In his report to the chief of Naval Operations concerning the search for the Akron, the commanding officer of the U. S. S. Portland stated that there was very little known data with which to initiate the search. The positions of the Akron crash were variously reported as from thirteen to twenty-four miles south of Barnegat Light Ship, twenty miles southeast of Barnegat Inlet Light, and merely twenty miles off Barnegat Light Ship. Even the SS. Phoebus, reporting the crash, had only an uncertain dead reckoning position to give of the location of the crash.
The search, then, at the outset, presented the problem of covering hundreds of square miles of ocean in an effort to locate enough wreckage to diminish the size of the search area to permit profitable dragging. At the same time, as the search progressed, the search area was also being increased in size—miles to seaward to discredit the possibility that a portion, at least, of the Akron had been blown out to sea.
At the conclusion of the tenth day of the search, some sixty pieces of wreckage, or “clues,” had been recovered; enough to define an area of approximately 400 square miles suitable for dragging operations.
Even this area was too large to permit bringing the search to a swift conclusion. But what about these pieces of wreckage, these clues? In themselves, drifting about the ocean and acted upon by the local winds and tides, didn’t they offer a possible solution to the problem of reducing this draggable area?
The idea was that since each piece of wreckage had originated from the Akron, it ought to be possible to reconstruct a drift track for each clue, and, plotting it in a reverse direction from where the wreckage was recovered, discover that many of these tracks crossed in a “fix” on the ocean’s surface, under which the Akron would be located.
To reconstruct the path of a piece of drifting wreckage, in direction and distance traveled, four factors had to be considered: the effect of the tides, the length of time the wreckage had been on the surface, the effect of the wind, and the effect the size of a piece of wreckage would have on its mobility.
According to the coast pilot for the Jersey coast area, the tidal currents were mostly those set up by the wind. Beyond the 20-fathom curve there had been reported, at various times, a counter-Gulf Stream current of about a quarter-knot strength, setting to the south and southwest. Principally, however, the tide followed the winds.
The longer a piece of wreckage had been on the surface, the greater distance it would travel. And, of course, a large piece, riding higher in the water, would be more affected by the wind and travel faster than a smaller piece almost submerged. It would have been impossible to classify each piece of wreckage in terms of its drift velocity, and the differences in velocity thus encountered were not considered.
The effect of the local winds was, however, carefully considered. From the log were obtained the direction and velocity of the true wind for each hour, from the time of the crash to the tenth day of the search. Using the time of the crash as the point of origin, the wind for each hour was plotted vectorially, in direction and force on a plotting sheet. From the continuous wind curve thus plotted, it was a simple matter to pick off from it the average wind acting upon each piece of wreckage from the time of the crash to the time that wreckage was recovered. This line, in direction, represented the direction of the average wind; in length, its velocity. Obviously, it also represented the path, in direction, in which the wreckage had drifted.
However, as has been recorded, it was impossible to determine the velocity of any object through the water; and it was quite as impossible to determine when an object had come to the surface so as to compute the length of time it had been exposed to the wind. But in numbers there is safety, and it was thought that though many errors would exist in plotting these drift tracks, out of about sixty, so plotted, enough would be reasonably accurate to insure some degree of success with this method. Then too, disregarding the drift of these objects, the plotting of their drift tracks in direction only would dispose of considerable error.
The location of each piece of wreckage recovered was accurately plotted and numbered on a chart. From each object was drawn its average directional drift track, as determined from the vector curve of the existing winds; these tracks, of course, being plotted in a reverse direction, towards the point from whence the object had drifted.
When these lines were plotted no single fix or locus could be identified. But it was ] discovered that practically all of the more prominent intersections could be enclosed in and determined a circle seven miles in diameter!
Considering the means to the end, the results of this method were particularly gratifying. Because it is so fraught with errors, with conjectures, and with probabilities, perhaps little faith should be placed in the results. The fact remains, however, that four days later the Akron was located, 1.3 miles from the center of this 7-mile position circle.
*
Preparation for war, rightly understood, falls under two heads,—preparation and preparedness. The one is a question mainly of material, and is constant in its action. The second involves an idea of completeness. When, at a particular moment, preparations are completed, one is prepared—not otherwise. There may have been made a great deal of very necessary preparation for war without being prepared. Every constituent of preparation may be behindhand, or some elements may be perfectly ready, while others are not. In neither case can a state be said to be prepared.—Mahan, Interest of America in Sea Power, Present and Future.