There are ""lies, damn lies, and statistics,"" Norris writes (quoting Disraeli); and the statistics on the safety of air-travel are just that--damned lies. Though the record shows very few accidents (in 1978, 99.99991 percent of US airline flights were completed without fatality), there are many unreported near-disasters--and the accidents that do occur, journalist Norris argues, are the result of lax safety standards, gross error, and deliberate, criminal secrecy. Yet he is not entirely convincing. He supports his charges with a veritable encyclopedia of disasters narrated in head-shaking prose sprinkled with such phrases as ""one-way ticket to the grave."" True, he shows that most accidents are due to error or negligence, and in hindsight could have been prevented. But he fails to reveal patterns of negligence. He merely proves what we all know: that the human beings who design, build, maintain, fly, and control aircraft sometimes make stupid mistakes and try to hide them. The safety procedures and regulations that attempt to root out human error and dishonesty do not eliminate all possibility of accident; but they are among the strictest of their kind (in contrast, for example, to controls in the nuclear power industry)--and Norris' occasional suggestions for improvement center on the benefits of certain British safeguards, not wholesale reforms. Here and there he does pinpoint legitimate cases of criminal folly: the DC10 cargo door scandal; the flagrant disregard, by foreign controllers, of the rule to speak only English; the failure to build new planes out of expensive, fire-resistant materials. Many of his charges, on the other hand, are exceedingly thin (e.g., that airlines are flying with less than a safe margin of fuel--to lighten weight and thus save fuel). The scary stories, however, keep the pages turning--and call to mind again, the headline accidents that did occur.