Surprising though it may seem to those in less well-endowed countries, scientists in the United States have good reason to complain of ageing and overcrowded laboratories, obsolescent instruments and daunting costs in maintaining up-to-date specialized facilities. In a biennial survey conducted across the disciplines last year by the National Science Foundation, fully two-thirds of institutions reported overcrowding. And while they said that nearly one-quarter of their research space needed major renovation or replacement, they also said they had put off over $11 billion in such work for lack of funds.

Given a major decline in defence funding in universities, for example, some suffering in the physical sciences, though regrettable, is not surprising. But in the face of these numbers, and in this era of $2-billion jumps in the budget of the National Institutes of Health, it is startling to consider that this agency's main programme for funding such construction is spending just $30 million this year. Similarly, a shared-instrumentation programme delivers only $35 million for costly instruments not covered by individual research grants. To be fair, the $15.6-billion agency also directs some $3 billion to universities in indirect research costs, some of which can be drawn on for construction and instrumentation. But, especially for smaller and poorer schools, these funds hardly meet the sizeable bill for getting projects off the ground and instruments bought.

Senator Tom Harkin (Democrat, Iowa) has the right idea in a bill that would designate $750 million for the National Institutes of Health to spend on construction and renovation at its grantee institutions in the next two years (see page 621). Individual scientists may worry that this would eat into their piece of a finite pie. But that would be short-sighted in a golden age in which the pie is growing considerably every year. For the promise of biomedical research to be realized, its underpinnings have to be maintained. Congress should back Harkin's proposal, while not losing sight of the needs of other sciences too.