Oct.12,4955 Dear Joshua, Last monts:,and the beginning of this one were spent on » course of statistics 7 have organised and thus I was unable to answer you earlier. Thank you very much for your letter.Let me deal first with the part of khis it that deais witn the paper on resistance. Journal. I would certainly agree for venetics, Vacterio- logists who may be interested cun receive it as a reprint.The paper is certainly no less geneticul than Luria and Delbritick's,whien did appear on Genetics, Title: yours much better. Aduresses of the author s: if you like to give as a heading both the Istituto “3ieroterapico Milanese and your Dept., it would have the atwahkage of psyine for my reprints.I heve no idea how mucn they woula cost,but I suess it would already be in the range where one likes to have them paid by some Institution. @khaxnisea,xyausganxkgsep we may indic.ite in the aemmowledgments that tne «ork was done,or partly done,wgile I was working in your Department on a Rockefelier fellowship. Order of authorsnip : it is very kina of you to sive me priority. However, T still believe tnut tne iuea Was more imoortant than the experiment and the iuea was yours. Introduction .Your text is 0.%.,I1 have only some objection to tne use of the word vehement, just because it is somewhat funny in Italian (your second line of first page).T have no substitute to sugrest,so,unless you have some brignt idaea,it is just as well to keep it. Would you not quote,at the sane place, also the 3...pesiun on Adaptation of tne Society of Ceneral Microbio- ogy? I am afraid T heve no exact reference,but just tse proots of the book, Your page_3.T have no special reason for wRing preferring the differentéak growth rate,us you do, or the relative growth rate of the mutant,as I did,except that the latter method was used for calculating the values in table 4 ,and that pe:haps the ie- lative growth rate may be less sensitive to caanges in environuental conditions than tne difference in absbfute e&rowth rates*#* as to for- mula (1) of my paper,page 4; I agree it is a poor Looking one, but perhaps it shows the derivation.Tnae introduction. oF Ry instead of et permits to use directly experinental data;stch 28 T),r,,and R for the calculation of the k values: T have used the arproximations you sugrvest for the calculation of the k-values, but the approgima- tion is not good enoupn for the last two cycles when rjuvand r ure relatively high. Also, forget:the experisnents do not eiva easi- ly a clue as to absolute growtn rates,but only of relative ones, I shoula like to suggest tuat your middlepiece, page3,be substitu- ted as given in my appendix 1 to this letter,This also contains a proposed chunge fbxt would end the introduction. I felt one should add probably fie “2 footnote, whet is there given about tne effect of fluctuation of the initial number of resistants in the inoculug. Nothing else Bbout tne introduction.Now,ahout your remur § to we bod, of the 2 REPEL s tL. gntgrely agree about deletion of tne SS sup- pr ssor{’t T *1s0 agree about your adcition on stability of the strep tomycin resistant.It is absolutely rieht that best and fertile,or rather,as you suggest, positive,should be defined exactly as you sugeest. pe7 0.k.sp.8 one mica he peancet reconstruction _/ experiments with mixtures of etebe ion and resistant mutants of E.coli, ceppeliini (1954) obtained = waxy closely similar value." to be added at the end of the last paragraph, page 6. 2s Table 4, Cycle 3. I am sorry there was a mistaxe, m should be 2, instead of 2.5,and therefore E = 6.5 . The k value is correct as f&iven,. O.k. about 11 the rest concerning tne epxerimental part. Answeringe your further questions about it : Table 9 does refer to, table 6, as you indicats:,and not to table 7. In tuble 9 it is/étiture with 814 which scould have « double errdéw (or better, boldface itulics,as you sug est). Stability is being tested. Tne mutentg is stiil resi- stant at the moment; =~ shali send you a note sbout it as soon as EK an expe.iment sim’ lar to yours is completed. Discussion, I h.ve a few comments about some points thut do not seem clear enough. I do not understand exactly what you mean by the two sentences : (middle of page 1,discussion): To sive the obtained result,ete. and : whether envipomuentel differences among the cul- ture tubes ete. Could you expbain