163 



wall of the peristomes. The marginal cancelli are fairly large, and are at first separated from 

 one another by narrow septa with sharp edges. In the more central regions the cancelli cease 

 to be separated by angular septa, and their openings are reduced by the development of a 

 concave iris-like cryptocyst. No ovicell seems to be present in this specimen. 



Of the specimens from Torres Straits, N" 6 has nearly all its zooecia isolated and 

 quincuncial, only one or two near the centre showing some disposition to assume a radial 

 arrangement. Their peristomes are greatly elevated centrally, but fall off sharply laterally, so 

 that the orifices have practically no lateral or distal raised wall. The central raised part of 

 the peristome is either truncated, the edge forming a segment of a circle without spines; or 

 its edge is fringed out into from three to five long delicate spines. The cancelli have their 

 openings much reduced throughout the colony. Spinules are borne by the outer walls of the 

 peristomes, as in 394. P. The marginal lamina is in this case free and upturned at its edge, 

 so as to form a saucer; and, as has been noticed in some of the other specimens, its surface 

 has a finely granular texture. In the other Torres Straits specimen (81) it is for the most part 

 closely adherent to its substratum. 



The specimen from Tosa, Japan, has a close resemblance, in most respects, to N° 6 

 from Torres Straits. The internal edge of the cryptocyst of the cancelli may be very finely 

 denticulate; a feature which is barely indicated in 394. P. ('Siboga'), but is well marked in 

 several of the other specimens. None of the colonies show any of the pin-like spines which 

 are so conspicuous a feature of L. novae-zelandiae. The Tosa colony shows a well marked 

 ovicell (fig. 5), which has some resemblance, in form, to that of certain species of Bercnicea, 

 being transversely elongated and giving off an ooeciostome [o) from the middle of its marginal 

 border. The ooeciostome is a simple tube, which at first runs horizontally, but soon curves, 

 so that its simple, nearly circular ooeciopore faces directly frontally. The ovicell, which exhihits 

 no tendency to run into lobes between the zooecia, is more or less outlined by a series of 

 ordinary cancelli, but its roof is not covered by typical cancelli. This part of the ovicell is, 

 however, traversed by irregular, low ridges, between which are a moderate number of pores, 

 each pore being as a rule sunk at the bottom of a slight depression. The roof of the ovicell 

 is more convex than the adjacent part of the zoarium. The remainder of the central part of 

 the colony is occupied by a somewhat similar convex surface, which is, however, simply 

 granular and does not seem to be porous. I cannot distinguish an ooeciostome in connexion 

 with this part, and I am unable to decide whether it represents a second ovicell or not. There 

 is nothing to show whether its cavity is continuous with the ovicell, which has a porous roof; 

 but it is not impossible that it my be merely a part of that ovicell. 



Of the three specimens on the Japanese slide J.Q, one has no ovicell; but in each of 

 the others the central part of the colony is occupied by an ovicell of the type already described, 

 except that it sends out short lobes, peripherally, between the more centrally placed zooecia. 

 In one of these colonies the ooeciostome is very distinct ; and in the same colony the ridges 

 on the roof of the ovicell are produced, here and there, into spinous projections. The cancelli 

 in these specimens have their openings hardly reduced by cryptocyst. Some of the zooecia are 

 arranged in uniserial rays, the zooecia of which may be contiguous. 



163 



