When I've had a Phal keiki, I usually wait until there are roots evident and
then cut the stem 1-2" above and below the keiki. I dust the cuts with
sulfur. The stem makes it easier to handle and anchor the keiki in a new
medium.
I've had some dendrobium keikis produce keikis especially lodigesii where
you can find several generations at any one time. Unless you want new
plants, I'd suggest removing the keikis as they are probably (I'm not
positive of this) sapping strength from the parent plant. I believe you will
find new growth starting on the parent plant shortly afterward. That is, if
the plant is healthy. Some plants produce keikis as a kind of
self-preservation thing if the parent plant is under severe stress for some
reason and soon to go under (root loss, fungus, bacteria, etc). If you want
a massive plant, plant several together in a basket. I have a Den nobile in
a 14" basket. I planted 3-4" plants in it and it is now a massive and
spectacular display piece (about 5-6 years). You don't say how big your
keikis are, but I've removed and potted some as small as 2" if they have
decent roots. They sit for a little while and then produce a new shoot which
grows several times the size of the keiki.
As for epidendrums, I think the same holds true as for the nobiles. If you
want a tall and spreading plant, you can leave the keiki on. For a fuller
plant, plant the keiki back in the parent pot or pot several together in a
new plant for a bigger specimen. I don't ever see a good reason for leaving
a keiki on a plant unless you want to propagate more of that plant. There
may , of course, be some exceptions I'm not thinking of off hand, but I
think that's a good general rule.

"Ted Byers" wrote in message
.. .
I now have three plants with keikis.

One is a phal (P. Sogo Smith), and it has two developing at the top of the
inflorescences (which had been cut just above the highest node that hadn't
flowered). Now, when it comes time to remove it, do I do so by cutting
between the keiki and the inflorescence supporting it, or do I cut the
inflorescence just below the keiki, leaving part of the inflorescence
attached to the baby? And after this, what is likely to happen with the
remaining inflorescences if they are left in place (and would you
recommend
leaving them in place?)?

The second is a nobile type dendrobium. In this case, I am tempted to
leave
the keikis in place, and just let them make the plant larger, assuming
that
doing so will do no harm. Part of the reason for this is that the keikis
on
this plant are the only sign of new growth on it. But Iam curious: Do
keikis produce keikis? I ask because there is, on the larger keiki, a
small
growth that is clearly developing into a new shoot but it does not have
any
roots that are clearly associated with it. If keikis do not produce
keikis
themselves (at least until mature), the only other explanation I can see
is
that this keiki is being born with two shoots. Does that happen often?
All
of my other nobile type dends have new growths but no keikis. I know what
to do with these things to produce flowers, but, in the case of the
specimen
with keikis, what can I do to stimulate it to produce new pseudobulbs
(without putting the keikis at risk)? This, and a second nobile type dend
need to be repotted soon (and I haven't decided whether to use coir or
chc -
they are presently in moss), but I am wary of setting back the new growth
on
the one, and the keikis on the other. Both seem to be quite robust, but I
have no idea what shape their roots are in (I only bought them a few
months
ago).

The third is an epidendrum, and it is putting out what look to be a number
of keikis as well as what looks like new shoots from the higher nodes.
Here
too, I am tempted to leave the keikis in place, or order to have a larger
"specimen". I spoke to a vendor at the RBG orchid show about this, and he
confirmed my suspicion that the lower of these growths are keikis. He
also
said that there is no good reason not to leave them in place. It seems to
me that doing so will create a visual impression of a rapidly increasing
number of stems, all, or at least most, of which in due course will have
clusters of flowers.

Cheers,

Ted