One of the most careful investigators into this subject was the accomplished geologist Mr. J. W. Foster, whom I met at Chicago, and who was at that time President of the American Association for the Advancement of Science. He told me what had been the results of his work, and accompanied me to parts of the outlying Illinois prairies, to examine the character of the earth near the surface. He had, a short time previously, carried out a series of experiments relating to the physical geography of the Mississippi valley, and had given his attention to the composition of the upper formation of the prairies on both sides of the river. He had arrived at the conclusion that the absence of trees upon them was not caused by fires or by the character of the ground, but was the consequence of the conditions of rainfall, temperature, climate, and exposure, all acting in a direction opposed to that natural order of things in which trees would be inclined to flourish. On the way south from Mankato, I observed that the banks of the Des Moines river were thickly timbered, and that on the borders of the various rivulets which had cut their way through the ground to a considerable depth, there was usually a flourishing growth of trees, chiefly consisting of oak, black walnut, basswood, and maple varieties. In all these instances the trees were sheltered, but on the exposed prairie immediately adjacent the ground was bare, and without a vestige of any signs of trees or shrubs. In that particular region it seems therefore probable, that the exposure to gales of wind sweeping over the plateau, may be the principal cause of trees not being able to take root and live. There are, however, other prairies equally bare of timber which are not so exposed. Upon those situated near the water shed or dividing ridge of the tributaries of the Missouri and Mississippi, it may be the quantities of water lying permanently within a few feet below the surface which prevent any growth taking place. No single theory seems to explain the facts, but from my own observation, I am inclined to think that the chief deterring influence is the nature of the soil.[40] There is an unexplained problem respecting the growth of trees in certain parts of North America, which has received much local notice. My attention was directed to the subject when I was in the neighbourhood of Lake Simcoe in Upper Canada. An Englishman, who had established a homestead there and made clearings in the forests, told me he had observed that after the old trees had been felled, new shrubs and timber of an entirely different character grew up in their place. There was nothing in the nature of the surrounding woods which, to his mind, could account for the change, as there were no trees of the class that had taken root existing in that part of the country.