The review of Stanley Prusiner's autobiography (G. Mallucci Nature 508, 180–181; 2014) suggests that the idea of an infectious protein was first put forward by Tikvah Alper and colleagues (Nature 214, 764–766; 1967) and by John Stanley Griffith (Nature 215, 1043–1044; 1967). This perpetuates a common myth.

Alper concluded from radiation-inactivation data that the agent that causes scrapie, a neurodegenerative sheep disease, does not depend on either a nucleic acid or a protein to replicate, favouring an earlier suggestion that it might be a replicating polysaccharide.

Griffith opens his paper by crediting the idea that the scrapie agent is a protein to an earlier paper by Alper and colleagues (T. Alper et al. Biophys. Biochem. Res. Commun. 22, 278–284; 1966), and also to I. H. Pattison and K. M. Jones (Vet. Rec. 80, 2–9; 1967). In fact, this earlier Alper paper does not contain the word 'protein'. Griffith's second claim is correct. Pattison and Jones made their suggestion because the techniques they used to purify the scrapie agent were the same as those used to purify small basic proteins.

This myth probably persists because the key 1967 papers are not freely accessible online, making it harder for today's busy scientists to check their facts.