224 E. A. ANDREWS. 



Immediately anterior to the anus the rectum is median (Fig. 21) 

 and the fin is on the right side ; but before the atriopore is reached 

 it approaches the median line and appears as a true median structure 

 as far as its free portion is concerned. The basal part, however, 

 even when the free edge is median, extends up to an attachment 

 to the muscular region of the right, much as is seen in Figure 21. 

 Even its free part is median only for a short space and soon pro- 

 jects again from the right side, becoming thus continuous with the 

 right metapleuron posterior to the atriopore, as is shown in Fig. 

 15, where the left metapleuron is seen in a similar fin-like state 

 and the posterior tip of the atrioporal spout is enclosed between 

 the two. 



The departure of the median fin from a true median position is 

 correlated with the breaking through of the median rectum : anterior 

 to the anus the fin becomes more nearly median in proportion as 

 the rectum rises up from its superficial ventral position or is more 

 and more enclosed by the downward extension of the lateral muscles, 

 passing from the opening seen in Fig. 22 through a stage shown 

 in Fig. 21 to a position shown in Fig. 15. Here, however, the 

 atrial formations again interfere with the median position of the 

 fin, and it becomes lost in the right metapleuron. 



The continuity of right metapleuron and fin suggests that the 

 fin may not be primarily a median structure, but one of a pair of 

 longitudinal folds of which the left forms only the left metapleuron 

 while the right is more extensive and comes to lie in the median 

 plane, as a locomotor organ, whenever the atrium and digestive tract 

 are not at the surface to render such a position difficult. 



The idea that the median fin is but a form of the right meta- 

 pleuron would not harmonize with the view suggested by Lan- 

 kester (23), that the double ventral fin of amphioxus may be the 

 fused epipleura, but that view would require re-stating if we accept 

 the formation of the atrium as recently described (26). 



As we have seen the median ventral fin of the caudal process 

 continue around the end of the notochord, thence along the whole 

 dorsal median line and finally become continuous with a median 

 ventral fin beneath the anterior end of the notochord, nowhere 

 showing any trace of a double or paired character, we might expect 

 that it would become continuous with the right metapleuron 



