BUT, and here is where it gets interesting
for me , if you code 
pyr,pur,pyr there is a real danger
of dimers from the outside pyr
in first or 3rd positions
and a possible pyr next to it
in either the  37 or 33 position.

Such that
33 = pyr
34,35,36 = pyr, pur, pyr anticodon
37 = pyr

thus you have dimer potential in 33 - 34
or 36 -37 position.

Thus it would be highly unlikely for an anticodon
to safely have pyr, pur, pyr coding -

(or for that matter its codon! on the mRNA)

But it seems the bases in tRNA have evolved
to solve this.
In position 37 there is always a purine.
Thus 36-37 can never be two pyrimidines
and there can never be a dimer between
this outside base and position 1 in the anticodon.

On the other end you have 3rd position wobble
at base 34, plus a guaranteed 2 pyrimidines
at position 33 and 32. (Perhaps this weak end
is why there is wobble)

***
Follow up.
Also note that wobble in position 33,
 seems to allow for
more ways to code with purines - thus
preventing that possible dimer problem
between positions 33 outside the codon
and 34 in the wobble position.

I, Inosinic acid has a purine base and can pair with U,A, or C
G, a purine can pair with U, C
U, a pyrimidine can pair with either purine A or G


Summary
I think it is more and more clear that dimers
were a major selection force on the genetic code.
There are too many coincidences if it isn't.
And that somehow the mostly purine coding
was associated with hydrophilic amino acids
while only that coding completely free of UV
dimers (whatever environmental conditions that
suggests) coded for hydrophobic amino acids.

Comment?
 >>
---
þ RIMEGate(tm)/RGXPost V1.14 at BBSWORLD * Info{at}bbsworld.com

---
 * RIMEGate(tm)V10.2áÿ* RelayNet(tm) NNTP Gateway * MoonDog BBS
 * RgateImp.MoonDog.BBS at 6/21/04 6:10:52 AM