5 
of programs are occupied by the instructions which do not relate directly 
to mathematics, but relate mainly to the method by which we give as wide 
flexibilities as possible to the computer. Computer has extremely high 
fidelity, but it is obstinately unflexibe and we have to compensate this 
obstinacy by the programming technique and systematic organization of 
their library. Smooth but flexible utilization of the above organization 
is attributed to the organization of the staffs of the computation section. 
The time for computation much differs according as the combinations 
of the above conditions. At first, our computer has access time of about 
1 mili-second, and it needed theoretically about ten minutes per model 
for longest case of computation, that is from correction of photo-coordi 
nates to the final result through block adjustment of strips. But 
actually in the average of whole one year it needed about half an hour. 
Two years later we introduced high speed core memory which has the access 
time of about 5 micro-seconds. It needed about 5 months to convert the 
system of programs to meet to the new device and the net time of computer 
was reduced to about one third of the before. But at the average, it seems 
that the actual time was scarcely reduced to one half of the before. 
Programs and the system of programs have to be improved always, adding 
new ideas maily obtained from experiences. At the beginning, a few 
personnel were working for every kinds of jobs, for programmings, operations 
of computer, preparations, administrations and theoretical investigations. 
But after one year, the section of computation was organized and quite 
recently it seems to have been established firmly. According to this 
organization, 4 personnel are working for the improvements of program 
(not only for aerial triangulation, but also for many other programs), 
6 operators are working to move the computer 400 hours in a month and 2 
punchers for imput data. Since in our system all data are obtained direct 
ly from comparator, we need comparatively small amounts of imput data to 
be punched. Quite large amounts of prints of results are produced daily 
and filing operations of these prints may not be ignored as a job of 
operators and punchers. Administrative businesses require 2 other personnel 
and for maintenance of computer one technician is provided. Those are all 
very young people about 20 years old except one electronic technician. 
Adding one supervisor and two sub supervisor our organization is composed 
of. 
Now, our computer works not only for analytical aerial triangulations, 
but also for every kinds of computations that which in our company have