MELONECHINUS. 37fl 



ambulacral plates which typically occupy this position in all regular Kchini ip. sJ: text -f 

 40-51, p. 80). The existence of two plates in an area on the peristome adorally implies that 

 Melonechinus in an early stage had only two columns of ambulacral plates in an area (repre- 

 sented by this one row of two plates), as in the young, probably, of all other regular Ki-liini 

 (Plate 3, fig. 7). Passing aborally, the ambulacral plates of the pcristome <-on-i>t of many row- 

 in each area; also the number of plates in each row increases to the periphery nf the area, where 

 there are more plates in a row than in the basicoronal row of the corona proper. In the inter- 

 radial areas of the peristome there are some non-ambulacral plate-, three in area A, and two 

 or more in areas E and I. In the reconstruction (Plate 56, fig. 8) I have assumed that three is 

 the typical number and have shown such in the two areas figured. The number is not of 

 apparent significance, but it is important that this peristome representing the family has both 

 ambulacral and non-ambulacral plates, and that the ambulacral consist of two plates in a row 

 adorally, passing to more plates in a row aborally. The Archaeocidaridae is the only other 

 family of Perischoechinoida that has both ambulacral and non-ambulacral plates on the peri- 

 stome, and in this feature both families make an approach to the character -ecu in the Cida- 

 roida (text-figs. 46-48, p. 80; pp. 82, 85). 



The apical disc is small in Melonechinus multiporus. In six specimens quite free from 

 matrix and with this structure in place the apical discs measured proportionately from 13 to 

 16% of the diameter of the tests (pp. 362, 367). The apical disc has all ten ocular and genital 

 plates preserved, or part of them, in numerous specimens (Plate 54, fig. 5; Plate 55, figs. '_'. 3; 

 Plate 56, fig. 6; Plate 57, fig. 2; Plate 60, figs. 1, 2; text-fig. 163, p. 149). The ocular- are 

 small and imperforate, and separate the genitals, meeting the periproct with the whole width 

 of each plate and ventrally cover the ambulacra and laterally the interambulacra in part on 

 each side, as best seen in Plate 56, fig. 6. The genital plates are wide and high, and have 

 typically three or four pores each (Plate 57, fig. 2). In one case (Plate 60, fig. 2, area G) a 

 genital has only two pores and apparently no pores have been oblitered. Not infrequently 

 one or more pores are obliterated so that the number existent has to be judged somewhat 

 from the position of those found; also pittings of erosion are sometimes confusing to distingui>h 

 from actual pores. Dr. Keyes figured a specimen with a single genital pore in a plate which 

 also shows numerous madreporic pores. I have not seen one genital pore only, nor madreporic 

 poreS in this species or any other species of the genus (p. 363). The periproctal plates are 

 unknown in this species, which is peculiar, as so many excellent specimens are known. The 

 structure, however, was doubtless similar to that of M. stewartii (Plate 56, fig. 1). 



The lantern for the genus is known only in this species, and only partially here (Plate '^. 

 figs. 9, 10). I can add little to what was previously shown by Meek and Wort hen. The 

 lantern is inclined, pyramids wide-angled, foramen magnum moderately deep, and the teeth 

 extend ventrally a considerable distance beyond the tips of the half-pyramids (p. 363). 



