5 
it. There 
japer, but 
ree optical 
ed stereo- 
luplicated. 
ious) idea 
ecise inea 
rthing de- 
ly and as 
it cannot 
g operator 
ler has to 
ps, taking 
lence with 
omparator 
pic corres- 
:ator, once 
3corded as 
antage lies 
is a great 
th pictures 
re pictures 
ences with 
at be more 
ring equip- 
and these 
are prima 
the mark- 
monocular 
negligible 
be greater 
ar methods 
e measure- 
ention and 
.e when we 
dor who is 
Of course 
bservations 
ie telescope 
important data for aerial triangulation is obtainable without any marking of the 
pictures at all. The transfer of rotations through a strip of photographs (they 
cannot in any event be transferred laterally with conventionally flown strips) 
depends only, or can be made to depend only, upon the relative orientations which 
do not require marked points. The transfer of heights certainly requires some 
knowledge of the scale but unless the ground is very accidented a not very pre 
cise scale is required and the marking necessary need not be of the highest preci 
sion. For medium and small mapping where a desired accuracy is more difficult 
to obtain in height than in plan *, it is true to say that it is of prime importance 
to make the measures in such a way that the very best possible heights are obtained 
and would not this be by stereoscopic measurements to unmarked, or, at any rate, 
to points marked on once picture only? 
The whole problem of transfer of position and scale both along and across the 
strips is, in fact, adequately solved by marking only one picture of an overlapping 
group by a circle of a size exactly that of the stereoscopic mark in the instrument. 
The marked point need not be placed precisely at a particular point on the picture, 
provided that it is in an area which has good stereoscopic character, and that its 
size is carefully controlled to ensure accurate monocular bisection without distract 
ing the operator when he makes his stereoscopic measures. In what way, both 
operationally and in precision, can the monocular measurement of fully prepared 
pictures be an improvement upon this? Let us hope the discussion on this paper 
will tell us. 
Before leaving the subject, perhaps a word should be said about three-table stereo- 
-comparators. They were presumably produced to enable strips to be observed 
without any point marking at all; but, while the author has had no experience 
with the use of these instruments, they do seem to have a more valuable use than 
this which, as we have seen, is not strictly necessary. An instrument with a third 
table does enable observations to be made on the photographs of adjacent strips 
without upsetting the routine observations of the strip in hand and this has much 
to recommend it. It is, of course, necessary that the instrument be designed so 
that any arbitrary point on one photograph can be brought into stereoscopic corres 
pondence with an arbitrary point on any other. Whether the cost of such an 
instrument can be recovered by the increased efficiency is a matter simply for 
experiment under working conditions. 
There are advocates for monocular measurement of pre-marked points; for ste 
reoscopic observation in twotable stereo-comparators. It would be gratifying if 
the discussion on this paper could throw some light on the reasons for the existence 
of these three groups. 
•ked photo- 
unless the 
ivocally on 
y fact that 
* At 1/50,000, 1 mm is 50 m on the ground. Who would seriously complain of an error in 
absolute position of 1 mm on the map? Rut an error of 50 m in height would not be tolerated.