Sir

We have counted the numbers of errata, corrigenda, corrections and addenda published in all 2004 Nature print issues, across all sections of the journal. During the past year, Nature has reported flaws in 32 peer-reviewed research papers, of which 24 were corrigenda (author corrections) and 8 were errata (journal corrections); there were also 2 addenda. Although all these corrections were published last year, 14 of the erroneous papers were published before 2004. Within the other sections of the journal there were 14 errors reported.

Worryingly, in 14 out of 34 cases Nature failed to attach an amendment page with the online PDF of the original paper. Nature has also failed to provide an amendment notice with the abstract or HTML version of several research papers (7 times out of 34). Similar problems arise for corrections to non-research items such as News stories.

With the widespread practice of accessing, printing and circulating PDF files through the Internet, it seems advisable to take this matter seriously.