Almost 32 years ago, I wanted to find out whether iron wires floating on water placed at different inclinations in relation to the Meridian would retain their position, and return to the same situation and direction in which they had been set, as William Gilbert says. While I was observing attentively, an amazing spectacle was offered to me, which had not been noticed hitherto: some extremities of the floating bodies rushed to each other to unite closely while others separated as occurs for a magnet and iron. Thus, stimulated by this novelty, I tried again using other corpuscles, straws, leaves of trees, and countless other bodies. I wished very much to understand the cause of this effect. After countless experiments, I noticed that these opposite operations resulted from the edges of water round adjacent floating bodies, some raised, some depressed. I took two copper sheets, thinner than paper, V and X (Table 8.​1, Fig. 10). At their centres C and L, I applied two straws CD and LM which I welded with wax perpendicularly to the planes of the sheets. I then set the sheets horizontally on the water of a vessel FRSO. When the floating sheets were depressed below the surface of the water, they were surrounded by water edges EA, GB and IN, KO. Then I made two wooden chips Y and Z both of an equal thickness of about half a finger breadth (Table 8.​1, Fig. 11). I also adapted to them perpendicular straws. When they were set on water, sloping wedges EA, GB formed round their circumference, above the surface FHO of the water. After this preparation, I dextrously pushed with my fingers the top D of a straw to move the small sheet V towards X, while retaining this firmly before it came into contact with the other sheet V. The distance between the sheets was less than one finger breadth. Then, firstly, I saw that the two sheets V and X spontaneously moved towards each other. Although retained by the small effort of a finger and prevented from reaching each other, afterwards they joined quickly, not less than previously. In the process of their union, the water edge GHN by which the sheets were separated previously levelled off completely. I then turned to the wooden chips Y and Z which were equally immobile and inert when distant from each other by more than a finger breadth. Drawn closer, the chips suddenly moved towards each other to join. This process was different from the previous one in that the two raised wedges GB and IM, not only did not level off nor did they move down to the surface of the subjacent water but, on the contrary, the intermediate space and the cavity BHI were filled completely up to the top BI. Finally, I joined the copper sheet V with the chip Z (Table 8.​1, Fig. 12) and I saw that, whenever I drew them closer at a distance of less than one finger breadth, not only they did not unite but, on the contrary, they rapidly moved away from each other and separated as if they abhorred their mutual view and vicinity. Experience thus shows that drawing closer and joining of sheets occur only when the water edges are similar, i.e. when both are raised above or when both are depressed below the surface of the water. When they are different, one being depressed below and the other raised above the surface of the water, separation and escape of the chips occur. In all these operations it was found that, if one of the sheets is retained fixed and immobile, or rather if it is in the orifice of a vessel, the other sheet free and not retained either joins or avoids the contact with the former which is immobile. Since both float freely over the fluid, the movement is common to both bodies. There is a difference, however: the smaller and less ponderous body joins or escapes the other more quickly while the displacement of the larger sheet is slower and more nonchalant. Such is the true and accurate story of this surprising effect. Thus, I do not wonder that the true cause of this effect was not proposed since the history of this operation was not obvious and was not noticed perfectly. It could be observed clearly and evidently only by way of the mentioned small sheets which I devised.

I reported this experiment to friends many of whom are still alive, in Sicily as well as in Rome; afterwards, in 1655, to the Serene Grand Duke Ferdinand of Florence, to Prince Cosme of Tuscany and the best patron of the arts, and to the very wise Leopold, Cardinal Medici, who kindly wrote in his recent letter that he remembered well my demonstration and the explanation which I presented at this time. Later, after the institution of the Accademia del Cimento, I publicly showed this experiment to its very learned members and to countless famous gentlemen of different nations to whom the spectacle of the best experiments of this Academy was offered, under the chairmanship of the Serene Cardinal.

Beside these novelties, I observed another one resulting from a very different cause. For its understanding one must review at first the fairly common effect of two perfectly flat and smooth small sheets of glass which are joined closely together and hold with such pertinacity that, if the upper one is raised horizontally, the contiguous lower sheet follows and is pulled and sustained hanging as if it were glued and bound to the upper one. If the upper glass sheet is a little inclined to the horizon, the lower one suddenly glides towards the sloping part of the upper plane without parting from the upper sheet. Always adhering to the upper sheet, it descends impelled by the natural instinct by which heavy bodies attempt more and more at reaching the centre of the earth the way they can, i.e. through an inclined path since the direct and perpendicular path is impeded.