52 THE CYSTIDEA 



(Br) which may attain 2/3 the length of the theca, without signs of a 

 joint. The spine is subcircular in section, and has no groove and no 

 accessory plates ; none the less it may have served as an arm, i.e. as the 

 bearer of a tentaculate extension of the water-system, and of a ciliated 

 path to the mouth. The mouth, anus, and hydropore were probably 

 situated in the integument uniting the two sides of the thecal opening 

 that seems to stretch between the spines. Haeckel (1896) supposes that 

 the Apus-foTra. of the Anomalocystidae was correlated with locomotion, 

 and that the stem was a locomotor organ a suggestion by no means far- 

 fetched ; but his statement that the anus was ad-columnal must have 

 been based on Pleurocystis (p. 64), which, though belonging to a different 

 order, simulates Placocystis in many features. 



ORDER 2. Rhombifera, Zittel (1879, emend.) 



Cystidea in which radial symmetry affects the food-grooves, 

 and, in the more advanced families, the thecal plates ; probably also 

 the nerves and ambulacral vessels, but not the gonads. The food- 

 grooves are exothecal, i.e. are stretched out from the theca on jointed 

 skeletal processes (brachioles). These either are close to the mouth 

 or are removed from it upon a series of ambulacral or subambulacral 

 plates not derived immediately from thecal plates, or are separated 

 from the oral centre by hypothecal passages passing beneath teg- 

 minal plates. The stereom and stroma become arranged in folds and 

 strands at right angles to the sutures of the thecal plates ; in higher 

 forms the stereom-folds are in part specialised as pectini-rhombs. 



The chief reason for the establishment of this order is the 

 recognition of a distinct line of development in the skeletal 

 structures bearing food-grooves. That this represents a true phylo- 

 genetic series is confirmed by the structure of the test, although 

 there may be some indefiniteness in this respect shown by the 

 earlier genera. The difficulty of classifying forms at the parting of 

 the ways is not one to be lessened by advance of knowledge. We 

 note also the gradual decrease in number of thecal plates, their 

 increasing subjection to a radial symmetry, and greater develop- 

 ment of a stem. Hence, as is to be expected in a natural classifica- 

 tion, there is a far greater difference apparent between extremes 

 in the same series, say, the Callocystinae and the Echinosphaeridae, 

 than between the latter and the Diploporite family Sphaeronidae, 

 which constitute initial forms of different series. The radial 

 symmetry or actinism of the order is trimerous, pentamerous, or 

 hexamerous, but may undergo secondary modification through 

 atrophy of a ray (e.g. Comarocystis, Lepadocrinus). 



FAMILY 1. ECHINOSPHAERIDAE. Rhombifera in which the thecal plates 

 are numerous and indefinitely arranged. So-called pore-rhombs are de- 

 veloped, but no pectini-rhombs. Brachioles confined to neighbourhood 

 of mouth, unbranched. Stem, when present, not composed of a single series 



