By this time — such is the dizzying speed with which Stem Cells and Development goes about its business — a second version had been submitted with the offending paragraphs rewritten. But Professor Parker was unmoved. It remains unpublished.

The outcome is that neither version now exists for other researchers to read, which is a serious shame. Doubts had already been cast on the claims it made, principally by Dr Allen Pacey of the University of Sheffield, an expert on reproductive biology, who said the Newcastle lab's products were too abnormal to be called sperm. But even if Dr Pacey is right, the Newcastle team had done something interesting. But what, exactly? Without the paper it is hard to tell.

Scientific fraud is increasing, and retractions have increased roughly tenfold in the past 20 years. Plagiarism is responsible for about a quarter of cases, and some journals now use special software to detect if authors have cheated in this way. But the numbers remain tiny — 95 retractions last year from 1.4 million papers published. Either scientists are very honest, or extremely hard to catch.

The episode probably tells us more about publishing than it does about science. Why the Newcastle team chose Stem Cells and Development, one of a stable of journals published by the US entrepreneur Mary Ann Liebert, is puzzling. It is possible the paper had been rejected by more prestigious journals. Maybe the journalists should have shown more scepticism: certainly, given the prominence of the original stories, the retraction deserved at least a mention.