310 PLUMB LINE AND SP. LEVEL. 



tclefcope, to pafs over a point near the eye end. And in this 

 manner a plumb-line was applied to it by Mr. E. Troughton, 

 with his ufual accuracy. 

 Defcribed by Dr. This way of levelling the axis of a tranfit telefcope has fince 

 been defcribed by Dr. Ufher in the Tranfactions of the Royal 

 I rim Academy. 

 Laid afide. T ne Dr. fpeaks unfavourably of it, but many of his objec- 



tions are ill founded, and it is probable that he never tried it. 

 This method of applying the plumb-line gave me much fatis- 

 faclion at that time, but fome years afterwards it was totally 

 Spirit level of laid afide for a ground fpirit-level. The fpirit-level which I 

 with Sexes 1 " *' now u ^ e * s g round to a radius, of 430 feet, with two indexes 

 graduated to two made of ivory, each about one-third of a minute in length/ 

 feconds. graduated to every two feconds. Thefe are made to Hide 



upon the level, that they may be brought over the ends of the 

 bubble at pleafure ; and no other adjuftment is neceiTary. 

 Improved me- My method of ufing the fpirit-level is not only more con- 

 thod of ufing it. ven i ent tna n any o^er that I have met with, but the axis of 

 an inflrument may alfo be examined by it without altering 

 the perpendicular fcrew, which is a matter of fome confe- 

 quence; becaufe an inflrument newly adjufted is fubject to 

 vary for fome time after, by the changes which take place in 

 the temperature of the air ; confequently the adjufting fcrews 

 ihould never be touched, except to make fome neceflary 

 alteration. 

 Obferve one end My method is as follows. Having hung on the level, Aide 

 reverfe the kvel an mdex over one end or> tne bubble, and after it has remained 

 and obferve at reft about a minute read off from the index the number of 

 different fta- feconds a P ointed out by the end of the bubble, then take off 

 tion is the error} the level, invert the ends, and hang it on again, and if the 

 to which bring f ame enc j of the b UD bi e come to the fame divifion (a) as be- 

 the inflrument. „ , .., , ^.^, , - .. i • Vi n 



fore, the axis is level. But if the end of the bubble reft at 



any other divifion (b) on the index, then half the difference 

 will be the error ; and by means of the perpendicular fcrew, 

 bring the end of the bubble half way between a and b, and 

 the axis will be level without further trouble. 

 If thepofition The pofition of the axis may be alfo determined at the 



t e in a edb d y et t e he fame time > ^ the ° ther end ° f the bubble > and if the IeVel 



other end of the be a good one, the refults given by the two ends will feldom 



fttnceu'fddt differ fo much as half a fecond. 



half a fecond. In 



