nice.

Thanks but what is teh answer  ? My original spiel seems to have been chopped  here it is in full .. All almost obvious to most .  

                                   Galactic rotation curves  ??

Given the importance and longelevity  of  the galactic rotation curve problem  it is probable that most, if not all,  prosaic explanations, apart from the existence of dark   matter,  have been  considered and rejected. This educated layman has limited access to ,and knowledge of,  the literature and thus the suggestion, given here ,might well   have been dead for  decades.

On this limited knowledge it appears that most, if not all,   rotation curves are actually measuring the tangential speeds of interstellar gas rather than the star field itself.

Question 1 . Is it possible to measure the small   ( 1: 10^5 ) Doppler shifts  directly from Fraunhofer lines in stellar spectra  Has this been done ? 

Again on limited knowledge one can imagine that these measurements require the relatively narrow emission lines from relatively cold but ionised gas, far away from stellar cores  . Thus the H alpha and 21cm line  data   dominate the literature.

Question 2 Are rotation curves  measured using these two lines exclusively and  out of necessity, because direct stellar observations are technically too difficult/ impossible? 

Assuming  all the rotation data is based on gas then there is perhaps  a fairly obvious explanation of the puzzle; there is no reason to think ionised gas  has to have the same velocity field as the stars, ,particularly in the outer reaches of galaxies. Ionised gas clouds, bombarded by radiation , create their own weak  magnetic fields and thus the dynamics of the gas cloud is always the result of both gravitational and EM( MHD) forces. Close to the  dense hub, gravity clearly dominates,  but with increasing radius the gravitational field  strength  falls off  roughly as1/r^2  . The self induced EM forces, on the other hand ,need not have any strong dependence on radial position  . Thus in the outer regions, gas dynamics could increasingly deviate  from stellar dynamics through the combination of gravitational and EM forces  of comparable magnitudes.

 Of course it remains to  be seen if MHD can spirit up  the required speed deficits of  ~  100 K m/s  in the outer reaches of a galaxy. Given all the gas is initially launched at high speeds  into interstellar space the required speed might not be impossible    

I do not have answers to all your questions, but some of the references in following article can answer some of your questions and those references can also lead you to other references. There is plenty of material on arxiv.com

https://arxiv.org/abs/0805.2233

Thanks Thierry  I know your  views  and am intrigued .   My question (s) is related  but far less fundamental  and actually almost sociological eg   has anybody actually measured stellar speeds directly. H alpha  from gas is used both because it is a nice sharp line  AND allows measurements way out where there few stars !!. I cannot imagine that gas speeds  are identical to stellar speeds. Have we all simply assumed that gas is  doing the same as stars   because the only force acting is gravity . That cannot be the case in a plasma  at around 6000K . If somebody can tell me that direct doppler studies of the stars   themselves has been done and the rotation curve is the same  I have no case . I just cannot find that evidence .    .    

I cannot imagine that gas speeds  are identical to stellar speeds. Have we all simply assumed that gas is  doing the same as stars   because the only force acting is gravity .

Chris, you don't need to imagine or assume anything. Numerical simulations confirm that gas clouds and stars orbiting at a given radius from the centre move at similar velocity. 

That cannot be the case in a plasma at around 6000K .

Why not? If you heat a cold molecular cloud up to a 6000K plasma, you have increased its internal energy but you haven't shed its orbital angular momentum.

So far  nobody has been able to tell me where to find   direct Doppler  measurements  of star speeds. Rather I am asked  to trust  simulations and  the influence of  entrainment via viscosity ( whose coefficient in an  interstellar  plasma space is likely   totally unknown.??)  in  order to make Star speed "similar" to gas speed . To be honest  it looks more like a  house of cards now than last week . Clearly gas dynamics is significantly different from that of the stars  , witness the relatively high gas temperature way out on the edge of a galaxy. and the observed  tangled morphology of gas clouds . Just one direct observation would shut me up .. please!!

only so many objects you can observe in a field at once, you need to worry about fiber crowding, and when observing, you usually need to plug in all the fibers into pre-drilled holes by hand.

Integral Field Unit Spectrographs are gradually coming onstream, and they resolve these issues.

The following summary paper by Beckman et al. illustrates observational agreements and disagreements between spiral galaxy rotation curves derived from ionized gas and stars (Figures 4-8), with helpful explanations of the observations, and physical reasons for the differences.  Much of the kinematic differences are present in spirals having larger bulges compared to those with much smaller ones:

https://www.slac.stanford.edu/econf/C0307073/papers/JB.pdf

Many Many thanks .The paper extended my  "enlightenment "  enormously . That is to say  yes Galactic dynamics is really very complicated and the ( don't laugh  please)  the realisation that stars form from pre-existing gas and thus take on the  angular momentum of that gas , at least initially,.rather than  a collection of stars eventually ejecting enough  particles  to subsequently form gas clouds. The gas - stellar system  is strongly  linked from the very beginning.    

Dear Chris

“Again on limited knowledge one can” also suspect that the idea of a one-sided Universal Gravitational attractive force or “Free Fall”, does not seem to tally up. No wonder this idea leads to superluminal speed, infinities, singularities giving rise to dark/black cosmic monsters and so on! This also begs the question, if a “super massive black hole” exists at the heart of the galaxies, why does not the whole galaxy turn into a "super^n massive black hole"?  

