On the other hand, based on our first
rainbow estimate of latente linkage, *fl1* seemed to be one possible
marker for it. By one cross and four backcrosses we obtained the floury-1
versions of 2 Tuxpan lines, 2 Cateto lines, 1 ETO flint line and 1 San
Luis Potosi line, derived from the same source as in MNL 61:27-29. These
floury versions were crossed to a latente source and used as recurrent
parents. In this material we husked the tips of the ears. In the SLP backcrosses
there was a segregation for color, so color acted as dominant. It appeared
in all its progenies, which were plants selected by means of a Steady State
Porometer Licor-1600 apparatus in the previous year. The plants were under
severe atmospheric stress by heat and very low relative humidity. The data
are presented in Table
1 including also kernel row numbers and the analysis. An unspecified
linkage *Krn Fl1* is readily visible comparing the factorials *Krn*
x *C* with *Fl* x *C* (*C* = uncolored) but the data
can not be explained by only one pair of alleles. In *Fl C* factorial
one sees many more colored (39%) in the *fl1* line than in the *Fl1*
(19%) and again the variation can not be explained by only one pair of
alleles. The models which explain the data are presented in Table 1, giving
nearly perfect fit for *Krn Fl1* and a reasonable one for *Fl1 C*.
The sum of the probabilities in each 2x2 factorial is one. To correct it
to the population size n=87, we can multiply them by 87 or correct the
calculated values in the end. In the case of *Fl Krn* for example,
if *Fl1 Krn2* (see MNL 61:27-29) linkage is q, and *Fl Krn12*
linkage is p, the joint probability of *Fl1* bringing both together
is (l-p)(l-q), and so on. Taking the logarithm likelihood expression including
a, b, c and d, differentiating in relation to p and q, and multiplying
by minus one to simplify we arrive at

and p=30.0__+__10 and q=32.2__+__10,
the errors being calculated by Fisher's indirect method (see MNL 62:37-38),
with survival equal to 1 for all entries, a very rough estimation, we guess.
The inclusion of allometric coefficients did not improve the results.

For *Fl1* and colour we did *Fl1
ltp* recombination value (*ltp* = latente p color) equal q and
*Fl1 p2* (*p2* because it is similar to existent *p1*) as
p, and *p2 fl1* is p=30.0__+__10 and q=33.3__+__11.

In maize x teosinte studies kernel row numbers should be counted and this model applied.

The high frequency, at least in our reported work of these types of situations, suggests that in QTL x RFLP programmes these models should be included in the computer in order not to lose important results.

The systematic occurrence of flavone
glucoside factors linked to *Krn* suggests that these blocks may be
the triggers for a system of horizontal resistance to pests (see MNL 60:33-34).

Return
to the MNL 64 On-Line Index

Return to
the Maize Newsletter Index

Return to the
Maize Genome Database Page