EARLIER TERTIARY (EOCENE AND OLIGOCENE). 779 



quartz sands have not been observed. Carbonaceous materials are generally; present as frag- 

 ments of plants, as vegetal ooze in greater or less proportion to the other constituents, and as 

 distinct coal beds. Carbonate of iron is frequently an integral constituent of the rocks. 



In color they are, when fresh, generally bluish gray, shading to brownish black. They 

 weather to buff tints, which are usually dull. The coarser and more massive varieties' form beds 

 20 to 100 feet thick, in which bedding planes are not distinguishable. The finer deposits are 

 thinly laminated and carry abundant leaf impressions, which occasionally interlap with one 

 another so as to form a mass of leaf fraigments. * * * 



In many of the sandstones silvery-white mica has developed as a secondary mineral, but 

 they exhibit no other indication of metamorphism. The coals, on the contrary, being chemically 

 more sensitive, have undergone metamorphism to a greater or less extent through loss of 

 combined water and concentration of fixed carbon. * * * 



The stratigraphic relations of the Puget series are not determinable within the area under 

 discussion, since the strata nowhere come in contact with older sedimentary rocks. Sixty 

 miles northward, on the Skagit River, is a contact between similar coal-bearing strata and 

 older metamorphic schists, described in an earlier report as possibly a surface of deposition 

 or of faulting. Examination of this locality in 1895 led to the discovery of small pebbles of 

 the schist forming a basal conglomerate in the sandstone beds next the contact, which was 

 therefore a surface of deposition during a transgression by the sea. Fossils found in limestone 

 under the schists are, stems of crinoids of Carboniferous or Triassic age, whereas the coal-bearing 

 sandstones of this locality are assigned by Knowlton to the Eocene on the evidence of numerous 

 leaf impressions. 



The age of the Puget formation has been in doubt because of the obscurity of strati- 

 graphic relations, the general absence of marine fauna, and the indeterminate character of the 

 flora. * * * 



A preliminary examination of the fossil plants enabled Knowlton to report that the lower 

 beds of the series are Eocene, whereas the upper beds may be of Miocene age. The floras from 

 horizons several thousand feet apart in stratigraphic range are so distinct as to afford means of 

 correlating separate strata of the Puget formation. * * * 



The measured sections of the Puget group exhibit total thicknesses of 5,800 feet on Green 

 River, 5,500 feet on South Prairie Creek, and 5,480 feet in Carbon River Canyon. None of 

 these measures is complete. In each instance the lowest stratum is of the Puget group outcrop- 

 ping on an anticline, and the highest is the limit of exposure where the rocks pass under later 

 formations. These sections overlap, and there are also higher beds exposed on South Prairie 

 Creek above the hmit of the measured sections. These considerations justify the inference that 

 the thickness of the Puget group may probably be 9,000 feet or more. 



C. A. White/^" who gave the name Puget to the strata described by Willis, 

 quotes Newberry on the resemblance of the flora to that of the "Laramie" and 

 conaments on the moUuscan faunas. of the two as follows: 



The fauna to which one instinctively turns for the purpose of zoological comparison with 

 this Puget estuary fauna is that of the Laramie group. Such a comparison is especially 

 suggested by the known floral relations of the two groups of strata, their presumable con- 

 temporaneity of origin, and the nonmarine character of the moUuscan faunas of both. Upon 

 making a comparison, however, important zoological differences appear. It is true, there are two 

 species of Corbicula in the Puget fauna that are so closely like Laramie forms as to suggest 

 specific identity upon casual examination, but the differences between the two faunas are 

 strikingly shown by the family and generic characters of the other members of the Puget fauna 

 as compared with the Laramie fauna. 



For example, a species of Teredo has been found in the Puget group, but no member of 

 the Teredinidse has yet been found in the Laramie. Two species of the Puget fauna are referred 

 to the TeUinidse, but no member of that family has yet been found in the Laramie. But the 

 generic form which gives an especially unique character to the Puget fauna is that of Batissa. 



