"The coloured granular matter is undoubtedly a product of true secretion; and the cells in which it is found must be regarded as true secreting cells. These cells are themselves frequently to be seen as secondary cells in the interior of parent cells, from which they escape by rupture, and then, falling into the somatic fluid, are carried along by its currents, until, ultimately, by their own rupture, they discharge into it their contents.

"We have no facts which enable us to form a decided opinion as to the purpose served by this secretion. Its being always more or less deeply coloured, and the fact of its being abundantly produced in the digestive cavity, might suggest that it represented the biliary secretion of higher animals. This may be its true nature, but as yet we can assert nothing approaching to certainty on the subject; indeed, considering how widely the cells destined for the secretion of coloured granules are distributed over the walls of the somatic cavity, it would seem not improbable that the import of the coloured matter may be different in different situations; that while some of it may be a product destined for some further use in the hydroid, more of it may be simply excretive, taking no further part in the vital phenomena, and intended solely for elimination from the system." [24]

Here we have very definite statements by a highly trained observer of the distribution of colour in the whole of these animals, and of the conclusions he draws from them.

Firstly as to the colour itself. We find it true colour—brown, pink, carmine, vermilion, orange, lemon-yellow, and even emerald green; a set of hues as vivid as any to be found in the animal kingdom. It is difficult to conceive these granules to be merely excrementitious matter; for in such simple creatures, feeding upon such similar bodies, one would hardly expect the excretive matter to be so diversified in tint. Moreover, excrementitious matter is not, as a rule, highly coloured, but brown. Thus, we see in the Rhizopods the green vegetable matter which has been taken in as food becomes brown as the process of assimilation goes on; and, indeed, colour seems almost always to be destroyed by the act of digestion.

Still, it by no means follows that this colour, even if it is produced for the sake of decoration, as we suggest, may not owe its direct origin to the process of digestion. The digestive apparatus is the earliest developed in the animal kingdom, and in these creatures is by far the most important; the cœlenterata being, in fact, little more than living stomachs. If, then, colouration be structural, what is more likely than that the digestive organs should be the seat of decoration in such transparent creatures?

Secondly, as to the distribution of the colour. We find it "frequently forming a continuous layer upon the free surface of" the endoderm, in the "spadix of the sporosac," and in the "bulbous terminations" of the canals, that colour is best developed. In other words, the colour is distributed structurally, and is most strongly marked where the function is most important.

Prof. Allman gives no hint that the colour may be purely decorative, and is naturally perplexed at the display of hues in such vigour; but if this be one of the results of the differentiation of parts, of the specialization of function, then we can, at least, understand why we find such brilliant colour in these creatures, and why it is so distributed.

As an illustration of the Tubularia we have selected Syncoryne pulchella, [Fig. 2, Pl. VI.], and its medusa, [Fig. 1]. The endoderm of the spadix of the hydranths is of a rich orange colour, which becomes paler as it descends towards the less highly organized stem. Medusæ are seen in various stages of development, and one, mature and free, is shown. In these the manubrium, and the bulbous terminations of the canals are also seen to be coloured orange.

In these medusæ we find the first appearance of sensory organs. They consist of pigment-cells enclosed in the ectoderm, or outside covering; and are singular as presenting the first true examples of opaque colouring in the animal kingdom. They are associated with nerve cells attached to a ring of filamentous nerve matter, surrounding the base of the bell. In some important respects the pigment differs from that in other parts of the animal. It is more definite in structure; and the whole ocellus is "aggregation of very minute cells, each filled with a homogeneous coloured matter." [25] These ocelli, and similar sense organs, called lithocysts, are always situated over the bulbous termination of the canals. The pigment is black (as in this case), vermilion, or deep carmine.

Plate VI.