Furthermore, Morison had sailed the Bahamas route in Mary Otis in 1940 and identified to his satisfaction the locales of the log. The question was closed, and it remained so for 40 years.
But Morison was wrong on three counts: The question would not go away; the track he sailed and proposed will not bear scrutiny; and the landfall was not Watling. Of that, I had no doubt. Other dissenters, such as Edwin and Marion Link and Pieter Verhoog and Robert Fuson and Robert Power and Arne Molander, found great fault with the Watling route—but I found their alternatives flawed in other ways. There is something to be said for Sherlock Holmes’s dictum that “when you have eliminated the impossible, whatever remains, however improbable, must be the truth.”
When, in 1980, the Society for the History of Discoveries took a bold step and reopened the landfall question with a reading of Verhoog’s 1954 paper recommending East Caicos, and papers followed advocating Grand Turk and other islands, I knew that at last “the game is afoot.” It was time to sic the hounds after Sherlock’s grand improbable. My thought was to sic the computer. The log is filled with quantifiable data—bearings and distances influenced by wind and current and sailing speed. A computer could run the variables out thousands of times.
Our first need was for a new partial translation of the diario. NATIONAL GEOGRAPHIC commissioned an expert in old Spanish documents, Dr. Eugene Lyon, to provide a literal, line-by-line translation from photocopies of the Las Casas manuscript. It was the first in English in many years, it was superb, and it became our bible.
Then I turned to another old friend and colleague, Luis Marden, who had twice sailed the southern route across the Atlantic in his ketch, Bounty, with his wife, Ethel, a first-class mathematician. They agreed to tackle the task of rendering the transatlantic log in terms of day-by-day actual position. The Mardens quickly made two astonishing, and crucial, discoveries. The story of how their corrected daily plots lead to Samana Cay begins in the next column.
In 1984 a major computer firm, Control Data Corporation, joined the Columbus team in the figure of Vice President Robert Lillestrand, an Arctic-navigation expert who became intrigued with the Columbusmystery. He assigned Carla Ryti and Scott Devitt to create the Columbus Research Tool (CRT)—an interactive program marrying a Control Data Cyber 170/865 computer with a color display in which the geography of the Bahamas had been digitized, and in which it was possible to sail courses electronically and instantaneously from any point to any point. An early run through using the CRT quickly discounted tracks from Grand Turk as not consonant with the Columbus log.
So—how does one go about discovering where the best accommodation in barcelona is? And how can you prove it once you have found it? Long ago Morison proposed three com-monsense tests. First, the transatlantic track ends there. Second, a track drawn backward from an undisputed point in Cuba ends there. Third, the island must fit the description in the log.