precisely
‘umental
initially
s elimi-
ormulas,
l to the
uence of
tal COr-
1stment,
type of
S paper
oraph is
carried
respond-
eometric
his, the
e intro-
'o perties
"mation,
rvature,
ne little
laxes in
je asked
'elations
at they
; under
egarded
applied
as we
typical
ered by
ence of
e factor
he type
)bserver
iblandit iiti liaison dus...
Owing to the fact that the x introduced into the final connection
is à x already corrected, the above mentioned correlations of a geometrie
nature do not find any more occasion to be applied, whereas the corre-
lations connected with the parallaxes observations can still furnish a
useful contribution. After that we note that the planned extension of
Roelofs’ method to the solar triangulation mainly consists of an appli-
eation to the triangulation method utilizing bz component.
We add that thesolar triangulation, according to its original pro-
cedure, works with the component bz at the center of the model and
can make use of both the cameras for the elimination of the residual
parallaxes on wo and o. The present application of Roelofs method re-
quires, instead, that the component bz be used according to the tra-
ditional way, along the area of triple overlap, and admits only the
movement of the eamera newly introduced.
Moreover, while the solar triangulation, on the acknowledgement of
the principle of the “non transmissibility’ of errors ”’ restores the ori-
ginally computed solar values of w and ¢ in the photograph ‘whieh re-
mains in place for the formation of the new model, this present application,
in which the new camera is the movable one, calls for the renounce
to that restauration. This derogation to the solar praxis, which is in
contrast whith the most important feature of the solar method, can
be avoided by operating whith the old camera instead of the new one.
The analytical examination of this case, which, practically, must be
followed, will be made afterwards. (1)
In this paper all the suggestions and supporting relations established
by Roelofs will be followed.
2. — Quasi observations.
The following seven quantities are introduced as ** quasi observations ^
to be adjusted. In the pair i — 1, i:
«€; = the lateral tilt of the photograph i
x; = the swing of the photograph i
(1) It is to he noted that also in the above mentioned last procedure the photograph
Which is in connection with the preceding model undergoes changes which represent a break
between the models. This interruption of continuity is the most important objection to the
applieability of the correlations method, which rely now only upon the smallness and rare-
faetion of the changes.
See, also, paragraph 5-3- (b).
A EG