REPORT OF THE DIRECTOR I916 181 



Ceratodictya, in its ontogeny, may properly be construed as ele- 

 mentary rather than senile; that is to say, the constrictions of the 

 surface do not, in the ontogeny of the sponge, appear to be pre- 

 ceded by any other conformation of the skeleton. 



The specimens in hand are long, slender, twiglike skeletons, with 

 very prominent horizontal rings, all of which, in the best preserved 

 of the two examples, appear to be duplicate on the periphery; that 

 is to say, these rings stand out prominently and abruptly; they are 

 compressed, and the surface is grooved in such a way medially, 

 as to make each ring a duplicate one. On the second and longer 

 specimen this duplication is not obvious, or at least is no more 

 than suggested, but the preservation of this individual is bad and 

 the absence of the crown groove may be due entirely to this fact. 

 These annulations are separated by smooth and subcylindrical sur- 

 faces, which are rather long, as they have fully twice the width of 

 the annulations themselves, and this is rather more than the interval 

 in the other species of the genus. As preserved, the specimens 

 show a sinuous or gracefully curved stock or twig, very slender 

 and scarcely tapering from one end to the other, so that the 

 original sponge must have been of extraordinary length in pro- 

 portion to its diameter. The shortest specimen measures lOO mm, 

 and carries 12 annulations; the longer specimen is 180 mm, and 

 carries 23 more or less recognizable annulations. The average 

 diameter of the shorter and better preserved specimen is 15 mm. 



The matrix of this specimen is Chemung sandstone of somewhat 

 coarse grain, and no trace is left of the reticulum. The larger 

 specimen is buried in the sand, but the smaller is represented only 

 by an external mold. The block which carries this smaller specimen 

 has also a very indistinct trace of either another individual or 

 the continuation of the first, bent at a very sharp angle. The 

 locality of these specimens is given by Mr Armstrong as '' about 6 

 miles southeast of Erie, Pa." 



HYDNOCERINA gCn. nOV. 



It seems necessary, for purposes of reasonably precise definition, 

 to introduce a new generic term here for species of. these sponges 

 which combine the characters of Hydnoceras and Ceratodictya in 

 such a way that they can be safely referred to neither. I think 

 the group which is represented by the species herewith described, 

 must be looked upon as a departure from the normal development 

 of the elementary types of these sponges. The variability in their 



