COMPAEATIVE MOEPHOLOGY OF THE GALEODID.^. 369' 



On reaching the median ventral blood-sinus, instead of flowing u]i between the 

 original inter-diverticular blood-passages, it was now required to flow through the 

 maze of branched tubules. Such a change as this necessarily affected the respiratory 

 invaginations, which originally j)rojected into these inter-diverticular blood-passages 

 (PI. XXXIII. tig. 12). The dispersal of these regular streams into a (Uffuse streaming 

 among the tubules no doubt played some part in causing the respiratory invaginations 

 to develop into long tubules branching freely among the tissues. The same principle 

 applies to the cephalothorax, where the blood-streams are much divided by the tissues 

 and had therefore to be sought out by the respiratory invaginations. 



In Scorpio, though a certain number (5) of the primary abdominal diverticula persist, they are no 

 longer simple, but consist of a mass of branching tubules. This change necessarily affected both the 

 circulatory and the respiratory systems, in a manner, however, strikingly different from what we have 

 described in Gakodes, although both are but modifications of the same original system. 



The blood is propelled forwards into a series of membranous sinuses closely surrounding and pene- 

 trating the brain. On leaving the brain anteriorly, it is conducted along apparently membranous 

 channels accompanying the principal nerves, both the nerves to the limbs and the long ventral ganglionic 

 chord. These channels are not ordinary blood-vessels, but appear to be rather guiding-membranes ; 

 that which accouijianies the ventral chord, however, presents remarkable histological features which show 

 it to be an organ of some unknown physiological significance which deserves investigation. This 

 remarkable channel runs dorsally to the nerve-chord, here and there sending a branch between the 

 chords, which opens into the ventral lacunar system. The physiological connection between this vessel 

 and the nerve-chords is seen at the ganglia. Over these latter the vessel spreads out and sends branches 

 mto their interior. The vessel then runs on above the chords to the next ganglion. The object of this 

 and of the other vessels accompanying nerves is apparently to pi'ovide them and the ganglia with a stream 

 of oxygenated blood. In Gakodes we find the nerves invariably accompanied by large trachere. 



From this system of nerve blood-channels, the blood escapes into the general lacunar system of the 

 body, either directly or, as in the case of that accompanying the nerve to the third leg, after first passing 

 through the coils of the coxal glands (12), As in Galeodes, the blood passes through the neural arch 

 of the diaphragm and then runs along a median ventral blood-sinus (PL XXXIV. fig. 9, vs). 



From this sinus it no doubt originally ran up between the primary segmental diverticula along the 

 dorso-veutral muscles (PL XXXIII. fig. 12) to the heart, being aerated by the respiratory invaginations. 

 The present segmental divisions of the abdominal alimentary system and the persistence of the veno- 

 pericardial strands (PI. XXXIV. figs. 1 and 9, on the left) show clearly that this primitive arrangement 

 persisted long enough to allow the respiratory invaginations to become highly speciaHzed. With the 

 increasing specialization of the distensible abdominal alimentary system, the inter-diverticular passages 

 became squeezed up and the blood hud to flow out laterally from the median sinus, passing across and 

 not along the respiratory invaginations, on their way. In adaptation to this change, the respiratory 

 invaginations developed in process of time a system of laminated air-chambers (the '' lung-books "). 

 Between these laminje (PL XXXi\. fig. 9, right) the blood, on its way up to the pericardium, would in 

 future have to pass. The inter-diverticular blood-passages gradually closed up, and the infolded 

 membrane degenerated into the strands above described, the veno-pericardial muscles (PL XXXIV. 

 fig. 9, left) . 



This system, however, would no longer suffice when the primary diverticula broke up into branched 



tubules, a change which in Scurpio proijably took place much more recently than in Galeodes. Each of 



these alimentary tubules must be supplied with blood. It is obvious that they cannot receive it, as they 



do in Galeodes, from the stream flowing along the ventral median sinus, for in Scorpio all this blood is 



SECOND SEllIES. — ZOOLOGY, VOL. VI. 49 



