The great progress that has been made in naval architecture and marine steam engineering demands improvement in the instruments for navigating the magnificent products of human ingenuity as exemplified in the modern men-of-war.
The solarometer claims to meet this demand by providing a method of making astronomical observations independent of the visibility of the sea horizon.
The primary object of the solarometer is this feature of its artificial horizon, which is so combined with the astronomical values of declination, hour angle, azimuth and altitude in relation to the observer's latitude and zenith that all the elements of the nautical astronomical problems are solved.
The mechanical solution of the problem as in the solarometer becomes more of a practical necessity as a result of hourly opportunities to astronomically determine the ship's position and compass errors; not because of any defects in the mathematical computations from altitude by the sextant, but in order to save the time these calculations involve.
If observations were taken once or twice every hour, they would involve a total of from six to eight hours daily work with sextant, alidades and logarithms to obtain the results of those observations, and this total time would be so distributed throughout the 24 hours that there would be very few spare minutes left for the navigator to do anything else than observe and calculate the results of his observations.
The solarometer obviates elaborate logarithmic calculations and combines in itself a pelorus; so that it furnishes a complete solution of the entire problem to ascertain ship's position and compass errors in the space of time ordinarily required to observe the altitude by the sextant and take its bearing with a pelorus.
The general principles upon which the solarometer is constructed may be concisely stated to consist of a series of circles representing the nautical astronomical triangle supported upon a constant level base which locates the position of the observer's zenith in that triangle.
There is a definite relation between all the five quantities of declination, latitude, hour angle, altitude, and azimuth, such that with each and every variation of the value of one or more of these quantities, the others have corresponding values. The variations of all these quantities cause an infinite variety of possible values to the astronomical triangle, and all are beautifully illustrated by the solarometer.
The Nautical Almanac gives the right ascension and declination of a heavenly body on a circle which passes through the poles. The book of azimuth tables for the same time and place gives the position of the same body or a circle which passes through the observer's zenith. These two publications give the exact position of the same body on two different circles for the same time and place. If the body is on two circles at the same time it must be at their intersection, and if a telescope be fixed at this intersection of the circles representing those of the astronomical triangle, it follows that a body cannot be seen in the axis of that telescope without making this system of circles show the hour angle, elevation of the pole and azimuth, or the observer's longitude, latitude and the ship's compass errors.
This is not a new principle in astronomy but merely a novel point of view, involving a mechanical movement in unison of two variable systems of coordinates, viz.: declination and hour angle, with latitude and azimuth, having their junction marked by the body's zenith distance or altitude. The axis of the telescope lies in the radius of the concentric circles, and that radius makes an angle with the artificial horizon equal to the altitude of the body observed. This mechanical arrangement of the circles is such that the movement in unison of the declination circle around the poles with the azimuth circle around the zenith makes the axis of the telescope follow the apparent path of the heavenly body observed in the sky from its rising, to the meridian, to setting.
The solarometer is an instrument mounted on a constant level base so arranged that the motion of the ship will not be communicated to the instrument. The constant level base consists of a metal stand about 30 inches in diameter and 4 feet high, with openings underneath to admit an electric storage battery and other implements. On top within the stand there is supported a large cast iron bowl on gimbals with a forged steel ring. The cast iron bowl is hemispherical. The bottom of the bowl has an extra mass of metal cast with it so as to make it swing on its gimbals with a pendulum effect. The bowl is lined with porcelain and contains mercury and a float. The float is made in two parts and consists of two cast iron bowls, concentric hemispheres. The inner bowl is filled with lead and carries a steel column, rigidly secured at the centre of the bottom of the float and accurately fitted at right angles to the flat surface of the bottom of the inner float. This steel column projects three inches above the horizontal plane of the top of the bowl. In the center of the bottom of the inner float there is a small spherical recess to admit the ball of a ball-and-socket bolt by which the two bowls constituting the compound float are joined. The bottom of the outer bowl of the compound float is also flat and has a socket in its center to admit the lower ball of the ball-and-socket bolt. The bottom part of this socket is closed by a wrought iron plate whose exterior surface completes the spherical shape of the bottom of the compound float leaving a rectangular horizontal space in the bottom. There are two sets of horizontal rectangular spaces in the sides of the inner bowl of the compound float. All of these horizontal rectangular spaces in the different parts of the compound float are made accurately at right angles to the steel column in the center of the inner bowl, with the object of securing a horizontal flotation of the compound float in the mercury and carrying the steel column absolutely vertical. The instrument is carried upon the steel column.
The experience with the solarometer on the American mail steamer New York has demonstrated that this arrangement of the compound float will not operate onboard of high powered vessels, where the vibrations are excessive. In the North German Lloyd steamer Weimar all the vibrations were compensated by this arrangement of floats united together by a ball-and-socket bolt.
The object of making this compound float was to increase the weight of the float by the mercury it contained and to compensate for the vibrations that might be communicated to the mercury in the outer bowl.
It was plainly evident that when the mercury in the outer bowl was agitated as much as the surface of water when boiling the various impulses coming from the agitated mercury were communicated to the inner bowl and gave that bowl a tendency to move in all directions at the same time, and it therefore did not move in any. Any one who agitated the surface of the mercury in the outer bowl would see that the mercury in the inner bowl was not disturbed, and be convinced.
But these concerned only the effect of the vibrations horizontally while the effect vertically remained to be seen by actual experience when mounted on a high powered steamer with excessive vibrations.
By examination and experience on board of the steamer New York on a run from New York to Newport News, on January 20, 1895, the vertical vibrations were found to be communicated to the float with double effect.
The impulses vertically upward appear to be fully annulled by freely rising against the air, but the impulses downward instead of meeting each other in their coarse along the curved spherical sides of the float and bowls and becoming neutralized, because equal and opposite, met the ball-and-socket bolt and imparted motion to it.
The ball-and-socket bolt was more or less inclined from the vertical and the vibrating impulses coming down on one side met that bolt sooner than those which came down on the opposite side. This bolt then received the successive shocks on one side and alternately received the shocks on the opposite sides. The bolt then communicated all these shocks to the float with double the effect.
These shocks were irregular and made it impossible to determine when the sun or star was in the axis of the telescope for a period of 20 seconds of time.
The ball-and-socket bolt was removed, the vibrations were absorbed and the objects seen in the telescope no longer jumped across the field of view. Before the bolt was removed the inner float had a constant tendency to revolve in the bowl so that the bolt, upon receiving these vertical pressures, acted somewhat like a rudder.
The special function of the ball-and-socket bolt is obtained by small balls or rubber tubing floating on the surface of the mercury. Larger balls are also floated upon the mercury of the outer bowl. These balls keep the float and inner bowl nearly concentric; they do not carry any of the shocks due to the vibrations and they avoid the sudden send of the float when the ship strikes a huge wave in a heavy sea. The ball-and-socket bolt will no longer be used.
This arrangement of the float in the gimballed bowl provides not only for preventing the motion of the ship from being communicated to the instrument, but also insures the instrument being carried absolutely vertical. The gimballed support of the outer bowl and its pendulum motion compensates for most of the motion due to the rolling and pitching of the ship. The vibrations caused by the throbbing of the engines are absorbed by the mercury in the inner bowl.
The principle involved by using a float in mercury is that which Prof. Michelson used in his apparatus to determine the velocity of light. He claims that the truest level that can be obtained under any circumstances for accurate astronomical work is that secured by floating a block of granite in a tank of mercury.
The float is free to move about in the mercury. The top surface of the inner part of the float is covered with a compass-rose, marked to quarter points and graduated in degrees from zero at north to 180 degrees south in each hemisphere. The top edge of the outer bowl is marked with lubber lines indicating the plane of the ship's keel. At these two points in the plane of the keel of the ship, or parallel therewith, are two pointers, which are hinged and lightly rest upon the compass-rose to indicate the direction in which the ship is headed.
As the compass-rose on the inner float will be carried about so that its center will rarely ever be exactly coincident with the center of the bowl, it is necessary to note the position of the pointers of the lubber line and transfer the plane of the lubber line to the plane passing through the center of the float as is explained by reference to the diagram of the compass-rose which contains the following instructions : "At the instant of an observation, note the points of the compass-rose that lie in the plane of the ship's keel, or of the lubber line.
"If diametrically opposite points are in that plane, the forward point will show the true direction in which the ship is heading at that instant. If the plane of the lubber line passes through points not diametrically opposite, apply to the forward point the half difference between 180 degrees and the sum of those points in degrees. This corrected reading of the forward point will be the true direction of the ship's head."
