ACANÏHASCIN^:. 133 



Ol" irregular in shape, a fact whicli can easily be observed under 

 the hand-lens and may be depended upon in concluding that the 

 dermalia are diactins and not forms with cruciate paratangentials. 



The gastralia are, with some exceptions, rough hexactins. 

 In the exceptional cases, they are represented by pentactins 

 intermixed with some stauractins {A. cactus), or by pentactins 

 and stauractins with occasional tauactins and diactins {S. heterad- 

 inus), or by diactins alone [S. pleoî'haphides, It. plumodigitatus). 

 The diactinic gastralia may insensibly intergrade with hypogastral 

 diactins. In the pentactinic forms the unpaired ray is always 

 directed away from the gastral cavity. In the hexactinic forms 

 the free proximal ray is often developed to a greater length than 

 any of the other rays. The cruciate paratangentials may form, 

 as they constantly do in several species, a continuous quadrate- 

 meshed gastral layer covering the excurrent canalar apertures, 

 but in several other species they are normally so sparsely present 

 as to leave in the eudosome gaps by which the excurrent water 

 passes out freely into the gastral cavity {A. cactus, S. doivlinyi, 

 S. solidus, S. tubulosus, S. japonicus). Also in cases in which 

 the gastralia are diactins {S. pleorJiaphides, R. jjlumodigitatus), a 

 continuous gastral layer is not developed, and hence the excurrent 

 canalar apertures are left freely open. 



A basidictijonal plate (PI. XII., fig. 37 ; PI. XV., fig. 12 ; 

 PI. XVIIL, fig. 14 ; PL XXL, fig. 12; etc.) seems to be possessed 

 by all the species that are firmly attached at base to the liard 

 substratum. The dictyonal framework discovered l)y F. E. Schulze 

 ('gg) in the buds of II. mirabilis, the like of which will be 

 described l)y me under S. glaher, I hold to be identical with the 

 plate in question. For the significance I attach to the structure, 

 the reader is referred to my Contrib. III., i). 24, foot-note. 



