| : ’ 4 . 7 Thess ISTITUTO SIEROTERAPICO MILANESE “SERAFINO BELFANTI~ Ente Morale Aggregato alla Universita di Milano 1 | 10/7/53 Via Darwin, 20°- MILANO Dear Joshua, ‘ , FEL. 36.646 - 0.677 - 51.757 - 33,023 - 33.917 ‘Your last letter just received. eanwhile proofs have come of our joint paper; I have accepted all your corrections (except & few minor ones, and these mainly for typographical reasons;they are marked NO on the wenuseript). I an. thefefore sending back your pages, so that you can recon strhute the full manuscript; I hed always forgotten to send the second copy. Re the st. tistical issue : I have dropped the relevs nt peragragn of little importance.I do not think I was wrong,though I certainly oversrted _ the importance of the ‘hypothetical model leading to extra Yoissonian varian- ce.Just to explain me better than before with an extreme example: supoos you seed 10° cell, »each from a parallel culture,on drug platessand each cul- ture contains 10 gise. 10 cells onsaverege,of some peculiar type ysay 2X snakes or some other odd type of cells. Suppasexkhe Actually,their number Will vary from cultu-e to culture even in the most strictly identical con= ditions,acco ding to a Poisson ;there uay be cultures which have 5 snakes, others which have 20 etc. Suppose,moreover, that the probability of a normal cell getting resistant is O,while snakes have a probs bility of 10% of get- ting resistant by sone lamarckian mechanism, There will be on average 1 re-~ sistant "mutant" per culture,but their number will vary from culture tio culture with a vyetience higher than 2 Poisson,winmemexkke like a Poisson su- psrimposed on another (Neyman's contagious distribution » I think 3 no’ exact quotation at. hand). The scheme could be e#sily extended, for inst.znce to a . continuous variation of the probability of giving resistance for different tvpes of mxtr ofd (rere) forms.It is essential that the, odd type of cells, which can become resistant,are rare;otherwise thefw aype ges- er culture will show little variation fron culture to culture. ‘I am enclosing cultures of W 5&3~-not checked here in the last ,two or three years,I am @f@fraid,but it could not improved upon if found unsa- ‘tisfictorysI hope it will meet your needs. “Also enclosing strain 8 (MjFTNyT) and an FY recombinant from it x W 945. The att rocan be infected but with — considerable aifficultysit still segrezates/in he progzeny. These strains were promise’ months ago, and not sent, Sorry. They stayed on my table and got lost among sheets. at Madis ; Thank you very much for your generous offer éonce:ning: my staye I shali see what I shel! be able to do about improving on dates of my coming. I have written to the Rockefeller,who answered an an uncompromising way* WHO is a good idea, but I have no;g00da strings to pull there.Fulbright is enother chance, for the journey only. . 7 ++.Pomerut says he will probably come to Eyrope next autumn and may be able -.to get some decision at thst time. Too many appiicants probsbly. , I had a hope of restarting some cytology, by taking movies of © -, @n. afr plate. at the microscopes This project has to be stoppe? until Octo- ber. I hope we shall ‘be able to clhaborate on the sybolosy beforme my coming there. I have considerevls” objectioy 3. to lookins wyself through a microscope because i get. easily boredg ut the idea thet I might have a camera looking for me has revived. my ‘interest for its” Jinks! coming allowed me to restart on the work of formal genetics of onkrd 26, Ty, the Jast five or six weeks it has been possible to work at full “oo Rabe kLbhough f had very little time to spend in the lab.I have often op- portunities for good technical help,and this time it could be exploited _ completely because of Jinks" constant supervision. Some ten thousand proto- _trophs or more were thus sored end the work is not yet ‘finished. The pro-_ "gramme