Anyone who thought the human body had reached its limits will have been thinking again after witnessing young Ye Shiwen's extraordinary - and controversial - Olympic swim last weekend. The 16 year-old took gold in the women's 400m medley, shaving a full five seconds off her personal best time. Most pertinently, Ye broke the world record by more than a second and her final 50m was faster than that of the men's gold medallist, the great American hope, Ryan Lochte.
In its way her performance was as impressive as that of the extraordinary Usain Bolt. In the Beijing Olympics the fastest man on Earth smashed his own 100m world record and then, a year later, at the World Athletics Championships in Berlin he demolished the time he had set in China - 9.69 seconds - by more than a tenth of a second. To put this in context, it had taken 10 years for the record to fall by a similar amount, and it was more than 30 years before that that the 10-second barrier was broken for the first time.
This all begs the question, how fast, how far can the human body go? In the absence of genetic manipulation and, of course, illegal doping of the conventional chemical or genetic (see below) kind, is there a limit to how far the human physique can be pushed?
Because with records tumbling by such large increments, the suggestion is that there is some way to go. Might we see a 100m below nine seconds? Might Ye Shiwen knock another second or two off her own extraordinary time? Indeed, are there any limits?
The story of ever-improving human performance is not a simple one. In some sports, progress appears to have stalled, if not halted altogether. The most extraordinary example of a long-standing record was that set by Bob Beamon, an American long-jumper whose leap of 8.90m, set in Mexico in 1968, was not bettered until 1991 by Mike Powell. And his jump has not been beaten since.
Sebastian Coe's 800m record time of 1:41.73, in Florence in 1981, was not beaten until 1997 and the current record is just seven tenths of a second faster than the time Coe posted 30 years ago. Meanwhile, Michael Johnson's 400m record has stood since 1999.
Perhaps the most stubborn of the current "frozen records" is that of the late Florence Griffith-Joyner in July 1988, who set a new world 100m best of 10.49 seconds, a time that no woman has come within a seventh of a second of since. (Known as Flo-Jo, her career was also dogged by accusations of drug-taking, although nothing was ever proven.)
But in other categories the times keep tumbling, notably in the sprints. The men's marathon record too, currently standing at 2:03.38 looks ripe for attack, although no woman has come close to breaking Paula Radcliffe's time of 2:15.25 for nine years.
There are several reasons why athletic performance has, in general, made a steady improvement since organised and properly timed competitions began more than a century ago. Training is immeasurably better. Coaches now have a wealth of data at their disposal, from physiology to diet and metabolism. Today's professional athletes train far longer, harder and more effectively than those who competed when London hosted the Games in 1948.
General levels of nutrition are better, as is the kit. This makes a lot more difference in some sports (such as cycling) than it does in "purer" fields of endeavour such as road running. But better shoes, more aerodynamic clothing and scientifically designed track surfaces have probably been responsible for shaving a few tenths of a second off, say, the 100m and 200m sprint times.
The most important reason why today's athletes are running faster, jumping longer and throwing further than those of a century ago, however, is that there is simply a much bigger pool of people from which they can be selected.
A century ago the world's population stood at 1.8 billion; today it is nearly four times greater. Until recently the vast pool of potential athletic talent in the developing world simply had no access to the training and competition opportunities that athletes from countries like Jamaica, Kenya and Ethiopia have today.
But can this continue forever? "The rate of improvement will definitely start to decrease," says Prof John Brewer, a sports scientist at the University of Bedfordshire and author of the London 2012 Training Guide to Track Athletics. "And records will be broken by smaller and smaller margins."
No "unmodified" human is ever going to run 100m in eight seconds, he says, or a marathon in 90 minutes, but that doesn't mean there is not room for some improvement. Events in which women have only been allowed to compete relatively recently, for instance, such as weightlifting, may see records set on a regular and frequent basis for some time as competitors flood into the sport.
Four years ago, a Stanford University study compared the running performance of humans and that of two other species: dogs and horses (which, like us, take part in races). The study found that speeds among the top greyhounds and racehorses plateaued in the Seventies, whereas human males continued to get faster. The conclusion of the study was that the fastest a human will ever run 100m is 9.48 seconds.
