QUADRANT. 



fcrew, and turned (lowly by the thimib-fcrew or nut, till 

 this coincidence is jierfedt, and then the lever mull be made 

 fail again ; tiie rcliedlcd horizon will be inverted when 

 viewed in the manner here defcribed. Wlien the inflrument 

 is ufed in a back obfervation on the land, which is not likely 

 to happen often, the adjuflmcnt will require iome artilicial 

 aid: by Dollond's method, an index is applied to the back 

 hori/.on-glafs, by whicii it may be put parallel to the fore 

 hori/.on-glafs, by the third adjnftment a;)plied thereto, and 

 then, by a graduated arc of 90°, inferted on purpofe, the 

 index of the back horizon-glals is fet perpendicular to its 

 former fituation, and confequently is made perpendicular 

 to the plane of the fore horizon-glafs. By Mr. Blair's 

 method, the under edge of the index-glafs is ground and 

 poliflied, fo as to be at right angles to the plane of the 

 glafs, and hence the back hori/.on-glafs is adjuiled, by 

 making the direft and reflefted horizons agree, while the 

 vernier Itan'ds at zero on the limb, exa&ly as is done in the 

 third adjuftment. But we are to fuppofe the inftrument 

 not furnifhed with either of the above named auxiliary 

 means ; when this is the cafe, a long level fpace mull be 

 chofen in fome common or park, where three Itaves can be 

 put up in the fame ftraight line exaftly, and at an interval 

 from each other of not lefs than 500 yards, but if 600 or 

 700 the better ; then, the vernier being truly placed at 

 zero, let the plane of the odlant be held horizontally witli 

 its back horizon-glafs exailly on the top of the middle 

 ftaff, and let one of the rods be viewed direftly through 

 the back fight-vane, then if the other rod, feen by reflec- 

 tion, coincide with it, the pofition of the back horizon- 

 glafs is true, but if not, it mull be made fo, by f'e tail- 

 piece behind, as before direfted ; if in fixing the tail-piece 

 a derangement Ihould take place, which cannot be exactly 

 done away by a fecond or third trial, then the index-error 

 for the back obfervation mull be afcertained. 



This error may be obtained by reverled obfervations by 

 two methods ; thus, let the coincidence of the two extreme 

 ftaves be made from the middle llaflF as juft directed, by 

 amotion of the vernier, inllead of the tail-piece of the back 

 horizon-glafs, and read the quantity moved forwards or 

 backwards ; then turn the downward face of the frame up- 

 wards, and repeat the operation, and read again ; now, if 

 thefe two readings are found to be one on the limb, and the 

 other on the arc of excefs, half of their difference will be 

 the error fought, -f or — ; but if they both are read or. 

 the fame fide of zero, half their fum, -f or — , will be the 

 error ; when they are both read on the limb, or if the greater 

 quantity be read on the limb, when the lefs is on the arc of 

 excefs, the fign will be — , and -vice verfd. By the fecond 

 method, inllead of reverfing the plane of the inftrument, 

 let the obferver reverfe his pofition, making the right-hand 

 ftaff to be the left-hand one, and "vice "verfd, and then let 

 him repeat his firfl obfervation, with the fame face of his 

 inftrument up that he had in that, and the fame refult will 

 follow, that was obtained by reverfing the plane of the in- 

 ftrument. Thefe two modes of obtaining the error in 

 queftion, by reverfed obfervations, may be made to check 

 each other, and will then give an average of the two refuks, 

 for the error to be applied to each fubfequcnt obfervation, 

 taken with the back horizon-glafs in its latt determined 

 pofition. 



