Stretching space

[DrGeoffG]

REDUCTIO AD ABSURDUM

You wrote in this discussion: "Although the galaxies themselves do not
expand (because the
gravitational attraction of their mass prevents space from
expanding),
space does stretch between one galaxy and the next. "

This does not seem reasonable to me. Take for example a spiral galaxy.
Its outer regions are spiral arms, getting less and less dense away
from the centre. And then again, there is the cloud of globular
clusters. Where does the galaxy (non-expanding) bit end and the space
(expanding) bit start? After all, its gravitational field has no bound
- it's field obeys the inverse square law in Newtonian space, which
means (I understand) that in General Relativity the surrounding space
is distorted ('curved') - however little - no matter how far away one
goes. So, any gas molecule in inter-galactic space belongs (in a
gravitational sense) to one galaxy or another. And although inter-
galactic space is a very good vacuum, it is not completely empty. Thus
there is no hard boundary to a galaxy. So how do you separate the
region of space which expands and that which doesn't?

And what about those close galaxies, where one sees a narrow thread of
matter joining the two? Does this configuration as a whole remain
fixed while the surrounding space expands? What about the space
surrounding the interconnecting thread - can this expand, but not the
thread itself?

To take your balloon analogy, galaxies cannot be represented by
pennies. The force holding a penny together stops at it's boundary -
the gravitational force does not. Only a black hole has an 'event
horizon'.

Convince me!

Geoff