Prof Brewer is perhaps more optimistic, expecting that we will "certainly" see a marathon run in below two hours some time in the next 10 to 15 years; the 9.5-second 100m barrier will be broken, "but it will take at least 100 years to take off the next half a second". For the 800m race we can expect a time below 1:40 before the end of the decade.
But from then on we can expect a record drought, with long-standing best-evers such as Flo-Jo becoming increasingly common. This will cause a problem for event organisers and, especially, the sponsors, who need a constant drip-feed of record-breakers to stimulate interest.
The solution, says Prof Brewer, will probably be a switch to timings measured in thousandths-of-a-second, replacing the 0.01 second increments used today (which in turn replaced the tenth-second measures used until the late Seventies). Whether a record "broken" by this amount is meaningful is open to question; at sprint speed, 0.001 second is equivalent to the time it takes for a crease in an athlete's bib to cross the line before its wearer.
A confounding factor is the vexed issue of drugs. Many of the medals won in the Eighties, for instance, now look tarnished after a flood of revelations that followed the break-up of the Soviet empire and in particular the dismantling of the sporting behemoth that was East Germany, in terms of population probably the most successful sporting nation in history.
Prof Brewer is diplomatic, saying that "we can speculate there were some records set at a time when the drug testing regimes were not as stringent as they are today". The IOC has rejected calls for a wholesale reallocation of medals awarded in the 1976 and 1980 Olympics, probably the "dirtiest" in modern times, calling the whole issue of retrospective doping penalties a "bottomless pit".
The war on doping is more of an arms race than an actual conflict; the dopers are usually a couple of years ahead of the testers, which is why blood samples are now kept for many years in case a test is developed that can retrospectively catch cheats. The big new fear is genetic doping - the injection of altered DNA, smuggled into the body using viruses (a similar technique to that used in gene therapy). In some animal studies it has been possible to increase muscle bulk and improve stamina by tinkering with genes, although there is no hard evidence that this has been used in humans.
Ultimately, when we face an era in which records stand for half a century or more, athletes will face a choice.
The desperation for perfection means athletes are pushing their bodies way beyond what most of us would consider to be the limit. Witness the naturally extraordinary musclebound physiques sported by some of the athletes in London this week - the weird legs of German cyclist Robert Forstemann, the "water wing" musculature of Italian swimmer Fabio Scozzoli, and Michael Phelps's almost alien bodyshape.
It takes thousands of years for evolution to create faster, better humans. So in the meantime, with such lucrative rewards on offer, the temptation will be for a handful of athletes to go down some very dark alleyways indeed.
Athletes who are naturally adapted for sporting greatness
Phelps's body is extraordinary. His arm span (6ft 7in) is greater than his height (6ft 4in), giving him a valuable competitive edge. His size 14 feet are more like flippers, his large arms and hands work as propulsive paddles, and his thin hips, flat back, long torso and short legs allow for perfect balance in the water.
At 6ft 5in, the Jamaican sprinter's height gives him a huge advantage over his rivals, as taller, leaner athletes excel in short-distance events like the 100m. And while most people have 50 per cent "fast-twitch" muscle fibres (quick contracting, easily fatigued muscle tissue that generates high power), Bolt has a total of 80 per cent.
Sir Steve Redgrave
In his sporting prime, Sir Steve Redgrave stood at 6ft 5in and weighed more than 100kg (around 151/2?st). His broad shoulders and long arms helped him excel as a rower. Though suffering ulcerative colitis, and diabetes mellitus type 1, Redgrave's formidable strength allowed him to compete until the 2000 Sydney Olympics.
The 15-year-old artistic gymnast is only 4ft 10in. Weighing in at 35kg (5st 5lb), she is the smallest and youngest member of Team GB and is fondly referred to as "Twiggy". Tunney's slight frame allows her to change positions quickly while on the high bars. Thankfully, her "rather large kneecaps" haven't provided any unfair advantages so far.
Wiggins's large heart sets him apart from other cyclists, as it allows more oxygen-rich blood to reach his muscles. An intense training regime encourages his muscles to use this extra oxgen efficiently. But it's a genetic trait that gives Wiggins his most lethal advantage - his lung capacity, inherited from his cyclist father, is much higher than normal.