ECHINOIDEA. II. 73 



importance for the animal, since it is hardly concei\'able, how the aiiimal could get the right position 

 again, if it were tiiriied over, the short spines being hardly able to set up the thick body with its 

 heavy contents. In the \ounger specimens the ventral side is not so flat, but the spines are here com- 

 paratively larger and in so far better adapted for keeping the bod)- in the right position. 



Some measurements are given here of the three best preserved specimens; the two larger ones 

 have only a few small holes in the test; in the third one the neck is broken, bnt the specimen is 

 otherwise well preserved. The other specimens are represented by separated anterior and posterior 

 ends. — All the measurements are in mm. 



Total length Width of -head Width of neck- 

 37 6 3 



26 47 2-5 



22 4-8 3 



The structure of the test (PI. VI. Figs. 17, 19, 21. PI. VII. Fig. 5) is essentially the same as in 

 P. phiale, the main difference lies in the extreme elongation of the inner piates of the bivium. The 

 labrum is rather broad and very long, 6-5'""' in the specimen of 26"'™ length. The inner piates of the 

 ambulacra I and V are both well developed, narrow and very elongate, joining at their outer end 

 the second plate of the corresponding series, the bivial ambulacra thus being uniuterrupted. Each of 

 the inner piates carries a single tube-foot', the same is the case in the ambulacra II and IV, the 

 edge of the invagination thus being provided with 8 rather well developed and distinct tube-feet. The 

 piates I. a, II. a, IV. a V. b thus do not carry two pores, but their relative size is in conformity 

 with the general rule. The inner plate of the interambtilacra i and 4 is distinct, but, as in phiale, 

 separated from the corresponding second piates by the widened ambulacral piates II. a. 2 — 3 and IV. 

 b. 2 — 4. The labrum joins at its outer end the two ambulacral piates I. a. 2 and V. b. 2. The\' are al.so 

 very much prolonged, no less than /■5""' in the specimen of 26'"™ length, somewhat widened in the outer 

 half. As in pliiah they show the curious feature of being split up at the oral and aboral end bv a 

 longitudinal line proceeding from both ends a long way iuto the piates; these two lines, however, do 

 not join in the middle (PI. VI. Fig. 21), the piates thus being undivided. In this figure the outer part 

 of the oral end of this plate is seen to be prolonged a considerable distance along the labrum to meet 

 the first ambulacral plate; in the largest specimen it is — as far as I am able to discern it — not pro- 

 longed orally down the side of the labrum, in the smaller one it is somewhat prolonged, but not so 

 much as in the specimen figured here. The sternum is situated very far back; it is not so much pro- 

 longed, 5'"" long in the specimen figured; the episternal piates which are comparatively rather short 

 (3""")) reach the point of the anal snout. The epiproctal piates are three on each side, viz. 5. a. 6 — 8 and 

 b. 7 — 9. The abactinal piates of the odd interambulacrum are not distincth- prolonged, their nmnber 

 therefore being larger than usual, 15 or 16 in the specimen figured of 26""" length. The bi\ial ambu- 

 lacra begin on the abactinal side at the anterior end of the test, being thus very little separated from 

 those of the trivium, which occupy the usual position at the anterior end of the test. The posterior 

 paired interambulacra (i and 4) are very curiously modified (F'ig. 13). The piates a. 2 and b. 2 are 



I In V. a tliere is exceptionally no pore in the specimen figured (1*1. V. Fig. 21). 

 The Ingolf-F.xpedition. IV. 2. lO 



