Instrument	:	Theodolite type T2
Manufacturer	:	Wild HeerBrugg
Country of origin	:	Switzerland
Manufacturing year	:	1971

I've obtained this theodolite in 2001. The instrument is of Swiss origin and was one of the best theodolites of its time. This is a state of the art instrument with, for its time, unprecedented accuracy (apart from the T3 and T4 models).
The manufacturer, Wild Heerbrugg merged with the Cambridge Instrument Company plc creating the new Leica Holding B.V. group on April 2nd, 1990.

Both the horizontal and vertical circles of the instrument are divided in gon (400 divisions in a full circle) and can be read down to 0.0001 gon (0.3 arc seconds) (and estimated one further digit, although that is beyond the instrument's accuracy).

In July 2008 I managed to lay my hands on a Roelofs Prism, an attachment to take sun shots. These sun shots are used to determine true north in the field, which can be useful when using a transit as with those one needs to know the magnetic deviation. The attachment slides over the telescope and is fastened using a clamping screw. The Roelofs Prism superimposes four images of the sun onto each other (see image 7). The areas where these images overlap will be brighter than the rest and form a perfect target to aim at with the cross-hairs. As the attachment simply slides over the telescope, the measurement has to be repeated while transiting the telescope and turning the instrument 180 degrees in order to eliminate any centring errors.

Figure 9 shows the centring errors of a series of eight observations. In figure 10 the observations are corrected for the average centring error. The final result is the combined accuracy of instrument, software, clock and surveyor. Note that the observations were taken at the worst possible moment: around noon. Best moments for sun observations are just after sunrise or just before sunset as the horizontal angular movement of the sun is at the lowest around that time (and hence the accuracy is less influenced by timing errors). As can be seen from the graph the errors of sun azimuths are typically less than 0.005 degrees (about 18 arcseconds). The average error was a very lucky 0.0000gon (around the time of observation the sun's horizontal angular speed was 0.004 gon/sec) with a 0.0025gon standard deviation (1sd, 68%).