But there is another lesson of the 1930s. It is that although capitalism survives it is capable of retreating behind a protectionist shell, at great cost to global prosperity. This is a real danger today. The "Buy American" provisions in President Barack Obama's fiscal boost are an ominous sign. The impulse to resort to protection when economic hardship suddenly strikes is, of course, always present. But there is today a dangerous new factor which magnifies the threat. The leaders of some of America's largest corporations have already joined up with organised labour (the AFL-CIO) to urge Congress to impose tariffs against imports from countries (such as China, for example) which are understandably unwilling to bear the heavy costs of an obligation to curb their carbon dioxide emissions. There is considerable support in Europe, notably within the European Commission and in France, for a similar approach.

It is essential, both in the US and in Europe, that this is resolutely rejected. The first and most important requirement for the future of capitalism is the preservation of globalisation, and the massive benefits it confers on mankind, in particular in the developing world. There are, inevitably, costs of globalisation; but they are hugely outweighed by the benefits. So resistance to protection, whatever arguments may be used in its favour, must be rigorously maintained. Nor is this an exclusively economic argument. It is a moral imperative, as well. Moreover, a trade war with China could well have unpredictable, and potentially highly damaging, political consequences.

But will capitalism need to change in the future? Again, the lesson of history is that the answer is "not really". The economic cycle is endemic and inescapable, and everyone (with the exception of prime minister Gordon Brown) has always known this. What the current crisis does underline, however, is that a cyclical downturn associated with a collapse of the banking system is by an order of magnitude worse than a normal cyclical downturn.

So there does need to be a change to the banking system. In a nutshell, we need to return, in all major financial centres, to the separation of commercial banking from investment banking that was enforced in the US under the 1933 Glass-Steagall Act, until it was repealed by President Bill Clinton in the 1990s. This is all the more important since we now live in an age in which the acquisition of wealth appears to count for more than reputation.

Achieving this will not be easy or popular in banking circles, but it can be done. We have time to get it right: this is not firefighting, but fireproofing.

The overriding reason why this separation is essential is straightforward. It is only a commercial banking crisis that poses a systemic risk and can lead to the sort of mess we face today. It is folly to allow core banks to be in a position where they can be brought down by exciting but highly risky investment banking activities. But the idea that this can be prevented by judicious regulation of investment banking activities is a chimaera. In the real world, that is not possible: either the investment bankers will outsmart the regulators, or the regulators will respond with damaging overkill.

Thus investment banks should be left to their own creative devices, and subject essentially only to the discipline of the marketplace. This leaves a much more limited, and practicable, but still absolutely essential, role for bank supervision and regulation: namely, to ensure that the core commercial banking system is thoroughly sound and adequately capitalised at all times.

It is worth adding that it is the capital adequacy regime, and not primarily interest rate policy, which needs to be responsive to asset-price bubbles.

What else (other than the maintenance of what passes for world peace) is needed to ensure that capitalism survives (as it will) and prospers (as it should)?

There is a danger in many parts of the world, and certainly in the UK, to imagine that since this is a global problem it requires a global solution, so the overriding need is for a global agreement. This may sound statesmanlike, but it is in fact a dangerous delusion. The overriding need is for the authorities in each country to put their own house in order.

The threat from terrorism is an instructive parallel. Terrorism is indeed a global problem, and international co-operation is clearly desirable. But that in no way diminishes the overriding duty of national governments to do what is necessary to protect their own people.

The same applies to financial regulation. As the Basel II bank capital rules clearly showed, international agreement is slow in arriving and, when it does arrive, it is likely to prove inadequate. As far as the UK is concerned, Mr Brown's decision, as chancellor, to scrap the strengthened system of bank supervision I put in place in 1987 and replace it with a system that has proved largely dysfunctional was not very clever. Without waiting for global agreement, however desirable that may be, we need to, and can, do a great deal better.