JUNE 5, 1902] 
NATURE 
14 
For convenience the whole arrangement is tilted at an angle | 
of 45°, and the light illuminating the plates reflected by mirrors 
m from a window at the back of the observer. 
The setting of the plates may be effected by turning a 
micrometer to the inclination of the base by means of a graduated 
circle, and making both sets of R-lines agree in inclination and 
height with the micrometer wires. The second micrometer is 
then set by making its wires parallel to the vertical R-lines on 
either plate. 
The vertical R-lines are combined by the microscopes, but 
the horizontal lines only when the distance between the centres of 
the pictures is equal to that between the microscope object glasses, 
In making a measurement, the plates are moved by the slow- 
motion screws on the slides of their carriers and of the stage 
until the zero square of one microscope fits a zero square of the | 
corresponding plate and the zero wires of the other microscope 
coincide with a pair of vertical R-lines on the second plate. The 
points in the field of view may then be bisected without disturbing 
the zero settings. 
The coordinates of a point on the plate are given by the 
direct readings of the micrometer heads added to the value of 
the R-line considered. The stereoscopic difference results from | 
the difference of the x’s on the two plates. } 
Range of the Method.—In practice, the range of the method is 
limited by the blurring of distant detail by light diffused in the 
atmosphere. This ‘‘aérial perspective” is reduced by the use of | 
orthochromatic plates and an orange screen cutting off the rays 
of shorter wave-length which form the blue haze, but even then 
| within a confined space, as when mapping hidden valleys. 
Fic. 4. 
the effective range would probably not exceed some 5 miles, or 8 
kilometres. 
On the other hand, the difference in phase of the objects 
would prevent their ready combination at distances less than 
three to four times the length of the base. The view would then 
correspond with that of a model seen with the eyes at a distance 
of 10 inches from the nearer edge. : 
Let 2é be the length of the base and a the angle subtended by 
it at adistancey. Then: 
ay =A cot Es 
2 
ay _6 da 
Ba Y 2sin2% 
2 
ao. 
sin a 
Let 1/1ooth of an inch or 0'25 mm. be the admissible error on 
the plan, 8 kilometres the limiting value of y and Aa=20’". 
the scale of the Canadian photographic surveys, 1/40000, the 
maximum error allowable will be 10 metres at $8 kilometres, or 
Ay/y=1/800. Then a=4° 27’ and 24=620 meters. 
By increasing the base to 2 kilometres, a maximum possible 
accuracy at 8 kilometres of 1/2500 of the distance, or 3 metres, 
would be attained, but the area mapped would be reduced to a 
narrow strip. 
With the base of 620 metres,!the area mapped with a plate of 
NO. 1701, VOL. 66| 
On | 
diameter equal to the focal length of the lens would be contained 
between the limiting circles, at 8 and 2°5 kilometres, shown at 
@ and # (Fig. 5), and would amount to 22 square kilometres on 
either side of the base, or more correctly to that portion not 
masked by the nearer topographical features. 
The error in x will be due to that in y and that of the x-co- 
ordinate on the plate. We may write: 
(Ax)?= G a)+(% ay). 
With a lens of 150 mm. focal length and an error of ‘025 mm. 
in the plate xs, the maximum error is, for the base and the scale 
of plan considered, 5 metres, or on the plan o'12 mm 
The errorin height is given by the same expression. At the 
maximum distance, the second term cannot exceed (1/4 Ay)? if 
the difference in height between the base and the distant points 
does not exceed 2000 metres. In absolute amount the total error 
for points at extreme distances would be + 2°75 metres. 
The contour lines should then, in the case already considered, 
be accurate to 0°25 mm. on slopes greater than 15°, but the 
actual accuracy will be reduced to some extent by the uncertainty 
of the correction forrefraction. This correction, combined with 
that for curvature, can be applied at sight from a small table 
with y-argument. 
By reducing the base, pairs of photographs may be taken 
The 
> 
ye 
Fic. 5: 
method can also be combined to any extent with the ordinary 
methods of photographic surveying. It would be of particular 
advantage in the mapping of large areas of mountainous country. 
UNIVERSITY AND EDUCATIONAL 
INTELLIGENCE. 
OxFoRD.—The following examiners have been appointed in- 
the Natural Science Schools:—Mr. William B. Croft, Pem- 
broke College (physics), Dr. Alexander Scott (chemistry). Dy, 
Leonard E. Hill (animal physiology), Mr. John Watts, Ballio} 
College (chemistry)—vzce Mr. Elford, resigned. 
THE 237th meeting of the Junior Scientific Club was held on- 
May 29, in the museum. Mr. A. F. Walden, New. College, 
demonstrated a new method of distinguishing between calcium: 
and strontium. Mr. E. A. Cockayne, Balliol, exhibited a 
| natterjack, and Mr. Lattey, Trinity, read a paper on the 
| occurrence of natural gas in England. 
Mr. Davin Rogerrson, lecturer in electrical engineering 
at the Bradford Municipal Technical College and formerly 
assistant lecturer at the Glasgow and West of Scotland 
Technical College, has been appointed professor of electrical 
engineering at the Merchant Venturers’ Technical College, 
Bristol. 
THE annual exhibition of work from schools and classes of the 
School Board for London will be opened by Lord Reay on 
