AN UNDESCRIBED A CRANIATE. 229 



of the primary rods called by Spengel(3O) the " Biigelstucke " are 

 very large and prominent ; ventrally, however, the endostylar plates 

 could not be demonstrated by maceration or by section. Such sec- 

 tions give the usual marked difference between the primary and 

 secondary bars, but on account of their small size would make 

 difficult any attempt to trace the exact course of the blood vessels 

 as determined by Lankester (23) or by Spengel (3O). 



In the digestive tract of this Bahama amphioxus there is a 

 marked zone of small, crowded cells that take up stain readily and 

 form a dark band posterior to the stomach region as indicated in 

 Fig. 5. This is found in the Florida specimens of JBranchiostoma 

 caribceum, but appears not to have been noticed in the European 

 Amphioxus. 



The atrial chamber has the usual longitudinal, ventral ridges 

 that have been regarded as of excretory function, but a long search 

 for the nephridial tubules found by Boveri(3l) was without suc- 

 cess, probably owing to the small size of the specimens and the 

 distortion due to preservation. The live specimens were not ex- 

 amined for this purpose. Surface preparations of preserved speci- 

 mens of the Florida B. caribceum showed nephridia like those 

 obtained by Boveri by this method, but none were made out in the 

 small Bahama animals. This atrial chamber differs from that of 

 European forms in the extent of its posterior portion. Thus while 

 Lankester (23) finds a single pouch running back of the atriopore 

 "as far back as the anus " there are here two short pouches, the 

 right longer than the left, that seem to have pushed themselves 

 back on either side of the atrial outlet and then, very soon, stopped 

 growing. Both are shown in Figure 15, which is posterior to the 

 base of the atrioporal spout, though showing its tip ; further back 

 Figure 21 shows no representative of the atrium. 



In the Florida B. caribceum there is a large post-atrioporal 

 pouch on the right, but it does not reach as far as to the anus. 



In the youngest individuals taken, the floor of the atrium is but 

 imperfectly formed, the two sub-atrial ridges being incompletely 

 fused on the median, ventral line. 



Here there is an interesting segmented arrangement of the 

 muscle fibers such as has not been noticed before, I believe. The 

 transverse muscle of the atrial floor (Fig. 14) begins as a series of 

 2 



