T II E 



T II K 



the telescope on the sceoud object. It is clear ihc mo- 

 tion of th repcating-table has me: 1 the tclc- 

 -..ipe to it original direction, without altering t!io read- 

 inir of llii- circle: nnil that if tin- be turned 

 on the fc-c ' l'> |K motion alone, without dis- 

 tributing the circle, the difference between Ihc 



MC new readings Biul the preceding mean will 

 also be the angle required. Hy continuing the process, 

 the mnglc may be measured a- often as the observer 

 pleases. It i/ evident tliat nil reiidings-off. except the 

 first and la&t, are superfluous, save as ( -hecks, or 



he means of estimating the ai : tin- tinal 



..mid terminate after a whole number 



\olutions as nearly as possible, when the cxccntricity 

 of the repeating-table will be eliminated, a matter of 

 possible importance if the object* are near ami the repeat- 

 mg-tahle carelessly made. or. if the objects arc pretty dis- 



.uid this caution mperflUOM, when the \ 

 nearly at the divisions at which \ou set out. which gets rid 

 ol or at !ca>t diminishes any errors of division. The latter 

 condition is however rather a speculative than a pi i 

 one. As the error of division is divided by the num. 

 observations, and the casual error of obsei -\ation only by 

 the square-root of the same number, it is evident that a 

 moderate number of repetitions in our excellently-divided 

 circles will reduce the error arising from mal-div ision to a 

 much smaller quantity than that which belongs to the 



-ual error of observation. 



The essential condition of repetition is, that the rnotipn 

 of the theodolite shall not disturb the repcating-table. 

 The motion of the latter therefore should be as heavy as 

 will admit of nicety in the tangent-screw, while the mo- 

 tion of the parts which move with the telescope should 

 be as liijlit and free as is consistent with firmness. There 

 no difficulty whatever in effecting both 

 point- : but lest any error should arise from repe- 

 tition, we should recommend a careful observer to deter- 

 mine his ancles by two scries. one by always moving 

 the telescope and its tangent-screw forward, and the 

 repcating-stand and its screw backwards; and another, 

 by reversing the process. If the two results agree, a- 

 tlicy should do within the limits of casual observation. 

 the" mean is probably free from all other error; and if 

 they do not, the observation should be repeated and \aried 

 until the quantity and probable law of the error i- 

 tained. \Ve should then be able to say decidedly where, 

 when, and under what precautions repeal inir was a sale 

 as well as a convenient and economical piir-ess which at 

 present is rath, i/iirfxlin. unless the decision be 



supposed to be again*) all repetition, to which we do 

 not 1 



The lurcgoing description has been confined to a form 

 of theodolite which is not in ordinary use, though from its 

 simplicity and power it is well adapted to the purpose of 

 explanation. The common theodolite is generally car- 

 lied by a pair of parallel plates, fixed on a three-legged 

 i-Iaff. The lower of these circular plates is screwed 

 upon the staff, and has an aperture above the 

 The upper plate has a strong descending shank which 

 through this aperture. A button of a 

 . d on the end of the shank, the cur- 

 vature ii] id rubs against the under surface of 

 the Ii ich is doiii'-shaped to tit it. Four 

 ass through the upper plate and abut with 

 Iheii ' Ihe lower plate. When the 



s are turned tl <! until the button 



and I '-n which it rubs are brought into 



.M-t. To level the theodolite, set the levels 



lei to a diagonal pair of the parallel 



plates. Then screw one pair until you come to a I" 



and by releasing one screw and up the other, but 



not very tight, set the corresponding level hori/ontal : 



leaving'this pair and taking bold of the other pair set the 



:nl level also right, and if the first !<. el \a< 

 it probably will be a little, restore its position ! 

 up the proper screw. Turn the telescope hall' round and 



't the error, half by tin- parallel pla'. :id the 



other half by the lev el-adjustments tin It is 



deirable thnt, when the final adjustment is made, the 

 screws should bite pretty hard, otherwise tin 

 chance that the upper plate will turn a little dun 

 observation. Tin* objection would I to the use 



Qf parallel plate* where grea nicety U required: they arc 



cr very convenient and of very ready use, an.l 

 haps if the screws are strong and tl" 

 to give the tcV^-opc three or four turns round i-i the 

 tion he means to observe, before Marling, und 

 move the ' Lhe same v 



avoided. The first object observed should alw.v 

 observed at the end of the service, m order to sec w],. 

 there has been any change in the original position. I 

 of the screws rci-t ill a notch, perhaps the tendency to 



nay be wholly overcome. 



