SOME NEW DEVONIC FOSSILS 169 



convex and very long, terminating posteriorly in a more elevated 

 triangular area. The prevailing aspect of the cranidium and gla- 

 bellar lobes is that of narrowness and length, particularly in the 

 distance between the nuchal furrow and the frontal lobe. The 

 nuchal furrow is broad and low and the occipital ring broad, flat 

 and arched. 



Spines. .While the general surface is tubercled and some of the 

 tubercles become developed into short spines the major spines are 

 as follows: three pairs, one in front of another on the crest of the 

 glabella; these are of great size and strength and deeply curved 

 backward. They seem to be all of about the same length. In some 

 of our specimens the posterior pair curves backward in a long arch 

 to and beyond the posterior margin of the head, but in a younger 

 specimen they are shorter. The middle and anterior pairs are quite 

 as long. In section these great frontal spines or hooks are not 

 circular but somewhat flattened on the opposing faces though 

 rounder on the outer surface and narrow fore and aft. 



On each lateral lobe where widest, and just above the dorsal fur- 

 rows, is a spine of less hight than the foregoing and apparently 

 erect and these are flanked in front by a much shorter pair. These 

 five pairs seem to be all there are on the glabella except for the 

 spinous tubercles in the occipital area. 



The occipital ring bears at its edge on the axis a series of long 

 curved flat or vertically compressed spines, one at the middle and 

 one diverging from the axial spine, at each side. These are neither 

 as long nor as large as those of the frontal lobe but they must have 

 reached back over several of the thoracic segments. The occipital 

 ring is deeply contracted at the dorsal furrows and where it expands 

 again beneath the cheeks it extends out on each side, into a flat 

 but straight and slender spine larger than the others. This makes 

 five spines on the neck ring, 15 in all on the cranidium, seven pairs 

 and one axial. It would be natural to expect others on the palpe- 

 bral lobe but these seem to be wanting. 



The other parts of the species are represented by portions of 

 free cheeks which indicate that these ran out into short, thick and 

 narrow genal extensions with a row of rather small spines along 

 the occipital margin, while just outside of the eye near the margin 

 there was a very large, long and recurved hook like those of the 

 frontal lobe. There is still some uncertainty as to the exact details 



