158 THE VOYAGE OF H.M.S. CHALLENGER. 



the contrary, may be regarded as combining the characters of the two groups ; in 

 Agassizia the posterior lateral ambulacra having the usual petaloid structure, while the 

 anterior pair of ambulacra are only petaloid in the posterior half, the anterior half retain- 

 ing the characteristic features usually found only in the anterior ambulacra, and this 

 wenus having in addition the embryonic features of a globular test and ambulacra flush 

 with the test ; in Hemipneustes the double structure of the petals extends to all the 

 lateral ambulacra. In Echlnocardium, Breynia, and Lovenia we find, with the internal 

 fasciole, that the structure of the pores is again simple, while in all the ambulacra the 

 petals correspond to that of the group to which Hemiaster and the like belong. 



Yet in all the young of true Spatangoids I have had occasion to examine, the ambu- 

 lacra consist of simple pores, extending from the apical system to the actinostome, the 

 change which characterises the groups thus spoken of taking place very gradually with 

 advancing age. The pairs of pores such as we find in the Cassidulidse can be traced 

 directly to the affinity of the CassidulidEe to such groups as Holectypus (which in their 

 turn retain features of the Desmosticha), and while the ambulacra assume a more or less 

 petaloid shape, yet the pores never come together into a single foramen. The earliest 

 known Spatangoids {CoUyritesf) retain this feature, and it is still found at the present 

 day existing to a certain extent in Homolainj^as, Argopatagus, and Genicopatagns, which 

 difi"er, however, from these earlier types in having the well-developed labiate Spatangoid 

 actinostome, and seem to hold from the structure of their ambulacral system much the 

 same relation to the Spatangina which Hemipneustes does to Ananchytes and Agassizia 

 to the Brissina. 



*Palceotropus loveni (PI. XXI. figs. 3-16 ; PI. XXXIX. fig. 33 ; PI. XLI.figs. 28, 29). 

 Palaeotropus Loveni, A. Agassiz, Proc. Am. Acad., vol. xiv. ji. 204. 

 The anal system is placed above the median line (PI. XXI. figs. 3, 5), above the angle 

 made by the curve of the posterior extremity of the test extending from the apex, and 

 that extending from the actinal surface along the subanal shield. The anal system is 

 elliptical, transverse (PL XXI. fig. 8), surrounded by an outer row of large plates, the rest 

 of the system covered by plates irregularly arranged. The plates of the apical system 

 are indistinct (PL XXI. figs. 12, 13) ; the madreporic body is prominent; there are three 

 genital openings. The larger primary tubercles are perforate, crenulate (PL XXI. fig. 

 16) ; and the test is covered by a dense minute granulation between the primaries and 

 secondaries. Although Palceotropus has the rudimentary aml)ulacral system, simple 

 pores perforating the primary ambulacral plates as we find them in Pourtalesia, it 

 yet has a simple compact apical system ; the ambulacra are not disconnected as in that 

 group at the summit by the encroachment of large distinct interambulacral plates 

 extending from the ambulacral area across the apical region so as to separate the bivium 