Newton needed a Third Law (To every action, there is an equal and opposite reaction) to make his mechanics work orderly in terrestrial Nature and get things running without any catastrophe. We have no singularities in terrestrial Nature except in mathematics (when you divide anything by zero) or in imagination (the reason we can make up a black hole in our mind). The only REAL singularity we can have is when a President of the most powerful country in the world goes singular! But in no other case this is true!

Newton forgot about his Third Law when he gave a gravitational theory of the cosmic bodies. He surrendered his responsibilities of doing good physics in the realm of the cosmos; in the hands of God, starting from the “first impulse” onward, to run things as He wished! For Albert Einstein “matter” and “motion” and hence “force/momentum” does not exist; only the abstract “spacetime” manifold, with tangible physical/mechanical attributes exists; but he does eventually bring in “matter and “motion” (not “force/momentum”) through the backdoor and like Newton has to depend on the good Lord to run things in the heavens for him.

How about conceiving (something) like a Newtonian Third Law in the realm of the cosmos, so that we can explain and understand the phenomenology of the cosmos (including the rotational curves and many more) in a similar and most certainly less definitive way that we do on mother earth, but hopefully without the cosmic monsters and the singularities!

Please see the reference below and many more from the same author about an infinite and eternal (dialectical) universe:

https://www.researchgate.net/publication/317040165_THE_CONCEPTUAL_DEFECT_OF_THE_LAW_OF_UNIVERSAL_GRAVITATION_OR_%27FREE_FALL%27_A_DIALECTICAL_REASSESSMENT_OF_KEPLER%27S_LAWS

Article THE CONCEPTUAL DEFECT OF THE LAW OF UNIVERSAL GRAVITATION OR...

Yes, this is the case.  There are many reasons for this.  First and foremost, making accurate radial velocity determinations to stars in external galaxies is not practical.  The stars (except in a few nearby galaxies) cannot be resolved individually and even if they could be, their light would be far too dim to allow for the high-resolution spectroscopy needed to determine an accurate radial velocity.  The problem of observing a proper motion is even worse.  Taking in the light of many stars simultaneously cannot yield "average" radial velocities because their spectral lines blend together and therefore do not yield any features that can be analyzed for accurate Doppler shifts.  Individual stars may also have peculiar motions of their own and may not share the general rotation curve.  

Measurements in the interstellar gas do not share such difficulties.  Peculiar motions in the gas phase, though they do exist (supernova remnants, etc), tend to damp out over time as gas clouds interact (collide) with one another.  Stars, once formed, interact gravitationally (without damping), but actual collisions or even close encounters between them are very rare.  Therefore, the gas motion is anyway more likely to reflect the average distribution of galactic mass, which is the astronomical use made of galactic rotation curves.  Furthermore, the spectral lines in the gas offer very sharp spectral features that yield accurate radial velocities.  

Often, the 21-cm line of atomic hydrogen is used to determine rotation curves, and this line offers the further advantage that it can be followed out to very large distances from the centers of galaxies: large distances where star formation is slowed, and light from stars may not even be detected (beyond the "visible" galaxy).  Orbital motions in galaxies are not very sensitive to any material lying at larger galacto-centric distances, and thus a determination of the entire galactic mass requires radial velocity measurements in the gas farthest from the central galaxy.  

Thanks  . I am sure you understand  my  query. If we  have not not really measured star  velocities directly  we have to assume  

1. since the stars started as gas  they take on similar angular momenta  to their parent  

fine

2. damping= friction irons out any gas - star differences. 

3. Assume simulations are equivalent  to  honest to goodness  direct measurement

 But for the past  N million years the dynamics  of gas & stars  has been different . Stars dominated by gravity  everywhere but gas  with  extra ? radiation pressure  and internally generated MHD  forces .Out in the far reaches  can we be sure these extra forces lead to negligible differences in dynamics . Surely No until  we actually measure star velocities directly .

With this in mind the jump to  assuming the existence of dark matter  strikes me as Ptolemaic  or at best premature .... and we don't  seem able to find the stuff in the solar system     

I think that you misunderstand stellar and gas motions, as they might be used to determine galactic mass distributions.  There is nothing sacred about stellar motions that make them better than gas to begin with.  Both the gas and the stars have tangential velocities that are consistent with the large-scale distribution of gravitational mass, to the best of our knowledge.  The interstellar gas in spiral galaxies would simply "fall" toward the central regions if left without adequate "rotational" velocity, and there is no good reason to assume that it could have gotten in such a condition to begin with..

There is therefore no reason to make your assumption #1 when analyzing gas-derived rotation curves.  It does so happen that stellar systems, when formed, probably take on similar velocities to their parent gas clouds, but this is a needless assumption.    Also, individual stars in a young system might take on peculiar velocities from their formation, such as having formed in a binary, or even a multiple-star system, or behind an outgoing shock from a recent supernova remnant.   

Also your assumption #2 is needless because it does not matter about the stars being the same as the gas, if measuring the gas is sufficient to begin with.  

As to your statement #3, the observed Doppler shifts of the 21 cm atomic H line are "honest-to-goodness" measurements, and even stellar measurements depend of spectral line shifts (except for very close stars in our Galaxy where proper motions can be measured).  As to these "other forces" that you mention (radiation pressure, MHD, supernovae), these are very small-scale relative to the sizes of spiral galaxies, and they generate no significant net tangential force.  In the outer regions of spiral galaxies, where the rotation curves yield their most powerful results, there are very few stars (they are dark out there), meaning that there are very few, if any, forces perturbing the gas dynamics.

In short, you focus too much on stellar velocities as "truly" reflecting gravitational mass.  The gas may be better to begin with when determining rotation curves.