or ill founded, that it is liable to rot linen, and the preference is given here to the starch of arrowroot.  It remains to be seen, however, what estimate will be formed of this starch in England, for if it should prove an esteemed variety, there can be no doubt of its proving a highly profitable cultivation.  Cassava grows readily in almost any soil, and when the drainage is tolerable, two crops of the sweet variety can, it is stated be grown in a year.  I have seen it growing luxuriantly in the light soils of the interior, as well as in the stiff clay soils of the coasts.  It is considered an excellent preparatory crop in new and stiff land, on account of its tendency to loosen the soil.  Were the bitter variety fixed on, the preparation of Casareep might be combined with the preparation of starch; and as that substance is one of the most esteemed bases for the preparation of various sauces, it is probable that this might turn out the most profitable part of the produce.  At all events, bitter cassava would have this advantage over all other starch-producing roots, that the juice of the roots could be turned, to account as well as the starch.

Of all the plants mentioned in the list, starch is most readily separated from the arrowroot, in consequence of the tissue being more fibrous, and yielding little or no cellular tissue requiring to be run off the starch.  Time and water are thus saved in the process, and were the fibrous residue pressed and dried, it could probably be turned to good account in the manufacture of paper.

In respect of facility of preparation, the plantain starch, though of excellent quality, ranks lowest, for the flesh-colored tissue in which the starch is embedded is somewhat denser than the starch, and settles down under it, and it is not a little difficult to arrange the process so as completely to separate the finer parts of this matter from the starch, and hence its color is never perfectly white.

Yield of starch-producing plants per acre.—­On this subject, as already remarked, I do not at present possess sufficiently accurate data.

In England ten tons of potatoes are not unfrequently produced per acre; now assuming 15 the per centage of starch, there would be a yield of one-and-a-half tons per acre, which, at the-lowest quotation, 28s. a cwt., would give L42 per acre; and were the starch to rank with that prepared from wheat, it would produce L40 per ton, or L60 per acre.  In the thorough drained land of Demerara, and under a good system of cultivation, I have no doubt that ten tons of cassava could easily be grown, and if it yielded 25 per cent. of starch, it would be a return of 21/2 tons, or of L62 10s. per acre, reckoned at the price of potato starch.