ECHINUS GRACILIS. 295 



markings of the older specimens are scarcely laid out even towards the 

 abactinal region, where in the older specimens they are most distinct. 

 The main vertical rows of the primary .tubercles, both of the ambulacra! 

 and interambulacral, are very prominent; miliaries are almost absent, the 

 anal system of the youngest specimen being covered by five plates, one 

 of which occupied nearly the whole of the anal system. This species at- 

 tains a considerable size ; specimens are in our collection measuring 65.8 in 

 diameter, and another 78 uim in height, exceeding somewhat the transverse 

 diameter. 



In Echinus, as in Toxopneustes, we find in the younger stages (PL VII 

 f. 2, 4) the same unbroken vertical arrangement of the pores, taking next a 

 vertically arched form, still connected, and then assuming the arrangement 

 of the adult. In these genera the anal system is at first covered by one 

 plate, and undergoes changes similar to those of Strongjdocentrotus, by the 

 addition of four smaller plates (PL VII. f. l), and so on (PL VII f. 3, 5, 6), the 

 original subanal plate retaining long a greater prominence. The miliaries 

 are formed in these genera as well as Strongylocentrotus by radiating ridges 

 arising from the base of the primary tubercles, forming a sort of star (PL VII 

 f. i2) ; then they swell at the distal extremity, forming a set of club-shaped 

 spokes round the main tubercle ; these are little by little separated from it, 

 and become independent elliptical tubercles at first, and then miliaries or 

 secondary tubercles. The ten large buccal plates of the actinal membrane 

 are the first to appear. Small plates (in genera in which they are found in 

 the adult) are next formed between them and the teeth, while afterwards 

 they cover the whole membrane, as in Toxopneustes and Trigonocidaris, — 

 a mode of growth of the actinal buccal plates totally different from that of 

 the buccal plates of Cidaris, where new plates are added next to the coronal 

 plates of the test, carrying the poriferous zone along with them. 



In young specimens of Psammechinus microtuberculatus we have a struc- 

 ture similar to that of the young of Echinus proper (PL VIII f. 19- U). 

 The poriferous zone is vertical, there being no tendency as yet, in specimens 

 measuring 5 mm ' in diameter, for pores to form arcs of three pairs ; the tuber- 

 cles are already closely crowded together, but no more than is the case in 

 young specimens of E sphaera of the same size, where we see, perhaps better 

 than in any other species, the impossibility of separating Psammechinus as 

 usually understood from Echinus proper as has been maintained by Dr. Liitken. 

 In Hipponoe the lateral spreading of the pores takes place very rapidly 



