Royal Astronomical Society. 807 



glass ; and upon reversing the object-glass, with micrometer attached, 

 the two bars will again occupy the same position in space. Conse- 

 quently there will be no additional interruption oflight if the support 

 of the prism-reflector of the eyepiece be two bars in the vertical 

 planes which pass through the micrometer- bars, carried by crooked 

 projections from the tube in which the object-glass turns (and there- 

 fore not reversed with the object-glass). And as the prism-reflector 

 intercepts a small portion of the object-glass excentrically, a corre- 

 sponding portion equally excentric on the opposite side must be in- 

 tercepted by a small plate carried by the two bars, in order that the 

 difi^raction-disturbance of the star's image may be symmetrical. One 

 lens of the eyepiece will be below the prism-reflector, and one close 

 to its vertical, or nearly vertical, face (unless it be thought preferable 

 to produce the effect of lenses, by grinding the faces of the prism- 

 reflector to spherical forms) : the remaining lenses of the ej'^epiece 

 (namely, the field-glass and the eye-glass) will be fixed in a tube, 

 entirely exterior to the object-glass and therefore causing no addi- 

 tional loss of light, carried by a crooked projection from the tube in 

 which the object-glass turns. These crooked projections permit the 

 micrometer-heads and reversing-handle, &c. to pass, in the reversion 

 of the object-glass. 



Notwithstanding the great simplicity and compactness of the 

 essential parts of this instrument, the Astronomer Royal thinks it 

 desirable that it be so arranged that a double observation can be 

 made at each transit of the star. It is necessary for this purpose 

 that two wires be fixed in the micrometer plate, at an interval cor- 

 responding nearly to the double zenith distance of the star. By 

 fixing temporarily in the immediate field of view of the eyepiece a 

 cross of wires, or by planting a microscope for the opcasion above 

 the micrometer frame, the interval between these two wires may be 

 found very accurately in terms of the revolutions of the micrometer ; 

 and by fixing other wires on the micrometer plate at intervals as 

 nearly as possible equal to that interval, and by using a series of 

 microscopes fixed for the occasion, the micrometer-scale intervals 

 between all these wires may be very accurately found. And by 

 turning the object-glass and attached micrometer to a position 90° 

 distant from either of the positions of observation, and observing the 

 transits of zenithal stars over all the wires, the intervals in arc may 

 be found. The combination of these will give the best possible in- 

 formation on the value of the micrometer- scale, and on the intervals 

 of the wires. 



A simple micrometer might be used for the observation ; it would, 

 however,have these disadvantages; that the micrometer must be read 

 between the two observations, and that the observer could not use 

 the same hand in the two actions upon the micrometer head. The 

 Astronomer Royal proposes a more complex micrometer, in which 

 the micrometer B, to which the bisection- wires are attached, is car- 

 ried by and has its screvf-appui in a micrometer A ; and the micro- 

 meter A has its screw-appvi in the cell of the object-glass. There 

 is no difficulty in so arranging this that the movement of micrometer 



X2 



