BRITISH I'AL^OZOIC ASTEROZOA. 



but slightly altei'iiating, and in consequence we find that both the flooring and 

 covering plates, which are in direct association with the podia, are not exactl}' 

 opposite, but arranged alternately (see p. 19). 



(v) It is possible that the covering plates were finger-shaped. A comparatively 

 slight opening of the plates of this shape would allow the podia to be easily thrust 

 out of the groove (compare with the shape of adambulacralia in any drawing of the 

 ambulacral groove of an Asteroid, as e.g. Text-fig. 4, p. 12, or Text-fig. 18, p. 20). 



(vi) A terminal unpaired plate at the end of each ray supported a terminal 

 unpaired tube-foot. Later a pigment spot developed at the extremity of the foot, 

 which consequently became an " eye " (compare the embryological researches of 

 Ludwig, 40). 



A creature of this type which left its stalk and acquired a " freely moving " 

 habit, would be a primitive Asterozoon. 



The flooring plates then arch over the ambulacral groove, and are generally 

 known as the ambulacral ossicles or ambulacralia. 



The covering plates are at their base, and are known as adambulacral ossicles 

 or adauihuJacralia. 



The stalk in the adult forms appears to have been resorbed before starfishes are 

 met with as fossils. 



If this account of a primitive Asterozoon be compared with my remarks in a 

 paper published in 1904 (73) it Avill be observed that my views have lieen con- 

 siderably luodified. The paper had as its primary object the furtherance of the 

 view that the Echinoidea were derived from an Asteroid ancestor. This led me to 

 criticise certain of Bather's views, and I stated inter alia (p. 44) that "it is im- 

 possible to suppose that the complicated ambulacrum and fixed skeleton of 

 Edn'odster gave birth to the primitive simple ambulacrum and movable skeleton 

 of the Asteroids." Literally, of course, this statement is correct, for the pores 

 between the flooring plates, structures which Bather compares with the ambulacral 

 pores of Asteroids (9), are not present in the early Asterozoa and must be secondary 

 modifications; again, the complete roofing over of the groove by covering plates 

 cannot be primitive. 



My studies upon fossil Asterozoa, however, now compel me to recognise 

 that in respect to the argument originally advanced I overstated my case. It 

 is now evident that Bather Avas correct when he asserted that the ambulacral 

 groove of the Asteroidea was built fundamentally upon the same plan as that of 

 Eilrioaster. This is seen most clearly when we compare the structures of the 

 hypothetical ancestor with that of EJrioaster. It is not difficult to imagine a more 

 primitive Edrioasteroid which would show, at any rate, near relationships with the 

 ancestral Eleutherozoa. 



