394 THE GORGONIANS OF THE BRAZILIAN COAST. 



In a former paper (1869) the writer proposed to reduce Phyllogorgia and 

 Hymenogorgia, as well as Xiphigorgia, etc., to mere sections of the modernized 

 genus Gorgonia, because the spiculation is characteristic and essentially idenf 1 

 in all these groups, while the mode of growth is various and of secondary im- 

 portance. 



The discovery of a variety (lacerata) of P. quercifolia, described below in 

 which some of the branchlets of a single specimen remain nearly terete as in 

 typical Gorgonia; while others are flattened, as in section Pterogorgia ■ others 

 slightly winged, as in section Xiphigorgia; with many that are broadly winged or 

 foliated, as in Hymenogorgia, confirms that view. (See Plate XXX fig 4 

 and PL XXXII, fig. 2.) 



However, out of deference to the views of other writers, I have here retained 

 Phyllogorgia , but so emended as to include Hymenogorgia, for our specimens show 

 no essential distinctions, even in the matter of branching of the axis. The only 

 distinction made by Edwards and Haime was that the branchlets anastomose in 

 Phyllogorgia and do not in Hymenogorgia. I find anastomoses, also, in the 

 typical variety of the latter. 



Phyllogorgia quercifolia (Dana) Verrill. 



Gorgonia quercus-folium Ehr., Corall., p. 143, 1834. 



Pterogorgia quercifolia Dana, U. S. Expl. Exped., Zooph., p. 647, 1846. 



Hymenogorgia quercifolia Val., op. cit., p. 13. M.-Edw., Corall., vol. I, p. 181, 1867. Hartt. 



op. cit., pp. 62, 196, 210, 1870. 

 Gorgonia quercifolia Verrill, Trans. Conn. Acad. Sci., I, p. 359, pi. IV, figs. 1-lb, 1868. 



Plate XXX, figure 3 (general figure of No. 1514, reduced). Plate XXXII, 

 figure 1 (one of the " leaves" enlarged). Plate XXXIII, figures 1, la (spicules 

 of the same). 



Variety quercifolia. 



The most typical specimen of this variety is 245 mm. high by 272 mm. broad. 

 It is flabelliform and nearly all the main branches are partially separate and 

 terminate in flat, deeply lobed, leaf-like expansions, closely resembling the 

 strongly lobed leaves of certain oaks, but many are irregular. 



The axis itself can be seen in outline, and especially when wet, by translu- 

 cency, against a strong light. The stem divides subpinnately into a number of 

 divergent primary branches and these send off numerous slender, divergent, 

 alternating branchlets, many of which fork, each ultimate division ending in a 

 lobe of the leaflets. In some places the small twigs anastomose, forming irregular, 

 often rather large meshes, filled in completely by the ccenenchyma. Thus tne 

 distinction between Hymenogorgia and Phyllogorgia, based on anastomoses o 

 the axis in the latter, does not hold good. 



The spicules are yellow and purple, rather larger than in the related forms. 

 The larger are roughly warted, acute double-spindles, with about four gra e 

 whorls of warts on each end. With these are stouter, similarly warted double- 

 spindles, grading into closely warted double-cones. 



