April 2, 1947.
Dear Luria- poe

Your manuscript arrived this morning, and I hawe read it once
quiakly. Allow me to offer my cengratulations on a well-written and
very useful paper. I have seme comments which are proveked by the
paper, but are not intended as criticisms ef it, except that
in the bibliegraphy, references 91 and 92 should de interchanged if
the text is to remain wnaltered. Yeu may have everlooked this in
cerrecting proof.

The phenomenon of phase variation deserves some genetic interpreta-
tion, In Salmonella, the alternate phases may be contrelled by the
intermutable allelomerphic states of a single gene, but in addition
the composition ef each phase can be altered separately, presumablg
by mutational loss. The phenemenon probably cunnet be elucidates without
recombinational analysis,

AS to the mechanism of 'Dauermedifications'’ I think the recent work
ef Kimball on acquired resistance te antibody in Psramecium is the
Clearest answer, and may be generally avplicable.However, one should
net expect to find ewidence for oyteplasmic systems in studies on
matational varfations fer this reason. Variation controlled by gene
mutation will eccur fellowing a single discrete act: gene mutation,
Cyteplasmic variation would demand the simul&&neous occurrence of
Many ‘mitations' if the entire set ef plasmagenes is to be altered. It
would appear mere prefitable to investigate natural groups which have
been genetically isolated from ene another fer evolutionarily signifi-

cant periods. The only exceptions are when differential growth rates
of cells and plasmagenes can be established.
Finally, I weuid suggest that the eredit for the first clearcut (albeit
unsuccessful) genetic appreach to bacterial gene recombination be
given to Sherman and Wing, o§ whose work . Gewen and Lincoln's is
essentially designed. : |

Thahk yeu for your very kind treatment of our work’ in this laboratory.
In answer te yeur question as te the multiplication ef the'sperophyte!
there is evidence suggesting that the sperophyte undergees immediate
reductien, and is certainly not capable of prologkd proliferation.
This evidence is based on the fact that the ayeete can be made to occur
in agar (i.e. the mutants are Gown BEE separately, washed and mixed
immediately before plating) so that all the prototroph recombinants
originating from a given fusion will be localized, in the smam same
celony. The homogenfety of such colonies with respect to the segrec ation
ef Lac and V has been studied; only a small page of the colonies aze
net essemtially homnceneous, and contain aifferent segregation types;
the possibilities ef accidental contamination with this frequency are
such that ne definite conclusion cah be drawn; a similar result ed
also be anticipated if there were four viable mx preducts of meiosys
as in spermatogenesis. The question will have to be restudied under
cenditions such that contamination can be minimized, (Conceivably, also
the pretetroph is an exceedingly rare recombination type, so that only
a small fraction ef segregants will be ef this type, and the chances for
2 'sporophytes’ derived frem the same zygote giving pretotrophs small;
this is not likely.) . .

Just noticed: I think your acceunt of Beivin's work is slighatly in
errer; according to his reprints and cerrespondence, transformation
te C2-S is net accemplished; either C1-R pr C2-R can be transformed te
Ci-S, however. His work is summarized in an additional reference:
Acta Helvetica Chim, 29:1338 1946. The claim that an enzymatic change is
transformed is subject te qualification, as he has written subsequently

that the same change may eccur Spontaneously.
I hope that you will not take amiss these unsolicited comments; please
rugard them as ‘peripatetic ruminations. !

You mentioned the development of a method for enumeratag the
‘lethal gene loci' in phage: permit me to guess what it is: to study
the ratio of plaques te the number of irradiated phage particles
absorbed per bacterium in the 12-14 system. KXMWOXMEXEMAK The frequency
of 'recombinationg types! in mixed infected bacteria being known or
100%, each two-particle infection will yield a plaque unless there is
a lethal mutagxion common to beth of them. Or, alternatively ta measure
the dose of radiation that is necessary te block out all the loci of
one phage tyne as determined by the failure to recover recombinants
when infections mixed with another marked type are studied.

Thank you for the cartocn which yeu enclosed with the manuscript.
It would be too simple to counter all criticism, however, with this
type cf defense; I have not heard too much of it at any rate, and mostly
it is justified.

I'll send the manuscript back shortly, after I've had a change to
study it more carefully.

Yours sincerely,

Joshua 7ederberg.

aa