I first made the drawing for the governor with the weight hanging to the slide. Mr. John McLaren, a machinist who had done good work for me, when I showed it to him said, “Why don’t you turn your weight upside down and put it between the arms?” I was not long in acting upon this suggestion, and that made the Porter governor complete. I had it described and illustrated in the Scientific American. They took a photograph of it as photographs were taken in those days—that is, they sent their artist up to make a sketch of it, and this sketch (shown [here]) and description will be found in the Scientific American of October 9, 1858. This governor has never been changed by me except in the shape of the counterpoise.
I believed the mathematics of my advisers to be sound, and that the perfect action of the governor was obtained entirely by the long driving-joint, which I supposed would have enabled the 36-lb. balls at 50 revolutions per minute to do just as well as 1-lb. balls at 300 revolutions, but I never tried the experiment.
In that belief I remained for 50 years. Now, at the age of over 80 years, after long rest from business activities, in revising these reminiscences for publication, the idea has first occurred to me, and has grown into a conviction, that my advisers were wrong here as they had been in every other respect. They overlooked the fact that the angular velocity of the driving-joint increased equally with that of the balls, so that the ratio between them would remain constant. The law that the driving force required increases as the square of the speed imparted applies only to the original source of power, as, to the force of the steam exerted in the cylinder of an engine, the motion of the piston remaining the same, and to the transmitting belts or gears whose speed also remains the same. At all these points the force exerted must increase as the square of the speed imparted; but this does not apply to the pressure exerted in the governor joint. Its speed does not remain the same, but increases with that of the balls. So, while the centrifugal force of the balls, changes in which produce the vertical movements of the counterpoise, varies as the square of the speed, the force required to be exerted in this joint to drive the balls, and which produces the friction to retard these movements, does not increase at all, whatever the speed of revolution may be. This fact, unobserved by me or any one else so far as I ever heard, has all the time been the secret, a pretty open secret when once seen, of the surprising combination of sensitiveness and stability in the action of this governor which has led to its general use, and at which I myself have never ceased to wonder because I was ignorant of its cause. This, however, was not the only time that I builded better than I knew.
I can imagine some persons, after having read the above explanation, to say, some of them perhaps flippantly, and some possibly sneeringly, “To a properly educated engineer this is obvious at a glance.” I think it will be so hereafter, but has it been so hitherto? If any one will produce the record of its observation I will cheerfully yield to him the priority and will congratulate him upon it.
Some things, however, make me doubt if this observation has ever been made. At the London Exhibition of 1862 this governor attracted much attention from its novel appearance, rapid rotation and remarkable action. Many engineers spoke to me about it. In their conversation I observed two things: first, no one ever asked me a question, but every one explained its action to me; and second, while each had an explanation of his own to make, they all agreed in a fundamental respect. Their minds ran in the same groove. They considered the governor only in its theoretical action. No one ever took notice of the incident of friction, which was the controlling factor. An improved governor was in their view one contrived in some way to free the governor from the limitation to its action, which is imposed by the law of the conical pendulum, and every one explained to me how my governor was adapted to do this.
The following illustrates this universal view among English engineers:
In the Appendix to the 10th edition of Rankine’s “Manual of the Steam-engine and other Prime Movers,” published in 1882, one reads as follows: “Isochronous governors. The ordinary governor is not isochronous; for when, in order to adapt the opening of the regulating-valve to different loads, it rotates with its revolving pendulums at different angles to the vertical axis, the altitude of the cone assumes different values, corresponding to different speeds. The following are expedients for diminishing or removing this defect.
1. Loaded Governor (Porter’s).—From the balls of the common governor, whose collective weight is (say) A, let there be hung by a pair of links of lengths equal to the pendulum arms, a load, B, capable of sliding up and down the spindle, and having its center of gravity on the axis of rotation. Then the centrifugal force is that due to A alone, and the effect of gravity that due to A + 2B; consequently the altitude for a given speed is increased in the ratio A + 2B : A, as compared with that of a simple revolving pendulum; and a given absolute variation of altitude in moving the regulating-valve produces a smaller proportionate variation of speed than in the common governor.”
That is the whole of it. Respecting this I have to say:
1st. The vertical motion of the counterpoise (variation of altitude), if the links had also a single joint at the bottom, could not be either more or less than twice that of the balls, which equal lengths of the arms and links give also in the common governor, so in this respect the governor is no improvement.