412 



SCIENCE. 



[N. S. Vol. XXI. No. .533. 



the beginning and the end of the revolution 

 determined the average rate from which 

 were found the positions which the tele- 

 scope should have occupied for the inter- 

 mediate times. A comparison with the ob- 

 served positions gave the periodic error. 

 It was possible for the worm to be defective 

 either in the thread itself or in the mount- 

 ing. Rowland makes the remark that the 

 correct mounting of a screw is more diffi- 

 cult than making the screw. It seemed 

 to be so in the present case. If the screw 

 was mounted eccentrically, we might ex- 

 pect a great improvement if it were allowed 

 to engage only lightly with the R. A. wheel, 

 being held in position by a strong spring. 

 Slide 4 exhibits the resulting curves. It 

 is seen that when the worm was adjusted 

 to engage only lightly with the R. A. wheel, 

 the oscillation extended through nearly 

 eight seconds, but that, when brought into 

 contact, the range was hardly two seconds. 

 Thus the error of tliis particular worm was 

 eccentricity. 



A further test was made to determine 

 if any intermediate adjustment of the 

 worm with respect to the R. A. wheel would 

 be advantageous. Slide 5 gives a set of 

 curves for the various settings, beginning 

 with the position in contact and ending 

 with the worm very lightly engaged. Curve 

 1 is best and curve 3 is worst. That the 

 position in contact is not always best is 

 shown by slide 6, giving results for the 

 worm that proved satisfactory. In this 

 case the action was quite erratic when the 

 worm was in contact. 



A further test was made of the eleven- 

 inch Draper telescope to determine the 

 best adjustment of the worm with respect 

 to the R. A. or sector wheel. Slide 1, 

 which has already been seen, shows the ap- 

 pearance of the trails at what may be con- 

 sidered an average adjustment. In slide 

 8 we have the result when the worm is in 

 contact, and in slide 9 when the position is 



adjusted to give the least error, found by 

 trial. Probably in all instruments, one 

 may, without any process of reconstruc- 

 tion, find by experiment where the periodic 

 error is much decreased. 



The foregoing has been in the nature of 

 an abstract of an investigation in progress 

 rather than a detailed account of work 

 complete. It is possible that a careful 

 study of carves representative of the ac- 

 tion of the driving-worm will suggest an 

 improvement in cutting the thread. As 

 the experiments are easily made, I hope 

 that other observers will test their instru- 

 ments. A comparison of the results ob- 

 tained with a greater variety of instru- 

 ments would be of interest and might lead 

 to a better understanding of the entire 

 subject. 



The Reflex Zenith Tube: C. L. Doolittle. 



In 1851, or thereabout, the instrument 

 having this designation was installed at 

 Greenwich. The maker was Mr. Simons, 

 the designer Mr. J. B. Airy. The immedi- 

 ate object in view was the observation of 

 y Draconis, a star which has been followed 

 at Greenwich with some kind of zenith in- 

 strument since the time of Bradley. 



The principle is briefly as follows: The 

 telescope is fixed permanently, with its 

 axis vertical as nearly as may be. Below 

 the objective at a distance nearly half 

 that of the focal length is placed a basin of 

 mercury. The rays from a star at the 

 zenith after passing through the objective, 

 are reflected from the mercury surface 

 and brought to a focus immediately in 

 front of this objective. By means of a 

 micrometer, the frame of which is firmly 

 attached to the cell of the objective, with 

 the plane of the reticule passing through 

 the focus, the zenith distance of a star 

 culminating within ten or fifteen minutes 

 of the zenith may be measured. Finally 

 a diagonal reflector brings the ray to the 



