E, MAKSHALLr. — TRABECUL.E. 161 



SHALL, '75). Thomsoint ('68, p. 120) described the Hexactinelli- 

 dan ' sareode ' as ' small in quantity, very soft, probably semifluid.' 

 Marshall {I.e., p. 15o) called tlie same of HoUcnia ' zahfliissig, 

 liell nnd durchsichtig ' with granulation, from which condition 

 that of Hijcdonema should have differed l)ut little. These remarks, 

 so far as they go, are essentially in accoi'd with my conception 

 of the nature of Hexactinellidan trabecule, notwithstanding the 

 fact that those writers at the time had a very vague and incom- 

 plete knowledge of the histology of their subjects. 



I believe the trabecular system in its entirety corresponds 

 both morphologically and physiologically to the system of 

 collencytes in that simplest form of the sponge connective-tissue, 

 the coUenchyme. There is every reason to assume that the said 

 collencytes, at least in certain sponges, anastomose with their 

 branching processes and thus place the various histological 

 elements of the body in protoplasmic continuity (Sollas, 

 '88, p. XLlv). In tlie He.xactinellids, that continuity of the 

 connective-ti.s3UC cells is present in an unusually marked 

 degree. 



Where is then the intercellular matrix which always occurs 

 in connection with the true connective-tissue? I believe it is 

 totally undeveloped in the Hexactinellida. As a matter of fact, 

 I find in the entire trabecular system no part which in genei'al 

 appearance or in the reaction against staining reagents can be 

 compared to it. 



The absence of the intercellular matrix is of fundamental 

 importance, as it goes a long way towards explaining the peculi- 

 arities of the Hexactinellidan soft parts. To it, in the first place, 

 are directly due the thinness of the trabeculse and the extraordinary 

 development in inveise proportion of the lacunar spaces between 



