'One of the grand points in controversy (a controversy indeed chiefly carried on between practice and speculation) is that of _deep ploughing_. In your last volume you seem, on the whole, rather against that practice, and have given several reasons for your judgment which deserve to be very well considered. In order to know how we ought to plough, we ought to know what end it is we propose to ourselves in that operation. The first and instrumental end is to divide the soil; the last and ultimate end, so far as regards the plants, is to facilitate the pushing of the blade upwards, and the shooting of the roots in all the inferior directions. There is further proposed a more ready admission of external influences--the rain, the sun, the air, charged with all those heterogeneous contents, some, possibly all, of which are necessary for the nourishment of the plants. By ploughing deep you answer these ends in a greater mass of the soil. This would seem in favour of deep ploughing as nothing else than accomplishing, in a more perfect manner, those very ends for which you are induced to plough at all. But doubts here arise, only to be solved by experiment. First, is it quite certain that it is good for the ear and grain of farinaceous plants that their roots should spread and descend into the ground to the greatest possible distances and depths? Is there not some limit in this? We know that in timber, what makes one part flourish does not equally conduce to the benefit of all; and that which may be beneficial to the wood, does not equally contribute to the quantity and goodness of the fruit; and, _vice versa_, that what increases the fruit largely is often far from serviceable to the tree. Secondly, is that looseness to great depths, supposing it is useful to one of the species of plants, equally useful to all? Thirdly, though the external influences--the rain, the sun, the air--act undoubtedly a part, and a large part, in vegetation, does it follow that they are equally salutary in any quantities, at any depths? Or that, though it may be useful to diffuse one of these agents as extensively as may be in the earth, that therefore it will be equally useful to render the earth in the same degree pervious to all? It is a dangerous way of reasoning in physics, as well as morals, to conclude, because a given proportion of anything is advantageous, that the double will be quite as good, or that it will be good at all. Neither in the one nor the other is it always true that two and two make four.'