86 MOLLUSCOIDA. 



a sedentary life. For tlie divergent position of the anal aperture 

 and the looping of the alimentary canal in the Brachiopoda, we 

 refer the reader to p. 81. Among the most typical features of the 

 MoUuscoida are the peculiar development of the cephalic region, 

 the presence of a horseshoe-shaped, tentacle -bearing lophophore 

 having a distinct part of the body-cavity belonging to itself, the 

 so-called lophophoral cavity (circular canal of the Gymnolaemata, 

 smaller brachial sinus of the Brachiopoda), and the possession of 

 a fold (epistome) above the mouth, into which another part of the 

 body-cavity is continued (epistomal cavity, larger brachial sinus of 

 the Brachiopoda). The relations of these parts of the body-cavity 

 inter se and to the lower cavity which contains the intestine are 

 in some respects still obscure. As centre of the nervous system, 

 we have a ganglion lying above the oesophagus, which in the 

 Brachiopoda, however, is less massive than the ventral ganglion. 

 The excretory organs are represented by a pair of nephridia, which 

 at the same time function as ducts for the genital products. (In 

 Phoronis australis and in Bhi/nchoneUa there seem to be two pairs 

 of nephridia; cf. on excretory organs of the Bryozoa, p. 56.) 



While we thus find far-reaching anatomical agreement between 

 the adults of the various divisions grouped as MoUuscoida, the 

 similarity between the larval forms and their metamorphosis is less 

 evident. "We shall, however, have to lay stress upon the fact that, 

 in the larvae of the three types to be distinguished {Actinotrocha, 

 Bryozoan, and Brachiopodan larvae), the posterior section, by means 

 of which the fixation of the adult is accomplished, first appears 

 in an invaginated condition. Tliis invagination in the Actinotrocha 

 (Fig. 4 C, ic, p. 7) which yields the body-wall of the posterior 

 section of the body, may be homologised with the sucker-invagination 

 of the Ectoprocta and with the pedal section of the Brachiopodan 

 larva, which is sunk into the mantle-fold. If we maintain this 

 homology, we shall have to proceed to explain the metamorphosis 

 of the MoUuscoida from the Actinotrocha as the most primitive 

 larval forn]. The Actinotrocha type is distantly allied to the 

 Trochopliore larva, being distinguished from the latter chiefly by 

 the presence of a true coelom. We have pointed out above (pp. 57 

 and 78) that the derivation of the Bryozoan and Brachiopodan 

 larva from an Actinotrochan form still presents some difficulties. 

 We may well feel inclined to compare the posterior end of the 

 body, which in the MoUuscoidan larva appears in an invaginated 

 condition, and by means of which fixation takes place later, with 



