MONOGRAPH OF THE EXISTING CRINOIDS. 285 



Totcometra paupera., E njthrovietra australis, Psathyrometra antarctica, Florametra 

 asperrima, and Florometra laodke; or (2) it may broaden into a plate, either 

 directly after its first appearance, or after it has taken on more or less of the rod- 

 like character. 



An intermediate type is seen in such forms as Florometra serratissiTna, Floro- 

 Tiietra asperrima, Hypalovietra dejecta, Hathrometra prolixa, and THchonietra 

 cubensis, wliere more calcareous matter is deposited on the distal end of the plate 

 than elsewhere, and this is largely laid down in a plane parallel to the surface, so 

 that a broadened and flattened end, often with large incisions or pierced with one 

 or several holes, results. 



If this additional deposition is extensive and all in one plane, a fan-shaped 

 form, as seen in Sarametra tmerhdis, derived through the type found in Penta- 

 metrocrinus vanans, results; and if both ends are similarly modified, an oval form, 

 as seen in Zenometra columnaris. 



After a considerable growth the spicule may fork or branch, so that a large 

 and stout triradiate or branched spicule results; spicules of this type are usually, 

 perhaps always, intermixed with others of simpler form; they are seen well illus- 

 trated in Neocomatella pulchella and in Nemaater rubiginosa. 



If the branching first occurs when the spicule is smaller, greater complexity is 

 found and one or two complete meshes may be formed, as in Amphivietra discoidea 

 and in Heterometra bengalensis; and if it occurs when the spicule is very small a 

 more or less definite plate with indefinite borders results, as in Pontiometra an- 

 dersoni and in Tropiometra 7nacrodkcus. 



But instead of first undergoing more or less enlargement as a spicule, the 

 original spicule may almost from the start take the course of the spicules which 

 form the bases of the first plates to appear in the larvae; that is to say, it may almost 

 immediately develop into a fine netAvork, spreading out over the perisomic regions 

 and entirely inclosing them; the cells in this network, however, are never so small 

 as those in the primary plates. 



When this occurs the relative distinctness and mobility of the outer portion of 

 the lappets will determine whether there is a single series of large plates, as in 

 Promachocrinus kerguehnsis, or in Comatilia. iridometriformis, or whether there 

 are distinct side and covering plates, as in the species of Calometridse, Thalasso- 

 metridse, and Charitometridse. 



Though the side plates in the Calometridse, Thalassometridse, and Charito- 

 metridse show great uniformity, and their structure is usually very complicated, 

 there is not the slightest reason for supposing that they are phylogenetically con- 

 nected with the adambulacral plates of any other groups, or that they are anything 

 else than the result of the secondary calcification of the perisome, which has been in- 

 fluenced by purely local factors. 



Each side plate is the external covering of a definite unit, composed of the 

 inner portion of an ambulacral lappet and the associated tentacles and sacculus. 



In their most simple form the side plates in these families are low and approxi- 

 mately oblong in shape, M'ith rounded corners, as seen in Asterometra macropoda 



