I wish to defend 'the dictionary view' of scientific theories against Miss Hesse's attack and to argue that she has insufficiently considered her assertion that phenomenal statements do not 'constitute a basic empirical language which is independent of all theory'. I agree that the word ' dictionary' is likely to mislead in this context and that the 'theory-loading' view of phenomenal statements, much in vogue among Angry Young Philosophers, contains a grain of truth. But, on the whole, I am a reactionary in these matters. So reactionary am I that I would be prepared to talk of observation statements, but I defer to Miss Hesse in this. However, I should like to introduce another term. Miss Hesse means by 'phenomenal statements' (' almost always') those which' do not describe what happened on a particular occasion to a particular observer, but what always happens and will happen on sufficiently similar occasions to all normal observers' (p. 15). Since the testing of theories is done by individuals who can only be in one place at a given time, I wish also to refer to 'test statements '. These are whatever singular statements are used, in the last resort, to test a theory, are directly verifiable and, at their simplest, are those singular statements which can be deduced from Miss Hesse's general phenomenal statements together with singular existential statements. Depending on the context, test statements may contain technical terms of some scientific theory or they may be in the simplest common-sense language understandable by all normal people. In happier circumstances I should call these last, at least, 'observation statements '.