4 Herbert L. HcnvJcins — On the Holectypoida. 



tubercles and graimles, is a sliglitly depressed groove that extends 

 over about four of the interambulacral plates. This groove is 

 generally smoother than the surrounding parts of the test, and the 

 granules occurring in it are often unlike those outside. In Pygaster 

 they are usually abundant and minute, most of them being transversely 

 elongated. In Holectyptis, and particularly in II. depressus, they are 

 relatively few in number, of fairly large size, and guttate in plan, 

 tending to radiate from the areolte of the main tubercles. In Discoidea 

 there is hardly any difference between the granulation of the groove 

 and tliat of the surrounding regions. 



At the places where the ' suprabranchial groove' crosses the 

 transverse sutures there are developed elongate depressions. Three 

 such hollows commonly occur in each groove, those nearest to the 

 peristome being usually the most deeply excavated. They are 

 completely devoid of granules, and in the case of Discoidea are each 

 surrounded by a smooth, ovoid area, flush with the test surface. 

 The pits are more deeply cut in Coenholectypus serialis than in any 

 other form I have examined (PI. I, Fig. 1). Here they are more 

 or less pyriform in outline, the pointed end being directed towards 

 the adradial suture. 



The fact that the shallow grooves into which these depressions are 

 trenched seem to originate at the branchial slits, makes it probable 

 that they must have some function connected with the exter'hal 

 branchiae. However, the distance to which the grooves and pits 

 extend from the peristome is far greater than one would expect the 

 branchioe themselves to traverse. In most regular Ectobranchiates 

 the 'gills' are small bunches of vascular tubes that hardly project 

 beyond the peristoraial membrane, and they are protected from the 

 radioles and pedicellarise of the adjoining plates by a smooth lip-like 

 area. A similar ' lip ' is present in the Holectypoida, so that it seems 

 probable that the nature of the ' gills ' was also similar. 



The curious guttate and elongate granules of the suprabranchial 

 groove may perhaps have supported pedicellarise of a special kind 

 that would act as ' gill-rakers ' ; but this supposition of course 

 involves the capability of such organs for reaching the branchiae. 

 But I can suggest no function or meaning for the depressions along 

 the sutures in this region. Their invariable presence, and frequently 

 considerable depth, seem to indicate that they are structures of some 

 physiological importance. They resemble the sutural excavations 

 common among the Temnopleuridas (these occur in at least one 

 Holectypoid genus, Coptodiscus). Whatever may be the value of the 

 pits in those forms, the restriction of these depressions to the supra- 

 branchial groove must indicate a more localized function. They are 

 certainly quite different from the ambulacral pits above described. 

 The latter are 07i the phites, not between them, and contain a granule, 

 whereas the suprabranchial pits are particularly smooth. 



To sum up : the Holectypoida (except Conulus) possess two series 

 of excavations on tlie test near the peristome. One series, ambulacral 

 in position and regularly situated with regard to the triads, may 

 possibly have lodged sphaeridia, although there are many difficulties 

 met with in such a view. The other series is interambulacral, 