Owing to the peculiar notation of the degrees on this compass rose constantly increasing from zero to 180 degrees from left to right, and then decreasing to 170 degrees on the other side, there is an apparent exception to the rule. For instance, in case the lubber line should pass through a line north 10 degrees east and north 175 degrees east, the half difference between 180 degrees and the sum of these two points would be 2 ½ degrees; and the forward reading would be north 7 ½ degrees east, and the after reading would be north 177 ½ degrees east These are evidently not diametrically opposite points. In this and similar cases, where the plane of the lubber line runs nearly north and south, one point should be read north so many degrees east, and the other north so many degrees west, or in this instance one would be read north 10 degrees east and the other north 185 degrees west, and according to the rule, half the difference between 180 and 195 degrees, or 7 ½ degrees, would give the ship's head; in this case, north 2 ½ east. The compass-rose has the figure 190 on the inner circle opposite 170, to indicate where the notation should be greater than 180 degrees, and the rule given on the compass-rose is thus made to apply under all circumstances.
The spherical shape of the bowls and float has a beneficial effect in compensating for the motion of the ship. The rolling and pitching makes the bowl move around the mercury. The weight of the mass of mercury, its consequent inertia, and the almost frictionless contact it makes with the porcelain surfaces tend still further to keep the mercury at rest. The shape of the displaced volume of mercury in the outer bowl is spherical, no matter how the bowl may move around the mass of mercury, and hence all currents that might be otherwise set up in the mercury are obviated, and the use of mercury is peculiarly well adapted for this purpose. It must be remembered that the bulk of the motion of the ship is compensated by the gimballed support of the bowl, and the percentage of that motion remaining to be compensated by the mercury is very small.
The only motion of the ship that cannot be compensated is that due to yawing or bad steering. Practice at sea, however, will soon enable an observer to follow that motion by swinging the float correspondingly. With the vibrations due to the engines there is an effect upon the observer's eye which causes the image of the body observed to oscillate slowly in the field of the telescope. In the American Line steamer New York the solarometer was mounted on a light steel deck within six feet of a windlass on the same deck. The vibrations of that deck when that windlass was in operation were extremely violent, but the image of the sun's disc always remained perfectly clear and well defined, but the disc moved about 1/8 of its diameter back and forth across the cross-hairs in the telescope, requiring a little judgment to determine when the disc was in the axis of the telescope. On one occasion, in the U.S.S. Montgomery, a simple float without the extra bowl and inner mercury was used, and then the vibrations of the engines at times made the sun's disc so indistinct that it could not be observed.
In taking observations, the float is necessarily disturbed by handling the instrument, but it soon comes to rest, and the evidence that it is at rest is plain from the observation of the heavenly body at rest in the telescope. In the first designs of the instrument, the horizon circle had eight spirit-level bulbs counter-sunk in the metal, but the inertia of the spirit bulbs was such that they continued to oscillate after the body observed was seen to be perfectly at rest in the telescope.
The instrument mounted on the constant level base consists of five circles with a supporting bracket and a telescope. The five circles represent the meridian, equator, declination, horizon, and latitude.
The meridian circle M and the equatorial circle E are joined rigidly at right angles to each other, and each has a pair of trunnions at points 90 degrees from their junctions. A hemispherical bracket A, with a hollow cylindrical sleeve, fits on the vertical spindle of the float and has bearing plates in its upper extremities to receive the trunnions of the equatorial circle E. These bearings are adjustable vertically and laterally by means of adjusting screws underneath and on the side of the bearing blocks. The trunnions of the equatorial circle E are diametrically opposite at 6 hours or 90 degrees from the plane of the meridian circle. It is graduated into hours, minutes, and to 30 seconds of time from zero to 12 hours on each side of the meridian.
One of the equatorial trunnions is hollow to admit a tube that projects inward and carries two horizontal arms with vernier plates to lie against the graduated surface of the meridian circle M. A pendant arm is attached to the outer end of the tube carrying these verniers, and has near its lower end bearings against two adjusting screws fixed within the frame of the supporting bracket A. On the opposite side there is a vise-clamp to secure the trunnion of the equatorial circle E in any fixed position. A pendant arm on this trunnion fits in between the frame of the bracket A, and abuts against a spring and micrometer screw to permit small movement of the clamped trunnion and to accurately set the equatorial and meridian circles to any desired elevation of the pole of the meridian circle M.
The meridian circle 31 is graduated to degrees, minutes and seconds of arc, from zero to 90 degrees in each quadrant.
The declination circle D is concentric with the meridian circle M and revolves on the trunnions at the poles around the equatorial circle E. This circle is graduated into degrees, minutes and seconds of arc. It carries a covering ring that supports the opposing verniers on the graduated face of the declination circle. A sliding block on the circle carries the verniers 180 degrees apart. A block with set screw and tangent screw permits accurate setting of the declination verniers at any desired point. The graduations are from zero at the equator to 90 degrees at the poles.
In the radial plane of the zero of the upper declination vernier block there is a socket tube which receives a pendant pin from the telescope block on the bracket K.
The declination circle revolves around the poles of the meridian circle and over the face of the graduated equatorial circle E. Two verniers are attached to the sides of the declination circle, on opposite side of the equatorial circle. These verniers fit close to the graduations of the equatorial circle and enable those graduations to be read to one second of time. The notation of the graduated equatorial circle is such that the zero of the vernier on one side of the declination circle comes opposite the mark for 12 hours when the vertical plane passes through the meridian circle and the common center of all. The numerical notation of the graduation is simply slewed around to accommodate the space required for the metal of the declination circle.
A graduated horizon circle is attached to the bottom of the cylindrical sleeve of the supporting bracket A. Above this circle a hemispherical bracket B revolves around that sleeve and carries two flat arms with opposing vernier plates to move over the graduations of the horizontal circle Z. The upper extremities of the hemispherical bracket B have bearing shafts supporting the hemispherical bracket K which revolves vertically from the horizon to the zenith, at the same time that the supporting bracket B revolves around the horizon circle Z at the bottom of the instrument.
A block containing a right prism is centered in the telescope bracket K in the radial plane of the common center of all the system of circles. The block has three circular openings for the tubes of the telescope. One tube having the object lens has its axis accurately fitted in the radial plane of the concentric circles. A second tube at right angles to the first and adjustable carries the eye piece and reticule of cross-hairs. The tube can be focused for observations of the sun and stars. The eye piece can be fitted with adjustable shade glasses and a spiral slot-way focuses the reticule of cross-hairs in the eye piece. A third tube at right angles to the first and on the opposite side of the circle carries a small incandescent electric light of ½ candle power, 4 volts and one ampere. This electric lamp is in a tube with ventilating holes and a frosted glass shade to send a diffused light through the prism across to the reticule of cross-hairs, which is illuminated without affecting the clear visibility of the star that is being observed. A sliding rheostat in the circuit of the lamp controls the degree of illumination. Underneath the telescope block in the radial plane of the axis of the telescope there is a pendant pin that fits in the socket tube of the vernier block of the declination circle D. This pendant pin is in the radial plane of the axis of the telescope and the common center of all the circles. The tube of the telescope and the pendant pin are all fitted with adjusting screws to be mounted accurately.
At one of the poles there is a clamp screw and micrometer screw which confines the declination circle in any position. A clamp and micrometer screw are also attached to the revolving sleeve at the bottom of the hemispherical bracket B, so that this bracket may be accurately confined at any desired position in azimuth. A clinometer is attached to one of the compensating weighted arms of the telescope bracket K. This clinometer is graduated to degrees and merely serves to indicate the approximate altitude of the body observed, for the purpose of making allowance for the refraction of the atmosphere. The horizon circle at the bottom is not concentric with the other circles, but is parallel to the horizontal plane which passes through the common center of all the circles.
In one of the first designs of the solarometer, this horizon or azimuth circle was placed in the horizontal plane of the center of the concentric circles. In that design, the vertical circle passing through the zenith was a hemispherical arc supported by sliding shoes on the horizontal circle. It carried the telescope block which moved in altitude by sliding up and down on that circle. This has been replaced by the telescope bracket K, which occupies a position at right angles to that formerly occupied by the vertical circle. This obviates the friction of the block moving up and down the vertical circle, and also obviates the friction of the sliding shoes on the horizon circle. The former movement is replaced by the shaft bearings of the bracket arm B, and the latter movement by the revolution of that bracket B on the sleeve. The original design was a more perfect representation of the actual conditions of the astronomical problem. The substitution affords a better mechanical arrangement.
The union of the several concentric circles capable of moving in unison around their own shaft bearings produces a peculiar series of motions that can only be clearly understood by an examination of the instrument in all its positions. The movement of the bracket B in azimuth causes the declination circle to revolve around the pole while the telescope bracket revolves in altitude in its shaft bearings and around the pendant pin of the declination vernier block. Four different revolutions are thus involved, ma king the several circles combine to cause the axis of the telescope in its motion to point and follow the apparent path of the heavenly body that may be observed in the sky.
All the bearings and movable parts of the instrument are adjustable. All the graduations are on silver, and each circle has opposing verniers, so that if there be any eccentricity it may be eliminated by taking the mean of the readings of the two verniers.
The circles are made of composition gun-metal; they are channel-barred with strengthening blocks at junctions. The instrument is as light as possible consistent with the requisite strength for durability.
