THE ROCKEFELLER INSTITUTE
FOR MEDICAL RESEARCH

66TH STREET AND YORK AVENUE
NEW YORK 2i. N. Y.

Hovember 21, 1952

Dear Josh:

As a consequence of the postulated segregatational lag in transduction the
transinductions should not be clonal, and I thought it of some import to study
this in another way, and if it were so when and how long should also be answered,
The preliminary experiments verify the conclusion amt, They were set up as follows:
At time zero a number of EiB gal and NSA plates were spread with a known number of
bacteria and phage ( gale, S” FA). The rest is obvious, respreading some plates
after intervals of incubation with streptomycin as necessary and washing others for

A
37 minutes inclding the initial lag. In order for any differences in growth rate

. gave
growth ratese fixe and one-half generations mex observed amt a generation time of

tux between phage infected and non-infected bacteria to be equilibrated out § the initial
multiplicity was about two. At about five’senerations the number of transinductions
doubles while prior to that it is pretty level considering the large variance to be
expected. With streptomycin a surprising thing happens, the number of zero points

is almost as high as those after one generation. This iausxtmkig means that there must be

a long delay in SM action which is both advantageous and disadvantageous. We can't study
the phenomic lag but then again phenomic lag doesn't much enter into our scoring of

the repositioned transinduc&Sions.. A good unrelated phage resistance marker would help.
Do you think that replica plating is efficient enough to do further work with that
procedure, or is it only useful to say that something is clonal and not vice versa?

Of course reconstruction experiaments must be done to show that growth rates are
equivalent and the lowera&ng of the growth rate of the cell in the process of transductior
ruled out by use of sufficient number of characters.

Best regards Sincerely,

Pale