448 PEOF. p. ir. DF^^CAN ON- THE STPvUCTUEE OP 



adoral and inner shoulder of the crenulation on the boss, and to pass 

 obliquely actinally and towards the median line of the ambulacrum. 



The line forms an arch, the chord of which is the transverse 

 suture at the actinal edge of the compound plate. The other line 

 starts from the abactinal and inner shoulder of the crenulation, 

 and passes obliquely abactinally and towards the median line. 



Both of these well-marked lines are sutural, are between certain 

 plates, and they both reach the median or vertical suture of the 

 compound plates. The adorally situated line is the aboral limit 

 of a primary plate, just as the aboral line is the limit of the 

 aboral primary plate of the combination. A line can be traced from 

 the first (the most abactinal) peripodium of the compound plate to 

 the part of the crenulation at the aboral and outer shoulder of the 

 boss ; and there is little doubt of the continuity along the abactinal 

 edge of the mamelon of the whole actinal line. The last peripodium 

 but one of the compound plate, counting from the abactinal peri- 

 podium, has a line passing from the adoral pore obliquely inwards 

 and towards the nearest shoulder of the boss, and this corresponds 

 with the aboral edge of the actinal primary plate of the combina- 

 tion. Hence there is an adoral primary with a bent or curved edge, 

 and there is also a similar actinal primary. There is also a middle 

 primary, the direction of the arching of the plates being opposite. 



In the compound plates with six peripodia there is an aboral 

 primary, an adoral primary, and there are also lines of suture 

 passing to the base of the mamelon from the 2nd, 3rd, & 4th perip^odia 

 (besides those from 1 & 5). But, as in the instances of the com- 

 pound plate with five peripodia, there are no signs of sutural lines 

 passing down the inner flank of the boss corresponding to the 2nd, 

 3rd, & 4th lines on the outer or interradial side. 



There are, then, three demi-plates to the compound plates with 

 six peripodia, and therefore there is a third primary plate, as there 

 must also be in the combination, with five peripodia. 



The position of the third primary is readily made out on the part 

 of the tubercle-bearing plate near the median line, for it must relate 

 to the expansion that exists between the inner ends of the aboral and 

 adoral plates. But its position near the peripodia is a matter of 

 doubt, in consequence of the homogeneous condition of the mamelon. 

 The direction of the part of the plates close to the peripodia is, 

 however, a somewhat correct guide where to look for the primary 

 between the others. Certainly the first pair of pores is in relation 

 to the aboral primary, and it is clear that the second pair is in a demi- 

 plate. In most plates the narrow part between the demi-plate just 

 noticed and the next line in adoral succession, passes either directly 

 transversely or slightly actinally and towards the boss, and it is this 

 direction which opens out two possibilities regarding the position of 

 the middle primary. 



In the compound plates with five component plates, the peripodia 

 (fig. 23, nos. 1, 2, 3, 4) have sutural lines represented by grooves 

 passing from their adoral pores to the mamelon, but not so the fifth. 

 Of these lines, nos. 2 & 3 pass up the boss to the crenulated groove 



