27 



To tropus glaebosa : inconstans var. globosa and var. niacaudrina, angulala, 964 a-b, 

 966 a-c, 98, 6l2a, a. o. 



I shall now have to show that not only cunctatrix, papillosa, g2a-l etc. (trop. tuèerosd) 

 cannot be specifically separated, as little as wilsoni, dysoni, areolata, 426 a-c etc. (tropus pyra- 

 midalis), but that specimens of one tropus gradually pass into those of another, nay that some 

 specimens might be placed in one tropus as well as in another. If this fact be demonstrated, 

 it will be evident, that the tropi as such do not represent as many species. 



What are the chief differences, by which specimens like cunctatrix or 426 c! (tropus 

 tuèerosa) are distinguished from, say 426 b-c, wilsoni, etc. (trop. pyramidalis)} At first sight it 

 seems easy enough to say what is the difference between cunctatrix and wilsoni, viz. the 

 presence or absence of spinispirae. However: the type of wilsoni (B.M. 33) is destitute of 

 spinispirae altogether. But in other specimens of wilsoni (B. M. 15, B. M. 22, B. M. 33) spinispirae 

 occur; I lay stress on the fact that in all these three specimens the size and the shape of the 

 spinispirae are different, as is shown on PI. X, tig. 5 and on PI. XII, figs. 3 and 4. 



One of the characteristic spicules of cunctatrix is the large robust spinispira. All the 

 specimens of cunctatrix in the British Museum possess these spicules (PI. X, fig. 6 1— X ; fig. 7 u-C, etc). 

 Comparing these figures with the spinispirae of wilsoni B. M. 15 and B. M. 22 the difference 

 strikes us at once. But this difference is not so conspicuous between wilsoni B. M. 32 (PI. X, 

 fig. 5 3— X) and cunctatrix B. M. 10 (Pb X, fig. 2 jt— v). Careful examination of the spinispirae 

 of several specimens, both of cunctatrix and of wilsoni teaches us how very variable they are 

 and we finally arrive at the conclusion that the difference between some specimens of cunctatrix 

 and wilsoni is certainly not greater than that of various specimens either of the one or the 

 other. The difference between the spinispirae of the above mentioned specimens (PI. X, figs. 2 

 and 5) is not larger than between the former (B.M. 10) and another cunctatrix, say B.M. 3; 

 on the contrary [Cf. PI. X, figs. 2 and 7]. If this be so, how shall we distinguish cunctratrix 

 and wilsoni as two different species? It is surely quite impossible on account of the robust 

 spinispirae, after all the most important character. An examination of the tylostyli leads to the 

 same conclusion. If we try to find a distinctive character in the external appearance or structure, 

 we meet with exactly the same difficulties. On PI. III, fig. 5 I have illustrated specimen g2c 

 and in fig. 1 specimen 426 b; they are typical examples ot trop. tuèerosa and trop. pyramidalis 

 respectively. The external appearance of these two iorms is quite different. In the former we 

 see an ellipsoid mass more or less regularly covered with tubercles or warts; in the latter there 

 is a large pyramidal cone provided with smaller cones on the slope, especially at the base. 

 However, there are many transitions: in g2a, 92 b, 92 d, 92 i some cylindro-conical elevations 

 of various size are visible, which are devoid of tubercles. On the other hand we see in 4260 

 several low warts, in 426 c only one or two, in 426 a and 426 b none. Likewise we observe that 

 cunctatrix B.M. 13 is covered with numerous tubercles all over; in B.M. 10 (PI. III, fig. 3) 

 the tubercles are distinct, but they are localised and large areas of the sponge surface are 

 entirely devoid of them. B.M. l is "pyramidal in form . .. with a smooth surface"; still it is 

 called cunctatrix'. — If we examine longitudinal sections of 426 a-c we see a wide central canal, 

 which narrows towards the top. Such large central canals are not seen in 426 d or 92 a-!; 



