CLOUD OBSERVATIONS IN VICTORIA. 268 
The lenses are “ Zeiss” anastigmat, series IV, No. 3, focal 
length 4:41 inches, beautifully paired. : 
The instrumental adjustments are—(lst) To make the optical 
axes of the cameras truly vertical ; (2nd) To make one of the 
central lines of the reseau at both stations coincident with the 
base-line. 
The first is made in two steps, as follows :—A small telescope, pro- 
vided with a micrometer, is mounted vertically on an independent 
tripod, looking downwards, with its objective a few inches above 
the photographic lens and central with it. A small dish con- 
taining mercury is then placed between the camera and the 
telescope, and by means of a Bohnenberger eyepiece we observe 
the reflected image of the telescope wires, and cause them to 
coincide with their direct image. We have thus made the colli- 
mation axis of the telescope truly vertical. The mercury dish is 
now removed, and a plate containing a developed image of the 
reseau is placed in the camera in the usual way as if it were 
exposed for taking a cloud picture. 
The centre of the plate is now illuminated by means of a 
small electric lamp, and we look at it through the telescope. 
If its image is coincident with the intersection of the wires in 
the telescope, then the collimation axis of the camera must be 
parallel to that of the telescope, and, therefore, truly vertical. 
If it is not coincident, the required rectification is made with 
three levelling screws provided for the purpose. 
The second adjustment was made approximately in the first 
instance by stretching a long wire between two poles on the base 
line, and orientating the camera simply by the eye, and sub- 
sequently, for further and more accurate verification and rectifi- 
cation, simultaneous photographs of the sun were taken on every 
possible opportunity. The azimuth adjustment is also verified by 
the condition that the co-ordinate, at right angles to the direction 
of the base line, ought to be the same in both pictures. 
Both these adjustments are very constant, and there has hardly 
been any necessity to make rectifications since last July. 
With ordinary care, these adjustments, once made, remain 
practically constant for a very long time. 
The are value of 1 millimeter on the plate is about 35’. 
The co-ordinates can be measured off the negatives, even 
roughly, with compass or scale to within a couple of tenths of a 
millimeter, and, if the cloud points selected for measurement were 
sufficiently well-defined, the probable error of the co-ordinates 
might be reduced to half that amount. 
Such accuracy, however, is not attainable, owing to the nature 
of the objects photographed. The pictures have no sharp out- 
lines ; their structure is evanescent, and can hardly bear any 
magnification. 
