STITUTL DTEROTERAPIGO MILANESE
SEMAFINO DELEANTD

ENTE MORALE AGGREGATO ALLA UNIVERSITA DI MILANO
MILANO - VIA DARWIN, 20

TELEF,. 30640-30677 -31757 -32823 -33917 TELEGR.: SIEROTERAPICO ~- MILANO U.P.t, ©. MILANO N, 48709

Milano,. May 19, 1953

 

Dear Joshua,enclosed is the final draft of the joint paper,to be
published on Chain's symposium. When your comments came I rewrote
the paper taking your suggestions into account,and resisting as much
as . could the temptation of altering the text while polishing it up.
Jinks read the whole thing and suggested a number of alterations of
the place Sr prepositions and conjunctions -of course,in my own
part. You probably overrate my capacity Zor writing English . The
only paper which was written entizxely by myself is the one on Here-
dity, chloramphenicdb-resistance ;3/ Re WeuieO. paper ewes ~ owe much

to the kindness of Yr.Ercoli,formerly director of this Istituto and
now working at Armour's.Unfortunately ,due to a mistake,his help in
the correction of the manuscript was not acknowledged in the paper.

I hope you will find the present work tolerably good.However,
there is a chance to improve it,wince Chain,to whom 1 e.plained on
Monday last that you had not seen the final version,promised tq send
me proofs. ~{ do not know when proofs will actually come,so it would.
be best if you could send me a copy of the manuscript with your dor.
rections if as soon as you can manage. Two copies enclosed so that
you can keep one. m,

The first part was altered more than the rest,in order to follow
your suggestions . 1 dropped Law's quotation (Texas Repotts Biol.&
Med. 10:571) because it is not quite clear whether his test is a flue-
tuation test or a correlation between relatives test;it is probably
a mixture of the two,since wight of tumor/dependx both on absolute ~
number of resistant cells (fluctuation test) and growth rate of indi-
Vidual mutations(correl.betw.rel.). Some hihliog. references may be
incomplete (e.g., Catlin's). As to the negative binomial story,1 have
kept it (but am ready to drop it in proofs). I think the variance.of
numbers of resistants wauid in parallel cultures would hay¥e-s-vericnee
be definitely greater tnan the mean if the probzbility of becoming
resistant were #4¢gkey variable from celi to cell(proportional to
cell size,for instance,to give an unlikely physiological example).
However, i am not a mathematician erhaps because of this reason
. fully agree with you as to thelpersuading power of indirect mitne-
mitical proofs,including kinetics; alt noggin yin the end,all proofs
may be put in biometrical terms,axmepkxkkxk Phe kinds of proofs of
which replica plating is a splendid example need not be put in bio-
metrical terms .to be convincing,and derive vart of their strenghh
from this fact] and cannot consult anybody here on the issue,but you
may in Madison, The same kind of srgument for deriving say, eb+nemtald
a negative binomial,far has been employed by Greenwood and Yule to
explain the non-poissonian distribution of deaths by accidents in in-
dustry. They assumed a variible degree of accident proneness
in the populationgand if this degree were vari:ble according to
2 distribution of chi square tyne (Eulerian or vearson type IIT)
the observed distribution of number of deaths per year should
be the negative binomial,which in fact they found. Of course, the
extra parameter permits a wide adjustability and the hypothesis
has no biological interest.It serves to stress that ,when testing
for mutation and selection versus physiological adaptation, the
latter cannot be denied by the simple assumption of a Poisson
distribution ,and the finding that this will not fit. Therefore
I believe the fluctuation test is nok test at all. I have kept
its original designation to avoid the introduction of further
nomenclature. :
I nope you will take a free hand in the discussion of the manu-
script. As to reprints I have noted your wishes.
Work : Calef,working here, has crossed coli B to K-l2sthe fre- .
quency of recombineztion is low,but higher with Hfr. (Demeree will
ssl now be xkigxks unexcusable if he does not carry out genetic ana-
deep hy ses of his setratne mutants). The interesting thine is that cros-
we sing a Tr-coli B to Hfr S¥ on minimal sm tryptophane,the Tr- sr
| recombinants are all Hfr ! Both when crossed to K-l2 F- and to
wew, | coli B. Coli B behaves as F— and when crossed to K-12 F+ gives
wy {F- progeny;it cannot be infected. These results are highly preli-
ye minary and T ope they will stand further trials,

ene have been quite convinced by your idea of a book on Genetics
Bey" ond Chemotherapy. Would there be a chance of writing it jointly.

wert? if I could come to the USA for a period?

X Perhaps next summer may-be (1954) will be my first opportunity,
for trying to get a leave. I em bound here both by routine and by
acadenic duties,and summer is unfortunately the best time.If you
spend at least part of it in Madison,? wight gee in the meanwhile
a few other places off interest,while you are away.June-October is
the longest stretch of time I cam think of ,in view of my present
duties. You wrote some time ago that there may be a chance of get-
ting local funds for my livang expenses;it would be easier if
had to get money only for the journey,or little more. I should li-
ke to come with my wife to give me a help in the lab anc in the
library,but would ld@ave the rest of the family home. If you let
me know your, opinion, ond {£ ean confirm the prospects of a leave,

+ ghall ed&beke “fhe Rockeféller first,and then perhaps .H.0.
Yours 1

4A