so great as to lead one to the belief that they are the result of mere guesswork. It is related of an old commanding officer in times gone by, that noticing on the chart what he thought to be a rock near which the vessel would pass, he told the navigator to have the course changed at once. The navigator replied, " I think, sir, there is not much danger in that rock," and with his handkerchief brushed from the chart the fragment of snuff that had fallen from the old gentleman's beard. We all wish that the charts could be as easily cleared of all doubtful dangers, and commanding officers as easily relieved from the many cares and doubts that they must feel when in their neighborhood. I have heard many adverse criticisms and objections raised against putting on charts these numerous doubtful reports. Many of the reported dangers, without doubt, have an existence only on paper. But who can tell without the strictest search which one shall be retained and which erased? It is a very easy matter to expunge a shoal from a plate; but it may be rather difficult to explain to the captain who has wrecked his vessel on it afterwards that its removal was warranted on sufficient grounds. I would not advocate placing a new report on a chart without a probability that the danger referred to really exists, and that its position is approximately correct. But, once deemed of sufficient importance to be placed on a chart, it ought not to be removed until an exhaustive search has proved its non-existence with almost absolute certainty. The difficulty of proving beyond doubt that a shoal, even in a well-known harbor, does not exist, is very great, as can be shown by many examples. How much more difficult is it to determine this in mid-ocean! The Rodgers shoal in the China sea was searched for during four days, but, although the original report gave bearings of points on shore, and the boats sounded and swept with a chain all around the spot, no danger was found, until when the vessel was about to get underway, a ripple on the water close aboard was examined and found to be over the rock, almost exactly on the intersection of the bearings given. The vessel and her boats had been many times within a very few feet of the rock without having been able to discover any signs of it. The Brilliant shoal, in about latitude 23° south and longitude 170° east, was discovered in 1847 and located. Captain Durhan, R. N., made a careful search for it, running many lines over a surface sixty miles in diameter about the position, without being able to discover any signs of shoal water, although sounding with 250 fathoms of line. This

was thought to be conclusive evidence of its non-existence, and the British Admiralty expunged it from their charts. It was not long, however, before a French and an English vessel reported it as very near the original position given. It was replaced on the chart, and no doubt was expressed as to the fact of its existence and the correctness of the location. Recently, another English vessel.has searched for it, but without success, and again it is marked "Position doubtful." Even on our own coast, within the past year, in Long Island sound, a rock has been discovered having over it 10 feet of water. It was searched for during many days without success; but finally it was discovered when, on getting underway, the centre board of the surveying schooner grazed its surface.

There are many instances where examinations have been made, and the shoal expunged on evidence of non-existence, derived from careful search over large extent of surface, and from frequent soundings with deep-sea lead in two and three hundred fathoms of water. The evidence, as far as the lead is concerned, is really of not much value when we remember how abruptly rocks rise from great depths. The Coast Survey steamer Blake found depths of over 2000 fathoms within 2} miles of the Bahama banks; and in some places on the line of soundings the inclination was 45°. Small pinnacles are often met with inshore, and sometimes with such inclination that the lead will not rest on their surfaces. One instance of this is met with near Armstrong bay, on the south coast of Australia. The pinnacle has only one fathom over its apex, and presents so small a surface to the waves that the sea rarely breaks thereon, and the lead finds no resting place. Cases of submersion or subsidence are sometimes found. The Gorringe bank was originally reported as having 3 fathoms of water; but the Gettysburg found 32 fathoms to be the least depth. Broken coral and shells and well-rounded pebbles were brought up, indicating that erosive wave-action had at one time exercised its influence on the summit of the shoal. The Hunter island was reported to have been situated in latitude 15° south, and in about longitude 166° east. It was described as of volcanic origin, well-peopled and cultivated, corresponding in most particulars to the island of Nui-fu, situated in about the same latitude, but in 176° west longitude. Many searching parties have been sent to look for it, but without success; two or three years since, the British Admiralty issued a notice that probably the two islands mentioned were identical, Hunter island having, through a typographical error, been transferred from west to east longitude.

