22 HEMIASTER CAVEKNOSUS. 



be different stages of growth ; H. aiistralis corresponding to tlie younger 

 stages of II. caveruosus, with its deeply sunken lateral ambulacra. Unfortu- 

 nately sufficient material is not available to settle this question, as the 

 species of Hemiaster from South America are extremely rare in the collec- 

 tions I have examined. Very small specimens of H. Philippii are quite 

 cylindrical (.75""" in length) ; they go through changes of form very similar 

 to those described in the growth of Brissopsis up to the youngest stages 

 figured here, when the gradual sinking of the ambulacral petals commences. 



Lat. 37° 42', South; Long. 56° 20', West. 44 fathoms. 

 Lat 51° 26', South; Long. 68° 5', West. 55 fathoms. 

 Off Cape Uos Bahias (East Coast Patagonia). 55 fathoms. 



Hemiaster cavernosus 



! Tripylus cavernosus Pllli.., 1845, Wieg. Archiv. 



! Hemiaster cavernosus A. Ac, 1872, Rev. Ech., Pt. L p. 132. 



One large specimen collected at Port Gallant. 

 Tripylus excavatus 



! Tripylus ezcavatus PniL., 1845, AVieg. Anh. 



With the southernmost specimens of Agassizia excentrica Avere collected a 

 couple of small Spatangoids, which I am unulile to determine satisfactorily. 

 They resemljlc, in general outline and appearance, Agassizia, ))ut without 

 having the characteristic structure of the anterior lateral ambulacra of that 

 genus ; the course of the peripetalous and lateral fascioles is the same, as 

 well as the general proportions of the ambulacral petals. 



The changes undergone during growth in Hemiaster Philippii, which have 

 been described above, seem to point to the probability of these small Spatan- 

 goids being the young of Tripylus excavatus. They have all the structural 

 features of the genus, except the deeply sunken lateral ambulacra. The 

 largest specimen, measuring about 7°""' long, diam., differed from the smaller 

 one by having slightly longer posterior lateral pctaloid ambulacra, already 

 slightly depressed ; while in the younger specimen the lateral ambulacra 

 were flush with the test, as in Agassizia, and in the Aoung of II. Philippii. 

 It seems probable, therefore, that in genera with depressed or deeply sunken 

 petals the young have, in their earlier stages, petaloid ambulacra flush with 

 the test. In Schizaster canaliferus the lateral posterior ambulacra are but 

 slightly sunken, even in the largest specimens, while the anterior ambulacra 



