As another commenter has noted, they wouldn't then be galaxies. And the question of whether there are (or how common and at what mass) galaxyless dark matter ov

As another commenter has noted, they wouldn't then be galaxies. And the question of whether there are (or how common and at what mass) galaxyless dark matter overdensities (think of it as a clump, in the literature they are called "halos") is still open. One of the problems with the current best model of the universe (Lambda CDM - a universe that is primarily make up of some cosmological constant driving expansion and cold dark matter whose gravity drives structure formation) is that it predicts that there should be far more of these dark matter overdensities than we see galaxies. This is the "Too big to fail" problem - https://arxiv.org/abs/1103.0007 https://arxiv.org/abs/1103.0007 is a pretty readable description. Basically the model predicts lots of dark matter clumps orbiting the milky way and we don't see enough satellite galaxies to populate them. So there are now two options: 1) Galaxy formation becomes stochastic at lower dark matter halo masses - and so there are lots of these galaxyless overdensities of dark matter. 2) There is something wrong with our model and there might not be these overdensities. 3) The milky way is an outlier. I think the short summary is that, at very low masses (say ~1e6 solar masses) we wouldn't be surprised if not all dark matter halos form galaxies. At high masses (1e12 solar masses, roughly that of the milky way halo) we would be very surprised if there weren't galaxies. But in between we are not sure. Edit: To be clear, these structures would still likely contain baryons (normal matter), they just haven't formed stars (or would have formed only a very few stars).