1920] SKELf:TAL Element in Lithistid Sponges 59 



part of the stalk having an especially long shaft, as is often the case 

 in such sponges. In the superficial part of the stalk there are abun- 

 dant triaenes with short shaft, and mingled with these are the likewise 

 abundant amphitriaenes. The latter spicules, in which the length of 

 the shaft is 160-540ft, are exceedingly variable, scarcely two alike, and 

 they form a close series grading over from amphitriaenes, quite like 

 the spicules I have described above, to the short-shafted triaenes. 

 These observations of the late Professor Lendenfeld securely establish 

 the origin of the amphitriaene. Amphitriaenes were already known 

 in the two other species of the genus, Amphitethya stipitata (Carter) 

 and in the sponge from Amboina designated by Topsent Tetilla mer- 

 guiensis Carter (Topsent 1897, p. 437). In the former Sollas 

 noted (1888, p. 49) that ''the amphitriaenes sometimes are reduced 

 to simple triaenes." In the latter, the detailed facts, such as similar- 

 ities in size and precise shape, convinced Topsent that the amphi- 

 triaenes are modified triaenes. Amphitethya is a genus with sigmata, 

 and thus falls in the Sigmatophora. 



There is still another non-lithistid tetraxonid sponge in which at 

 least a step has been made toward the transformation of the triaene 

 into the amphitriaene. This is Ancorella paulini Lendenfeld (Len- 

 denfeld 1906, p. 248) from the Indian Ocean. The spicules referred 

 to consist of a shaft, one millimetre or less in length, which bears at 

 one end three cladi, projecting slightly downward as in an anatriaene, 

 and at the other end a similar single cladus, extending out from the 

 shaft at an angle in the same general direction as the cladi at the op- 

 posite end. The sponge is classed in the Astrophora by Lendenfeld, 

 who regards its microscleres (microxeas) as derived from streptasters. 



It may be regarded as certain that the presence of the amphitriaene 

 in these several cases is not due to inheritance. The distance of the 

 sponges from one another in the classification, expressing their general 

 dissimilarity, negatives this idea. The comparison of Amphitethya 

 with Jereopsis is especially instructive. Each has plenty of close rela- 

 tives without amphitriaenes, and the two sponges fall in different 

 suborders of the Tetraxonida. To be sure the Lithistida may not be 

 a natural suborder but a polyphyletic group, some members of which 

 have been derived from non-lithistid tetraxonia with astrose ■ 

 microscleres (Astrophora), and others from non-lithistid tetraxonida 

 with sigmata for microscleres (Sigmatophora). Even in this case 



