390 F. R. Cowper Reed— The Oenns Trinuclens. 



just in front of the glabella owing to its encroachment on the fringe. 

 The pits, which are of subequal size (except in the third in front, 

 Avhich is there smaller), are sunk in somewhat indefinite radial sulci in 

 front ; all the pits become larger laterally and posteriorly, with a loss 

 of the regular radial arrangement by the intercalation of pits in the 

 first row, and are mostly alternate. At the genal expansions many 

 much smaller pits are developed inside the main rows. Thus the 

 fringe seems identical with that of T. elowjatus. 



Smith's var. arcuatus (op. cit., p. 122, pi. viii, figs. 5, 6) is imperfectly 

 known and somewhat unsatisfactorily defined, but it has no close 

 relation with T. fimhriatus of Murchison, as Portlock thought, if it is the 

 form mentioned by him as intermediate between his T. latus and this 

 species. In the so-called var. expanms (Smith MS.) the lower surface 

 of the fringe is unknown, and on the upper plate there are (as iu 

 var. arcuatus) only 2-3 ])its in the radial grooves in front, but the 

 I'adial arrangement and grooves are lost at the sides ; an additional 

 row of small pits is here inserted between the first and second, and in 

 all three rows the pits finally alternate, but the pits do not increase 

 so markedly in size when the rows are traced posteriorly as they do in 

 T. Portlocki, and the pits which are developed on the genal expansions 

 are smaller, closer together, and more numerous. 



The characterization of the fringe of these Tyrone forms is not 

 marked, and the separation of the varieties from one another by this 

 means is not obvious. It appears probable that the}' are not fixed or 

 well established varieties, but merely individual vai'iations of one 

 extremely unstable species, as will be further discussed in the sequel. 



Passing to T. Nicholsoni,^ which has been shown elsewhere to be allied 

 to T. concentricus, we note that the outer band of the inferior plate of 

 the fringe bears two rows of equal radially arranged pits, while the 

 inner concave band behind the well-marked girder bears two rows of 

 pits in front, increasing to three rows at the sides. All these inner 

 rows have their pits arranged in tlie same regular radial fashion and 

 on the same radii as those on the outer band, and are all of equal or 

 subequal size (PI. XIX, Figs. 4, Aa). 



The upper plate has similarly four rows of pits arranged in depressed 

 regular radial lines, but not in distinct sulci, with a fifth inner row 

 inserted at the base of the cheek posteriorly. These pits correspond 

 with those on the Inwer lamella, and their arrangement agrees precisely. 



T. seticornis. — The British forms attributed to this species show 

 several variations of more or less importance in the character of the 

 fringe botli on the upper and lower plates. 



In specimens from llhiwlas the lower plate shows two regular 

 rows of equal-sized pits, not lying in sulci but radially arranged, on 

 the outer band, while on the steeply inclined concave inner band there 

 are 3-4 rows of smaller pits in distinct radial sulci corresponding in 

 number, and these sulci have corresponding ones on the upper plate, 

 but they are there less pronounced and they hold five pits in front of 

 the glabella and 6-7 at the sides. 



In specimens from Applethwaite (Coniston Limestone) the two 



^ Reed, Geol. Mag., Dec. V, Vol. YII, p. 212, PL XVI, 1910. 



