The joy of museums such as the Walter Rothschild Zoological Museum at Tring, the Sir John Soane's Museum and the Pitt Rivers, is that they are as much 'museums of the museum' as they are museums of today: preserved reminders of the collections, displays and attitudes of the past.

But is this what we want for all Britain's museums? Is it a good thing that visitors from the 1870s would notice little difference in the National Gallery, if we could transport them forward through time, today?

Because although the building is bigger, thanks to the 1991 Sainsbury extension, the collection - though most visitors don't realise it - is substantially the same.

Charles Saumarez Smith, the gallery's director since 2003, describes the collection as 'largely shaped by what was considered great works of art in the 19th century. So, for example, we don't have anything from the United States, Australia or Russia, because those were not part of the "great tradition".'

Despite high-profile purchases, most recently the Madonna of the Pinks, the fact remains that a combination of sky-high prices for Old Masters, and the dwindling amounts of money in the gallery's acquisitions budgets, means additions are few and far between.

If the National Gallery is an extreme example of a collection that is effectively a representation of Victorian taste, it is not alone. The splendid collections on display in the new World Museum Liverpool - though benefiting from modern interpretation and interactives - are also fundamentally Victorian.

The period of Liverpool's great collectors was the period of its greatest wealth: a reflection of the city's leaders, full of civic pride, enjoying the fruits of trade and empire. (The situation was so different by the 1980s that the city council - far from buying things for its museums -was actively trying to get rid of them. National government was forced to take over funding the museums in 1986.)

Of course, not every local authority has taken the view of that of 1980s Liverpool - and certainly not its current council. But the pages of Museums Journal have been filled with tales of cuts and closures - so it is hardly surprising that funds for purchases are so low.

Figures collected by Museums Journal show that some major hub museums, such as Bradford Museums, have as little as £4,000 a year, which slowly amasses (and of course there are many smaller museums that have nothing at all).

Others have an endowment to fund acquisitions: Leicester City Museums has about £70,000, while the Manchester Art Gallery's endowment has shrunk from £180,000 in 1995 to £135,000 today.

These may not look like insubstantial funds, but prices have soared well beyond the retail price index, not only for art objects, but for a whole range of items, including archaeological pieces and even natural history specimens.

While Renaissance in the Regions barely mentioned acquisitions, its pre-decessor - the 1963 Rosse report - bemoaned the state of acquisition funds. It recommended significant increases in funds such as the V&A Purchase Fund.

But although the fund reached a peak of £1.7m in 1994/5, new research by heritage consultant Adrian Babbidge reveals that by 1999/2000 the fund was back, in real terms, to the 1960 levels of which the Earl of Rosse was so critical. The Prism Fund, designed to help the science sector, has also remained stuck at a mere £250,000.

'People don't like saying there is a crisis around acquisitions, but in my opinion there is one looming,' says Jack Lohman, the director of the Museum of London. 'Of course, curators love objects, and however much money you have, they would go out and spend it, so it is absolutely correct to think carefully about what you are acquiring and why.

But equally, museums are not supposed to remain static: times change, people change. In the Museum of London, for example, we have a major challenge ahead in telling the history of social diversity of this city.'

The situation among national mu-seums is hardly better. The British Museum has cut its budgets from almost £1m ten years ago, to £100,000. The National Gallery's budget has also shrunk from almost £3m to nil, while the National Museum of Science and Industry's (NMSI) budget has reduced from £250,000 to £50,000.

Even allowing for the substantial donations of items to NMSI, it seems hard to square these sums with its remit to collect across the expanding field of science and technology.

Of course what has also happened is that the vast majority of museums rely on the Heritage Lottery Fund (HLF) and the National Art Collections Fund (NACF), which raises money from its members. In the past ten years, the HLF has given almost £130m towards museum acquisition - about a tenth of its total spend on museums.

The NACF, which has provided about £49m in its 100-year existence, currently provides £3m-£4m a year. But neither of these are government funding streams, using taxpayers' money - however much government seems to regard lottery money as a capacious purse.

The remarks, back in 1915, by one of the founders of the NACF that the fund should be used to save out-of-the-way interesting things that might otherwise be missed and not things that state ought to purchase seems prescient.

The NACF is now leading a campaign - Living and Giving - to encourage the Treasury to give tax breaks for lifetime donations, an attempt to reinvigorate a philanthropy that has also declined in the past century.

But while government action is clearly needed if it is to fulfil its obligations to continue collecting the national heritage, government is not entirely to blame.

In the early 1990s, a number of national museums - led by the Natural History Museum and the Science Museum - argued for greater control over the way they spent in grant-in-aid.

In 1992/3 the ring fence that protected acquisition funds was dropped - a move that David Barrie, the director of the NACF, describes as a 'bad day - it is a terrible shame that it was welcomed for a time, but now it is obvious it meant museums had to choose between fixing the roof or buying an artefact'.

Nicholas Serota, the director of the Tate, is equally resigned: 'I had a struggle through much of the 1990s with my trustees to maintain our acquisition budget. I managed it to an extent, but not in real terms.'
Other, more intellectual, factors were also coming into play in the 1990s.

These questioned what was being collected and for who. In 1997, David Fleming - then the director of Tyne and Wear Museums - wrote attacking 'object fetishism'.

This was a wake-up call to those curators who believed in collecting for collecting's sake, and reflected the views of curators trying to make museums responsive to their communities. In the art and academic world, things were changing too.

'Approaches to art history put a premium on ideas rather than objects,' Serota recalls. 'Academics started to feel that the collection was more interesting than the individual, the theme more interesting than a single career. All of this ties back to the social context: it was right to welcome some of these developments but it has had the unfortunate side effect of discouraging collecting as a skill.'

Helen Wilkinson, the author of the Museums Association report Collections for the Future, says: 'The debate about collections must move beyond an inaccurate dichotomy… too many people in museums still make the false assumption that to place emphasis on the importance of collections is to reject modernising and progressive initiatives.'

And there's the rub: if museum and gallery staff believe in the power of collections to inspire and shape people's lives, and believe they can display and interpret them in ways that will draw in more people from wider sections of society, then they cannot abnegate their responsibilities to create dynamic collections.

Society is in constant flux, it cannot be represented - like recreated Victorian street scenes - by collections preserved in aspic.

Jane Morris is a freelance journalist