(b) Changes in the Interior of the Bean.—What is the relation between the comparatively simple fermentation of the pulp and the changes in the interior of the bean? This important question has not yet been answered, although a number of attempts have been made.
As far as is known, the living ferments (micro-organisms) do not penetrate the skin of the bean, so that any fermentation which takes place must be promoted by unorganised ferments (or enzymes). Mr. H.C. Brill[2] found raffinase, invertase, casease and protease in the pulp; oxidase, raffinase, casease and emulsinlike enzymes in the fresh bean; and all these six, together with diastase, in the fermented bean. Dr. Fickendey says: "The object of fermentation is, in the main, to kill the germ of the bean in such a manner that the efficiency of the unorganised ferment is in no way impaired."
From my own observations I believe that forastero beans are killed at 47 degrees C. (which is commonly reached when they have been fermenting 60 hours), for a remarkable change takes place at this temperature and time. Whilst the micro-organisms remain outside, the juice of the pulp appears to penetrate not only the skin, but the flesh of the bean, and the brilliant violet in the isolated pigment cells becomes diffused more or less evenly throughout the entire bean, including the "germ." It is certain that the bean absorbs liquid from the outside, for it becomes so plump that its skin is stretched to the utmost. The following changes occur:
(1) Taste. An astringent colourless substance (a tannin or a body possessing many properties of a tannin) changes to a tasteless brown substance. The bean begins to taste less astringent as the "tannin" is destroyed. With white (criollo) beans this change is sufficiently advanced in two days, but with purple (forastero) beans it may take seven days.
(2) Colour. The change in the tannin results in the white (criollo) beans becoming brown and the purple (forastero) beans becoming tinged with brown. The action resembles the browning of a freshly-cut apple, and has been shown to be due to oxygen (activated by an oxidase, a ferment encouraging combination with oxygen) acting on the astringent colourless substance, which, like the photographic developer, pyrogallic acid, becomes brown on oxidation.
(3) Aroma. A notable change is that substances are created within the bean, which on roasting produce the fine aromatic odour characteristic of cocoa and chocolate, and which Messrs. Bainbridge and Davies have shown is due to a trace (0.001 per cent.) of an essential oil over half of which consists of linalool.[3]
(4) Stimulating Effect. It is commonly stated that during fermentation there is generated theobromine, the alkaloid which gives cacao its stimulating properties, but the estimation of theobromine in fermented and unfermented beans does not support this.
(5) Consistency. Fermented beans become crisp on drying. This development may be due to the "tannins" encountering, in their dispersion through the bean, proteins, which are thus converted into bodies which are brittle solids on drying (compare tanning of hides). The "hide" of the bean may be similarly "tanned"—the shell certainly becomes leathery (unless washed)—but a far more probable explanation, in both cases, is that the gummy bodies in bean and shell set hard on drying.
We see, then, that although fermentation was probably originally followed as the best method of getting rid of the pulp, it has other effects which are entirely good. It enables the planter to produce a drier bean, and one which has, when roasted, a finer flavour, colour, and aroma, than the unfermented. Fermentation is generally considered to produce so many desirable results that M. Perrot's suggestion[4] of removing the pulp by treatment with alkali, and thus avoiding fermentation, has not been enthusiastically received.
Beans which have been dried direct and those which have been fermented may be distinguished as follows: