On the site of Lars Bruzelius the specification of the brace winch can be found as it is drawn up by Capt. Jarvis in 1890. The specification speaks for the greater part for itself. Yet there are a number of items which are outstanding:

        1.    In his specification Jarvis is bracing a whole mast with one winch, by putting every other yard on the winch, the lower yard, the upper topsail yard and the upper topgallant yard. The remaining yards are rove conventionally with rope braces. The manually operated yards follow the winch operated yards when bracing. This means the rope braced yards had to be eased off and hauled in during the operation.
    In later years, however, you always see the three lowermost yards are being put on the winch: the lower yard and both the topsail yards. The reason for this might be that Jarvis, when writing the specification about the winches, did not realise (yet) that the brace winch could contribute in the safety of the crew.
    When the complete mast is braced by the winch the crew always has to man the manually operated yards for easing and hauling. But at times when that job becomes dangerous, when a loaded vessel is sailing in a gale and green water is coming over the lee rail frequently, one is mainly bracing around yards with furled sails! After all, the topsails are the last to be taken in. Therefore it makes more sense to put the bottom three yards, with the topsails in between, on the winch which, by its location amidships, is out of reach of the most massive green water.

bark Padua, by..
There is something about this painting...,
something that is not in accordance with the text.

        2.    The braces are carried from the yard-arms to the mast behind it and from there guided by blocks to the winch. In Germany, according to Middendorf (see next page), an arrangement was used where the braces of the lower yard runs to the rail, from there to guiding blocks under the top and thereupon to the winch.

        3.    The hauling parts of the braces are rove on the smallest diameter of the drums, and not as the picture below shows, on the largest diameter. It is crucial to understand this.

This is a wrong example of roving a brace winch

 
        4.    The last but one paragraph of the specifiction holds the point of the chapters to follow: the coning of the drums being proportional to the differences in length of the braces when square and when braced up sharp on either tack and the size of the drums being proportional to the actual lengths of the braces.

        5.    Capt. Jarvis does not mention any other dimensions then the distance of masts apart and the size of the yards. No vertical dimensions, no trusses or parrals, no brace pendants etc. Perhaps he only calculated the length of the braces roughly, the fine tuning of the winch had to be done in practise anyhow, a foot of rope wire more or less should not make any difference in the sizes of the winch. And perhaps he just measured the length of the braces on the ships he was sailing on.
Jarvis also does not explain clearly how he achieves the coning of the drums to match with the length of the braces. Middendorf, in his book Bemastung und Takelung, is more explicit about that. He dedicates two paragraphs to the brace winch.

 

Continue.