204 NEW YORK STATE MUSEUM 



The orad end is toward the left and it will be seen that the covering 

 pieces a increase rapidly in height with age while at the same time 

 they increase very little in thickness. The row back of these is seen 

 in part, though much out of focus. The upper surface of the two 

 rows presented a shallow channel into which the long and solid wing 

 plate, whose under surface is shown in figure i of the same plate, 

 fitted tightly, as is shown by the impressions made by these covering 

 pieces on the under side of this wing plate. Thin sections of the 

 wing plates show them to be homogeneous in their nature and not 

 formed by the fusion of smaller pieces. These plates serve to lock 

 the covering pieces and with them make a very high and solid 

 covering over the food groove. 



Directly under the covering pieces are seen the outer edges 

 of part of a row of adambulacrals, one member of which has been 

 marked b. The sutures between the covering pieces and the 

 adambulacrals can not be clearly seen on account of a thin veil- 

 like band of calcite which seems to indicate that the brachioles 

 were attached to the side of these plates, helping also to make a 

 solid structure of a pseudambulacrum. The veillike band but 

 slightly obscures the openings into the food groove. The rem- 

 ''"ftants of brachioles which still adhere to the specimen at the left 

 '^'"sHow that the food and water channels crossed the outer faces 

 '•'^bf the wing plate and covering pieces at an angle of about 25 

 '= degrees with the edge of the deltoid and on arriving at the open- 

 ings into the food grooves turned abruptly and ran down to the 

 edge of the deltoid (pi. i, fig. 2, d) at an angle of 90 degrees or 

 parallel with the deep vertical channels which run down between 

 the adambulacrals. These channels lead to the openings into 

 the hydrospires. The lower portion of the outer edge of an 

 adambulacrum presents a flat face against which rested one or 

 more of the basal plates of a brachiole. Above this the outer 

 face of an adambulacral becomes narrower and more rounded. 

 This is the region where so much light was admitted through 

 the rather thick section drawn for figure 2 of Bulletin 107, page 

 105, and which suggested "brood chambers." Taking the evi- 

 dence of the cross section and that now before us we may safely 

 conclude that water passing down the brachioles could enter 

 any one of a series of openings into the food groove and that 

 surplus water, or water deprived of its food content, could pass 

 down any one or more of the vertical side channels and find an 

 entrance into the hydrospires. On page 114 of Bulletin 107 



