SepteaPber 24, 1953

Dr. AlLan,3erapbdu
DeparMat of B8&eFiolOgy
Univerrlty of yicrhtgan
AM Arbor, tih.

  R*er rreparate aover, as requested In your letter of the ad,
       \*
Z GAB CSW&@Q$~~$~ lLl2 derivativea, W-6 atd W-U.77 for ala88 experimsnts

      fn, per ths eaalosed sheet.

for other 8uggestiom. I wuld 8tron&.y reaommnd that

you try replim plating (8.8. for the moring of the recombinant8 in

the aross%ng experimnt) and indire& 8ele&ion of a phage-re818txint

tautant (e.g. B/l) a8 P proof of pm-adsptim 88at8tion.
  Aropoulpfngtokttlude avar&amsanalpfs(afterLuriaand

Delbrtlak?) If so, aay I suggest you perfect the experbeat 8nd aammr

Xinshelwood `II obj8ctiom by ass@.q~ 8snple8 in the f lra t serlss, ami

show- a oorrelatlon in a aeaond series twing a hQh and low tube 88

inoaula In muabars that should, and should not, carry ow$r a precsxisting

nuts& That is, a sort of general teat of heritability of the response

to tha phsge.

Beat lutk to your lgilitwy aareer.  It may turn out to be a more

prodwth eacperienoe thqou would have any preesnt reason to hope.


AnaJ4rsis of variance in bacterial mutation: heritability.

(Exwplary figures from table 2, Exp. 15, L&D 1943)

Assay saq~~lea from each of 10 tubes. I3 you find extreme8 such a8 A) 5 Bfi per .05 nil
                             set of

ami Bj 165 B/l per .05 n&, set up a eeaonddseries of tubes with inoctia a8 fo~ons:

       Al:  .ool per tub8
   Bl /t#:  ,001 per tub8
   B2 cd/:  .00005 per tube.

Series Al ani 82 should re?mble the ln%Lal distr             mutant8lnths

iaoculura) while most of the tubes in Bl

a prior mutant ia the inmulufn, CD.
din,g to the actual spread of numbers,           on the scope of the initial

sries.

P.S. So far as

J.L.