FOSSILS FROM THE UPPER GREENSA^D. 103 



after their embedment in the sand. The specimens of Siphonia 

 tulipa also are filled with the sandstone matrix and have the 

 appearance of having been phosphatised in situ. They resemble, 

 indeed, those which occur in the upper part of the Upper Green- 

 sand of the Isle of Wight. Lastly, there is a cast of Ammonites 

 rostratus in a sandstone matrix without any phosphatic matter, 

 though this species has never before been recorded as an underived 

 fossil from so high a position in the Upper Greensand. 



It is at present beyond my power to explain all these anomalies ; 

 why the same species should occur in different states of mineralisa- 

 tion ; why some specimens seem to be derived from an older 

 deposit while others do not ; and why this nodule bed at the very 

 summit of the Greensand in Dorset should yield an assemblage of 

 fossils which in other parts of the country are not found together. 

 I can only state the facts and leave the matter for future 

 elucidation. 



It is, however, the phosphatic fossils which create the difficulty : 

 apart from the occurrence of Ammonites rostratus in the sandstone 

 the fossils which occur as shells or as sandstone casts are such as 

 might be expected at this geological horizon. Most of them are 

 found in the bed of sand which forms the uppermost member of 

 the Upper Greensand at Warminster, and which has long been 

 celebrated for the number of its fossils and their excellent state of 

 preservation. There cannot be a doubt that the majority of the 

 fossils found in Dorset sandstone belong to the fauna of the zone 

 which Dr. Barrois has called the zone of Pecten asper. 



Thus if we exclude the fossils which only occur as phosphatic 

 casts, the total number of named species from this sandstone is 60, 

 and of these, 45 occur at Warminster, or about 75 per cent. The 

 number of species found in a phosphatic condition is 38, and of 

 these, 22 occur in the Gault, or as derived phosphates in the 

 Cambridge Greensand ; this is only about 58 per cent., a 

 proportion which confirms the statement already made that the 

 phosphates are not all derived fossils, for they are not all members 

 of the fauna of the Am. rostratus zone. 



