140 
The measurement of the coordinates of a point being made 
independently on each plate, although simultaneously, it will be 
a sufficient condition for the viewing apparatus to make corre- 
sponding portions of the two pictures combine with or without 
change of perspective. 
Using a magnifying optical system to view the pair of plates, 
the condition for distinct vision is that the two images of any 
point appear in a corresponding plane of vision, so that the 
visual rays meet in space. This condition evidently remains 
satisfied when the images are magnified, or when they are 
brought nearer together along a line parallel to that joining the 
nodal points of the two eyes, and for different distances 
between the viewing lenses or the eyes, since in all these cases 
the lines joining the two images of a point remain parallel to the 
eyes. 
Surveying Camera.—The essential features are a camera on a 
theodolite base, and a telescope with its line of collimation at 
right angles to the optical axis of the camera, so that by changing 
pivots the orientation of the pair of plates is not affected by errors 
of inclination, collimation or graduation. 
The photographic plate is pressed, during exposure, against a 
back frame in the focal plane of the camera lens, by a spring 
contrivance, similar to those used in other surveying cameras, 
which permits the shutter of the dark slide to be drawn and 
replaced. The réseau is hinged in front of the plate, its correct 
register being determined by geometrical contacts. It is impressed 
Plan 
Elevation 
upon the plate by exposure to sky light reflected through he 
camera lens and then moved out of the way for the expose 
the picture itself. A graduated front slide is used to displace the 
horizon line by moving the lens, but in normal circumstances lt 
is set at the zero of its scale. Fig. 2 shows the general arrange- 
ment of the instrument. 
Conditions to be satisfied —One ee 
sufficiently satisfied in construction, 1s that the front slide He 
parallel to the vertical réseau lines. Any _ defect in this 
respect is eliminated by determining the origin of the veel 
coordinates and the focal length for different readings of the 
scale. ; ; 
nT he camera adjustments are : (1) Plane of réseau to be vertical. 
(2) Horizon line of réseau to be horizontal. _These adjustments 
are made with the aid of a level, fitted with a Bohnenberger 
ye-piece. ; ; 
the auxiliary level having been placed directly in front of the 
camera and its line of collimation made horizontal, the vertical 
axis of the camera is set vertical by reference to the level of the 
vertical circle. Then (1) is effected by turning the camera in 
altitude with the footscrews, and in azimuth, until the cross wires 
of the level coincide with their image reflected from the silvered 
back surface of the réseau when the bubble of the longitudinal 
level on the camera is adjusted to the centre of its run, 
Replace the front slide and lens and set again the vertical axis 
NO. 1701, VOL. 66] 
instrumental condition, 
NATURE 
[JUNE 5, 1902 
vertical, (2) is now effected by making the ends of the horizon 
line of the 1éseau coincide with the cross wires of the level in two 
positions, using for the purpose the side capstan-headed screws in 
the base. The transverse level on the camera is then adjusted, 
and the longitudinal level made perpendicular to the vertical axis 
by means of the front capstan-headed screws under the camera. 
The theodolite adjustments, effected by ordinary methods, are : 
(3) Horizontal axis made perpendicular to vertical axis. (4) For 
collimation. (5) Horizontal axis made parallel with optical axis 
of camera. An approximate adjustment of (5) is sufficient. 
Instrumental Consiants.—These are (1) the zero of the front 
scale, (2) the zero of the réseau and coordinates of the R-points, 
and (3) the focal length. They may be determined in the usual 
manner, but it is convenient to first make the centre R-point 
coincide with the zero of the réseau coordinates by collimating 
directly upon the réseau plate when adjusting the camera with the 
help of an auxiliary level as already explained. In that case the 
lens requires to be adjustable horizontally as well as vertically. 
The focal length / is found from the measurement of exposed 
plates containing the images of well-defined points of which the 
angular distances are known. Call a the angle between two 
points of which the horizontal coordinates are a@and 4. Then: 
CO, | =e, 
J tana’ N 4tan?a 
Measurement of the Plates.—It is unnecessary in a preliminary 
note such as this is to enter into the construction of the measuring 
apparatus in much detail, as a description of actual instruments 
with examples of their use may fitly be givenin a subsequent paper. 
A suitable machine would generally resemble those which have 
been used for the measurement of celestial photographs, and like 
such may be of various types. 
Fic. 3. 
In the type now considered, the plates are set side by side at an 
inclination corresponding to that of the base line and at heights 
| such that corresponding R-points are horizontal. Both plate- 
carriers can slide about ina horizontal direction on a stage formed 
of a sheet of plate glass ¢ (Fig. 3), which itself can be moved verti- 
cally by a double rack and pinion. Any small error in the setting 
of the plate and in the fitting of the slides will be automatically 
corrected by the position of the eyes in front of the eye-pieces of 
the viewing microscopes and by their power of accommodation, 
and does not affect the accuracy of the measurements. E 
The measuring microscopes are of low power and include in 
their field at least one clear R-square of I centimetre side. Their 
distance apart is adjustable to suit the eyes of the observer. One 
is fitted with a pair of micrometers at right angles capable of 
rotation in order to bring the horizontal and vertical wires 
parallel to the R-lines. The other is similarly fitted, with the 
exception that one horizontal micrometer is sufficient. The runs 
are adjusted on a scale. 
The centres of the plates are separated to a sufficient distance 
by introducing in each microscope a pair of prisms of total 
reflection / (Fig. 4). 
The micrometers might also be used in the position of the plates, 
giving more room for the screws and greater facility in the 
reading of their heads, and the plates themselves set further 
back, behind an additional lens, as in the Cambridge measuring 
machine recently described by Mr. Hincks (Monthly Nottces, 
Ixi. p. 444). ; ’ 
The zero wires form a frame fitting an R-square, as in Sir 
David Gill’s machine used at the Cape Observatory (Monthly 
Notices, lix. p. 61). 
