_ YALE UNIVERSITY
_ OSBORN BOTANICAL LABORATORY
_ NEW HAVEN, CONNECTICUT

Se EK ‘March 3, 1947.
Dear Dr. Mather: ee

eo 2 ee tn pa gy te oa aonk 2.
This should perhaps have been written as a postscript: te
my letter of the Sth ult., since certain, remarks. therein, will
have to be revised. ...0 #78 os. 6 fe 2h oe a Deve ef Leo
~ fhe results. of the cress B-P--X T-1-B,-LacrV* de, not support - .
the map order B, 3; BM: Lac: V:P:2L. which seetied: te: be suggested by.
the comparisons ef the crosses BM X TLB,V* and BM X TPV™. The

   

data are: .
BHP-T+L+B,+Lactv® X B+P+2-L-B,-Lac-V". (The P here does not
Lae’ recombine with P of TP).
, -R +R -5 +5
Protptrophs: 25 0 27 4
Be + Bt 62 2 41 1
(these are types from B, plates.
B,~ ca 10X B+ )
l 1
Be + Bt T4 LO 69 12.

 

‘(frem B plates— but B--not more
than about 1/2 as frequent

as B+ ) mt nd Bt *

There are two respeets in which these data do net fit the
map erder assumed above:

1. #S is mich too frequent / uly ppb? ppede thay

2..B- is much too infrequent. tel,

fatal to the hypothesis of linear

linkage, since the recombination value for V in the cress BM X TP.”
may depend on the map order BM--P-V-T, on the assumption that the °
recombinations between V and T...P depend en two crossovers. fhe |
numerical data are not inconsistent with the theory (even in the .-
absence of interference) particularly if T is to some extent to the

However, these data are not

right of L. |
: AB.
This gives a@ map: : + ge FF LS eo
co _ 0 ff 15 10 iS FS OR Ath fa PVET.
B ce eBMc cece LAC cect sPesesceeeVocecs cba? - fy > & .
: _$ t ob , Ab wm Re

 

A .
In this cress then, the Kmauam crossevers required are: wee

(A.D is -S; A.E is -R ; D=E)

  

Pretotrophs:A.DE
or A.B.C.DE A.B.G.D is +S; A.B.C.E is 4R
By" : DE or B.C.DE Be F.DE w FA, €.pe A8C.DE
~o ALP A =f ory

id nae
tote oe The

 
It would not appear to be too worthwhile to attempt to maxe too
precise estimations of absolute distances from such data es these
until we learn more avout the possibilities of intereferences, which
could certainly throw one wide of the marx.

Another approach to the problem of linearity has suggested itself,
and J am working on it now.

It should be possible to find ‘cross-over suppressor! stocks after
intense irradiation. These would be presumablg based on the occur-
rence of inversions. It should be a relatively simple matter to detect
inversions covering region A of the map, since crosses involving

such sttécks could yieid By-- but no proti®phs. (using, eg BM X TLB, - )

If such can be fuund, one’ can also detetmine whelifer, if gual,

of the regions £,C,D,or E are involved; a group of such stocks could
provide evidence as to the linear plotting of regions of sup -ressed
interchange and tnerewy indicate the linearity the gene arrangéwent.
first, the inversions will have to be deceobed.7.

What news on the new addition of your ‘Statistical Analysis?!

Sincerely,

 

a