Robert Hooke's Inventions

Robert Hooke (1635–1703) came up with a great many novel and interesting ideas and inventions; whether one favours or opposes patents, one should therefore give at least some consideration to the effect of the patent system on his case. He lived in an era when useful innovation was much less widespread than it is today; when natural philosophy was still new and the scientific method was still being developed; and when the patent system was a good deal less sophisticated, being still a royal prerogative, bestowed through the favour of the King and generally best applied for, therefore, by obtaining the services of some one blessed with the King's favour to plead for the patent. There was also no international propagation of patents — the King conferred the patent so as to promote the exercise of the technology within the economy on which he got to levy taxes. Indeed, foreigners were not even eligible to hold patents.

Balance Wheels

Hooke was one of the players in the development of half-way decent pocket watches (for truly good ones, the world had to wait another century). The pivotal invention which made this practical was the spring balance, invented in the second half of the 1600s by (at least) both Hooke and Huygens. Hooke's side of the tale reveals all the complications attached to the economic exploitation of novel ideas.

Background: commercial significance

Ships at sea need to know where they are if they are to avoid disaster. Running aground on shallows can wreck a ship in any weather; in fog, it is all too easy to run into rock one would otherwise have seen. When sailing along a coast, it is generally possible to keep track of progress and thus of location; even when out of sight of land, it is possible to estimate the ship's motion and thereby keep track of location to moderate precision, especially with the help of a lodestone (which we now call a magnet). However, the precision of such dead reckoning decreases with distance; the uncertainties and errors add up from day to day, making the method singularly unreliable for long voyages. In an age of global trade by sea, the resulting risks were of the utmost importance.

It is possible, from astronomical observations alone, to determine latitude (how far North or South one is), at least when the sky is clear. At the same time, one can determine the local solar time; from which one can determine one's longitude (East — West position) if only one also knows the time at some place of known longitude — this difference in time is 15 degrees of longitude for every hour of time. Thus a reliable watch mechanism promised to solve the longitude problem which was one of the most commercially and militarily important issues of the day: indeed, when Harrison finally solved it (about a century later), it was kept secret as a matter of national security.

Background: basic clock-making

The (light and delicate) spring used in regulating the rate at which a watch runs should not be confused with the (stout and brutal) spring which generally powers the mechanism. The driving spring of a watch (or clock) applies a torque (twisting force) to a cog, which drives other cogs and so round to the cogs that turn the hands.

Of course, the driving spring delivers more force when it is fully wound than when it is nearly exhausted; but that can be evened out by the use of a fusee. The spring is mounted on a spindle inside a cylindrical drum which turns on the spindle due to the force of the spring. A fine chain (one could as readily use a thread or wire) is anchored to the drum at one end and, when the spring is fully unwound, wrapped around the drum. The other end of the chain is attached to the base of a cone-like structure which shares a spindle with the cog that actually drives the rest of the clockwork. The cone has a track spiralling around it, presenting a surface parallel to the axis at radius varying from small, at the apex, to large at the base. The peg that one turns (generally with a key) to wind up the watch shares its spindle with the cone. As it turns, it pulls the chain off the spring's cylinder and onto the cone, first filling up the broad start thereof and so winding inwards along the track on the cone's surface.

The result is that the driving spring's force pulls at variable radius upon the axle of the chain's cone, which is shared with the cog by which the main sequence of cogs is driven. When the spring is fully wound, its large force is applied to a small radius near the apex; when the spring is almost exhausted its small force is applied to a large radius near the base. If the cone's variation of radius is correctly arranged, the force of the spring times the radius at which the chain pulls on the cone can be arranged to be constant. Thus a spring, whose power is variable, may deliver constant torque to a clockwork mechanism. I do not know whether Hooke and his contemporaries had this degree of sophistication.

However, even when the spring is persuaded to deliver a constant torque to the sequence of cogs from it to the hands, one must have some way to regulate the speed at which the cogs turn. Reliance on friction alone does not provide good precision — it is too dependent on external influences and the state of wear of the watch. In any case, in the interests of making the watch durable, one wishes to minimise wear and tear on its parts, to which end one avoids friction as far as possible.