Examination of the Inftrument. — Thefe feveral adjullments 

 being finiflied, the obferver might proceed to make his 

 obfervations, provided he could rely on the fl<ill and credit 

 of the maker of his inftrument ; but if not, he would adl 

 prudently to examine the materials of which it is compofed, 

 before he relies on its performance, and alfo to put to fome 



tell the accuracy of the divifions on the vernier and limb, 

 as well as the care that has been taken in felefting glafles of 

 uniform thicknefs, and of good polifh. The exadlnefs of 

 the total arc may be afcertained from comparifau with 

 meafurements taken with a circle, or iome well-known fupe- 

 rior inftrument ; or by careful trigonometrical meafurcment 

 on fome level and extenfive ground, where the fides of the 

 triangle can be accurately meafured, for which purpofe 

 Troughton's new chain of five-feet links is admirably cal- 

 culated : but the intermediate divifions may be examined by 

 the vernier itfelf, by ftepping the arc with it ; and if all 

 parts of the arc are found alike divided, when examined by 

 the coincidence of the extreme ftrokes of the vernier, tJie 

 intermediate ftrokes of the vernier itfelf may laftly be ex- 

 amined by the ftrokes on the limb taken in various places. 

 Ur. Mackay recommends a table of correftions to be made, 

 in cafe of errors being detedled in the dividing, but we 

 fliould rather recommend new divifions from a good engine, 

 the inferting of which can now be performed at a trifling 

 expence. Whe;i the planes of the mirrors are examined, 

 try if a candle reflefted appears in two images ; if it does, 

 the two faces of the glals are not parallel, and confequently 

 the glafs is not fit for its purpofe. 



To try whether the fnrfaces of the mirrors be perfect 

 planes, bring two dillindl dillaiit objefts into good contadt, 

 and let them be fjen at the upper edge of the filvcred part ; 

 then move the inllrumcnt in its own plane, and move the 

 image and body along the line of ieparation or edge ef the 

 filvered part, and it the coincidence is not difturbid the 

 plane is perfuct, but not othcrwife. Alfo, when an obferva- 

 tion is taken of the fun, and the vernier has been fixed by 

 the fixing fcrew, read the altitude, as the dark glafs, or 

 glafies, may have been ufed ; then remove them from tlie 

 focket that holds thf^m, and r>;verfe their planes, by putting 

 their remote faces nearell ; and if, after this change, the 

 fame altitude is given, tlie planes are parallel, otherwife half 

 the difference mull be applied, in all fuch cafes, as an error of 

 the coloured glajfes. This examination is beft made exaftly 

 at noon, when the fun's altitude is not fenfibly changed dur- 

 ing the examination ; unlefs a good chronometer or regula- 

 tor, duly regulated, and put to exaft time, be at hand ; 

 for then the fucceifive examinations may take place at equal 

 diftances from noon. Or the fame thing may be done by 

 means of a luminous objeft placed at a diflance, and fo ele- 

 vated, that an artificial horizon may give its double altitude 

 before and after the coloured glafies are reverfed in pofition. 

 The befl artificial horizons are either a veffel of pure mer- 

 cury, with a roof of good glafs framed over it to prevent agi- 

 tation by the wind ; or otherwife a piece of black glafs, well 

 polifhed, and placed on fcrews of adjuftment, with a good 

 fpirit -level in a glafs tube, fo ground, that it will reverfe in 

 pofition, and will place the glafs in a perfeft level in the 

 two requifite direiflions, at right angles to each other. 



It may be proper to examine further if the two holes in 

 the fight-vane are fo made that the fun or other luminous 

 body dazzles the eye more in looking through one, than in 

 looking through the other, the intention beijig that one fhall 

 take in more of the filvered part, and the other more of the 

 unfilvered part of the glafs, fo as to accommodate the quan- 

 tity of diretl lightto the ftrength of the eye. 



lllujlration of the Ufe. — When Hadley's odlant, or fextant, 

 is ufed for altitudes at fea, the fun, moon, or ftar, as the 

 cafe may be, muft be viewed in the way that the eye can beft 

 bear, with or without the dark glaffes, telefcope, &c. as 

 occafion may require, and as experience will didlate ; and in 

 a fore obfervation the body .obierved muft have its image 

 brought gradually down, fo as to be in exadi contad with 



the 



