If all this has been carefully done, so that the mould in the pot shall not have been shaken free of the rootlets, the seedlings will not even _know_ it has been transplanted. Its growth will not be delayed for a day, instead of two or three months; and by the time the dry season comes, the tap-root will have descended far enough to imbibe moisture. Another plan to effect the same object. Instead of pots, use coarse bamboo open wicker-work baskets. The split bamboo forming the said wicker-work about half an inch wide, the interstices about one quarter of an inch square. Let the diameter of the basket be the same at top and bottom, viz., 9 inches; the depth of the basket 10 inches. When the seedlings in the nursery are large enough to enable you to select a good class of plant, transplant one into each basket previously filled with soil.[30] This being done when the plants are very young, and there being _then_ no difficulty in taking them up with earth attached to their short tap-roots and rootlets, they will scarcely be thrown back at all. Being near water they can also be well tended. Put basket and all into the vacant hole at the beginning of the rains, and fill up as directed for the pots. The interstices will allow the feeding rootlets to pass through, besides the basket rots quickly under ground, so quickly it cannot impede the plant. Seed is not sown at once in the baskets as in the pots, because the baskets would not last so long. Even putting the seedling in it during (say) February, the basket, with the occasional watering necessary, will, more or less, have rotted before it is put into the hole. I have concluded a contract for ten thousand pots and five thousand baskets at half an anna each for both kinds. Two pice, to ensure the filling up of a vacancy, is not a large outlay. Since writing the above I have had experience of both the above plans. The pot system is far the better, and answers very well.[31] I am now trying to improve this still further by making the pots a little larger, and placing a thin inner lining of tin inside each about half an inch from the sides. This space is first filled with sand, then the pot is filled with mould, and the tin pulled out. The same tin will therefore do for any number of pots. The seed is then put in. I think by this plan if, when about to plant, the mould in the pot is well wetted, that it, with the seedling, can be turned out whole in one piece, and then put in the hole _without_ the pot. The same pots would then answer year after year, and the expense would be quite nominal. If well done, the seedling in this, as in the former case, would not even _know_ it had been transplanted.[32] FOOTNOTES: [29] By “best class” I mean the most indigenous class. [30] Mind again this be of the same nature as the garden soil. [31] The baskets are too frail; being often wetted, they fall to pieces before the planting time.