Is there a compendium of peer-reviewed journals that have been shown to be wrong? I've heard of scientific journals that only have required two reviewers. How often does rubbish get by; not just typos and simple mistakes, but fundamentally unsound rubbish? Is there a remedy process for these journals? At my research internship, during research we would follow an assertion back through three or four references to find that it was a (more or less) baseless intuitive prediction made in the 60's that just kind of propagated through the literature to today.