This fifth volume of the collected works explores 'the semantic character of scientific discourse.' It opens with a new essay by Professor Halliday in which he looks at the power of language to make meaning, and addresses the question 'how big is a language'? In the essays that follows Halliday argues that there is no single register of science, but there are numerous scientific discourses. Looking at the history of scientific discourse, it is possible to see new strategies evolving which are grounded in processes of metaphor. These grammatical metaphors increase the power that a language has for theorizing and hence it is within these metaphors that much of the power of language resides.