Go back to article: Giovanni Canestrini’s models of Leonardo da Vinci’s friction experiments


Leonardo’s contributions to the understanding of friction were remarkable, and undoubtedly based on some kind of experimental investigation; it is hard to see that he could have deduced the independence of friction force from contact area (which is counter-intuitive), or obtained quantitative values for the coefficient of friction in any other way (Hutchings, 2016). Kemp (2006) has commented that Leonardo’s methods of investigation were ‘an untidy mixture of deductive and inductive reasoning, habitual observation, hands-on intervention, “thought experiments”, “drawn experiments”, actual experimental testing and analogy’. It is impossible to be sure whether any of his sketches and notes relating to friction represent real experiments that he actually performed, or were illustrations of concepts or of thought experiments[21]. But the discussion above firmly suggests that the sketches used as the basis for Giovanni Canestrini’s models either, despite being concerned with friction, fell into the latter category (in the case of Model A), or have been completely misrepresented as relating to friction experiments at all (in the case of the four other models).

Since the late 1920s numerous models have been constructed and exhibited, based on Leonardo’s ingenious designs of machines, machine elements and measuring instruments, although no other models relate to the investigation of friction. Notable are those created for the 1939 and 1953 Milan exhibitions, including the models discussed above, whose history has been reviewed by Giorgione (2015). The Science Museum has its own smaller collection, produced for the 1952 celebration in London of the quincentenary of Leonardo’s birth and originally exhibited at the Royal Academy of Arts (Bennett, 2015). Models produced in the USA for exhibition in Los Angeles in 1949 were subsequently acquired by IBM and this collection, later augmented, formed the basis of travelling exhibitions from the 1960s to the 1980s (Landrus, 2013). Many of the ex-IBM models together with more recent additions now form a large collection in the Museo Leonardiano in Vinci, while some others are on permanent loan to the University of Technology Sydney, Australia.

The rediscovery of the Madrid Codices provided further inspiration for mechanical models in the 1970s. Several constructed in the 1980s for the Museum of Fine Arts in Montreal were included, together with a couple from the MUST collection, in an exhibition at the Hayward Gallery, London in 1989 (Kemp and Roberts, 1989). The Museo Galileo (Istituto e Museo di Storia della Scienza) in Florence holds a rich collection dating from the 1980s and 1990s, and many of these were included in the only previous exhibition of Leonardo’s engineering at the Science Museum (The Art of Invention: Leonardo and Renaissance Engineers, 15 October 1999–24 April 2000, see Galluzzi, 1999). There are also several commercial ‘Leonardo da Vinci museums’ worldwide, each with its own collection of models of varied fidelity to the original sources. As discussed by both Bennett (2015) and Giorgione (2015), construction of models based on Leonardo’s drawings involves considerable extrapolation and interpolation from the sometimes sparse detail of the original source material.

Interest in Leonardo da Vinci and in his technical achievements remains remarkably high. In this context it is understandable that one might wish to display models to illustrate Leonardo’s experiments on friction. It is perhaps natural to want to base such models, as Canestrini did, on original drawings that show perspective views and are visually more attractive than the rough two-dimensional sketches found elsewhere in the notebooks. But the acid test must surely be whether the viewer can see an answer, or even see that there might be an answer, to the question ‘How might Leonardo have used this for his friction experiments?’. Unfortunately, Giovanni Canestrini’s models fail that test.



I am very grateful to Professor Jim Bennett for his encouragement, helpful discussions and valuable suggestions – including, particularly, the idea that the main value of Leonardo models may lie in the detailed examination of the drawings they enforce, which may yield new insights (Bennett, 2015). Dr Susan Mossman kindly arranged to make measurements of the exhibit while it was at the Science Museum.

Component DOI: http://dx.doi.org/10.15180/160602/004