turc rose to 37.3"C. on the 8th dn~r and headache nnd backache were present. In Cnsc 0 there \vns a rise to 37.2"C. on the 10th da\.. In Case 15 the tenipcrature was 37.2"C. on the 6th day after vaccination. 9.58 YELLOW FEVER VACCINATION The Leucocytic Reaction fo Vaccination To secure evidence on the nature of the immunizing reaction to vac- cination and make possible comparison of the blood cell changes with those observed by Berry and Kitchen (1) in attacks of yellow fever, the following study of the blood of the vaccinated persons was car- ricti out. Total leucocyte counts were made each forenoon, except for a few omissions, beginning 2 or 3 days before vaccination. The results are presented in Table V. The immediate reaction of the white blood cells, observed 4 to 6 hours after the injection of the virus suspension, was in most instances an abrupt rise. This ws followed by a decrease, noticed the next morning. Over the ensuing several days thcrc occurred an additional and more gradual fall, interrupted occasionally by irregular rises, until the count was in most casts below normal. Of 16 persons vaccinated, 14 showed some degree of Icucopcnin between 7 and 10 days after vaccination. In 5 cnscs the low point was below 4,000; in 5, between 4,000 and 4,999; in 4, l~etnwn 5,000 and 6,000. In the remaining 2 casts it was over 6,000, but in 1 of them several counts ncre omitted. Diffcrcntinl white cell counts were made on the first 9 persons vaccinated. They rcveakd no constant cleviation from the normal in the ratios. In 5 of the 9 cases there was an incrcnsc of ncutrophilcs a few hours after vaccination, at the height of the irnmctliate reaction to the foreign tissue. The lcucopenia was due chiefly to a decrease in the number of ncutrophilcs. The changes in the numbers of Iympho- qrtes and monocytes wcrc not suIt;ciently consistent to be significaiit. :\ dcgrcc of lcucopcnia \vas obscrvccl \vhich was in general less than that seen in yellow fever, but the time of occurrcncc was consistent with that in yellow fever if allowance be made for an incubation period. The time from vaccination to Ieuco- pcnia would SCCIII to correspond npprosimatel~~ to the time between onset of illness and Icucopcnin, plus the incubation period. :\s leucopcnia may occur in serum sickness (23), we considered the possibility that the low white cell cou~~ts observed after vaccination might be late reactions to the injection of foreign tissue rather than effects of the virus. Against this intcr- prctation is the observation that definite leucopenia was not observed in any of 3 persons immune to ~rellow fever by reason of attacks of the disease, who were inoculated with mouse-brain tissue as controls, although they showed the usual immediate rise after the injection of the foreign tissue (Table V). Two of these persons (S. 1:. K. and AI. T.) rcccived normal tissue and 1 (IV. A. S.) infective tis- sue. The amount of brain tissue injected subcutaneously into \V. A. S. and S. I;. K. was approximately the same as that given in Case 11 (0.001s gm. per kilo), and the suspension was in immune human serum and \vas unfiltered. To R/I. T. was given a filtcrccl suspension in normal human serum containing 0.003 gm. of virus per kilo minus the loss bar filtration. The small number of these controls still left Total Lczuocyle Cowts nf 16 PusotIs T'accixated agaiusl Yellow Fez,er a?rd of 3 Pcrsom Ciwn Control Injections of Mouse- Brain Tissue Case so. ZZZ I I 1 D. I{. \V. 2 E. II. 3 L. 11. T. 4 \`. G. 5 T. I'. II. 6 I,. II. 51. 