PKXTACKROS BOYSTI. 81 



The arms are well produced. R = 80 mm. and i- zi= 18 mm., the major radius 

 being thus about four and a half times the minor radius. Tlioy taper gradually 

 to the extremity. The breadth of the I'ay about the fourth supra-marginal ])late 

 is irS mm. The height of the ray at the same spot is almost exactly the same. 

 Tho ravs are steep-sided, and consequently appear almost square in cross section. 



The supero-marginalia are adjunct throughout almost the whole length of the 

 rav, for only one or two single radialia enter the base of the ray. At the base of 

 the ray they are flat and slightly rhomboidal. They possess an anterior indenta- 

 tion on their inner surface and are about 3"5 mm. in Ijreadtli. They gradually 

 diminish in size distally and at the same time become distinctly swollen. They 

 numlier al)()ut twenty-ciglit. 



The infero-marginal plates are appioximately of the same size and number as 

 the supero-marginals. Both series imbricate slightly. The ornament of these 

 plates consists of a number of fine granules in the centre, while there is a distinct 

 margin without granulations. 



Between the supero- and infero-marginal plates a few scattered granides repre- 

 sent a slight development of the intermarginalia. 



Tlie adambulacrals are a series of small ol)long i)latL's. They border the int'ero- 

 marginals from aliout the eleventh supra-marginal onward. They are much worn, 

 and but slight traces of their armature I'emain. Al)out five adambulacrals occupy 

 the same length as two infero-marginal plates. Proximally there is a single row of 

 small plates which separate the two series. 



Only a few scattered ossicles of the actinal surface of tlie disc remain. 



Locality and Stratigraphical Position. — Upper Chalk, Kent. 



BemarliS. — Valette ('Bull. Soc. Yonne,' 1U02) has described a number of 

 species of starfishes from the Senonian of the South of France. The remains are 

 found as scattered ossicles. Some of these are grouped by Valette as a new 

 species which he calls P. senoiiensls. They arc noticed by the author to resemble 

 P. Boijsii except that they are smooth and therefore do not have the ornament 

 possessed by P. Boijsii. Valette regards this absence of ornament as rendering 

 them specifically distinct from P. Boi/sii, as other ossicles found in close 

 proximity still possess the ornament. In view of the vagaries of the way in 

 which solution may occur, I cannot admit this contention and consider that it 

 is much more probable that tlie ossicles at one time possessed ornament and 

 were identical ^\^th P. Boijsii. All the other ossicles except those of the so-called 

 Arthraster senonensis {fide infra, p. ii2) were identified with English Cretaceous 

 genera, which would support this contention. 



