292 SYNOPTIC COLLECTION. 



The boundaries of the segments would in time become 

 definite and the segments themselves be capable of mov- 

 ing freely upon one another. While this may be an 

 explanation of the origin of segmentation, there is as yet 

 no evidence to prove it in embryology. Granted, how- 

 ever, that this explanation be correct, it is doubtful 

 whether the flattened, unsegmented stage of development 

 is represented by adult living worms of to-day. The 

 unsegmented Nemerteans and Turbellaria (p. 317) are 

 considered by many naturalists as the nearest living 

 representatives of ancestral forms, but there are so many 

 reasons for considering these as secondary and not primi- 

 tive groups (see p. 317) that we prefer to place them as 

 terminal branches rather than trunk forms of the great 

 genealogical tree of Vermes. 



The process by which at the present time a many-seg- 

 mented worm arises from a trochophore is shown more 

 plainly in Polygordius than in any other worm so far 

 described. For this reason we shall consider it here, but 

 provisionally, since it is not yet clear whether Polygordius 

 is a primary or a secondary form. As a rule such uncer- 

 tain species are omitted in this Guide, and an exception 

 is made in favor of Polygordius only because it illustrates 

 the subject far better than any other known worm. Its 

 trochophore, familiarly known as "Loven's larva" (PI. 

 727, fig. i, seen from the side), is free-swimming and 

 similar to the trochophore of Mollusca. It is spherical 

 and transparent. The mouth (fig. i, at the left) is on the 

 ventral side and in front of it is the pre-oral lobe, and 

 also the pre-oral band of cilia. Parallel with this band, 

 and just back of the mouth, is another band of cilia which 

 are much shorter than those in front. This band is seen 

 more plainly in fig. 2, which is a drawing of an older larva. 



The digestive system consists of a mouth, stomach, and 

 intestine ending in the anus which is at the posterior end 

 (figs, i, 2). Soon the trochophore begins to elongate 

 (fig. 3) by adding segments to the posterior end, while the 