7 S. R. 8 I'. J. C. 9 G. \\`. x. 10 J. P. c. 11 J. S. C. 12 31. H. 13 J. H. n. 14 C. R. E. 15 H. E. H. 16 J. H. P. Cor.:rcl Conlrol Control iv. A. s. 11. T. S. F. I;. -I - - Perssn -4 --`; , I J- 1,30( 3,60( -31-2 -1 - - __ `10.100,9,800 4,100,6,500 8,5508.000 i. 200's. so0 10,300 6,OOOi i,JOO 7.100 lO,SOO! 9,&OO a,;00 10.350 8.4OC 6,400 7,100 i.SOO 7,400'7,60( 7,ooc 9,400 9,800 8.60( 4. ioo 8,30( DAYS after vaccination or injection of mouse-brain tissue j-000 7,200 5,450 1 5,6iS 5,550 7,350 7,150 8,450 i, 300 9,150 9.300 8.100 6,200 6,850 7,200 7.050 9,800 10,300 6.600 5.200 9, loo 8,400 - 5,350 6,2OQ6.200 5,800 7,400 5,350 9.200 0.050 * Counts in this column were made on the morning of, and preceding, the vaccination. t Counts in this column v,-ere done from 4 to 6 hours folloGng vaccination, except in Case 15 in which the interval was 1 hour. $ The lowest leucocyte count in each case is underlined. 960 YELLOW FEVER VACCINATION some doubt as to the significance of the leucopenia, and further evidence was there- fore obtained by making total white cell counts on monkeys. Four monkeys which received an amount of filtered virus-containing mouse brain per kilo of weight equal to that most commonly used in vaccinating persons, showed a definite leucopenia with low points from 3 to 6 clays aft&`the injections. Two other monkeys given normal mouse brain and 1 control animal which was not inoculated, showed irregular fluctuations in white cell counts, but no definite leucopcnin. The evidence suggests strongly that the icucopenia observed after vaccination is a reaction to the virus rather than to the mouse-brain tissue. The cffcctivcness of the vaccination of persons against yellow fever was measured by carrying out intrapcritoncal protection tests with their serum in mice (21), and in all 19 casts, including those of the 3 persons vaccinated in Nigeria and Brazil with materials sent from this laboratory, a substantial active immunity was demonstrated within a month after vaccination. As is shown in Table IV the scra of the 16 persons vaccinated in this lnhorntory ncquirctl power to protect mice against yellow fever virus in from 7 to 21 tln\ls nftcr vaccination. 1'11 e serum of each had failed to protect before vacf`ination. As the nctivc immunit>~ had been produced 1~~. injl.i-tions of virus fixed for mice, it was important to find out whcthcr the sc'r;t of the vnccinatcd persons would pro- tect also against strains of swallow fc\~r \rirus other than those adapted to these animals. The scrn of the first 12 persons vaccinx!cd ivere therefore tested in rl~c~sm monkeys against the Asibi strain of yellow fevc:r virus from monkey source, a strain known to be virulent for monkey and man. In every case the serum pro- tected cnch of 2 monke!-s against death, and in most cxses against fever also, when injected intrnperitoneall!, in the amount of 1.5 cc. per `-:ilo of body weight. That passive immunitJ7 pln~ved no part in the obscr\red protection by the sera of vaccinated persons was prolretl bar testing the sern of the first 3 persons, by the pro- tection test in mice, approsimatel>- 2, 6, and 24 hours after vaccination. Two negative results and 1 inconclusive result were obtained for each time interval. Moreover, in most cases, including Cases 15 and 16 in which the serum was given in larger dosnge, there 1~3s eviclcnce :Igainst the prcscncc of passive immunity in the negative and inconclusive results 1xtwec11 the time of vaccination-and the first posi tille. The rise of the protective power of the sern of vaccinated persons and the subsequent fluctuations were followed by titration in mice. Protective power of swa of ~WSOXIS vaccinated against yellow fever or immunized by illness Titration by intraperitoneal protection test in mice Weeks after vaccination 2 4 6 8 10 12 14 16 18 `20 22 24 26 28 30 Case numbers in Roman numerals FIG. 1. Titres of human sera taken at various intervals after vaccination or after attacks of yellow fever. 