Dr. H. Kauff (??i the Genus Hindiaj Dune. Ill 



that even in certain perfectly globular Astylospongida, likewise 

 destitute of a stomachal cavity, the canal-system is of pre- 



Fig. 4. 



cisely similar nature, consisting merely of radial tubes • the 

 system in ILindia therefore is not quite so isolated as Prof. 

 Steinmann states. 



In the Astylospongidaj also the individual element and the 

 tectonics of tlie skeleton seem to be in an analogous way 

 governed by a centripetal mathematical law. 



As regards the mode of fossilization and preservation of 

 the Ilindicey the specimens now before me from Tennessee 

 St. Petersburg, and the island of Sylt are completely silici- 

 fied, but the skeleton in all is quite hollowed out ; but while in 

 the first-named the outermost layer of^ the trabecula3 is ure- 

 served as a thiu siliceous husk (at any rate a thin layer suffi- 

 cient for the preservation of the form), and the original outlines 

 and forms of the skeleton thus still appear in all their dis- 

 tinctness, this husk is deficient almost without exception in 

 those from Sylt, which therefore are true matrices, and the 

 spaces originally occupied by the skeleton are more or less 

 enlarged and eroded. Nevertheless there can be no doubt 

 that the Sylt fossil is identical with those from America and 

 Russia : leaving out of consideration the skeletal husks all 

 the appearances and conditions are the same ; in both radial 

 fractured surfaces show the same radial six-sided rods, which 

 are the infiUings of the tubes, and on the angles of these the 

 same alternately-placed little tubercles, and the cross bars 

 uniting the rods, wliich are due to the infilling of the mesh- 

 spaces *. The interstices between these granules and cross 

 bands, as well as the narrow cavities bounding the infillino-s 

 of the radial tubes in tangential sections, exactly agree in tlie 

 course, form, and size of the nodal distances with the skeletal 



* See F. Romer, ' Tennessee,' pi. ii. fig. 2 h, and Hinde, ' Catalogue,' 

 pi. xiii. tig. 1 a. 



