UNIVERSITY OF CAMBRIDGE es
ue PRE; 2SSOR R. A. FISHER, Sc.D., F.R.S. DEPARTMENT OF GENETICS o WHITTINGEHAME LODGE ws

MISS M.F.E SPEYER, B.Se. . : foe Lo Dos 44 STOREY’S WAY
Research Assistant and Secretary ° nee ney Cee ey ge ve ME a7 7

 

Dear Lederberg, 7 . : lye.
Thank you for your letter, 3 and for strain 123 de 3 which ‘arrived & a “ong

 

ago 3 4 too was surprised to find no effect, ‘in ‘crosses. T have made no ‘great. progress

aPle et I CMEW CGP kb. Se ae

  

with 185; however, growth requirenents have been found to ve methionine and jysiney , -
: w relay

but with a lag ‘of some a8 Yaoes" this cee

 

on which 143 grows ax

correspond with your findings ? ee : fee —_ an

 

As I wrote you in my last letter, I had some | troubles: with recombina- ~

Ade Aner oe

tion which did not occur as usual, in the last nopths of 1949. I wamit thought qt had a
found a reason for that; but have no more been able to reproduce ‘the failure of recom-

bination,once it Started reappearing again. Ehis meant | a “considerable | waste of ‘time 3s

” ¢ : iv
: . yok

ft an glad it is Howe over now.

me ra om ~

Summarizing the results. of a work some have been of Little ‘encourage-

= et at ° _

os

ment, some others more interesting: here are some details that might interest you 2

1. Hfr_ . Results of crosses Hfr : x Her were ‘surprising : no Efr, ‘in the PREERI !

.

@ am repeating them now. The interpretation of a Dauermodifkation : is always ‘trying.

26 Mating. Hfr proved disappointing under this point of ‘view § nothing ‘definite

has resulted. Some syntophic growt.,which seems ‘unavoidable in ‘mixtures, makes the

observation more difficult, but even so it should be possivlé to see ‘soitething. This

‘

failure may be of some interest in relations to your new hhypothebis of small male“

gametes, which I take from Davis's paper on BMG 25 and in ‘this connection Tr ghaald ‘aike

aen FM yta Sve se
to mote two facts,none of which has- mach Weight per ‘se, “but they “may give rise ‘to

Or Let a
further developments.Cin some Hfr crosses | with few cells of one strain, ‘oie ‘Bees “more

pe

meconbenants,sin Some exveriments, than colonies on controls” with complete of the rarer

strain the other fact is that with ‘microscopical observation in phase contrast, 1500 x.

. Bes ver cds are
ae pean 6

one definitely sees with Hfr crosses some very saall ‘motile elenents,yh : ob Es youd “dee
o..@iving importance it

scribe like $free flagella". I entirely agree ‘that/wak/these facts may “seem “foolish

Be: nye Faye ote eT erry pete ee

at “this stage. The way ‘is probably ‘repetition of “Davis's” experiment with larger filters

and a more e- “ficient strain like Hfr I am " Seneciee ‘trying’ something in “this. ‘line.

whee a ee ae eT be Ore Pe ee WIth hones oot wlod cP Fah

ASE

Phe Te (ae) font fs 7 STO - ddlute & ied aT wit . be

tor Cors | while ergs ety ol “fee wali A “peel,”

ow pe r ~* fie ae a bay heer oe
~ oo. ; ee Te, a ot none fae ares oo

 

 

 

 
3. Maps - I am looking forward to your paper on segregation announced on Genetics;
I feel I am perhaps the onkty one who still believes in linearbty,but I had some result:

which pointed to a possible way out of the mess. I am inclined to think thatv&he data

collected so far (I have seen also Newoombe's data on ST) can be explained on the hypo-
' thesis of linearity only if either a major chromosome mutation has occurred in the
building of Bane or T-L=B, = -,ywhich is not unlikely with use of X-rays » or selection
of prototrovhs introduces a biasxof ‘some sort - not revealed showever,from reciprocal
crosses} otherwise, lineatrty seems tntenable. The first-hint for a chromosome mutatior
came from the outrrossefof W 677 and W 705 with WV 836 - In the two cases, the rela-
tionships between Gal and Lac are reversed using W 677, Gal is unlinked with Lac,
using VW 70 it is tery closely linked. The markers of W 836 are closely linked between
themselves, slightly on the right of M. Gal of 677 and 70§ seem allelic(and not alleli:
to Gal of W 583,which is linked with Lac on the left of it). The aasiest interpresatio:
seems that thare is an inversion bets with break points left of M and left of Lac, the
orders being : VW 677 : By Gal Hl Lac vy LT ’ and W 70§: ‘By M Gel Tac Va: “the: normal
order being the last | one. Many other markers are linked with Gal : Kyl,Mal of W 677
(not allelic to these of W 708, unfortunately) Ara and S,and should all be within the
inversion. The results witl be : a) in the cross Bi x W 677, or W 836 x W 6TTs markers
within the inversion will recombine only with double c.o. (oaa crossovers being normal
ly inviable) giving rise to the observed mess of combi netions 3 bd) there will be an
apparent ,and partly possibzly real keiveen negative. interference between Bol ana
MeLac,as is, in fact,found . Akso other results follow. Bossibly vart of the difficult
of"diploids"may be due to random segregation of acentrics ? The agreement of data
with theory is only qualitative,so far; it is gifficult to collect enough data, and

it is difficult to test such hypothesis only on the basis of agreement with etpectatio
in view of ignorance on interference . Tr am trying other ways,now, and should I cone
to more yeikixhtexcans final conclusions about it, I should like ‘perhaps to ask you

the earlier strains T. etc., to $eace back the history of the mutation. But it is
definitely tooezrly now. At present, I should need instead a replacement of W 826,
lost in an accident,and Tg 3 I should also dike to have an original K-12; I should

very much appreciate a sending of them,and perhaps also strain Y+10,as I am using
as TLB,- a J 909 reverted fog Gal.
4.Antigens. Differences of antigenic type between K-12,971113, 123are too small
to be of value. However,two and perhapd three strains yantigenically different, and

fert interfertile have been recently found,and serological analysis is in progress 3;

I am developing convenient markers and hope to be able to ship them to you soon. Such

strains show also some degree of interfebtility with the three mentioned above.
Thank you for the very interesting details of your "divloid"work .
Serndarrg wsrcder” aparnd -ovWr
T am onelebing ofZprints of the letter to Fature; unfortunately Ix did not correct
reference to
the proofs, and the alterations you sugsested about/Professor Tahun, which was insufri-

cient, could not be done. I apologise for this . I am also Padgsading offprint-ef the
abstract of ig Stockolm paper, taken from the Proceedings. This paper was quoted by
you in your review on Bacterial Variation ; unfortunately, in the Abstracts,where you
must have taken it from,only my name was given ,and not that of my coworker Visconti.
This mistake was corrected in the Proceedings. I am adding this, in case it happened

to you to quote again the same paper.

A Cambridge statistician, N.”.J.Bailey, has produced sone nice methods to deal

with selection of ptototrophs,estimation of map distances, viabilities etc. He believes

that some ofahis methods may be identical to those you have employed for the analysis

of the data of your 1947 paper on Genetics,and would be grateful if he could know more

of those methods. Is it possible to get,from Yale University library, a copy of your

disserétion ?
Your sincerely

Peng Gal -