Subsurface Laboratory Methods 111 



and the number of species present in samples. In peat investigations, pollen 

 of tree species are usually the only ones used and the conclusions are usually 

 based upon a count of 200 fossils. In peat seldom more than a dozen plant 

 species are present and frequently the total number of tree pollen species is 

 not more than six in any one level. In coal and shales studies the paleontologist 

 uses all types of spores and pollen and seldom restricts his studies to tree 

 pollen. For this reason the number of species is often several times the number 

 encountered in peat. Consequently, it is desirable to examine and count a 

 greater number of individuals. What this count should be is not established 

 and differs widely among workers. . . . 



Graphic treatment of the counts is probably the best method of demon- 

 strating plant microfossil correlations. Line graphs were early used in peat 

 pollen studies to show paleoecological succession, but these have been largely 

 replaced by bar graphs of several types. . . . 



Two types of graphs can be constructed for correlation purposes. These 

 are channel-sample graphs and horizon-segment graphs. The former is the 

 better type for direct correlation purposes, if the strata involved are not 

 more than several feet thick. In thick coal seams, channel sample correla- 

 tion becomes difficult. Where possible, a coal seam should be divided into 

 segments separated by shale or pyrite partings. These partings are often of 

 extensive areal extent and make good natural boundaries. In shale de- 

 posits lithological types should be used to limit the extent of the sample. 

 In order to determine paleoecological succession within a rock member, 

 contiguous thin vertical samples must be studied. The graphed results will 

 show the percentage of the fossils at successive levels. Close correlation of 

 comparable measured horizons has not yet shown exact percentages of identical 

 species, but successional trends are usually indicated, if the section is viewed 

 as a whole, and as such have paleoecological and stratigraphic value. 



How similar the percentages of each fossil species must be in strata to 

 indicate correlation is a question of considerable pertinence. Experiments 

 with seams of coal have shown that the dominant fossils will frequently vary 

 between five and ten percent in a 200-fossil count, if the samples are collected 

 several miles apart, or the maceration process has not been uniform. In the 

 first instance, areal distribution of the ancient vegetation apparently is a 

 factor, or the coal seam was thicker or more completely represented at one 

 locality than at another. In the second instance, where uniformity of macera- 

 tion is not attained, fossils with various thicknesses of spore or pollen coat, 

 or great differences in size, will appear in variable percentages. An effective 

 method of combating such problems is to divide the sample in the course of 

 its preparation and allow additional time for each portion of the sample. 

 Uniformity can be checked with test slides by using the corrosion of certain 

 species of fossils as an index for uniformity of maceration. 



In conclusion, it might be stated that results have been attained with 

 fossil spores and pollen which show conclusively that all coals and shales thus 

 far studied can be assigned within the limits of geological periods, and that 

 various strata can be separated from each other by their spore and pollen 

 facies, or specific abundance of various species. It would seem that plant micro- 

 fossils have great future scientific and economic value as the science develops 

 and broadens. 



Grass Seeds 



Several years ago Elias "^ published a noteworthy paper pertaining 

 to prairie-grass seeds of the Late Tertiary deposits of the Prairie States. 



Elias, M. K., Tertiary Prairie Grasses and Other Herbs from the High Plains : Geol. Soc. America 

 Special Paper 41, pp. 1-176, 1942. 