In the earlier designs, the instruments were made of alloyed aluminum. Different alloys were tried and none were found to be suitable. At sea the aluminum seemed to absorb salt from the air; at least small white crystals were found on the rings every day, while none could be seen on the yellow metal of certain parts. The oxidization of aluminum is white, and these crystals may have been particles of oxidized aluminum, but they had a salty taste, were easily removed, and were found in greater quantities at sea than in port away from the sea air. The hard silver aluminum alloy was also found to be too soft for the bearings, and it was necessary to bouch the bearings in that metal in the shop after the limited use of the instrument for observations in Fauth & Go's yard in Washington. Experiments were made with various aluminum alloys in a simple salt water bath, and after three months in the bath they were pitted as much as a zinc rod in a Le Clanche cell, with a year's service. This experience has conclusively demonstrated the worthlessness of such alloys of aluminum as were tried, for any instrument of precision. The weight of the solarometer made of aluminum alloy was 21 pounds, or only 7 pounds less than the weight of the present channel-barred composition gun-metal instrument.
In making the broad statement that an observer can accurately determine a vessel's position and compass error by the solarometer at all hours when anything is visible in the sky, there are certain limitations. It is not claimed that anything can be done with the solarometer which may not be solved by mathematics.
In observing bodies on the prime vertical when the effect of an error in the latitude has but little or no apparent effect on the computed hour angle, it will be extremely difficult to find the latitude by the solarometer at such a time, and in the same manner the time and azimuth will be extremely difficult to determine accurately by observations of a body crossing the meridian near the zenith.
As the solarometer must be supported at some point on the equatorial or meridian circle, it has been found to be least disadvantageous to support it at six hours from the meridian. These supports prevent observations from being taken at these hour-angles, but as the altitudes will be low and the effect of refraction large, the time is unfavorable.
But these limitations do not apply if the body is at any appreciable distance from those directions, and the times when the observations of visible bodies are not advantageous are comparatively so few that, generally speaking, the broad statement ought to be admitted.
To use the instrument the observer has a chronometer regulated to Greenwich mean time, a nautical almanac, book of azimuth tables, refraction tables, and blank record books with forms for observations of the sun and stars.
METHOD OF OBSERVING.
For Observations of the Sun.—Find in the nautical almanac the sun's declination and equation of time for the Greenwich time of the observation, adjust the vernier of the declination circle D to the declination for the Greenwich apparent time. Set the latitude vernier to the approximate latitude of the place. Point the telescope to the sun by revolving the telescope arm and bracket to its altitude, and swing the float in the bowl, to find the sun's disc in the axis of the telescope. The axis of the telescope should be set a little west of the sun, and the instrument allowed to come to rest on its float while waiting for the sun to move westward in azimuth and altitude to appear in the exact axis of the telescope. Note the time by chronometer (or watch compared with the chronometer regulated to Greenwich mean time) the instant the sun is in the axis of the telescope. The arrangement of cross hairs circumscribing the sun's disc enables the observer to readily determine when the center of the sun's disc is exactly in the axis of the telescope. At the same instant the position of the pointers in the plane of the lubber line must be noted on the compass-rose of the float.
This observation with an assumed latitude is a correct altitude of the sun's center, but the hour angle would be in error corresponding to the error in the assumed latitude, precisely as in an ordinary time sight taken with a sextant if worked with an erroneous latitude. When the latitude is not known, it is therefore necessary to take a series of observations to determine both the latitude and hour angle, or the true local apparent time. Proceed then to take a second observation, setting the latitude 10, 20, or 30 minutes (according to the amount in which the latitude may be in doubt) north of the latitude used at the first observation, and take a second observation as before, noting the chronometer time, hour angle and azimuth for record. After recording, take a third observation, setting the latitude 10, 20, or 30 minutes south (according to the amount in which the latitude of the first observation may be in doubt) of the latitude used in the first observation, note the chronometer time, the position of the pointers of the lubber line at the instant of the observation, and read the hour angle and azimuth for record. The three observations with these different latitudes must then be compared to determine the true latitude.
Take the difference between the chronometer times of the first and second observations and the difference between the solarometer hour angles of those observations; find the second difference between these two intervals.
Take the difference between the chronometer times of the first and third observations and the difference between the solarometer hour angles of those observations; find the second difference between these two intervals.
Take the difference between the latitudes of the second and third observations, or the difference between the extreme latitudes used.
We then have the proportion that, as the sum of the second differences is to the second difference of the first and second observations, so is the extreme difference between the latitudes used to the correction to be applied to the latitude used for the second observation. Similarly as the sum of the second differences is to the second difference of the first and third observations, so is the extreme difference of latitude used to the correction to be applied to the latitude used in the third observation.
Each of these proportions will give the true latitude. Proceed to take a fourth observation, using the true latitude thus found to find the true hour angle. In every case note the change in latitude due to the ship's run in the interval of observing and make proper allowance for that change.
The following series of three observations by the solarometer will explain this subject clearly.
Chronometer Time. | Hour Angles. | Latitude used. |
1st observation, 10h 7m 35s | 11h 21m 20s | 38° 55' N. |
2nd observation, 10h 10m 19s | 11h 28m 30s | 39° 10' N. |
3rd observation, 10h 14m 57s | 11h 22m 5s | 38°30' N. |
Comparing the first and second observations, the difference between chronometer times equals 2 m., 44 s., or 164 s., and that of the hour angles equals 7 m., 10 s., or 430 s.; the difference between these intervals is 430s. — 164 s., or 266 s.
Comparing the first and third observations, the difference between chronometer times equals 7 m. 22s., or 442 s.; that between the hour angles 45 s.; and the difference between these intervals is 442s.—45s, or 397s. The sum of the second-differences is 266 + 397 = 663.
We have then the proportion that, as the sum of the second differences is to the difference of intervals of the first and second observations, so is the difference between the highest and lowest latitude used to the correction to be applied to the latitude used in the second observation; or, as the sum of the second-differences is to the difference in the intervals of the first and third observations, so is the difference between the highest and lowest latitudes used to the correction to be applied to the latitude used in the third observation.
In this example we have the two proportions, viz.
663:266 = 40:16'; and 39° 10' — 16' = 38° 54'.
663:397 = 40:24'; and 38° 30' + 24 = 38° 54'.
In this example, the latitude used in the first example was nearly correct, and it will be noticed that when, as in the second observation, the latitude was too high, the difference between the solarometer hour angles was greater than that between the chronometer times. And again, when the latitude used is lower than the true latitude, as in the third observation, the difference between the solarometer hour angles is less than that between the chronometer times.
It follows as a rule that in any series of observations of the same body, if the difference between the solarometer hour angles is greater than that between the chronometer times the latitude used is too high; and if the difference between the solarometer hour angles is less than that between the chronometer times the latitude used is too low.
By keeping this rule in mind, a skillful observer will soon be able to discover error in the assumed latitude and find the correct latitude and the corresponding correct hour angle readily.
Having found the latitude, take a fourth observation with that latitude and find the hour angle as before, which will be the exact local time. These four observations, which can with practice readily be taken in 10 or 15 minutes, determine with an accuracy within two miles the observer's latitude and local apparent time, and hence his longitude; and by comparing the compass course at the instant of the last observation with that indicated by the lubber line on the compass-rose of the solarometer, the error of the compass on that course will be accurately ascertained.
This method of determining the latitude and hour angle is based upon the fact that by the movement of the telescope in altitude and azimuth, the altitude is affected by the elevation of the pole and the distance of the telescope from the meridian. If the latitude is too high or too low, the movement of the telescope will not follow the plane of the path of the movement of the body in the sky, but will incline thereto, either higher or lower, according to the error in the elevation of the pole, or the latitude used, and the differences between the chronometer time interval and the solarometer time interval, indicated by the hour angles, show the deviation of the movement of the telescope from the plane of the path of the body in the sky.
Another method for determining the latitude and hour angle is to take one observation with an assumed latitude, and then take a second observation at five or ten minutes later; if the hour angles differ exactly five or ten minutes, then the assumed latitude is the true latitude. If they differ unequally, set the latitude north or south accordingly; take a third observation, and five or ten minutes later take a fourth observation; if the chronometer time intervals and the difference between the hour angles of the third and fourth observations are the same, the latitude last used is correct, and so is likewise the hour angle of the last observation. If the chronometer time intervals and hour angle intervals of the third and fourth observations do not correspond, correct the latitude again, corresponding to the difference, and take two more observations, and proceed in this manner until both the latitude and hour angle are correctly obtained.
In taking and reading the observations, the opposing verniers should generally both be read. If there is any discrepancy between the readings of the two opposing verniers on the same circle, the mean of the two must be taken to eliminate any eccentricity caused by dust, etc., in the bearings around which the circles revolve.
