| Instrument | : | Demi-cross (Demi-arbaleste) |
| Manufacturer | : | N. de Hilster (original unknown). |
| Country of origin | : | The Netherlands (Original also) |
| Manufacturing year | : | 2007 (Original: at least from 1621) |
After finishing the spiegelboog reconstruction I wanted to reconstruct another navigational instrument that no longer exists. A few were on my list and one of them got first: the demi-cross. It is not known who invented the demi-cross, but the shape clearly shows resemblance with the first backstaff Sir John Davis presented in his book Seaman's Secrets in 1595.
While researching another instrument on my list, the hoekboog, I came across drawings and an extensive description of the demi-cross. The instrument has similar features as the cross-staff, but has an asymmetrical staff and uses a horizon vane and sight vane sliding along it. The cross is only half a cross (hence the name) with on it a shadow vane that can be positioned at four different levels.
Although the instrument has been described for at least 70 years by Dutch authors, none of them survived. As none are left I made this reconstruction using the information various authors left in their books.
After a year of research, building the instrument took me about 120 hours. Now the instrument should be a 95% accurate reconstruction of the original. A second specimen of this instrument has been made as part of a 17th century navigation set for a museum in Hirado, Japan.
On November 5th, 2007 a field test was carried out in order to judge the usability of the instrument and to compare it to the Davis Quadrant, which must have been the competitor to this instrument back in the 17th century. I was joined in this test by Nico Duyn, Jan Jonker and Jaap Ypma, all experienced navigators. Compared to the Davis Quadrant the demi-cross was harder to handle due to its size and susceptibility to wind, but it performed much better. When calculating using all the recorded data we find an average error of only -2.0 arc minutes with a standard deviation of 4.9 arc minutes. When calculating using the data from 12:18 onwards (see graph) the average error is slightly worse with -2.6 arc minutes, but at 2.3 arc minutes the standard deviation is much better. The Davis Quadrant had an average error +13.7 arc minutes with a standard deviation of 8.5 minutes for all the data and +10.5 arc minutes error with a standard deviation of 5.1 arc minutes as from 12:18.
Materials used on this reproduction are ebony for the staff and pear wood for the vanes. In addition to that brass is used for the spring in the cross and the screw in the shadow vane. The horizon vanes have a white painted area.
I'm very grateful to Diederick Wildeman and Anton Oortwijn (both Scheepvaartmuseum Amsterdam) and Sjoerd de Meer (Maritiem Museum Rotterdam), as they supplied me with lots of information on this instrument and other contemporary instruments by allowing me to study period literature.
If you like to know more, don't hesitate to contact me.