M. ^. A. Zittel on Fossil Uexactlnellida. 421 



bution of the recent Hexactinellida*. According to this, these 

 sponges occur from about 65° N. lat. to about 50° S. lat., and, 

 indeed, at the very considerable depth of from 500 to 4000 

 feet. The recent Hexactinellida are therefore strictly inhabi- 

 tants of the deep sea. 



This fact may justify us a priori in coming to the conclu- 

 sion that the fossil forms also probably existed under similar 

 circumstances. And, in fact, considering the mode of occurrence 

 of the fossil Hexactinellida we arrive at the same result. 



In this respect, indeed, the Palaeozoic forms give us no 

 information. The genera at present icnown in the Silurian 

 [Astylospongia, Pakeo^nanon, Protachilleuvij Eosponqia^ Pro- 

 tospongici, Calathium, Tracliyum^ Archceocyaihiis^ Acantho- 

 spongia^ and ? Acestra) differ in essential characters from the 

 later Hexactinellida, and possibly required other conditions of 

 life than their successors. In the Devonian and Carboniferous 

 Limestone and in the Dyas our order is as yet represented only 

 by the very insufficiently investigated genera Steganodictyon 

 (Devonisiia.), Acanthospongia (Carboniferous), and Bothroconus 

 (Dyas). In Canada ArchcEocyathus and Eospongia occur in 

 the Potsdam, Calathium and Trachyum in the Quebec group, 

 consequently in the very oldest Silurian ( = Cambrian) forma- 

 tions. In Tennessee there are numerous examples of Astylo- 

 spongia and Pcdceomanoh in the Middle Silurian Limestone ; 

 whilst in Europe the former genus is found in Esthland and 

 Gotland in Upper vSilurian deposits, and all over Northern 

 Germany in diluvially transported material. In general the 

 sponges are associated with Brachiopoda, Trilobites, and Ptero- 

 poda, and occur in deposits to which one may equally well 

 ascribe a pelagic or a littoral character. 



From tlie Trias and Lias no Hexactinellida are at present 

 known ; and even in the Dogger there are only isolated ex- 

 amples of the genera Tremadictyon and CraticuUiria^ and, 

 indeed, in the Inferior Oolite, in the Bath and Kelloway 

 grou]j. 



The Upper Jura exhibits a rich development of Hexacti- 

 nellida, but only where it occurs in the form of limestone. 

 The " Spongitenkalk " of the White Jura y and 2, as well as 

 the so-called Birmensdorf beds in Poland, in the Swiss and 

 Frencli Jura, and in the neighbourhood of Niort, are the chief 

 localities for Jurassic Hexactinellida and Lithistida. They 

 also appear quite isolatedly in the Corallian deposits of 

 different localities. They are, liowevcr, completely deficient 

 in the muddy or sandy littoral formations of Northern France, 

 * J..r. |). ].■■)(). 



