Popular Science Monthly/Volume 29/August 1886/Good Time and its Ascertainment
By Professor ISAAC SHARPLESS.
THE natural divisions of time are the year and the day. The week is arbitrary, being probably derived from considerations first suggested by the first chapter of Genesis. The month, though originally intended to be the time from one new moon to the next, has, of necessity, departed from this idea, in order to make an even number in the year. The decade and the century are purely artificial, deduced from our system of numbering. But the day and the year, the one derived from the reappearance of light and darkness, the other measuring the round of the seasons, are universally adopted units of time, suggesting themselves alike to cultured and savage, and which we can not think will ever be superseded.
The year is the time of the revolution of the earth around the sun. Its measure is most easily obtained by the reappearance of the sun at the same altitude in the sky. Every one knows that it is higher in summer than in winter. If the circle of the earth's equator were extended right out from the center of the earth into the sky, it would cut out a circle there which is called the celestial equator. Now, the sun crosses this line in the spring northward, arriving at its greatest altitude in the middle of summer; thence it descends, crossing the line southward in the fall, and reaching its lowest point in midwinter. The ancients, by measuring the length of the shadow cast by a vertical stick on different days of the year, arrived at surprisingly correct results as to the length of the year. In 450 b. c, Democritus asserted the year to be 3651days long, which is within about eleven minutes of the truth. Another ingenious device for the same purpose was that of the Egyptian astronomers, who set up a wheel parallel to the plane of the equator. When the sun was in this plane, the shadow of the sunward side of the wheel would be exactly intercepted by the other, and the interval between two such occurrences would measure the year. Owing to the fact that the sun does not "cross the celestial equator in the same place each year, this year which measures the seasons is a few minutes shorter than the exact time of the earth's motion around the sun.
To measure the day troubled the ancients much more. It is, perhaps, a common idea that the shadow of a vertical rod cast by the sun is always exactly northward at twelve o'clock noon. Any one desirous of trying this can easily do so, and he will find that such a shadow would be sometimes eastward and sometimes westward of the meridian-mark at noon. Moreover, he will find that the time between two passages of the sun over his meridian is not the same, so that, if this time were taken as the day, there would be no uniformity. It was, however, the recognized day in most countries till comparatively recent times. In France, when in 1816 the change was made to our present system, there were fears of a disturbance among working-people, lest the abolition of the sun-day should somehow increase their hours of labor. It met the approval of the watch-makers, however, whose customers had hitherto complained that their watches would not keep pace with the sun, not knowing that this would be impossible for a good watch.
The time of the rotation of the earth on its axis can not be measured directly from the sun, for the reason that the earth is moving around it. We must have some external point, fixed with reference to the earth, by which to measure it. The stars afford such points. By noticing the time between two successive passages of a star over our meridian (our meridian being, as is well known, the semicircle in the sky passing from the north to the south point of the horizon directly overhead), we would obtain the exact time of the earth's completing one spin on its axis. This time, which is about four minutes less than our ordinary day, is called in astronomical parlance a sidereal day, and, divided in the ordinary manner into hours, minutes, and seconds, is known as sidereal time. It has no direct relations to ordinary life.
Through all the time that the earth is making one turn on its axis it is advancing around the sun in the same direction. So it takes this extra four minutes to bring the same meridian under the sun again, after making a complete revolution. Hence we have our solar day. Again, since the forward motion of the earth is not uniform, as well as for another cause, which is too intricate to mention here, the solar days are not, as we have said above, of equal length. So the device is adopted of ascertaining their average through the year and calling it the mean solar day. This, subdivided into hours, minutes, and seconds, is mean time—the clock-time of ordinary life.
If, therefore, it is desired to find correct time from a sun-dial, or by any method depending on the sun, the correction from apparent to mean time must be made. At four instants during the year this correction is zero. At other times a quantity, amounting at its greatest to about sixteen minutes, must be added to or subtracted from sun-time. For several days in the early part of November the sun is on the meridian more than a quarter of an hour before twelve o'clock. Our present system is not exact sun-time, but sun-time so modified as to be adapted to the current wants of our existence. It is uniform, because it is based on the time of revolution of the earth on its axis, which has not varied, if at all, more than one sixtieth of a second in the past twenty-five hundred years. But the common day is not the exact time of the earth's revolution, nor is the common year the exact time of its motion around the sun.
