286 



length of each cardinal spine a little more than one half of the greatest 

 breadth of either valve without the spines. Ventral valve regularly con- 

 vex from beak to front, though the nasute forms are most prominent 

 anteriorly along the median line ; umbonal region compressed ; beak 

 small and raised very little above the general level of the hinge line ; area 

 transversely elongated and very narrow in the direction of its height, 

 with a small equilateral foramen in the centre. Dorsal valve concave, 

 with a perfectly straight cardinal margin, an extremely minute beak and 

 a hinge area much narrower than that of the ventral. 



" Surface marked by numerous, but comparatively distant and, for the 

 most part simple, radiating raised lines, which increase by intercalation 

 and alternate at unequal distances with from one to five (or perhaps more) 

 shorter and much smaller ones, the whole being crossed by extremely 

 minute and close set concentric striations, and bj a few more or less dis- 

 tant lines of growth. Characters of the interior unknown. 



" Collected at several localities on the northern portion of the east shore 

 of Lake Winnipegosis, in the district of Saskatchewan and in the adjacent 

 part of the Province of Manitoba by Mr. J. B. Tyrrell in 1889, (but 

 previously found loose in this vicinity by Mr. D. B. Dowling in 1888,) 

 also on the shores and islands of Cedar Lake, and on the Saskatchewan 

 below Cedar Lake, by Mr. J. B. Tyrrell in 1890. At each of these local- 

 ities it is apparently abundant and often associated with Isochilina 

 grandis, var. latimarginala, Jones. 



" The specimens consist either of natural moulds of the exterior of the 

 shell or of casts of the interior, in a compact fine grained dolomite, and 

 in no case is there any vestige of the actual test remaining. In several 

 of these natural moulds, however, the minutest details of the surface orna- 

 mentation are well preserved, and it is from wax impressions made from 

 two of these moulds that the figures " on Plate 24 were drawn. 



"The species is apparently most nearly allied to the Strophomena Leda 

 of Billings,* from division 3 of the Anticosti group of the Island of Anti- 

 coiti (which Mr. Billings correlates with the Llandovery of England and 

 with the Clinton of the State of New York), but seems to differ 

 therefrom in its much larger size, and in the greater proportionate length 

 of its cardinal spines. Both it and S. Leda are evidently what Professor 

 H. S. Williams t would call ' geological mutations ' of the ' race which 

 began in Strophomena alternata in the Trenton stage,' but they form a 

 marked exception to his statement that in the American race of the S. alter- 

 nata type, the slender mucronate points at the terminations of the hinge 

 line first appear in the Tully limestone." 



*Geological Survey of Canada, Palaeozoic Fossils, vol. I, 1865, p. 120, figs. 98 and 99. 

 tSee his paper on "The Cuboides Zone and its Fauna," in Bull. Geol. Soc. America, 

 publisheOMay, 1890. 



