SEGMENTAL INDIVIDUALS 119 



being during development. Also that this entailed the 

 development of the canal system as a closed circulatory 

 system. The causes of mouth-formation are dealt with in 

 another chapter. 



Naturally, in order to drive its contents along, such a 

 closed circulatory system would need something new in 

 place of ancestral bell-contractions, or in addition to the 

 muscular segmental contractions representing these ; and 

 it may well be that the closure itself called forth the driving 

 mechanism. For the recurring accumulation of waste- 

 products within the closed system would stimulate contrac- 

 tion in the vessel-walls, and cause the development of the 

 regularly contracting hearts of the worm. In harmony with 

 our theory, the dorsal and ventral vessels would represent 

 the linking up of ancestral radial canals in straight lines, 

 during development. 



It would seem that the primitive ancestor of the worm 

 was not called on to develop gills as was that of the fish, 

 hence the probable derivation of the hearts is very apparent. 

 In the fish, however, as we shall see, we have the medusoid 

 circular canals represented in the branchial arches, but 

 owing to gill formation they do not surround the digestive 

 tube as they do in the worm. 



The nervous system of the earthworm can be interpreted 

 in a similar way to the circulatory ; that is, its ground plan 

 is the same as if the circular nerve-cords in the rims of 

 successive medusoid bells in series had their ganglia united 

 serially to form ganglionic chains. In the earthworm, 

 however, only one longitudinal chain has developed, but 

 in compensation there is central nervous control. This is 

 exercised by two large ganglionic masses united together 

 above the pharynx and situated on the well-developed 

 circum-pharyngeal collar mentioned and figured on page 115. 

 It is in this nerve-collar that one can recognise the medusoid 

 nerve-ring circling round the rim of the bell. In the medusoid 

 the ring bears ganglia at regular intervals, and in the worm's 

 supra-pharyngeal ganglia we seem to have these massed 

 together at one definite spot. In medusoids the nerve- 

 ring surrounds, at a distance, the manubrium ; in echino- 

 derms, which would seem to be of medusoid ccelenterate 

 derivation, the nerve-ring surrounds the gullet also, though 



