ECHINOIDEA. II. i6i 



not a constant feature; in one case the anal Ijranch was quite distinct on one side and not at all disc- 

 ernible on the other. 



The peripetalons fascicle is narrow and elon<;ate. The petals, which nia>' be rather deepened 

 are so directed as to form a cresceut-shaped figure on each side — the character hitherto thought 

 characteristic of the genus Toxohrissiis (conip. below, p. 166-71. The posterior petals are iconfluent , the 

 posterior interambulacrum forming onh- a narrow separatmg bridge, with the primary tubercles not 

 beginning before about halfway out, whereas in lyrifera and alta the primary tubercles begin close 

 behind the apical system. In the inner (median) series of the posterior petals the large pores are found 

 only in the outer half, from about the 9th, whereas in lyrifera and alta the large pores begin near the 

 inner end, from the 4th — 6th. The odd interambulacrum is very narrow on the part between the peri- 

 petalous fascicle and the anal area. The madreporic plate is scarcely longer than in the two other 

 species, but it is somewhat narrower. There are four genital pores in the usual position. — Spines, 

 tubefeet and spicules do not afford any distinguishing characters. 



The pedicellarite are very richl\- developed; globiferous, rostrate, tridentate, ophicephalous and 

 triphyllous pedicellarise have been found. The globiferous pedicellarice occur, rather surprisingly, in 

 two very different forms. One form (PI. XVIII. Figs. 20, 24) has very elongate, narrow' valves, ending' 

 in two long, somewhat diverging, inward bent teeth. The valves are clad in a rather thick coat of 

 skin; the stalk is very short. Length of head ca. 1-5 — i-S""".' The other form (PI. XVIII. Figs. 5, 9, ig) 

 is like the type found in alta, biit there are generally only two teeth on either side of the terminal 

 opening. The stalk has a rather small circlet of thorns below. Length of head ca. o-s™"'. It may be ex- 

 pressly noticed that I have found both kinds of globiferous pedicellarise in the same specimen, though 

 certainly not in all of them. It nia}- not seem unreasonable to suggest that both kinds of globiferous 

 pedicellarise ma>' also prove to occur in other species, as e. g. Br. alta and columbaris (in which latter 

 species the slender form occurs, as I have been able to prove on specimens examined in the U. S. 

 National Museum). — The tridentate j^edicellarise likewise occur in two distinct forms. One form 

 (PI. XIX. Figs. II, 33) has very elongate, slender valves, very wide apart, joining only for a very short 

 space at the point. This outer part is widened, with finely serrate edges; all the rest of the blade is 

 quite narrow, with smooth edges or with one or a few teeth near the outer end. A few crossbeams 

 are generally found in the lower part of the blade. The valves are almost straight. Length of head 

 up to ca. 1-5™™. The neck is very well developed, the stalk long and slender. The other form (PI. XIX. 

 Figs. I, 28, 32) is ver}- like that found in Br. alta, and may likewise occur four-valved. There are large 

 teeth in the lower, somewhat narrowed part, whereas the outer part, where the \'alves join, has the 

 edges finely serrate. In smaller specimens of this form the valves join for a considerably larger part 

 of their length, only one or two large teeth occurring in the lower narrowed part. I have not seen 

 specimens of this form larger than 1-2™'" length of head. Neck and stalk as in the first form; the 

 neck maj- be much longer than in the specimen figured, in which it is somewhat contracted. — The 

 rostrate pedicellarioe (PI. XIX. Fig. 5) remind one very much of the slender form of tridentate pedi- 



1 A quite similar fonn of globiferous pedcellana; was describetl and figured by Dr. de Meijere (Siboga-Ecbinoi- 

 dea. p. 189. PI. XXIII. Fig. 474), from some specimens wrongl}- referred to Brissopsis lusonica; through the kindness of Pro- 

 fessor M. AVeber I have received one of these specimens and can thus state definitely that it is not Rr. lusovia but a new 

 species, which I intend to describe in Part II of the Siam-Echinoidea. 



Tlie Ingolf-Hxpedition. IV'. 2. 21 



