Roots are of course intended to absorb or suck in water. If there is a drain in the soil or a place where water collects, the roots will grow towards that place. Very often they form a dense spongy mass of fibres which may almost choke the drain. Along a riverside one can often find great fibrous masses of tree roots near the water. But how does the root learn that the water is there and turn away from its original track to find it? It certainly does so! Then again, Herr Lilienfeld has recently shown that roots seem able to turn away from poisonous materials in the soil and to seek out and grow towards valuable and nutritious substances. He found that peas, beans, sunflower, and other roots were very sensitive to different substances in the soil, and were directly attracted by what was good for them and turned aside from what was unwholesome. This property and the power of growing towards water probably explain the mysterious sense of direction alluded to above, for roots will take a line which has not been exhausted by their neighbours.[36] [36] Lilienfeld, _Beihefte z. Botan. Centralblatt_, Band XIV., abth 1, pp. 131-212. The facts were denied by Newcombe and Rhodes, _Bot. Gazette_, 36, 1904. But of all these wonderful properties, the most remarkable is the way in which roots find their way past stones and other obstacles in the soil. They insinuate themselves into winding cracks and crawl round stones with an ingenuity that makes one wonder if they can possibly be without some sort of intelligence. It is the very tip or end of the young root that seems to be responsible; for if, in the course of its journeyings underground, it should strike a stone or something hard, the root does not grow on and flatten itself. But some sort of message is sent back from the tip to the growing part which is a short distance behind it. After this message has been received, the growing part begins to curve sideways, so that the tip is brought clear of the obstacle and can probably proceed triumphantly upon its way. The inexplicable part is that the growing part which curves has never been touched at all, but simply answers to the message from the tip.[37] [37] If the growing part itself touches a stone it curves round the stone, not away from it--the reverse of the reaction at the tip! This is perhaps the most reasonable and intelligent behaviour found in the whole vegetable world, and it is not surprising that Darwin compared the root-tip to a brain. These extraordinary responses fill one with astonishment, but there are others still more interesting and remarkable. It will be remembered that we have already shown how different the soil is at different levels. The subsoil, soil, and uppermost layers are all quite different from one another.