Eduard Suess (1831-1914) is one of the most widely misunderstood and miscited authors in the history of tectonics mainly because of the nature of his writings in which very detailed local descriptions are tightly interwoven with novel theoretical interpretations. Two short publications by him, an abstract with the title Ueber den Aufbau der mitteleuropaischen Hochgebirge (On the structure of the middle European high mountains) published in 1873 and a letter he wrote to the editor of the English translation of Das Antlitz der Erde, William Johnson Sollas and which was published as the 'Preface by the Author' to the translation in 1904, may be taken as guides to probe his thinking on tectonics by finding the continuous thread running through his publications pertaining to tectonics. In fact, it is quite impossible to understand what his basic tectonic picture was without being familiar with the 1873(a) abstract, which is never cited in the literature and has not yet been examined by historians of geology. After having read it, one has to understand then why Suess stuck to the contraction theory. The answer to that is in his letter to Sollas. Basically, Suess saw that mountain-building was a consequence of motions of discrete rigid to semi-rigid lithospheric blocks moving independently with respect to one another. While a block moved to shorten its frontal part, it caused extension in its wake. Such motions of independent blocks Suess likened to the motions of ice floes in drifting pack ice. When he considered global stratigraphy, he realised that the main transgressions and regressions were global and it was them that governed the dominant character of the stratigraphic time-table. Changing the capacity of ocean basins was the only way, Suess thought, to bring about transgressions and regressions. To do this, Constant Prevost's model of global contraction (not Elie de Beaumont's, accepted by Dana and Le Conte) provided the best mechanism. Prevost's model worked so well for stratigraphy that Suess felt that it had to be right also for tectonics. To use Prevost's contraction for mountain-building and rift-making, Suess had to assume different depths of detachments and irregular regions of attachment of one storey to the other along such detachments. Qualitatively, Suess' tectonic model was the best ever offered before plate tectonics and plate tectonics preserved many of its basic elements and even details of some of them.