M. K. A. Zittel o)i Fossil Lithtsftdce. 121 



of the cup is to be regarded in this light. Sometimes both 

 surfaces are similarly constructed, and the canals penetrate 

 from both sides into the skeleton, either as simple, at first 

 somewhat curved, tubules, or dividing as they pass inwards 

 into two or three branches. Such an abundant ramification 

 as Hackel has described in the Leucones I have never ob- 

 served in the Lithistidaj. Penetrating canals, traversing the 

 whole thickness of the wall, are also deficient in the whole of 

 the second group ; but there are certainly cases in which the 

 canals only terminate immediately beneath the opposite sur- 

 face (Chenendopora). 



The two canal-systems, running in opposite directions, are 

 not always equally developed. Very frequently one surface 

 bears oscula, measuring 4-5 millims. or still more, and either 

 depressed [Hyalotragos, Ghenendopora) or prominent and mar- 

 gined [Verruculina, Einstomella^ Macandrewia^ Azorica), 

 while the other is merely covered with fine pores. The one 

 system is then reduced to a capillary net, whilst the other 

 principally provides for the efflux (and perhaps also the influx) 

 of water. In general, in the cup-shaped sponge-bodies, the 

 larger oscula are situated on the inner surface ( Verruculina, 

 Gorallistes, Macandrewia) ; but the contrary case may also be' 

 met with {Leiodermatium) . If both surfaces are beset with 

 larger oscula {Leiodorella^ Amphithelion) , a conclusion maybe 

 arrived at from the size of the orifices as to the development 

 of the canal-system. 



The third modification of the canal-system appears only in 

 genera with a vv^ell-developed stomachal cavity of cylindrical, 

 trochiform, or some similar shape. If we regard the wall of 

 the stomach as the inner surface of a cup-shaped sponge-body, 

 all that has been said as to the course of the canal-system of 

 the preceding group applies also to the present one. The 

 ostia of the radial canals opening towards the stomachal cavity 

 are distributed either in series or quite irregularly. The 

 canals penetrating from them into the wall are curved in a 

 somewhat undulated manner, rarely straight; towards the 

 outside they gradually diminish in size, at the same time 

 sometimes forking into a few branches. Similar canals origi- 

 nate in the interior of the wall, and take their course in a 

 radial direction outwards, where they open at the surface in 

 larger or smaller ostia. The genera GyUndrojyhyma, Phyma- 

 tella, Galymmatina, Meqaltthista, &c. possess a canal-system 

 of this kind. 



In a fourth group of globular, pyriform, top-shaped, or 

 cylindrical sponge-bodies, usually with a narrow central 

 cavity, straight (sometimes capillary) canals run in a horizontal 



