yet, even so, hari seldon could not repress the surge of satisfaction that he felt as he entered his laboratory.

how things had changed.

it had begun twenty years earlier with his own doodlings on his second-rate heliconian computer. it was then that the first hint of what was to become parachaotic math came to him in a cloudy fashion.

then there were the years at streeling university, when he and yugo amaryl, working together, attempted to renormalize the equations, get rid of the inconvenient infinities, and find a way around the worst of the chaotic effects. they made very little progress, indeed.

but now, after ten years as first minister, he had a whole floor of the latest computers and a whole staff of people working on a large variety of problems.

of necessity, none of his staff-except for yugo and himself, of course-could really know much more than the immediate problem they were dealing with. each of them worked with only a small ravine or outcropping on the gigantic mountain range of psychohistory that only seldon and amaryl could see as a mountain range-and even they could see it only dimly, its peaks hidden in clouds, its slopes veiled by mist.

dors venabili was right, of course. he would have to begin initiating his people into the entire mystery. the technique was getting well beyond what only two men could handle. and seldon was aging. even if he could look forward to some additional decades, the years of his most fruitful breakthroughs were surely behind him.

even amaryl would be thirty-nine within a month and, though that was still young, it was perhaps not overly young for a mathematician-and he had been working on the problem almost as long as seldon himself. his capacity for new and tangential thinking might be dwindling, too.

amaryl had seen him enter and was now approaching. seldon watched him fondly. amaryl was as much a dahlite as seldon's foster son, raych, was, and yet amaryl, despite his muscular physique and short stature, did not seem dahlite at all. he lacked the mustache, he lacked the accent, he lacked, it would seem, dahlite consciousness of any kind. he had even been impervious to the lure of jo-jo joranum, who had appealed so thoroughly to the people of dahl.

it was as though amaryl recognized no sectoral patriotism, no planetary patriotism, not even imperial patriotism. he belonged-completely and entirely-to psychohistory.

seldon felt a twinge of insufficiency. he himself remained conscious of his first two decades on helicon and there was no way he could keep from thinking of himself as a heliconian. he wondered if that consciousness was not sure to betray him by causing him to skew his thinking about psychohistory. ideally, to use psychohistory properly, one should be above worlds and sectors and deal only with humanity in the faceless abstract-and this was what amaryl did.

and seldon didn't, he admitted to himself, sighing silently.

amaryl said, "we are making progress, hari, i suppose."

"you suppose, yugo? merely suppose?"

"i don't want to jump into outer space without a suit." he said this quite seriously (he did not have much of a sense of humor, seldon knew) and they moved into their private office. it was small, but it was also well shielded.

amaryl sat down and crossed his legs. he said, "your latest scheme for getting around chaos may be working in part-at the cost of sharpness, of course."

"of course. what we gain in the straightaway, we lose in the roundabouts. that's the way the universe works. we've just got to fool it somehow."

"we've fooled it a little bit. it's like looking through frosted glass."

"better than the years we spent trying to look through lead."

amaryl muttered something to himself, then said, "we can catch glimmers of light and dark."

"explain!"

"i can't, but i have the prime radiant, which i've been working on like a-a-"

"try lamec. that's an animal-a beast of burden-we have on helicon. it doesn't exist on trantor."

"if the lamec works hard, then that is what my work on the prime radiant has been like."

he pressed the security keypad on his desk and a drawer unsealed and slid open noiselessly. he took out a dark opaque cube that seldon scrutinized with interest. seldon himself had worked out the prime radiant抯 circuitry, but amaryl had put it together-a clever man with his hands was amaryl.

the room darkened and equations and relationships shimmered in the air. numbers spread out beneath them, hovering just above the desk surface, as if suspended by invisible marionette strings.

seldon said, "wonderful. someday, if we live long enough, we'll have the prime radiant produce a river of mathematical symbolism that will chart past and future history. in it we can find currents and rivulets and work out ways of changing them in order to make them follow other currents and rivulets that we would prefer."

"yes," said amaryl dryly, "if we can manage to live with the knowledge that the actions we take, which we will mean for the best, may turn out to be for the worst."

"believe me, yugo, i never go to bed at night without that particular thought gnawing at me. still, we haven't come to it yet. all we have is this -which, as you say, is no more than seeing light and dark fuzzily through frosted glass."

