REPORT
ON THE COAST SURVEY
Page:
1
2
3 4
5 6
7
8 9
10
To
obviate this source of error, Professor Bache designed an apparatus,
which was constructed in the office of the Coast Survey, under
his direction, and with which the subsequent bases have been measured.
In this apparatus the measuring bars are arranged upon the plan
of compensation for the effects of temperature, which was first
used by Colonel Colby in Great Britain, and afterwards in this
country by Mr. Borden, in the Trigonometrical Survey of the State
of Massachusetts; but the mode of obtaining the compensation is
entirely different. A principle is also introduced, which, though
of sufficiently obvious importance when once suggested, had not
been previously recognized; or had not, at least, been previously
applied in practice. It is matter of observation that if a bar
of brass and a bar of iron of equal dimensions be exposed to the
same source of heat, they will not always undergo equal changes
of temperature in equal times; but that, after a short interval,
a difference will be found to exist, the amount of which will
depend on the different conducting properties of the two metals,
their different specific heats, and the different powers of absorption
possessed by their surfaces. Accordingly, two such bars, if of
equal cross-sections, will not form a system compensating during
progressive changes of temperature, however perfectly te compensation
may be ultimately effected, when the temperature shall have become
stationary, and uniform throughout the system. In order therefore
that a combination of such bars may answer the purpose aimed at,
their sections must be so proportioned to their conducting power
and specific heat, that their fluctuations of temperature under
different exposures may go on at precisely the same rates. As
to the absorbing powers of their surfaces, that is easily made
identical, by giving to both bars the same external coating.
The lever of contact and level, first used
in the adjustment of standard measures by Bessel, has been adapted
to this apparatus, the contact between two sets of bars being
made by a blunt knife-edge against a plane, both of agate. Those
who desire to study this beautiful and delicate arrangement
are referred to the Coast Survey Report for 1854, where it is
described and figured. The mechanical arrangements by means
of which the bars are adjusted in line, height and inclination,
are also worthy of admiration; and such is the rapidity with
which the apparatus can be used, that on favorable ground a
mile has repeatedly been measured in a single working day of
ten hours. It is so arranged that the measurement may be made
at inclinations up to 3° with the same facility and accuracy
as on level ground.
The accuracy of its results has been practically
tested by repeated measurements of the same distance, from which
it appears that the probable error in measuring a mile is about
two hundredths of an inch. This base apparatus was designed
in 1845, and first used in 1847 on Dauphine Island, on the coast
of Alabama. Since then base-lines have been measured with it
in North and South Carolina, in Florida, and in Maine. All these
measurements have been conducted by the Superintendent in person,
aided by some of his assistants. The average length of the primary
base-lines measured thus far is about six miles.
Many laborious and delicate experiments were
requisite to perfect the adjustments of the apparatus. Before
and after the measurement of each base-line, the length of both
sets of bars, which is six metres or nearly twenty feet, is
carefully compared with a standard bar of iron. The rate of
expansion of this standard has been ascertained by experiment,
and its length at 32° Fahr. has been verified by means of
six single metres abutting endwise, each of which has been compared
with an authentic standard metre– one of those originally
made by the French committee on weights and measures. The possession
of this valuable standard, which was brought to this country
by Mr. Hassler, and by him presented to the American Philosophical
Society, was one of the chief reasons for using the metre as
the unit of length in the Coast Survey, especially in the absence
of any legally established standard of the country. The length
of the English yard, which is our customary unit of length,
has but recently been definitively settled in Great Britain;
and the fact that the dimensions of the earth have hitherto
been best known in metres, makes the latter measure one of great
convenience in geodetic operations. The distances in the Survey
are, of course, also given in yards, and miles, for common use.
The comparisons of length are made by means
of Saxton’s pyrometer, an instrument of great simplicity,
and affording the highest degree of accuracy. A small mirror,
movable about a vertical axis, reflects into the field of a
telescope, placed at a distance of 20 feet, the divisions of
a long curved scale fixed directly under the telescope; by means
of a small abutting piece attached to the mirror, any difference
in length to be measured causes a corresponding rotation of
the mirror, by which different scale-divisions appear on the
wires in the telescope, and a greatly magnified indication is
thus obtained by an index equal in length to twice the distance
of the mirror from the scale. The quantity to be measured is
thus readily magnified 2,000 times, without any loss from inertia
and flexure, which would be unavoidable in obtaining a similar
result by means of mechanical levers.
The necessity for the greatest attainable accuracy
in such operations is apparent when we remember, that an error
in the base-line will affect the whole distance depending upon
it by triangulation in the same ration: thus, if the base-line
were in error by but its ten-thousandth part, a distance of
100 miles depending on it would be in error, from that source
alone, by one hundredth of a mile or 53 feet.
