WARD.J CctLLECTIOXS FKOM THE FORT UNION GKOUP. 547 



raxis bivalvis, Plate XXXI, Fig. 3) was abundant, having spiral stria- 

 tious, as if twisted. This fucoid always exhibited a tendency to split 

 open longitudinally into two equal valves, and many of the segments 

 lay around in halves, the plane of division being always smooth and even 

 and passing directly through the center of the specimen. Only a small 

 collection was made at this point. 



The characteristic fossil of the Cracker Box Creek beds was a species 

 (or two very closely related species) of Viburnum ( V. usiierum, Xewby., 

 Plate LXIV, Figs. -1-9, V. Xewhcrriamim, Plate LXI V, Figs. 10-12, Plate 

 LXV, Figs. 1-3), which, however, differs very much from the abun- 

 dant forms of Clear Creek and does not occur there, nor does the 

 Clear Creek form occur at Cracker Box Creek, although the two locali- 

 ties are only five miles apart and very similarly situated. On the right 

 bank of the valley occurred beds containing Populus leaves, masses of 

 Ta.vodium Europivum, not elsewhere met with, and an abundance of 

 both Equisetum and cane (Aruudof), the latter very large. On the 

 left bank occurred the principal Viburnum bed, and in this a few other 

 plants were found. 



The rock in which the specimens from this locality were embedded is a 

 highly- calcareous marl, sometimes amounting to argillaceous limestone 

 and slightlj' ferruginous. At certain points it is of a dark blue color, 

 sometimes nearly black, and in one fossiliferous bed the outer portion of 

 a small hutte which was cut through by a gulch was cf a red color, like 

 that of Iron Blutf, while the interior was blue or dark. This was of 

 course due to combustion of the carbonaceous matter, the efl'ect of 

 which had not penetrated to the center of tiie butte. This combustion 

 did not affect the character of the iilant impressions, but the unburned 

 portion was much more easily worked and much heavier. In a few of 

 the oxidized buif specimens from this place, the peculiar diagonal mark- 

 ing, so striking at Iron Bluff, apjjears. It seems in these cases to occur 

 on the large gramineous culms. 



The several localities on the Yellowstone River above described were 

 all visited by Dr. C. A. White and his party the year previous, and tiieir 

 stratigr.ai)hical position determined; but, nevertheless, wherever it was 

 possible I observed and collected the molluscan forms, which, however, 

 were very rare. The following shells accompany my collections and 

 have been kindly named for me by Dr. White : 



From lion Bluff: Sphaerium (planum t) ; Physa (Canadensis?). 



From Bunis's Ranch: Acroloxus minutiis. 



From Seven Mile Creek: Ironstone bed: Viviparus (species indeterminable) ; Unio 

 (species indeterminable); scale of a gar. Sparganinm bed : Spb;eriuiu (species inde- 

 terminable). 



From Clear Creek: Physa Canadensis, Whiteaves, ined.: Helix (Patnla) (species 

 uiidescribed). 



From Cracker Box Creek : Viviparus prudeutius, White ; fragments of gasteropoda. 



Very few fossil plants were collected during the journey that was 



