The Great Wall, discovered in 1989 by M.J. Geller and J.P. Huchra, is the largest observed structure in the universe. It's a 'wall' or sheet of galaxies 500 million light years long and 200 million light years wide, and only 15 million light years thick, and it separates two gigantic spatial voids containing almost no galaxies at all, only some vast, diffuse clouds of hydrogen. The discovery of enormously large-scale structures such as these have caused many physicists to admit that their understanding of the early universe is, at best, very seriously flawed, because up until their discovery, almost everyone was betting their house on a uniform distribution of galaxies throughout the universe. In fact, the exact opposite has proved to be the case: galaxies, clusters of galaxies, and even superclusters (clusters of clusters) have turned out to be distributed in gigantic, regular patterns throughout space.
In "Cosmic Voids and Great Walls", written in 1991, John G. Cramer summarizes the problem with the existence of structures as large as the Great Wall:
"The vast size of the new large-scale structures is troubling, because in the accepted age of the universe, about 12 to 18 billion years, there isn't time for such large objects to form."
Cosmologists have tried
shoehorning these discoveries into their existing theoretical structures, by hypothesizing different kinds of
dark matter that could be responsible, or by putting forward the idea that the
Big Bang explosion contained irregularities which were sustained as time went by, resulting in '
clumping' of galaxies and clusters. However, each of these attempts to account for the Great Wall and other structures, and for the 'bubbly' appearance of
intergalactic space altogether, run into their own kinds of problems. For example, irregularities in the Big Bang would not explain the uniformity of the universe's background
microwave radiation. Some cosmologists are trying to piece together models which contain both
cold dark matter, which may explain the stability of galaxies, and
hot dark matter (
neutrinos), which may explain the larger-scale structures. However, this approach seems inelegant to many theorists, who are only willing to hypothesize a certain number of
invisible agents without starting to feel a little silly.
Another idea has been to resurrect Einstein's hated cosmological constant, a concept which (in its more modern form) effectively states that the 'energy density' of the universe can fluctuate locally, becoming slightly positive or slightly negative, causing matter to drift towards or away from other matter in a manner not explained by gravitation. There is some experimental evidence to support this idea - see the Casimir Effect - but really, no one knows anything at this point. Cramer simply says that "cosmology has just become much more interesting", and it may be that we have to wait for M-Theory to mature and develop before we can explain the existence of the Great Wall. It's reassuring to know that, throughout human history, just as science appears to be reaching a complete and unified theoretical understanding in any field, reality intrudes and something is discovered which overturns all existing theory and requires all the work to begin again from experimentation upwards. In this case, computer modeling is becoming indispensable to cosmologists, as most of these enormous structures could not have been observed or inferred by normal means.
References:
http://www.science-frontiers.com/sf067/sf067a08.htm
http://www.npl.washington.edu/AV/altvw47.html
http://www.damtp.cam.ac.uk/user/gr/public/gal_lss.html
http://archive.ncsa.uiuc.edu/Cyberia/Cosmos/WallsVoids.html