962 YELLOW FEVER VACCINATION The results of the titrations are brought together in Fig. 1. We are indebted to Dr. T. P. Hughes and Dr. II. Moskins for making the titrations, and to Dr. J. S. Cunningham and Dr. J. H. Bsuer for assist- ance in preparing the many serial dilutions required. For comparison, the titres of sera from persons who have passed through attacks of yellow fever arc shown in Fig. 1. The G recent cases are mentioned in the publication by Berry and Kitchen (1). For the sera of the 3 early cases we are indebted to Colonel J. F. Siler, Chief Health Officer of the Canal Zone, and Dr. L. B. Bates, Chief of the Board of Health Laboratory, Canal Zone. A. W. C. had yellow fever in 1901 in Cuba in one of the experiments of the Army Commis- sion under Walter Reed. J. H. S. and J. C. had their attacks in the Canal Zone in 1905. From Fig. 1 and Table IV it appears that, as a rule, the protective antibodies became demonstrable by tests in mice during the 2nd week after vaccination and increased to a maximum in tl;e following 2 or 3 weeks. After that, they remained at about the` same level for at least 6 months, most of the titres lying between l/64 and l/256.' The average titre was a little lower than the average for the sera from rcccnt cnscs of yellow fever, but the clii`fcrcncc may not have been significant in view of the smnll number of these recovered cases. Of the sern from persons \vho had had yc~llow fever from 26 to 30 years ago, one had a titre at the tom, of the range for vaccinated persons, another's titrc was at the bottom of this range, and the titre of the third was much lower. Tlrc 11rfccti:liLy of t/x Blood nftcr Vaccination If there were yellow fever virus in the blood of vaccinated persons during the process of immunization and the mosquito vector were present, there would be risk of transferring the infection and starting an epidemic. This risk would be serious if the virus were from monkey source, but would seem to be remote if it were the neurotropic strain fixed for mice. Earlier in this paper we have mentioned the difficulty 1 Since the completion of this article titrations of the sew of ,7 of the WC- cinntcd persons have been made from 5 to 10 months after vaccination. In Case 6 a titre of l/S WE obtained 27 weeks after vaccination. In Cases 5 and 11 uncompleted titrations ~110~~ tliat the titrcs lrill be below l/32. W. A. SAWYER, S. F. KITCREN, AND W. LLOYD 963 of carrying infection with the fixed virus from one monkey to another by the transfer of blood. It is, however, theoretically possible that the fundimental changes which have taken place in the virus during its adaptation to mice could be reversed under unusual conditions, and it cannot be too strongly urged that the living yellow fever vaccine virus should not be sent into regions in which yellow fever is absent but its vector is present, or introducetf into other countries without the sanction of the government 2xeaIth authorities. To determine whether virus was present in the blood of vaccinated persons several tests were made. Dr. Max Theiler cooperated by mak- ing many of the tests in mice for infectivity. The intervals bctwccn vaccinntion and the taking of blood were 2, 6, 2-1, and 48 hours in Cast 1 ; 2, 5, and 2~1 hours in Case 2; 2, 5, :tn(l 2-l hours, and 7 days in Case 3; 3, 4, 6, and 7 tla!:s in Cnsc 15; nnd 2 and 3 cla~rs in Cxx 16. Each specimen WLS dcfibrinatcti nnd injected intrnccrebrnlly into 12 or more highly susceptible mice. In most c;tscs the l)loocl was centrifuged xncl the serum injeclcd into onc set of 12 mice anti the ~~ltctl corpuscles into another. In Case 16 .only the serum was used. In no cast was virus rccoverecl. There were n few deaths which might have been due to the lrirtts, but in those inst~nccs in which subinoculntion wns possible or 1iistologic:tl csatninntions of the brain were made, the diagnosis of yellow fcvcr uxs rulctl out. In certr~iti cases some of the blood was injected also into rhcszss monkeys: 10 cc. of the wnshcd 1~100~1 cells of the (,-hour specimen in Case 1, nnd 3.5 cc. of the tlctibrinatccl -I-tlnj. qwcimen :ttttl 5 CC. of the \~:~hed b10od ~11~ of the 7-tlity spcci- mcti in Case 15. l<;tch of the 3 tnonkq5 succumbed Inter to test inoculation. I<\*iclcntl!- the!, hat1 not bectt irntttitttizctl and the i+rus had not been present in the injcctcd blood. It scemcd clear that in e\`cry instance the virus was absent from the blood of the x~~rcinntc~l persons tested, or at least was very small in a1110u11 t. On :lccount of the remaining uncertainty the following cspcrimcnts in n~onke~s were undcrtxken. (a) :\ ~~l~c~.