First there was SciGen for generating gobbledygook papers that sounded like computer science. Now there is Mathgen, a service for producing LaTeX-formatted papers on mathematics that are complete nonsense.
The blog That's Mathematics written by Nate Eldredge reports that one such paper has been accepted by the "journal" Advances in Pure Mathematics, published as Open Access along with over 200 others in a wide range of fields by SCIRP. At least someone tried to read the manuscript – there are a few editor's notes, requesting, for example, a proof of one of the more ludicrous theorems. But after the author wrote back the paper was accepted, contingent of 500 € being deposited in the bank account of the journal.
These junk journals are giving Open Access a bad name. They spam our inboxes with invitations to submit articles on an extremely wide range of topics. They offer, for a price, an international publication with an ISBN or ISSN, just the thing for administrative bean-counters who define quality as being the quantity of publications There does not appear to be an easy way to stop them.
If you read German, there is an article worth reading in the Zeit Online collecting many of the cases that have been discussed here and elsewhere, as well as some wonderful new ones such as Mathgen. Eldredge has also published a randomly generated book, Galois Knot Theory.