proposed route crosses existing roads or railways. To avoid having to reset 
models, the siting of these additional cross sections should be agreed by 
inspection of the air photographs before work commences. 
When cross sections are observed directly, it is again necessary to 
adopt the profiloscope technique. Using the result of the latest scheme stage 
horizontal alignment calculation, a plot is prepared of the reference line, 
stations and normals along which the profiloscope will be guided. It is still 
common to find observations recorded by hand, this being probably no more 
expensive than automatic recording of XYZ coordinates followed by computer 
conversion into measures of offset and level. But, as noted before, manual 
recording and manual data punching are operations from which it is difficult 
to exclude human error. 
As to the style in which cross section data must be presented, this 
is dependent on the design system that has been chosen. There are two basic 
styles - one in which levels are taken at standard offsets from the reference 
line (the offsets need not be recorded since they are written into the program) 
and the other which permits levels to be taken at change-of-slope points (here 
the offsets, being irregular, must be recorded). At least one design system 
admits a combination of both styles and this has proved rather troublesome 
because photogrammetrists are prone to forget to insert the special character 
which denotes a change of style. This omission is far from obvious when 
examining the data, and has the unfortunate effect of causing the computer to 
confuse offsets with levels. 
Circumstances permitting, mapping and cross sectioning (or DTM 
heighting) will proceed simultaneously, the guide sheets for the profiloscope 
serving also as mapping sheets. But it should not be thought that because of 
this onlyone field party will be needed to check both. The engineer will be 
anxious to receive the first few miles of cross sections so that earthwork 
calculations may begin, and therefore it must be expected that two parties will 
be required: an advance party, to complete obscured parts of the priority cross 
sections, followed towards the end of the mapping by the main party. 
- ll -