In recent years , upper limits on the cosmic microwave background ( CMB ) anisotropies combined with predictions made by theories of galaxy formation , have been extremely powerful in ruling out purely baryonic dark matter ( BDM ) universes .
However , it has recently been argued that the absence of a prominent second peak in the anisotropy spectrum measured by the BOOMERanG-98 and MAXIMA-1 experiments seems to favour a \Lambda BDM model when combined with standard Big Bang Nucleosynthesis ( BBN ) constraints .
In this Letter , we further investigate this result showing that , using the CMB data alone , a purely baryonic adiabatic model of structure formation seems unlikely if the universe is flat ( \Omega = 1 ) .
Combining the CMB data with supernova type Ia ( SNIa ) data renders purely baryonic models inconsistent with flatness at high significance and more than 3 \sigma away from both the BBN constraints and the HST key project result h = 0.72 \pm 0.08 .
These results indicate that only a radical revision of cosmology with ad hoc properties could bring baryonic models such as those advocated by MOND enthusiasts back into agreement with current observations .