It would be hard to find a cosmologist today who does not believe that the vast bulk of the Universe (ninety-five percent or more) is hidden from our eyes. We review the evidence for this remarkable consensus, and for the latest proposal, that the mysterious dark matter consists of as many as {\em four separate ingredients}: baryons, massive neutrinos, ``exotic'' dark matter particles, and vacuum energy, also known as the cosmological constant (Lambda). Of these, only baryons fit within standard theoretical physics; the others, if their existence is confirmed, will mean rewriting textbooks. New experimental evidence has recently appeared for and against all four components, so that the subject is in a state of turmoil and excitement. The past three years in particular have seen the fourth (vacuum) component come into new prominence, largely at the expense of the third (exotic dark matter). We conclude our review by exploring the possibility that the energy density of the vacuum is in fact so dominant as to leave little room for significant amounts of exotic dark matter.

Evidence for a positive cosmological constant $\Lambda$, derived from the Ly$\alpha$-forest in high-resolution spectra of quasars, leads to a closed, low-density, $\Lambda$-dominated universe. The analysis is based on the assumption of a universal shell structure expanding predominantly with the Hubble flow. Supporting evidence comes from two pairs of very wide absorption lines in the spectra of two quasars separated by 8 arcmin on the sky. These results contradict the higher values of the density parameter $\Omega_{M,0}$ derived, for example, from clusters of galaxies by the pure gravitational instability theory. Implications thereof and for the amount of non-baryonic dark matter in the universe are discussed with some $\Lambda$-dominated models, in particular with $\Omega_{M,0} = 0.2$ and with pure baryonic models. The evidence points to low-density, closed models with spherical metric, expanding forever.

We present a new explanation for the very wide absorption features in quasar spectra. In our model, a very wide absorption feature will originate, when the line of sight crosses a bubble wall tangentially. We demonstrate this on the quasar pair (2138-4427), (2139-4434). Both show two very wide absorption lines in their spectra at the same redshift. The bubble wall model can explain these observations in low density Friedmann-Lemaitre models with spherical metric. It contradicts models with euclidian or hyperbolic metric.