402 ' DR ARTHUR T. MASTERMAN ON THE 



is no torsion. I have here used the terms radius and inter-radius to make the process 

 more clear ; the more accurate expression in speaking of stage D is to describe the 

 stone-canal as lying some little distance along the dorsal surface, but its position can 

 be better identified as in an imaginary 4/5 inter-radius. From the above it will be 

 seen that if we place a starfish (e.g., Cribrella) on its oral surface, the larval sagittal 

 plane will be horizontal and the larval right side uppermost ; if we wish to determine 

 the larval antero-posterior axis, we must place the starfish with the madreporic inter- 

 radius away from us, and the opposite arm towards us ; the antero-posterior axis of the 

 larva will then start anteriorly in the inter-radius immediately to the left of the 

 madreporic inter-radius, and pass along the radius opposite it, immediately to the 

 right of the main radius. 



Theoretical Considerations. 



How far does the development of Cribrella justify us in drawing conclusions in 

 regard to questions of phylogeny ? The fact that the ccelom arises by separate anterior 

 and posterior elements places this species in a unique position amongst asterids, and 

 the great length of the development, uninterrupted by larval feeding or the vicissitudes 

 of pelagic life, is favourable to a correct repetition of the phyletic stages. Naturally, the 

 loss of mouth and anus must be regarded as secondary. The objection may be urged 

 that it is rash to attempt phyletic speculations upon an experience of one type only, 

 but as long as one is not led into matters which are specially connected with other 

 groups, this objection does not appear to apply. There must have been an unbroken 

 chain of living organisms from the ccelenterate to the modern starfish. Does Cribrella 

 help us to realise the structure of the early links ? 



The Bilateral Ancestor. 



From the preceding it will not be difficult to sketch the main features of the bilateral 

 ancestor. We are certainly justified in regarding as precocious and ccenogenetic 

 phenomena the enlargement of the organs of the left side, and in correcting this 

 asymmetry we naturally bring the central ccelom back to the median line and the 

 pore-canal to the left. Again, we regard the absence of mouth or anus and the 

 persistent continuity of the various elements of the ccelom with the mesenteron as 

 purely embryonic characters. We are then left with a bilateral organism with a large 

 pre- oral lobe, with a simple alimentary canal, a ventral mouth below the pre-oral lobe, 

 and a posterior anus. The ccelom was segmented into a pre-oral unpaired part, a small 

 central vesicle over the oesophagus, a pair of lateral cceloms and a pair of posterior 

 cceloms. Locomotion was probably effected by cilia, in all probability arranged in 

 definite tracts in relation to the mouth. I think there are few who will not allow this 

 much as a legitimate inference from the facts. When we commence to inquire whether 

 the lateral cavities possessed tentacles we find all conjecture. If they did, it is obvious 

 that the right lateral cavity has lost them. The only evidence in favour of this view 





