It sometimes happens that in well charted waters it is desirable to plot the ship's position at frequent intervals, or continuously; and it may be desirable to know the ship's head independently of the compass, the error of which, from lack of recent observations or by change of latitude, may not be sufficiently known.
This apparatus will accomplish both of these objects instantaneously with an accuracy greater than that of ordinary cross bearings, and but little less than that of sextant angles; both of which methods require time, and which give the position only, not the course.
The principles employed are as follows: assuming the ship at rest, her position is determined approximately by two compass bearings, more accurately by three. If the three bearings cross in a point the deviation is assumed to be 0. If they do not so cross, it is possible to find both the true position and the deviation (assuming always that the chart is correct). The first impulse is to locate the true position in the center of the triangle formed by the intersection of the three bearings. This may or may not be correct. It is reasonably correct if the objects are equally distant from the ship and are at approximately equal angular distances apart. But the true position may be outside the triangle altogether. The rule is, facing successively each object, the true position is on the same side of all the lines of bearing successively, that is, it is to the right or to the left, of them all (facing the object), and its distance from the lines of bearing is proportional to the distance of the objects. This will be clear from Figs. 1 and 2. In each case the dot is the true position. The full lines are the compass bearings, the dotted lines the magnetic bearings (also true bearings if the variation is assumed as 0, which it may be in the discussion), and the angle at each object between the full line and the dotted line is the deviation, that is, it is the same angle at each object. Also each dotted line must be on the same side of each full line, facing the object, since the deviation is always the same way (easterly in the figures). So it follows that the distance of the true position from each line of bearing is proportional to the distance of the object, being that distance multiplied by the sine of the deviation. The true position is inside the triangle in Fig. 1, outside in Fig. 2. Also in Fig. 2, the true position is nearest to the line of bearing of A, the nearest object, next nearest to that of B, the next nearest object, and farthest from that of C, the farthest object. The above is on the assumption always that the chart is correct, and that an unknown deviation is the only cause of the divergence of the bearings.
A useful practical way of obtaining the true position and the deviation is the following, which was used successfully aboard the Massachusetts in Newport Harbor, when cloudy weather and fog prevented azimuths. Three objects were observed by the standard compass. The bearings were then drawn on the chart. If they intersected in a point, the deviation on that heading was assumed as 0. If they did not, then the angles between the bearings were assumed as the correct angles between the objects (which they evidently were, the ship and compass being at rest), and these angles were then applied to the arms of a three-arm protractor, by which the true position was plotted. Lines were then drawn from the true position to the three objects. The angles at the objects should have been the same, that is, the deviation. If they were not the same the mean of the three was taken as the deviation.
The present apparatus is intended to accomplish all the above instantaneously, and whether the ship is at anchor or under way. Suppose a chart is fastened to a pivot-top table and turned until its meridian lies in the actual meridian. Then suppose three sight-vanes are pivoted on three objects as shown on this chart and are then sighted at the actual objects. If the chart is correct and truly in the meridian, the three sight-vanes will intersect in a point, which is the ship's position. If we could always get the chart in the actual meridian, two sight-vanes would be enough. The third vane gives a practical means of accomplishing this object. On placing the chart somewhere near the meridian, the vanes will be observed to cross in a triangle instead of a point. Now turning the table-top slightly one way or the other, the triangle will grow larger or will grow smaller. The table-top is finally turned until the triangle reduces itself to a point. This is the ship's position, and may be dotted on the chart. A pointer on the edge of the table shows the ship's head with regard to the meridian. A comparison of this with the ship's head by compass gives the compass error; or, if the variation is known, then the deviation. This operation is instantaneous and may be made continuous.
The actual apparatus is three slotted sight-vanes with a perforated loose clamp pivot through the slots. Each vane has a small hole near the end which may be placed over a pin on the chart through one of the objects selected for bearings. The best form of pin is made like a thumb-tack, with the shaft extending through the head of the tack. The lower end of the shaft is pointed and is stuck through the chart into the top of the table, which should be of soft wood. The upper end becomes the pivot for the vane. The table-top is round and pivoted on ball bearings. The edge is graduated in degrees like the compass. There is a pointer that shows the ship's head, or fore-and-aft line.
The operation in use is as follows: The chart is fastened on the table with its meridian in the north and south line. Either the true meridian or the magnetic meridian may be used. Then three objects being selected for bearings, a pivot thumb-tack is placed on each one. The vanes are then placed on the pivots and an observer for each vane sights his vane on the object. The vanes slide on the loose clamp pivot, which preserves the common intersection, and the table turns automatically until the meridian of the chart lies in the actual meridian. The degrees as read from the edge of the table give the course (true or magnetic), and when compared with the compass give the error (or deviation). When the observers are all on, one of them calls "mark," the table and vanes are left at rest, the ship's head is noted by compass and by table, and the ship's position is dotted through the center of the clamp pivot. Or a tracer pencil may be fitted through the center of the clamp pivot which will thus trace the course. The true course may be read at any time from the edge of the table independent of the compass.
The apparatus has worked very satisfactorily aboard the Chattanooga. The results when using an accurate chart and a compass with a known deviation, have come out within a degree or a half-degree. It is quite interesting to watch the table settling itself to the course without apparent assistance. The observers simply keep their sights on, and what effort they use in so doing turns the table to the correct heading.
The photograph shows the apparatus in use. There are three observers, and a fourth stands ready to read the compass, or dot the position through the center of the clamp. The present apparatus was made aboard ship and is much heavier than necessary. By using a stiff bronze, and having the work done by an instrument-maker, the apparatus could be made materially lighter, and hence would take much less room on the chart. The lighter and smaller the parts, the less interference of pivots, and hence a wider range of application. This is a matter of mechanical detail.
A number of uses suggest themselves. Finding the compass error has already been mentioned. It would also be useful in obtaining tactical data. In pilot waters, if the chart and compass are both reasonably accurate, the board can be kept to the ship's head from the compass, one vane may be dispensed with, thus giving more space and range, and the ship's course may be traced continuously from the other two vanes. It would then be a matter of watching the ship trace her course on the chart. The apparatus would find a wide field in the running of lines of soundings, as the lines could be kept continuously, even without ranges. For boat work the water would have to be smooth, but for ship work this would not so much matter. These and other uses that will readily suggest themselves would warrant adding this apparatus to the navigator's outfit.