THE APODOUS HOLOTHUEIANS 



131 



reous particles, is to be regarded as ancestral. Whether the tentacles of the 

 earliest synaptids were digitate or pinnate cannot be positively determined; 

 the earliest tentacles of the Synaptidas with pinnate tentacles (see Clark, '98(7 ) 

 are simple and then pinnate, and of the Chiridotinag (see Ludwig, '98&) are 

 digitate almost from the start; in Labidoplax, however (see Semon, '88), the 

 tentacles are simple before becoming digitate. On the whole, it seems probable 

 that the earliest synaptids had 10 simple tentacles and a few more or less 

 wheel-shaped deposits. From this form have diverged two quite distinct 

 branches, one losing the wheels and developing anchors, the other developing 

 various forms of wheels. The interrelationships of the genera might then be 

 represented as follows: 



Polyplectana 



Euapta 

 Synaptula 



Leptosynapta 



Anapta 



Polycheira 

 Chiridota 

 Toxodora 

 Achiridota 



Myriotrochus 

 Trochoderma 



Rhabdomolgus 



10-tentacled 



Ancestral Form 



It will appear then that Rhabdomolgus is the nearest living representative 



of the ancestral stock, though not necessarily close to it, while Polyplectana, 



Protankyra, Polycheira, and Acanthotrochus are the most highly specialized 



forms of their respective branches. 



Geographical Distribution. 



In considering the distribution of the foregoing genera and species we may 

 well consider simply the littoral regions proposed by Ortmann {Grundziige tier 

 Marinen TMergeographie, 1869), so few of the Synaptidae are truly abyssal 

 forms and these almost exclusively of the genus Protankyra. We are at once 

 struck by the fact that the Indo-Pacific region is easily the most important 

 geographical area for synaptids, no less than fifteen genera having representa- 

 tives there, and ten of these (six monotypic, however) may be considered 

 as distinctly Indo-Pacific genera. The Atlantic Boreal subregion has repre- 