In taking observations of the sun at low altitudes, it is necessary to allow for refraction. A clinometer is attached to one prolonged arm of the telescope bracket K to indicate the angular altitude of the axis of the telescope. Refraction causes the sun to appear higher than it really is, and the observer should observe the center of the sun in its true position and not in the position where it appears elevated by the refraction of the atmosphere. Tables showing the effect of refraction at various altitudes are given in the books of navigation, but in observing with the solarometer the observer may apply this correction in the altitude when observing, and to do this, he must observe a star or the center of the sun's disc as much below the axis of the telescope as the refraction makes it appear above its true altitude. The small central square in the axis of the telescope is a square whose dimensions are equal to two minutes and forty seconds (2' 40"), which is the refraction for an altitude of 20 degrees. One half of that space is the refraction for 36 degrees, and two spaces would be the refraction for an altitude of 10 degrees. The clinometer shows the altitude approximately; and, with practice, a skillful observer can allow for the effect of refraction at different altitudes and varying conditions of the atmosphere.
This method of observing the sun or a star as much below the axis of the telescope as the refracted rays of light make it appear above its true position is theoretically correct, but it is difficult to introduce into practice.
When the amount of the refraction is known, it is much more difficult to allow for it than to observe the body exactly in the axis of the telescope, and the effect of neglecting or accurately allowing for the refraction is much greater than is generally supposed. The habit of applying refraction to observed sextant altitudes is so fixed that the effect of neglecting it has not been generally considered by practical men.
Tables are in preparation showing the effect of refraction on the hour angle at different altitudes for different latitudes and polar distances. These tables have so far been completed only for the latitudes of the transatlantic steamer routes, and they show regularity in the decreasing error of the hour angle up to within a certain azimuth, when the error begins to increase and reaches a second maximum close to the meridian.
This peculiarity is explained by the fact that the change in altitude is greatest on or near the prime vertical, and that when on or near the meridian, it takes the body much longer to change its altitude by the small amount of the refraction, than when nearer the prime vertical to change very much more in altitude.
The table for a latitude of 40 degrees north, polar distance of 100 degrees, shows an error of 30 seconds in time at an altitude of 10 degrees, when the hour angle read from the solarometer was 4 hrs., 31 min., 8 sec, without allowing for refraction of 5' 20", azimuth 112° 33' instead of 112° 29'. At an altitude of 20 degrees, azimuth 123° 5', the hour angle was 16 seconds too small. At an altitude of 30 degrees, azimuth 138° 11', the hour angle was 12 seconds too small. At greater altitudes the error in the hour angle began to increase until at an altitude of 39 degrees, azimuth 166° 14', it was 18 seconds too small; 43 min., 22 sec, from the meridian instead of 43 min., 40 sec. The error in hour angle near the meridian is much greater and in a measure indeterminate.
Observations for time, on or near the meridian are not reliable, especially under these circumstances, and the effect of refraction must then be considered almost entirely upon the latitude.
By setting the declination at a polar distance decreased by the refraction multiplied by the cosine of the hour angle, a fair degree of accuracy will be obtained by observing the body in the axis of the telescope to find the latitude on or near the meridian.
In using these tables, the observer notes the chronometer time when the body is exactly in the axis of the telescope and the position of the plane of the lubber line on the compass-rose. He then reads the azimuth, hour angle, and altitude from the clinometer attached, and corrects the hour angle for the refraction according to the correction given in the tables. He then sets the declination circle to the correct hour angle and finds the true azimuth corresponding thereto from the azimuth circle. The difference between the azimuth observed and the correct azimuth must be applied to the compass-rose to find the corrected heading of the ship.
In finding both latitude and longitude from a series of observations by second differences, each of the observations should be corrected and second differences worked from the corrected hour angles to find the true latitude, as explained.
The effect of parallax is so small that it is ignored in observations by the solarometer; observations of the moon are not recommended on account of its horizontal parallax.
OBSERVATIONS OF STARS.
For observations of stars, a small electric lamp shining through frosted glass, illuminates the cross hairs in the eye-piece without affecting the clear visibility of the star in the telescope. The magnifying power of the telescope is such that the usual bright navigational stars are clearly distinguishable, more so than with the naked eye.
For observations of a star, find in the Nautical Almanac the star's right ascension and declination, and the right ascension of the mean sun, all of which are to be corrected for the Greenwich time of the observation. Set the declination vernier on the declination circle D to the declination of the star, and the latitude vernier to the approximate latitude of the observer. Revolve the telescope in altitude and azimuth, and find the star in the axis of the telescope in the same manner as described for observations of the sun. When the latitude is not known, take a series of observations in the same manner as described for observations of the sun.
The facility with which a star can be found in the field of the telescope is remarkable. It is much easier to find the star to be observed in the field of the telescope than to point a spy-glass to see a star when standing on land, especially if the declination circle is set to the approximate hour angle.
To read the arcs with facility, a small hand telescope is provided in which there is a side tube containing a miniature electric lamp, which throws a light from within the reading telescope through its object lens on the vernier, which is thereby clearly illuminated.
In recording the observations, the four quantities—declination, latitude, hour angle, and azimuth—as read from the solarometer should be compared with those same four quantities in the book of azimuth tables, enlarged and extended. If there is exact correspondence between the observed four quantities read from the solarometer and those computed in the book of azimuth tables, the observer has positive proof that his result is accurate.
For observations of Polaris, or the polar star, note the right ascension and declination of Polaris in the Nautical Almanac, and see if the right ascension is such that the star is above or below the pole. If the star is above the pole, set the declination circle vernier at the north declination 88° 44' 15". But if the star is below the pole, set the declination vernier at its sub-polar declination, 91° 15' 45", or as given in the Almanac.
If the star's hour angle is not approximately known, it may be readily ascertained by adding the right ascension of the mean sun (as given in the Almanac and corrected for the Greenwich mean time) to the approximate local mean time lo find the right ascension of the meridian; then add or subtract the star's right ascension from the right ascension of the meridian to get the star's hour angle.
As the declination circle revolves on the polar axis the observer must not attempt to revolve the declination circle by means of the telescope bracket or the azimuth clamp when the telescope is clamped by the declination set-screw near the pole. In this position the change in azimuth can be but very small, and the leverage of the telescope bracket is liable to throw the instrument out of adjustment. Any change in hour angle with the telescope near the pole must be made by turning the declination circle in hour angle directly, and allowing the movement of the declination circle to move the telescope bracket B in azimuth.
The clamp screw which sets the declination circle at any hour angle ought not to be used for any observations; it is provided for use in adjusting the instrument, and should not be used when taking observations.
Having set the declination of Polaris on the solarometer, and the latitude vernier to the observer's approximate latitude, put the declination circle to the approximate hour angle of Polaris and turn the telescope by swinging the float horizontally until the star is seen in the field of the telescope. Bring the star in the axis of the telescope by working the micrometer screw of the latitude clamp. This will give the altitude of Polaris, and, since the position of the declination circle in hour angle mechanically causes the pole of the solarometer to be as much elevated or depressed from the altitude of the north pole as the star's position in its path around the north pole deviates from the north pole, the pole of the solarometer will have the same elevation as the true north pole, and the vernier on the meridian circle will show the observer's true latitude.
Observations of Polaris by the solarometer have invariably been found to be exact, much more so than any meridian altitude of the sun by a sextant. It is claimed that a good observation of Polaris under favorable conditions will give the true latitude within one mile. Observations of Polaris where cross bearings of lighthouses were available, as in the Chesapeake bay, invariably indicated the latitude as accurately as those by cross bearings, and agreed perfectly.
DETAILS OF MANUFACTURE AND ADJUSTMENT.
In the manufacture of the solarometer all the circles are turned on the lathe and are made absolutely concentric. The equatorial circle E and the meridian circle M, after being securely united together at right angles to each other are revolved in the same lathe. These circles are graduated after they have been thoroughly tested to be concentric and accurately at right angles to each other. They are placed in the adjustable bearings on the hemispherical bracket A and fitted on the lathe. The declination circle is fitted on its polar trunnions and adjusted to revolve around its polar bearings, exactly concentric with the meridian and equatorial circle. The telescope supporting arm B is placed on the shaft under the hemispherical bracket B and adjusted to revolve concentrically around that shaft. The azimuth circle is fixed rigidly on the bottom of the shaft under these brackets and adjusted to be accurately in the horizontal plane at right angles to the vertical axis passing through the common center of the rings. The telescope bracket is fitted in its bearings on the ends of the hemispherical arm B in the horizontal plane passing through the common center of all the rings. The telescope block with its projecting pendant pin is fitted into the socket tube on the declination vernier block, and the telescope block is centered in the telescope bracket K. All parts are thus assembled and ready for final accurate adjustment.
APPURTENANCES.
Each solarometer is furnished with a set of implements, viz.:
First.—A reading telescope to be held in the hand. This is designed to be held close to the eye and close to the vernier to be read, and its magnifying power is such that the fine graduations of the verniers are clearly legible and sharply defined. It is made of aluminum and can be readily cleansed. For use at night, a small tube is set in on one side at an angle opening towards the object lens. This tube contains a miniature electric lamp of 1 candle power, 4 volts, and requires 1 ampere current. A small aluminum shield is set in the main tube of the reading telescope and prevents the electric light from being thrown into the observer's eye. The light clearly illuminates the vernier and enables it to be read with even better facility than by daylight; under circumstances, the observer will find it advantageous to use the electric light in day time. The arrangement is such that the lamp and its shield do not interfere with its use by daylight. All verniers are accessible to the reading telescope in all the various positions of the different circles.