In a pendulum clock, the speed is controlled by having a cog in the sequence from spring to hands regulated by the pendulum. The pendulum can be made to cause some mechanism to rock backwards and forward at a dependable frequency, and this can be used to (for example) allow one tooth of the regulated cog to pass in each cycle, thereby regulating the speed of the cog.

For example, in 1671, William Clement invented an escapement, known as the anchor, comprising a curved bar with two triangular prongs on its ends that intrude into the gaps between the teeth of the cog, preventing the cog from turning. The separation of the prongs is equal to the gap between one tooth of the cog and the centre of the gap between two others. The bar is so placed that there is always at least one of its prongs in the cog's way, but it is pivoted about its middle, so that either of them can be swung out of the way. As the bar rocks one way, its prong at one end gets out of the way of the cog's teeth, allowing the cog to turn until a tooth strikes the other prong; the bar then rocks the other way, bringing the first prong back into the way of the cog's teeth and moving the second prong out of the way of the tooth it obstructs, allowing the cog to again turn, but only as far as will cause a tooth to strike the first prong. Each cycle of the anchor's rocking motion thus lets the cog turn by one tooth. With a little care, the mechanism for doing this can be so arranged as to transmit a small amount of force, from the regulated cog (driven, indirectly, by the main spring) to the mechanism that restricts its motion and, thus, to the pendulum, so as to ensure the pendulum keeps swinging (care must be taken to do this without, in the process, disrupting the regularity of the pendulum's swing).

This technology (and its less sophisticated predecessor, the verge, as modified by Huygens to use a pendulum) was well understood for pendulums, but a pendulum is very sensitive to movement of the frame from which it hangs; this makes a pendulum clock unsuitable for a pocket watch or, indeed, a ship-board clock.

The important property of a pendulum is that it swings with a period that scarcely varies with the size of its swing (and some clever games with catenaries can be used to eliminate such little variation as there is). This is because it is (to a fairly good approximation) a simple harmonic oscillator. Any system with a stable equilibrium position, in which the parts move without friction and there is a restoring force, tending to move the parts towards that equilibrium position, will be a simple harmonic oscillator if the size of the restoring force is proportional to the size of the displacement from equilibrium – but this was not known (or at least proved) until after Newton and Leibniz developed the infinitesimal calculus.

More recent watch mechanisms use a fine plane spiral of wire as a spring; its outer end is mounted on a small light cog (the balance wheel) and its inner end is held fixed. The wheel, when pushed off its equilibrium position, then spins back and forth, oscillating at a regular frequency. Some lightweight cogs and gearing then transmit the wheel's motion to a bar which rocks to and fro, just as in a pendulum clock, to regulate the motion of the main sequence of the watch's cogs. Such a mechanism lends itself well to the task of regulating a watch. This is likely more sophisticated than the mechanisms Hooke and Huygens devised.

The Invention

What was needed was a mechanism that would reliably oscillate with a fixed period, that could be used to serve in place of a pendulum in a mechanism otherwise essentially the same as was used by clocks. Hooke discovered that the force a spring exerts is proportional to its extension (or, indeed, compression) – ut pondus sic tensio or ut tensio sic vis. This (though he did not know it) is exactly the pre-requisite for a spring to be a simple harmonic oscillator. Hooke noticed that a spring, displaced from its equilibrium position and released, oscillated with a regular frequency, so realised it could be used to regulate a watch.

In the early 1660s Hooke clearly researched the possibilities of using springs to regulate watches, leading to an attempt, in 1664, to enlist some men of influence in a scheme to get him a fourteen year royal patent for his (claimed) solution of the longitude problem. This fell through for failure to devise a contract agreeable to all parties. Hooke clearly had, and demonstrated to the others, a watch regulated by a spring: but, since he kept its details secret, we cannot know what form the spring took.

Also in 1664, Hooke gave a public lecture in which he expounded on twenty ways of using a spring to regulate a watch, and indicated that this could be done in a hundred ways. It is not clear how many, if any, of these methods he actually tried, let alone found the means to make work reliably. He described the use of counter-rotating balance wheels (on a common spindle, linked by some cogs) to ensure that shaking of the watch would not interfere with its working (because, in so far as the shaking turned one wheel with its motion, it would also turn the other against its motion) and it seems likely the watch he wanted to patent employed such a contrivance.

