The material we have today indicates that shell building was known and practised in the Mediterranean in classical antiquity. Comparative material shows that the‘intermediate stages’between shell and skeleton shipbuilding are in most, or all, cases shell building with some skeleton influence. The date when‘the skeleton idea’first struck a boatbuilder is not established. There are indications that this happened in the Mediterranean. No definite proof of skeleton work is found in the Mediterranean earlier than the Yassi Ada wreck, where a bottom shell may have been equipped with a skeleton on which topsides were built. The shell-skeleton division is a working tool for research on shipbuilding history, and should be used critically. The wrecks from Blackfriars I and New Guy's House indicate a kind of skeleton approach to shipbuilding in northern Europe (Marsden, 1967). If the 'skeleton idea’existed in the Mediterranean in classical antiquity, especially in warships, proof will turn up sooner or later. In the meantime, publishers of shipwrecks should take into consideration that we know very little, and that all details count. Field work on the existing tradition in the Mediterranean should be pursued. The lesson learned in Scandinavia is that the archaeologist may theorize as much as he wants, but the boat-builders will be able to give definite answers, if he takes the trouble to ask them, and watch them at work.