~z~~ tnonkcy was inocul;ttcd sulxutancously ivith 0.3 cc. of Vaccine 4 per kilo of body weight. It receivctl in this way, on the basis of body weight, the usual nmottnt of immune serum usecl in xxccinating persons, and 10 times the usunl amount of x.irus. I3lood was taken for test 2 hours and 20 hours Inter, and each sample W;IS injcctal intmccrcbrally into 24 mice, 12 of which received defibrinntcd blood nnd 13, scrutn. The results were like those obtained in the tests of human blood. Only- n few mice died, and in 1 instance subinoculation excluded yellow faver as the ause of clenth. 964 YELLOW FEVER VACCINATION (b) Two rlmus monkeys were inoculated with the usual amount of virus but only one-tenth the usual quantity of immune serum. The blood of one monkey was tested after 1, 3, 5, 7, and 9 days, and that of the other after 2, 4, 6, 8, and 10 days. In each test 12 mice were inoculated intracerebrally with serum from de- fibrinated blood. The specimens taken 2,3,4, and 5 days after inoculation of the monkeys caused heavy mortality in the mice, and the cause of death was shown to be yellow fever by subinoculation and histological examination. The virus was not recovered from the specimens taken after other periods. (c) Of a group of 4 ~IIESZLS monkeys, 2 were given subcutaneously 0.5 cc. of immune human serum per kilo of body weight. 3 hours later all 4 monkeys were inoculated subcutaneously with 0.03 cc. of Vaccine 6 per kilo. In this vaccine they received the usual amount of virus, but no immune serum. Of the 2 monkeys which had received immune serum, one had no reaction and the other developed fever. From 1 to 7 days after vaccination each monkey was bled on alternate days, 1 of each pair on the even days and the other on the odd days. Twelve or more mice were inoculated intrncercbrnIl>~ with cnch blood specimen. The mice which had rcceivcd blood taken 1, 2, and 3 days after inoculation from the monkeys which had not been given immune strum shovel a high mortality, and the virus was recovered from each group of mice. The virus was not recovered from the blood of the monkey which had received immune serum and virus and had shown no rise of temperature, and there were very few suspicious deaths among the mice receiving this blood. The mice inoculated with blood from the monkey which had been given immune serum and virus but had nc\erthclcss reacted by showing fever, also had a low mortality, but the \rirus was rccolrered from 1 mouse which had been given blood drawn 3 days after vaccination. It seems that the virus was present in \ery small amounts in the blood of the monkey which had not been completely protected by the immune serum. Daily counts of the leucocytes in the blood of each of these 4 monkeys were made. There was leucopcnia in each case, with the lowest number of cells from 4 to 6 days after inoculation. In thcsc tests we were unable to recover the virus from the blood of persons or monkeys vnccinatetl with fixed virus and sufficient immune serum to prevent fever, but the virus was found to be abundant in the sera of monkeys given fised virus ant1 much less than the usual amount of immune serum. I'irus was present in extremely small amount in the blood of 1 monkey which developed fever after receiving almost enough serum. It is concluded that virus may circulate in the blootl after vaccination if the amount of immune serum injcctctl is not ntlcqu:ltc, and that this dnngcrous occurrence may be prevcntcd by using immune serum in excess of minimnl roquircmcnts. W. A. SAWYER, S. F. KITCHEN, AND W. LLOYD 965 The Nature of the Intmmizing Reaction In our experiments it was shown that yellow fever virus could be mixed with proportionately large quantities of immune serum and kept at room temperature during several hours of preparation without losing the power to produce