The Principles of Palaeontology. 1G3 



from each other. A shell normally spiral or turbinate may in 

 developing become simply conical, making the transition by a 

 cowl-shaped form ; or it may uncoil so as to become a straightened 

 tube, or again the later volutions may entirely cover those 

 preceding as is the case in the Cyprasidae. Lastly the shell 

 passing beneath the mantle may regress and more or less com- 

 pletely disappear. These same phenomena are found in all the 

 types of Gasteropoda ; Prosobranchia, Opisthobranchia, Pulmo- 

 •nata, and Heteropoda. 



These processes of evolution may be compared to those which, 

 in the Cephalopoda, manifest themselves at notably different 

 epochs, in the two very distinct groups of the Tetrabranchiates 

 and the Dibranchiates. Forms more or less completely uncoiled 

 and precisely parallel have appeared among the Tetrabranchiates 

 of the Silurian, and the Ammonitidae toward the Cretaceous 

 epoch. It is thus that the Baculites of the Maestrichtian 

 reproduce the Lituites of the Silurian. It would seem that the 

 same law of deformation of the normal type presided over the 

 evolution of these forms and announced their approaching 

 decadence. ^ , 



Of the irregular Echini some are provided with jaws, others 

 are destitute of them. No transition term exists between the 

 two types as regards these important organs. But in regard to 

 the exterior form, gradual modifications appear in the two 

 groups to such a degree that for a long time the groups of the 

 Gnathostomes and Atelostomes were confounded. 



The Corals, both the perforate and the imperforate, between 

 which no transition exists, display also a certain number of 

 simple or colonial forms which are reproduced in the two groups 

 with a parallelism sometimes so complete as to make us doubt 

 whether the division should be made thus between the perforate 

 and imperforate, or whether we should consider as allied to 

 each group of imperforate corals a corresponding group of 

 perforate forms derived from them, perhaps, by regression. 



The same remark is applicable to the Foraminifera, perforate 

 and imperforate, which often present exactly the same exterior 

 forms. 



It may be seen from the foregoing remarks, that when 

 we seek to establish the real afl&nities of the various groups^ 

 that is to say, their genealogical tree, great attention must be 

 paid to these phenomena of convergence and parallelism, and it 

 must be kept in mind that the same causes have sufficed to 

 produce the same modifications among beings which in other 

 respects had no immediate kinship one with the other. 



Aberrant and synthetic types. — A second important exception 

 to the principle of correlation is drawn from the fact that the 



