Speciks ok thk CuKMirNo Gkoup. 141 



irregular a character as to convey the impression that the frond miglit have 

 conformed its shape to the irregidarities of the surface ui)on which it grew. 

 Yet our knowledge of the mode of growth in living siliceous sponges does not 

 favor this conception ; rather, that the frond was attached at one extremity, its 

 expanded surface free and, consecj^uently, its irregularities not due t<j such 

 conformation, but altogether idiocratic. These irregularities may be roughly 

 classed as great and small. The former are the main depressions and eleva- 

 tions of the frond ; the latter are groups of nodes or pits upon the surface of 

 the former, which are probably of similar nature to the tufted node of 

 Hydn^ocp:uas and other genera. On the cxi)osed surfivce of the sjHjcimen in 

 hand, these nodes are expressed as pits generally narrow and elongate rather 

 than circular, and seldom occurring singly. 



The reticulum is fine and sharp but shows only the ultimate and penulti- 

 mate quadrulation, spicular bands of higher order being wholly obscured. 

 The direction of the spicular lines indicates no convergence to any single 

 point; those of the horizontal series curve as the contour of the surface 

 requires, and seem to approximate a circular curvature in both the upper and 

 lower portion of the specimen. The upper area is the better defined, and near 

 the center of the general curvature of the spicular bands is the best defined 

 group of small nodes upon the surface. We may thus infer a remote pos- 

 sibility of the attachment of the frond at more than one point, whence the 

 growth of the spicular skeleton progressed. 



This species has been observed in but a single specimen having a length 

 of 190 mm. and a greatest width near the middle of 90 mm. 



The interpretation of the obscure genus Cryptodictya, as well as Aci'ino- 

 DicTYA, is made clearer by the aid of Hallodictya. In Actixodictya, the 

 evidence of spicular reticulation is restricted to coarse, ridge-like bands which, 

 though seeming to make, at times, quadrangular meshes, usually appear to 

 be without much order, so that, by itself considered, there might be a possi- 

 bility of misconstruction in regarding them as representing the actual mesh- 

 work of the sponge. In Cbyptodictya, this condition is the more obscured 

 by the actual obliteration of nearly every trace of the spicular skeleton except 

 the tufts at the tips of the nodes. In this genus, however, the siu-face nodes 

 are quite as strongly developed as in Hallodictya. All three genera agree in 

 their explanate, irregular and undefined mode of growth. 



Locality. Hallodictya Sciensis occurs in a compact siliceous sandstone at 

 Scio, Allegany county. (Collection of E. B. Hall.) 