a substantial active immunity after injec- tion into monkey or man. On superficial consitlcration this experience wo~~lcl seem at variance with the frequent observations of others that mixtures of viruses and large amounts of immune serum are without immunizing power. J. II. Bauer (24), in nn unpublishctl cspcriment cited with 11is permission, kept 5 per cent of yellow fcvcr virus (infectious monkey serum) in immune human serum nt 37oC. for 2 hours. lrc then injcctctl the mixture into 10 monkcg in amounts ranging from 0.001 to 3.0 cc. `I'hcrc Jv\`crc no renctions, nnd all but 1 of the monkeys died Inter of yellow fever nftcr test inoculation. `l'hc virus in the mixture hncl evitlcntl~~ been complctcly nc~ltrnlixcd anti hnti lost its immunizing power. In an- other csperimcnt hc sl~owcd that a misture of 1 cc. of immune serum with 0.1 cc. of virus (infectious monkey serum), nftcr esposurc to room temIie%turc in the African tropics for 4.5 minutes, would produce active immunit>~ when injected into 3 mon- lazy, hut that n misturc of the sonic amount of strum and only 0.01 cc. of virus woultl produce no immurlit!~. I~.Yj~~~ri~~~~~~~i :I .--`l`\vo 10 per cent suspensions of the fixed yellow fever virus were nl:L(lc, one in normal human strum and the olher in immune human serum. ?`hey were centrifuged ant1 filtered through 3 Seitz disc. Each suspension wns divided into Tao lots, one of which 1~3s kept c2t 2S"C. and the other at 3oC. The mktures kcI>t at 2S"C. I\TI-C tested after -1, 2-1, 2nd 96 hours, and those at 3oC. after 24 and 96 hours. In cxw~~ test, a monkq* was inoculated with 0.3 cc. of the suspension per kilo of hotly weight, and in ill1 esccpt the tests of tlic k-hour specimens 0.03 CC. of the material ~9s injected intrnccrcbrxily into cxcli of 6 susceptible mice. All the monkc\x which rcceivcd mixtures containing normal serum and the one given tlic 4hour specimen containing immune serum became highIy immune, for their blood serum tlevcloped :L high titre for protective antibodies, ranging from l/G4 to l/512. ~Ioreover, ,211 thcsc monkeys later survived test inoculation, except 1 \rhich did of other cz~use than yellow fever. Every mouse inoculated with the snmcf mixtures died, presumabl>r of yellow fever. The monkeys which received the mixture containing immune serum, after it had been exposed to tem- 966 YELLOW FEVER VACCINATION peratures of 28oC. or 3oC. for 24 hours or longer were not immunized, for all died of yellow fever after test injections, and their serum failed to protect. The mice inoculated with the same mixtures survived, except 1 in each of 2 sets. From this evidence it is apparent that the virus in a filtered 10 per cent suspen- sion of fixed virus in immune serum was not neutralized in 4 hours at 28'C. but that it was entirely neutralized and inert after 24 hours at 28" or 3oC. E~pcrinrc~/ B.--A filtered 10 per cent suspension of fixed yellow fever virus in normal monkey serum was divided into two parts. To one was added an equal part of normal monkey serum and to the other an equal part of immune monkey serum. In testing these mixtures in monkeys, 0.06 cc. per kilo of weight was in- jected subcutaneously. Each mixture was placed at 37oC. and tested immediately and after 0.5, 2, and 6 hours. All the mixtures without immune serum, and also those which contained immune strum but which had been kept at 37oC. for 2 hours or less, produced immunity in monkeys and killed nearly all the mice inoculated inlr;~ccrcl~r:tlly. But t hc mislurc of immune strum nn(l fixctl virus, which had been kept at 37oC. for 6 hours, permitted all but 1 of the 6 mice inoculated to sur- vive. Two monkeys injected with the same mixture survived test inoculation, but the titrcs of their scra wcrc low (1 /,JZ L ?ncl l/6-1), while the titrcs of the sera of the other monkeys in the cspcrimcn t were about l/256. One of the 2 monkeys showed the typical leucopenin and the other did not. It appears that the immune serum in this espcrirnent had not neutralized the virus in 2 hours but had neutralized it almost, but not. quite, complctcly in 6 hours. Esfic~vir~c~111 C:.