In article ,
Claudio Jolowicz wrote:

"Stewart Robert Hinsley" schrieb im Newsbeitrag
...

Pear and apple are both haploids with 17 chromosomes, probably descending
from an ancient polyploid. The structure of their chromosomes seems to be
very similar, as you said that little modification has occurred in it,
caused by duplications, deletions or inversions, during their evolution
from that polyploid, and (I guess) little enough to cross them.

What Stewart said is that apple and pear are both *diploid*, with *haploid
number* of 17 (i.e. diploid number 34).

As another unnatural effect of domestication, some apple cultivars are
triploid, and produce nonfunctional pollen. I've occasionally wondered
whether there are triploids with nonfunctional ova, but good pollen, but
we never hear about them because nobody wants an apple tree that doesn't
produce apples, no matter how good the pollen is. Come to think of it,
people do seem to want fruitless fruit trees as ornamentals. Maybe there'd
be big bucks in a fruitless ornamental crabapple. No squashed rotten
crabs tracked into the house, or people slipping on them and suing you.

The apple and quince belong to another subfamily, Maloideae. Apples
(there's many species) are classified in genus Malus. Quince is
classified in genus Cydonia, which is monospecific, quince being Cydonia
oblonga. There are many other genera in Maloideae, and it seems unlikely
that Cydonia is the most closely related genus to Malus. However, the
phylogeny of Maloideae appears to be confused; many intergeneric hybrids
occur, and genus boundaries are disputed. (I see that the taxonomists
have been chopping Sorbus into pieces.) The pears, genus Pyrus, are
often thought of as the more closely related genus to Malus. I'd guess
at Chamaemeles as the genus most closely related to Cydonia.

Pears are grafted onto quince to dwarf them, but not only are grafts of
European-descended pears onto the Siberian Pyrus ussuriensis not long
lived, but some cultivars of European pears are incompatible grafted
to each other. Is graft compatibility of any value in estimating
relatedness?