The important bit is this here:
And that is it! Hornell stops there, and goes on to describe a different boat. The man who declares shunting to be unseamanlike doesn't think it's worth bothering with an explanation of where, when and how the cross beams and outrigger are detached and re-attached again, beyond how they are lashed. I don't see how it could be done at sea, unless in a dead calm and first taking down the rig. Given a choice, I would prefer shunting to changing tack by rebuilding at sea.But the most remarkable divergence is that these outriggers are made to unship instantly by a very simple device and can be rigged out on the other side of the boat, thus avoiding the unseamanly custom of the Sinhalese who can never say which end of canoe is the head without looking at the direction they may happen to be sailing !
The Kilakarai boats can therefore employ a rudder fixed at the sharp curved stern by the orthodox pintles and gudgeons. The details of the stays led to the outrigger are readily seen in the sketches given. The device for attaching the booms inboard on the one side and to the outrigger on the other is a form of the Spanish windlass, well known to sailors and shipwrights for exercising force in bending a plank into position and holding it there till secured; the principle is that of the surgeon's tourniquet. In the present instance the loop end of a ring of rope or grommet is passed through a hole either in the gunwale or in the outrigger float, the boom pole is laid over this and the looped ends of the grommet are brought up at each side and over the pole; the end of a short rod or stake is passed through the two loops and then by the simple device of twisting the two loops round one another by means of the rod, the two main parts are bound together with the greatest possible tightness. The free end of the tourniquent rod is then seized to the gunwale or the boom as the case may be, and if the various parts be sound this lashing will maintain attachment under any ordinary violence. Its chief advantage lies in the rapidity with which it can be operated. A couple of seconds suffice to release the boom and scarcely more are required to reship it.
An obvious solution is not to rebuild at sea. Peter Evans also offers a description of another one-way outrigger boat the Yatra Dhoni: http://www.tackingoutrigger.com/yatra_dhoni.html. That is a trading vessel sailing in predictable monsoon winds. I don't see how that applies to the Kilakarai outrigger, which seem to be smaller vessels. Also, Hornell has this to say:
As I understand it, among the modifications making outrigger boats more useful for general purposes are the permanent bow and stern, a stern-hung rudder with gudgeons and pintles, and the outrigger that is detached and re-attached on the other side when you want to sail on the other tack. How does that make the boat more practical?yearly they spend some six months in north Ceylon waters fishing chank shells. Hence their relations are most intimate with Ceylon and this may
have determined the outrigger canoe as the type of boat adopted here for general fishing purposes. But while the true Ceylon outrigger in its typical form is largely used for trolling by these men,the boats being imported ready for work from Ceylon [...] while agreeing in the adoption of the outrigger principle, have shown marked ingenuity in modifying it in several notable details, whereby it has become much more useful for general purposes
I have failed to find any more information on this. Does anyone else know more?