INTRODUCTION. 19 



of preservation of their material, have described these depressions as pores for the 

 passage of ampullse. 



This mistake caused many of the early Asterozoa to appear to be more nearly 

 allied to the true Asteroidea than they really are. 



The form of the primitive ambulacralia is important, as we can readily derive 

 from it the more advanced ambulacralia found in the more specialised groups. 



(b) As to the Disjwsition of the Primitive Amhulacralia. — We have seen that both 

 in Recent Asteroidea and in Recent Ophiuroidea the ambulacralia are in opposite 

 pairs. Many observers have noticed that in the older forms the ambulacralia are 

 not exactly opposite but alternating with each other. Bronn (14), in one of the 

 earlier classifications of the Asteroidea, used this characteristic to separate the 

 Encrinasterige or Asteroidea with alternating ambulacralia from the Euasterige or 

 Asteroidea with opposite ambulacra. Stiirtz (76-80) made a group of the Ophio- 

 Encrinasterige for Palaeozoic " Ophiuroidea " which had their vertebrae in unfused 

 alternating halves (see also p. 43). 



Gregory (27) and Jaekel (34) regard this alternating condition as the primitive 

 one. 



Bather (3) also showed that Edrioaster had alternating flooring plates [ambu- 

 lacralia], and that the Edrioasteroidea are alone among known Pelmatozoa in pre- 

 senting a type of ambulacrum from which the Stellerid type could readily be 

 derived. 



The earliest strata from which Asterozoa are obtained, the Ordovician of 

 England and America, contain forms which have their ambulacralia exactly oppo- 

 site and forms with alternating ambulacralia. I intend to follow the views of 

 Gregory, for it appears to me that the most primitive forms had ambulacralia 

 which are irregularly alternating. From this primitive stock two sets of forms 

 arose, namely, those with opposite ambulacralia and those with definitely alter- 

 nating ambulacralia. This latter series in Ordovician times Avas almost entirely 

 confined to Ophiuroid-like wriggling forms, as, e. g., species of Aspidosoma, Goldfuss, 

 Protaster, Forbes, and Tmiiiaster, Billings. Later, the conditions of existence led to 

 the extinction of the families with alternating ambulacralia and the survival alone 

 of families with opposite ambulacralia. This latter arrangement appears to have 

 had the greater mechanical efficiency. 



(c) As to the Origin, of the Ophiuroidea. — Many previous observers, Stiirtz 

 (76-80), Gregory (26), Jaekel (34), and Sollas and Sollas (71), have shown that 

 certain Palaeozoic Ophiuroidea present primitive features which indicate the 

 relationship of the class to the Asteroidea. Thus, in these early forms (1) the com- 

 ponent halves of the vertebras are not fused, (2) the ventral shields are not present 

 but the ambulacral groove is directly open to the exterior. In consequence it has 

 been customary to derive the Ophiuroidea directly from the Asteroidea. 



In this monograph I have adopted the view that there is a class which is more 