Second.—Two miniature electric lamps are furnished for reserve supply in case of breakage.
Third.—A striding level to fit upon the bearings of the hemispherical arm A is furnished for use in adjusting the instrument, to see if it is carried absolutely vertically upon the float.
Fourth. —One spare set of spider cross hairs, and one reticule of cross lines engraved on glass. These reticules are not to be inserted or adjusted unless the vessel's position is accurately established. Though exceedingly delicate, it is rare that it will be necessary to use the spare reticules.
Fifth.—Two shade glasses, which fit over the eye-piece for the observations of the sun.
Sixth.—One screw-driver.
Seventh.—Two small steel capstan bars for making adjustments.
Eighth.—Two electric storage batteries, with suitable lengths of insulated conducting wire, switches, etc. One of these batteries is to be kept underneath, within the stand of the base. The other battery is charged and kept in reserve to replace the one in use within the stand when exhausted.
ADJUSTMENTS.
First.—Adjust the bearings of the equatorial circle F on the adjustable bearing on the horizontal plane of the hemispherical bracket A. The equatorial circle is placed horizontally, and the meridian circle exactly vertical. The bearings are adjusted with the equatorial circle perfectly horizontal, by means of the adjusting screws.
Second.—Clamp the declination circle D accurately in the vertical plane of the meridian circle. Loosen the latitude vise clamp V on the hemispherical bracket A and revolve the three circles E, M, and D, together in the bearings of the equatorial ring. In revolving, the declination circle D must always remain in the vertical plane of the meridian circle M, and it is so adjusted.
Third.—Place the declination circle in the plane of the meridian circle M with the pin of the telescope set in the declination vernier socket and adjust the telescope block so that the axis of the telescope is equidistant from the bearings of the telescope bracket K on the supporting hemispherical arm B. Loosen the declination clamp screw and revolve the telescope in altitude from the horizontal plane to the vertical and the horizontal plane on the opposite side, adjusting any deviation of the axis of the telescope from the vertical plane of the meridian circle.
Fourth.—Set the azimuth circle with its opposing verniers at zero and 180 degrees, loosen the declination clamp, and revolve the telescope in altitude, and adjust the bearings for any deviation of the hour angle from the plane of the meridian circle. The declination circle must, in this case, remain in the plane of the meridian circle, though unconfined. Any deviation will be corrected by adjusting the height of the telescope bracket bearings.
Fifth.—Point the telescope to a distant fixed point, and adjust its axes by revolving the loosened verniers and latitude clamps for various positions in the plane of the meridian circle M.
Sixth.—Place the instrument upon a level stand on a pier, put the axis of the telescope in the zenith of the instrument. A collimating tube with a cross hair is fixed above the instrument in the vertical plane of the axis of the center of the instrument. By looking in the eye-piece of the telescope, the collimating wire must appear in the center of the cross-hairs in all positions in which the arm B may be revolved.
Seventh.—The circles are set at every lo degrees of latitude and declination, and every 10 minutes of hour angle; and for all the various positions which the azimuth circle will occupy due to these combinations, the azimuth must coincide with the computed azimuth in the book of azimuth tables.
Eighth.—The instrument is mounted on a solid pier and observations are made of the heavenly bodies, and adjustments tested.
Ninth.—The instrument is next placed upon the float and secured thereon, with the plane passing through the zero and 180 degree points of the azimuth circle, exactly coincident with the plane passing through the north and south points of the compass rose on the float. A striding level is put upon the recesses in the hemispherical bracket A and the instrument is revolved in different positions in azimuth and altitude to determine if it is thoroughly counterbalanced in all of its parts, and is carried absolutely vertical by the float.
Tenth.—The instrument is finally mounted on its constant level base and tested by taking observations of the sun and stars at different declinations to determine the longitude and latitude of the observer, in the yard at Fauth & Co.'s works. The result must be exact.
In the manufacture of the stand, bowls and floats, careful and accurate mechanical workmanship is also requisite. The bearing blocks of hardened steel are accurately fitted, and the forged steel ring and bowl are accurately balanced on their gimballed bearings.
After the float is floated in the mercury the level flotation is tested by a sensitive spirit level centered on the spindle on which the instrument is mounted. The float swimming in the mercury is revolved, and after it ceases to oscillate the bulb of the long spirit level must come to perfect rest while the float continues spinning in the mercury perfectly horizontal. The long sensitive spirit level is placed in all positions in different horizontal planes on the spindle that supports the instrument, and the level flotation of the float is tested in all positions both at rest and when spinning freely in the plane. All the floats are adjusted to meet these requirements and the accuracy of the constant level base is established beyond all doubt.
Whenever the float is touched by the hands the level flotation is necessarily disturbed, but it comes back with remarkable celerity.
Actual experience is necessary to convince many of the fact that the motions of the ship are thoroughly compensated by the constant level base. When a huge wave strikes the ship, the float will be by its inertia thrown with some degree of violence against the sides of the bowl, but experience thus far obtained has shown that this rarely happens and only interferes with observing for a brief interval. The motion due to yawing and bad steering cannot be compensated; the helmsman should be cautioned to keep the ship steady on her course and much of that motion will usually be avoided. Any navigator will have no difficulty on this account, since all such yawing will necessarily be so limited that the heavenly body will not be lost to view in the field of the telescope, and an observer can easily judge by the regularity of the oscillations of the heavenly body when it is in the axis of the telescope. Such judgment is equally necessary in observations with a sextant and much more difficult than with the solarometer, whose cross hairs enable the observer to judge this with facility and accuracy.
On board the U.S. Cruiser San Francisco the solarometer, after having been adjusted, was raised from its spindle and a fresh compass-rose placed on top of the float. It was subsequently found to be out of adjustment, and the results of a great many observations obtained by Lieutenant A. Ward and other officers, while agreeing uniformly within a few seconds, were about 2 minutes and 18 seconds in error in longitude, and with this difference that the longitude given by the forenoon observations was too far east by two minutes and 18 seconds, while the longitude by the afternoon observations was too far west by the same amount.
The error arose either because the instrument did not set vertically, or the arms of the latitude verniers were not horizontal. The striding level is furnished with each solarometer to determine if the instrument is properly carried on the float, but upon examination this spirit bulb had become useless on account of the evaporation of the alcohol from an almost invisible crack in the graduated glass tube.
As the San Francisco was about to sail for Europe and the striding level could not be replaced in Newport, the horizontal latitude verniers were adjusted by the amount of the error in the latitude ascertained by a meridian altitude of the sun.
The error in the latitude illustrates the well known effect of error in latitude on the longitude, and at the same time demonstrates the accuracy of the instrument and its thorough accordance with the mathematical principles.
In the sketch let the line W A C B E represent the equator, the line NP C the meridian, the lines P T E and P T' W the plane of the declination circle with the pole at its proper elevation, and the lines N T B and N T' A the plane of the declination circle of the solarometer out of adjustment, having the pole at A' instead of being at P. The true hour angles were C P E and C P W for forenoon and afternoon observations. The solarometer out of adjustment gave those hour angles as C N B and C N A. T and T' indicate the altitudes of the sun at those observations.
The error in the elevation of the pole having caused this difference in the longitudes on different sides of the meridian. This effect is the same as that upon which Table XXXVIII of Bowditch is calculated; the effect of an error of one minute in the latitude upon the longitude.
In order to obtain accurate results, it is necessary to read the graduations and verniers accurately. If any of the circles are set carelessly and erroneously, it is absurd to expect correct results with any instrument of precision.
The instrument is, when once mounted, kept on its base constantly available for observations. It is not carried about, and therefore the most frequent cause of damage to instruments is obviated. Intelligent handling of the instrument to take observations cannot injure it, and its durability under these conditions is assured much more than is that of instruments which are packed in boxes and carried about after having been used.
THE SOLAROMETER DECK OBSERVATORY.
