414 M. K. A. Zittel on Fossil Hexactlaellida. 



Hexactiuellida without taking into consideration tlie skeletal 

 structure and the canal-system. 



jSTevertheless the general form of the bodj must not be 

 neglected in a natural system as a secondary aid. If the rela- 

 tionships of a Hexactinellid have been established by the 

 investigation of the skeleton and the canal-system, the external 

 appearance generally furnishes excellent characters for the 

 distinction of the genera and species. 



Hackel and Carter, who certainly cannot be charged with 

 undervaluing the microscopic structure, and still less with 

 overestimating the value of the external appearance of the 

 sponge-body, have principally distinguished the genera within 

 the larger groups, both of Calcispongite and Hexactiuellida, by 

 their external form. But what is justifiable in the case of 

 living sponges will also be admissible in that of the fossils. 



The general form of the sponge-body, the constitution and 

 thickness of the wall, the size, form, and position of the cen- 

 tral cavity, the mode of union in the polyzoic forms, furnish 

 valuable data for classification. Special interest also attaches 

 to the mode of fixation of the sponge to tlie bottom. The 

 want of a so-called root, or its nature when present, sometimes 

 serves for the recognition of the different genera. 



In the fossil Dictyonina beard-like roots, composed of long- 

 isolated siliceous spicules, have not yet been detected with cer- 

 tainty. Generally the root forms a stalk-like prolongation, a 

 nodose or lamellar dilatation, or a ramified base to the sponge- 

 body. It consists of siliceous elements, which either agree 

 more or less with those of the rest of the sponge-body in form 

 and arrangement, or the tissue of the root shows a striking 

 difference from the true latticework. The lattice-structure 

 becomes indistinct, and the root consists of long parallel sili- 

 ceous fibres usually without an axial canal, the production of 

 which from Hexactinellid tissue is generally indicated only 

 by the transverse unions which occur at more or less regular 

 intervals. 



I must reserve a detailed description of all the characters 

 hitherto mentioned for a special work upon the fossil sponges 

 which occur in Germany. For this comprehensive mono- 

 graph a portion of the text and the figures of the microstruc- 

 ture of almost all the genera of Hexactinellida are already 

 completed. I have been forced upon this larger work in part 

 by the lamentable condition of palseospongological literature, 

 but partly also by a feeling of responsibility, which urges me 

 not merely to erect the rough framework of a system upon the 

 sure bases obtained by a new method of investigation, as in 

 the present memoir, but to hand over the building also in a 



