2 Herbert' L. Hawkins — On the HoledyiDoida. 



occur on this plate, and therefore are situated adorally to the areolae 

 of the large tubercles. 



In P. ' semisulcatus' (PL I, Fig. 2), counting from the peristome, 

 two triads in each column of the ambulacra are normal, but the 

 succeeding five show the structures in question. On the lowest plate 

 of each of these five triads, almost impinging upon tlie peripodium, 

 is a minute, crescent-shaped depression. The horns of the crescent 

 point obliquely adorally and adradially, and between them, where 

 the depression is becoming shallow, there is a minute granule. As 

 far as can be seen, this granule is similar in shape and size to those 

 of the ordinary parts of the plates. But whereas all the ordinary 

 granules may be said to have an axis at right angles to the test 

 surface, these slightly sunken ones have an axis sloping away from 

 the peristome, directed more or less into the depression on whose 

 brink they stand. On the side of each depression, away from the 

 peripodium, there is constantly a small tubercle, the persistent 

 occurrence of which is noteworthy in view of the apparent 

 irregularity in distribution of such tubercles in other regions. 

 It seems to indicate that the structures associated with the depression 

 were sheltered beneath a specially placed secondary radiole. 



Owing to the superficial nature of the depressions very well- 

 preserved specimens are necessary for their perception, but I have 

 been able to satisfj' myself that they are constant in number and 

 position in five examples of P. ' semisulcatus ' in my collection. 

 It seems likely that their characters are similar in P. ' tunbrella ' , 

 but in the material at my disposal they are not very clearly shown. 



Pits of a like distribution and nature occur in Holectyptis depressus 

 and H. hemisphoincus, but their numbers seem to be variable, even 

 on the ambulacra of a single specimen. This irregularity may, 

 however, be due to imperfect preservation. 



In Coenhohctypus serialis the corresponding pits are miich more 

 pronounced (PI. I, Fig. 1). The only specimen in my collection 

 (now transferred to the geological department. University College, 

 Heading), of which the peristomial region is exposed, is slightly 

 weathered, and all but the larger surface features have been 

 destroyed. In spite of this the pits are clearly visible, particularly 

 after the specimen has been stained with dilute Indian ink. As 

 in Pygaster, two more or less complete triads (counting from the 

 peristome margin) seem to be devoid of pits. These when they 

 occur are fewer in number than in tbe Oolitic genus, there being 

 only three in each area, except in amb. V, where there are four, 

 two in each column. They show no signs of being crescentic in 

 shape, but seem roughly elliptical in plan, and uniformly depressed 

 throughout. jS'o granules can be seen in them, but since all such 

 structures have been destroyed on the rest of the test, it is fair to 

 assume that they were present. The distribution of the pits is 

 interesting. There are two in columns la, Ila, Illb, IVa, Va & b, 

 and one only in columns lb, lib. Ilia, IVb. It will be noticed 

 that, except in area V, the occurrence of the pits is in strict 

 acf-ord with the order of plate- development as determined by Loven 

 {Etudes, p. 13). 



