The Belgian bridge builders Charles Marcellis (1798-1864) and Arthur Vierendeel (1852-1940) were jack-of-all-trades in the 19th century, but both were mostly known as bridge designers, trying to have a grip on new structural possibilities. Though the mechanical behaviour of their bridges is very different, Tom Peters already noticed the visual resemblances between Marcellis' girder and box bridges made out of pierced cast-iron plates (1835-1860) and Vierendeel's development of the bridge type named after him (1890-1940). The non-simultaneous yet very similar evolution of these two characters is a duet with consonants and dissonances. Marcellis was an industrialist whereas Vierendeel was an engineer, professor and self-made art critic. Marcellis had imported the idea of cast-iron girder and box bridges from England (e.g. from Fairbairn and Stephenson) and he did not shrink from calling this a Belgian system to erect bridges. Vierendeel on the other hand, after having seen bridge collapses where the diagonals were hardly deformed, developed a simplified arithmetic method to calculate a beam that consists only of a series of rectangular frames, a system that still finds use in design problems today. This recurring pattern of engineering feats is the connecting thread between Marcellis and Vierendeel in this paper. Within a time frame of 100 years both men stood in a fascinating period on new materials (transition from cast iron to steel) and new calculation methods (transition from elementary formulas and trial-and-error testing to full understanding of secondary stresses and mechanical behaviour of materials).