Smith tapped the table impatiently. "I'm quite aware of the distances across the Atlantic Ocean, Lieutenant."

"Ah, but are we really concerned with distances, Mr Smith?" Ramage's voice was bantering now, and Smith wondered hurriedly why there had been such emphasis on the word. "We are trying to find something in the Atlantic Ocean, Mr Smith, so aren't we really concerned with areas?"

When the Postmaster nodded warily, Ramage said: "I hope you'll take my word for it that, because of the uncertain direction of the wind, a packet could sail some 250 miles either side of the regular route. In other words a packet on its way from Falmouth to Barbados could be lost in a rectangle measuring 4,200 miles by 500 miles. That" - he glanced at the paper - "is an area of more than two million square miles."

Smith said nothing.

"We'll ignore the leg from Barbados to Antigua, and say that for the 900 miles from Antigua to Jamaica the packet could be twenty-five miles either side of the direct course," Ramage continued. "A rectangle 900 miles by fifty comprises 45,000 square miles."

Smith was now jotting down the figures, and Ramage paused for a moment. When he saw the Postmaster had stopped writing he said: "Now for the voyage home from Jamaica. It's roughly 3,750 miles from the Windward Passage to Falmouth, and allowing the 250 miles either side of the direct route gives us nearly two million square miles. For the round voyage, Falmouth, Barbados, Antigua, Jamaica and back to Falmouth, we get" - he glanced at the paper again - "a total of more than four million square miles. Four million and twenty thousand, to be exact," he added: Smith was a man who would like exact figures. He waited while Smith wrote them down.

"Now, in good conditions a lookout at sea might sight a ship at ten miles - it'd be unlikely, but I'll be generous with the figures. That means he is looking from the centre of a circle twenty miles in diameter and scanning an area of about 300 square miles. Since a packet can be lost anywhere in more than four million square miles of ocean, I admit it's only of academic interest to divide it by the 300 covered by the lookout, but - I'm using the precise figures now - the answer is 13,400. Tell me, Mr Smith," Ramage said quietly, "where would you start your investigation?"

Smith smiled amiably, already regretting his sharp tongue. The lad was right, and he was prepared to admit it. "I'd start right here, Lieutenant, sitting in that very chair and asking me questions!"

As if to emphasize his change of heart he removed three paperweights and put the "Outward Bound - Lost" and "Inward Bound - Lost" piles of paper in front of him, with "Auckland" on top. Tapping them, he said: "All that's known about the losses is written here."

"Yes," Ramage said gently, "but first I want to know how the Packet Service functions."

The question still puzzled Smith: his whole life had been so wrapped up in the Service he could neither credit that there could be people who knew nothing about it nor really know how to begin describing it.

"The packets," Ramage prompted. "How long does the average passage take?"

"Forty-five days out to Jamaica, via Barbados, and thirty-five days back, sailing direct."

"Who actually owns the ships?"

"They're owned individually, usually by the commanders - by the commanders and their business associates, in some cases."

"And the Post Office charters them?"

"Yes, Lombard Street hires the ships."

"And the crews?"

"They are employed and paid by the Post Office: the commander and the entire ship's company."

"Even if the commander owns the ship?"

"Yes, he's paid a monthly wage as well as the charter fee."

"Who stands to lose if a packet is captured? Or pays for repairs if she's damaged in action?"

"The Post Office pays. Of course, the conditions are all set down in the original charter agreement, but in effect the Post Office carries the insurance."

"How many packets serve the West Indies?"

"Normally there are sixteen - that's the number needed to maintain a regular fortnightly service."

"And the losses in the whole war so far?"

"Thirty-two. Not all of them bound to or from the West Indies, of course. Twelve were lost in the first four years of the war. After that there was a lull, although towards the end of '97 three more were lost in a month. Then losses were irregular - until this year. We've lost nine so far, all West Indies packets."

"Where do the replacement packets come from?"

"Several new packets are building to Post Office specifications," Smith said, "but we are having to hire temporary vessels to make sure we have ten available."

"This year's losses - you have details?"

Smith sorted through his pile of papers and extracted one sheet.

"Here's the list."

Ramage saw that the Princess Royal had been lost in February from the Leeward Islands, the Cartaret from Jamaica homeward-bound in March, the Matilda also in March from Falmouth for the West Indies, three more in May, all homeward-bound from Jamaica, and two outward and one homeward-bound in June.

"You don't have the actual positions where they were captured?"

Smith shook his head. "The only information sent to me is given there."

"Out of nine, five were homeward-bound from Jamaica," Ramage said, slowly scanning the neat writing, "one homeward-bound from the Leeward Islands, and only three outward-bound from Falmouth..."

"That is correct," Smith said.

"Seems strange," Ramage mused, reading the list again.

"What does?"

"So many lost on the way back."

Smith shrugged his shoulders. "Easier to catch 'em going back - that's obvious!" Really, he thought, the youngster looks sharp enough but he doesn't seem to know much about the way these damned French privateers lurk around the islands!

"Why is it easier?" Ramage asked, his voice disarmingly innocent.

"Well," Smith said pompously, "far be it from me ... But obviously the privateers just hang around the Windward Passage! Probably waiting in the southern Bahamas."

Ramage folded the list and tapped the table with it. Quite reasonably, Smith was assuming the losses were due to privateers, yet assumptions at this stage were dangerous. "But they don't know the date a packet is likely to sail from here, do they?"

"Of course not! I hardly know myself until the last moment. It all depends when one arrives."

"So if they wanted to be sure of catching the Falmouth-bound packets, privateers would have to patrol all the obvious places all the time?"

"Obviously," Smith said, with something approaching contempt in his voice. He's recovering from the effect of those millions of square miles, Ramage noted wryly.

"But surely the mails bound for Jamaica would be more valuable? Anyway, no less valuable."

Smith shrugged his shoulders; the young fellow seemed to be asking questions just for the sake of it.

"The Jamaica packets," Ramage said. "They all come here from the Leeward Islands after calling at Barbados and Antigua. None comes direct from Falmouth?"

Smith nodded.

"So in effect we can picture two highways," Ramage said, running a finger along the table-top. "One goes from Falmouth across the Atlantic to Barbados, up to Antigua, and then right across the Caribbean to Jamaica, and the Jamaica packet sails along that, delivering and collecting mail at the various islands until she arrives here in Kingston about forty-five days after leaving Falmouth. The other runs north-east from Jamaica up through the Windward Passage between Cuba and Hispaniola and out into the Atlantic and back to Falmouth and is used by the homeward-bound packets which take about thirty-five days to reach England."