PAST HISTORIES OF PLANT FAMILIES 167 



do also countless impressions and trails of animals. In 

 this portion of the study of fossils it is better to have a 

 healthy scepticism than an illuminating imagination. 



Diatoms, with their hard siliceous shells, are natu- 

 rally well preserved as fossils (see fig. 121 ), for even if the 

 protoplasm decays the mineral coats remain practically 

 unchanged. 



Diatoms to-day exist in great numbers, both in the 

 cold water of the polar regions and in the heat of hot 

 springs. Often, in the latter, one can see them actually 

 being turned into fossils. In the Yellowstone Park they 

 are accumulating in vast num- 

 bers over large areas, and in 

 some places have collected to a 

 thickness of 6 feet. At the 

 bottoms of freshwater lakes they Fig. 121.— Diatom showing the 

 may form an almost pure mud Double siliceous coat 



of fine texture, while on the floor 



of deep oceans there is an ooze of diatoms which have 

 been separated from the calcareous shells by their 

 greater powers of resistance to solution by salt water. 



There are enormous numbers of species now living, 

 and of fossils from the Tertiary and Upper Mesozoic 

 rocks; but, strangely enough, though so numerous and 

 so widely distributed, both now and in these past periods, 

 they have not been found in the earlier rocks. 



In one way the diatoms differ from ordinary fossils. 

 In the latter the soft tissues of the plant have been re- 

 placed by stone, while in the former the living cell was 

 enclosed in a siliceous case which does not decompose, 

 thus resembling more the fossils- of animal shells. 



Bacteria are so very minute that it is impossible to 

 recognize them in ordinary cases. In the matrix of the 

 best-preserved fossils are always minute crystals and 

 granules that may simulate bacterial shapes perfectly. 

 Bacillus and Micrococais of various species have been 



