154 NEW YORK STATE MUSEUM 



rectly on settling down from the free larval state, or conjunction 

 may be formed by attachment of the annelid larva after the growth 

 of the coral has well progressed; in both cases the growth of the 

 latter engulfs the former save at its tentacled aperture. 



The coral Zaphrentis or Cystiphyllum and the worm Gitonia 

 corallophila. I give this latter designation to what appear to be 

 chiefly straight worm tubes found in simple cyathophylloids such 

 as those mentioned. The worm has attached itself at any stage of 

 the coral growth and quite often its tubes are found projecting in 

 considerable number from the calyx of the coral disordering the 

 septa by its thickened stereom and taking just the position most 

 advantageous to its feeding with the help of the coral's tentacles 

 [pi. 2, fig. i]. Often these tubes seem to puncture the thecal walls 

 of the coral where actually they have become overgrown or left 

 behind by the increase of coral substance. It is not usual to find 

 both of these conditions in one corallite. Plate 2, figure 3, shows 

 a Zaphrentis with a series of small worm apertures at its base; 

 figure 2 is an enlargement of the thecal wall of Zaphrentis with two 

 apertures one of which shows distinctly the wall of the tube ; figure 4 

 is a Cystiphyllum with apparently short-lived worm tubes established 

 at different growth stages of the coral. ■ In figures 5, 6 of the same 

 plate are two views of a tube both ends of which seem to open into 

 the calyx of a small Zaphrentis. If I interpret the growth of this 

 worm correctly it started almost concurrently with the coral and 

 like the worm on Pleurodictyum kept both ends up. It will be seen 

 by examination of these figures that the course of the worm tube is 

 singularly erratic ; -both branches have kept close to the margin of 

 the calyx, one has come pretty straight up, while the other in its 

 late stages made almost a half circuit of the calyx. 



All the examples above cited are from the Onondaga limestone of 

 the Lower Devonic. 



The corals Monticulipora and Stromatopora and the worm 

 Gitonia sipho. These compact, stony, massive sti'uctures covered 

 with thousands of arms reaching out for new supplies of 

 nourishment, seem to have especially invited the settlement of 

 straight tubed worms which, for convenience, are designated as 

 Gitonia sipho. 



A very striking example is that illustrated in plate i, figure 4, 

 where the coral has overgrown the face and eyes of a moulted head 

 shield of the trilobite Dalmanites and a series of worms has started 



