Thanks, Della. Just added a couple more after I found in this paper:
https://www.researchgate.net/p...
The paper is also well worth a read, since it shows that using triploids in hybridising is not as simple as crossing with a tetraploid. After reading it and other documents (plus the invaluable posts here) on polyploidy, I reached the following conclusions:
2n X 2n = 2n (3n on very rare occasions)
2n X 3n = Typically nothing (since triploids are almost always pollen sterile)
2n X 4n = Mainly 3n (sometimes small numbers of 2n and 4n)
3n X 2n = Typically nothing unless pollen parent matches pod parent’s dominant chromosomes (see below), in which case small numbers of 2n and occasionally 3n, 4n or aneuploids.
3n X 3n = Typically nothing
3n X 4n = Small to moderate numbers of 3n and 4n
4n X 2n = Mainly 3n (sometimes small numbers of 2n and 4n)
4n X 3n = Typically nothing
4n X 4n = Moderate numbers of 4n
For triploids (the key in the paper is Table 3):
L/A (3n) X A, L/A (4n) or O/A (4n) -> should produce some seed
L/A (3n) X L -> won't work
L/O (3n) X L, L/T (4n) or L/A (4n) -> should produce some seed
L/O (3n) X O -> won't work
O/T (3n) X O, O/T (4n) or O/A (4n) -> should produce some seed
O/T (3n) X T -> won't work
O/A (3n) X A, O/A (4n) or L/A (4n) -> should produce some seed
O/A (3n) X O -> won't work
For those working with triploids, the paper is well worth the read.