220 LAMARCKIAN HYPOTHESIS OF THE 



of land from tlie original centre in Germany, or borne by 

 currents from the mouths of the Germanic rivers, — the theory 

 of Mon. C. Martins, — then must we also hold that that de- 

 velopment took place since the times of the boulder clay, and 

 that fucoids and confervas became dicotyledonous and mono- 

 cotyledonous plants during a brief period, in which the Pur- 

 pura lapillus and Turritella terehra did not alter a single 

 whorl, and the Cyprina islandica and Astarte borealis re- 

 tained unchanged each minute projection of their hinges, and 

 each nicer peculiarity of their muscular impressions. Crea- 

 tion would be greatly less wonderful than a sudden transmu- 

 tative process such as this, restricted in its operation to groupes 

 of English, Irish, and Manx plants, identical with groupes in 

 Germany, when all the various organisms around them, such 

 as our sea-shells, continued to be exactly what they had been 

 for ages before. A process of development from the lowest 

 to the highest forms, rigidly restricted to the flora of a coun- 

 try, would be simply the miracle of Jonah's gourd several 

 thousand times repeated. 



I must here indulge in a few remarks more, which, though 

 they may seem of an incidental character, have a direct bear- 

 ing on the general subject The geologist infers, in all his 

 reasonings founded on fossils, that a race or species has ex- 

 isted from some one certain point in the scale to some other 

 certain point, if he find it occurring at both points together. 

 He infers on this principle, for instance, that the boulder 

 clay, which contains only recent shells, belongs to the recent 

 or post-Tertiary period ; and that the Oolite and Lias, which 

 contain no recent shells, represent a period whose existences 

 have all become extinct. And all experience serves to show 

 that his principle is a sound one. In creation there are many 

 species linked together, from their degree of similarity, by 

 the generic tie ; but no perfect verisimilitude obtains among 

 them, unless hereditarily derived from the one, two, or more 