which I was thus able to set to work was the following : obtain, from Hfr x TLB,-sugar narked ‘erosses, F- phogeny of all combinations. for Met, Lac,Az pyAva(linked with TL) and Gal (linked, to Lac,left of it). Make | orosses in all possible combinationskselecting always f6rom M-ST x ii+s8 on minimal streptomycin,efter infection with F+ of either or both parents. ' An important by-product of this “research was the following.When Gal+ - recombinants from HfrGal+ x W 945 (Gal-F-) were selected for(they are very | a euze,certainly below oned | "and possibly much rarer), HE1. these recombinants proved to be Her{Ro°mRtter whet they were for other ‘markers Thus, Frr can segregate reg ularly,more- ‘or less like any other locus; put it is usuaklkyx wn SO * _ . eliminated, and so are the markers linked with 7? elimination being tese le drastic,for markers nome distant from it,like or Bat. | - , This prompted me to advance the ‘following “Sneory, A single point (it re. may be. a short region,but of ae or this later) of the F+ chromosome is eli- D pinate fa is very meet ear ee We cen bp eS tbe" (without any really “Serious reason at this time). If this is true,irrespective of whether an F+ or an afr is crossed to an F-,the fo? lowing consequences would be drawn : : : 1) our crosses. lead to a map (with a single chromosome.) it-St-Bée-Gal- Lac 4 AgeAra. There are difficulties of pairing especially at “Bén-Gal, ‘less in the rest (see also Rothéels data, TL- supplemented. cross )5 paréjng varies signi- _ Zicantly from line to line. mo 2) The standard cross M-F+ x TLB -F} is consistent with this theory. The 4 point left of Lac,mapped as M ,is actually the centromere, which must al- ways be that of the F- parent. The best support of the theory comes from |, Newcombe and Nyk jme's data. tarker lap’ as follows i M-Xyl-Mal-S-Gal- Ara-Bbn— -Lac-T. —BD. Only double and quadruple crossovers between w and Cen will be -viable.No quadruple is foundsdoubles show some ‘slight negative interference’ (incomplete pairing? this seems uNavoidable on ahy theory). Similar results | from your cross A,table 5, ,CSHS 1951; cross B is B+ x F+.; also from my cress” 1950, Boll.15.m. . 3) Data from our JG paper Consistent with the theory,assuming map B, (af)= : t+ oKyl- -Mal- S-Cen-Gal-Lac- L(Ara)T. In crosses TLB -STF+ x TLBL+F-, double c.o- necessary. in region xyeftxaf right of. Cen for recombination of Lac or Gals therefore very. mrere (rarity accentuated by pairing troubles). s - hore. deta cane from F+xF+ crosses ‘in _ M-S?T programme. Here, both; F+xF+ and Hert P+ are better than F+xF- or Hfr-F- crosses;yieldsare as bigh or. higher than F+xF— or HfrxF- crosses,and elimination or pairing troubles Beem smaller. This is,I think, the best point now available for postzygotic ~""" elamination®#inex. Since an HfErxF+ or F+xF+ cross works with less elimina-— ‘one -ssumes F and Herr occupy the same locus) and elimination of the F+ (correlated. These conclusions may be affected(or be made easier)by the “fact that there is here selection for the S-Mal region gusually eliminated. ISTITUTO SIEROTERAPICO MILANESE “SERAFINO BELFANTI* Ente Morale Aggregato alla Universita di Milano Via Darwin, 20 - MILANO 3 __ TRL, 30.640 - $0.677 - 31.787 - 32.823 - 33.917 tion than when/ohe parent meting as F-,the conclusion seems unescapable - . . : s e that elimination must be postzygotic anit /the consequences of the interaction between the two parental centromeres,unless one assumes that the F- parent “e@vokes the formation of gametes withxa from the F+ strain,which have lost sgements,while the F+ do not. Possibly,infection of the F~(centromere,if centromere(and linked markers which have not crossed over) are strictly Two points need consideration now. Is Gal of W 945,which is linked with Hfr,the same as Gal ? I have no Gal. in my collection. From Rita,Rome, I had a Lembda®Gal-straif, W1294 5 wkieh