4 
efficiency seems to be fluctuated very easily. 
The efficiency of A7 seems to be far more better than KRC, but 
this is a mere appearance. The main reasons are in the incompleteness 
of read out device of comparator and operator’s unskillfulness. It is 
sure that in near future it will reach the same level as the case of A7. 
Fig.3 is the statistics for man-hour of marking of points on dis 
positive in the period from July 1961 to March 1964. With the settling of 
electronic computer June 1961, we organized the section of marking and 
after 2 months the man-hour per model reduces from 1.2 hours to 0.8 hours 
and gradually reduces until 0.6 hours in one year and a half. Here and 
there the curve fluctuates mainly because young new operators join to 
the job. This inclination is most conspicuous in the period from May 
1963 to September 1963» This is due to the increase of pricking device 
and participation of young new graduates. From the curve it is suggested 
that at steady state the man-hour may reach 0.5 hours per model. Number 
of points in one model are in the average about 10, so that 3 minutes 
for one point will be considered as the utmost efficiency including the 
miscellaneous minor jobs. 
In Fig.4 we see the statistics for preparation work. This work 
is most important in the whole system of practice and is done by skilled 
specialist. Fig.4 shows that it needs about three quaters of an hour 
per model. The important thing in this phase of work is the selection 
and nomination of the points to be measured on paper prints. It is includ 
ed, naturally, the other jobs in this phase, for example, the general plan 
ning, scheduling, determination of the method for identification, its 
actual planning, preparation of dia-positives etc.. In the Fig.4 we see 
that in some month it needed extraordinary long time for preparation. 
This is due to the large amounts of works for which our client requires 
only aerial triangulations, not including plotting works. This kind of 
work needs some additional jobs which may not be necessary in the work 
including plottings by our selves. 
A little we shall mention about the time necessary for computation. 
The time for computation depends, at the first place, on the scale of com 
puter. But actually it depends on the quality of the raw materials. And 
the organization of the computing section is the third factor which effects 
the efficiency of computation. The kinds of equations for solving the 
problems are only insignificant factors and there are little needs of 
considerations for the selection of the kinds of equations. The main parts