Hi
Whould you consider ancient texts such Tacitus : Agricola and Germania, to be a primary or secondary source ?
Ancient historians don't really divide source material up like that. We may refer sometimes to primary material, and secondary literature is always just secondary literature, but the nature of an ancient historian's sources is fundamentally unlike what a modern historian, with his neat distinctions between primary and secondary material, works with. Setting aside ancient histories that are primarily material due to an absence or scarcity of literature, ancient history is effectively a literary history. In the absence of diaries, letters, and government documents the ancient historian mines historiography, oratory, poetry, and so forth, sometimes contemporary to the events described but more typically following them, sometimes by some great distance. Moreover, ancient historiography is a literary genre, lacking the strict methodology and academic restraints of modern historiography. Therefore, the way that even the historiographical sources are treated has to be fundamentally different. There's an old thread on here somewhere that insists that the historian's work is not really detective work, but in many cases the work of an ancient historian really is closer to a detective's than a modern historian, since we have to piece together information from scattered, fragmentary literary sources, and try to get them to work with archaeological and epigraphic evidence before we can even get to the "proper" work of history.
An ancient historian would not call the Agricola and the Germania (which are two separate works, not one) "primary" sources except as a sort of shorthand, or if he were working on Tacitus as a historiographer, not as a source of history. We refer to "the sources" or "the texts," further breaking them down from there according to somewhat shifting categories of reliability and usefulness based on how contemporary they are, what sort of literature it is, and so forth. More important to the work of the ancient historian is the nature of a source. We distinguish between literary evidence, epigraphy, and papyri instead of understanding everything that isn't modern, academic, secondary literature as "primary"