512 Causes and Course of Organic Evolution 



feature of the Annelida (Chfetophora) is the presence of chaitte 

 or bristles, on some or most of the segments of the body. So, 

 if the Annelida have had a rotiferan ancestry, we should prob- 

 ably be able to trace in some of the latter both annulation 

 and suggested segmentation of the body, as well as the rudi- 

 ments of chsetse. As to the former we need only cite Notom- 

 mata pilarius and auritay Taphrocampa annulosa and Albertia 

 intrusor or naidis to prove that alike annulation and com- 

 mencing segmentation are exhibited in several genera of roti- 

 fers (fig. 21). Setae, typical of the cephalic tentacles of many 

 rotifers, are often elongated and aggregated into patches of 

 two to five. But further, in Copeiis and the Triarthrida, spines 

 or groups of spines, some of which greatly recall those of the 

 Annelida, are set into the body wall as genuine chsetse, and 

 are moved by appropriate muscles. 



Again in the evolution of the Annelida from the Rotifera, 

 we might expect still to find rather advanced larval annelidan 

 stages that would aid us in bridging over the gap which may 

 seem now to separate the two. One such from amongst many 

 is presented in the larva of Eupomahis which we compare 

 in the adjoining page (Fig. 21 a, b, c) with a compounded dia- 

 gram of Pterodina from the Rotifera, and with Polygordius 

 from the Annelida. These speak for themselves. But an 

 important change passed through in this evolution process 

 must have been the retention of rows of cilia, left from a primi- 

 tively holotrichous condition, at the same time that these were 

 retained across areas of advancing segmentation. Alike in the 

 mature animal Dinophilus, and in the larva of the annelidan 

 Ophryotrocha puerilis figured (Fig. 21 e), the so-called telo- 

 trochal larva, typical of Mysostoma and of the adult rotifer 

 compounded above, has become a polytrochal type by reten- 

 tion of the ciliary rows and commencing segmentation. But 

 neither in Dinophilus nor in the larva are setae as yet observed. 

 Segmentation from before backward is then gradually estab- 

 lished, as i^ointed out by Korschelt. 



Thus there arises a polysegmental condition, which in many 

 points still largely retains rotiferan characters. Such exists 

 now in that simple group of fresh-water annelids, the Aph- 

 aneura, which A. Sedgwick has suggested should be placed 

 amongst the Archiannelida. It along with Dinophilus sug- 

 gests that many genera once existed which formed graded and 

 perfect transitions from primitive rotiferan and trochophore 

 types to planarian worms on the one hand, and to highly seg- 

 mented annelidans on the other. 



