OF PLANORBIS AT STEINHEIM. 51 



It is quite curious that under these circumstances, such a close agreement could 

 have been obtained as is exhibited in the preceding table. In this table, however, I 

 have taken the liberty to transform the word sand into shell-sand, otherwise Sandberger's 

 table is given literally. This was done, because in my own notes and sections all hard 

 sand is put down as shell-sand, whether entirely composed of shells and their fragments, 

 or largely made up of detritus. 



The sections from 1-8 inclusive, were taken from the precipitous sides of the various 

 pits, which are located on the Geological Map. The measurements were taken with 

 a rule or tape in millemetres, from the face of the strata, and in no case estimated. 

 Nevertheless, the thickness of the limestone partings are quite often exaggerated, 

 in order to make them of an appreciable thickness in the printed sections, where, 

 if reduced to y^^, they would in ■ some cases be too thin to show the cross-bars 

 which indicate their lithological character. 



It may be said that there is general prevalence of the clay layers below in the lower 

 parts of Sections 4, 6, and probably also in 2, 3, and 7, 8, if they had been penetrated 

 sufficiently ; and that these clay bands are for the most part destitute of fossil shells. 

 Above the fish layers c the sand predominates until we reach formation m, when the 

 clay again appears in excess. It will be observed that this coincides with the three great 

 faunal groups which may be made, the period of the PI. Steinheynensis which is rare 

 above the fish formation c ; of PI. irochiformis, which is so abundant from formation d to 

 I, inclusive ; and of PI. oxystomus which is equally characteristic of formations m to o, 

 inclusive. This is apparently contradicted by the prevalence of clay in formations _/" to i, 

 inclusive, in sections 3 and 4. We can, however, account for these as unfossiliferous 

 partings between the layers of shell-sand, corresponding to the limestone partings of the 

 the other sections. 



Notwithstanding this apparent correspondence between the kind of bottom obtaining 

 in the lake at any one time and the species of shells living at that time in the 

 waters, there are many failures which prevent the application of this rule in detail. 

 Thus, the oxystomus series, which is prevalent in the clays of formation m, in Sections 3, 

 7, and 8, are equally prevalent in the sands of the corresponding formations in the 

 Cloister Pit, and in the Coarse Limestone of the Lower Period. PL trochiformis, 

 also, which appears usually only in the shell-sands, is very abundant in the upper 

 clays of Section 8, and the formation p, of Section 7. All species seem to be found 

 in about equal abundance in the limestone partings, but are rarer and often entirely 

 absent in the thin clay layers, or partings described above, and in some of the thicker 

 beds. 



Nevertheless, the general tendency of the observed facts, leads to the conclusion 

 that the entire series of animals found below formation m preferred to inhabit localities 

 in which sand was being deposited, or had sandy bottoms, and those from formation m 

 upwards were either equally well fitted for clayey or sandy waters, which seems the 

 most probable conclusion, or preferred to dwell in clayey localities. 



This conclusion derives additional probability from the fact that the return of the 

 trochiformis fauna, which occurred in formation x, in the East Pit, Old Pit, and New 

 Pit, was accompanied by a deposit of shell sand. 



