ORGANIC DEPENDENCE AND DISEASE 47 



uals and it is safe to say that the worm is present in the 

 majority of examples. It is usually easy to determine its 

 presence on inspection of the tentacular surface of the coral 

 by the contrast between its round tubes and the angular 

 coral cells. All the specimens here figured to show the con- 

 volutions of the worm have been drawn from actual prep- 

 arations. 



The history of the combination in P. styloporum is as 

 follows : At the close of the free-swimming larval stage the 

 coral, in fully eight cases out of ten, selected and attached 

 itself to a dead or living shell of the common gastropod 

 Loxonema hamiltoniae.^ Directly upon fixation or even 

 actually contemporaneous with it was the attachment of the 

 larval worm upon the gastropod and alongside the incipi- 



1 The selective attachment of such lens-shaped coralloid stocks seems to have 

 acquired directiveness with the progress of time. At any rate we have a sug- 

 gestive intimation of this in the very common Chaetetes lycoferdon (Prasopora 

 simulatrix) in the Trenton limestone of the Ordovieian^ which is a stony coral 



Fig. 21. Basal surface of the solid bryozoan colony, Prasopora selwyni, which 

 has attached itself to the brachiopod Plectambonites. Trenton limestone 

 (Ordovieian), Ottawa. 



of quite the same shape and habit of growth as these Pleurodictya. This is 

 found attached sometimes to brachiopod (especially Plectambonites sericea) 

 and as often to gastropod shells which were the abundant exuviae of the sea 

 bottom. More often perhaps it is fastened to some casual stone or other hard 

 object, but among all of which I have taken note there seems to have been no 

 obvious preference by majority. 



