INVERTEBRATE PALAEONTOLOGY. 481 



outer side of the whorls, though our specimen is so worn that they arc 

 nearly obliterated. 



Locality ami position. — Near the forks of Cheyenne River, Dakota ; in 

 the upper part of the Fort Pierre group of the Upper Missouri Cretaceous 

 series. 



Heleroceras tni-tiini, M & H 



Plate 22, figs. 4, a, b, c. 



FTclicoceras tortum, Meok and Hayden (1858), Proceed. Acad. Nat. Sci. Pbilad., X, 54. 

 Hctcroceras tortum, Meek (1864), Sinithsoniau Check-List N. Am. Crot. Fossils, 25. 



Of this species, we have one entire septate volution, which evidently 

 belonged to a sinistral spiral shell, the whorls of which are rounded, and 

 increase so as nearly to double their diameter each turn. They are coiled 

 in such a manner as to be disconnected by a free space, equaling from one- 

 third to one-half the diameter of each succeeding whorl below ; while the 

 umbilical cavity left within the coil is less than the diameter of the largest 

 volution. The surface is ornamented by two rows of rather depressed nodes, 

 passing around below the middle of the outer side, and small annular costse, 

 which sometimes bifurcate at the nodes. 



The siphuncle is small, and presents the remarkable peculiarity of gradu- 

 ally changing its position in passing from the smaller to the larger extremity 

 of the fragment studied; that is to say, that at the smaller end of the spe- 

 cimen it occupies exactly the middle of the outer side, but, in passing around, 

 it gradually curves upward, so that, by the time it reaches the larger extrem- 

 ity, it comes out on the summit of the whorl. It is also worthy of note that 

 the lobes and sinuses of the septa, as might be expected, follow this curve of the 

 siphuncle, so that it would seem that this whorl not only forms an ascending 

 spiral curve, but is also apparently twisted upon an imaginary axis within 

 itself; this, however, was almost certainly not the case throughout the entire 

 length of the shell, but probably occurred only in this particular whorl, pre- 

 paratory, as it were, to the peculiar deflection of the succeeding, or non- 

 septate portion. 



The septa are not very closely crowded, though rather complex ; the 

 lobes and sinuses (which scarcely differ in size and form on opposite sides of 

 the siphuncle) being variously branched and subdivided. The siphonal lobe 

 is comparatively small, oblong, about one-third longer than wide, and orna- 

 mented at the extremity by two nearly equal branches, each of which has 

 two or three small, more or less digitate, subdivisions. The first lateral sinus 

 (il u 



