*I*

The Lauriston lies to the buoy with her head northwest, which is not a heading Renn asked for. It is the one the river has given him.

He takes the chimney at Vauxhall with her standard compass and writes down what the compass says. The chimney's bearing from this reach is painted in white on the dock wall below it. The mark was put up before Renn was in the trade and has gone thin at the edges since. The difference between the wall and the compass is his whole morning. On this heading her compass is out 15 degrees westerly.

He writes it against northwest and waits for the tide.

She has been afloat since April and the paint in her cabin is still soft enough to smell. An iron ship is built on a slip that lies one way, and she is hammered there for the better part of a year, and she keeps what the hammering puts into her. The Lauriston's head lay north-northeast on the stocks at Whiteinch. Renn knows how those berths lie and roughly what comes off them, and the figure he has written is close to the one he had in his head coming out in the boat.

The flood slackens. She starts to come round on the buoy, not quickly and not evenly, and he takes the chimney again each time she gives him something to take it on. He gets a heading a little east of north, then north by east, then a long wait at northeast while the tide makes up its mind, then a run of them close together as she goes through east. The book has him doing this on 32 equidistant points. He does it on what the tide gives him, and writes the time beside each figure so that later he can see where she hurried.

By slack water he has 17 observations, unevenly spaced, covering rather less than half the compass, and a sheet of paper with a bad shape on it.

He plots them on a Napier's diagram in the deckhouse with his coat still on. The diagram exists for this: a ruled sheet on which headings that came in crooked can still be made to lie on a curve, and the curve then gives him the headings he never observed. It is a clean curve. That is the first bad news.

Her semicircular deviation runs to 23 degrees easterly a little south of east, and falls away and reverses through the compass as it should. There is some quadrantal in her as well, but she carries spheres on her binnacle brackets and they are doing what they are for.

He sits with the sheet on his knee. He does not need to work it out to know what it is.

II

The 23 degrees are made of two things that look identical on his sheet and behave nothing alike.

Part of it is her own iron, put into her at Whiteinch, and it will be very nearly the same off Rangoon as it is off the landing stage. The rest of it is not in her at all. It is induced in her upright iron by the earth pulling downward, and it is only ever as strong as that pull is where she happens to be floating. In the Mersey the pull is strong. At the magnetic equator there is none of it. Past the equator it runs the other way.

He can take all 23 degrees out this morning with magnets in the binnacle trays. An hour's work. The card would come out zero on every heading, and any man in that dock who looked at it would see a job perfectly done.

The magnets would still be exactly what they are when she is off Brazil. The other part would not. It would fall away as she went south, and by the time she was on the Line there would be nothing left of it, and the magnets he set this morning would be pulling her needle 9 or 10 degrees off with nothing to pull against. Past the Line it grows. She crosses the magnetic equator twice going out and twice coming home.

Or he can fill the tube.

The tube is already on the front of her binnacle. It came with the binnacle, Thomson's new pattern, bought when she was fitted out: brass, spheres on the brackets, a place under the bowl for the heeling magnet, and on the fore side a vertical tube. There is nothing in the tube. The maker sold them the fitting and nobody has put anything into it.

What goes in it is upright soft iron, the same as her own upright iron. It takes its strength from the same downward pull, it weakens where that pull weakens, and it dies at the Line at the rate her error dies. It follows her south. Nothing else on that binnacle does.

The cost of it is this. It does not take out the whole 23. It takes out the part that travels, and he corrects what is left with magnets, fewer of them, and the card goes aboard with figures on it.

Then there is the thing he cannot do. He sits with that a while in the deckhouse. To cut the bar to length he has to know how much of the 23 is hers and how much is the earth's. There is no observation he can make in this river that separates them. Here they point the same way and add together. To take them apart he would have to swing her again where the earth pulls differently: well south, months from now, by somebody else.

So he estimates it, from the yard and her build and the ships he has had off those berths before. He is good at this. It is still an estimate, and the length of iron he cuts is that estimate made solid.

III

Doig has the sheet flat on the cabin table and is reading it properly, which not every master does. Hesketh is at the end of the table with the ship's papers and has not touched the sheet at all.

Renn tells him what is in the 23 and what happens to each part of it between here and Rangoon. Doig follows it the first time.

"So you want to leave error on my card," Doig says.

"I want to leave error on your card that gets smaller as you go south instead of larger."

"How much?"

"Five, at the worst heading. Less by the time you're below the Canaries. Not much of anything by the Line."

"Against nothing at all today."

"Against nine or ten at the Line, the other way, and growing after."

