or THE TEST OE THE TEMNOPLETJRID.E. 347 



will be foimd to be blunt audeveu sliglitly concave; andtliis loss 

 of substance is to accommodate the pit (Plate VIII. figs. 1 &2, a, 6). 

 Opposite to this concavity is the horizontal suture between the as- 

 sociated interradials ; and there is loss of test on their edges close 

 to the commencement of the suture. The margins of the sutures, 

 both vertical and horizontal in direction, are faintly grooved ; and 

 in large specimens there is a distinct shallow pit on the vertical 

 series, between each larger pair of pits (fig. 2, c) . 



The shallow pit is a mere depression ; yet it occupies what 

 might have been solid test, and it diminishes the amount of plate- 

 edge available for junction or suturing with its fellow. 



The line of junction between the interradials and the tentaculi- 

 ferous plates of the ambulacra is marked, not only by a pit at the 

 commencement of the horizontal interradial suture, but also by 

 four, and sometimes five, well- developed and deep pits at the 

 junction of the sutures of the small ambulacral plates. Hence 

 on the surface of the test the interradial area is attached to the 

 ambulacral by a series of very small processes, across which a faint 

 vertical sutnral line can be seen. 



III. The Sutures of Salmacis sulcata. 



This development of the pits appears to relate to the very re- 

 markable sutures of the species. 



The sutures of the interradial plates are four in number. Each 

 plate is joined by its abactinal edge to the plate above ; by its 

 actinal edge to the plate below ; by a lateral median angular 

 suture to the plate at its side ; and by a lateral suture to the 

 plates of the ambulacrum close by. The first two sutures are 

 horizontal, and the others are more or less vertical in direction. 



On carefuUy separating two interradial plates of a vertical 

 series, along the horizontal suture, the tissue connecting their 

 edges (that is to say, the suture) is seen. The abactinal edge of 

 the lower plate is minutely, but very distinctly, marked with from 

 two to four rows (from without inwards in the normal jDOsition 

 of the test) of hollows or sockets. The rows are separate, and 

 extend over the breadth of the edge not quite close to the ends. 

 The sockets are limited, deep, and are special structures in the 

 reticulate calcareous tissue of the plate (Plate VIII. fig. 5). The 

 test is brilliantly w^hite ; and although the sockets can be seen 

 wnth a hand-lens without difficulty, some careful arrangement 

 of the light transmitted by the bull's-eye, or reflected from 



