WHITEAVES. ] DEVONIAN FOSSILS OP MANITOBA, ETC. 261 



to that of the plate of which they form a part, and project coiisideralJy 

 beyond its margin, but in two of the rays the genera] direction is \'citioal, 

 and in the other two horizontal or lateral. All the rays live thickened at 

 their junction with each other, striated longitudinally, and tapei- gra- 

 dually to an acute point. The horizontal or lateral rays are longer than 

 the vertical, and the former ai'e cvir\ed slightly inward and downward. 

 The distal rays, or those vertical rays which are directed forward, incline 

 slightly inward, and the proximal rays, or those vertical rays which are 

 directed backward, incline as slightly outward. 



The rays of the ' spicules ' of immediately adjacent plates interlock in a 

 very regular manner, as shown by Dr. Schluter in his figure of S. meya- 

 ra2]/i j.s* and as specially indicated in iig. 1 a, of plate xxxiii, in which the 

 dotted lines represent the relative position of the hexagonal plates. The 

 distal ray of each ' spicule ' passes behind or inside of the proximal ray of 

 the ' spicule ' immediately above it. The lateral rays of each ' spicule ' 

 pass between the distal and proximal rays of the ' spicules ' lying to the 

 right and left of it, while the right lateral ray of each ' spicule ' passes 

 above the left lateral ray of the corresponding ' spicule ' in the second row 

 to the right of it. 



In the narrow portion next to the base the ' spicules ' appear to 1 le par- 

 tially amalgamated and are less clearly defined, so that this part of the 

 fossil often presents a longitudinally ribbed appearance, the ril)s being 

 rounded oi' flattened and convergent posteriorly. 



At the summit the distal rays of the last two or three ' spicules ' in 

 each longitudinal row are prolonged and convergent and ultimately meet 

 together at its apex. These prolonged summit rays are hollow and flat- 

 tened laterally and the central summit area formed by them is about four- 

 teen millimetres broad at the base, in the only specimen (tig. 1.) in which 

 it is preserved. 



In attempting to free the organism from the matrix, the rock in these 

 specimens from Dawson Bay often carries with it the outer ci ivering of 

 hexagonal plates. The result of the fracture of these plates frofn the 

 ' spicules ' of which they formed a part, is shown in tig. 1 a. When the 

 rays of the ' spicules ' are not present, which is often the case, their 

 original shape and position are indicated by corresponding moulds of their 

 exterior in the matrix. The spaces between the spicules, also, are repre- 

 sented by raised ridges crossing each other at right angles and directed 

 diagonally across the specimen, thus giving the peculiar reticulated appear- 

 ance shewn in tigs. 1 and 4. 



The maximum breadth of the largest specimen collected, which is too 

 imperfect to show the height, is fifty six millimetres. The most perfect 



'' Zeitschr. der Deutsch. geol. gessellsch. , 1887, vol. XXXIX, pi. i, fig. <>. 



