"First, two assumptions. One: the natural unit of time on any world is the day—that is, the time it takes the planet to rotate on its axis . . ."

"Assuming it rotates," somebody tossed in.

"That was my second assumption. But the only cases we know of where there's no rotation—or where the orbital period equals the axial period, which amounts to the same thing—occur when a small body orbits close to a far more massive one and is swamped by gravitational tidal effects, like our Moon. For that to happen to a body the size of a planet, the planet would have to orbit very close to its parent star—too close for it to support any life comparable to our own."

"Seems reasonable," Caldwell said, looking around the table. Various heads were nodding assent. "Where do we go from there?"

"Okay," Hunt resumed. "Assuming it rotates and the day is its natural unit of time—if this complete table represents one full orbit around its sun, there are seventeen hundred days in its year, one entry for each."

"Pretty long," someone hazarded.

"To us, yes: at least, the year-to-day ratio is big. It could mean the orbit is large, the rotational period short, or perhaps a bit of both. Now look at the major number groups—the ones tagged with the heavy alphabetic labels. There are forty-seven of them. Most contain thirty-six numbers, but nine of them have thirty-seven—the first, sixth, twelfth, eighteenth, twenty-fourth, thirtieth, thirty-sixth, forty-second, and forty-seventh. That seems a bit odd at first sight, but so would our system to someone unfamiliar with it. It suggests that maybe somebody had to do a bit of fiddling with it to make it work."

"Mmm . . . like with our months."

"Exactly. This is just the sort of juggling you have to do to get a sensible fit of our months into our year. It happens because there's no simple relationship between the orbital periods of planet and satellite; there's no reason why there should be. I'm guessing that if this is a calendar that relates to some other planet, then the reason for this odd mix of thirty-sixes and thirty-sevens is the same as the one that causes problems with our calendar: That planet had a moon,"

"So these groups are months," Caldwell stated.

"If it's a calendar—yes. Each group is divided into three subgroups—weeks, if you like. Normally there are twelve days in each, but there are nine long months, in which the middle week has thirteen days."

Danchekker looked for a long time at the sheet of paper, an expression of pained disbelief spreading slowly across his face.

"Are you proposing this as a serious scientific theory?" he queried in a strained voice.