THE INDUSTRY OF ALU NATIONS.

dates back to a venerable antiquity of origin. The greatest arc yet measured isthat of Eastern Europe, which, undertaken in 1816, has now progressed so as toinclude 25J degrees of the meridian from Ismail on the Danube to Fiuglenessin Einnmarken. A provisional calculation of this has already been made, theRussian part of the operation embracing 20° 3D being under Struve, the eminentastronomer, and having been throughout in high favor with the Czar. TheSwedish and Norwegian portions of 4° 49' were under the charge of Hansten andSdlander. A Southern prolongation through Turkey and the Archipelago toMount Ida on Crete, has been talked of hopefully (though, of course, it isat present impossible), and this would make the entire arc amount to 36°,or one-tenth of the earth’s circumference. The Indian arc of 21° 21' be-tween Cape Comorin and Kaliana, was commenced under the East India Com-pany by Major Lambton, in 1802, and has since been extended with a liberalityand accuracy which, in so difficult a country, are particularly commendable. Itis hoped that this arc which has now crossed the Himalayan range of mountainsmay still be extended north to the Arctic Ocean, as the Russian Emperor is un-derstood to favor this enterprise, which of course means that he is ready to pro-vide the means for its continuation through Asiatic Russia. This would give farthe longest and best conditioned arc ever measured, or even possible on the earth,being about 60 ° in length. We need not more than enumerate the smaller arcsmeasured by Lacaille and Cassini, in France; Boscovich, in Italy; Mason and Dixonin the United States; Lacaille, at the Cape of Good Hope; Condamine, in Peru;Maclear, in South Africa; and the geodetic results of the surveys of Switzerland,Holland, Bavaria, Baden, Wurtemburg, Hesse Darmstadt, Hanover, Brunswick,Upper Italy, Prussia, Austria, Denmark, and Sweden, and the British NorthAmerican provinces. The United States Coast Survey has not yet published itsgeodetic results. Its triangulation is however connected over an arc from Port-land, Me., to Cape Henry, Ya., of about 7° in latitude and 6^° in longitude. Withinthis range are two portions—one on the Chesapeake Bay, and the other from Nan-tucket to Blue Mountain, in Maine, where the triangles deviate but little on eitherside of the meridian. The extreme latitudes of this arc are found by a preliminarycomputation to correspond well with Bessel’s elements, a result of the more presentimportance, as the Russian arc requires a diminished value of his compression.When the triangulations of all the Atlantic sections are connected, an arc of about0°2 of latitude will be embraced, and the arcs of parallels are of unusual extent.The series of stations in this scheme of triangulation will have their latitudes andlongitudes determined more in detail, and with greater aggregate precision, thanbelongs to any kindred operation now executed. We now propose to give in somedetail, an account of the latitude and longitude instruments and methods used inthis great national work.

To determine the latitude of a station, it is only requisite to measure the zenithdistance or altitude of some known stars at their culmination or passage acrossthe meridian. For approximate results, this is one of the simplest problems ofpractical astronomy. But where the highest accuracy is required, it becomes oneof great delicacy. A second of arc being equal to about 100 feet on the ground,is a great error when a comparison is to be made with a triangulation, the sides ofwhich are correct probably to a single foot. The measurement of angles in themeridian, on the graduated arcs of portable instruments is so much affected by un-certainties of refraction and instrumental errors, that the minutest accuracy is onlyto be reached by nullifying these uncertainties and errors as far as practicable. Aseries of careful experiments on various instruments was made by Prof. Bache,and the result has been that the coast survey observers quite concur in preferringthe zenith telescope, used according to the method originated by Oapt. AndrewTalcott, late of the U. S. Corps of Engineers. Trial was first made of a two-feetvertical circle, and of some eighteen-inch repeating circles, made by Troughton &Simms, and of Gamby’s six, ten, and twelve-inch repeating theodolites. In all ofthese cases, the amount of instrumental error was so great as to indicate the ad-vantage of adopting larger instruments. Prof. Bache then procured from Simmsa transit instrument, a zenith telescope on Talcott’s plan, and a zenith sector, onthe plan devised by Airy for the ordnance survey, and described in the Astronom-ical Society notices. Telescopes of forty-five inches focal length were employedin all these instruments.

The transit, mounted in the prime vertical was found to give a ,/few good results,but clouds so far interfered with observing corresponding eastern and westerntransits, as to lead to the abandonment of this instrument for field use where theresults must be obtained in a limited period of time. Airy’s zenith sector, afterfull trial, has been also essentially superseded as a field instrument by Talcott’szenith telescope. Its great weight and the excess of labor in observing with it,have caused it to give place to the far lighter and more manageable zenith teles-cope, which yields results with more facility and equal accuracy.

Visitors at the Crystal Palace can see, adjacent to the office of the superintend-ent of the building, a beautiful specimen of the zenith telescope, made for thecoast survey, by Mr. Wurdeman, of Washington, and containing all the latest im-provements and modifications which experience has indicated. This instrument,as it stands, is essentially American, both in its construction and in its manner ofapplication.

