How one may be deceived in evolving a chain of facts is evident from much that we read in scientific journals.

Here is an illustration. At Kinnened in Sweden a boat was recently dug out of a bog. It was a "dug-out," and six feet across the edges. From this the inference is drawn that the trees of that region "must have been 20 feet in circumference".

In Alaska the captain of our vessel gave us "half a day "to follow the receding Davidson glacier. The rest of the party took a roundabout but water-worn way, but the writer of this thought he could save time by crossing the narrow but wood-covered peninsula that separated the glacier's track from our landing place. It was a mistake. The immense thickets of Shallon three or four feet high, and the terribly prickly Panax horrida, or "Devil's walking cane," took several hours to get through. It was worse than an Eastern blackberry thicket. The glacier's trail was reached, but that was all. The others got to the glacier's edge, but the unfortunate battler with woods had to stand the laugh of the more successful ones. But he had his reward. In his tramp through the woods he found a "boat yard." The Indians had all left to see the big ship and "trade," but there was their work in progress. The trees were of alder, which grow to an immense size for alders on the bottoms. The logs in preparation were not measured, but from memory were between two and three feet. But when finished were much wider than this. They were stretched apart. As was gathered from a half-breed afterwards, who spoke fair English, the logs after being hollowed, are filled with water.

Red hot stones are then put in the water till the temperature is high. Then braces are wedged across, much as the butcher uses a gambrel to spread a sheep apart, and the boat remains in that form when it cools. Therefore the width of an Indian dug-out is no guide to the width of the tree that furnished the log.