The Fire Island base, on the south side of
Long Island, and the Kent Island base in Chesapeake bay, are
connected by a primary triangulation. These are the bases alluded
to in a former part of this report, where it is mentioned that
the Kent Island base has served to verify the others. This Kent
Island base is five miles and four-tenths long, and the original
Fire Island base is eight miles and seven-tenths. The shortest
distance between them is two hundred and eight miles; but the
distance through the triangulation is three hundred and twenty.
The number of intervening triangles is thirty-two; yet the computed
and measured lengths of the Kent Island base exhibit a discrepancy
no greater than four inches.
On some portions of the coast, where the pressing
wants of commerce made it desirable to commence the work in
advance of the regularly primary triangulation, local surveys
have been executed based on preliminary baselines, which were
measured with wooden or iron bars, and some simple auxiliary
apparatus, generally devised by the assistants having immediate
charge of the work. Some of these contrivances have been described
in the annual reports of the Superintendent, and will be found
very useful to persons wishing to execute local surveys with
some degree of accuracy. See particularly C.S. Report, 1857.
All such preliminary measurements will in time be superseded
by measurements with the compensating apparatus above described.
MEASUREMENT OF ANGLES.– In measuring
the angles of the primary triangulation, theodolites of various
sizes and construction are employed. The largest instrument
of this class belonging to the Coast Survey has a circle of
30 inches in diameter, graduated five minutes, and reading to
single seconds of arc on three micrometer-microscopes; it is
provided with a telescope of great power, which is partly supported
by springs within the upright columns, so as to relieve the
bearings in a great degree from friction. The weight of the
whole upper or movable portion is in great part secure the latter
against wear and to obtain a very free motion about the axis.
This instrument was designed by Mr. Hassler, and constructed
by Troughton & Simms, of London. It is used by the Superintendent
in the execution of the primary triangulation of the north-eastern
part of the coast, where the country presents considerable natural
elevations and admits of long lines of sight, ranging from twenty
to eighty miles in length. The accuracy of pointing and reading
attained with this instrument is such, that, notwithstanding
the varying condition of the atmosphere, the probable error
of a single measure of an angle is about one second and a quarter,
which in the mean of thirty such measures, the usual number
taken, is reduced to about one-fourth of a second. The observations
are made under all conditions of visibility of signals, varying
from a well-defined and steady image, to a nearly obliterated,
trembling blur. For each observation, the appearance of the
signal observed has been carefully noted; and the means have
been thus obtained for a discussion of the results with reference
to the conditiosn of vision. The results arrived at justify
the conclusion that the mean of observations taken in various
states of the atmosphere, is probably preferable to the mean
of such as have been obtained under circumstances apparently
the most favorable; there being in the former case less of correction
required to fulfil the geoemtrical conditions of the work; while
a given limit of probable error can generally be reached by
that method, in much less time than by the other. This result
may be readily accounted for, by supposing that, under the conditions
which appear most favorable to accuracy of observation along
a line of considerable extent, there may regularly occur a slight
lateral deflection, which, in the mean of observations taken
under different conditions of the atmosphere, is sensibly eliminated.
A theodolite nearly similar in construction
with the one described, but having only twenty-four inches diameter,
is likewise used in the primary triangulation. Besides these,
repeating theodolites with circles varying from twelve to six
inches are employed in the different classes of triangulation.
On the Atlantic coast south of New York, and on the shores of
the Gulf of Mexico, where the country presents no considerable
eminences, and is in general densely wooded, the instruments
frequently require to be elevated on wooden structures, in order
to overcome the curvature of the earth, and to get above the
most heated statum of air. These structures, which range from
twenty-five to fifty feet in height, have been constructed on
various plans by the assistants of Professor Bache, and have
proved amply steady for measurements with the repeating instruments.
The signals generally employed to mark station-points
are straight poles, supported by a tripod, and surmounted by
a cone or bell-shaped cap of bright tin, on which the reflection
of the sun shows a luminous spot to the observer. On the long
lines of the primary triangulation, however, heliotropes are
used in addition to the single-poles. The heliotrope, invented
by Gauss, is an instrument in which a mirror of a few square
inches surface is so mounted in connexion with a telescope,
that the reflection of the sun may be thrown to any desired
direction. It is very effective, and it is not unusual to observe
on heliotropes on the summit of mountains 80 or 90 miles distant,
when the outlines of the mountains are invisible through the
hazy atmosphere.
