696 WATER SYSTEM. 



and interambulacral systems, except the abactinal part of the latter which 

 lies between the petaloid ambulacra. Miiller has made a calculation that in 

 an Echinanthus rosaceus of ordinary size there must be as many as 100,000 

 locomotive pores ; they have been figured in a young specimen on Pi XI f . 

 The ambulacra! galleries are readily seen on the inner side of Clypeaster 

 subdepressua (PL XP.f. 4), Echinanthus rosaceus (Pi XP.f. ;). in Mellita 

 (PI XII". f. !-■;). and in Encope (Pi XII". f. 1 ; Pi XII'. f. 4). Con- 

 nected with the ambulacra] system of the Clypeastroids is the forked pro- 

 longation of the actinal ambulacra! edge of the test forming a tube with 

 several openings, which give passage to small tentacles identical with those 

 of the furrows proper. Agassiz considered them as the probable openings of 

 gill passages, but neither .Miiller nor myself have been able to find in living 

 Echinarachnius anything but the tentacles mentioned above. 



The course of the water-system branches is well shown in Echinanthus 

 (PI. XP.f. .'i. and in Mellita (PL XII". f. 2), seen from the actinal side; 

 when seen from the interior, the branches of the water system on the acti- 

 nal floor are shown in Pi XXX. lower fig. Although the tentacles of the 

 petaloid portion of the ambulacral zone are apparently so different, they are 

 in reality only modified sucking-tentacles; in the earliest stages of some 

 of the Spatangoids and Clypeastroids exaniiiied.no difference exists, previous 

 to the format ion of the petals, between the tentacles of the abactinal and 

 actinal part of the poriferous zones. 



Echinarachnius is interesting as showing, even in the oldest specimens 

 examined, traces of a sucking-disk together with a great development of the 

 base of the tentacle, which characterizes the so-called abactinal gills of the 

 CI3 peastroids and Spatangoids. The young tentacles are at first all provided 

 with sucking-disks (but without a terminal rosette); as fast as the petals 

 commence to form, by the spreading of the abactinal part of the poriferous 

 /ones, the pores not yet becoming conjugated, the tentacles assume a conical 

 shape (Pi XXXI. f. <;) ; with advancing age the tips elongate (PL XXXI. 

 f. 4), and finally are capable of considerable expansion and contraction inde- 

 pendently of the fiat basal part of the tentacle I Pi XXXI. J. 5, ? -9). During 

 this growth the terminal portion has constantly been thickening, assuming 

 more or less the appearance of a sucking-disk (Pi XXXI. f. 12). As in the 

 Desmosticha and in all Echini, no matter how widely separated the conju- 

 gated pores of the petaloid portion of the ambulacra may be, the flattened 

 base of the suckers forms but a single tentacle (Pi XXXI. f. 13), with two 



