3 8 DISCOVERY REPORTS 



of the colony, where branchlets bud, the terminal parts of the branchlets very soon touch and also 

 coalesce to form new meshes. Terminal branchlets may have a diameter of only about i mm., but they 

 soon grow thicker, and the diameter of the branches generally increases to 3-4 mm., in some cases 

 even to about 5 mm. The meshes commonly have diameters of some 3-8 mm., and the size of the 

 meshes is fairly constant in older parts of the colony. During growth the branches become oval in 

 transverse section, the greater diameter being perpendicular to the main plane of the flabellum. In older 

 parts of large colonies the greater diameter may often exceed twice the shorter diameter in length. 



Accessory fans are common in older colonies. Such secondary fans diverge more or less strongly 

 from the plane of the mother fan and are attached to the latter along the greater part of one side. In 

 some cases a secondary fan may attach itself by its entire periphery to the mother fan, so that a closed 

 ' basket' is formed, generally with finer meshes than in other parts of the old colony. The photographs 

 on PI. II, fig. 2 and PI. Ill, fig. 1, convey the general features more clearly than any description. 



The largest dried colony has a continuous accessory fan formation along the entire lateral and upper 

 peripheral part, and the colony appears about to split along a line some 3-7 cm. within the margin 

 into two somewhat diverging fans. 



Under the microscope the surface of the colony gives a general impression of being finely reticulated. 

 However, when cleaned with sodium hypochlorite the coenosteum becomes rather glossy, and its 

 surface then exhibits a more vermiculated structure strongly reminiscent of a fingerprint (PI. Ill, fig. 2). 

 In places this vermiculated structure, owing to greater irregularities in the lines, presents a more reticu- 

 lated appearance, but in all cases the surface keeps its semi-glossy character and does not give the 

 'lump of sugar' impression of the Eu-Errina species described by Broch in 1942. 



The internal structure of the coenosteum is almost homogeneous, approaching to a certain degree 

 the compact porcellaneous consistency of the labiata group of Errina. On the other hand, the central 

 part with somewhat thicker branches is lighter in colour than the peripheral layers, and in some cases 

 the axial part is all but white. In accordance with the comparatively porcellaneous appearance, the 

 coenosteum is rather hard, and grinding is not so easy as in most of the Eu-Errina species. It must be 

 added that there are no coarser canals in the oldest and thickest branches. 



No correlation between the position of gastropores and dactylopores can be traced. The gastropores 

 are rather small, their diameter at the opening measuring only 0-30-4 mm. The pores are also not 

 particularly deep, and the well-developed gastrostyle is easily observed externally in colonies cleaned 

 with sodium hypochlorite (Fig. 3). The gastrostyle (Fig. 5) resembles a pine cone with rather bristling 

 scales, although it is comparatively compactly constructed. The cone is placed on top of a fairly high 

 pedicle with smooth walls. Occasionally a gastrostyle was observed having an almost needle-shaped 

 prolongation of the apical part of the cone. 



It is impossible to discern any regularity in the distribution of the gastropores in peripheral free 

 branchlets. In older branphes, on the other hand, the great majority of gastropores are found along 

 the lateral sides facing the meshes, whereas the anterior and the posterior sides of the branches (the 

 flabellum) have only a few gastropores. The numbers are also smaller on the posterior than on the. 

 anterior side of the colony. 



It has already been mentioned that no regular correlation between gastropores and dactylopores can 

 be traced. Nevertheless, from one to three or four spines are almost always formed at the opening of 

 the gastropore, each spine with one or more dactylopores (Fig. 3). In some cases such spines are 

 wanting, and in others dactylopores occur in great numbers between the spines and independent of 

 them. On the peripheral branchlets spines are numerous and equally developed on all sides. When the 

 branchlets grow thicker, the spines seem to diminish, or to disappear little by little, first on the posterior 

 and subsequently on the anterior side of the colony. Along the sides of the meshes, however, spines 



