although apparently larger in proportion to the size of the body. The rostrum is, however, clearly a new structure. How can we explain its origin, at the anterior edge of an Apus-like head?

In

According to our theory the anterior antennæ once pointed backwards, as do those of Apus. Ceratiocaris, however, we find them almost at the anterior end of the head. It is not difficult to show that this migration would almost necessitate the formation of a rostrum.

One variation on the primitive Apus type would certainly be a species using their antennæ forwards as organs of sense. Just as, in Apus, the eyes travelled forwards, so, in process of time, the antennæ might tend to move forwards, but, by way of protection for these, at first, delicate organs, we may suppose them to have moved forwards in slight grooves on each side of the median line. As they moved forwards they may have become more and more developed, not only as sensory organs, but as appendages, until they projected freely from the front (as typical Crustacean antennæ). The rostrum is the remains of the middle wall between the two grooves. It is clear that such grooves could not exist on the under surface of the head of an Apus without forming a primitive rostrum. According to this view, the rostrum was originally a necessary accompaniment of the migration of the antennæ from the sides of the labrum to the front of the head. The articulation of this rostrum was a secondary acquirement not in itself difficult to imagine.

This view explains the morphological significance of the rostrum, as the protective point for the more or less delicate antennæ, arising, not per se, but as the further development of the tip of the middle piece between the two depressions along which the antennæ travelled forwards.

From all that remains then of these primitive Nebalidæ we see a sufficient resemblance to the Apodidae to form a very striking confirmation of our theory. We see in them true transition forms between Apus and the higher Crustacea; the fossils showing very clearly one of the first steps in this transformation, and one of the most needful for success in the struggle for existence, i.e. the gradual migration of the antennæ to a frontal position near the eyes.

The many points of likeness between Apus and the Macrura will already have struck every reader of the first part of this book. The detailed deduction of Astacus from Apus on the lines here laid down would be a most interesting and profitable study.

Starting, then, from our theory that Apus, owing to its likeness to an Annelid, must be one of the racial forms of the whole group, we have been able, with varying success, to show that all ancient Crustaceans are clearly related to Apus, and that all the chief groups of the modern Crustacca, with the probable exception of some of the Ostracoda, can be more or less clearly deduced from Apus. An attempt to derive the modern forms from the Apodidæ in detail would

be the work of a life-time and would fill many volumes, but we believe we have established our theory beyond question, and have shown for the first time how a natural system of the Crustacea may be built up by taking Apus as the key to the original Crustacean form.

SECTION XVI

PERIPATUS AND THE TRACHEATA

BEFORE closing this essay, in which we have endeavoured to prove that Apus is an almost ideal transition form between the carnivorous Annelids and one large division of the Arthropoda, viz. the Crustacea, it is but fit that we should briefly refer to Peripatus, which is acknowledged to be a transition. form between the Annelids and the other division of the Arthropoda, viz. the Tracheata, in which we include the Myriapoda, Hexapoda, and Arachnida. It cannot but add to the interest of this book if we dwell upon this point for a short time.

The accepted fact that both divisions of the Arthropoda are derived from Chatopods, the chief cause of the transformation being the same in both, viz.: the use of the parapodia as appendages for mastication and locomotion, accounts for the resemblances in the organisations of the Crustacea and Tracheata which have led to their being placed side by side as Arthropods. There are, however, striking differences in their

morphology which stand obstinately in the way of attempts to establish a close relation between them.

Has not our derivation of Apus and the Crustacea from a bent Annelid supplied us with the clue as to the essential morphological difference between the Crustacea and the Tracheata, leaving out of sight for the moment the trachea and the Malpighian tubules which are confined to the latter?

The Annelid which gave rise to the Tracheata started, as did the Crustacean-Annelid, by using its anterior parapodia as mouth parts, but, unlike the latter, it did not bend round its anterior segments to browse in the manner described in the opening sentences of this essay, but remained straight. The fusion of segments to form the head was, in the TracheatanAnnelid, axial, the mouth remaining at the anterior end of the body.

In such an axial fusing there is nothing to fix the number of segments to form a head common to all the Tracheata, whereas in the Crustacea the bending round of the five segments marked off this region of the body as the head for all time.

The difference between the number of the cephalic appendages of the Crustacea and the Tracheata is to be referred to the fact that with the mouth at the anterior end of the Annelidan body it was impossible to bring so many pairs of parapodia into the region of the mouth to function as mouth parts as in the Crustacea, where its ventral position allows of the arranging on each side of a large number of parapodia as jaws.