Huygens devised a coiled spring. He communicated to the Royal Society by a letter – which the secretary, Oldenburg, read to the assembled members on 28 January 1675 – that he had a new invention relating to watches, and included (as was common at the time — Hooke did the same for his law) an anagram which encoded the fact that he was using a coiled spring. The details of Huygens's spring are unclear, but we may surmise that it at least loosely resembled the mechanism described above that was subsequently widely used.

Hooke claimed priority over Huygens, and convinced himself that Huygens had learned of his invention from Oldenburg (whose job, as secretary, was to communicate the Royal Society's activities to its overseas members). However, in 1675, Hooke clearly did not have a working watch based on a spiral spring: he spent much of that year working with Thomas Tompion – the father of English watch-making – to produce such a working model. Hooke had a widely noted habit of exaggerating his claims and of declaring everything one might hope for from his inventions without actually going to the trouble of overcoming the practical issues that stood between his idea and the applications declared; so it is reasonable to treat his claims of priority with skepticism.

Oldenburg and Lord Brouncker attempted (on Huygens's behalf – as a foreigner he could not hold a patent) to persuade the King to grant them a patent; Hooke worked with Tompion on a rival bid, and Hooke endeavoured to persuade the King that Huygens's invention was a derivative of Hooke's earlier work, going back to 1658. Ultimately, the King granted no patent and the London watch industry blossomed.

Relevance to Patents

If Hooke had had the modern patent system to work with, he would have taken out a patent in the 1660s covering the use of any mechanical contrivance that used the regular oscillations of a spring to control the speed of clock-work. He would not need to have made a working watch: when Huygens finally produced one, he would have had to pay Hooke for the privilege of using the idea. But the modern patent system is broken – it no longer demands a working model and (at least in fields familiar to me) the standard of examination is not such as to inspire the confidence of practitioners of the practical sciences and arts.

In reality, the prospect of getting a patent meant that Hooke had (possibly exaggerated) hopes of becoming wealthy from his invention. To do so he needed a patent, to get which he needed the backing of someone with influence in court (where someone in the modern era would need funding to properly file and pay for a patent). To obtain backing he needed partners, but he didn't dare reveal the invention to them because he didn't yet have the protection of a patent – he feared they might steal his idea. His associates were amenable to a plan to contractually organise the joint venture so as to get round that: but he and they could not agree on a contract. The sticking point, in 1664, was the prospect of incremental improvement: they held that, if someone subsequently improved on Hooke's idea, the patent (if obtained) should pass to them; Hooke objected that it is easy to enhance a good idea and he should not lose out when someone did.

Negotiations failed and the plan was scrapped. Hooke decided to keep the details of his watch secret: (at least if it worked) this delayed the introduction of a better watch mechanism by at least a decade. One may blame the uncertainties attached to the patent process (hence arguing for a more stream-lined system): or blame the hope of being granted a patent (thus arguing for the abolition of patents). None the less, he did describe the basic idea in a public lecture in 1664, but no watch maker pursued it. Perhaps the watch-making industry could, given the details of Hooke's clearly inadequately developed watch, have improved it in a short while to make better watches, had he but published the details: but equally, given what happened to some of Hooke's other ideas, he might have published and been ignored (though this might have been due to his publishing, as here, only the basic idea, not the details – a cynic might say he published only enough to establish bragging rights in the event of the idea being made to work). In any case, the tale clearly illustrates that it is one thing to have a nice idea, quite another to put it successfully into practice.

As a long time collaborator with Tompion, Hooke probably did (albeit indirectly) benefit from the flourishing of the English watch industry, though not as much as a patent would have allowed him to. A patent might well have slowed, or distorted, the industry's growth: though, unless the patent was as overly broad as modern ones all too often are, I suspect the industry's innovations in watch design would have rapidly produced watches not covered by the patent.

The case illustrates various issues relevant to consideration of the patent system (or, indeed, anything similar):

Written by Eddy.
$Id: Hooke.html,v 1.2 2005-06-28 12:17:41 eddy Exp $