Let us consider the problem. Suppose for convenience that the earth is divided up into “squares,” as nearly, at least, as you can consider a globe to be so marked out. Let us suppose that it has been agreed to draw on it from pole to pole 360 lines of longitude, commencing with one through say Greenwich Observatory as a starting-point, and going right round the earth till you come back to Greenwich again. Also suppose that there have been drawn a series of circles parallel to the equator, but going up at equal distances apart towards the poles. Let us have 179 of these circles, so as to leave 180 spaces, _a_ to _b_, _b_ to _c_, etc., from pole to pole. This will divide the earth up like a bird-cage into squares, as if we had robed it in a well-fitting Scotch plaid. The length measured along the equator of the side _p q_ of each square at the equator is taken as exactly sixty nautical miles (apart from a small error of measurement, which makes it in actual practice 59·96). This is equal to sixty-nine and a quarter English statute miles. The side of the square leading towards the poles _q s_ would also be sixty nautical miles were it not that the earth is not truly spherical, which introduces a slight error. We may, however, roughly say that at the equator each square measures sixty nautical miles each way. [Illustration: FIG. 64.] As we get towards the poles the squares become rectangular figures, with the heights of latitude still sixty nautical miles, but the widths becoming smaller. Thus in England our squares measure _p q_ = 37 nautical miles and _q s_ = 60 nautical miles. Now of course we can see at once that it is easy at any place on the earth’s surface to find your _latitude_ by a simple observation of the sun at noon, if you know the day of the year, and have got a nautical almanac. For by an instrument called a sextant you can measure the angle he appears to be above the horizon, and then, as you know from a nautical almanac the angle he is above the equator, you can soon determine your place _A_ on the globe. Or at night, if you measure the angular distance that the polar star _P_ is from the zenith, or point exactly over your head—that is, the angle _P O Z_—you can subtract it from a right angle and get your latitude, _A O E_, at once. [Illustration: FIG. 65.]