In order to protect the instrument from the wind and weather, and enable observations to be taken under all conditions of rolling and pitching at sea, a peculiar observatory is provided as shown in the illustration. This observatory consists of a sheet-metal cylinder six feet in diameter and twelve inches high, which is screwed by wood screws to the deck. The upper part consists of two parts joined together. The lower part is a cylinder six feet in diameter, fitted with rollers and clips to revolve on top of the concentric cylinder secured to the deck. This is three feet high and six feet in diameter. Upon this cylinder the upper part is supported. The upper cylinder is cylindrical, five feet in diameter, three feet high, with a flat conical top. The upper part sets with one of its surfaces in the vertical plane of the lower cylinder, so that the center of the upper cylindrical part is six inches in rear of the center of the lower cylinders. A door is made to fit in the rear surface of the two upper parts. To get inside, it is necessary for the observer to step over the lower cylinder that is secured to the deck. The front hemispherical section of the upper part is fitted with two movable shutters, which can be closed entirely or opened to any extent up to 180 degrees. On the ledge in front of the opening there is an adjustable wind screen which is three feet wide and eighteen inches high, fitted with two curved bracket arms so that it may be adjusted to any position from the vertical to the horizontal. Within the observatory, under the ledge, beneath the wind screen, there is a box to contain books and implements. The electric storage battery to furnish the light for the telescope and to read the circles is stowed within the stand of the instrument. The arrangement of the observatory permits an observer to observe a heavenly body that is abeam, with the ship rollings and the wind from abeam, so that the wind screen will protect the instrument from being heeled over out of the vertical plane, and the top of the observatory will not obscure the body being observed. Or in case the body to be observed is ahead, and the wind is ahead, while the ship is rolling deeply, the wind screen will protect the instrument from the wind, and the opening of the shutters will prevent the body from being obscured by the sides of the observatory. In this house, which can be revolved so that a body can be observed in any direction from the ship, the instrument is always available for observations.
In the U.S.S. San Francisco, the solarometer has a peculiar cover of the same dimensions as the top of the base. It is made to open at the top and drop down and rest on the rim of the base, and to turn on rollers around the top, using one of the hemispherical lids of the top for a wind screen.
This arrangement is much smaller than the regular deck observatory, but is inadequate when there is any wind. In the solarometer observatory, the observer may enter and close the door and when he opens the sliding wings of the slit there is no draught of wind blowing through, and no effect of the wind upon the instrument.
The primary object of the navy is to train officers and men to fight the ships in time of war, but while training, much useful hydrographic and meteorological work has been done by the navy, and such work will always be done in time of peace.
The solarometer will be found to be a most useful aid in such utilitarian work. For surveying coasts, investigations of ocean currents, running lines of sounding in surveys and for ocean cables, and every phase of scientific research so well done by the navy, it is evident that the solarometer meets a demand for accurately locating the ship's position and compass error astronomically at all hours of the day or night, independent of the visibility of the sea horizon.
PRECAUTIONS TO BE OBSERVED IN USING THE SOLAROMETER.
First.—The clamp screw on the shaft of the declination circle must not be used to set the declination circle at any hour angle except in the plane of the meridian circle at twelve hours when examining the adjustment of the instrument. When observing, this clamp must invariably be kept loose.
Second.—To change the altitude of the telescope the movement of the circles in azimuth, altitude and hour angle simultaneously must be done by moving the clamp block of the azimuth vernier in azimuth, except when the declination vernier is set near the pole, as is the case in observations of Polaris. In this case turn the declination circle and see that the azimuth clamp is loose.
Third.—The azimuth clamp and hour angle clamp on the shafts of the declination circle should be kept loose when changing the latitude or the elevation of the pole of the solarometer.
Fourth. —When not observing, the instrument should always be kept covered by its chamois skin cover to prevent dust, etc., from getting in on the bearings.
Fifth.—The float must not be used under any circumstances as a receptacle for any implements, pencils, knives, screw-drivers or adjusting bars, shades, etc.
Sixth.—The instrument may be cleaned by using a soft camel’s hair brush and piece of chamois skin. The instrument must be kept carefully dry and clean, but no vigorous rubbing to remove rust or oxidization on the surfaces of the circles should be attempted.
Seventh.—In case it should be necessary to dismount the instrument from the float, the set screw at the bottom, by which it is screwed to the column on the float, must be carefully unscrewed, and the instrument raised by grasping it under the azimuth circle and carefully twisting it slightly while raising it vertically. Before dismounting, the latitude should be set at zero, the declination at zero, and the hour angle in the plane of the meridian circle. The upper circles will then all be vertical, and all the clamp screws should be set taut.
To put it in its transporting box, the instrument should be lifted by grasping with both hands under the two hemispherical arms A and B. It should be lowered in the box vertically, holding it with both hands and fitting it on to the spindle in the bottom of the box. After the box is packed with the wooden side blocks, the space around the instrument should be filled with wads or bunches of ordinary newspapers. Excelsior and similar material is not suitable, as small particles are liable to get in the shaft bearings. Bunches of soft newspapers make the best packing in the box for transportation by rail.
The importance of a constant available means of determining the compass error on board the modern steel vessels cannot be too strongly emphasized. Experience shows that no compensation of the compasses for the magnetism of the ship, nor any determination of magnetic effect upon the compass, will hold good for any length of time, or for great changes of position. The most constant observations are necessary in order that any confidence can be reposed in the compass, and even then its indications must be regarded with suspicion.
The compass error is obtained at present by time azimuth of the sun, observed by the alidade on the standard compass. This compass is now (in most approved patterns) a liquid compass in a bowl on gimbals. The alidade is fitted with sight vanes and a prism in which the observer is with difficulty obliged to find the sun, and then note the instant with his watch or chronometer. The observation is made in connection with a sextant to find the local time, since to find the sun's true bearing it is necessary to know the local time or the sun's hour angle.
The same results are obtained by such observations of the moon or stars, but in order to ascertain the true bearing of the moon or star observed, it is necessary to know that body's hour angle. This hour angle can, however, only be ascertained by elaborate calculation from a sextant altitude of the heavenly body. Sextant altitudes at night are, as has been stated, impracticable, and consequently at present it is only during sunshining daylight that compass error can be ascertained.
The solarometer thus incidentally accomplishes what is sought to be done by both the sextant and alidade, with the advantage of being available at night or in foggy weather, and obviating the necessity of any elaborate calculation; and besides this has the great additional advantage that whatever may be the result indicated by the solarometer, the observer can always know positively if his observations and results are right or not.
The inventor takes this opportunity to express his grateful appreciation of the assistance of brother officers and others in developing and perfecting this instrument.
Mr. G.W. Gail, of Baltimore, Md., generously promoted the enterprise by financial aid. The U.S. Lighthouse Board gave opportunities to test the original design on the steamer Violet, Mr. Malster, the builder of the cruiser Montgomery, further tested it on her trial trips. The director of the North German Lloyd Steamship Company gave facilities for sea trials on two transatlantic voyages in the S.S. Weimar. The Maritime Association of New York allowed the free use of the Maritime Exchange for an exhibition and lectures on the solarometer to the maritime community of New York for a period of two weeks. The American Line of U.S. mail steamers gave free passage to guarantee officers to try the solarometer on two voyages. The Cunard Steamship Company and the French Compagnie Generale Transatlantique have ordered solarometers on trial. The United States Navy Department has encouraged its development throughout, and put one on board the cruiser San Francisco.
This gratifying experience demonstrates the interest and general desire for the success of an instrument which can do what is claimed for the solarometer.
A thorough discussion and critical examination of the details of the solarometer and its appurtenances is earnestly invited in order that any imperfections may be revealed, and a perfect modern navigating instrument be evolved.
DISCUSSION.
The Chairman:—I have listened with great interest and pleasure to the description of the solarometer by its inventor, Lieutenant Beehler. This beautiful instrument which he exhibits to us represents, as I know, his patient labor for many years, and I desire to extend to him my congratulations on his success. Any plan or device to aid the navigator touches us very nearly, and we must regard it as of the highest importance. Although the solarometer is still in an experimental stage, the favorable reports of its practical working give promise of its future usefulness. The point of superiority in the instrument which especially strikes me is that it enables the navigator to ascertain the ship's position by observations of a heavenly body when the horizon cannot be seen. This is done by measuring from the zenith which is determined by the instrument. This property of the solarometer gives it a wider field than the sextant, and makes it available in cases when the horizon is obscured and the sun is shining overhead. I should think it would be especially valuable on many occasions when vessels are approaching our coast in foggy weather, for at such times the sun often shines out for a few minutes although the horizon does not clear. It also increases the accuracy and value of night observations, which are usually uncertain on account of the badly defined horizon.
I am doubtful about the extreme accuracy which is claimed in the results of the observations with the solarometer, but even if it only enables the navigator to determine his position by astronomical observation within a few miles, when otherwise his only guide would be the dead reckoning, it will afford him most material aid in these times when the magnificent ships already afloat frequently run more than five hundred miles a day.
The fact that with the solarometer on board the American Line S.S. New York when she was proceeding at high speed, the instrument was unreliable on account of the vibrations, does not seem to me to be a serious objection. If it is important to determine a ship's position accurately, she can always be slowed down or stopped long enough to take an observation, and the result will probably be a saving of coal and time as well as a greater degree of safety.
If the solarometer proves to be a navigating instrument of practical value, with which a ship's position can be determined by astronomical observations with reasonable accuracy, when without its aid this could not be done, the inventor can justly feel that he has added greatly to the security of life and property on the ocean, and he will deserve the thank of all seafarers. I believe that a careful trial of the instrument under the conditions of actual service will prove its worth.
