408 



SCIENCE. 



[X. S. Vol. XXI. No. 533. 



•<is tlie following ones. It has accordingly 

 •been assigned the vvt. 1, the four remaining 

 walues being given the wt. 2. Combining 

 laeeordiug to these weights, we find for the 

 mean, 



20^^540 ±.0055 



I wish this to be regarded as the defini- 

 tive value of this constant as derived from 

 the zenith telescope observations extending 

 from December, 1889, to December, 1903. 



The Test of a Transit Micrometer: John 

 F. Hayford. 



When, in connection with an astronom- 

 ical transit as used for time observations, 

 a transit micrometer and chronograph are 

 substituted for a system of fixed lines in 

 the diaphragm, a telegraphic key and a 

 chronograph, the observer is relieved of 

 the necessity of operating the key at, or as 

 soon as possible after, each of the several 

 * instants of transit of the star across the 

 fixed lines. Instead, he is required simply 

 to keep the star image bisected continuously 

 iby the movable micrometer line during its 

 progress across the field of view. In the 

 new process of thought the element of time 

 ^enters only in an indirect manner. Hence, 

 with a transit micrometer the personal 

 equation becomes so nearly zero, and its 

 variation so nearly zero, that it is difficult 

 to prove that they are not both absolutely 

 .zero. The personal equation is one of the 

 most serious sources of error in all time 

 determinations and determinations of right 

 ascension. The destiny of the transit 

 micrometer is to produce a decided increase 

 in accuracy in this class of observations 

 without increase of effort or cost. 



The observation of star transits by means 

 'ol a movable transit line was first suggested 

 in 1865 by Director Carl Braun of the 

 Kalocsa Observatory. He believed that 

 it was necessary to have the movable line 

 driven by clockwork. He failed to con- 

 struct a satisfactory apparatus. 



Repsold, the well-known iustrument- 

 malcer, was the first to suggest in print, in 

 1889, that no clockwork is required. He 

 constructed a hand-driven transit microm- 

 eter with which excellent results were se- 

 cured. 



The Prussian Geodetic Institute put the 

 Repsold hand-driven transit micrometer 

 into use on portable instruments in making 

 telegraphic longitude determinations in 

 1891, and has continued its use to the pres- 

 ent time. In all, it has been used in ten 

 longitude determinations. 



Utilizing the published past experience 

 with transit micrometers Mr. E. G. Fischer, 

 chief of the Instrument Division, Coast 

 and Geodetic Survey, designed and con- 

 structed in the winter of 1903-4 the tran- 

 sit micrometer which is before you, and 

 which is adapted for use on the transits 

 ordinarily used in longitude determina- 

 tions. 



It is a hand-driven transit micrometer. 



It is so well designed and constructed 

 that in the extensive tests, to which I will 

 refer in a moment, it never required the 

 slightest change in adjustment, not even of 

 the pressure of the contact spring, and not 

 a single record was ever lost on account of 

 any failure of the transit micrometer to 

 operate properly. 



A peculiar and important feature of 

 this transit micrometer is an automatic 

 switch which operates, without the slight- 

 est attention from the observer, in such a 

 manner that a record is made on the 

 chronograph for the middle four turns of 

 the field, and for those turns only. This 

 positively identifies those four turns, keeps 

 the chronograph sheet clear, and enables the 

 observer to practise following the star dur- 

 ing the earlier part of its transit without 

 aflPeeting the chronograph sheet in any 

 way. 



In March, April and May, 1904, this in- 

 strument was tested by 75 time sets on 18 



