30 HYATT ON THE TERTIARY SPECIES 



and use the word force as all of us are obliged to do for a cause, which we can recognize, 

 but cannot see. This battle of a force from within, against a force from without, is 

 perhaps the strongest argument for natural selection which can be adduced. But it also 

 shows that natural selection, even if it is as general as is claimed, is the expression for a 

 series of results caused by the action and reaction of these forces, one upon the other. 



The views of Wagner upon the necessity of isolation for the production of new forms, 

 have specially interested me in connection with the Steinheim shells, since they show cer- 

 tainly the effects of isolation in an extreme case. Although the Steinheim basin is an 

 evident exception, and although it is much more difl&cult to apply the law of isolation in 

 ordinary cases, still Weissmann has shown us how this can be done, and how the pre- 

 vention of indiscriminate crossing, "amixie," may be called in to assist isolation in the 

 production of new forms. This view appears to me to receive the strongest possible con- 

 firmation when applied to the differential characteristics of the Steinheim Series, since, 

 as I have pointed out, the four species were selected from a number of varieties of PI. 

 levis, after these had migrated into this isolated basin where their surroundings were 

 favorable, and the field unoccupied. But the uniformities presented by these series appear 

 to me to indicate the existence of other causes, which must be taken into account if the 

 phenomena presented by them are to be fully and correctly explained. 



The law of acceleration in the inheritance of characteristics is very often difficult to 

 recognize among shells, but not fortunately among the Steinheim species. The rounded 

 whorl of PL steiJheimensis, bcgius to be confined to the early stages in the PI. Steinheimensis, 

 and in those like figs. 4-19, line c, pi. 1, which have angvdar whorls in the full-grown, it is 

 found at still younger stages, and finally disappears altogether in PI. discoideus, with 

 some exceptions, and is replaced in the young by a form like that of PL tenuis, fig. 4. 

 This form is again replaced in the young of the extreme forms of PL discoideus 

 and PL trochiformis, by a more spiral form resembling the adult of PL elegans, the whorl 

 has the same flat, angular upper side, and rounded lower side. The increasing spirality 

 of the adults is carried back into the easy stages in the same way and at the same time 

 as the other characteristics. In PL crescens, this acceleration is less noticeable 

 than in other forms, but even here the inheritance of the acute, angular, outer side 

 of the whorl takes place much eaiiier in P. crescens, fig. 15, than in the transition 

 form, fig. 14. In the young of PL discoideus and trochiformis, all traceable derivation 

 from PL Steinheimensis is lost, being crowded out by the angular sulcatus or elegans fonn. 

 This also occurs in PL triquetrus var. turhinatus, fig. 20, when compared with true PL 

 triquetrus, fig. 19 and P. minuius. 



The inheritance of striae in the first sub-series proceeds according to the same law, but 

 the presence of numerous hybrids makes it difficult to follow it out. 



The small sac-like shell which occupies the apex, belongs to the later embryonic 

 stages, and as such is present in nearly aU Gasteropoda, except such as retain a still earlier 

 and flatter or more open form, and all the Ammonoids and Belemnoids, and, though absent 

 in the Nautiloids, its former presence is indicated by a scar which occupies the apex of 

 the whorl. 



I have not a sufficient acquaintance with the local peculiarities of the Planorbidge in 

 other habitats to determine whether there is any general modification in all of the shells, 