Commander Wm. Bainbridge-Hoff, U.S.N.:—There are now several refinements in the arts, applicable to ships and navigation, which permit such an instrument as is the solarometer to become of great value. In fact Mr. Beehler has gone a long way towards supplying a necessity which the increasing speed in ships demands.
The passing out of sails in vessels on the great trunk routes of the sea, the solving of the problem of steadiness of platform, and the ability to steer to within small angles by means of steam make the use of the solarometer possible. Now there are no sails or spars to interfere with an observer's vision. Great steadiness of platform is scientifically procured through our present knowledge of the value of the distribution of weights in ships, while mechanical helm appliances, together with the inertia of large ship masses, make steering much more accurate than formerly.
In the last thirty years the speed at sea of steam vessels has been doubled. In the next thirty years, this speed may be doubled or trebled. If this proves so, the voyager of 1925 may see a speed of a mile a minute at sea. Again, we find the 2000-ton ship of 1855 has become has become the 15,000-ton ship of 1895; will it not become in the year we have indicated, a vessel of at least 75,000 tons, and cost perhaps in the unnumbered millions of dollars, and hold a countless number of souls? If a ship of this size and speed becomes the packet of the future, it becomes then necessary to know hourly where the ship is, and when you are without a horizon for more than a third of the time of her run of three days through fog, haze or mist, and for half the time certainly through the phenomena of night, it makes such an instrument as the solarometer of priceless value. In these days there will br. carried an astronomer who, like a pilot, will be ever on the watch, while a "ticker" on the bridge or in the captain's room will keep the commander at all times advised of the ship's position.
Commander F.A. Cook, U.S.N:—It would be presumptuous for me to attempt to criticize, after a merely cursory examination, an instrument which has been perfected after years of patient and intelligent study and labor. I do not intend to do so, nor can I see wherein it can be criticized from my present knowledge of it.
The solarometer is a success, and has been so proven from experiment and use at sea. It is built upon a simple and plain theory. It is an astronomical triangle mounted upon a pedestal whose base is in the plane of the horizon, and its practical use depends upon its ability to maintain a constant level when mounted on the deck of a vessel at sea. It is in this ability that the ingenuity and patience of the designer is shown.
In clear weather the work of the navigator is truly "plain sailing." It is when the horizon is obscured that his anxiety begins and constantly increases as he approaches the land. Who that has navigated, but has often experienced the provoking and perplexing condition of a clear sky overhead, but an obscure and unreliable horizon? Just such conditions more often confront one in the vicinity of land. The solarometer meets these conditions perfectly, gives the position and corrects the compass. It is of inestimable value, and ought to be found upon all ocean craft. Unfamiliarity with its use should be no argument against it. The designer frankly admits that in extreme cases of vibration in high powered steamers, and in heavy seas the solarometer cannot keep its level. Under these conditions, if it did, it is quite certain most observers would not. The correct thing to do under such circumstances would be to stop for the observation,—a small concession indeed to make for so important a result.
I sincerely trust the inventor may receive his just reward by finding his instrument universally adopted and in successful use.
Commander A.D. Brown, U.S.N.—The mechanical solution (at sea) of the astronomical triangle has long been greatly desired by the navigator. Numerous instruments tor the purpose have been devised, but all have failed owing to the fact that the plane of the horizon could not be determined by them with a sufficient degree of accuracy. In the solarometer, as described by Lieutenant Beehler, this difficulty appears to have been entirely overcome. If the constant level of the base of this instrument can be preserved, as appears to be abundantly proven by the method of manufacture and the actual use of the machine, no further argument for its adoption would seem to be necessary. In these days of ships which are themselves huge floating magnets, the means of frequently and accurately determining the compass error has become an absolute necessity. This is unquestionably afforded by the solarometer, and its adaptation to stellar observations is a feature which should greatly commend it, as with its use it will be no longer necessary to go blindly along at high speed for ten or twelve hours every day without any means of ascertaining the true course and distance.
The navigators of the world are indebted to various officers of the U.S. Navy for many improvements in the exercise of their art, but it would appear to have been reserved for Lieutenant Beehler to devise an instrument which will render assurance doubly sure, will make the navigation of a ship a pleasure rather than a task, and will remove an immense weight from the commanding and navigating officers of all vessels that have what we may well call "the automatic position and compass error finder," as a portion of their equipment.
Lieut.-Commander J.G. Eaton, U.S.N.:—Although I have studied the solarometer with a view to offering some technical criticisms, I shall confine the few remarks I offer more to the practical side of the question and content myself with stating that I fully believe that the defects which now exist are entirely due to the difficulties of construction, and not to faults in the theory.
Every navigator, vexed and perplexed by dim horizons, must have prayed for some point d'appui more reliable than the shifting line where sky and waters meet. A visible zenith would prove a blessing to every sea observer who seeks to establish his position. Lieutenant Beehler has originated and developed in his solarometer an instrument which practically confers on the observer a horizon that neither land nor fog can obscure.
I have not seen the solarometer at work afloat, but I can speak of its accuracy on shore, and the results laid before you in this paper will show you that the instrument is reliable, and that by its use we may eliminate the unreliable true horizon, and gives in place thereof, an always determinable circle of reference.
The mechanical portions of the instrument are so well adjusted that even the preliminary trials have shown great accuracy. The results on the sea show that the results may be relied upon within the limits of the usual errors of observation. The correction of the errors due to refraction presents the greatest difficulty. The true path of a heavenly body, and this is the only path for which the coordinated circles can be used, must always differ from the apparent paths due to refraction. The empirical method of correction, that of using the squares of the reticule, appears crude and unsatisfactory for refined observations. No better method presents itself, but the use of the one advocated must detract from the value of the observation.
Unquestionably, as the solarometer becomes generally used, slight changes in sections, conducing to greater steadiness, with consequent facility of observation will suggest themselves. The instrument is still, though developed, in its experimental stage, and faults due to design and want of poise will be corrected as they are recognized.
As it stands to-day, the solarometer marks a wide departure from our present methods, and its results, creditable as they are, fail to show its great superiority over horizon sextant observations.
Lieutenant Beehler has invented, and I may add, well nigh perfected, an instrument which bids fair to revolutionize the accustomed methods of ascertaining a ship's position at sea. I do not regard it as essential, or indeed wholly practicable, to determine the latitude and longitude by single observations. But as to the essential points, the solarometer can do all that the sextant can, and besides, can be accurately used where the sextant, deprived of the horizon, is useless.
The great value of such an instrument to a man-of-war will be appreciated. The importance of correctly establishing the ship's position at any time, day or night, has greatly increased with the speeds now in use. Cases will readily occur to naval officers, where swift descents upon an enemy's coast or fleet must depend for their success, upon the correctness of the departure. Any instrument of navigation which eliminates the uncertainties of the day horizon, and gives us one at night which can be relied upon, will prove of inestimable value.
I fully believe that the solarometer does, or will do these very things.
Lieut.-Commander Richard Wainwright, U.S.N.:—The advantages of the solarometer as may be gathered from Lieutenant Beehler's paper, are the constant level or base, which provides a means of taking observations independently of the visibility of the sea horizon, and the use of a mechanical means of solving the problems in place of solving them mathematically.
Many attempts have been made to solve the problem of ascertaining the latitude and longitude at sea when the horizon was obscured or invisible. Several attachments to the sextant have been patented, but none have proved very successful when brought into practice, although correct in theory. If Mr. Beehler has solved the problem and his instrument is capable of being handled practically, as well as being theoretically correct, his invention will be of great value to the mariner. The difficulty heretofore has been to handle a necessarily delicate instrument with sufficient skill to produce good results under the different conditions prevailing at sea. Some of us may remember that Professor Chauvenet held that there was no reason why as exact observations should not be taken with a sea horizon as with an artificial one. He stuck to his proposition until on a practice cruise, when he undertook to take sextant observations while the vessel in which he was cruising was under the influence of a choppy sea. Then he expressed doubts as to the possibility of ever getting reliable observations at sea.
Mr. Beehler says: "The solarometer obviates elaborate logarithmic calculations and combines in itself a pelorus; so that it furnishes a complete solution of the entire problem to ascertain the ship's position and compass error in the space of time ordinarily required to observe the altitude by the sextant, and take its bearing with a pelorus." This sentence is misleading, as, in the first place, there are no elaborate calculations necessary in order to ascertain the ship's position from sextant observations; the calculations are simple and occupy but little time when the navigator is in practice. Again, the nature of the problem is such that the instrument cannot obtain, under ordinary circumstances, good longitude results and good latitude results from the same heavenly body at the same time. Either the body will be nearer the prime vertical and the hour angle will be good while the latitude will be bad, or it will be nearer the meridian, when the opposite will be the case. If we examine the method of taking observations it will be found that when the latitude is unknown it is necessary to take four observations before the correct hour angle is ascertained, and then, in spite of the inventor's claim, the latitude must be in doubt. To take the observations and compute the results, it is necessary to take out the same quantities from the nautical almanac and from Bowditch as when the sextant is used, except the dip, semi diameter and four logs. In addition, it is necessary to make a small calculation from the readings of the compass rose to obtain the true course. When four observations are required, I believe it would be as well, if not better, to work a Sumner with the instrument.
