Les instruments du calcul savant > Instruments d'intégration conservés au musée des arts et métiers


Integraph with transfer of direction by cog-wheels, system Abdank-Abakanowicz/Napoli
Maker: P. Barbier & Cie., Paris; instrument no.: none (presumably one-of-a-kind); c1885 (see Abdank-Abakanowicz 1886, 39)
Inventory: CNAM, inventory no.
Details: Entry CNAM: 1900
References: Abdank-Abakanowicz 1886, 44-48 and Abdank-Abakanowicz 1889, 37-44 (with illustrations)

Between 1880 and 1889 Bruno Abdank-Abakanowicz (1852-1900) tried out a substantial number of different mechanisms in order to solve the problem, so important for integraphs, to transfer a direction extracted from a given curve on a knife-edged wheel. A fruitful cooperation came into being around 1885 with David Napoli (1840-1890), the chief inspector and director of the workshop of the French Eastern Railway (inspecteur principal des chemins de fer de l'Est et chef du laboratoire de cette Compagnie; Abdank-Abakanowicz and Napoli had met 1883 at the Vienna Exhibition). It seems that Napoli had the idea to use cog-wheels in order to solve the problem of transfer of direction.

Fig. 37, Abdank-Abakanowicz 1886, 46 = Fig. 39, Abdank-Abakanowicz 1889, 40

The construction plans shown give a plan-view of the instrument. The cog-wheel pair to the left always remains at its place, but its wheels are turned by the bar (shown in an oblique position in the drawing). An axis with quadratic section transfers this movement on another cog-wheel pair, which is allowed to glide along the axis, so that its distance relative to the first pair may change. The axis with its quadratic section keeps the cog-wheel pairs in absolutely the same direction; thus the knife-edged wheel mounted below the second pair will have the same direction as the tracing bar. Moving the instrument in the direction of x (in the drawing: to the bottom) will produce the integral curve as the trace of the knife-edged wheel. A writing pen is mounted next to the wheel and will produce a more visible trace of the integral curve. Due to the general construction of the instrument the ordinates of the integral curve are shifted a certain distance with respect to the original curve.

Fig. 38, Abdank-Abakanowicz 1886, 47 = Fig. 40, Abdank-Abakanowicz 1889, 41