not be injured by the pressure, whilst those which are soft will be almost reduced to pulp.  Therefore, the rollers of the mill should be always placed horizontally, and upon passing the trees lengthways through the mill, the pressure will be uniform and the fibre uninjured.  In this manner, pass the different sorts of layers separately, and the produce will be about four pounds of fibre from each tree.  The stalks of the branches of the plantain give the best fibre, and a large quantity, as compared with the body of the tree; 100 lbs. of the stalk will give 15 lbs. nett of fibre.  In general, if a tree will give 4 lbs. nett of fibre, the stalks will give 1 lb. out of the 4 lbs.  The stalks ought also to be crushed separately, because they are harder than the exterior layers of the tree.  About 3,000 trees may be passed through the mill in a day.  Whilst the experiments were in progress it was ascertained that with a single horse, 100 plantain trees on an average were crushed in twenty minutes, giving five minutes rest for the horse.

Fermentation.—­This operation may be performed in several ways.  If the trees are allowed to ferment upon the spot after being cut, a great saving will occur in respect of carriage; this matter ought to be carefully studied, because, on an extensive scale of manufacture, it is of serious importance.  It is found that the trees when cut and heaped up, are subject to a drainage of juice, which, having a tanning property, discolors those pieces which lie at the bottom; hence much time is consumed in afterwards restoring the fibre to its natural color.  The cut plants should be removed from the stumps of the trees, and then placed in heaps, shaded from the sun by laying the leaves over them.  They will take several weeks to ferment.  To pursue this process in the immediate vicinity of the establishment, would give rise to many inconveniences, in consequence of the very large space of ground that would thereby be occupied.  Fermentation requires a mean temperature.  A tree cut down and exposed to the sun, would be nearly dry at about 30 deg. centigrade, showing a result quite different to that which ought to be obtained; whilst a tree placed on a wet soil, and open for the fresh air to circulate between the plants, covered at the same time with its own leaves, and shaded by the foliage of the plantation, would be decomposed at the desired point of about 22 degrees.  The different modes of fermentation require the same proportions.  If the cut plants be covered with a thick layer of earth, they will not decompose in six months; but if, on the contrary, they are covered slightly, so that they may receive the freshness of the earth, and the heat of the air, they will decompose in six weeks.  It is the same with the fermentation of alkaline baths.  Baths at only one degree will produce decomposition, whilst baths at three degrees will not produce any decomposition.