and their Relationship to the Corals. 117 



ciiliar forms of the flatly dilated net-like stocks are j^roduced 

 by the repeated concrescence of the branches and anastomosis 

 of their cavities, so in the sponges there are found stocks not 

 only dilated and reticulated, but even twisted up into a coil, 

 whilst at the same time their branches, i. e. " persons," grow 

 together and anastomose at their points of contact. Among 

 the Calcispongiae these labyrinthic coils become so dense, 

 especially in the Nardopsidce and Tarromtda?^ that the inter- 

 spaces between the adult ^' persons " have been frequently 

 taken for the internal cavities of their communicating canal- 

 system. Thus, for example, Kolliker describes the interstices 

 and fissures between the densely united branches of the stock 

 of his Nardoa si^ongiosa (our Torrus spongiosus) as " eiferent 

 canals," and the internal vibratile canal-system (the cavities 

 of the branches) which occurs in this as in many other sponges, 

 as '' a network of ciliated canals, such as has hitherto been 

 seen in no sponge," 



The most remarkable results are produced by continued 

 concrescence of the " persons " in the genera Nardoa^ Nar- 

 dopsis, and Coenostoma^ which I have therefore comprised in 

 the distinct order of the Cojnosyca. In these, after the attain- 

 ment of maturity, the stomachal cavities or " flues " of the 

 different " persons " which comjjose a stock, and which have 

 been produced by lateral gemmation from one " person," open 

 together finally into a single cavity (a common " excurrent 

 tube ") which opens outwards by a single orifice (a common 

 mouth). As the mature sponge in this case possesses only a 

 single mouth-opening, it is apparently only a single '"'"person^"' 

 hut in reality a true stocky i. e. a cormus composed of several 

 "persons." In youth each "person" possesses a proper 

 mouth-opening, until it subsequently becomes united with its 

 neighbours, and forms, together with these, a common mouth- 

 opening. 



If we are to distinguish these wonderful animal-stocks the 

 " persons " of which, by excessive centralization, have given 

 up the most essential part of their individuality, the mouth, 

 and in place of it have acquired a common stock-mouth {cor- 

 mostoma), from the primitive polystomatous cormi,by a parti- 

 cular denomination, they might perhaps be fittingly named 

 Coenobia. The oldest form of starfish (Tocastra), which, ac- 

 cording to my hypothesis as developed in the ' General Mor- 

 phology,' is also the primitive stem-form of the Echinoder- 

 mata, would have to be regarded as a coenobium of this kind. 

 If, in accordance with this phylogenetic hypothesis, the primi- 

 tive form of starfish actually represented a stock of annulated 

 worms (persons) which had formed for themselves a common 



