] The next step is to build a real detector and then, if
] that works, to convince the authorities to support
] widespread deployment. A system that scanned every lorry
] coming into the United States would, the researchers
] estimate, cost about $1 billion. That is a lot of money
] in most contexts, but not all that much in the context of
] nuclear defence. The existing missile-defence programme
] has already cost over $130 billion, and is scheduled to
] spend at least another $50 billion in the next five
] years. And unlike missile interceptors, which on February
] 13th spectacularly failed their second test in two
] months, detectors at America's borders should actually
] work. Madness, it seems, may give way to science.