I think it probable, however, that the island has sunk beneath the sea. The searching parties have found two or three shoals near its reported position, and one part of the original description of the island is so unique as to throw a doubt on the validity of the conclusion of the British Hydrographic Office. The little fingers of the left hands of all the inhabitants had been cut off at the second joint, and their cheeks were perforated. The first is a remarkable peculiarity that would undoubtedly attract the immediate attention of explorers, but nowhere else can I find mention of this maiming of the hand.

As to the best method of searching for dangers, I can offer no new ideas. There is no improvement upon the old plan of running on ranges, having one or more buoys ready to drop at a moment's notice. In deep-sea searching, the wire-sounding apparatus of Sir William Thompson, and that of Commander Sigsbee, are acknowledged by all to be not only useful as auxiliaries, but also absolutely necessary, if valuable results are to be expected. When a shoal is found it should, if possible, be examined from a boat. Even if the boat cannot go over the shoal, it can go near enough to be sure that the discoloration is not due to the presence of animalculae, or to breaking water carried on by the tide or currents. If the breakers are small, the boat can pull close around them and ascertain at once if there is an extension of the danger, a service that the ship could not satisfactorily perform on account of her greater draught. If time can be spared it would be well to take, not a single set of observations, but a series of sets on both sides of the meridian, or a series of three or four twilight observations, for Summer's lines. In the report, the officer should state definitely the point on which the chronometer errors depended, both before and after the discovery.

The Mercator projection for the construction of charts is the one in general, and I may say, almost universal use. It possesses an advantage which is deemed by most seafaring men an essential to a chart; any straight line drawn makes the same angle with all meridians. It has on the other hand no two consecutive miles of the same length, except on the parallels. The rhumb-line represents a part of a spiral, or practically a small circle of the earth. For a chart of the world, or for long distances, it is probably better to use Mercator's projection; but for small extents of surface, say 10 or 15, or at most, 20 degrees of longitude, the polyconic projection is far preferable, for the following reasons: 1st. It distorts to a less degree the configuration of the land or bottom; 2d. It has a scale

which may be taken from any point, on any meridian (preferably the middle one), and which may be used on any portion of the chart; and 3d. On the polyconic projection, a straight line is almost exactly part of a great circle of the earth; and by sailing on it we travel on the shortest possible line between the two points.

There are many objections named to the polyconic projection, and some of them I will mention. Recently, a commanding officer in the Gulf of Mexico noticed, I suppose for the first time, a chart on the polyconic projection. He was bound to New Orleans, and, calling his navigator, he asked how it was that the chartmakers could have made such a mistake as to curve the parallels. Another objected because he could not lay out a course on it with confidence, fearing every time that his landfall would be miles away from his port. Another said he could not plot a bearing to a distant peak, because the projection was curved and his bearing would be straight, forgetting that his line of vision was a straight line and followed in the plane of a great circle, and that on Mercator's projection all lines, except the meridians and equator, represent small circles of the earth. Another objected to this unwelcome innovation because he could not plot a position with a parallel ruler from the margin of the chart; and another, because he was accustomed to Mercator's projection. I suppose I have heard scores more of as reasonable objections, and probably you all have heard similar ones. If we are always to adhere to what is old, out of respect to the wisdom of our fathers, we shall never improve; but, perhaps, in future generations, when pounds, shillings, and pence, and ounces and pounds, are things of the past, eleven and a quarter degrees to a point of the compass will have gone too; and men who go to sea will not be afraid to use a chart simply because it is on the polyconic projection.

Without going into minute details, the simple conic projection is this Imagine a cone surrounding the globe, its apex coincident with the earth's axes produced, and its surface tangent at any given middle parallel. Upon developing the cone, this parallel will be the arc of a circle having a radius equal to the slant side of the cone from the apex to the point of tangency. Parallels are struck from the same centre, and the meridians are straight lines. lengths of the degrees of latitude are true. are in excess of the true below the middle the true above the parallel.

It will be seen that the The degrees of longitude parallel, and are less than

There are many modifications of this projection, the most valuable