StIM'CTI'KK <>I' SKKMVroN. 



»l 



the sponge as, in PhywspongUi Da^moni, they are scattered over the entire 

 length of the pi'iucipal vertical paragastrul strands. The relation of their form 

 to the noimal hexactins may ])e uTiderstood by observing that the sides of the 

 iiead of the anchor not extended into 

 spines are much swollen, forming ab- 

 breviated representatives of the rays 

 of the third axis. With these anclioi-- 

 ate rods are others, occurring in Phys- 

 ospoNGiA, Clatiirospongia iiud Lvno- 

 DiCTYA, which may be termed detnes; 

 formed of flattened rods with alternat- 

 ing broad triangular processes at the 

 sides (figure 5). The anchorate and 

 clemate I'ods are found together, but only 

 in the vertical fascicles. The spicules of 

 the outer or dermal Hvall of the reticulum 

 are seldom retained in place, and the 

 precise position of some of the forms is 

 to a considei-able degree a matter of con- 

 jecture. Here there are no smooth rods, 

 anchorate spicules or clemes, or if the 

 first are ever present, they are small and 

 of rare occurrence. The mesh of the 

 outer wall is, as already observed, com- 

 posed of stauractins, principally pentac- 

 tins, which lie at the angles of small 

 quadrate areas (see figure 4). These 

 pentactins are usually smooth-rayed but 

 not infrequently they bear thorns or long 

 spinules (figure 6 (9)). In these pen- 

 tactins a single low node near the intersection of the axes is frequently present 

 and represents a sixth ray. The surface of the «piadrules set off by the prin- 

 cipal pentactins is covered Avith, or filled by, isolated spicules of various 

 forms. It is not easy to conclude, except from comparison -with living 

 hexactinellids, which of these spicules are referable to the dermal layer and 

 what portion of them to the interstitial or parenchymal layer of the 

 reticidum. On account of the thinness of the entire sponge wall, the dis 

 tinction of such an intermediate layer is perhaps a purely conventional one. 

 It is usually extremely difficult to discriminate between the outer and 

 inner layers of the skeleton especially when under compression, but the 



FiouRK 5. Spicules of Dictyospongidip. 



1. Terminal portion of a small vertical rod in Motto- 



dictya osc\daUt. x 400. 



2. Cleme of Physospon(/ia Dawtoni. x 60. 



3. Part of cleme of AcUeodictya marsipus. x 400. 



4. Group of anchorate spicules, Pkysospongia Daumoni. 



x60. 



