Assuming then, he tells us, that the small quantity of potash required could be supplied by the soil, and that the pulp is returned to it, the loss by the crops could be fully supplied by 100 lbs. of castor cake and 10 lbs. of bones per acre.  Then if we require much more from the plant than the production of crop (for we expect it, in addition, to grow wood for the succeeding crop, and during this process the plant grows much superfluous wood, besides suckers, which have to be removed), it must be remembered that all primings and superfluous wood are left on the land.  What there is actually carried off it is really very small in quantity.  Why, then, it will naturally be asked, is it necessary that so much manure should be present in the soil if we wish to grow good coffee and have continuously good crops, and why is it that if manuring is neglected you will soon find that it is only the rich hollows that are able to maintain the coffee in good condition and produce good crops continuously?  To such questions no distinct answer can be given, and we can only conjecture that coffee, when it wants its food, must, for some unknown reason, have a considerable supply at hand.  There is, however, one test which, I think, always shows conclusively whether this food is present in the quantity required to supply the needs of the plant.  Just before the hot weather the coffee trees throw out a small flush of young wood.  Now if the trees have given a fair average crop, and at the same time have a good show of bearing wood for the next season’s crop, and are also throwing out a good supply of vigorous young shoots, then you may be sure that your land is well fed.  But if the trees throw out no young shoots at that time, or very few, then you will know that your land is not as well fed as it ought to be.

It might naturally be supposed that I could furnish some guide to the planter, from our experience in Mysore, as to the quantity of manure that should be put down, but I regret to say that I am unable to do so, as I know of no estate where a regular and continuous system of manuring has been carried out.  But in North Coorg, and very close to the Mysore Border, the continuous practice on Mr. Mangles’s Coovercolley Estate of 500 acres gives a fairly approximate idea of what can keep an estate in a well-fed condition.  There the practice has been to put down every third year from 7 to 10 cwt. of bone-meal an acre, and one-third of a bushel of cattle manure, and, besides this, composts of pulp, mixed with top soil and lime.  Now this is the finest estate I ever saw.  The coffee was even and of a beautiful colour, and when I saw it towards the end of 1891 there was a fair crop of coffee on the trees, and an ample supply of young wood for the following crop, and the land was so well fed with nitrogen that an experimental application of nitrate of soda to a part of the land had produced no perceptible effect on the trees.  From what I have previously