To determine the latitude of a place, by Talcott’s method, pairs of north andsouth stars are selected from the star catalogues, with opposite zenith distances ofless than 25° each, the difference between these distances for any pair not exceeding about ten minutes of are. The stars of a pair shouid culminate successivelywith an interval of from one to twenty minutes time, to provide for reading andreversing the instrument. Having thus selected and arranged his pairs of northand south stars, the observer determines approximately his meridian, and marks itwith the stops provided on the horizontal limb; he then sets the instruments bythe level vernier to the mean zenith distance of the pair to be first observed,adjusts the level horizontally, and waits the coming of the first star into thefield. He then moves by its screw the horizontal wire, until it covers the star atits culmination. The micrometer and level scale are then read, and the telescopeis turned 180° until checked by the stops, when the transit of the second star ofthe pair is observed in like manner by means of the same, or another horizontalwire. By comparing the two readings of the micrometer and level scale, the dif-ference of zenith distance for the star-pair is found. The values of the micrometerdivisions are readily determined with exactness by several direct methods; andthe relative value of a unit on the level scale is easily ascertained, though this isusually converted into arc, and the readings applied as corrections. The effect ofrefraction is only that due to the difference of zenith distances, and its uncertaintyis almost totally overcome, especially as the two observations on a pair are sep-arated by so short a time. In the instruments now used, the probable error of asingle observation is only half a second of arc, while the probable error of northpolar distance for a star in the British Association Catalogue, is l - 4 of a second,though the Greenwich Twelve-Year Catalogue gives a considerable number ofbetter positions, the probable error being only 0 //m 6. From this lack of precisionin the catalogue position of stars, it is better to multiply pairs observed, than ob-servations on each pair. In determining the latitude of an astronomical stationof the coast survey, from twenty-five to forty pairs are observed with three to fiveobservations on each pair. From these observations, the latitude is derived withas much accuracy as is thought necessary, though if the star catalogues should beimproved, and the effect of proper motions of stars accurately eliminated, the in-strument as it stands, could go considerably further with no greater labor. The ac-tual number of observations made is such as to place the error of observation belowthat of the places of the stars. A full study of this subject cannot fail to establish thesuperiority, on practical and scientific grounds, of this American method of latitudedeterminations. It may as well be stated here, that this is not at all alike Gauss’smethod, as has been supposed by a high astronomical authority in France.

That the best mode of determining longitude differences, is wholly and peculiarlyAmerican, is a clear fact in the history of science. Though Capt. Charles Wilkes,U. 8.N., made the very first use of the telegraph for determining differences of lon-gitude, yet the labor and credit of giving practical shape and development to this me-thod, is due to Prof. Sears C. Walker, whose recent death science still mourns. Hiscareer in charge of the department of longitudes in the coast survey, was onemuch redounding to our national reputation and to the advantage of science. Thefaoility with which American ingenuity in mechanical contrivance, met the con-ditions for automatic clock records, when Mr. Walker had pointed them out, andthe skill with which accuracy and simplicity were combined in the several record-ing devices of Walker, Locke, Mitchel, Saxton, and Bond, were truly characteristic.The history of the several arrangements used, the discussion of their defects andadvantages, and, indeed, the whole subject of mechanical or automatic recordingof astronomical observations, is of much interest and extent. To treat it intelli-gibly, would require much space, and is rendered superfluous by the full informa-tion easily accessible in documents and journals. When we compare Mr. Walker’slast labors, in which the determination of longitude differences attains almost theaccuracy of the best latitude results, with the early history of the longitudeproblem, in which the rude accuracy demanded by practical navigators, couldscarcely be reached, we see a progress of the most conspicuous kind.

Longitude being convertible into time, the problem of longitude differences, isin fact but a question of difference between local times. How can the local timesof two distant stations be compared ? is the fundamental question. Simultaneousobservations of an instantaneous event, like an eclipse of a satellite of Jupiter, oran intermediate signal light flash, were among the earliest and obvious modes, as thedifference between the local times of observation would give the longitude differ-ence. The transportation of chronometers between stations is another obviousmode, the value of which depends entirely on the number and perfection of thechronometers thus compared with the two local times. Again, the theory of themoon’s motion being supposed to be perfectly correct, observations on its time ofculmination, or passage across the meridian, would give a means of knowing thedifference between the time used in computing the lunar tables and the local time.Or by observing the instant in local time of the occultation or covering of a knownstar by the moon’s disc, and comparing this with the lunar table time for the sameevent, the time difference results. The modes at present in use for accurate lon-gitude observations, are those by chronometers, by transits, eclipses, and occul-tations, by moon culminations and by telegraph. The lunar theory is of such

exceeding difficulty and complexity, that the combined resources of physical as-

[Continued on page 181 .