more at length? As to your que- stion on kam aata,i am giving whet I have about it in appendix 2. Discussion page 2,middle : Your sentence : "nen eitner method of indirect selection should ap car to fail,i shoula at least be s..own that it is applicable to tne recovery 0f/% Eoteu elones artificially adued to the tested vopulation" T would aid at the end of it (if this is really what you meant): "if the nerative result oF indirect selection has to serve us valid proof ug:‘inst preadaptation". Al. the r st is 6.k. About acknowledgments to Fockefe__e1,terhaps you will find a better wording then I sugsested above. I um certain- ly srateful to them for having taken me to Madison. Your acknowledg- metts to NCI,NIn etc. are of course ok, As to the rest of your l:tter,and in particular your mention of the possibility that i may take’a job in the States. Needless to say,ig am a:preciating gweatly your interest on my behalf.Tf a full professorship were available here,I thin«x I would prefer it to one elsewhere,essentially because it gives a sreat liberty,ewten if t:.e Salary is meagre (on american standards,und also,to soie extent,on Ttalian stanuards).No fuli prof'ship in Genetics aas been available since 1947,and it is not yet @tear when one will be available, The- refore my willingness to stay here decreases with time.I nave not fixed a critical point for its decrease,or rather IT am still above the eriticul point. However,l should like to be kept informed of good opportunities, Byga,.° if any turn up. I had forgotten : fis.2 (or rather,what was fis.2,and is kh now fig.l1) is enclosea.T have more copies if useful.T think the lubbts with words stuck on it will not be noted in the cliché;tnis is tne stuncvard way in which German journuls used to prepare their figures for publication. . I have nothing interesting in the lab,or at least nothing suf- ficiently clear or safe. Anda you? Ali the best to you and Estner,from the both of us. # Moases Ate, ond Appendix 1 ,page 2 inoculated with more than on@resistant mutant. Since the expected enrichments are calculated on tae assumption that only one ressstant mutant is inoculated int» the positive tube, a correction may be ne- cessary,but the following urgument shows that the correctiog is small (and therefore was neglected) xmskmespresent . The proportion of tubes out of n inoculated with a single sample gxskxix containing m bacteria and distrisuted equally to them (so taut euch tube receives n/n bac- teria on aver.ge) , which xaggivg,(on uveraging different experiments) receive 0, 1,2... resistant mutants is given by the expansion of the binomial : 1 We select only the"positive"tubes, and therefoie we must corect tne average number of bacteria per tube, n/n ppggine only tne nositive tubes into account.The proportion of meEeatian tubes is,from the above formula : (aS m +t oof Coop ton positive n-l.m lpositive tubes therefore is 1 - (——) n the expected number of bacteiia per Fersine tube is then : osiGve a pen tt =—1 ; je App. tf. 4 nP1- npr - &h*] 1 hem A Aer which exceeds Yne by « small anount, e.g. for me= 2, n =10,the most extreme instance here recorded,the correction factor is 1,055. and tha (t hope what given in the footnote is correct;I should have a cnance getting a more compiete reasoning before roofstage,anyhow.§ 17 a be 4 : we / a pr 7 SFT Appendix 2 Josoua, you ask about more data or coments on the "heritability of the variance". This is all I have about it. The jackpot culture (No.41),containing 10 resistants/ml was expan- ded by seeding 1 ml of it in 100 ml. of broth, Pma.fvash.sulturax masukapi.in?tt¥ni care tor. faxakO.daxa Exactly the same was done with culture No, AOD emt mutants control, From each/culturé, 10 new cultures were prepared,each by adding 1 ml to Penassay tubes of 7 mls.The whole cultwres were then plated on streptrycin agar,and tne -ollowing counts were obtained: From cult.41: 135,142,181,123,177,111,138,149,140,144; average 144 variance 461 Hrs 28.808) From culture 42: 0,0,8,2,0,0,0,1,0,1 average 1.@ variance 62 A second enlargement of jackpot and of cultmwre 42 were made from tae 100 ml cultures,about 290 days later. Cult.41 gave an avera- ge count of 22 resistunts /ml,and culture 42 an average of 0.3 (ave- rages based on 3 plates). Ho you think you can use this data? If so,I leave to you where to introduce théx information. .