OCULAR AND GENITAL PLATES. 93 



To show the regularity of sequence of incoming of ocular plates in the Centrcchinoida, 

 it may ho put briefly as follows. In the total 48,541 specimens of this order examined and 

 tabulated, G,235 have all the oculars exsert. In 3,300 specimens there is one ocular insert, and 

 of these in 3,279 cases, it is either ocular I or V, that is, 99.36% are correct by rule. Only 21 

 specimens are aberrant, and these have either ocular IV (text-fig. 140, p. 134) or II insert, or in 

 three instances ocular III alone is insert. In 35,184 specimens, two oculars are insert, and of 

 these it is I and V, the bivium, in 34,881, or 99.14%, are correct by rule. In all of the 303 

 exceptions, one of the two plates insert is either I or V. In 97 of these, oculars I, IV are insert, 

 and in 76% of them genitals 4, 5 are fused, as in text-fig. 144, p. 134, mechanically excluding 

 ocular V from the periproct. These may therefore be properly considered cases of an incom- 

 plete I, V, IV insert. In 56 of the exceptions, oculars I, II are insert (text-fig. 141, p. 134). 

 This character was found only in the Echinidae or in the nearly allied Strongylocentrotidae and 

 is a species character in Gymnechinus pulchellus and rohillardi (text-figs. 177-179, p. 105). In 

 82 of the exceptions, oculars V, IV are insert (text-fig. 142, p. 134). This is especially common 

 in the Arbaciidae, and in no case was an ocular excluded bj' the fusion of genitals. In 63 cases, 

 oculars V, II are insert. This combination was seen in the Saleniidae, Echinidae, Strongylocen- 

 trotidae, and Echinometridae, and in all of the cases in the families Echinidae and Strongj'lo- 

 centrotidae genitals 5, 1 were fused, excluding ocular I (text-fig. 146, p. 134), so that this 

 combination may properly be considered an incomplete I, V, II, which character is common 

 in those families. One specimen has oculars I, III, and four have V, III insert, which are 

 sporadic variants. 



In 2,917 specimens there are three oculars insert. Of these in 2,584, or 88.58%, it is oculars 

 I, V, IV, that is, the bivium and left posterior plate of the trivium. Of the 333 exceptions, 

 in 300 cases the order is I, V, II insert (text-fig. 145, p. 134) ; that is, the right posterior plate 

 of the trivium is insert instead of the left posterior, as usual. These may be considered as right- 

 handed specimens. This combination was not seen in the order Cidaroida, or in the families 

 Centrechinidae and Arbaciidae where three plates are frequently insert, but it is a feature of 

 the families Saleniidae, Stomopneustidae, Echinidae, and Strongylocentrotidae. It is espe- 

 cially frequent in Sphaereddnus granulans and Tripneustes esculentus. When three plates are 

 insert in the Centrcchinoida, it is typically the bivium and usually the left, but occasionallj' 

 the right, posterior plate of the trivium, and this rule is adhered to in 98.87% of the 2,917 

 cases observed. There were only 33 exceptions to the above. In 17 of these, oculars V, IV, 

 II are insert, but ocular I was always excluded by the fusion of genitals 5, 1 (text-figs. 148, p. 134 

 and 196, p. 169), so that this combination may reasonably be considered an incomplete I, V, 

 IV, II insert. In one case oculars I, IV, II are insert (text-fig. 147, p. 134), with genitals 4, 5 

 fused, excluding ocular V, evidently also an incomplete I, V, IV, II insert. In addition, four 

 of the cases are I, V, III insert, a sporadic variant, and eleven are cases of V, IV, III insert. 



