April 186, 1953 Dear Spice! Esther and I were greatly distressed to hear of your bout with Brucella. We hope we ahali not to have to wait too long to hear of your permanent, complete and prompt recovery. You indicated this was a iub. infection: do you know how it happened? Evidently we were only lucky aot to have picked up some other bugs around Atlanta! Witness Ike's recent attack of fcod polsoniag (picked up at Augusta, justi prior to his polimy speech a couple of days ago-— or do you get to hear such details in your press? I will adalt one fish platter at Atiaiuta left me feeling rather dubious, bus I terminated the expariaent before the outcome was well-defined. fo tara to in vitro Salmonelloses, I agrsa with you oumpistely aout the fuidiiity of chasing every wili-of-the-wisp of sross-reactions.’ I think, however, tnat Felix’ atatemente (as opposed to his practice; are rather extreme ani that we have to take and use the K-W scheme for whatever 1t da worth. It would be very helpful if somaone more erlilice. tha& K, but stiil sympathetic with the purposes of the scheme would review the whole question of the serological structure of the group. I suspect either JT or PRE could do a Zine job, if they could be persuaded to do it. [I note, by the way, that a b/1,2... crosa-reaction to a titer of 1:200 Is recorded in Edwards & Bruner's Kentucky ciroular]. I domm't want to gc into this sort of thing ayself any more ddeply than 1 have to,w settle ambiguities that may arise in other problems. The only real point of the c-c' story is its bearing on monophasicity, and the selective or inductive effects of seruns. Ohiously, one has to be well aware of evan the minor cross-reactions in using serums for this purpose. Edwards, at, the moment, seems to suspect that c' is the somatic antigen, which would put the formalin effect mre or less in iine with your H9OLK. I am not too deeply concerned about this, except that in ay single trial c' appeared to bs tharaclabile. The more in- portant point is that this reaction will explain the effectiveness of some serums and not others in provoking the co phase from kunzendiorf. As time passes, my recollection for the details of your sxperimaats becomes rapidiy diamer. But was not the O-inagglutinability of your 901K only part of the story? I thought you had a IV-V serum which agglutinated 901K, but did not aggiutinate alcoholised or 9-901 in control titrations. Have you identified thia reaction? The anomalous 1,2 phase story gets more complex daily. There are two java strains to keap in :ind, N25 and NJ7. On a single occasion, each of these strains have given rise to i,2.. phases, apparently serologically identically with the second phase of typical paraB. N25 ph2 is Edwards #157, has never shown any alternative phages in serum selections (/}.423 /405; /2), and in all tranaductions, to and fro, behaves as if it were Hy Hp", ao that, for example, #157 x S. miami (/a,5) gives 1,2:1,5. I have not done ach with N25 ph 1, but one experiment is consistent with ite being H,>. N97 ph 2 has not been atudied serologically as fully es #157, but appears to be the same, However; it fairly readily gives back a ph 1 when selected in 1,2 serum. These phases however show a strong but variable reaction with 333 which is unlike the original N25. as well as b/) This may not mean too mech: both the original N25 and the return bz33 will give rise to 233 phases more or less regularly after prolonged selection in b serum. The sequences can be diagrammsd: I would not place too mich emphasis on the . irreversibility of phases until the possibilities Ip fone of interfering cross-reactions are worked out. wT AD N25 bz gath, aes ‘All of this would be entirely consistent with ib oO wv a set of mtations from one allele to another. + ada One could in fact put the b complex down as: 5 a ~—l Ib ess b.233.1,2....., with the components usually expressed in the order given. This would have nothing in particular to do with the typical phase variation of b:1,2 in paraB. I for- got to mention that N25.2 also behaves as Ht, age, in —x S. miahds. But last week I got anotner startling result which has been very upsetting: SW 623 (= TM ~-x S666, i:—) —x N25 gave an i transduction whose second phase proved to be b. Aa with N25.2--1, the b phase was not, however, reversible. But one still would represent this bea as iV. V XII dsb. For a time I took this as evidence that N25 is genotypically Reage? yH.°, but with a very sluggish phase variation. This was see a however , by W25——x S. miami, and the result, IX XII b:1,5 shows that the b phase pf N29 is a phase 1 homologue. We have the situation, therefore, where the b; 1,2; and (by some indirect evidence) 233 phases of MZ N97 are all homologous with the a phase of miami and tie i of TM, is “6? ab all Hy alleles. This does not account for the b:i anomaly, unless N&x N25 is H, , which gaems unreasonable. Some further tests have to he made with the b:i stock. 4G determine whethe: i4 i. actually undergoing typical phase variation, i.e., a shift between “ave loci. Sco fur there is no specific evidence of the homology behavior of the b and 1 factors. It may be that too much emphasis 4s being placed an tha single cecurrence cf this tyoc, an’ tnat it plays no real part in the scheme: I have to see whether it can be obtained reproducibly, We've just received the SGM preprints for the adaptution symposium. I'n really quite sur- prised at the numerical strength of the direct adaptationists. I hone someone pointed out Hinshelvood's improper application of : ladirect. selection, (galley 15). It is a little hard tc see hew 2x10" cells could hava given visibly single colcniss!¢ It is nowhere mlear whether he had actually identified a mutant clore in his first plating; his second plating, in which one cf a hundred colonies gave rise to 2 mtsait mst certainly have been a completely independent mutation occurring late in the Jevelopment of this colony. On this basis it is not surprising that