160 ECHINOIDEA. II. 



points. On the lower edge of this limb the muscles of the stalk are inserted. At the upper end the 

 stalk is somewhat pointed. The head is ea. o-6 — cr8 mm , the stalk ca. i — rj™" long; the part above the 

 limb may be considerably longer than in the figured specimen. The rostrate pedicellarioe (PI. XIX. 

 Fig. 7) have long and slender valves, only slightly curved, except towards the point; the edge may 

 be quite smooth or more or less serrate; the point not much widened, finely serrate; the neck is 

 generally well developed, no limb on the stalk. Length of head ca. i mm . The tridentate pedicellarise 

 (PL XIX. Fig. 24, 26, 27) differ rather much in shape according to size, but only one form can be 

 distinguished. Large specimens (up to ca. o - 8 mm head) have short stalk and neck, and may be 3 — 4- 

 valved. The valves are rather wide apart, joining only for about the outer third of the length 

 of the blade. In the lower part the blade is narrow, more or less keeled on the dorsal side. The 

 edge is coarsely and more or less irregularly serrate; the serrations are generally bent outwards. No 

 meshwork in the bottom of the blade ; often a few crossbeams unite the edges in the lower part, 

 just above the apophysis. The outer part of the blade, where the valves join, is somewhat spoonshaped 

 widened, the edge being finely and regularly serrate. The basal part is rather small, with smooth 

 or faintly serrate edges; the apophysis is smooth. The short stalk is rather thick and compact, with a 

 rather distinct milled ring below. — Small specimens generally have a long neck and a longer, slender 

 stalk, consisting of distinct longitudinal fibres connected by crossbeams; the valves join in almost 

 their whole length, and may have a single large serration in the lower part, or this part may be 

 quite smooth, the edge otherwise being as usual finely serrate. The blade is simply leafshaped. — The 

 triphyllous pedicellarise (PL XVIII. Figs. 4, 11) may be rather variable in shape, but otherwise do not 

 present special features. — The sphaeridiae with rather numerous longitudinal ridges (PL XVIII. Fig. 22) 

 which is, however, scarcely a constant feature. 



In the « Blake. -Echinoidea, PL XXVI. Figs. 7 — 8 Professor Agassiz figures an elongated type ■■• 

 of Br. lyrifera, which differs very considerably from both lyrifera and alta through its confluent petals. 

 After having examined a number of specimens of this form — I am especially indebted to Professor 

 Rathbun for sending me several specimens to Copenhagen for study — I can show beyond doubt 

 that this form is not at all a mere local form of Br. lyrifera, but a very distinct species, which I shall 

 describe here under the name of Brissopsis atlantica n. sp. 



The general shape of the test is shown in PL III. Figs. 6, 10, 17. Also the figures cited of the 

 « Blake -Echinoidea show it rather well, only the posterior end of the test is generally almost vertical, 

 not sloping as in the Fig. 8, but there is some variation in this respect. The test is upon the whole 

 rather low, rising somewhat towards the posterior end ; the width is rather variable (see below, p. 162). 

 The actinostome is considerably more distant from the anterior border of the test than is the case in 

 Br. alta; it is more as in lyrifera, but on the other hand the labrum is less prominent than in that 

 species. The narrow posterior prolongation of the labrum does not reach the second adjoining arabu- 

 lacral plates. The first ambulacral plate reaching within the subanal fasciole is the 6th, and there are 

 generally 4 pairs of pores enclosed within the fasciole; sometimes, however, only 3 pairs are included, 

 which case may be found also in large specimens, while already in the smaller specimens 4 pairs of 

 pores may be found within the fasciole. I have also seen a specimen with 4 pores on one side and 

 3 on the other within the fasciole. Anal branches from the subanal fasciole may be distinct, but it is 



