SMITHSONIAN EXPLORATIONS. I92S T5 



sufficient funds with which to carry on the work and develop the 

 most effective methods of exploration. 



The new methods adopted in my work of the past season at Mel- 

 hourne I found most effective hoth in collecting and keeping separate 

 the fossil bones of the different formations and in the critical study 

 of the geologic structure of the two principal beds involved, as well 

 as in studying the character of the contact plane between them. 



The sedimentary de])osits under investigation at the Melliourne 

 locality are relatively shallow, averaging not more than 8 or lo 

 feet in thickness. They lie everywhere on an uneven floor of con- 

 solidated Coc|uina layer made up of marine shells. They are quite 

 distinctly divided into two layers or beds. The upper one averages 

 about 3 or 4 feet in thickness ; where undisturbed by recent 

 stream beds or other excavation it is composed principally of alter- 

 nating layers of rather coarse, loose sand, swamp-muck and leaf- 

 mould, the muck deposits being heaviest at the top, the interstratified 

 leaf-mould and sand layers prevailing in the lower porti(in of the 

 upper bed. The lower l)ed is composed of a more compact and purer 

 sand mass which shows little stratification, except in its low^er portion, 

 where there is some evidence of faintly defined alternating layers 

 and lenses of coarser and finer sands intermixed in places with a 

 sandy swamp muck. This geologic structure is almost identical with 

 conditions observed at Vero. Hence in describing them I have used 

 Dr. Sellards' designations of the three geological levels. The under- 

 lying shell deposit is known as " No. i bed," the lower .sandy deposit 

 as " No. 2 bed " and the upper one as " No. 3 l)ed." 



In former explorations, excavations of very small extent were the 

 rule, with occasional development of larger areas in which a com- 

 plete section of the deposit was exposed and worked on a more or 

 less perpendicular face. This method gave a good general idea of 

 the structure of the beds but only such as could be studied in cross 

 section. In consequence some wrong interpretations of the origin of 

 the beds were made, and some important features of structure were 

 overlooked. 



For example, it was rather generally accepted that the unevenness 

 of the contact plane ol)served between the No. 2 and No. 3 beds was 

 due to stream-channel erosion, and that the deposits of both these beds 

 were principally due to stream-channel action. On this assumption 

 it was suggested that the entire deposit was of relatively recent origin, 

 the presence of a more ancient fauna being accounted for by assuming 

 that they were derived from worked over and redeposited older beds. 



