WATER SYSTEM. 701 



When taken in connection with other structural features, the general aspect 

 of the plates and spicules of the ambulacral tubes may be of some value in 

 tracing the affinities of genera ; but the mode of formation of all the calca- 

 reous deposits of Echinoderms, as large open limestone cells, is subject to 

 such remarkable variations in the same species, that characters drawn from 

 the spicules of the ambulacral tubes of the ovaries and of the alimentary 

 canal must be used with great caution, as the merest inspection of the fig- 

 ures of the terminal plates of a few of the leading genera will readily 

 show (PI. XXX VIII.). The variations are still more striking if we com- 

 pare the spicules from the shafts of the tubes (PI XXXVIII. f. 4-6, 7-9, 

 16, 17, 30, 31, 34, 35) of a few of the species figured. The same extreme varia- 

 tions occur in the spicules from the ovaries and from the alimentary canal of 

 the same species (PI. XXXVIII. f. 2 a ' c , U a '\ 12 a - k ). Valentin, and subse- 

 quently Stewart and Herepath, have called attention to the systematic value 

 of the spicules. An additional number of spicides from the ambulacral 

 tubes have been figured by Perrier in his Memoir on the Pedicellarise, and 

 the probable passage of the terminal disk of the Desmosticha into the spic- 

 ules of the shaft of the tube intimated for one family (the Diadematidae). 

 He has also, as Midler had previously shown so well, called attention, as a 

 new feature, to the absence of a terminal calcareous disk in Petalosticha. 



In the Desmosticha the ambulacral pores pierce the ambulacral plates ; 

 the same is the case also in the Clypeastroids and Petalosticha, only in the 

 petaloid portion of the ambulacra they are placed in the sutures between 

 adjoining plates, and towards the extremity of the petals the pores gradually 

 pass from the sutures to the central part of the plates. In the Desmosticha 

 and Petalosticha we never find more than one double pore for each ambula- 

 cral plate ; this is by no means the case in the Clypeastroids, in which, as 

 was first shown by Midler, the number of ambulacral pores increases with 

 age as fast as the ambulacral plate increases in size. The phyllodes of the 

 Cassidulidae form no exception to this rule among the Petalosticha, for Mid- 

 ler has plainly pointed out that it is owing to the unequal development of 

 the ambulacral plates, — two small ones and a large one alternating, — that 

 the phyllodes are developed. It is the connection of the primordial ambu 

 lacral plates in series of three or more plates (secondary ambidacral plates, 

 but originally all of the same value), which has served to a greater or less 

 extent to distinguish genera among the Desmosticha. See analysis of the 

 plates of the ambulacral system (PL VI). Among the Clypeastroids and 



