TJRASTERELLA RUTHVENI. 141 



the arm referred to, all the specimens show the forms in oral view. Text-fig. 91 

 is a plan of the disc and one reconstructed arm of the holotype. The arm used in 

 the reconstruction is the longest arm shown in the photograph (PI. X, fig. 4). 

 Probably this represents the extreme length, but this is not certain. The other 

 four arms are broken off short. The disc is as usual small (r: 2*5 mm.). The 

 arms are long (R : 27 mm. at least) and slightly petaloid at the base. The mouth- 

 angle plates are of about the same comparative size as in U. grayse, and they are 

 succeeded by forty-seven adambulacralia. Each adambulacral is crowned by a 

 sharp ridge placed not medially but distinctly distalwards. The distal position of 

 the ridge gives the adambulacralia a peculiar appearance, making them look like a 

 succession of cups fitting into one another. The ridge is continued well on to the 

 lateral surface. Pustular elevations for the attachment of spines may be seen on 

 some of the proximal ridges. A number of the adambulacralia have broken away 

 from the right side of this arm, thus allowing one to obtain a good view of the 

 groove. The inner steep sides of the adambulacralia are well exposed, as also are 

 the noses. The usual slope may be also observed. In fact the appearance is 



o: 



Text-fig. 91. — Plan of ossicles on the oral surface of Urasterella ruthveni. 0., odontophor. x 3 



exactly as in U. grayse. The ambulaeralia, on the other hand, look at first sight 

 quite different from those of that species, for the ambulacral ridges are very sharp 



and steep and appear to run straight across the ossicle. Really the same l-shaped 



ridge is present in both species, although the short arm of the I bordering the 



ambulacral canal is not prominent in U. ruthveni. If the ossicles are slightly 

 displaced they appear just as in Schuchert's drawing of the ambulaeralia of 

 U. grandis (85, pi. xxx, fig. 3) which is obviously a drawing of the ridges 

 of successive ossicles without sufficient representation of the plates to which they 

 are really attached. 



The odontophors may be seen both in the holotype and in fl/510. They fit very 

 closely into the interradial angles and reach well up towards the apical surface. 



Casts from Bl give good views of the plates of the mouth-region. The 

 arrangement is almost precisely similar to that described for U. grayse. The groove 

 is widely open in this specimen and in one of the arms of B2. No point beyond 

 those already stated seems to call for mention. 



a/921 is a smaller specimen than the holotype. One of the arms is undoubtedly 

 present in full length, and hence gives the exact relationships between the major 

 and minor radii (see below). 



19 