Another contrivance which is to be found in aliuc. 

 theodolites is much more objectionable. Tin 



to save himself addition or Mihtra'-tion. an 

 quires an adjustment by which be can turn the whole 

 circle about and bring the telescope upon the 

 object, the vern. prcv iously set to zero. There 



is therefore a motion with a clamp and tangent- 

 for this purpose, which, as the clamp has usually a very 

 short bearing, is particularly liable to yield and 

 destiny all accuracy. To remedy this unnecessary evil, 

 id or watch t'eleseopc, as if is called, is attached to 

 this part of the instrument and brought to bear upon a 

 \vell-detincd object. Any motion or wriggling of the zero- 

 clamp is betrayed by the watch-telescope, and when an 

 angle is taken, it imi>t be first ascertained whether the 

 watel. keeps its position, and the position if 



turbed must be restored to the zero tangent-sen 

 the observation is finally made. In some theodolites made 

 I<T the Indian survey.' under the diree; lonel 



Kvorr-st. the zero and slow-motion damp take the form of 

 a repeating-tahle. and may be so applied. It would he, 

 safer to have this motion n'uide considerably heavier than 

 in the patterns we have seen, and if the instrument is 

 to fall into clumsy hands the watch-telescope miirbt > 

 be added for greater caution. Such a theodolite would, ,-n 

 far as we can judge, have no limit to its accuracy, except 

 that depending on the diminutive telescope. 



For many purposes of surveying it is desirable that the 

 telescope should allow of being considerably clevar. 

 depressed, and that means should be civ en for measuring 

 this angle with considerable accuracy. A circle, or por- 

 tion of a circle, is then fixed upon the telescope a\i 

 the necessary verniers and level may be secured by a tail- 

 piece or otherwise to the support. If the vertical an:;! 

 to be measured as accurately as the horizontal angle 

 instrument becomes an altitude and azimuth circle. [Cm- 

 CLE.] Hut such instruments are rarely applied t 

 measurement of terrestrial angles. The! direction of the 

 meridian was determined in the Ordnance Survey by ob- 

 serving Polaris at its greatest elongations I-'., and \V., and 

 taking the middle of the two readings for the direction of 

 the north. Hence the telescope required all the tiansit 

 adjustments except that for azimuth [TRANSIT], and was 

 considerably elevated above llic circle. Though the 

 results were upon the whole satisfactory, yet we greatly 

 doubt the prudence of ascertaining this fundamental and 

 delicate point from such an instrument, or of risking the 

 steadiness of the telescope supports by raising Hi. 

 much above the body of the instrument. It would have 

 been better, we conceive, to have determined the direction 

 of the meridian bv a series of careful tiansit observations, 

 using more optical power with greater steadiness, and to 



'.i-pt the theodolite to its proper office, that of mea- 

 suring horizontal angles. hieing the height of 

 the telescope support-. The great theodolite had origi- 

 nally a semicircle fixed to the axis of the telescope, for 

 measuring altitudes and depressions. This has since, very 

 pro;-' .'.ml a whole circle substituted. 

 When a theodolite i> merely used for survey ing. Ihc 



"pe requires only a moderate vertical range. Mr. 

 Troughton fixed a portion of a circle t which may be more 

 properly called a slice than a sector) to one or two ol his 

 12-indi theodolites, and this construction is often found in 

 Other makers. The telescope is thus kept lower, the in- 

 strument is firmer, and the larger radius gives the portion 

 eming advantage over the entire circle of 

 smaller radius. There is I 



v in a portion of a circle, and we should prefer a 

 I sort Of compromise, giving the n-h an ele% 



i'id allow a vcitieal circle of about half the dinien- 



nf the horizontal circle: if Hie t the meri- 



dian u to ! determined by this instrument, the sii| 

 . must be at least no high as to gee 2 or 3" above the lati- 



