SIPHONOTRETID^. 619 



The opening of the foramen in the interior of the valve is on the low median ridge at a point 

 three-fifths of the distance from the area to the anterior end of the median line of the splanch- 

 noccsle (visceral area). 



Dorsal valve unknown. 



Oiservations. — This species is known only by the type specimens of Matthew. These 

 include a matrix of the ventral valve, represented by Plate LXXXIV, figure 5b, a cast of the 

 interior of a broken ventral valve, figure 5, a fragment of the exterior of a ventral valve showing 

 the foraminal furrow, and a fragment of the cast of the outer surface. As stated under the 

 description of the genus, Matthew [1893a, p. 276] mistook the matrix of the ventral valve for 

 the interior of a ventral valve, and the cast of the interior of a ventral valve for a dorsal 

 valve. The true relations of the specimens are readily seen by comparing them with the speci- 

 mens of closely allied species on Plate LXXXIII." 



The type material was collected from a layer of rock that is identical with, or is nearly at 

 the same horizon as, a layer at the same locality carrying numerous specimens of Trematoholus 

 pristinus {=Protosi'phon pristinus Matthew). I would consider the latter a synonym were it 

 not for the position and size of the vascular sinuses in the ventral valve, which are located well 

 out toward the oxiter margin (compare PI. LXXXIV, fig. 5, with PI. LXXXIII, fig. 2c), while 

 in T. pristinus they are much nearer the median line of the valve. The interior of the ventral 

 valve of T. insignis is more like that of T. Tcempanum (Matthew). (Compare PI. LXXXIV, 

 fig. 5, with PI. LXXXIII, figs. Ic, li.) Until a larger series of specimens representing T. insignis 

 is collected the diagnosis of the genus must depend largely upon the material representing T. 

 Tcempanum. 



Formation and locality. — Middle Cambrian: (301 [Matthew, 1895a, p. 123]) Sandstones of Division lb2 of 

 Matthew, on Hanford Brook, St. John County, New Brunswick, Canada. 



Trematobolus kbmpanum (Matthew). 

 Plate LXXXIII, figures 1, la-o. 



Protosiphon hempanum Matthew, 1897, Geol. Mag., new ser., dec. 4, vol. 4, pp. 70-71, figs. 1-4, p. 70. (Described 



and discussed as a new species.) 

 Protosiphon hempanum Matthew, 1898, Trans. Roy. Soc. Canada for 1898, 2d ser., vol. 4, sec. 4, No. 2, p. 131, PI. I, 



figs. la-f. (Text is almost an exact copy of the preceding reference. Figs, la-d are copied from figs. 2, 1, 3, 



and 4, respectively, of the preceding reference.) 



General form subcircular to rounded ovate, biconvex, with the dorsal valve about one- 

 third more convex than the ventral; outline of ventral valve obtusely acuminate and dorsal 

 valve broadly rounded posteriorly. Beak of ventral valve marginal above a well-defined area; 

 just in front of the beak a narrow furrow starts that marks the path of advance of the pedicle 

 opening; this furrow is usually from one-fourth to one-third the length of the shell and grad- 

 ually widens and deepens up to the oval pedicle opening; strong lines of growth, with the 

 curvature of the posterior side of the opening, mark the stages of growth and advance of the 

 opening; these are beautifully shown by the cast of the furrow and opening illustrated by 

 Plate LXXXIII, figure lo. The furrow and pedicle opening are entirely in the layers of the 

 shell and do not cause a projection on the interior of the valve; the openmg is compressed 

 in some specimens, but it is usually of good size, as is shown by a cast of it that extended from 

 the outer to the inner surface; this is represented by figure le. The opening is of the same 

 character as that of Scliizamhon, except that it is through a thick shell. 



Dorsal valve proportionally a little shorter than the ventral; beak marginal. 



Surface marked by a few concentric lines of growth; the inner surface appears to have 

 been smooth. The specimens are preserved in a fine-grained sandstone, and usually in the 

 form of casts; a few show portions of tlie shell, and these indicate that it was of a calcareo- 

 corneous nature and built up of several rather strong layers or lamellae. This is also indicated 

 by the cast of the pedicle furrow and opening (fig. lo). 



a I sincerely regret being obliged to differ so greatly with my Mend Doctor Matthew on the interpretation of the types of this species and of 

 the genus Trematobolus. 



