(679). The entire alimentary system must next be taken away, and with it necessarily a considerable portion of the reproductive network. A view will thus be obtained of the attached ends or roots of the branched segmental organs. These roots will be found to equal the alimentary caeca in number, and therefore that of the feet which are situated posteriorly to the proboscidiform oesophagus. They appear under the character of pyriform tubuli, commencing or ending in a single external orifice. Internally they are lined with ciliated epithelium, the cilia being large, dense, and acting with great force and vigour. The current raised by these cilia sets up on one side and down on the other. The ciliated epithelium ceases at the point where the primary branches divide. All the rest of the organ is unciliated, and filled with the reproductive products. This portion is elaborately branched, the branches twining round the diverticula of the stomach. No microscopic object can be more beautiful than a portion of this tubular network.

The individual tubes are bridled on one side and glandular on the other.

A similar structure is exhibited by the male tubes*.

(680). Tubicola

The two preceding orders of Annelidans are erratic: but in the third we find creatures inhabiting a fixed and nermanent residence that encloses and defends them. This is generally an elongated tube, varying in texture in different species. Sometimes it is formed by agglutinating foreign substances, such as grains of sand, small shells, or fragments of various materials, by means of a secretion that exudes from the surface of the body and hardens into a tough membranous substance; such is the case of Terebella Medusa (fig. 132.) In other cases, as in the Serpula contor-tuplicata (fig. 131), the tube is homogeneous in its texture, formed of calcareous matter resembling the shells of certain bivalve mollusca, and apparently secreted in a similar manner. These tubes are generally found incrusting the surface of stones or other bodies that have been immersed for any length of time at the bottom of the sea; they are closed at one end, and from the opposite extremity the head of the worm is occasionally protruded in search of nourishment. It must be evident that, in animals thus encased, the character of the respiratory apparatus must be considerably modified*; instead, therefore, of the numerous branchiae appended to vicinity of the neck, where they form fanlike expansions, or ramified tufts, so arranged as to be most freely exposed to the surrounding medium. The mouth, placed at the origin of the tentacular cirri, is a simple orifice closed with a valve-like flap or upper lip, but is unprovided with any dental structure. The alimentary canal is generally a simple and somewhat capacious tube that traverses the axis of the body; but in some species, as in Sabella pavonina, it assumes a spiral course, making close turns upon itself from the mouth to the anal aperture, which is always terminal.

* Dr. Williams, Phil. Trans. 1858, p. 134.

Serpula contortuplicata.

Fig. 131. Serpula contortuplicata.

* M. de Quatrefages gives the following resume of the various modifications met with in the respiratory apparatus of the Annelidans: -

"1. Respiration general and entirely cutaneous (Lumbriconereis).

"2. Respiration cutaneous, but confined to certain segments (Choetopterus).

"3. Respiration cutaneous, but confined to certain points of each segment (Nereis).

"4. Respiratory organs taking the form of a simple caecum or bladder (Glycera).

"5. The branchiae characterized more and more by the formation of a canal in connexion with larger or smaller lacunae.

"6. These branchiae may be distributed all along the body (Eunice sanguinea).

"7. They may be confined to a certain number of segments placed towards the middle of the body (Arenicola, Hermella).

"8. They may all be placed at the extremity of the body, so as to form a double tuft".

"The breathing is accomplished," says Dr. Williams, " in every species, the Earthworm not excepted, in strict conformity with the aquatic principle. No known Annelid respires on the atmospheric model. In every Annelid the blood, though variable the segments of the body which we have found in the Dorsibranchiate order, the respiratory tufts are all attached to the anterior extremity of the creature, where they form most elegant arborescent appendages, generally tinted with brilliant colours, and exhibiting, when expanded, a spectacle of great beauty. In some species, as in that represented in fig. 131, there is a remarkable provision made for closing the entrance of the tube when the animal retires within its cavity. On each side of the mouth is a fleshy filament resembling a tentacle; but one of these (sometimes the right and sometimes the left) is found to be considerably prolonged, and expanded into a funnel-shaped operculum that accurately fits the orifice of the shell, and thus forms a kind of door, well adapted to prevent intrusion or annoyance from external enemies.

(681). The curious habitation of the Terebella Medusa is constructed by cementing together minute shells and other small bodies (fig. 132.) In neither case is there any muscular connexion between the worm and its abode; so that the creature can be readily drawn out from its residence in order to examine the external appendages belonging to the individual segments of its body. When thus displayed (fig. 133), the modifications conspicuous in the structure of the lateral oars are at once seen to be in relation with their circumscribed movements, and offer a wide contrast to the largely-developed spines, setae, and tentacular cirri met with in the Dorsibranchiata. In the upper part of the body, rudimentary protractile bunches of hairs are still discernible, but so feebly developed that their use must evidently be restricted to the performance of those motions by which the protrusion of the head is effected; while upon the posterior segments even these are obliterated, the only organs attached to the rings being minute foot-like processes adapted to the same office. The tentacular cirri, which were likewise distributed along the entire length of the Dorsibranchiate order, are here transferred to the head, where they form long and delicate instruments of touch, and, most probably, assist materially in distinguishing and seizing prey. The branchiae, likewise, are no longer met with upon the segments enclosed within the tegumentary tube, but are placed only in the immediate in colour, is non-corpusculated. The converse is true of the chylaqueous fluid, which in every instance abounds in regularly and determinately organized floating cells." "In the Annelida the function of respiration is discharged under two remarkably distinct conditions.