Two dipoles on one feed line?
 

Ditto the three previous responses.

My two cents worth is that the amount of interaction depends on how near
to one another the different dipole legs are.

If you run one dipole north/south, and the other east/west (i.e., 90
degree physical separation between the dipoles) there will be very
little interaction.

If the legs parallel one another, seperated with spacers of some sort(or
just hanging, one below the other), there will be more interaction.

If the legs are parallel, and separated by about 6 inches, there will be
more interaction than if the wires are parallel and separated by about
12 inches. If the wires are to parallel, putting a spacer close to the
feedpoint on each leg helps. I've seen articles on twinlead dipoles,
but the SWR bandwidth on them must be pretty sharp.

You can multiband the bands that are not harmonics pretty easily. For
example, 40 and 20 have more interaction than would 40 and 30, or 20 and 17.

I had a dipole for 80/40/20 up for a year or so. The SWR bandwidth was
sharp on all bands, but a tuner could handle it easily. I now have
seperate 80 and 40 meter antennas, but use multiband dipoles for 20/17
and 15/10.

QST had a wire yagi article about 2 years back based on three dipoles
for 20/15/10. I built one for just 15/10, works great, and I can pull
the dipole off quick if I need to.

Experiment away, but its easier to tune a multiband dipole for SWR in
its final location. Lots of FB antenna fun.

73
Keith, W4KAZ