242 INVERTEBRATE MORPHOLOGY. 



of setae. In Priapulus the intestine is almost straight and the 

 anus terminal ; but in Sipunculus and the allied genera, such as 

 Phascolosoma and Phascolion, the digestive tract is convoluted 

 and bent back upon itself, so that the anus lies on the dorsal 

 surface near the anterior extremity of the body. A " respira- 

 tory tree " is absent or rudimentary as a rule except in Pria- 

 pulus and allied genera, and the large prostomial lobe char- 

 acteristic of the Echiurese is lacking. The anterior por- 

 tion of the body, however, is capable of being iuvaginated by 

 means of strong retractor muscles (Fig. 109, dR and vR) into 

 the fore part of the digestive tract, forming the so-called in- 

 trovert. The extremity of this is provided with a circle of 

 finger-like or branched tentacles in the centre of which lies 

 the mouth, and which are supposed to have a respiratory 

 function, being richly supplied with blood. In Priapulus 

 these are absent, but at the posterior end of the body there 

 is a prolongation which bears papilla-like processes which 

 probably function as respiratory organs. 



Development and Affinities of the Gephyrea. The early development 

 of the Gephyrea resembles closely that of the Polychaeta, more especially in 

 the Echiureae. In this order a Trochophore larva is formed resembling very 

 closely the typical Polygordius trochophore, the similarity extending even 

 to a segmentation of the primitive niesoderm bands. In later stages this 

 metamerism of the mesoderm disappears, no traceof it being found in the 

 adult forms. In the Sipunculacea the larva differs from the Trochophore 

 in lacking the typical praeoral baud of cilia, though this may be weakly 

 developed in some forms, such as Phascolosoma. The postoral cilia are, on 

 the other hand, strong. A further difference is found in the absence of 

 metamerization of the mesoderm, which at a very early stage of develop- 

 ment forms a layer lining the interior surface of the body-wall, and also 

 covering the digestive tract and enclosing a coalomic cavity continuous 

 through the entire body. 



Notwithstanding these important differences there seems little room for 

 doubt but that the Kip/un'/ihis larva has arisen us an adaptation of the 

 typical Annelidan Trochophore still represented in the development of the 

 Echiureae. By these forms a close relationship is shown to the PolychaMa ; 

 and the Gephyrea are to be regarded as Polychseta which have secondarily 

 lost a metamerization originally present in the adult ancestors and still 

 represented in the Echiurus larva, but lost even in the larval stages of the 

 Sipunculacea. 



Since the discovery of the larval forms of certain Echiurid and Sipun- 

 culid forms there has been a tendency to regard these two orders as being 



