Abstract
It is argued that P. Ellworth's (1994) interpretation of William James's theory of emotions is not supported by his writings. On the basis of textual evidence and systematic considerations, the authors try to show that, in line with the traditional view, James's theory holds that emotions are identical with feelings of bodily changes, of which autonomic changes are by far the most important ones (i.e. the only ones necessary and sufficient for emotions). Furthermore, the authors argue that Jamses's question of the temporal priority of emotions versus bodily feelings makes good sense even if one acknowledges that these events usually show temporal overlap. (A)