54 



LO WER INVEIl TEBRA TES. 



sponges there occurs between the undoubtoil mosoilerin and the ectoderm di^itinit 



layers, the origin of which is uncertain. 



One of the most interesting points to tlie 

 naturalist lies hi the history of the skeleton and 

 its elements. Tliis consists of two parts, the 

 thread of binding substance of horn or keratode 

 and the hard mineralized spicule. All authors 

 apparently agree in considering the spicules as 

 mesodermic, but the origin of the threads has 

 not been so thoroughly worked out. Barrois, 

 however, considers them of ectodermal origin 

 in the silicious sponges, and the author has ex- 

 jtressed the same opinion regarding the fibres of 

 the horny sponges. In the ChalininiB the same 

 would also appear to be true. The skeletal 

 threads of Chalinula are surrounded by a 

 special membraiie, which I have seen in sev- 

 eral instances, and which may be called the 

 jierifibral membrane. This is composed of flat 

 ej)ithelial cells, either transparent or deejily col- 

 ored by granules. They somewhat resemble the 

 cells of the ectoderm in outline, but are longer, 

 fusiforni in outline, very closely set, and usually 

 sj)irally arranged around the fibre. These are 

 e\ idently the cells which secrete the threads, 

 ftnd in one section I followed this sheath and 

 found it continuous with the ectoderm. We 

 can thus readily account for the skeleton of 

 Chalinula by the presence of invaginated pro- 

 longations of the e])iderm which would natural- 

 ly follow and surround first the vertical threads 

 and then others arising in all directions, 'i'he 

 differences in the structure of the inner and the 

 outer portions of the fibres of the A])lysina, and 

 their often hollow condition, can only be ac- 

 counted for by this ex])lanation as well as the 

 fact that iu SjMiujia and its allies the centre of 

 the threads is frequently occupied by foreign 

 ni itter, carried in from the exterior by the in- 

 vagination of the ectoderm to form the sheaths 

 and subsequently enveloped by the liorny mat- 

 ter secreted. 



The form of the s]ncules varies greatly, and 

 affords good systematic characters. A few of 

 the forms are shown in the adjacent figure. 

 Some are pointed at one end, some have both 



extremities acute, while others may terminate at one or both ends like anchors. They 



may l)e smooth or variously laiobbed and ornamented. 



Fio .50. — l>itjereut forms of spoiige spicules. 



