This study extends systemic-functional linguistic accounts of explanation genres in school science books with a view to enhancing explicit, linguistically-based pedagogies, which see students' learning of science as concomitant with their learning to control the semiotic resources that construe scientific knowledge. The study has two main interrelated concerns. The first is a clarification and extension of recent functional linguistic work describing different types of explanation in school science books. The second is an account of the linguistic variation within and across such explanation types and a comparison of this variation across junior secondary school textbooks and information books used by children in the upper primary school. The analyses provide some linguistic evidence to support the distinction of at least three agnate explanation genres: transformational, realizational and orientational explanations. The comparative analyses of secondary texts in the same field enabled an interpretation of the relative effectiveness of linguistic choices in each text in contributing to its explanation of the phenomenon. One text in each of two fields, and two texts in the third field, clearly provided more effective explanations than the other texts. The fact that three of these four texts were from the same textbook series with the same authors, suggested that composing effective pedagogic explanations is an aspect of expertise which some authors exercise consistently. The analyses of the primary texts showed that the proposed accounts of schematic structure potential were applicable across school levels. Although significant inadequacies were identified in at least one text in each of the three fields, the remaining texts in two of the fields provided functional approximations to the most effective secondary texts. This seemed to be due to the capacity of the linguistic construction of these explanation types to accommodate the obvious recontextualizing strategy of keeping explanations brief in primary school books. However, this was not the case with realizational explanations of sound waves, which were problematic in al primary texts. On the whole the primary texts compared poorly with the secondary texts in scaffolding the novice readers negotiation of the specialised language of school science.