they do not form the basis of the charts. That basis is the variety of observations compiled and reduced in the careful and laborious manner already described, and which have then been examined from every standpoint to elicit whatever beam of information they might contain to illumine the ocean highways. Therefore, in justice to the amount of thought, care, and labor bestowed on these charts, it must be stated that they are not a rearrangement of old matter, but are essentially new, and from original sources.
The second series of charts (now in course of compilation) embraces a series for the whole Atlantic between the parallels of 60° north and 60° south. The observations of the direction of the wind on Maury's pilot-charts, to the number of about 2,600,000, or 300 years, will be embodied in them: the actual period over which these observations extend is from the year 1800 to 1855. But, as with the Pacific charts, the real groundwork of the Atlantic series is the various kinds of observations at present compiled in the careful manner already described from recent log-books and meteorological journals. In number, these observations will amount to about 650,000 hours, or 75 years. The actual period over which they will extend is from the year 1855 to 1881, the time when it is confidently hoped this set will be ready for issue. By far the greater part of the compilation is already done. The Atlantic charts, then, will contain hourly observations to the number of about 3,250,000, or nearly 375 years: in other words, if it were imposed on a single ship to collect this mass of data, she would have to cruise in all parts of the Atlantic during every month of the year, for a period of 375 years, without ever going into port!
As, however, these observations were collected by a multitude of vessels and during a continuous series of 81 years, several vessels were observing at the same time in different parts of the ocean. Surely, this is an abundance of facts that must render indisputable the information contained in the more frequented squares: more would be mere accumulation, without perceptibly affecting the mean results.
On some accounts it would be desirable to have the areas for which the information is classified smaller than 5° squares, as 1° squares, for example; but, again, there are objections, all but insuperable, to such a system:
1. To collect data for it sufficient to give trustworthy results, would require a fleet of cruisers almost as large as the combined merchant marine of the world—all to be assiduously engaged for many years. This is unattainable. Even with the inducements now offered, and notwithstanding that the undertaking is mainly for their benefit, only a very small percentage of all the masters of merchant-vessels will take the trouble to keep a meteorological journal with the requisite accuracy and care. Were it not for the excellent log-books of our ships of war, our knowledge of the phenomena of the ocean would indeed be most meager and inaccurate.