neiv Beccnt Pharetronid Sponges. 509 



of liorny sponges, I name the cement- or mortar-forming 

 cells of calcareous sponges "telmatoblasts"*. They are unfor- 

 tunately not well preserved in the present specimens. They 

 i'orui in places a compact layer of columnar cells, about 9 fx 

 liigh, with prolongations at the distal ends (PI. XIII. fig. 9); 

 the base of the cells is closely applied to the outer walls of 

 the tubular canals of the canal system. Tiie contents of the 

 cells are granular, but I was unable to make out the nucleus. 

 The telmatoblasts are evidently modified branched collencytes ; 

 in parts where they had not become columnar and aggregated 

 into a compact layer the cells were flattened and discrete, 

 though at the same time joined by branched processes ; in 

 this condition a nucleus was visible. There could be little 

 doubt that the latter cells were the same as the columnar 

 ceils, but in a different condition, because transitions could be 

 traced, and both kinds formed deeply-stained patches in 

 similar positions relatively to the skeletal strands. 



In wholly undissolved skeletal strands the stain of the 

 borax- carmine entered about halfway into the thickness of 

 the same; in half-decalcified strands needles and pillars of 

 the cement were seen separated by a clear space from the 

 columnar telmatoblasts. In a rapidly decalcified stained 

 section the meshes of the tubular network of canals had 

 very strongly contracted, and had imprisoned the telmato- 

 blasts, which likewise had become contracted almost to fine 

 hyaline threads, from wliich the stain had disappeared ex- 

 cepting from a small point (? nucleus) about the middle of 

 the lengtl). 



The central part of the summit of a mortar cell can here 

 and there be seen embracing the point of a stalagmite (or 

 stalactite). 



I hope in the course of a few weeks to have some fresh 

 properly preserved specimens of the sponge described below, 

 and to be in a position to give a more detailed account of the 

 nature of telmatoblasts. 



Reproduction . — Minchinella is hermaphrodite. The em- 

 bryos belong to the parenchymula type. PI. XIII. fig. 10 

 shows one 128 yu- in diameter, surrounded by a nutritive follicle 

 formed of a single layer of large cuboidal blastomeres. The 

 outer layer consists of a columnar epithelium, which surrounds 

 a central mass of large cells. The wrinkling and shrinkage 

 of the embryo is due to the mode of preparation. 



The spermatogonia are present in various stages of growth. 

 An early stage in which there has been a division into two 



* TiXuii, mortal'. 



