Bt 
JUNE 2, 1923] 
was then given a final adjustment in detail and the 
prints stuck down in place after removing all the 
elastics. When this work had been completed, some 
40 points, distributed over the surface of the mosaic, 
were measured up and compared with the Ordnance 
map, and it was found that there were no errors of more 
than 60 yards, except on the extreme northern edge of 
the part of the mosaic that was made without gyro 
control on the rudder. In this region errors up to 130 
yards were recorded. 
The scale to which the template had to be constructed 
came out at 1/19,930. The difference between this 
scale and that of the separate mosaics compiled from 
the same photos was possibly introduced during the 
systematic adjustments. 

NATURE 
747 « 

a good measure of control between very widely spaced 
ground- surveyed points. If, for example, the photos in 
these flights be taken at exac tly equal time intervals and 
the positions of the ends of the strips be known, the 
centre of each intermediate photo could be determined 
with considerable accuracy. 
We have in hand experiments upon a scheme for 
using these indication strips, together with a few long 
strips at right angles, to control the positions of the 1o- 
mile squafe units. We estimate, on good but not yet 
conclusive evidence, that, representing these prelimin- 
ary strips by elastic bands and stretching the frame so 
formed to fit control points, we could so distribute the 
errors that the 100 square mile units mosaics could be 
located in position within } to } a mile, even when the 


Fic. 2.—Showing method of compiling the 15-mile square photographic map to fit four control points. 
weights used to hold them in place after their final adjustment, 
shown. 
VARIATIONS OF HEIGHT. 
The country of which this map was made contains 
local differences of level up to 300 feet in several places 
and on the extreme southern edge rises to 500 feet above 
the lowest part. 
NAVIGATIONAL CONTROL. 
In some types of country, control points even of ro- 
mile spacing may not be economically obtainable. The 
complete absence of control points would not, as we 
have seen, seriously affect the accuracy of the indi- 
vidual mosaics as regards distortion, but it may leave 
their average scale in some doubt. There may, there- 
fore, be some difficulty, in these circumstances, in 
fitting the mosaics together and in controlling their 
relative positions. It will be remembered, however, 
that it was necessary, for the identification of the start- 
ing points of the mosaic strips, to begin a survey by 
laying down identification strips spaced 10 miles apart. 
The most economical way to do this, when mapping over 
large areas, would be to commence by laying down a 
series of long parallel strips to miles apart having a 
convenient length of, say, fifty miles. If care is taken to 
fly these indication strips very straight and at a con- 
stant speed, it may be possible to use them to give quite 
NO. 2796, VOL. 111] 
The prints are hidden by the 
ut the white elastic bands to which the prints are fastened are clearly 
The black elastic bands were added to facilitate systematié straining at right angles to the direction of the strips. 
control points are spaced as much as 50 miles apart. 
It will be noted that this method of locating position 
is unaffected by the presence of hills. 
We do not recommend using control points so widely 
spaced as this, but we are concerned to show that the 
mosaics could be located with moderate accuracy even 
when the control points are so far apart. 
The methods upon which experiments are in progress 
would not be limited to use with control points forming 
any particular pattern ; they could be used with any 
form of triangulation. If, however, triangulation is 
impracticable, as it may be in flat wooded country, we 
are informed that astronomical methods, carried out 
on the ground, with the help of wireless time signals, 
can be used to locate position within 100 to 200 yards. 
If this method could be employed, therefore, in conjunc- 
tion with aerial methods, it would be possible to push 
accurate mapping into unsurveyed country in which 
the ordinary ground - surveying methods, based on 
triangulation, are impractic able. 
TRAINING AND EQUIPMENT. 
The methods that we have described require con- 
siderable skill and special training on the part of the 
pilot and observer. If they are not adequately trained 
