REPORT OF THE DIRECTOR I9IO 243 



with cross sections of tubes belonging to the same system. As to 

 the character of these marks we must note that the boundaries show 

 only as carbonized lines and spaces in places made more distinct by an 

 internal tube lining of limonite-colored mud. There are no distinct 

 walls as in sections of bryozoan colonies. 



The fact that these tubes are commensurate in diameter with the 

 sutural canals would lead one to question as to whether or not they 

 were internal continuations of the exothecal canals. The fact that 

 they contain limonite-colored muds would negative this idea, for we 

 have supposed that b.vs. contain coelomic or other body fluids and 

 not sea water. The idea is also negatived by the direction taken 

 by the tubes, their curved character, and their great number. 



Another hypothesis is open for us. We have seen that the con- 

 ditions of existence favored a system of anal respiration. Such a 

 system would favor saclike extensions of the rectum that would tend 

 to become branching diverticula and so form respiratory trees very 

 like those possessed by Holothurioidea. These might be either homo- 

 generic or homoplastic. As the specimen died with the anal area 

 down, the intestine would press any structure between it and the 

 plate against the latter. In this case we should find portions of the 

 supporting tubes on their sides and variously bent. Surrounding 

 these would be a border of tubes whose tips only touched the inside 

 wall of the anal plate and these would give us the cross sections we 

 find. These tubes would contain more or less of the limonite-colored 

 muds of the bottom, drawn in by the last respiratory efforts of the 

 creature as we found them drawn in in Blastoidocrinus. This limon- 

 ite-colored mud is most often very suffuse and only faintly apparent. 

 In other places it has collected as little lumps but always in the 

 tubes. The tube walls are very poorly preserved and appear as if 

 partly ruptured or decayed. Their soft walls were often compressed,, 

 thus making them present angular outlines. 



Through the courtesy of the late Doctor Whiteaves the writer was 

 allowed to develop this portion of the specimen. From o.i to 0.2 

 mm of this surface was removed with a fine narrow file and 

 the area photographed again under the gum dammar mounting and 

 with an amplification of 10 dia. The result is shown in plate 6, fig- 

 ure 3. Another thin sheet was removed, but in order not to destroy 

 any of the sutures next r.post.R it ran from a thickness of o mm 

 on the right to between o.i and .2 mm on the left.. The uncovered 