"true enough."

"and what is it you think you see, yugo?" seldon watched amaryl closely, a little grimly. he was gaining weight, getting just a bit pudgy. he spent too much time bent over the computers (and now over the prime radiant)-and not enough in physical activity. and, though he saw a woman now and then, seldon knew, he had never married. a mistake! even a workaholic is forced to take time off to satisfy a mate, to take care of the needs of children.

seldon thought of his own still-trim figure and of the manner in which dors strove to make him keep it that way.

amaryl said, "what do i see? the empire is in trouble."

"the empire is always in trouble."

"yes, but it's more specific. there's a possibility that we may have trouble at the center."

"at trantor?"

"i presume. or at the periphery. either there will be a bad situation here-perhaps civil war-or the outlying outer worlds will begin to break away."

"surely it doesn't take psychohistory to point out these possibilities."

"the interesting thing is that there seems a mutual exclusivity. one or the other. the likelihood of both together is very small. here! look! it's your own mathematics. observe!"

they bent over the prime radiant display for a long time.

seldon said finally, "i fail to see why the two should be mutually exclusive."

"so do i, hari, but where's the value of psychohistory if it shows us only what we would see anyway? this is showing us something we wouldn't see. what it doesn't show us is, first, which alternative is better, and second, what to do to make the better come to pass and depress the possibility of the worse."

seldon pursed his lips, then said slowly, "i can tell you which alternative is preferable. let the periphery go and keep trantor."

"really?"

"no question. we must keep trantor stable, if for no other reason than that we're here."

"surely our own comfort isn't the decisive point."

"no, but psychohistory is. what good will it do us to keep the periphery intact if conditions on trantor force us to stop work on psychohistory? i don't say that we'll be killed, but we may be unable to work. the development of psychohistory is on what our fate will depend. as for the empire, if the periphery secedes it will only begin a disintegration that may take a long time to reach the core."

"even if you're right, hari, what do we do to keep trantor stable?"

"to begin with, we have to think about it."

a silence fell between them and then seldon said, "thinking doesn't make me happy. what if the empire is altogether on the wrong track and has been for all its history? i think of that every time i talk to gruber."

"who's gruber?"

"mandell gruber. a gardener."

"oh. the one who came running up with the rake to rescue you at the time of the assassination attempt?"

"yes. i've always been grateful to him for that. he had only a rake against possibly other conspirators with blasters. that's loyalty. anyhow, talking to him is like a breath of fresh air. i can't spend all my time talking to court officials and to psychohistorians."

"thank you."

"come! you know what i mean. gruber likes the open. he wants the wind and the rain and the biting cold and everything else that raw weather can bring to him. i miss it myself sometimes."

"i don't. i wouldn't care if i never go out there."

"you were brought up under the dome-but suppose the empire consisted of simple unindustrialized worlds, living by herding and farming, with thin populations and empty spaces. wouldn't we all be better off?"

"it sounds horrible to me."

"i found some spare time to check it as best i could. it seems to me it's a case of unstable equilibrium. a thinly populated world of the type i describe either grows moribund and impoverished, falling off into an uncultured near-animal level-or it industrializes. it is standing on a narrow point and topples over in either direction and, as it just so happens, almost every world in the galaxy has fallen over into industrialization."

"because that's better."

"maybe. but it can't continue forever. we're watching the results of the overtoppling now. the empire cannot exist for much longer because it has-it has overheated. i can't think of any other expression. what will follow we don't know. if, through psychohistory, we manage to prevent the fall or, more likely, force a recovery after the fall, is that merely to ensure another period of overheating? is that the only future humanity has, to push the boulder, like sisyphus, up to the top of a hill, only to see it roll to the bottom again?"

"who's sisyphus?"

"a character in a primitive myth. yugo, you must do more reading."

amaryl shrugged. "so i can learn about sisyphus? not important. perhaps psychohistory will show us a path to an entirely new society, one altogether different from anything we have seen, one that would be stable and desirable."

"i hope so," sighed seldon. "i hope so, but there's no sign of it yet. for the near future, we will just have to labor to let the periphery go. that will mark the beginning of the fall of the galactic empire."