The tendency of civilization seems to be to depart from these strict astronomical units, while all the time depending upon them for their ascertainment. And the recent changes of using "standard time" are in the same direction. Of course, every place, not just north or south of another, has a different noon. To prevent the confusion resulting from so many "times," our railroads have adopted as noon the mean times of certain standard meridians. These are taken just one hour apart, so that if the new time were universally adopted, the minute and second hands of all correct clocks would be the same over the whole United States, and the hour-hands would differ by one, two, or three hours. In England they have used Greenwich time over the island for many years, and our system is connected with theirs by using for our standard meridians those which are an even number of hours from Greenwich. In Philadelphia, for instance, which is situated on a standard meridian, the time is just five hours later; so that tidings of an event, happening at noon in London, if telegraphed immediately, will reach Philadelphia a few minutes after seven o'clock in the morning.
The objections to adopting this standard time, in some places, based on the inconveniences of having noon at some other time than when the sun is on the meridian, very much resemble those made in France when the Government substituted mean noon for apparent. In practice we never know when the sun is on the meridian, and if it gets there at 12.30 instead of 12, no one is the worse off, and the methods of living are readily adaptable to it.
Time being thus dependent on the facts of astronomy, its ascertainment is a part of the work of an astronomical observatory. The instrument used for the purpose is a transit-instrument. It consists of a telescope which is mounted, not to be pointed to any part of the sky, but to swing only in the plane of the meridian. It will point horizontally, north or south, to the zenith, and to intermediate points. A star in the east or west can not be seen by it. When it crosses the meridian, if the telescope is elevated to the proper angle, it will cross the field of view. To determine exactly what part of the field the meridian crosses, a spider-thread is stretched in the tube just in front of the eye-piece, which by a very accurate adjustment must be made to coincide exactly with the meridian. Just as the star crosses this thread, or, to speak more accurately, just as the particular meridian of the place passes under the star, the time must be recorded. As there is a possibility of an error in this, several spider-lines are inserted parallel to this central one, and symmetrically placed on either side.
The telescope is connected with an axis pointing east and west, working on the tops of two pillars set far enough apart to allow the telescope to swing between them.
Let us now go through the operation of "taking a transit." The observer, by means of graduated circles, points his telescope to the place in the heavens where he knows the star is to cross. He has his clock or chronometer by his side ticking seconds or half-seconds. A little lamp sends a ray into the tube of the telescope, so that he can see the spider-lines. With paper and pencil in hand he stations himself in front of the tube. The star enters the field of view and moves toward the first spider-line. He glances at the clock, catches the time by the second's hand, and counts the ticks. Three—four—five—six—the star has just crossed a line. Estimating the tenths of a second, he records the time on the paper. All this while he is noting the beats of the clock, and, when the star reaches the second-line, he is ready to record another transit, and so on through. The mean of all these times is the time of crossing the central line by the sidereal clock. But in the "Nautical Almanac" this time is given accurately, and a comparison of the two shows his clock error.
Instead of recording the transits by the "eye-and-ear" method above described, there is an easier way by simply tapping the key of an electric circuit at the time of transit. This makes a record on a "chronograph," which can be read at leisure.
A chronograph consists of a brass cylinder, on which is fastened a sheet of paper. This is placed with its axis horizontal, and is revolved uniformly by clock-work. A pen rests with its point against the paper, making a mark around it. By a slight longitudinal motion this mark does not come around into itself, but advances a trifle, being like the thread of a screw, running from end to end. A current from a galvanic battery is so arranged that every swing of the second's pendulum causes an electro-magnet to attract the armature to which the pen is attached, and makes a break in the mark. Hence there is a series of breaks separated by intervals of a second. When the observer notes a transit, he, by his key, makes galvanic connection and interjects another break in the line. The position of this break among the seconds tells when the transit occurred, the fractions of a second being readily read.
He thus knows sidereal time; a little reduction gives him the mean solar time of the place of observation, from which the time at any other place whose longitude is known is directly deduced.
His telescope, to avoid all possibility of error, must be in perfect adjustment. The axis must be level; it must point east and west; his spider-line must be correctly placed in the tube; the pivots of the axis must be of equal size and uniformly round, and the axis must not bend under the weight of the tube. All these sources of error are carefully guarded against, but, as human powers are finite and disturbing causes very plentiful, errors will be introduced in various directions. So he seeks to nullify these by taking many stars in different parts of the sky, and from the varying errors he deduces what part belongs to the clock and what to the instrument. Should cloudy weather continue for many successive days and nights, he has to fall back on his knowledge of the rate of his clock, which is kept under as uniform conditions of temperature and moisture as possible. There must be something radically wrong, either with the observer or his equipment, if he can not give the time of noon within a very few tenths of a second.