TRANSIT INSTRUMENT. 11 



25. TJte equatorial intervals of the wires were determined from 500 transits 

 of clock stars, and 32 transits of close circumpolar stars. No accident 

 happened to the wires during the four months of observations. The wires 

 were named a, b, c, d, e, the wire a being farthest from the micrometer-screw- 

 head. The equatorial intervals finally adopted were 



a = -f- 28-483") 

 b = + 14-413 | 

 c = o'ooo )> For micrometer east. 



d= 14-384 

 = 28-426J 



Consequently, to reduce the mean of the five wires to the center wire, the 



o 9 * o 1 60 

 correction, with micrometer east, is + s ; n KPD stars within 30 of 



the equator this correction has been taken as = + O s- 02 (micrometer E.) 



26. TJie value of one revolution of the micrometer screw was obtained from 

 very numerous observations of close circumpolar stars, chiefly of Polaris. (^ 

 Altogether there were made 64 separate determinations at different parts of 



the screw, each obtained from at least five bisections, all tending to show 

 that there was no sensible drunkenness or error periodical on one revolution. 

 The adopted value is 56"-251. 



The integer revolutions of the screw were read from a comb in the field 

 of view by means of the center wire ; they were numbered in the observation- 

 book so that the 20th revolution (20 r- 000) occupied the optical center as 

 nearly as possible, with increasing readings towards the head. Thus 22 r> 000 

 signifies that the system of wires was moved two revolutions from its central 

 position towards the micrometer-screw-head. 



27. The Error of Level of the transit axis was generally determined, with 

 the hanging spirit-level described above, once or twice with every batch of 

 time stars and with every circumpolar star observed for azimuth error. The 

 relative position of the pivots, both in level and in azimuth, was subject to a 

 regular diurnal variation. During the night the east pivot sank from 2" to 3", 

 and moved to the north about the same amount. The adjusting screws were 

 touched on 1874, October 5, after which date the level-error remained small 

 and steady. Several shocks of an earthquake on 1874, December 28, pro- 

 duced no sensible alteration in the position of the instrument. 



28. The correction to the level-error, as found by spirit-level, for inequality 

 of the diameters of the pivots, was found, by repeated reversions of the 

 instrument, to be on 



c 