The necessity of allowing for refraction is another difficulty with this instrument. It is evident that it would require a very highly skilled observer to allow for the effect of refraction by the position of the heavenly body in the telescope. Except when the body is both high in altitude and near the prime vertical, when the effect of refraction can be ignored; tables must be consulted, and the hour angle and azimuth corrected. Unless high in altitude and near the prime vertical, only stars that have a declination within the limits of the sun can be used, or very large tables must be computed. This refraction is a very bothering quantity for the inventor, for it varies with different states of the atmosphere and during low-lying fogs the quantity given in the table may be far from correct at a time when the solarometer should be most needed.
The inventor further says: "And besides, this has the great additional advantage that, whatever may be the result indicated by the solarometer, the observer can always know positively if his observations and results are right or not." This is incorrect. By use of the azimuth tables taking the hour angle observed, and the latitude and declination used, if the azimuth found in the tables corresponds with the one observed, the instrument may be said to be in adjustment. Should the observation be taken improperly or wrong quantities used, the error must be serious to be detected by the tables. The body might be observed slightly out of the center of the telescope, and if it were near the prime vertical, a latitude far from correct recorded without an apparent change in azimuth. The fact is with this instrument as with the sextant, the navigator may be sure that his results are in error. What he desires to know is within what limits he can rely upon his results. The main question to be determined for the solarometer is. Within what limits can it be relied upon, at night, or in the daytime, when the horizon is obscured and the sextant of no use? If the probable error is sufficiently small, the instrument is of great value.
I believe the claims made by the inventor are too large and that they are likely to retard the progress of the instrument in its way towards general adoption; for, when unexpected difficulties are encountered with a new instrument, it is liable to fall into disrepute. The value of the solarometer does not depend, fortunately, upon its doing away with elaborate calculations, producing absolutely accurate results or accomplishing impossible tasks. What must be known is, Can the instrument be used in ordinary practice, in spite of the vibrations, rolling, pitching and yawing of the ship? Will the many adjustments remain correct for a reasonable time within reasonable limits under ordinary conditions? These questions should be answered shortly with three instruments afloat, and, if answered favorably, the solarometer must be adopted by all vessels of sufficient size to afford a proper location. If the instrument prove reliable, it must serve to save enough coal to more than compensate for its cost in a few trips; and its value to commerce is beyond estimation when the additional security to life and property is brought in consideration.
Mr. John Martin:—Having made six transatlantic voyages in charge of a solarometer on board of the U.S.M.S. New York, my experience has demonstrated the practical success of this instrument, under all conditions, except when excessive vibrations of the ship disturbed the float to such an extent, that it was almost impossible to judge when the body observed was in the axis of the telescope.
The solarometer was mounted in the New York on the hurricane deck directly over the thrust bearings of the engines, and in that place the vibrations of the engines were transmitted directly to the deck upon which the solarometer was secured.
On the first round trip voyage from New York to Southampton and return, December 12, to December 29, there was so little clear sky that there was no opportunity to test the instrument. I myself had had but a few hours experience in observing with the instrument before I went on this voyage, and was not sufficiently familiar with the manipulation of the instrument to get star sights.
On three days at sea I got snap shots of the sun, and my results agreed to within six miles of those observed by the sextant; while in port at Southampton, and subsequently at New York, on January 1, I obtained perfectly accurate results by observations with the solarometer.
On the second round trip voyage, from January 2 to the 19th, experienced some clear weather on the eastward voyage, and obtained some good results, but in the long following seas, the engines raced to such an extent, that the excessive vibration often made it impossible to determine positively when the sun was in the axis of the telescope ; many of the observations were found to be erroneous, and were discarded, as they did not show agreement with determinations by the sextant. Such results as agreed with the sextant closely, also agreed in the comparison of the computed and instrumental values of hour angle, latitude, declination and azimuth.
The steamer New York went to Newport News, Va., to be docked on January 20, and on this trip the ball and socket joint of the float was removed, and thereby a great deal of the effect of excessive vibration was compensated. Under ordinary conditions, the removal of this bolt prevented much of the vibration of the ship from being communicated to the float, and I thought the only difficulty had been overcome.
I sailed again for Southampton on February 13, and obtained a series of observations, in which throughout the voyage I was enabled to ascertain the ship's position and compass error, except at times when the vibrations were such that the results were indeterminate.
Captain Jamison, commanding the New York, was much interested in the solarometer, but during these winter voyages the sun was so rarely visible, and so low in the bank of clouds and horizon, that he had not sufficient time or opportunity to become so familiar with the instrument, that he could rely on its indications: though by carefully comparing the readings of the solarometer with the computed value of azimuth in the book of tables, he has positive evidence of the correctness of results.
The method of allowing for refraction by observing the body in the telescope as much below its axis as refraction elevated the body above its true position, was practiced in the first two voyages, but on the last voyage, much better results were obtained by accurately allowing for the refraction and observing the body directly in the axis of the telescope.
Captain Jamison expects to use the solarometer through the spring and summer, when he will have more time to familiarize himself with its workings, and its value will become apparent in foggy weather during the coming spring and summer weather.
The position of the solarometer on board the New York, is unfavorable because that part of the ship is subjected to greater vibrations than any other. A much better position would be near the pilot house, but there, other reasons prevented its being mounted, and in order to get it on board at all, it was necessary to yield and place it where it was most convenient and least in the way.
Lieutenant G.L. Dyer, U.S.N.:—Mr. Chairman, I would like to hear some account of the inventor's own experience with the solarometer at sea.
Lieutenant Beehler, U.S.N.:—I made two round trips, transatlantic voyages, in the North German Lloyd steamer Weimar, from Baltimore to Bremen, in March and August, last year.
On the voyages I had solarometers which were subsequently greatly improved, but though imperfect instruments, I was able at times to obtain good results which demonstrated the practicability of the instrument.
In these instruments, by setting the arcs in positions corresponding to the computed values of the four quantities of declination, latitude, hour angle and azimuth, and waiting for the sun or star to be visible in the axis of the telescope, I could get accurate and reliable results. The defects of the instrument were removed in the subsequent design, but the experience demonstrated that the rolling and pitching motion of the ship was fully compensated by the arrangement of the float in the bowls.
A Hicks clinometer was fixed near the instrument, and observations were taken, and accurate results obtained, when the ship was rolled 20 degrees each way. At one time I tried to get observations when the ship was rolling deeply, 40 degrees each way, but the heel was so great that it was impossible for me to reach up so as to have my eye at the eye piece of the telescope. It would have been necessary for me to have been about 7 feet tall to reach over, but for all that, I could see the shadow of the sun shining in the field of the telescope, even when rolling 40 degrees each way. At this time, the Weimar was flying light, with very little cargo, and in a heavy sea. The instrument was mounted on the hurricane deck, about 42 feet above the water-line, and the rolling and pitching motions were fully compensated.
In reply to the criticism of Lieutenant-Commander Wainwright, in regard to the refraction, I admit that there is, and has been, considerable trouble with refraction, but no more, nor in fact as much, as with a sextant.
In developing the instrument, the question of compensating or correcting for refraction has been a serious one. I find that it is best to correct the hour angle by a correction from tables corresponding to the hour angle, polar distance and latitude, or altitude; these tables are not as long as at first seems probable, and can be readily applied. The accuracy of the table of refraction has been questioned, but the error is small. I was under the impression that the amount of moisture in the atmosphere would have a serious effect on its refraction, but I have in a letter from Professor Harkness the statement that in the most refined observations of astronomers the moisture of the atmosphere is neglected, the temperature and pressure determining its density and refractive power.
The criticism of the proof claimed by the agreement between the computed and instrumental values of hour angle, declination, latitude and azimuth, seems to overlook the fact that as with all other instruments of precision, accurate results can only be obtained by careful and precise work. If the body observed is seen exactly in the axis of the telescope, and if the graduated circles are accurately read, it follows as an axiom that if the computed and instrumental values agree, the result is correct, and the observer has proof of the accuracy of his observation. This feature has been used to obtain results with an imperfect instrument, and the practice has demonstrated the correctness of the claim.
In conclusion, unless there is any other point which I may explain, I desire to express my sincere thanks for the kind expressions of approval, and good wishes of those who have discussed the paper. In this connection, I feel deeply sensible of the favorable consideration I have met with on every hand. I have often heard it stated that the Hon. Navy Department does not encourage inventors. My experience has been quite the reverse, for I have had no reasonable request refused, and have been encouraged in every way. Your kind attention is another evidence of this feeling, and I thank you all sincerely for this most flattering reception.
The lecturer exhibited and explained the various forms of the instrument, beginning with the early imperfect ones and ending with that at present set up for trial and use in the Naval Academy grounds. After tendering a vote of thanks to the lecturer for a very entertaining and instructive lecture, the meeting adjourned.