5Q 



the openings do not correspond with wide canals of the nature described above (PI. V, figs. 6 

 and 8). In the second place we see, that some of them lead into comparatively deep holes or 

 furrows, others into very shallow ones. The structure of the wall is entirely different from that 

 seen to be the case in the "central" (incurrent) canals ; it corresponds, on the contrary, rather 

 with that of the external surface. I saw this distinctly in 931 b , where the position and amount 

 of spinispirae is in both the same. In both we find the yellowish pigment-cells (?). As far as I 

 could tracé the canal system doubtlessly excurrent canals communicate with those holes. They 

 must consequently be considered as belonging to the excurrent system. The apertures of the 

 ultimate excurrent canals open here, like in 1365 etc, by means of small procts, which are 

 perhaps slightly larger than the (incurrent) stomata on the free outer surface. The holes act as 

 cloacae; the often irregular terminal openings might be called oscula. Thus we see that in such 

 specimens at least part of the numerous large openings on the surface are not in-, but excurrent 

 ones. I find exactly the same arrangement in 1047 and others. As stated before, this specimen 

 1047 strikingly resembles Vioa florida of Lendenfeld. This author says (18975 p. 10S): "Von 

 dieser Kruste erheben sich clann machtige, zapfenartige Papillen, welche .... zu zylindrischen 

 Körpern auswachsen .... welche .... an ihrem abgerundeten Distalende eine Anzahl grosser, 

 unre^elmassigfer Löcher besitzen .... von denen weite Röhren ins Innere des Schwammes hin- 

 abziehen. Diese Röhren sind nicht Oscular-röhren sondern blosse Einsenkungen der ausseren 

 Schwammoberflache". It is evident what Lendenfeld means and that his "Einsenkungen" of 

 the external surface correspond to the holes I described above. In so far I fully agree with 

 Lendenfeld and can but confirm his correct observation. But according to my views there is 

 no evidence against their cloacal nature or, in Lendenfeld's terminology, against their being 

 "Oscular-röhren". Also I do not thirik they are really "Einsenkungen", due to deepenings or 

 excavations of the surface, but places which remained open between fused neighbouring processes. 

 We saw that a characteristic feature of tropus concrescens is the tendency neighbouring processes 

 exhibit to fuse. The examples I have given are evident; however, this fact once acknowledged, 

 we see the tendency likewise in specimens of other tropi, indiscreminately in specimens with 

 long or with short processes. This mode of growth fairly explains the irregularity or the outline 

 they often show and the shallowness or depth of the "holes"; hence it not unfrequently occurs, 

 that the pits are contorted, slitlike etc, giving finally the "meandering grooves or elongated 

 pits" which Dendy said to be characteristic for his Suberites inconstans var. maeandrina. In 

 fact we find the coalescens of neighbouring parts, leaving grooves, pits, furrows etc. between 

 them in numerous specimens, belonging to different tropi ; thus e. g. in B. M. 11 (PI. II, fig. 2), 

 1447 (Pk IV, fig. 4), in " Spirastrella insignis" , in S. pauis var. massalis, in 1949, 1947, 1975, 

 in B.M. 23, M.M. 63, 1341c, 305c, 612c, in " Vioa florida ", in 1047, 98, B.M. 9, M.M. 60, 

 M.M. 59, 1447 etc - etc - A very interesting example of this formation we have in 944a and 

 944b. Seen from above 944b shows pits like those in other specimens (PI. IV, fig. 5). But we 

 see on the left hand side of the drawing a short cylindrical process with a large central cavity. 

 In 944a (PI. I, fig- 2) we see several rather long cylindrical processes with a large central 

 cavity. At first sight one might compare these to the similar processes in specimens of tropus 

 tubulifera. Closer examination, however, teaches us that the wall of the cavity resembles, the 



