TYPE EOHINODERMA. 591 



in an Annelid the most anterior metamere is that which is apt to first come 

 into contact with new conditions of environment, and consequently it has 

 become specially provided with sense-organs for the perception of these 

 new conditions whether favorable or unfavorable; similarly the conditions 

 affecting the dorsal and ventral surfaces are different and consequently a 

 dorso-ventral differentiation exists; while on the other hand the conditions 

 affecting the two sides of the body are apt to be alike, and consequently 

 the differentiations which occur on each side of the median line are usually 

 alike. Bilaterality is then associated with a free-mode of life. 



Conversely, radiality is usually associated with indefinite axes of pro- 

 gression or with a fixed mode of life, and the idea suggests itself that the 

 radiality of the Eehinoderms may be the result of a fixation of the bilateral 

 larva. The majority of recent Echinoderms are, it is true, free forms, the 

 Crinoids alone being sessile, but it will be found that geologically the free 

 forms are the latest to appear, and that the Pelmatozoa are especially char- 

 acteristic of the Palaeozoic rocks. This would imply that the Crinoids are 

 to be regarded as the most perfect representatives among recent forms of 

 the ancestral types, an idea which is borne out by certain points in Echino- 

 derm development. Thus the calyx of the Crinoid is developed in the 

 posterior portion of the oval larva, the anterior portion being occupied by 

 the stalk; in the Brachiolaria and Pluteus larvsB the adult body is developed 

 in the posterior portion of the larva, the anterior portion undergoing 

 absorption, an arrangement which may be explained by supposing that 

 originally the adult forms which possess these larvae were stalked, but 

 secondarily acquired a free mode of life, and that as a result the stalk dis- 

 appeared, the body, corresponding to the calyx of the Crinoid, still develop- 

 ing in its accustomed position in the posterior portion of the larva. In 

 the Holothurians there is not apparently that distinction into anterior and 

 posterior regions in the larva which obtains in other forms, but it is to be 

 noted that in the young Cucumaria, for example, there is a well-marked 

 prsBoral lobe which later on disappears and may represent the partially- 

 aborted stalk region of other forms. 



Attempts have beeu made to trace out a phylogeny taking the Holo- 

 thurians as a primitive type and tracing them back to such forms as the 

 Gephyrean worms, the presence of respiratory trees in both forms suggest- 

 ing a possible affinity. Or again the later barrel-shaped larva with five 

 tentacles has been considered to represent a common ancestor for all the 

 various groups of Echinoderms, the term Pentactcea being given to the 

 hypothetical ancestor. Neither of these views, however, affords any clue 

 to the origin of the radiality, and the structure of the Holothurians departs 

 so widely from that of the other groups that it seems preferable to consider 

 them as most remote from the common ancestor. Furthermore as regards 

 the Pentactcea it is diflScult to understand how there should be so much 

 similarity in structure of the different forms if they all differentiated in- 

 dependently from this common ancestor. 



From the evidence at our disposal it seems far more logical to accept a 



