so 



DlCTYOSPONOIDiK. 



skeletons in a pyritized condition, and it is from this source that all our 

 knowledge of their ultimate composition is derived. As already observed, the 

 frame work of these sponges consists of bundles of rods crossing at right 

 angles. The majority of these are smooth and taper very gradually, not 

 necessarily extending for the entire length of the sponge. Close observation 

 has not yet shown them to have more than a single axis, that is, they are not 



FiouBE 4. Spicules of Dictjogpongidae. 



1. Part of a primary pentactin of Cleodictya Mohri, witli umbels in various positions, 



and fragments of ottier spicular rods, x 60, 

 8. Two angles of a quadrule in Phragmodictya catillifomiis, the primary hexactins or 



pentactins being in place, with fragments of the vertical and horizontal 



bundltis. X 50 

 S. Part of a quadrule, showing the vertical and horizontal bundles with parenchymal 



spicules scattered over the interspace, x 40. From an undetermined fragment 



at Crawfordsville, Indiana. 

 4. Some of the major and minor hexactins of Dictyoapongia f Marcetlia. x 30. 



crossed by other rays. Such rods may, therefore, be regarded as extreme 

 modifications of the typical triaxial or hexactinellid form, produced by the 

 abbreviation of two of the axes. The terminal portion of a very small rod 

 {Mastodictya osculatd) enlarged 400 diameters,'is represented in the accompany- 

 ing figure 5. In some genera, particularly in Piiysospongia, vertical bundles 

 of these rods, which lie on the inner or paragastral surface of the sponge, 

 are accompanied by comparatively short rods which terminate in an anchor 

 with two sharp flukes (figure 5). These had nothing to do with the fixation of 