----In a thirtl cspcrinicnt the same basic 10 per cent suspension of fixed virus in norm;ll serum ~~3s used as in Experiment B,-`&;d equal volumes of varying dilutions of the immune monkey strum used in that experiment were added. The final concentrations of immune serum in the mixtures were l/2, l/8, l/32, l/12& l/512, and 0. `I'hc mixtures were kept at 37oC. for 48 hours to per- mit maximum neutralization, and then each was injected intracerebrally into 6 mice. &Iistures with immune serum concentrations of l/32 or over permitted all, or all but 1, of the mice in each set of 6 to survive, but the mixtures with concen- trations of l/lZS or less, caused ,211, or all but 1, of the mice to die. It follows that the minimum concentration of the immune serum which would neutralize the 5 per cent content of fixed virus, minus the loss by filtration, in 48 hours at 37"C., lay between l/32 and l/128. From these experiments the following conclusions are drawn. ((I) Neutralization of yellow fever virus by immune serum takes place very slowly at room temperature. It must have been far from compIete in the misturcs of immune serum anti virus used t>y us in immunizing man and monkey, and it must have ceased when the mix- tures were dried. (b) When yellow fever virus is compIetcIy neutdizetl by immune serum NIL ~~jilr0, it loses its power to immunize or infect. W. A. SAWYER, S. F. KITCHEN, AND W. LLOYD 967 (G) The injection of a mixture of yellow fever virus and immune serum, under the conditions of our experiments, is approximately equivalent to the simultaneous injection of the two components. The observation that the degree of immunity produced by vaccina- tion was almost as high as that following the naturally acquired disease was contrary to expectation, as it had seemed probable that an amount of immune serum capable of inhibiting symptoms would also limit the immunizing reaction. To find out whether extreme quantities of serum would prevent or diminish the production of immunity the following experiment was performed. A monkey was inoculated subcutaneously with 6 cc. of immune monkey serum per kilo of body weight ant1 1 hour later it wx inoculntccl subcutaneously with 0.06 cc. (per kilo) of tlic snmc 5 per cent filtered suspension of fisetl virus in normal monkqr serum ns ~1s usctl in ICspcrimcnt 13. The amount of immune serum was 20 times the usu;~l amount usccl in vnccination and corresponded to 420 cc. for a man weighing 70 kilos. 1-l cla~5 after the injections, when pnSsZVc immunity was still present, the protective po\vcr of the monkey's serum was low (titre, l/16), and 62 clnys dtcr the injections the serum protected only when tested in full strength (titrc, l/l). TIail>~ lcucoc!.te counts showed that the usual leucopenia did not occur. Elcvntions of tcmpxlturc were observed and attributed to the excessive amount of strum injectccl. :\ test inoculntion of virus of the Asibi strain was given 7.3 d:l~3 after t Ilc injections. b,hcn some passi\re immunity was probably present. The monkc~~ sur\Tivcci. The injection of cstrcmcly hrgc amounts of immune serum may grcntly climinish7 ~1~1 perhaps prevent, the production of protective nntibotlics in the blooti after injectiorls of yellow fever virus. `l'hc minimal amount of immune serum, per kilo of body wejght, wliic-h will protect monlitxys against the highly virulent Asibi strain of ycl10w fcvcr virus nppears, from our experience, to be also the minimal :LIlIOunt nccdctl for protection of men or monkeys flg:riinst: the fisp4 strflin hnving 10w virulence for them. SUMMARY XND CONCTAUSTONS 1. After preliminary experiments in monkeys, 1.5 persons were activcbly inimunizecl l)y ;L single injection of 3, tlriccl !nislurc of living yellow fcvcr virus, fisctl for niicc, and hurnnn init*;:!nc sawn, with scpnratc injections of enough ;Idtlitionnl wr11rV 1L? t's' . `s Tit> !Ln* `