My own experience as to roots, in connection with deep pipe draining, is as follows:--I have never known roots to obstruct a pipe through which there was not a perennial stream. The flow of water in Summer and early Autumn appears to furnish the attraction. I have never discovered that the roots of any esculent vegetable have obstructed a pipe. The trees which, by my own personal observation, I have found to be most dangerous, have been red willow, black Italian poplar, alder, ash, and broad-leaved elm. I have many alders in close contiguity with important drains; and, though I have never convicted one, I can not doubt that they are dangerous. Oak, and black and white thorns, I have not detected, nor do I suspect them. The guilty trees have, in every instance, been young and free growing; I have never convicted an adult. Mangold-wurzel, it is said by several writers, will sometimes grow down into tile drains, even to the depth of four feet, and entirely obstruct them; but those are cases of very rare occurrence. In thousands of instances, mangolds have been cultivated on drained land, even where tiles were but 2-1/2 feet deep, without causing any obstruction of the drains. Any reader who is curious in such matters, may find in the appendix to the 10th Vol. of the Journal of the Royal Ag. Soc., a singular instance of obstruction of drains by the roots of the mangold, as well as instances of obstructions by the roots of trees. _Obstruction by Per-oxide of Iron._ In the author's barn-cellar is a watering place, supplied by a half-inch lead pipe, from a spring some eight rods distant. This pipe several times in a year, sometimes once a week, in cold weather, is entirely stopped. The stream of water is never much larger than a lead pencil. We usually start it with a sort of syringe, by forcing into the outlet a quantity of water. It then runs very thick, and of the color of iron rust, sometimes several pails full, and will then run clear for weeks or months, perhaps. In the tub which receives the water, there is always a large deposit of this same colored substance; and along the street near by, where the water oozes out of the bank, there is this same appearance of iron. This deposit is, in common language, called per-oxide of iron, though this term is not, by chemists of the present day, deemed sufficiently accurate, and the word sesqui-oxide is preferred in scientific works. Iron exists in all animal and vegetable matter, and in all soils, to some extent. It exists as protoxide of iron, in which one atom of iron always combines with one atom of oxygen, and it exists as sesqui-oxide of iron, from the Latin _sesqui_, which means one and a half, in which one and a half atoms of oxygen combine with one atom of iron. The less accurate term, per-oxide, has been adopted here, because it is found in general use by writers on drainage.