See, actuaries like large numbers. The bigger the group, the more they can read into the statistics. If we had a hundred forty survivors in a group of three hundred, that would have some significance. Fourteen hundred out of three thousand, that would be even more significant. A hundred forty thousand out of three hundred thousand, that would begin to suggest that the sample was composed of people who lived inChernobyl, or whose mothers took DES during pregnancy. It would really set the sirens wailing."

"I see."

"I've had some experience in direct-mail advertising. We tested everything. You have to. If we had a list of half a million names, and we did a test mailing to a thousand of those names, we knew we'd get the same response ratio within a point or two from the entire list. But we knew better than to send out a test mailing to thirty names, because the results wouldn't mean anything."

"Where does that leave you?"

"It leaves me impressed with the percentages, and never mind the size of the sample. I can't get past the fact that statistically we should have suffered four or five deaths and instead we took a hit three or four times as heavy. What do you make of it, Matt?"

I gave it some thought. "I don't know anything about statistics," I said.

"No, but you're an ex-cop and a detective. You must have instincts."

"I suppose I do."

"What do they tell you?"

"To look for special circumstances. You mentioned one man who died inVietnam. Were there any other combat deaths?"

"No, just Jim Severance."

"How about AIDS?"

He shook his head. "We had two gay members, although I don't believe anybody knew they were gay when the chapter was founded. I wonder if that would have made a difference. In 1961? Yes, I'm sure it would have, and when we stood up and recounted the most interesting fact about ourselves at that first meeting, that particular fact went unmentioned.