LaFiteau, writing before 1724, stated definitely that the Iroquois did not build any birch-bark canoes, but obtained them from their neighbors, and that the Iroquois elm-bark canoes were very coarsely built of a single large sheet of bark, crimped along the gunwales, with the ends secured between battens of split saplings. He noticed that the gunwales, ribs, and thwarts were of "tree branches," implying that the bark was not removed from them. The most detailed description was by a Swedish traveller, Professor Pher Kalm, who gave extensive information on the construction of an elm-bark canoe in 1749; this account is particularly useful when interpreted in relation to the spruce-and elm-bark canoes of the eastern Indians. It is upon the basis of Kalm's account that the procedures used to build an Iroquois war canoe have been reconstructed. The bark most favored by the Iroquois was that of the white elm. Next most favored was red elm, and then other barks--certain of the hickories and chestnut are mentioned in various early references. It was necessary to find a tree of sufficient girth and height to the first limbs to give a sound and fairly smooth bark sheet in the length and breadth required. If possible the bark was stripped from the standing tree; even after steel tools were available, felling was avoided for fear of harming the bark. Great care had to be taken in the operation, to avoid splitting or making holes in the bark, and often two or more trees had to be stripped before a good sheet of bark was obtained. In warm weather the bark could be removed without much difficulty, but in the spring and fall it might be necessary to apply heat; this was apparently done by means of torches or by the application of hot water to the tree trunk. When the bark was removed from the tree, the rough outer bark was scraped away; if the builder was hurried this scraping was confined to the areas to be sewn or folded. The bark was then laid on a cleared piece of ground, the building bed, with the outside of the bark up, so that it would be inside the finished boat. The building bed does not appear to have required much preparation; apparently not raised at midlength, it was merely a plot of reasonably smooth ground, located in the shade of a large tree if building was to be done in summer. It is not wholly clear from the descriptions whether the gunwales were shaped before or after being secured to the bark. However, extensive experiments in building model canoes show very plainly that it would be easiest to assemble the main gunwale frame and use it in building, after the fashion of eastern birch-bark canoe construction. With the main gunwales assembled, the stakes would be placed on the bed, the bark replaced, the frame laid on it and weighted, and the stakes then redriven in the usual way and their heads lashed together in pairs.