664 R. KIRKPATRICK. 



ovei- the upper edges of the calcareous walls and between 

 any two tubercles (PL 33, f5g. 1). The circular masses show 

 a lighter central area, or occasionally two such areas. Fiiriher, 

 from one to three nearly vertical bundles of fine needles are 

 present in or above each circular mass. Scattered mostly in 

 or upon the surface of the sponge are very minute, oval siliceous 

 rings. The siliceous spicules are not visible in glycei'ine pre- 

 parations, but clearly so in balsam ones. Thin specimens 

 decalcified whole show very clearly the darker node-like 

 circular areas of masses of fiagellated chambei'S joined to each 

 other by five to seven radiating spokes, the blank spaces 

 between the spokes having been occupied by the calcareous 

 tubercles. The under-surface of fairly thick decalcified 

 specimens presents an extraordinary appearance, such as has 

 never been observed before in any sponge. Hanging down 

 from the thin surface-layer of the sponge are closely packed 

 but separate parallel moniliform cylinders. The segments of 

 each cylinder vary in breadth from 75 to 100 n (or rarely 

 150 ^), and in length from 75 to 250/^. The segments are 

 joined to those above and below by narrow central strands, 

 which had passed through the holes or slits in the tabulee. 



Fig. 9 on PI. 32 shows a vertical section of a decalcified 

 specimen. The spaces between the cylinders extend from the 

 base of the sponge to the apex of the gaps left by the dis- 

 solved tubercles, not taking into account the fine organic 

 basis of the skeleton remaining after decalcification. The 

 upper part of the section, including the ectosome and part of 

 the choanosome, is extremely contracted, but below this the 

 fiagellated chambei's, bundles of spicules, and the cylindrical 

 masses of large granular cells are to be seen. 



P'ig. 3 on PI. 33 shows a decalcified vertical section of one 

 of the two expanded specimens; in the pai-t of the section 

 shown in the figure there happens to be only one mass of 

 crypt cells that had been covered by a tabula. All the other 

 basal extensions of soft tissue had filled in calcareous pits, 

 the floors of which had been formed by the surface of the shell 

 on which the sponge grew (see below, Section 3). 