Doig turns the sheet round and looks at the curve upside down for a moment. Renn has seen good men do that. Then he pushes it back.

"I have two mates and men who steer," he says. "On a card of zeros there is nothing for them to get wrong. On a card of figures a tired man applies the deviation the wrong way and doubles it instead of taking it off. I have seen a ship put on the wrong side of a light by a mate who was quite sure which way to add."

"You have," Renn says. "I can't answer that. It's true."

Hesketh has been waiting to be finished with. "She goes out at noon or she goes out at nine tonight," he says. "That is what the thing costs. The iron is nothing."

"The iron is four pounds and some for the brackets."

"The iron is nothing," Hesketh says again, not unpleasantly. He has a pen out. "What is it called?"

"A Flinders bar."

"Spell it."

Renn spells it and Hesketh writes it into the account, and it is the first time Renn has seen anybody write it down.

"We have paid for the binnacle," Hesketh says.

"You have paid for the tube. There is nothing in it."

"Sowerby would have her done by four."

"He would," Renn says. "He would be doing what he was asked."

Nobody answers that. Doig has gone back to the sheet.

"And if you have the length wrong?" Hesketh says.

"Then she carries less error south than she would have. Or a little more."

"Which?"

"I will not know which."

Hesketh looks at him for a second and then writes the figure into the account under the spelled word.

Doig makes up his mind in the manner of a man who settled it some minutes ago. "Fit it," he says, and then, to Hesketh rather than to Renn: "It costs me nothing to carry it."

Then he says the other thing, flatly, because it is not a concession and he does not intend it as one.

"Understand me. I will swing her myself by the sun when we have weather for it, and again before the Line, and what I get out of that is what I will steer by. If your bar is right my table will say so and I will not think about it again. If it is wrong I will have my own figures by then anyway."

"That is what you should do," Renn says, and means it, because it is.

IV

The iron comes off in the shop boat at two.

He lays the segments out along the deck by the binnacle: soft iron in lengths, and wooden distance pieces to go beneath them so the top of the bar stands where it must stand against the compass bowl. The tube takes what he gives it. What he gives it is the whole of the morning's judgment in one dimension.

He has it between 15 and 18 inches. Eighteen if the earth's share of the 23 is what he thinks it is. Fifteen if he is a little over on it.

He takes 15.

The reasoning is short and he has had it since the deckhouse. If he cuts long he corrects more of the induced part than is there, and past the Line an over-correction turns round and works against her, and goes on working against her the whole homeward passage. If he cuts short he leaves a little of it in her, and what he leaves shrinks all the way down. If he has guessed wrong there is no length that is right in both hemispheres, and 15 is the one that is wrong more gently.

He sets the segments, packs the wood beneath them, and puts the cap on the tube.

From the deck the binnacle looks exactly as it looked this morning.

The magnets are ordinary work and go quickly. Then he waits for the evening flood to bring her round again so that he can see what he has done.

She comes round from the other side this time and gives him headings the morning did not, and he takes the chimney at each of them until the daylight goes off the dock wall and he has to work by the lamps. By the end of it he has enough for the second curve.

Worst heading, 5 easterly. The curve is flat and low and there is nothing in it that will surprise anybody.

*V*

He writes the card in the deckhouse.

It is the shop's printed form: the ship's name at the head, the date under it, deviations against the points of the compass, east and west set out in their columns. He reads the figures off the Napier's diagram for the points he never had and enters them alongside the ones he did, and nothing on the form distinguishes the two. Nobody has ever asked him to.

There is no certificate. There is nothing to swear to and nobody to swear it to. The form is his shop's, the figures are his, and Doig will sign the book for the delivery of it the way a man signs for a thing he has paid for.

He puts it in the frame at the standard compass and stands looking at it a moment longer than he needs to.

What is on the card is true this evening. Her sub-permanent will go on shifting for weeks yet, fastest at the start, and some of these figures will be a degree out by the time she is past Ireland.

He writes the date on it.

Doig signs the book without reading it, hands it back, says goodnight and goes forward. Hesketh has already gone ashore with the account, with the word spelled out in it and the figure under the word.

VI

The tug has her a little before nine.

Renn is on the landing stage with his bag between his feet. It is not quite dark. The river is doing most of the turning; the tug only keeps her from going too far. She comes round slowly and steadies.

He brings her two masts into line and takes the bearing off the pocket compass, because his hands do it before he decides to.

341.

There is a figure against that heading on the card in her binnacle and he knows what it is without looking it up. He holds her while she stands on, and the angle does not change.

Then the Woodside boat comes across on its half hour and puts itself between them.