February 5, 1956
Dear Nertens

Thanks very much fer sending yeur precis. I wish I ceuld be mere help
at this stage, but it is written se cempressedly I weuld almest have tn
write it out in detail_in erder te discuss it _peint-by-peint, and I am sure
you want te de this task yeurself. When it gets tn that stage, I will be very
happy te give it the teetghormb treatment. Hewever, I think I ceuld get the
main drift ef your argument here, theugh yeu overestimate your reader's ability
te guess at the gaps in yeur description, tables, etc. The frllewing are then
seme preliminary suggestions; style and centent are intermixed,

Terminelegy of types., There is arme possible canfusien here between gene-
types, which govern several alternative potentialities, and reactiens. A table
(see appendix) weuld be the easiest way wut, and may help tn ayeid centradictiens
like "2" where immune cells praduce phage. I weuld alse defer the generalizatien
that "all strains... G2 ceatain heaclegeus genetic material". Yeu ceuld say
that Lweff had defined praphage as the sr&sr, and yeu were breadening the definitéen
te include that genetic material of a iymmag lysegenic or a sensitive bacterium
which oan be shewn te be brmelegeus with the geneme of a temperate phage. If yeu
are agreeable, I weuld like te cens&lt with yeu en the detakls ef this generalized
definition, in hepes ef having ene that wenld be cempatible with eur c-nceptien
of pre-lambda as equivalent to an exegencte in cur system. (We weuld need seme-
thing aleng these lines, that the phage geneme is that which is regularly included
in the infective parficle; the prophage is the correspending material which is
propagated in a lysogenic bacterium..., but I haven't worked this threugh in
detail), The logical alternative would be factitious, to accept Lwoff's defi-~
nitien (i.e. as the difference between a lysegenic and a sensitive bacterium,
which is perhaps one lecus of the prophage in the breader sense) and invent a
new term (latent gonephage]) for the breeder meaning. Please note:
abort is a v.i. according te my dicticnary.

Neither Larry nor I could see why immunity had to be subdivided inte heme
and heterc-loggeus, What is your definition, actually? If you have A(Plv2), does
this show homelegous or heterologeus immunity in re Plv9"? As they are written,
the definitions are tec complex: they lump the feet. ttern with the response
te infection so that I could not readily unta tY ngle them, and Ihad
better wait to see this pelished befere discussing further.

The paragyzph "The major point for consideration..."follows logically from
the definition of prophage. A°gain and throughout, it would be better if you could
spell cut your experimental results, and resebve as much of the discssuinn or the
end as pessible.

I don't agree with your definition of host—-induced-modification, if you
mean it as a generalization. How would it fit Luria's T4 case, for example? You
have every right to propose the hypethesia for the immediate examples, In general,
don't be in a hurry to pack everything into one sentence, and make sure you have
expressed each idea. I don't think you will have too many literary prohlems if
you do this, I am sure a large audience can understand the principles and the
experiments, but if I am already having so much trouble follewing yeu, you can
see that pti will have make sure that you are making complete statements of

your facts and ideas.

~~On Experimental: Should you document the dtrains by published references?
What does "P22 as given" mean, grown on A? TI assume you will be giving your phage
cross data in this paper, or at least prior to ita publication,

What does assorts at random mean for the output of Pl on B (p.3)? Does thie mean
half the cutput, or half the mutant (non-P1) output?

p.3 Strain B.... This 4a the heart of the stéry, and I am delighted it has worked
Cut so well, However, I think you have te show that C differge from A, e.g.» in

noy giving mutants under UV conditions, or at least net the typical pattern. Aren't
you going to mention this point in regard to A? Otherwise, the reader will wonder
whether A and Care not equivalent. As matters stand, one could argue that C has
merely lost a selective (compatibility) facter that brings out the mtants. I thought
you had delye@agentzed a B(Pliat) 3 if so you should get that particular mutant back,

Actually, I wonder if at least of the story wouldn't be cleaner if you used
strain A rather than B, and uv/uy to elicit recombination, rather than an exen vaguer
compatibility system. Of course, if you can tie everything together, all the better.

P-4 "D" had me puszled for a long time. I then realised you must mean that D was a
de~;ysegenised B(P5), net B(P1) as stated.

If you can degument your characterization of the phages as AA', etc., you should hee
a very neat story, Why is there an infinte series? There are nine theoretical pes—
sibilities, considering Just One marker per chromosome, If negessary, I would synthesize
bacteria that had just one mutant marker On each chromosome, and could threugh the
cycles with them, though I doubt this is necessary, You already haves phage that
corresponds to:

cc, cB! BCt 5B! Ac?

Perhaps a simpler approach still would be to dely@sgginize PLT22, and deal Only with it,
C and B. There is every expectation (if I understend correctly) that C and A will

have the same general medificational behavior, and differ only in a few marker alleles
that you are not concerned about here anyhow. This stock shonld, of cOurse, resemble
C. (This is a confusing point of Symbolology; 1+ might be better to reserve "C" for
P22, and call the delyeegeniszed B(P2) By or the like, (I'm not too happy about this
expedient either, )

This is about all I can do at this Stage. The two principal suggestions mkax are
to verify your C as CC! by uv-crosses (with the phages); to simplify the syetan by
using a delysogenized (I almest typed deloused) P22 in place of A, From this pair,
you should generate only 4 principle combinations, and a complete analysis should
show what happens in every combination of Phage and bacteria, When you get this rounded
Cut some more, I will be happy tof see it if you want to take the time.

T've been busy myself with Hfr types, having work&d out methods te pick out¢g mmy

more, It all started when I finally realised that Hayes' Hfr must be different in
segregation behavior from Cavalli's, and sure enough it as, I was x®umsm misled be
fore by the fact that Hayes’ stock had largely reverted to F+ (as Cavalli's used to),
While I was in the middle of this, and had been iselating some new cne's from UV'd
W-6, I got a note from Jaceb that they've been doing much the same thing, indeed believe
that #3 CROSSING of F+ is due to Hfr mutants, For a number of reasOns, I don't believe
this is true, theugh Hfr mutants might explain some excepticnally fertile clones,
What cOncerns me more is the reason behind the variability in segregation
pattern, which may be as simple as chromosome rearrangement. One reason I doubt
that Hfr mutants play a predominant role in F+ fertility is tik constancy of
elimination behavior of Fi (as tested qualitatively in diploids) as cOmpared
to the likely variability of many of the new Hfr's (which, however, still have
to be tested in diploids). The original F+ and HETo avatr are rather alike in
this respect, which is why the issue had not come up beffre here, but at least
we can nOw understand why FrancOis et al and we could not agree On details
of syngamic elimination, etc, The whole story is Obvicusly coming te a ravld
boil, now that the fancies of F+vectors have been dispesed cf,

Sincerely & best regarda,