The number of observations taken for the different
classes of triangulation is so apportioned to the quality of
the instruments used, and to the agreement of the individual
observations among themselves, that the error of the sum of
the angles in a primary triangle will not exceed three seconds,
or nine seconds in a triangle of the second order. These limits
of error, however, are seldom reached in fact. The average correction
required for an observed angle in the eastern section, where
the stations are natural elevations and the sides average over
forty miles in length, is about four-tenths of a second. In
the southern sections it is rather larger.
OBSERVATIONS OF LATITUDE.— In order to fix the situation
of the whole network of triangles on the surface of the earth,
it would suffice to determine the latitude and longitude of
some one point, and the direction with reference to the true
meridian of one of its lines– its azimuth– if the
figure of the earth were precisely known. But owing to the irregularities
of the latter, it is found necessary to repeat those determinations
very frequently, by which means the work is checked, and any
accumulation of error avoided, while at the same time the figure
of that portion of the earth covered by the survey is accurately
ascertained. Observations of latitude are made at numerous stations
of the primary triangulation. Different methods and different
instruments have been employed, and the results have been carefully
discussed in order to determine which are best and most available.
A two foot vertical circle, and several twelve-inch repeating
reflecting circles, by Troughton & Simms; and also six-inch
and ten-inch repeating theodolites by Gambey, were tried successively,
but the results were found to be liable to considerable instrumental
errors, and the degree of their accuracy proved to be insufficient
to repay the labor which had been bestowed upon them. Finding
a larger class of instruments to be indispensable, Mr. Bache
had recourse to the prime vertical transit, the zenith telescope
of equal altitude instrument, and the zenith sector; each provided
with a telescope of forty-five inches focal length.
The method of determining latitude by observations
of transits of stars over the prime vertical, first used by
Bessel, is admirable in theory, but in practice a great loss
of time and labor is often occasioned by clouds, which interfere
with obtaining corresponding sets of eastern and western transits;
and for operations in the field this method is found to present
less advantages than others.
The zenith telescope, or equal altitude instrument,
was first applied to the determination of latitude by Capt.
A. Talcott, late an officer in the United States army, and has
been remodeled and specially adapted to the purpose in the Coast
Survey, where it has become the favorite instrument on account
of the great number and accuracy of the results that can be
obtained in a given time, and the facility with which it is
used. It is employed in determining latitudes by measuring,
with an attached micrometer, the differences of zenith distance
of stars culminating within a short time of each other, at nearly
the same altitude, on opposite sides of the zenith. From the
star catalogue of the British Association may be selected any
desirable number of pairs of stars having the required relative
positions. If, after pointing the micrometer wire in the telescope
to a star at its meridian passage south of the zenith, and revolving
the instrument about its vertical axis, the telescope remaining
at a fixed inclination, we find that a star culminating north
of the zenith passes precisely on the same wire, we infer that
the zenith is exactly midway between the two stars, and consequently
that its distance form the celestial equator, or the latitude
of the place, is equal to the mean or half sum of the declinations
of the two stars observed. If, furthermore, the second star
does not pass precisely on the wire, but still in the field
of the telescope, we are able to measure the difference of zenith
distance by means of the micrometer, and correct by half that
amount the half sum of the declinations. By means of a delicate
level attached to the instrument we keep account of its deviation
from the true vertical. The whole observation thus consists
of pointing and reading the micrometer, and noting the level,
in each position of the instrument.
This method possesses also the great advantage
that the result is nearly free from the effects of refraction,
which enters only so far as the difference of zenith distance
is affected by it; the correction for refraction never exceeds
a quarter or a second.
The observations with the zenith telescope
being considerably more accurate than the positions of the stars
in the British Association’s catalogue, it has been preferred
rather to increase the number of pairs employed, than to multiply
observations on the same pairs. In practice, from twenty-five
to forty pairs of stars are observed, at a station; with from
four to six observations on each pair, on different nights.
The probable error of a single observation with the zenith telescopes
used in the Coast Survey is from three to five-tenths of a second;
while the probable error of the catalogue places considerably
exceeds one second. The Greenwich observations, and other recent
catalogues, frequently supply better places than the larger
catalogue just referred to; and thus contribute to improve the
result.
Prof. Bache has at various times communicated
to the American Association notes on the use of the instrument
in question, and at the Baltimore meeting read a paper showing
the remarkable result that when at a station the same stars
are observed by the same observers with different zenith telescopes,
and by different observers with the same instrument, results
almost identical were obtained; setting at rest the question
of any residual personal or instrumental errors. It may safely
be predicted that this simple and effective instrument is destined
to supersede all others for the determination of latitudes in
the field, and it is proper here to state that the recent assertion
of Mr. Porro, a French engineer, that the method used in America
is merely Gauss’ method of equal altitudes, is founded
on an entire misconception of the method and instrument as used
in this country.
- Top of Page -