SoLLAs — On VetvUna stalactites (0. 8.), 8^c. 487 



as distingnislied by their " imregelmassig astige Skeletkorperelien, 

 deren Aeste in einem knotig verdicktem Centrum zusammenstossen. 

 Da dieselben an ihren Enden nur massig verzweigt sind, so ensteht ein 

 maschiges Netzwerk, das in manchen Fallen grosse Aehnlichkeit 

 mit dem Gittirgeriist gewisser Hexactinelliden erhalt," &c. ( he. cit. i., 

 p. 23). In effect Zittel considered the skeleton as consisting of cor- 

 puscles, each having a solid centrum, with a variable number of rays 

 proceeding from it. These rays unite at their ends with those of 

 neighbouring corpuscles, to foiTQ a network deceptively like that of 

 some Dictyonine Hexactinellids. Thus Zittel at this time recognized 

 two kinds of nodes in the network — one furnished by the centrum of 

 the corpuscle, and the other by the union of the rays. 



Schmidt, in describing Yetulina, gives substantially the same ac- 

 count, only adding the important statement that " the centrum of 

 each corpuscle grows not only by the superposition of new layers, but 

 also by the more or less intimate additions to it of branches from 

 neighbouring centres, whereto a quantity of shorter curled or knobby 

 excrescences are associated." 



Zittel in his latest account finds only one kind of node in the net- 

 work, and recognizing in it a union of corpuscular rays, concludes that 

 the other kind of node furnished by the centrum is absent, and so is 

 led to regard the elements of the skeleton as simple rods forked at the 

 ends, i.e. as the rays of his previously described corpuscles without a 

 centrum. He thus re-defines the characters of the family as follows : — 

 " Skeletal elements, consisting of simple, generally straight, but some- 

 times curved rods, more or less strongly branched at the two extremi- 

 ties. The branched ends of several (4-9) neighbouring rods meet 

 together, and by their amalgamation form the nodes" (loc. cit. ii., 

 p. 276). 



In re-examining this question we may most conveniently commence 

 with an inquiry as to the true form of the skeletal corpuscles : Are 

 they centra with rays, or rods merely ? To determine this with cer- 

 tainty, it is best to boil a fragment of the sponge skeleton for some 

 hours with a strong solution of caustic potash. This dissolves the 

 opal of which the corpuscles consist ; and as it acts with most efiect 

 along the surfaces of contact or union between the corpuscles, it is 

 possible, by carefully watching the progress of solution, and arresting 

 it at the right point, to completely separate the corpuscles from one 

 another without damage to them by corrosion. As thus separated, 

 they are very beautiful objects — far more so than from an examination 

 of broken fragments of the skeleton one would expect. 



Amidst great diversity of form they constantly maintain a multi- 

 radiate character, and are never simple rods. For convenience of 

 description they maybe roughly sorted into three groups — which we call 

 types — 1, 2, and 3. Type 1 is the staple corpuscle forming the chief 

 mass of the skeleton ; it presents a rounded centrum, from which 

 two to five, more or less, cylindrical rays proceed ; these are given 

 off from the lower half, or two-thirds of the centrum ; they are seldom 



