November 4, 1952
Dear Bruce:

Perhaps I should have waited a few hours to post my last letter.
Yours of Oct. 30 arrived yesterday afternoon. I am sopry that my
misaddressed aetter never reached you. Enclosed is my carbon copy;
I have a photocopy of it, I suppose the original will eventually
return.

The notion of "hold-fast" is worth consideration-- we had speculated
before that E. coli lysogenicity involved the synapsis of a transduced
Lp locus along with lambda, i.e., that the induction of lysogenicity
is, in a sense, another transduction. The Vi-phage story sounds something
like phenotypic mixing (Hershey? Novick and Szilard). Anderson does not
seem to be acquainted with this work.

I am giad to hear that you are giving some priority to writing up the
Owork. The f-2 tests are how complete, as expected, the FA from a second-
generation 1 tranemites both b and } to SW-545. I don't think that I will,
after all, bother with this same analysis with another factor, The first
generation seems to be valid on the face of it. Re serological analysis,
the b-agglutinability of SW-5435 is still paradoxical. The cultures showing
it give a large proportion of rough colonies. These, however, give a perfectly
non-specific agglutination, while the broth from which they came still
reacts quite specifically with bh. Perhaps there is some sort of unmasking
of a latent b antigen in the incipient rough O form. FA( 666/588) finally
gave a swarm, but the count is about 30 tracks per dwarm. SW-666 + SW-534
give a nice demonstration of 0 + 0 celia giving a swarm, but strangely enough
this has been 1,2 Gal+ , indicating that the SW-534 was the transducee.
Conceivably there is another phage (from 534 growing on 666), or the initial
outpizt from 666 is remarkably high.

Boyd has sent all his lysogenic typhimurium, but these have not arrived.
Let us keep in tough about these phages. Spicer's phages don't seem to be
doing very mich yet; by all means let us have any group O forms, at your
convenience.

I am especially interested in the possibilities that may come out of
your para A Oform. If it also shows a linked transduction, it should be
also linked to the 543 factor. But it's not likely to be identical, since
the 543 transinductions are usually FA type, much less frequently b. (My
data on the proportions, even for comparable experiments, ate not homogeneous.
There may be some @hteresting environmental controls to be examined.) Have
you studied the allelism of the paraA O with 5437 If the linked transduction
is not a fluke (reversion?) it may be possible to orient the three postulated
factors-- I'd like to look at this umless you're already busy with it.

I'll be happy to send these other traneductions to SW-543, but I don't
really see the necessity of a detailed serological study on all of then.
The O form, the spontaneous and induced b's and i's certainly should be. I
have working arrangement with Edwards for fairly full verification of sero-
types, but if you think it should be done, let me know, Instead, how about
the suspicious b and j forms that have come up. I'll send you the lot, to-
gether with 665 and 666. Before I make up the package, let me know if there's
anything else you're likely to need.

Selection with Chi phage is far more efficient in liquid medium (ca.
1000-10,000 fold). With the addition of a known number of marked, reference
O cells, it may be feasible after all $p lock at the flares more selectively.
Also, selection against H by antiserum looks much more encouraging now that
the unexpected b-agglutinability of the 666 culture has been discounted.

The final pedigree:

LT—-2, x bbb

 

 

: a lets v
f 3
b only x bEL eK . .-
b” | 5 . oon
/ &\ _34 ye
b 3 © > : -
£ A 1 29 I

roy,

The ats conlarvidn the yoo Tears ‘f 2 aipanle be ae. CO}
Sanne hos pectinenk pov BAL. nuded , T he. Byte Slee Wed Cae ent

Lik Alone! a opr Dan