From what I have said as to the composition of castor cake, it is probable that white castor contains from 4 to 5 per cent. of phosphate of lime, and I desire to call particular attention to this, because oil-cake is usually regarded purely as a nitrogenous source of manure, whereas one of the oil-cakes commonly used—­i.e., castor cake—­contains an appreciable quantity of that phosphate of lime of which bones are generally considered to be the sole suppliers by the planter.  But it is evident that if we annually used 300 lbs. per acre of white castor, we should, even if it contained only 4 per cent. of phosphate of lime, be supplying six times the amount of lime and more than three times the amount of phosphoric acid removed by an average crop of coffee, and though the lime is liable to loss from waste, it must be remembered that the phosphoric acid is firmly retained by the soil.  It is important to remember that castor cake should, like bones, be mixed with a considerable quantity of fine top soil, so that the manure may be widely distributed through the soil.

Nitrate of potash, or saltpetre, is an extremely expensive manure, and not a desirable one, because the nitrogen in it is in a too quickly assimilable form, and is very liable to be lost in drainage.  But it might be used with effect, and in small quantities, for bringing forward supplies, and I am informed that for this purpose it has been used with advantage in Coorg.  I have used the nitrate of potash on my property—­an experimental amount only—­and it caused the trees to throw out strong and numerous shoots.  It should be bought in the form of pure nitre.

Nitrate of soda is also liable to the objection that the nitrogen in it is in a too quickly available form, and liable to be lost.  I have never used it on my property, but from observing its effect on an estate in Coorg, and the effect it had in causing the trees to throw out a fine supply of young wood, can see that it might be used with great effect in rapidly forcing forward worn-out coffee growing on an exhausted soil.  But if used for this purpose it should be backed up with a liberal supply of bones and castor cake, or of bones and farmyard manure, or bones and top soil, as, if not so backed up, the result would be unsatisfactory, if not disastrous, seeing that the nitrate of soda, if applied alone, would cause the plant to wring out everything that was available in the soil.  The application of nitrate of soda on the estate alluded to was at the rate of 2 cwt. an acre, and cost 21 rupees an acre, inclusive of the cost of application.  I saw the estate at the end of October, and the nitrate had been put down in March previous.  The wood it had been the means of producing was very good and strong, dark green, and abundant, and the effect of the nitrate was by no means confined to one season, for the effect of the nitrate put down the year previous was still apparent.  The land here evidently was short of nitrogen, and