Benzene is the cat burglar of air pollutants. It is both highly dangerous and difficult to catch, as it travels fast, leaving behind no trace of its presence. The Environmental Protection Agency (EPA) is supposed to police this petroleum product. But it is hard to regulate what you can't find, and the agency knows little about benzene's whereabouts.
''There have been something like 700 individual tests for benzene made in the US, ever,'' says Richard Dowd, a former acting assistant administrator at the EPA. ''That's nothing. Just nothing.''
This is one example of a larger problem: In many ways, government does not do a thorough job of keeping tabs on the state of the environment in the United States, experts say.
Earlier this year, a Conservation Foundation report concluded that US environmental monitoring programs were seriously inadequate. Budget cuts, the foundation said, were continuing to weaken our ability to measure air and water quality.
The problem is particularly acute with regard to toxics - nasty substances such as benzene and mercury that put the ''hazard'' into hazardous waste.
Toxics usually cannot be measured by the extensive network of stations that monitor more traditional pollutants. The US forward-warning system for water pollution, for instance - the National Stream Quality Accounting Network (NASQAN) - can't keep track of most toxic organic chemicals.
In addition, some special programs to watch toxics have disappeared in the wake of recent budget cuts, says William Drayton, a Carter-era EPA official. The agency's experimental, 165-station network for measuring airborne metals has been axed. An EPA effort that surveyed people to determine exposure to various toxics - and whose data helped convince regulators of the dangers of the pesticide DDT - has been greatly weakened, according to Mr. Drayton.
The US Mussel Watch, an EPA program that sent one scientist along the US seacoast to study the concentration of toxics in mussels, lost its federal funds in 1979. It survives in a few states.
There are also problems in some of the toxic measuring efforts that remain. Toxic dump owners, for instance, are supposed to monitor the ground water under their sites, to protect against leaks. But the draft of an internal EPA study indicates this program just isn't working. Eighty percent of affected dump owners are not complying with the law, the study says.
''The amount of monitoring around (hazardous waste) sites is insufficient,'' admits EPA administrator William D. Ruckelshaus. ''The law simply hasn't been implemented.''
Pesticides are another problem area. Although the Food and Drug Administration and other agencies analyze foodstuffs for traces of pesticide chemicals, there is little monitoring of pesticides in the air or water. A 1972 law required the EPA to produce a pesticide monitoring plan; the agency submitted it just this year.
In sum, information now being gathered gives the public ''only a glimpse of (its) exposure to toxic substances,'' according to the Conservation Foundation.
The situation is better, though still somewhat troubled, in respect to ''conventional'' pollutants such as carbon monoxide and dissolved solids.
Conventional air pollutants are measured by a network of some 3,600 monitoring stations around the country. While this network is extensive, its accuracy depends on the proper installation and maintainence of delicate scientific instruments. Five years ago, a General Accounting Office report found 81 percent of air monitors examined had at least one problem that could affect data reliability.
Recently, a survey of some 2,000 monitors found half of them to be in ''fair to poor condition,'' says William Becker, director of the State and Territorial Air Pollution Program Administrators.
The most coherent data on US water quality are provided by NASQAN, a system of river monitors run by the US Geological Survey. Although extensive, NASQAN was ''originally set up to measure stream flow, not pollutants,'' notes Edwin Clark, a senior associate of the Conservation Foundation.
Accordingly, stations are often located in the wild, far from where people actually use water, and are most effective only when measuring pollutants that don't degrade while traveling downstream.
''And for the most part, they don't monitor toxics,'' Mr. Clark says.