300 Canadian Record of Science. 



Both dorsal and ventral shields are so much like those of 

 P. Grouchi tliat if these Nova Scotian fossils liad been 

 found in West England we should have referred them to 

 the latter species. Perhaps the rostral plate may prove to 

 distinguish your form when it is completely known. One 

 piece of dorsal shield, in counterpart, shows the impres- 

 sions of the supposed branchial pouches on one side. 



" The pointed fragments may be Cephalaspidian cornua, 

 but are imcertain. 



" There is the typical Onchus Murchisoni, Ag. 



" Most interesting is one small fragment of Psanwiosteus, 

 with ornament identical with that of F. anglicus. (See 

 Traquair, Ann. Mag. Nat. Hist., Ser. 7, Vol. 11, 1898, p. 67, 

 pi. i., figs. 1, 2.) In this fossil the chambers of the middle 

 layer are larger than in our unique plate. 



" On the whole, I should place the McArras Brook beds 

 on the same horizon as the Lower Old Eed Cornstones of 

 the Hereford District of England, above the passage beds." 



In order to give a comprehensive view of the succession 

 of strata in this Knoydart formation, the following section, 

 carefully measured and prepared by Mr. Hugh Fletcher 

 of the G-eological Survey of Canada, in the year 1897, is 

 here given : — 



" From the mass of trap near the mouth of McAra Brook 

 the following is the section in ascending order : 



Amygdaloidal trap, probably Lower Carboniforous as 



described in lieport P. for 1886. 

 Measures concealed. On the left bank of 

 the brook trap is in the cliff", while on the 

 right bank there are indications of red ft. in. 

 stratified Devonian rocks oO 



1. Eed, argillaceous shale, more or less slaty, 



with coherent underclay full of rootlets. 



Dip 230732" (Magnetic) :! 



2. Eed, argillaceous slaty rock, not well seen . . 4 9 



