“All this while, there was only the swift motion of the cylinder round its axis, without any attrition, but now when I came to apply my hand to the lower part of the glass (so swiftly whirled about) and consequently to add attrition to the former motion; the threads presently began to change their direction, and all harmoniously pointed to the center of the circle, in whose plane the wire was placed, as in fig. 3. neither were they at all disordered or flung out of that position, by the wind occasioned by that violent motion (but as if there had been no such hurry of air about them) they still persisted in their central direction; I found I could by shifting the place of the attrition hither or thither, draw the threads towards this or that end of the cylinder; but yet they all still went uniformly converging towards some center in the axis of it; so that they formed themselves into a sort of conical surface. “Farther, if the wire with its loose threads was reverted, so as to encompass the lower part of the cylinder (as before it did the upper part) yet the effect still answered with the same exactness. For the threads were all erected into so many strait lines, still directing themselves towards a center in the axis of the glass. “Hitherto the axis of the cylinder was placed horizontally; in the next place I set it in a vertical position, so that it stood perpendicular to the plane of the horizon; in which case I made use of a wire hoop, which was necessary to be placed parallel to the horizon, that it might encompass the cylinder, in the same manner as the semicircular wire did before: only one small part of this wire was left open, to make way for the touch of the hand, which was to give the attrition. And the wire being thus placed, it was evident that the threads (without some external force to support them) must all flag and hang perpendicularly downwards. Yet, as soon as the motion and attrition were given, the threads presently began to be extended; and as if they were become stiff and hard, formed themselves into an horizontal plane; their loose ends pointing to a center in the axis of the glass, as before. “And thus (in all sorts of positions whatsoever, both of the wire and of the glass too) were the threads acted upon by a sort of centripetal force; to the laws of which they were always conformable.” See _Hauksbee_’s experiments, page 53 &c.