It will be noted that this observation has value for the purpose in question only in conjunction with certain tables in which the movements of Jupiter and its satellite are calculated in advance. This is equally true of the various other observations through which the same information may be obtained--as for example, the observation of a transit of Mars, or the measurement of apparent distance between the moon and a given fixed star. Before the tables giving such computations were published it was quite impossible to determine the exact longitude of any transatlantic place whatsoever. We have already pointed out that Columbus had only a vague notion as to how far he had sailed when he discovered land in the West. The same vagueness obtained with all the explorations of the immediately ensuing generations.
It was not until about the middle of the sixteenth century that Mercator and his successors brought the art of map-making to perfection; and the celebrated astronomical tables of the German Mayer, which served as the foundation for calculations of great importance to the navigator, were not published until 1753. The first _Nautical Almanac_, in which all manner of astronomical tables to guide the navigator were included, was published at the British Royal Observatory in 1767.
At the present time, a navigator would be as likely to start on a voyage without compass and sextant as without charts and a _Nautical Almanac_. Indeed were he to overlook the latter the former would serve but a vague and inadequate purpose. Yet, as just indicated, this invaluable adjunct to the equipment of the navigator was not available until well toward the close of the eighteenth century. But of course numerous general tables had been in use long before this, else--to revert to the matter directly in hand--it would not have been possible to make the above-recorded test in the case of Harrison's famous watch in the voyage of 1761-62.
In the days before the chronometer was perfected, almost numberless methods of attempting to determine the longitude of a ship at sea were suggested. There were astronomers who advocated observation of the eclipse of Jupiter's satellites; others who championed the method of so-called lunars--that is to say, calculation based on observation of the distance of the moon at a given local time from one or another of certain fixed stars arbitrarily selected by the calculator. Inasmuch as the seaman could always regulate even a faulty watch from day to day by observation of the meridian passage of the sun, it was thought that these observations of Jupiter's satellite or of the moon would serve to determine Greenwich time and therefore the longitude at which the observation was made with a fair degree of accuracy. But in practice it is not easy to observe the eclipse of Jupiter's satellite without a fair telescope; and it was soon found that the tables for calculating the course of the moon were by no means reliable, hence theoretically excellent methods of determining longitude by observation of that body proved quite unreliable in practice.
