98 NEW YORK STATE MUSEUM 



make an angle of 58° with the vertical ; the ratio of the length 

 of these outer walls to the distance across the invaginated area 

 in a specimen a little more than half grown is as 11 to 9, in older 

 specimens the difference is greater. 



The oral portion, when the high ambulacral ridges and anal 

 piece are in position, is four times as high as the aboral and rises 

 from it as a dome, maintaining a vertical outline for nearly one 

 third its hight. 



The published restorations of this species have been made 

 from fragments showing but very little of the aboral portion 

 and with the ambulacral ridges and anal piece broken from the 

 oral surface. These figures give the aboral portion nearly twice 

 the hight of the oral while in reality the oral has very nearly 

 four times the hight of the aboral. A side view of this species 

 seems therefore to be more strongly suggestive of Pentremites 

 than Mr Billings supposed. 



The plates of the aboral surface, while nearly 80 in number, 

 are arranged in four horizontal circlets. The first two are of 

 basals and radials as in crinoids. The plates of the third circlet 

 consist of 10 bibrachials and 13 interbrachials. The plates of 

 the fourth circlet are between 50 and 60 in number but while 

 the alinement is horizontal and not zigzag, the circlet is cut at 

 each radius by the upward extension of the bibrachials. 



The radials are so remarkably like the basals of certain Rhodo- 

 crinidae and are so abundant in Chazy deposits that it has seemed 

 best to describe then\ rather more fully than is usual. 



Basals. Basals unknown but probably very small and together 

 having but little greater area than a joint of the stem. The 

 " basals " of E. Billings, of which he says '' from another speci- 

 men it appears that there are at least three, if not four or five, 

 basal plates, and their form is remarkable," are not basals but 

 radials. All published statements concerning the basals have 

 been made on the authority of the quoted passage and are 

 without value. 



Radials. Radials five, many angled ; each having two ( ?) 

 proximal very short sides resting against the basals ; two long 

 sides where each plate meets its neighbor in this circlet; next 

 above this on the left is a side which supports the largest plate 

 of a lower row of two or three interbrachials ; on the right are 

 either two or three sides meeting the whole lower row of inter- 

 brachials ; the remaining two sides, which are distal, support the 

 bibrachials. 



