130 M. K. A. Zittel on Fossil Lithistidce. 



pears, or can only be exceptionally detected ; but even in the 

 latter case the four arms are always unequally developed and 

 differently branched, and they do not meet in tlie centre at 

 any definite angle. In certain genera, such as Doryderma 

 (PL VIII. fig. '6) J Lyidiumj and Heterostima, they are divided 

 into several unequal branches, which again may fork into a 

 few short and blunt lateral branchlets ; in others, such as 

 Megalithista (PL VIII. fig. 4) and Carterella, short branches 

 originate at the ends of the elongated and curved skeletal 

 corpuscles, diminish rapidly in thickness, and usually fork only 

 once, or at the utmost twice. Besides these, short knobby 

 processes spring from the main stem here and there. In the 

 genus IsorhapJimia the skeletal corpuscles acquire nearly the 

 form of simple, crooked, cylindrical spicules ; but their thick- 

 ened ends, often cleft into two short branches, prove them to 

 be true Lithistid elements. 



All Megamorina possess a simple axial canal, which some- 

 times traverses nearly the whole length of the main stem 

 (fig. 4), but without ever making its appearance at the ends ; 

 but sometimes only forms a short capillary central canal 

 in the middle of the branched skeletal corpuscle. 



The skeletal corpuscles eitlier compose alone the entire 

 inner skeleton of the sponge-body {IsorhapMnia^ Dorydermay 

 Lyidium^ Megalithista)^ or they are accompanied by much 

 smaller, strongly ramified siliceous elements [Heierostinid) y 

 which in their general characters agree with those of the next 

 group. The union of the large Megamorine corpuscles is 

 effected by the curved ramified ends applying themselves to 

 neighbouring skeletal corpuscles, and, indeed, frequently com- 

 pletely embracing them. 



A small group of Lithistidse, hitherto known only in the 

 fossil state, is characterized by its irregularly ramose skeletal 

 corpuscles, the branches of which meet in a nodosely thickened 

 centre (PL VIII. fig. 5). As the branches are only moderately 

 ramified at the ends, a meshed network is produced, which in 

 many cases shows a great resemblance to the latticed frame- 

 Avork of certain Hexactinellidas, and, indeed, may be confounded 

 therewith upon a hasty examination. By the furcation of the 

 4-7 usually smooth arms, these sponges, which I have pro- 

 posed to denominate Anomocladina, are proved to be true 

 Lithistida3. The genera Cylindrophyma^ Melonella^ Leca- 

 nella, and Mastosia are the representatives of this group, from 

 which, possibly, the Tetracladina have been developed. 



In the great majority of the Lithistidae the skeleton con- 

 sists neither of these Anomocladine corpuscles, nor of distinct 

 quadriradiates, nor of large, feebly ramified Megamorine ele- 



