53 



Generally the tylostyles are slightly curved. The variation of the base and the apex, 

 the shape of the "head" (phyma) l ) and all such details are better seen in the illustrations than 

 any description can give. The length of the spicules varies between 150 and 800 p. ; the maximal 

 diameter between 5 and 23 a. For details see the tabular view on p. 60 -). 



The development of the phyma is variable; on the whole it is distinctly present, but in 

 almost every preparation sorae spicules hardly possess it and the tylostyle becomes a style. 

 The apex is likewise variable and not unfrequently it is not pointed but rounded of. Finally it 

 occurs that thus both ends are rounded of so as to form strongyla. I found a few of these e. g. 

 in specimens B. M. 12, B. M. 15, B. M. 16, B. M. 22, B. M. 31, M. L. B. 132 ; further in 210, 305 c, 

 966. In 305 a there is a remarkable variety of style or strongyla (PI. XIV, fig. tf— e), which I 

 did not find in any other specimen. As the illustrations show, they are often very short and 

 thick, much thicker than the tylostyles. I cannot make out what these formations are. As we 

 find transitions to ordinary tylostyles I feel rather inclined to consider them as (pathologically?) 

 modified tylostyles. I don't know anything about the way how the latter originate in our sponge ; 

 whether built up by one cell or by more than one. It is often supposed that such spicules 

 "grow" normally as long as the socalled central canal remains "open". It is possible that in 

 305 a by some agens this canal is shut very early, at a time that the mother-cell or mother- 

 cells is or are still able to produce more layers of silica over the original "embryonic" spicule 

 and in this way give rise to the curious spicules. As I did not only find them plentifuily in the 

 preparations of isolated spicules but also in the sections of the sponge, it seems to me hardly 

 admissible that they are corpora aliena. It is worth our while to remark that the successive 

 concentric layers of silica are much more conspicuous in these abnormal spicules than in the 

 normal ones ; and secondly that in 305 a several obvious abnormalities in the spicules occur. 

 Although I did not find these spicules in other specimens, at least not to such an extent, it 

 must be remarked that sometimes a thicker, bluntly terminated spicule occurs, as e. g. in 210 

 and 305 c (PI. XIV, fig. 4). 



Another question is in how far oxea may belong to the complete set of spicules of 

 Spirastrclla purpurea. I found them in preparations of B. M. 8, B. M. 19, B. M. 23, B. M. 27, 

 B.M. 31 and in 1341c!, 1391, 1402, 1945- They have always quite another character than the 

 tylostyles or its modifications, and they are always scarce. I believe, therefore, that they do 

 not belong to the sponge under consideration, but are foreign, as so often occurs. I feel the 

 more inclined to consider them as corpora aliena since in 1391 I found also a few acanthoxea, 

 and in B. M. 44 strongyla, oxea, acanthoxea and tetraxons ; all however in very small quantity. 



The fact that in addition to tylostyli other spicules occur (oxea, styli or strongyla) is not 

 new, for previous authors mention the same. Thus, e. g. Ridley & Dendy say (1887 p. 231) 

 about Spirastrella massa (B.M. 31) that it "is also remarkable for the great irregularity in 



1) I propose the term "phyma" for the swollen base of a tylostyle. 



2) I have taken into account only my own measurements, which are calculated in the same way. This is done as follows. The 

 drawings are made with aide of an Abbe camera. With the same combination of lenses, length of tube etc. drawings were made of 

 the lines of a micrometer; the real distance of two lines known, it is easy to calculate how many times the draw ing is enlarged. 

 A quantity of drawings are made with Edinger's projecting apparatus (new system of Leitz); the calculation is of course the same. 



