While there had been some precocious excavations, such as the exploration of the Duntocher bathhouse in 1775 (Keppie 2004a) or Adam de Cardonnel's 1783 work on the Inveresk hypocaust (de Cardonnel 1822), the era of scientific archaeological research did not commence until the 1890s. In 1890 the Glasgow Archaeological Society set out to determine if the Antonine Wall really was constructed of turf, while the Society of Antiquaries of Scotland sponsored a series of excavations at forts, fortlets and towers along its course. Both societies continued to sponsor excavations during the inter-War years. The work was inevitably of its time, but the techniques and interpretations stood comparison with those of wider Roman scholarship, and in some cases (notably Curle's publication of Newstead; 1911) greatly surpassed equivalent undertakings elsewhere in Britain.
This period also saw significant works of synthesis. In 1911, Sir George Macdonald published The Roman Wall in Scotland, the first modern treatment of the Antonine Wall, and undertook research excavations in order to help check the line of the frontier; this was updated in a second edition (1934). The sculpture and inscriptions in the Hunterian Museum were published (J Macdonald 1897), as were the first of a continuing series of coin surveys (Haverfield 1899, 159-168; G Macdonald 1918). Research also began in earnest on the impact of the occupation on indigenous societies, notably with James Curle's magisterial corpus of Roman finds from non-Roman sites (Curle 1932a).