Septembeb 20, 1907] 



SCIENCE 



381 



which forms Thousand Creek Hill, and pre- 

 sents a steep escarpment to the east. 



The thickness of the Virgin Valley beds is 

 evidently between one thousand and two thou- 

 sand feet. The larger part of the formation 

 is composed of volcanic ash or tuff showing 

 a variable amount of induration. Character- 

 istic bad land structure has been developed in 

 many places. The formation may be divided 

 somewhat arbitrarily into upper, middle and 

 lower divisions. The lower beds are somewhat 

 harder than the others, and where bad land 

 structure occurs in them very steep faces are 

 frequently produced. These beds show strong 

 contrasts of coloration, varying from white to 

 green or bright red. The middle beds are 

 generally brownish or gray, and weather in 

 gently rounded knolls. The upper beds are 

 usually softer than the others and consist of 

 cream-colored ash. 



Maramal remains are quite common in por- 

 tions of the upper beds, and at a horizon which 

 is apparently in the lower division, though not 

 in its lowest portion. Only a very few frag- 

 ments of bones were found on the middle divi- 

 sion, but plant remains are very abundant at 

 this horizon. Large logs of beautifully petri- 

 fied wood are present in abundance, and near 

 the middle of this division stems and leaves 

 have accumulated in sufficient quantity to 

 form a thin lignitic deposit. 



Judging from the character and occurrence 

 of the fossil remains, large portions of the 

 upper and lower divisions of the formation 

 are of ajolian origin, or have formed in shal- 

 low, shifting lakes comparable to existing 

 lakes only a few feet in depth, such as are 

 not uncommon in the eastern Oregon region 

 at the present time. A part of the middle 

 division, particularly that portion containing 

 the lignitic deposit, has evidently formed in 

 or about a body of water. 



The mammalian remains obtained from the 

 lower horizon include a type near Chalico- 

 iherium or Moroyus, two types of horses, two 

 or more cameloid types, a canid and a num- 

 ber of other forms. Chalicotherium is repre- 

 sented by a number of characteristic bones, 

 including the peculiar phalangeal elements. 



Of the horses one is a form with short-crowned 

 molar teeth resembling those of the Miocene 

 Eypohippus. The other has short-hypsodont 

 molars and is evidently in, or close to, the 

 Miocene genus Merychippus. Of the camels 

 one species is represented by limb bones indi- 

 cating an animal of considerable size, and 

 evidently of the Procamelus type. The canid 

 is represented by a second lower molar appar- 

 ently differing from any type thus far de- 

 scribed from the Tertiaries of the Pacific 

 Coast region. 



The fauna of the upper beds includes a 

 mastodon, a horse, two camels, a large cat 

 and fragmentary remains probably represent- 

 ing a rhinoceros. The mastodon is a species 

 of considerable size, and is evidently not older 

 ■than the stage of the Mascall Miocene of the 

 John Day region. One of the camels is rep- 

 resented by a metapodial about as large as 

 that of Alticamelus of the lower Loup Fork. 

 The cat is known only from a second phalanx 

 indicating a large form with powerful ex- 

 tremities. 



Such paleontological material as is present 

 indicates that the upper division of the Virgin 

 Valley beds is probably of Miocene age and 

 not older than the stage of the Mascall Mio- 

 cene of the John Day region. The fauna of 

 the lower horizon, so far as known, is also 

 evidently Miocene, though it may represent a 

 slightly different phase. 



The fauna of the Virgin VaUey beds, par- 

 ticularly of the upper division, appears in a 

 general way to represent the same period as 

 the fauna of the ash and tuff formations in 

 Nevada, which have commonly been recog- 

 nized as corresponding to the Truckee Mio- 

 cene. If such is actually the case, there will 

 apparently be little reason for correlation of 

 the Truckee with the John Day, as has com- 

 monly been done, since the Virgin Valley beds 

 are near the age of the Mascall beds of Oregon, 

 and the John Day is separated from the Mas- 

 call Miocene by the Colimabia Lava formation, 

 as also by a considerable unconformity below 

 the Columbia Lava. 



As far as I am aware, the beds of the 

 Virgin Valley region offer more material 



